[Federal Register Volume 68, Number 100 (Friday, May 23, 2003)]
[Proposed Rules]
[Pages 28328-28603]
From the Federal Register Online via the Government Printing Office [www.gpo.gov]
[FR Doc No: 03-9737]



[[Page 28327]]

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

Part II





Environmental Protection Agency





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



40 CFR Parts 69, 80, 89, et al.



Control of Emissions of Air Pollution From Nonroad Diesel Engines and 
Fuel; Proposed Rule

Federal Register / Vol. 68, No. 100 / Friday, May 23, 2003 / Proposed 
Rules

[[Page 28328]]


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

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 69, 80, 89, 1039, 1065, and 1068

[AMS-FRL-7485-8]
RIN 2060-AK27


Control of Emissions of Air Pollution From Nonroad Diesel Engines 
and Fuel

AGENCY: Environmental Protection Agency (EPA).

ACTION: Notice of proposed rulemaking.

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

SUMMARY: Nonroad diesel engines contribute considerably to our nation's 
air pollution. These engines, used primarily in construction, 
agricultural, and industrial applications, are projected to continue to 
contribute large amounts of particulate matter (PM), nitrogen oxides 
(NOX), and sulfur oxides (SOX), all of which 
contribute to serious public health problems in the United States. 
These problems include premature mortality, aggravation of respiratory 
and cardiovascular disease, aggravation of existing asthma, acute 
respiratory symptoms, chronic bronchitis, and decreased lung function. 
We believe that diesel exhaust is likely to be carcinogenic to humans 
by inhalation.
    Today EPA is proposing new emission standards for nonroad diesel 
engines and sulfur reductions in nonroad diesel fuel that will 
dramatically reduce emissions attributed to nonroad diesel engines. 
This comprehensive national program will regulate nonroad diesel 
engines and diesel fuel as a system. New engine standards will begin to 
take effect in the 2008 model year. These standards are based on the 
use of advanced exhaust emission control devices. We estimate PM 
reductions of 95%, NOX reductions of 90%, and the virtual 
elimination of sulfur oxides (SOX) from nonroad engines 
meeting the new standards. Nonroad diesel fuel sulfur reductions of up 
to 99% from existing levels will provide significant health benefits as 
well as facilitate the introduction of high-efficiency catalytic 
exhaust emission control devices as these devices are damaged by 
sulfur. These fuel controls would begin in mid-2007. Today's nonroad 
proposal is largely based on EPA's 2007 highway diesel program.
    To better ensure the benefits of the standards are realized in-use 
and throughout the useful life of these engines, we are also proposing 
new test procedures, including not-to-exceed requirements, and related 
certification requirements. The proposal also includes provisions to 
facilitate the transition to the new engine and fuel standards and to 
encourage the early introduction of clean technologies and clean 
nonroad diesel fuel. We have also developed provisions for both the 
proposed engine and fuel programs designed to address small business 
considerations.
    The requirements in this proposal would result in substantial 
benefits to public health and welfare and the environment through 
significant reductions in emissions of NOX and PM, as well 
as nonmethane hydrocarbons (NMHC), carbon monoxide (CO), sulfur oxides 
(SOX) and air toxics. We project that by 2030, this program 
would reduce annual emissions of NOX, and PM by 827,000 and 
127,000 tons, respectively. These emission reductions would prevent 
9,600 premature deaths, over 8,300 hospitalizations, and almost a 
million work days lost, and other quantifiable benefits every year. All 
told the benefits of this rule would be approximately $81 billion 
annually by 2030. Costs for both the engine and fuel requirements would 
be many times less, at approximately $1.5 billion annually.

DATES: Comments: Send written comments on this proposal by August 20, 
2003. See section IX for more information about written comments.
    Hearings: We will hold public hearings on the following dates: June 
10, 2003; June 12, 2003; and June 17, 2003. Each hearing will start at 
9 a.m. local time. If you want to testify at a hearing, notify the 
contact person listed below at least 10 days before the hearing. See 
section IX for more information about public hearings.

ADDRESSES: Comments: Comments may be submitted by mail to: Air Docket, 
Environmental Protection Agency, Mailcode: 6102T, 1200 Pennsylvania 
Ave., NW., Washington, DC 20460, Attention Docket ID No. A-2001-28.
    Comments may also be submitted electronically, by facsimile, or 
through hand delivery/courier. Follow the detailed instructions as 
provided in section IX of the SUPPLEMENTARY INFORMATION section.
    Hearings: We will hold public hearings at the following three 
locations:

------------------------------------------------------------------------
 
------------------------------------------------------------------------
New York, New York, Park Central New York,     June 10, 2003
 870 Seventh Avenue at 56th Street, New York,
 NY 10019, Telephone: (212) 247-8000, Fax:
 (212) 541-8506.
Chicago, Illinois, Hyatt Regency O'Hare, 9300  June 12, 2003.
 W. Bryn Mawr Avenue, Rosemont, IL 60018,
 Telephone: (847) 696-1234, Fax: (847) 698-
 0139.
Los Angeles. California, Hyatt Regency Los     June 17, 2003.
 Angeles, 711 South Hope Street, Los Angeles,
 California, USA. 90017, Telephone: (213) 683-
 1234, Fax: (213) 629-3230.
------------------------------------------------------------------------

    See section IX, ``Public Participation'' below for more information 
on the comment procedure and public hearings.

FOR FURTHER INFORMATION CONTACT: U.S. EPA, Office of Transportation and 
Air Quality, Assessment and Standards Division hotline, (734) 214-4636, 
asdinfo@epa.gov. Carol Connell, (734) 214-4349; connell.carol@epa.gov.

SUPPLEMENTARY INFORMATION:

Regulated Entities

    This action would affect you if you produce or import new heavy-
duty diesel engines which are intended for use in nonroad vehicles such 
as agricultural and construction equipment, or produce or import such 
nonroad vehicles, or convert heavy-duty vehicles or heavy-duty engines 
used in nonroad vehicles to use alternative fuels. It would also affect 
you if you produce, import, distribute, or sell nonroad diesel fuel, or 
sell nonroad diesel fuel.
    The following table gives some examples of entities that may have 
to follow the regulations. But because these are only examples, you 
should carefully examine the regulations in 40 CFR parts 80, 89, 1039, 
1065, and 1068. If you have questions, call the person listed in the 
FOR FURTHER INFORMATION CONTACT section of this preamble:

--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                  NAICS      SIC
                   Category                       codes     codes                        Examples of potentially regulated entities
                                                   \a\       \b\
--------------------------------------------------------------------------------------------------------------------------------------------------------
Industry......................................    333618      3519  Manufacturers of new nonroad diesel engines.

[[Page 28329]]

 
Industry......................................    333111      3523  Manufacturers of farm machinery and equipment.
Industry......................................    333112      3524  Manufacturers of lawn and garden tractors (home).
Industry......................................    333924      3537  Manufacturers of industrial trucks.
Industry......................................    333120      3531  Manufacturers of construction machinery.
Industry......................................    333131      3532  Manufacturers of mining machinery and equipment.
Industry......................................    333132      3533  Manufacturers of oil and gas field machinery and equipment.
Industry......................................    811112      7533  Commercial importers of vehicles and vehicle components.
                                                  811198      7549
Industry......................................    324110      2911  Petroleum refiners.
Industry......................................    422710      5171  Diesel fuel marketers and distributors.
                                                  422720      5172
Industry......................................    484220      4212  Diesel fuel carriers.
                                                  484230      4213
--------------------------------------------------------------------------------------------------------------------------------------------------------
\a\ North American Industry Classification System (NAICS).
\b\ Standard Industrial Classification (SIC) system code.

How Can I Get Copies of This Document and Other Related Information?

    Docket. EPA has established an official public docket for this 
action under Docket ID No. A-2001-28. The official public docket 
consists of the documents specifically referenced in this action, any 
public comments received, and other information related to this action. 
Although a part of the official docket, the public docket does not 
include Confidential Business Information (CBI) or other information 
whose disclosure is restricted by statute. The official public docket 
is the collection of materials that is available for public viewing at 
the Air Docket in the EPA Docket Center, (EPA/DC) EPA West, Room B102, 
1301 Constitution Ave., NW., Washington, DC. The EPA Docket Center 
Public Reading Room is open from 8:30 a.m. to 4:30 p.m., Monday through 
Friday, excluding legal holidays. The telephone number for the Reading 
Room is (202) 566-1742, and the telephone number for the Air Docket is 
(202) 566-1742.
    Electronic Access. You may access this Federal Register document 
electronically through the EPA Internet under the ``Federal Register'' 
listings at http://www.epa.gov/fedrgstr/.
    An electronic version of the public docket is available through 
EPA's electronic public docket and comment system, EPA Dockets. You may 
use EPA Dockets at http://www.epa.gov/edocket/ to submit or view public 
comments, access the index listing of the contents of the official 
public docket, and to access those documents in the public docket that 
are available electronically. Once in the system, select ``search,'' 
then key in the appropriate docket identification number.
    Certain types of information will not be placed in the EPA Dockets. 
Information claimed as CBI and other information whose disclosure is 
restricted by statute, which is not included in the official public 
docket, will not be available for public viewing in EPA's electronic 
public docket. EPA's policy is that copyrighted material will not be 
placed in EPA's electronic public docket but will be available only in 
printed, paper form in the official public docket. To the extent 
feasible, publicly available docket materials will be made available in 
EPA's electronic public docket. When a document is selected from the 
index list in EPA Dockets, the system will identify whether the 
document is available for viewing in EPA's electronic public docket. 
Although not all docket materials may be available electronically, you 
may still access any of the publicly available docket materials through 
the docket facility identified in section IX.
    For public commenters, it is important to note that EPA's policy is 
that public comments, whether submitted electronically or in paper, 
will be made available for public viewing in EPA's electronic public 
docket as EPA receives them and without change, unless the comment 
contains copyrighted material, CBI, or other information whose 
disclosure is restricted by statute. When EPA identifies a comment 
containing copyrighted material, EPA will provide a reference to that 
material in the version of the comment that is placed in EPA's 
electronic public docket. The entire printed comment, including the 
copyrighted material, will be available in the public docket.
    Public comments submitted on computer disks that are mailed or 
delivered to the docket will be transferred to EPA's electronic public 
docket. Public comments that are mailed or delivered to the Docket will 
be scanned and placed in EPA's electronic public docket. Where 
practical, physical objects will be photographed, and the photograph 
will be placed in EPA's electronic public docket along with a brief 
description written by the docket staff.
    For additional information about EPA's electronic public docket 
visit EPA Dockets online or see 67 FR 38102, May 31, 2002.

Outline of This Preamble

I. Overview
    A. What Is EPA Proposing?
    1. Nonroad Diesel Engine Emission Standards
    2. Nonroad, Locomotive, and Marine Diesel Fuel Quality Standards
    B. Why Is EPA Making This Proposal?
    1. Nonroad, Locomotive, and Marine Diesels Contribute to Serious 
Air Pollution Problems
    2. Technology and Fuel Based Solutions
    3. Basis For Action Under the Clean Air Act
II. What Is the Air Quality Impact of the Sources Covered by the 
Proposed Rule?
    A. Overview
    B. Public Health Impacts
    1. Particulate Matter
    a. Health Effects of PM2.5 and PM10
    b. Current and Projected Levels
    i. PM10 Levels
    ii. PM2.5 Levels
    2. Air Toxics
    a. Diesel exhaust
    i. Potential Cancer Effects of Diesel Exhaust
    ii. Other Health Effects of Diesel Exhaust
    iii. Ambient Levels and Exposure to Diesel Exhaust PM
    iv. Diesel Exhaust PM Exposures
    b. Gaseous Air Toxics
    3. Ozone
    a. What are the health effects of ozone pollution?
    b. Current and projected 8-hour ozone levels
    C. Other Environmental Effects
    1. Visibility
    a. Visibility is Impaired by Fine PM and Precursor Emissions 
From Nonroad Engines Subject to this Proposed Rule
    b. Visibility Impairment Where People Live, Work and Recreate

[[Page 28330]]

    c. Visibility Impairment in Mandatory Federal Class I Areas
    2. Acid Deposition
    3. Eutrophication and Nitrification
    4. Polycyclic Organic Matter Deposition
    5. Plant Damage from Ozone
    D. Other Criteria Pollutants Affected by This NPRM
    E. Emissions From Nonroad Diesel Engines
    1. PM2.5
    2. NOX
    3. SO2
    4. VOC and Air Toxics
III. Nonroad Engine Standards
    A. Why are We Setting New Engine Standards?
    1.The Clean Air Act and Air Quality
    2. The Technology Opportunity for Nonroad Diesel Engines
    B. What Engine Standards are We Proposing?
    1. Exhaust Emissions Standards
    a. Standards Timing
    b. Phase-In of NOX and NMHC Standards
    c. Rationale for Restructured Horsepower Categories
    d. PM Standards for Smaller Engines
    i. <25 hp
    ii. 25-75 hp
    e. Engines Above 750 hp
    f. CO Standards
    g. Exclusion of Marine Engines
    2. Crankcase Emissions Control
    C. What Test Procedure Changes Are Being Proposed?
    1. Supplemental Transient Test
    2. Cold Start Testing
    D. What Is Being Done To Help Ensure Robust Control In Use?
    1. Not-to-Exceed Requirements
    a. NTE Standards We are Proposing
    b. Comment Request on an Alternative NTE Approach
    2. Plans for Future In-Use Testing and Onboard Diagnostics
    a. Manufacturer-Run In-Use Test Program
    b. Onboard Diagnostics
    E. Are the Proposed New Standards Feasible?
    1.Technologies To Control NOX and PM Emissions From 
Mobile Source Diesel Engines
    a. PM Control Technologies
    b. NOX Control Technologies
    2. Can These Technologies Be Applied to Nonroad Engines and 
Equipment?
    a. Nonroad Operating Conditions and Exhaust Temperatures
    b. Nonroad Operating Conditions and Durability
    3. Are the Standards Proposed for Engines of 75 hp or Higher 
Feasible?
    4.Are the Standards Proposed for Engines =25 hp and 
<75 hp Feasible?
    a. What makes the 25-75 hp category unique?
    b. What engine technology is used today, and will be used for 
the applicable Tier 2 and Tier 3 standards?
    c. Are the proposed standards for 25-75 hp engines 
technologically feasible?
    i. 2008 PM Standards
    ii. 2013 Standards
    d. Why EPA has not proposed more stringent Tier 4 NOX 
standards
    5. Are the Standards Proposed for Engines <25 hp Feasible?
    a. What makes the < 25 hp category unique?
    b. What engine technology is currently used in the <25 hp 
category?
    c. What data indicates that the proposed standards are feasible?
    d. Why has EPA not proposed more stringent PM or NOX 
standards for engines <25 hp?
    6. Meeting the Crankcase Emissions Requirements
    F. Why Do We Need 15ppm Sulfur Diesel Fuel?
    1. Catalyzed Diesel Particulate Filters and the Need for Low 
Sulfur Fuel
    a. Inhibition of Trap Regeneration Due to Sulfur
    b. Loss of PM Control Effectiveness
    c. Increased Maintenance Cost for Diesel Particulate Filters Due 
to Sulfur
    2. Diesel NOX Catalysts and the Need for Low Sulfur 
Fuel
    a. Sulfur Poisoning (Sulfate Storage) on NOX 
Adsorbers
    b. Sulfate Particulate Production and Sulfur Impacts on 
Effectiveness of NOX Control Technologies
    G. Reassessment of Control Technology for Engines Less Than 75 
hp in 2007
IV. Our Proposed Program for Controlling Nonroad, Locomotive and 
Marine Diesel Fuel Sulfur
    A. Proposed Nonroad, Locomotive and Marine Diesel Fuel Quality 
Standards
    1. What Fuel Is Covered by this Proposal?
    2. Standards and Deadlines for Refiners, Importers, and Fuel 
Distributors
    a. The First Step to 500 ppm
    b. The Second Step to 15 ppm
    c. Other Standard Provisions
    d. Cetane Index or Aromatics Standard
    B. Program Design and Structure
    1. Background
    2. Proposed Fuel Program Design and Structure
    a. Program Beginning June 1, 2007
    i. Use of a Marker To Differentiate Heating Oil from NRLM
    ii. Non-highway Distillate Baseline Cap
    iii. Setting the Non-highway Distillate Baseline
    iv. Diesel Sulfur Credit Banking, and Trading Provisions for 
2007
    b. 2010
    i. A Marker To Differentiate Locomotive and Marine Diesel from 
Nonroad Diesel
    ii. Diesel Sulfur Credit Banking and Trading Provisions for 2010
    c. 2014
    3. Other Options Considered
    a. Highway Baseline and a NRLM baseline for 2007
    i. Highway Baseline
    ii. Nonroad, Locomotive, and Marine Baseline
    iii. Combined Impact of Highway and NRLM Baselines
    b. Locomotive and Marine Baseline for 2010
    c. Designate and Track Volumes in 2007
    i. Replacement for the Non-highway Baseline Approach
    ii. Designate and Track as a Refiners Option in Addition to the 
Baseline Approach
    C. Hardship Provisions for Qualifying Refiners
    1. Hardship Provisions for Qualifying Small Refiners
    a. Qualifying Small Refiners
    i. Regulatory Flexibility for Small Refiners
    ii. Rationale for Small Refiner Provisions
    iii. Limited Impact of Small Refiner Options on Program 
Emissions Benefits
    b. How Do We Define Small Refiners for Purposes of the Hardship 
Provisions?
    c. What Options Are Available for Small Refiners?
    i. Delays in Nonroad Fuel Sulfur Standards for Small Refiners
    ii. Options to Encourage Earlier Compliance by Small Refiners
    d. How Do Refiners Apply for Small Refiner Status?
    2. General Hardship Provisions
    a. Temporary Waivers From Non-highway Diesel Sulfur Requirements 
in Extreme Unforeseen Circumstances
    b. Temporary Waivers Based on Extreme Hardship Circumstances
    D. Should Any Individual States or Territories Be Excluded From 
This Rule?
    1. Alaska
    a. How Was Alaska Treated Under the Highway Diesel Standards?
    b. What Nonroad Standards Do We Propose for Urban Areas of 
Alaska?
    c. What Do We Propose for Rural Areas of Alaska?
    2. American Samoa, Guam, and the Commonwealth of Northern 
Mariana Islands
    a. What Provisions Apply in American Samoa, Guam, and the 
Commonwealth of Northern Mariana Islands?
    b. Why Are We Treating These Territories Uniquely?
    E. How Are State Diesel Fuel Programs Affected by the Sulfur 
Diesel Program?
    F. Technological Feasibility of the 500 and 15 ppm Sulfur Diesel 
Fuel Program
    1. What Is the Nonroad, Locomotive and Marine Diesel Fuel Market 
Today?
    2. How Do Nonroad, Locomotive and Marine Diesel Fuel Differ From 
Highway Diesel Fuel?
    3. What Technology Would Refiners Use To Meet the Proposed 500 
ppm Sulfur Cap?
    4. Has Technology To Meet a 500 ppm Cap Been Commercially 
Demonstrated?
    5. Availability of Leadtime To Meet the 2007 500 ppm Sulfur Cap
    6. What Technology Would Refiners Use To Meet the Proposed 15 
ppm Sulfur Cap for Nonroad Diesel Fuel?
    7. Has Technology To Meet a 15 ppm Cap Been Commercially 
Demonstrated?
    8. Availability of Leadtime To Meet the 2010 15 ppm Sulfur Cap
    9. Feasibility of Distributing Nonroad, Locomotive and Marine 
Diesel Fuels That Meet the Proposed Sulfur Standards
    a. Limiting Sulfur Contamination
    b. Potential Need for Additional Product Segregation
    G. What Are the Potential Impacts of the 15 ppm Sulfur Diesel 
Program on Lubricity and Other Fuel Properties?
    1. What Is Lubricity and Why Might It Be a Concern?
    2. A Voluntary Approach on Lubricity

[[Page 28331]]

    3. What Other Impact Would Today's Actions Have on the 
Performance of Diesel and Other Fuels?
    H. Refinery Air Permitting
V. Economic Impacts
    A. Refining and Distribution Costs
    1. Refining Costs
    2. Cost of Lubricity Additives
    3. Distribution Costs
    4. How EPA's Projected Costs Compare to Other Available 
Estimates
    5. Supply of Nonroad, Locomotive and Marine Diesel Fuel
    6. Fuel Prices
    B. Cost Savings to the Existing Fleet From the Use of Low Sulfur 
Fuel
    C. Engine and Equipment Cost Impacts
    1. Engine Cost Impacts
    a. Engine Fixed Costs
    i. Engine and Emission Control Device R&D
    ii. Engine-Related Tooling Costs
    iii. Engine Certification Costs
    b. Engine Variable Costs
    i. NOX Adsorber System Costs
    ii. Catalyzed Diesel Particulate Filter (CDPF) Costs
    iii. CDPF Regeneration System Costs
    iv. Closed-Crankcase Ventilation System (CCV) Costs
    v. Variable Costs for Engines Below 75 Horsepower and Above 750 
Horsepower
    c. Engine Operating Costs
    2. Equipment Cost Impacts
    a. Equipment Fixed Costs
    b. Equipment Variable Costs
    3. Overall Engine and Equipment Cost Impacts
    D. Annual Costs and Cost Per Ton
    1. Annual Costs for the 500 ppm Fuel Program
    2. Cost Per Ton for the 500 ppm Fuel Program
    3. Annual Costs for the Proposed Two-Step Fuel Program and 
Engine Program
    4. Cost per Ton of Emissions Reduced for the Total Program
    5. Comparison With Other Means of Reducing Emissions
    E. Do the Benefits Outweigh the Costs of the Standards?
    1. What were the results of the benefit-cost analysis?
    2. What was our overall approach to the benefit-cost analysis?
    3. What are the significant limitations of the benefit-cost 
analysis?
    F. Economic Impact Analysis
    1. What is an Economic Impact Analysis?
    2. What is EPA's Economic Analysis Approach for This Proposal?
    3. What Are the Results of This Analysis?
    a. Expected Market Impacts
    b. Expected Welfare Impacts
VI. Alternative Program Options
    A. Summary of Alternatives
    B. Introduction of 15 ppm Nonroad Diesel Sulfur Fuel in One Step
    1. Description of the One-Step Alternative
    2. Engine Emission Impacts
    3. Fuel Impacts
    4. Emission and Benefit Impacts
    C. Applying 15 ppm Requirement to Locomotive and Marine Diesel 
Fuel
    D. Other Alternatives
VII. Requirements for Engine and Equipment Manufacturers
    A. Averaging, Banking, and Trading
    1. Are we proposing to keep the ABT program for nonroad diesel 
engines?
    2. What are the provisions of the proposed ABT program?
    3. Should we expand the nonroad ABT program to include credits 
from retrofit of nonroad engines?
    a. What would be the environmental impact of allowing ABT 
nonroad retrofit credits?
    b. How would EPA ensure compliance with retrofit emissions 
standards?
    c. What is the legal authority for a nonroad ABT retrofit 
program?
    B. Transition Provisions for Equipment Manufacturers
    1. Why are we proposing transition provisions for equipment 
manufacturers?
    2. What transition provisions are we proposing for equipment 
manufacturers?
    a. Percent-of-Production Allowance
    b. Small-Volume Allowance
    c. Hardship Relief Provision
    d. Existing Inventory Allowance
    3. What are the recordkeeping, notification, reporting, and 
labeling requirements associated with the equipment manufacturer 
transition provisions?
    a. Recordkeeping Requirements for Engine and Equipment 
Manufacturers
    b. Notification Requirements for Equipment Manufacturers
    c. Reporting Requirements for Engine and Equipment Manufacturers
    d. Labeling Requirements for Engine and Equipment Manufacturers
    4. What are the proposed requirements associated with use of 
transition provisions for equipment produced by foreign 
manufacturers?
    C. Engine and Equipment Small Business Provisions (SBREFA)
    1. Nonroad Diesel Small Engine Manufacturers
    a. Lead Time Transition Provisions for Small Engine 
Manufacturers
    i. What the Panel Recommended
    ii. What EPA Is Proposing
    b. Hardship Provisions for Small Engine Manufacturers
    i. What the Panel Recommended
    ii. What EPA Is Proposing
    c. Other Small Engine Manufacturer Issues
    i. What the Panel Recommended
    ii. What EPA Is Proposing
    2. Nonroad Diesel Small Equipment Manufacturers
    a. Transition Provisions for Small Equipment Manufacturers
    i. What the Panel Recommended
    ii. What EPA Is Proposing
    b. Hardship Provisions for Small Equipment Manufacturers
    i. What the Panel Recommended
    ii. What EPA is Proposing
    D. Phase-In Provisions
    E. What Might Be Done To Encourage Innovative Technologies?
    1. Incentive Program for Early or Very Low Emission Engines
    2. Continuance of the Existing Blue Sky Program
    F. Provisions for Other Test and Measurement Changes
    1. Supplemental Transient Test
    2. Cold Start Testing
    3. Control of Smoke
    4. Steady-State Testing
    5. Maximum Test Speed
    6. Improvements to the Test Procedures
    G. Not-To-Exceed Requirements
    H. Certification Fuel
    I. Labeling and Notification Requirements
    J. Temporary In-Use Compliance Margins
    K. Defect Reporting
    L. Rated Power
    M. Hydrocarbon Measurement and Definition
    N. Auxiliary Emission Control Devices and Defeat Devices
    O. Other Issues
VIII. Nonroad Diesel Fuel Program: Compliance and Enforcement 
Provisions
    A. Fuel Covered and Not Covered by This Proposal
    1. Covered Fuel
    2. Special Fuel Provisions and Exemptions
    a. Fuel Used in Military Applications
    b. Fuel Used in Research and Development
    c. Fuel Used in Racing Equipment
    d. Fuel for Export
    B. Additional Requirements for Refiners and Importers
    1. Transfer of Credits
    2. Additional Provisions for Importers and Foreign Refiners 
Subject to the Credit Provisions or Hardship Provisions
    3. Proposed Provisions for Transmix Facilities
    4. Highway or Nonroad Diesel Fuel Treated as Blendstock (DTAB)
    C. Requirements for Parties Downstream of the Refinery or Import 
Facility
    1. Product Segregation and Contamination
    a. The Period From June 1, 2007 Through May 31, 2010
    b. The Period From June 1, 2010 Through May 31, 2014
    c. After May 31, 2014
    2. Diesel Fuel Pump Labeling To Discourage Misfueling
    a. Pump Labeling Requirements 2006
    b. Pump Labeling Requirements 2007-2010
    c. Pump Labeling Requirements 2010-2014
    d. Pump Labeling Requirements Beginning June 1, 2014
    e. Nozzle Size Requirements or Other Requirements To Prevent 
Misfueling
    3. Use of Used Motor Oil in New Nonroad Diesel Equipment
    4. Use of Kerosene in Diesel Fuel
    5. Use of Diesel Fuel Additives
    6. End User Requirements
    7. Anti-Downgrading Provisions
    D. Diesel Fuel Sulfur Sampling and Testing Requirements
    1. Testing Requirements
    a. Test Method Approval, Recordkeeping, and Quality Control 
Requirements
    i. How Can a Given Method Be Approved?
    ii. What Information Would Have To Be Reported to the Agency?
    iii. What Quality Control Provisions Would Be Required?
    b. Requirements To Conduct Fuel Sulfur Testing.
    2. Two Part-Per-Million Downstream Sulfur Measurement Adjustment
    3. Sampling Requirements
    4. Alternative Sampling and Testing Requirements for Importers 
of Diesel

[[Page 28332]]

Fuel Who Transport Diesel Fuel by Tanker Truck
    E. Fuel Marker Test Method
    1. How Can a Given Marker Test Method Be Approved?
    2. What Information Would Have To Be Reported to the Agency?
    F. Requirements for Recordkeeping, Reporting, and Product 
Transfer Documents
    1. Registration of Refiners and Importers
    2. Application for Small Refiner Status
    3. Applying for Refiner Hardship Relief
    4. Applying for a Non-Highway Distillate Baseline Percentage
    5. Pre-Compliance Reports
    6. Annual Compliance Reports and Batch Reports for Refiners and 
Importers
    7. Product Transfer Documents (PTDs)
    a. The Period From June 1, 2007 Through May 31, 2010
    b. The Period from June 1, 2010 Through May 31, 2014
    c. The Period After May 31, 2014
    d. Kerosene and Other Distillates To Reduce Viscosity
    e. Exported Fuel
    f. Additives
    8. Recordkeeping Requirements
    9. Record Retention
    G. Liability and Penalty Provisions for Noncompliance
    1. General
    2. What Are the Proposed Liability Provisions for Additive 
Manufacturers and Distributors, and Parties That Blend Additives 
Into Diesel Fuel?
    a. General
    b. Liability When the Additive Is Designated as Complying With 
the 15 ppm Sulfur Standard
    c. Liability When the Additive Is Designated as Having a 
Possible Sulfur Content Greater Than 15 ppm
    H. How Would Compliance With the Sulfur Standards Be Determined?
IX. Public Participation
    A. How and to Whom Do I Submit Comments?
    1. Electronically
    i. EPA Dockets
    ii. E-mail
    iii. Disk or CD ROM
    2. By Mail
    3. By Hand Delivery or Courier
    B. How Should I Submit CBI to the Agency?
    C. Will There Be a Public Hearing?
    D. Comment Period
    E. What Should I Consider as I Prepare My Comments for EPA?
X. Statutory and Executive Order Reviews
    A. Executive Order 12866: Regulatory Planning and Review
    B. Paperwork Reduction Act
    C. Regulatory Flexibility Act (RFA), as Amended by the Small 
Business Regulatory Enforcement Fairness Act of 1996 (SBREFA), 5 
U.S.C. 601 et. seq
    1. Overview
    2. Background
    3. Summary of Regulated Small Entities
    a. Nonroad Diesel Engine Manufacturers
    b. Nonroad Diesel Equipment Manufacturers
    c. Nonroad Diesel Fuel Refiners
    d. Nonroad Diesel Fuel Distributors and Marketers
    4. Potential Reporting, Record Keeping, and Compliance
    5. Relevant Federal Rules
    6. Summary of SBREFA Panel Process and Panel Outreach
    a. Significant Panel Findings
    b. Panel Process
    c. Transition Flexibilities
    i. Nonroad Diesel Engines
    ii. Nonroad Diesel Equipment
    iii. Nonroad Diesel Fuel Refiners
    iv. Nonroad Diesel Fuel Distributors and Marketers
    D. Unfunded Mandates Reform Act
    E. Executive Order 13132: Federalism
    F. Executive Order 13175: Consultation and Coordination With 
Indian Tribal Governments
    G. Executive Order 13045: Protection of Children From 
Environmental Health and Safety Risks
    H. Executive Order 13211: Actions That Significantly Affect 
Energy Supply, Distribution, or Use
    I. National Technology Transfer Advancement Act
    J. Plain Language
XI. Statutory Provisions and Legal Authority

I. Overview

    Nonroad diesel engines are the largest remaining contributor to the 
overall mobile source emissions inventory. We have already taken steps 
to dramatically reduce emissions from light-duty vehicles and heavy-
duty vehicles and engines through the Tier 2 and 2007 highway diesel 
programs.\1\ With expected growth in the nonroad sector, the relative 
emissions contribution from nonroad diesel engines is projected to be 
even larger in future years. This proposed rule sets out emissions 
standards for nonroad diesel engines used mainly in construction, 
agricultural, industrial, and mining operations that will achieve 
reductions in PM and NOX emissions levels from today's 
engines in excess of 95% and 90%, respectively. Nonroad diesel fuel is 
currently unregulated. This proposal represents the first time nonroad 
diesel fuel will be regulated. We are proposing to reduce sulfur levels 
in nonroad diesel fuel by more than 99 percent to 15 parts per million 
(ppm). Taken together, controls included in this proposal would result 
in large public health and welfare benefits.
---------------------------------------------------------------------------

    \1\ See 65 FR 6698 (February 10, 2000) and 66 FR 5001 (January 
18, 2001) for the final rules regarding the Tier 2 and 2007 highway 
diesel programs, respectively.
---------------------------------------------------------------------------

    The proposed standards for nonroad diesel engines and sulfur 
reductions for nonroad diesel fuel represent a dramatic step in 
emissions control, based on the use of advanced emissions control 
technology. Until the mid-90's, these engines had no emissions 
requirements. As a comparison, cars and trucks have been subject to a 
series of increasingly stringent emissions control programs since the 
1970s. In terms of fuel quality requirements, nonroad diesel fuel is 
currently uncontrolled at the Federal level. EPA has already issued 
rules ending these disparities for diesel engines used in highway 
applications. Starting in 2007, these engines will meet standards of 
the same level of stringency as comparable gasoline vehicles, based on 
the use of advanced aftertreatment technologies and ultra low sulfur 
diesel fuel (containing no more than 15 ppm sulfur). This proposal is 
largely based on the performance of the same advanced aftertreatment 
technologies, and would bring nonroad diesel fuel to the same 15 ppm 
cap for sulfur that will be required for highway diesel fuel starting 
in 2006. We believe it is highly appropriate to propose dramatic steps 
forward in emissions standards and reductions in sulfur levels in 
nonroad diesel fuel. As discussed throughout this proposal, such steps 
represent a feasible progression in the application of advanced 
emissions control technologies, would achieve needed production of low 
sulfur diesel fuel to enable the advanced emission control 
technologies, the standards are cost-effective, and provide very large 
public health and welfare benefits.
    We followed certain principles when developing the elements of this 
proposal. First, the program must achieve reductions in NOX, 
SOx, and PM emissions as early as possible. This includes reductions 
from the in-use fleet of nonroad diesel engines. Second, as we did in 
the 2007 highway diesel program, we are treating vehicles and fuels as 
a system since we believe this is the best way to achieve the greatest 
emissions reductions. Third, the implementation of low sulfur 
requirements for nonroad diesel fuel must in no way interfere with the 
implementation and expected benefits of introducing ultra low sulfur 
fuel in the highway market, as required by the 2007 highway diesel 
program. Lastly, the program must provide sufficient lead time to allow 
the integration of advanced emissions control technologies from the 
highway sector onto nonroad diesel engines as well as the expansion of 
ultra low sulfur fuel production to the nonroad market.
    This proposal sets out new engine exhaust emissions standards, 
emissions test procedures, including not-to-exceed requirements, for 
nonroad engines, and sulfur control requirements for nonroad, 
locomotive, and marine diesel fuel. The proposed exhaust standards 
would

[[Page 28333]]

result in particulate matter (PM) and nitrogen oxide (NOX) 
emissions levels that are in excess of 95 percent and 90 percent, 
respectively, below comparable levels in effect today. They will begin 
to take effect in the 2008 model year, with a phase-in of standards 
across five different engine power rating groupings. New engine 
emissions test procedures are proposed to take effect with these new 
standards to better ensure emissions control over real-world engine 
operation and to help provide for effective compliance determination. 
Diesel fuel used in nonroad, locomotive, and marine applications would 
meet a 500 ppm cap starting in June 2007, a reduction of approximately 
90%. There are large benefits to taking this first sulfur reduction 
action, especially in the reduction of particulate matter from the in-
use fleet. In 2010, sulfur levels in nonroad diesel fuel (though not 
locomotive or marine diesel fuel) would meet a 15 ppm cap, for a total 
reduction of over 99%. While there are important health and welfare 
benefits associated with the reduction from 500 ppm to 15 ppm, the main 
benefit will be to facilitate the introduction of advanced 
aftertreatment devices on nonroad engines, which would in turn lead to 
significant benefits. We are also seeking comment on and seriously 
considering applying the 15 ppm cap to locomotive and marine diesel 
fuel.
    The requirements in this proposal would result in substantial 
benefits to public health and welfare and the environment through 
significant reductions in emissions of NOX and PM, as well 
as nonmethane hydrocarbons (NMHC), carbon monoxide (CO), sulfur oxides 
(SOX) and air toxics. We project that by 2030, this program 
would reduce annual emissions of NOX, and PM by 827,000, and 
127,000 tons, respectively. These annual emission reductions would 
prevent 9,600 premature deaths, over 8,300 hospitalizations, and almost 
a million work days lost, among quantifiable benefits. The overall 
quantifiable benefits of this rule would be approximately $81 billion 
annually by 2030. Costs for both the engine and fuel requirements would 
be significantly less, at approximately $1.5 billion annually.

A. What Is EPA Proposing?

    This proposal is a further step in EPA's long-term program to 
control emissions from nonroad diesel engines. The EPA has taken 
measures to reduce harmful emissions from nonroad diesel engines in two 
past regulatory actions. A 1994 final rule, developed under provisions 
of section 213 of the Clean Air Act, set initial emissions standards 
for new nonroad diesel engines greater than 50 hp (59 FR 31306, June 
17, 1994). These standards gained modest reductions in NOX 
emissions and are referred to as EPA's ``Tier 1'' standards for large 
nonroad engines. A subsequent final rule published in 1998 set more 
stringent Tier 2 and Tier 3 standards for these engines, as well as 
Tier 1 and Tier 2 standards for the nonroad diesel engines under 50 hp 
(63 FR 56968, October 23, 1998). Nonroad diesel fuel quality is not 
presently regulated by the EPA.
    We also expressed our intent in the 1998 final rule to continue 
evaluating the rapidly changing state of diesel emissions control 
technology, and to perform a review in the 2001 timeframe of the 
technological feasibility of the Tier 3 standards, and of the Tier 2 
standards for engines rated under 50 hp. This review was completed in 
2001 and documented in an EPA staff technical paper that confirmed the 
feasibility of those standards, finding that the number of potential 
control options had expanded since the 1998 final rule to include new 
technologies and more effective application of existing 
technologies.\2\
---------------------------------------------------------------------------

    \2\ ``Nonroad Diesel Emissions Standards Staff Technical 
Paper'', EPA420-R-01-052, October 2001.
---------------------------------------------------------------------------

    There are two basic parts to this proposed program: (1) New exhaust 
emission standards and test procedures for nonroad diesel engines, and 
(2) new sulfur limits for nonroad, locomotive, and marine diesel fuel. 
The systems approach of combining the engine and fuel standards into a 
single program is critical to the success of our overall efforts to 
reduce emissions, because the emission standards will not be feasible 
without the fuel change. This proposal is largely based on the 2007 
highway diesel program.
    We looked at a number of alternative program options, as discussed 
in more detail in section VI below and chapter 12 of the draft 
Regulatory Impact Analysis (RIA). For example, we analyzed a program 
that would require refiners to produce 15 ppm nonroad diesel fuel 
starting in 2008, with appropriate engine standards phased-in beginning 
in 2009. Many of these alternatives provided a very similar level of 
projected emissions control and health and welfare benefits as our 
proposed program. However, taking into account the need for appropriate 
lead time, achieving the greatest possible emissions reductions as 
early as possible, and the interaction of requirements in this proposal 
with existing highway diesel engine environmental programs, we believe 
our proposed program provides the best opportunity for achieving all of 
our goals, as described above, including timely and significant 
emissions reductions from nonroad diesel engines and the associated 
introduction of ultra low sulfur nonroad diesel fuel. We are asking for 
comments on the alternatives discussed in this proposal.
    The elements of the rule are outlined below. Detailed provisions 
and justifications for our proposed rule are discussed in subsequent 
sections and the draft RIA.
1. Nonroad Diesel Engine Emission Standards
    Today's action proposes standards for nonroad diesel engines 
ranging from 3 to over 3,000 horsepower. Applicable emissions standards 
are determined by year for each of five engine power band categories. 
For engines less than 25 hp, we are proposing new engine standards for 
PM (0.30 g/bhp-hr) and CO (4.9 g/bhp-hr) to go along with existing 
NOX standards beginning in 2008. For engines between 25-75 
hp, we are proposing standards reflecting approximately 50% reduction 
in PM control from today's engines applicable in 2008. Then, starting 
in 2013, PM standards of 0.02 g/bhp-hr and NOX standards of 
3.5 g/bhp-hr would apply. For engines between 75-175 hp, the proposed 
standards would be 0.01 g/bhp-hr for PM, 0.30 g/bhp-hr for 
NOX, and 0.14 g/bhp-hr for HC beginning in 2012. These same 
standards would apply for both engines between 175-750 hp and greater 
than 750 hp starting in 2011. These PM, NOX, and NMHC 
standards are similar in stringency to the final standards included in 
the 2007 highway diesel program and are expected to require the use of 
high-efficiency aftertreatment systems to ensure compliance. Thus, 
virtually all nonroad diesel engines after 2013 would likely be using 
advanced aftertreatment systems. We are phasing in many of these 
proposed standards over a period of three years in order to address 
lead time, workload, and feasibility considerations.
    We are also proposing to continue the averaging, banking, and 
trading nonroad emissions credits provisions to demonstrate compliance 
with the standards. In addition, we are proposing to include 
turbocharged diesels in the existing prohibition on crankcase 
emissions, effective in the same year that the proposed Tier 4 
standards first apply in each power category. More specific information 
on the proposed standards can be found in section III below.

[[Page 28334]]

    To better ensure the benefits of the standards are realized in-use 
and throughout the useful life of these engines, we are also proposing 
new test procedures and related certification requirements. We believe 
the new supplemental transient test, Constant Speed Variable Load 
transient duty cycle, cold start transient test, and not-to-exceed test 
procedures and standards will all help achieve our goal. This is a 
significant and important aspect of this proposal that would bring 
greater confidence and certainty to the compliance program.
    The proposal also includes provisions to facilitate the transition 
to the new engine and fuel standards and to encourage the early 
introduction of clean technologies. We are also including proposed 
adjustments to various fuel and engine testing and compliance 
requirements. These provisions are described further in sections III, 
IV, and VI.
2. Nonroad, Locomotive, and Marine Diesel Fuel Quality Standards
    We are proposing that sulfur levels for nonroad diesel fuel be 
reduced from current uncontrolled levels ultimately to 15 ppm, though 
we are proposing an interim cap of 500 ppm. Beginning June 1, 2007, 
refiners would therefore be required to produce nonroad, locomotive, 
and marine diesel fuel that meets a maximum sulfur level of 500 ppm. 
This does not include diesel fuel for home heating, industrial boiler, 
or stationary power uses or diesel fuel used in aircraft. We estimate 
there are significant health and welfare benefits associated with this 
proposed reduction, including reductions in sulfate emissions and 
reduced engine operating expenses. Then, beginning in June 1, 2010, 
fuel used for nonroad diesel applications (excluding locomotive and 
marine engines) is proposed to meet a maximum sulfur level of 15 ppm, 
since all 2011 and later model year nonroad diesel-fueled engines with 
aftertreatment must be refueled with this new ultra low sulfur diesel 
fuel. This sulfur standard is based on our assessment of the impact of 
sulfur on advanced exhaust emission control technologies and a 
corresponding assessment of the feasibility of ultra low sulfur fuel 
production and distribution. We are also asking for comment on bringing 
sulfur levels for locomotive and marine fuel to 15 ppm in 2010 and note 
that we anticipate beginning the process of developing new engine 
controls for these two sources in 2004. This proposal includes a 
combination of provisions available to refiners, especially small 
refiners, to ensure a smooth transition to ultra low sulfur nonroad 
diesel fuel.
    In addition, this proposal includes unique provisions for 
implementing the ultra low sulfur diesel fuel program in the State of 
Alaska. We are also proposing that certain U.S. territories be excluded 
from both the nonroad engine standards and diesel fuel standards. 
Similar actions were taken as part of the 2007 highway diesel program.
    The compliance provisions for ensuring diesel fuel quality are 
essentially consistent with those that have been in effect since 1993 
for highway diesel fuel, reflecting updated requirements that were 
included in the 2007 highway diesel program. Additional compliance 
provisions are proposed for the transition years of the program 
concerning the interaction of the nonroad, locomotive, and marine 
sulfur control requirements with existing highway diesel sulfur control 
provisions. These provisions could also help discourage misfueling of 
nonroad equipment utilizing high-efficiency aftertreatment devices. The 
proposed compliance requirements include provisions that would prohibit 
equipment operators from fueling their machines with higher sulfur 
fuels after completion of the shift to lower sulfur nonroad diesel 
fuels, regardless of the age of their equipment.

B. Why Is EPA Making This Proposal?

1. Nonroad, Locomotive, and Marine Diesels Contribute to Serious Air 
Pollution Problems
    As discussed in detail in section II and chapter 2 and 3 of draft 
RIA, emissions from nonroad, locomotive, and marine diesel engines 
contribute greatly to a number of serious air pollution problems, and 
these emissions would have continued to do so into the future absent 
further controls to reduce them. First, these engines contribute to the 
health and welfare effects associated with ozone, PM, NOX, 
SOX, and volatile organic compounds (VOCs), including toxic 
compounds such as formaldehyde. These adverse effects include premature 
mortality, aggravation of respiratory and cardiovascular disease (as 
indicated by increased hospital admissions and emergency room visits, 
school absences, work loss days, and restricted activity days), changes 
in lung function and increased respiratory symptoms, changes to lung 
tissues and structures, altered respiratory defense mechanisms, chronic 
bronchitis, and decreased lung function.3 4 5 Second and 
importantly, in addition to its contribution to ambient PM inventories, 
diesel exhaust is of specific concern because it has been judged to 
likely pose a lung cancer hazard for humans as well as a hazard from 
noncancer respiratory effects. The Agency has classified diesel exhaust 
as likely to be carcinogenic to humans by inhalation at environmental 
exposures. Third, ozone and PM cause significant public welfare harm. 
Specifically, ozone causes damage to vegetation which leads to economic 
crop and forestry losses, as well as harm to national parks, wilderness 
areas, and other natural systems. PM causes damage to materials and 
soiling of commonly used building materials and culturally important 
items such as statues and works of art. Fourth, NOX, 
SOX and direct emissions of PM contribute to substantial 
visibility impairment in many parts of the U.S. where people live, 
work, and recreate, including mandatory Federal Class I areas. Finally, 
NOX emissions from nonroad diesel engines contribute to the 
acidification, nitrification and eutrophication of water bodies.
---------------------------------------------------------------------------

    \3\ U.S. EPA (1996) Air Quality Criteria for Particulate 
Matter--Volumes I, II, and III, EPA Office of Research and 
Development, National Center for Environmental Assessment, July 
1996. Report No. EPA/600/P-95/001aF, EPA/600/P-95/001bF, EPA/600/P-
95/001cF.
    \4\ U.S. EPA (2002), Air Quality Criteria for Particulate 
Matter--Volumes I and II (Third External Review Draft). This 
material is available electronically at http://cfpub.epa.gov/ncea/
cfm/partmatt.cfm.
    \5\ U.S. EPA (1996) Air Quality Criteria for Ozone and Related 
Photochemical Oxidants. EPA Office of Research and Development, 
National Center for Environmental Assessment, July 1996. Report No. 
EPA/600/P-93/004aF. The document is available on the Internet at 
http://www.epa.gov/ncea/ozone.htm.
---------------------------------------------------------------------------

    Millions of Americans live in areas with unhealthful air quality 
that may endanger public health and welfare (i.e., levels not requisite 
to protect the public health with an adequate margin of safety). Based 
upon data for 1999-2001, there are 291 counties that are violating the 
8-hour ozone NAAQS, totaling 111 million people. In addition, at least 
65 million people in 129 counties live in areas where annual design 
values of ambient PM2.5 violate the PM2.5 NAAQS. 
There are an additional 9 million people in 20 counties where levels 
above the PM2.5 NAAQS are being measured, but the data are 
incomplete. Without emission reductions from the proposed new standards 
for nonroad engines, there is a significant future risk that 32 
counties with 47 million people across the country may violate the 8-
hour ozone national ambient air quality standard (NAAQS) in 2030, based 
on our modeling. Similarly, modeled PM2.5 concentrations in 
107 counties where 85 million people live are above specified levels in 
2030. An additional 64 million people are projected to live in counties

[[Page 28335]]

within 10 percent of the PM2.5 standard in 2030, and 44 
million people are projected to live in counties within 10 percent of 
the level of the 8-hour standard in 2030. Thus, our analyses show that 
these counties face a significant risk of exceeding or failing to 
maintain the PM2.5 and the 8-hour ozone NAAQS without 
significant additional controls between 2007 and 2030.
    Federal, State and local governments are working to bring ozone and 
particulate levels into compliance with the NAAQS through State 
Implementation Plan (SIP) attainment and maintenance plans, and to 
ensure that future air quality reaches and continues to achieve these 
health- and welfare-based standards. The reductions in this proposed 
rulemaking will play a critical part in these important efforts to 
attain and maintain the NAAQS. In addition, reductions from this action 
will also reduce public health and welfare effects associated with 
maintenance of the 1-hour ozone and PM10 NAAQS.
    Emissions from nonroad, locomotive, and marine diesel engines 
account for substantial portions of the country's ambient PM and 
NOX levels. NOX is a key precursor to ozone and 
PM formation. We estimate that these engines account for about ten 
percent of total NOX emissions and about ten percent of 
total PM emissions. These proportions are even higher in some urban 
areas, where these engines contribute up to 19 percent of the total 
NOX emissions and up to 18 percent of the total PM emissions 
inventory. Over time, the relative contribution of these diesel engines 
to air quality problems will go even higher unless EPA takes action to 
further reduce pollution levels. For example, EPA has already taken 
steps to bring emissions levels from light-duty and heavy-duty vehicles 
and engines to near-zero levels by the end of this decade. The PM and 
NOX standards for nonroad, locomotive, and marine diesel 
engines in this proposal would have a substantial impact on emissions. 
By 2030, NOX emissions from these diesel engines under 
today's standards will be reduced by 827,000 tons, and PM emissions 
will decline by about 127,000 tons, dramatically reducing this source 
of NOX and PM emissions. Urban areas, which include many 
poorer neighborhoods, can be disproportionately impacted by such diesel 
emissions, and these neighborhoods will thus receive a relatively 
larger portion of the benefits expected from proposed emissions 
controls. Diesel exhaust is of special concern because it is associated 
with increased risk of lung cancer and respiratory disease. EPA 
recently issued its Health Assessment Document for Diesel Exhaust.\6\ 
The Agency has classified diesel exhaust as likely to be carcinogenic 
to humans by inhalation at environmental exposures. State and local 
governments, in their efforts to protect the health of their citizens 
and comply with requirements of the Clean Air Act (CAA or ``the Act''), 
have recognized the need to achieve major reductions in diesel PM 
emissions, and have been seeking Agency action in setting stringent new 
standards to bring this about.\7\
---------------------------------------------------------------------------

    \6\ U.S. EPA (2002) Health Assessment Document for Diesel Engine 
Exhaust. EPA/600/8-90/057F Office of Research and Development, 
Washington DC. This document is available electronically at http://
cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060.
    \7\ For example, see letters dated April 9, 2002, from Agency 
Secretary of California EPA, Commissioner of NY State DEC, and 
Commissioner of Texas NRCC to Governor Whitman; dated January 28, 
2003 from Western Regional Air Partnership to Governor Whitman, and 
dated December 17, 2002, from State and Territorial Air Pollution 
Program Administrators and Association of Local Air Pollution 
Control Officials and Northeast States for Coordinated Air Use 
Mangement (and other organizations).
---------------------------------------------------------------------------

2. Technology and Fuel Based Solutions
    Although the air pollution from nonroad diesel exhaust is 
challenging, we believe they can be addressed through the application 
of high-efficiency emissions control technologies. As discussed in much 
greater detail in section III, the development of diesel emissions 
control technology has advanced in recent years so that very large 
emission reductions (in excess of 90 percent) are possible, especially 
through the use of catalytic emission control devices installed in the 
nonroad equipment's exhaust system and integrated with the engine 
controls. These devices are often referred to as ``exhaust emission 
control'' or ``aftertreatment'' devices. Exhaust emission control 
devices, in the form of the well-known catalytic converter, have been 
used in gasoline-fueled automobiles for 28 years.
    Based on the Clean Air Act requirements in section 213, we are 
proposing stringent new emission standards that will result in the use 
of these diesel exhaust emission control devices. We are also proposing 
changes to nonroad diesel fuel quality standards, under section 211(c) 
of the Act, in order to enable these high-efficiency technologies.
    To meet the proposed new standards, application of high-efficiency 
exhaust emission controls for both PM and NOX will be needed 
for most engines. High-efficiency PM exhaust emission control 
technology has been available for several years. This technology has 
continued to improve over the years, especially with respect to 
durability and robust operation in use. It has also proved extremely 
effective in reducing exhaust hydrocarbon emissions. Thousands of such 
systems are now in use, especially in Europe. It is the same technology 
we expect to be applied to meet the PM standards in the 2007 heavy-duty 
highway diesel engine rule. However, as discussed in detail in section 
III, these systems are very sensitive to sulfur in the fuel. For the 
technology to be viable and capable of meeting the standards, we 
believe it will require diesel fuel with sulfur content capped at the 
15 ppm level.
    Similarly, high-efficiency NOX exhaust emission control 
technology will be needed if nonroad diesel engines are to attain the 
proposed standards. This is the same technology that we anticipate will 
be applied to heavy-duty highway diesel engines to meet the 
NOX standards included in the 2007 highway diesel program. 
This technology, like the PM technology, is dependant on the 15 ppm 
maximum nonroad diesel fuel levels being proposed in this action in 
order to be feasible and capable of achieving the standards. Similar 
high-efficiency NOX exhaust emission control technology has 
been quite successful in gasoline direct injection engines that operate 
with an exhaust composition fairly similar to diesel exhaust and is 
expected to be used to meet the 2007 and later heavy-duty highway 
diesel standards. As discussed in section III, application of this 
technology to nonroad diesels has some additional engineering 
challenges. In that section, we discuss the current status of this 
technology as well as the major development issues still to be 
addressed and the development steps that can be taken. With the lead-
time available and the introduction of ultra low sulfur nonroad diesel 
fuel, we are confident the proposed application of this technology to 
nonroad diesels would proceed at a reasonable rate of progress and will 
result in systems capable of achieving the standards.
    This view is further supported by the fact that manufacturers are 
already working on developing high-efficiency aftertreatment devices in 
order to have them available for introduction on highway diesel engines 
by 2007. EPA issued a progress report in June 2002 which discussed our 
findings that industry was making substantial progress in developing 
these devices. Additionally, the Clean Diesel Independent Review Panel 
issued a report in October 2002 on similar

[[Page 28336]]

questions and concluded that, while technical issues remain, there were 
no technical hurdles identified that would prevent market introduction 
of high-efficiency aftertreatment devices on schedule.
    The need to reduce sulfur in nonroad diesel fuel is driven by the 
requirements of the exhaust emission control technology that we project 
will be needed to meet the proposed standards for most nonroad diesel 
engines. The challenge in accomplishing the sulfur reduction is driven 
by the capacity to implement the needed refinery modifications, and by 
the costs of making the modifications and running the equipment. Today, 
a number of refiners are acting to provide low sulfur diesel to some 
markets. We believe that controlling the sulfur content of highway 
diesel fuel to the 15 ppm level is necessary, feasible, and cost-
effective.
    Additionally, there are health and welfare benefits associated with 
the initial step of reducing the sulfur level of nonroad, locomotive, 
and marine diesel fuel to 500 ppm. This proposed action will provide 
dramatic, immediate reductions in direct sulfate PM and SO2 
emissions from the in-use fleet. As described in this proposal, we 
believe this fuel control strategy is a cost-effective air quality 
solution as well.
3. Basis for Action Under the Clean Air Act
    Section 213 of the Act gives us the authority to establish 
emissions standards for nonroad engines and vehicles. Section 213(a)(3) 
authorizes the Administrator to set standards for NOX, VOCs, 
or carbon monoxide, to reduce ambient levels of ozone and carbon 
monoxide which ``standards shall achieve the greatest degree of 
emission reduction achievable through the application of technology 
which the Administrator determines will be available for the engines or 
vehicles.'' As part of this determination, the Administrator must give 
appropriate consideration to cost, lead time, noise, energy, and safety 
factors associated with the application of such technology. Section 
213(a)(4) authorizes the Administrator to establish standards to 
control emissions of pollutants which ``may reasonably be anticipated 
to endanger public health and welfare''. Here, the Administrator may 
promulgate regulations that are deemed appropriate for new nonroad 
vehicles and engines which cause or contribute to such air pollution, 
taking into account costs, noise, safety, and energy factors. EPA 
believes the proposed controls for PM in today's rule would be an 
appropriate exercise of EPA's discretion under the authority of section 
213(a)(4).
    We believe the evidence provided in section III and the Draft 
Regulatory Impact Analysis (RIA) indicates that the stringent emission 
standards proposed today are feasible and reflect the greatest degree 
of emission reduction achievable in the model years to which they 
apply. We have given appropriate consideration to costs in proposing 
these standards. Our review of the costs and cost-effectiveness of 
these standards indicate that they will be reasonable and comparable to 
the cost-effectiveness of other emission reduction strategies that have 
been required or could be required in the future. We have also reviewed 
and given appropriate consideration to the energy factors of this rule 
in terms of fuel efficiency and effects on diesel fuel supply, 
production, and distribution, as discussed below, as well as any safety 
factors associated with these proposed standards.
    The information in section II and chapter 2 of the draft RIA 
regarding air quality and the contribution of nonroad, locomotive, and 
marine diesel engines to air pollution provides strong evidence that 
emissions from such engines significantly and adversely impact public 
health or welfare. First, as noted earlier, there is a significant risk 
that several areas will fail to attain or maintain compliance with the 
NAAQS for 8-hour ozone concentrations or for PM2.5 concentrations 
during the period that these new vehicle and engine standards will be 
phased into the vehicle population, and that nonroad, locomotive, and 
marine diesel engines contribute to such concentrations, as well as to 
concentrations of other NAAQS-related pollutants. This risk will be 
significantly reduced by the standards adopted today, as also noted 
above. However, the evidence indicates that some risk remains even 
after the reductions achieved by these new controls on nonroad diesel 
engines and nonroad, locomotive, and marine diesel fuel. Second, EPA 
believes that diesel exhaust is likely to be carcinogenic to humans. 
The risk associated with exposure to diesel exhaust includes the 
particulate and gaseous components among which are benzene, 
formaldehyde, acetaldehyde, acrolein, and 1,3-butadiene, all of which 
are known or suspected human or animal carcinogens, or have serious 
noncancer health effects. Third, emissions from nonroad diesel engines 
(including locomotive and marine diesel engines) contribute to regional 
haze and impaired visibility across the nation, as well as acid 
deposition, POM deposition, eutrophication and nitrification, all of 
which are serious environmental welfare problems.
    EPA has already found in previous rules that emissions from new 
nonroad diesel engines contribute to ozone and carbon monoxide (CO) 
concentrations in more than one area which has failed to attain the 
ozone and carbon monoxide NAAQS. 59 FR 31306 (June 17, 1994). EPA has 
also previously determined that it is appropriate to establish 
standards for PM from new nonroad diesel engines under section 
213(a)(4), and the additional information on diesel exhaust 
carcinogenicity noted above reinforces this finding. In addition, we 
have already found that emissions from nonroad engines significantly 
contribute to air pollution that may reasonably be anticipated to 
endanger public welfare due to regional haze and visibility impairment. 
67 FR 68242, 68243 (Nov. 8, 2002). We find here, based on the 
information in section II of this preamble and chapter 2 of the draft 
RIA, that emissions from the new nonroad diesel engines covered by this 
proposal likewise contribute to regional haze and to visibility 
impairment that may reasonably be anticipated to endanger public 
welfare. Taken together, these findings indicate the appropriateness of 
the nonroad diesel engine standards proposed today for purposes of 
section 213(a)(3) and (4) of the Act.
    Section 211(c) of the CAA allows us to regulate fuels where 
emission products of the fuel either: (1) Cause or contribute to air 
pollution that reasonably may be anticipated to endanger public health 
or welfare, or (2) will impair to a significant degree the performance 
of any emission control device or system which is in general use, or 
which the Administrator finds has been developed to a point where in a 
reasonable time it will be in general use were such a regulation to be 
promulgated. This rule meets both of these criteria. SOx and sulfate PM 
emissions from nonroad, locomotive, marine and diesel vehicles are due 
to sulfur in diesel fuel. As discussed above, emissions of these 
pollutants cause or contribute to ambient levels of air pollution that 
endanger public health and welfare. Control of sulfur to 500 ppm for 
this fuel would lead to significant, cost-effective reductions in 
emissions of these pollutants. The substantial adverse effect of high 
sulfur levels on the performance of diesel emission control devices or 
systems that would be expected to be used to meet the nonroad standards 
is discussed in detail in section III. Control of sulfur to 15 ppm in 
nonroad diesel fuel would enable emissions control technology that will 
achieve significant, cost-

[[Page 28337]]

effective reduction in emissions of these pollutants, as discussed in 
section II below. In addition, our authority under section 211(c) is 
discussed in more detail in Appendix A to the draft RIA.

II. What Is the Air Quality Impact of the Sources Covered by the 
Proposed Rule?

    With this proposal, EPA is acting to extend highway types of 
emission controls to another major source of diesel engine emissions, 
nonroad diesel engines. These emissions are significant contributors to 
atmospheric pollution from particulate matter, ozone and a variety of 
toxic air pollutants. In our most recent nationwide inventory used for 
this proposal (1996), the nonroad diesels affected by this proposal \8\ 
contribute over 43 percent of diesel PM emissions from mobile sources, 
up to 18 percent of PM2.5 emissions in urban areas, and up 
to 14 percent of NOX emissions in urban areas.
---------------------------------------------------------------------------

    \8\ For NOX and PM2.5 this includes all 
land-based nonroad diesel engines, but not locomotive, commercial 
marine vessel, and recreational marine vessel engines. Since the 
latter three engine categories are affected by the fuel sulfur 
portions of the proposal, they are included for SO2.
---------------------------------------------------------------------------

    Without further control beyond those standards we have already 
adopted, by the year 2020, these engines will emit 62 percent of diesel 
PM emissions from mobile sources, up to 19 percent of PM2.5 
emissions in urban areas, and up to 20 percent of NOX 
emissions in urban areas.
    When fully implemented, this proposal would reduce nonroad diesel 
PM2.5 and NOX emissions by more than 90 percent. 
It will also virtually eliminate nonroad diesel SOx 
emissions, which amounted to nearly 230,000 tons in 1996, and would 
otherwise grow to approximately 340,000 tons by 2020.
    These dramatic reductions in nonroad emissions are a critical part 
of the effort by Federal, State and local governments to reduce the 
health-related impacts of air pollution and to reach attainment of the 
NAAQS for PM and ozone, as well as to improve other environmental 
effects such as atmospheric visibility. Based on the most recent data 
available for this rule (1999-2001), such problems are widespread in 
the United States. There are over 65 million people living in counties 
with monitored PM2.5 levels exceeding the PM2.5 
NAAQS, and 111 million people living in counties with monitored 
concentrations exceeding the 8-hour ozone NAAQS. Figure II.-1 
illustrates the widespread nature of these problems. Shown in this 
figure are counties exceeding either or both of the two NAAQS plus 
mandatory Federal Class I areas, which have particular needs for 
reductions in atmospheric haze.
[GRAPHIC] [TIFF OMITTED] TP23MY03.000

    As we will describe later in this preamble, the air quality 
improvements expected from this proposal is anticipated to produce 
major benefits to human health and welfare, with a combined value in 
excess of half a

[[Page 28338]]

trillion dollars between 2007 and 2030. By the year 2030, this proposed 
rule would be expected to prevent approximately 9,600 deaths per year 
from premature mortality, and 16,000 nonfatal heart attacks. It is 
estimated to also prevent 14,000 acute bronchitis attacks in children, 
260,000 respiratory symptoms in children, and nearly 1 million lost 
work days in 2030. The reductions will also improve visibility.
    In the remainder of this section we will describe in more detail 
the air pollution problems associated with emissions from nonroad 
diesel engines, and the emission and air quality benefits we expect to 
realize from the fuel and engine controls in this proposal.

A. Overview

    The emissions from nonroad engines that are being directly 
controlled by the standards in this rulemaking are NOX, PM 
and NMHC, and to a lesser extent, CO. Gaseous air toxics from nonroad 
diesels will also be reduced as a consequence of the proposed 
standards. In addition there will be a substantial reduction in 
SOx emissions resulting from the proposed reduction in 
sulfur level in diesel fuel.
    From a public health perspective, we are primarily concerned with 
nonroad engine contributions to atmospheric levels of particulate 
matter in general, diesel PM in particular and various gaseous air 
toxics emitted by diesel engines, and ozone.\9\ We will first review 
important public health effects linked to these pollutants, briefly 
describing the human health effects and the current and expected future 
ambient levels of direct or indirectly caused pollution. Our 
presentation will show that substantial further reductions of these 
pollutants, and the underlying emissions from nonroad diesel engines, 
are needed to protect public health.
---------------------------------------------------------------------------

    \9\ Ambient particulate matter from nonroad diesel engine is 
associated with the direct emission of diesel particulate matter, 
and with particulate matter formed indirectly in the atmosphere by 
NOX and SOx emissions (and to a lesser extent 
NMHC emissions). Both NOX and NMHC participate in the 
atmospheric chemical reactions that produce ozone.
---------------------------------------------------------------------------

    Following discussion of health effects, we will discuss a number of 
welfare effects associated with emissions from diesel engines. These 
effects include atmospheric visibility impairment, ecological and 
property damage caused by acid deposition, eutrophication and 
nitrification of surface waters, environmental threats posed by 
polycyclic organic matter (POM) deposition, and plant and crop damage 
from ozone. Once again, the information available to us indicates a 
continuing need for further nonroad emission reductions to bring about 
improvements in air quality.
    Next, we will describe our understanding of the engine emission 
inventories for the primary pollutants affected by the proposal. As 
noted above, these include PM, NOX, SOX, Air 
Toxics and HC. We will present current and projected future levels of 
emissions for the base case, including anticipated reductions from 
control programs already adopted by EPA and the States, but without the 
controls proposed today. Then we will identify expected emission 
reductions from nonroad engines. These reductions will make important 
contributions to controlling the health and welfare problems associated 
with ambient PM and ozone levels and with diesel related air toxics.
    While the material we will present in this section will describe 
our understanding of the need for control of nonroad engine emissions 
and the air quality improvements we expect to realize, this section is 
not an exhaustive treatment of these issues. For a fuller understanding 
of the topics treated here, you should refer to the extended 
presentations in the Draft Regulatory Impact Analysis accompanying this 
proposal.

B. Public Health Impacts

1. Particulate Matter
    Particulate matter (PM) represents a broad class of chemically and 
physically diverse substances. It can be principally characterized as 
discrete particles that exist in the condensed (liquid or solid) phase 
spanning several orders of magnitude in size. PM10 refers to 
particles with an aerodynamic diameter less than or equal to a nominal 
10 micrometers. Fine particles refer to those particles with an 
aerodynamic diameter less than or equal to a nominal 2.5 micrometers 
(also known as PM2.5), and coarse fraction particles are 
those particles with an aerodynamic diameter greater than 2.5 microns, 
but less than or equal to a nominal 10 micrometers. Ultrafine PM refers 
to particles with diameters of less than 100 nanometers (0.1 
micrometers). The health and environmental effects of PM are associated 
with fine PM fraction and, in some cases, to the size of the particles. 
Specifically, larger particles (10 [mu]m) tend to be removed 
by the respiratory clearance mechanisms whereas smaller particles are 
deposited deeper in the lungs. Also, particles scatter light 
obstructing visibility.
    The emission sources, formation processes, chemical composition, 
atmospheric residence times, transport distances and other parameters 
of fine and coarse particles are distinct. Fine particles are directly 
emitted from combustion sources and are formed secondarily from gaseous 
precursors such as sulfur dioxide (SOX), oxides of nitrogen 
(NOX), or organic compounds. Fine particles are generally 
composed of sulfate, nitrate, chloride, ammonium compounds, organic 
carbon, elemental carbon, and metals. Nonroad diesels currently emit 
high levels of NOX which react in the atmosphere to form 
secondary PM2.5 (namely ammonium nitrate). Nonroad diesel 
engines also emit SO2 and HC which react in the atmosphere 
to form secondary PM2.5 (namely sulfates and organic 
carbonaceous PM2.5). Combustion of coal, oil, diesel, 
gasoline, and wood, as well as high temperature process sources such as 
smelters and steel mills, produce emissions that contribute to fine 
particle formation. In contrast, coarse particles are typically 
mechanically generated by crushing or grinding. They include 
resuspended dusts and crustal material from paved roads, unpaved roads, 
construction, farming, and mining activities. These coarse particles 
can be either natural in source such as road dust or anthropogenic. 
Fine particles can remain in the atmosphere for days to weeks and 
travel through the atmosphere hundreds to thousands of kilometers, 
while coarse particles deposit to the earth within minutes to hours and 
within tens of kilometers from the emission source.
    The relative contribution of various chemical components to 
PM2.5 varies by region of the country. Data on 
PM2.5 composition are available from the EPA Speciation 
Trends Network in 2001 and the Interagency Monitoring of PROtected 
Visual Environments (IMPROVE) network in 1999 covering both urban and 
rural areas in numerous regions of the U.S. These data show that 
carbonaceous PM2.5 makes up the major component for 
PM2.5 in both urban and rural areas in the western U.S. 
Carbonaceous PM2.5 includes both elemental and organic 
carbon. Nitrates formed from NOX also play a major role in 
the western U.S., especially in the California area where it is 
responsible for about a quarter of the ambient PM2.5 
concentrations. Sulfate plays a lesser role in these regions. For the 
eastern and mid U.S., these data show that both sulfates and 
carbonaceous PM2.5 are major contributors to ambient 
PM2.5 in both urban and rural areas. In some eastern areas, 
carbonaceous PM2.5 is responsible for up to half of ambient 
PM2.5 concentrations. Sulfate is also a

[[Page 28339]]

major contributor to ambient PM2.5 in the eastern U.S. and 
in some areas make greater contributions than carbonaceous 
PM2.510 11
---------------------------------------------------------------------------

    \10\ Rao, Venkatesh; Frank, N.; Rush, A.; and Dimmick, F. 
(November 13-15, 2002). Chemical speciation of PM2.5 in 
urban and rural areas (November 13-15, 2002) In the Proceedings of 
the Air & Waste Management Association Symposium on Air Quality 
Measurement Methods and Technology, San Francisco Meeting.
    \11\ EPA (2002) Latest Finds on National Air Quality, EPA 454/K-
02-001.
---------------------------------------------------------------------------

    Nonroad engines, and most importantly nonroad diesel engines, 
contribute significantly to ambient PM2.5 levels, largely 
through emissions of carbonaceous PM2.5. Carbonaceous 
PM2.5 is a major portion of ambient PM2.5, 
especially in populous urban areas. Nonroad diesels also emit high 
levels of NOX which react in the atmosphere to form 
secondary PM2.5 (namely nitrate). Nonroad diesels also emit 
SO2 and NMHC which react in the atmosphere to form secondary 
PM2.5 (namely sulfates and organic carbonaceous 
PM2.5). For more details, consult the draft RIA for this 
proposed rule.
    Diesel particles from nonroad diesel are a component of both coarse 
and fine PM, but fall mainly in the fine (and even ultrafine) size 
range. As discussed later, diesel PM also contains small quantities of 
numerous mutagenic and carcinogenic compounds associated with the 
particulate (and also organic gases). In addition, while toxic trace 
metals emitted by nonroad diesel engines represent a very small portion 
of the national emissions of metals (less than one percent) and a small 
portion of diesel PM (generally less than one percent of diesel PM), we 
note that several trace metals of potential toxicological significance 
and persistence in the environment are emitted by diesel engines. These 
trace metals include chromium, manganese, mercury and nickel. In 
addition, small amounts of dioxins have been measured in highway engine 
diesel exhaust, some of which may partition into the particulate phase; 
dioxins through out the environment are a major health concern 
(although the diesel contribution has not been judged significant at 
this point). Diesel engines also emit polycyclic organic matter (POM), 
including polycyclic aromatic hydrocarbons (PAH), which can be present 
in both gas and particle phases of diesel exhaust. Many PAH compounds 
are classified by EPA as probable human carcinogens.
    For additional, detailed, information on PM beyond that summarized 
below, see the draft Regulatory Impact Analysis.
a. Health Effects of PM2.5 and PM10
    Scientific studies show ambient PM (which is attributable to a 
number of sources, including nonroad diesel) is associated with a 
series of adverse health effects. These health effects are discussed in 
detail in the EPA Criteria Document for PM as well as the draft updates 
of this document released in the past year.12 13 In 
addition, EPA's final ``Health Assessment Document for Diesel Engine 
Exhaust,'' (the Diesel HAD) also reviews health effects information 
related to diesel exhaust as a whole including diesel PM, which is one 
component of ambient PM.\14\
---------------------------------------------------------------------------

    \12\ U.S. EPA (1996.) Air Quality Criteria for Particulate 
Matter--Volumes I, II, and III, EPA, Office of Research and 
Development. Report No. EPA/600/P-95/001a-cF. This material is 
available electronically at http://www.epa.gov/ttn/oarpg/ticd.html.
    \13\ U.S. EPA (2002). Air Quality Criteria for Particulate 
Matter--Volumes I and II (Third External Review Draft) This material 
is available electronically at http://cfpub.epa.gov/ncea/cfm/
partmatt.cfm.
    \14\ U.S. EPA (2002). Health Assessment Document for Diesel 
Engine Exhaust. EPA/600/8-90/057F Office of Research and 
Development, Washington DC. This document is available 
electronically at http://cfpub.epa.gov/ncea/cfm/
recordisplay.cfm?deid=29060.
---------------------------------------------------------------------------

    As described in these documents, health effects associated with 
short-term variation in ambient particulate matter (PM) have been 
indicated by epidemiologic studies showing associations between 
exposure and increased hospital admissions for ischemic heart disease, 
heart failure, respiratory disease, including chronic obstructive 
pulmonary disease (COPD) and pneumonia. Short-term elevations in 
ambient PM have also been associated with increased cough, lower 
respiratory symptoms, and decrements in lung function. Short-term 
variations in ambient PM have also been associated with increases in 
total and cardiorespiratory daily mortality. Studies examining 
populations exposed to different levels of air pollution over a number 
of years, including the Harvard Six Cities Study and the American 
Cancer Society Study suggest an association between exposure to ambient 
PM2.5 and premature mortality, including deaths attributed 
to lung cancer.15 16 Two studies further analyzing the 
Harvard Six Cities Study's air quality data have also established a 
specific influence of mobile source-related PM2.5 on daily 
mortality \17\ and a concentration-response function for mobile source-
associated PM2.5 and daily mortality.\18\ Another recent 
study in 14 U.S. cities examining the effect of PM10 on 
daily hospital admissions for cardiovascular disease found that the 
effect of PM10 was significantly greater in areas with a 
larger proportion of PM10 coming from motor vehicles, 
indicating that PM10 from these sources may have a greater 
effect on the toxicity of ambient PM10 when compared with 
other sources.\19\ Additional studies have associated changes in heart 
rate and/or heart rhythm in addition to changes in blood 
characteristics with exposure to ambient PM.20 21 For 
additional information on health effects, see the draft RIA.
---------------------------------------------------------------------------

    \15\ Dockery, DW; Pope, CA, III; Xu, X; et al. (1993) An 
association between air pollution and mortality in six U.S. cities. 
N Engl J Med 329:1753-1759.
    \16\ Pope, CA, III; Thun, MJ; Namboordiri, MM; et al. (1995) 
Particulate air pollution as a predictor of mortality in a 
prospective study of U.S. adults. Am J Respir Crit Care Med 151:669-
674.
    \17\ Laden F; Neas LM; Dockery DW; et al. (2000) Association of 
fine particulate matter from different sources with daily mortality 
in six U.S. cities. Environ Health Perspect 108(10):941-947.
    \18\ Schwartz J; Laden F; Zanobetti A. (2002) The concentration-
response relation between PM(2.5) and daily deaths. Environ Health 
Perspect 110(10): 1025-1029.
    \19\ Janssen NA; Schwartz J; Zanobetti A.; et al. (2002) Air 
conditioning and source-specific particles as modifiers of the 
effect of PM10 on hospital admissions for heart and lung 
disease. Environ Health Perspect 110(1):43-49.
    \20\ Pope CA III, Verrier RL, Lovett EG; et al. (1999) Heart 
rate variability associated with particulate air pollution. Am Heart 
J 138(5 Pt 1):890-899.
    \21\ Magari SR, Hauser R, Schwartz J; et al. (2001) Association 
of heart rate variability with occupational and environmental 
exposure to particulate air pollution. Circulation 104(9):986-991.
---------------------------------------------------------------------------

    The health effects of PM10 are similar to those of 
PM2.5, since PM10 includes all of 
PM2.5 plus the coarse fraction from 2.5 to 10 micrometers in 
size. EPA is also evaluating the health effects of PM between 2.5 and 
10 micrometers in the draft revised Criteria Document. As discussed in 
the Diesel HAD and other studies, most diesel PM is smaller than 2.5 
micrometers.\22\ Both fine and coarse fraction particles can enter and 
deposit in the respiratory system.
---------------------------------------------------------------------------

    \22\ U.S. EPA (1985). Size specific total particulate emission 
factor for mobile sources. EPA 460/3-85-005. Office of Mobile 
Sources, Ann Arbor, MI.
---------------------------------------------------------------------------

    In addition to the information in the draft revised Criteria 
Document, the relevance of health effects associated with on-road 
diesel engine-generated PM to nonroad applications is supported by the 
observation in the Diesel HAD that the particulate characteristics in 
the zone around nonroad diesel engines is likely to be substantially 
the same as published air quality measurements made along busy 
roadways.
    Of particular relevance to this rule is a recent cohort study which 
examined the association between mortality and

[[Page 28340]]

residential proximity to major roads in the Netherlands. Examining a 
cohort of 55 to 69 year-olds from 1986 to1994, the study indicated that 
long-term residence near major roads, an index of exposure to primary 
mobile source emissions (including diesel exhaust), was significantly 
associated with increased cardiopulmonary mortality.\23\ Other studies 
have shown children living near roads with high truck traffic density 
have decreased lung function and greater prevalence of lower 
respiratory symptoms compared to children living on other roads.\24\ A 
recent review of epidemiologic studies examining associations between 
asthma and roadway proximity concluded that some coherence was evident 
in the literature, indicating that asthma, lung function decrement, 
respiratory symptoms, and other respiratory problems appear to occur 
more frequently in people living near busy roads.\25\ As discussed 
later, nonroad diesel engine emissions, especially particulate, are 
similar in composition to those from highway diesel vehicles. Although 
difficult to associate directly with PM2.5, these studies 
indicate that direct emissions from mobile sources, and diesel engines 
specifically, may explain a portion of respiratory health effects 
observed in larger-scale epidemiologic studies. Recent studies 
conducted in Los Angeles have illustrated that a substantial increase 
in the concentration of ultrafine particles is evident in locations 
near roadways, indicating substantial differences in the nature of PM 
immediately near mobile source emissions.\26\
---------------------------------------------------------------------------

    \23\ Hoek, G; Brunekreef, B; Goldbohm, S; et al. (2002) 
Association between mortality and indicators of traffic-related air 
pollution in the Netherlands: a cohort study. Lancet 360(9341): 
1203-1209.
    \24\ Brunekreef, B; Janssen NA; de Hartog, J; et al. (1997) Air 
pollution from traffic and lung function in children living near 
motor ways. Epidemiology (8): 298-303.
    \25\ Delfino RJ. (2002) Epidemiologic evidence for asthma and 
exposure to air toxics: linkages between occupational, indoor, and 
community air pollution research. Env Health Perspect Suppl 110(4): 
573-589.
    \26\ Yifang Zhu, William C. Hinds, Seongheon Kim, Si Shen and 
Constantinos Sioutas Zhu Y; Hinds WC; Kim S; et al. (2002) Study of 
ultrafine particles near a major highway with heavy-duty diesel 
traffic. Atmos Environ 36(27): 4323-4335.
---------------------------------------------------------------------------

    Also, as discussed in more detail later, in addition to its 
contribution to ambient PM inventories, diesel PM is of special concern 
because diesel exhaust has been associated with an increased risk of 
lung cancer. As also discussed later in more detail, we concluded that 
diesel exhaust ranks with other substances that the national-scale air 
toxics assessment suggests pose the greatest relative risk.
b. Current and Projected Levels
    There are NAAQS for both PM10 and PM2.5. 
Violations of the annual PM2.5 standard are much more 
widespread than are violations of the PM10 standards. 
Emission reductions needed to attain the PM2.5 standards 
will also assist in attaining and maintaining compliance with the 
PM10 standards. Thus, since most PM emitted by diesel 
nonroad engines is fine PM, the emission controls proposed today should 
contribute to attainment and maintenance of the existing PM NAAQS. More 
broadly, the proposed standards will benefit public health and welfare 
through reductions in direct diesel PM and reductions of 
NOX, SOX, and NMHCs which contribute to secondary 
formation of PM. The reductions from these proposed rules will assist 
States as they implement local controls as needed to help their areas 
attain and maintain the standards.
i. PM10 Levels
    The current NAAQS for PM10 were established in 1987. The 
primary (health-based) and secondary (public welfare based) standards 
for PM10 include both short- and long-term NAAQS. The short-
term (24 hour) standard of 150 ug/m3 is not to be exceeded 
more than once per year on average over three years. The long-term 
standard specifies an expected annual arithmetic mean not to exceed 50 
ug/m3 averaged over three years.
    Currently, 29 million people live in PM10 nonattainment 
areas. There are currently 58 moderate PM10 nonattainment 
areas with a total population of 6.8 million. The attainment date for 
the initial moderate PM10 nonattainment areas, designated by 
operation of law on November 15, 1990, was December 31, 1994. Several 
additional PM10 nonattainment areas were designated on 
January 21, 1994, and the attainment date for these areas was December 
31, 2000. There are an additional 8 serious PM10 
nonattainment areas with a total affected population of 22.7 million. 
According to the Act, serious PM10 nonattainment areas must 
attain the standards no later than 10 years after designation. The 
initial serious PM10 nonattainment areas were designated 
January 18, 1994, and had an attainment date set by the Act of December 
31, 2001. The Act provides that EPA may grant extensions of the serious 
area attainment dates of up to 5 years, provided that the area 
requesting the extension meets the requirements of section 188(e) of 
the Act. Four serious PM10 nonattainment areas (Phoenix, 
Arizona; Coachella Valley, South Coast (Los Angeles), and Owens Valley, 
California) have received extensions of the December 31, 2001, 
attainment date and thus have new attainment dates of December 31, 
2006.\27\ While all of these areas are expected to be in attainment 
before the emission reductions from this proposed rule are expected to 
occur, these reductions will be important to assist these areas in 
maintaining the standards.
---------------------------------------------------------------------------

    \27\ EPA has also proposed to grant Las Vegas, Nevada, an 
extension until December 31, 2006.
---------------------------------------------------------------------------

ii. PM2.5 Levels
    The need for reductions in the levels of PM2.5 is 
widespread. Figure II-1 at the beginning of this air quality section 
highlighted monitor locations measuring concentrations above the level 
of the NAAQS. As can be seen from that figure, high ambient levels are 
widespread throughout the country.
    The NAAQS for PM2.5 were established by EPA in 1997 (62 
FR 38651, July 18, 1997). The short term (24-hour) standard is set at a 
level of 65 [mu]g/m3 based on the 98th percentile 
concentration averaged over three years. (This air quality statistic 
compared to the standard is referred to as the ``design value.'') The 
long-term standard specifies an expected annual arithmetic mean not to 
exceed 15 ug/m3 averaged over three years.
    Current PM2.5 monitored values for 1999-2001, which 
cover counties having about 75 percent of the country's population, 
indicate that at least 65 million people in 129 counties live in areas 
where annual design values of ambient fine PM violate the 
PM2.5 NAAQS. There are an additional 9 million people in 20 
counties where levels above the NAAQS are being measured, but there are 
insufficient data at this time to calculate a design value in 
accordance with the standard, and thus determine whether these areas 
are violating the PM2.5 NAAQS. In total, this represents 37 
percent of the counties and 64 percent of the population in the areas 
with monitors with levels above the NAAQS. Furthermore, an additional 
14 million people live in 41 counties that have air quality 
measurements within 10 percent of the level of the standard. These 
areas, although not currently violating the standard, will also benefit 
from the additional reductions from this rule in order to ensure long 
term maintenance.
    Our air quality modeling performed for this proposal also indicates 
that similar conditions are likely to continue

[[Page 28341]]

to exist in the future in the absence of additional controls. For 
example, in 2020 based on emission controls currently adopted, we 
project that 66 million people will live in 79 counties with average 
PM2.5 levels above 15 ug/m\3\. In 2030, the number of people 
projected to live in areas exceeding the PM2.5 standard is 
expected to increase to 85 million in 107 counties. An additional 24 
million people are projected to live in counties within 10 percent of 
the standard in 2020, which will increase to 64 million people in 2030.
    Our modeling also indicates that the reductions we are expecting 
will make a substantial contribution to reducing exposures in these 
areas.\28\ In 2020, the number of people living in counties with 
PM2.5 levels above the NAAQS would be reduced from 66 
million to 60 million living in 67 counties, which reflects a reduction 
of 9 percent in potentially exposed population and 15 percent of the 
number of counties. In 2030, there would be a reduction from 85 million 
people to 71 million living in 84 counties. These represent even 
greater improvements than projected for 2020 (numbers of people 
potentially exposed down 16 percent and number of counties down 21 
percent). Furthermore, our modeling also shows that the emission 
reductions would assist areas with future maintenance of the standards.
---------------------------------------------------------------------------

    \28\ The results illustrate the type of PM changes for the 
preliminary control option, as discussed in the Draft RIA in section 
3.6. The proposal differs from the modeled control case based on 
updated information; however, we believe that the net results would 
approximate future emissions, although we anticipate the PM 
reductions might be slightly smaller.
---------------------------------------------------------------------------

    We estimate that the reduction of PM levels expected from this 
proposed rule would produce nationwide air quality improvements in PM 
levels. On a population weighted basis, the average change in future 
year annual averages would be a decrease of 0.33 ug/m\3\ in 2020, and 
0.46 ug/m\3\ in 2030. The reductions are discussed in more detail in 
chapter 2 of the draft RIA.
    While the final implementation process for bringing the nation's 
air into attainment with the PM2.5 NAAQS is still being 
completed in a separate rulemaking action, the basic framework is well 
defined by the statute. EPA's current plans call for designating 
PM2.5 nonattainment areas in late-2004. Following 
designation, Section 172(b) of the Clean Air Act allows states up to 
three years to submit a revision to their state implementation plan 
(SIP) that provides for the attainment of the PM2.5 
standard. Based on this provision, states could submit these SIPs as 
late as the end of 2007. Section 172(a)(2) of the Clean Air Act 
requires that these SIP revisions demonstrate that the nonattainment 
areas will attain the PM2.5 standard as expeditiously as 
practicable but no later than five years from the date that the area 
was designated nonattainment. However, based on the severity of the air 
quality problem and the availability and feasibility of control 
measures, the Administrator may extend the attainment date ``for a 
period of no greater than 10 years from the date of designation as 
nonattainment.'' Therefore, based on this information, we expect that 
most or all areas will need to attain the PM2.5 NAAQS in the 
2009 to 2014 time frame, and then be required to maintain the NAAQS 
thereafter.
    Since the emission reductions expected from this proposal would 
begin in this same time frame, the projected reductions in nonroad 
emissions would be used by states in meeting the PM2.5 
NAAQS. States and state organizations have told EPA that they need 
nonroad diesel engine reductions in order to be able to meet and 
maintain the PM2.5 NAAQS as well as visibility regulations, 
especially in light of the otherwise increasing emissions from nonroad 
sources without more stringent standards.29 30 31 
Furthermore, this action would ensure that nonroad diesel emissions 
will continue to decrease as the fleet turns over in the years beyond 
2014; these reductions will be important for maintenance of the NAAQS 
following attainment. The future reductions are also important to 
achieve visibility goals, as discussed later.
---------------------------------------------------------------------------

    \29\ California Air Resources Board and New York State 
Department of Environmental Conservation (April 9, 2002), Letter to 
EPA Administrator Christine Todd Whitman.
    \30\ State and Territorial Air Pollution Program Administrators 
(STAPPA) and Association of Local Air Pollution Control Officials 
(ALAPCO) (December 17, 2002), Letter to EPA Assistant Administrator 
Jeffrey R. Holmstead.
    \31\ Western Regional Air Partnership (WRAP) (January 28, 2003), 
Letter to Governor Christine Todd Whitman.
---------------------------------------------------------------------------

2. Air Toxics
a. Diesel Exhaust
    A number of health studies have been conducted regarding diesel 
exhaust including epidemiologic studies of lung cancer in groups of 
workers, and animal studies focusing on non-cancer effects specific to 
diesel exhaust. Diesel exhaust PM (including the associated organic 
compounds which are generally high molecular weight hydrocarbon types 
but not the more volatile gaseous hydrocarbon compounds) is generally 
used as a surrogate measure for diesel exhaust.
i. Potential Cancer Effects of Diesel Exhaust
    In addition to its contribution to ambient PM inventories, diesel 
exhaust is of specific concern because it has been judged to pose a 
lung cancer hazard for humans as well as a hazard from noncancer 
respiratory effects.
    EPA recently released its ``Health Assessment Document for Diesel 
Engine Exhaust,'' (the Diesel HAD).\32\ There, diesel exhaust was 
classified as likely to be carcinogenic to humans by inhalation at 
environmental exposures, in accordance with the revised draft 1996/1999 
EPA cancer guidelines. A number of other agencies (National Institute 
for Occupational Safety and Health, the International Agency for 
Research on Cancer, the World Health Organization, California EPA, and 
the U.S. Department of Health and Human Services) have made similar 
classifications. It should be noted that the conclusions in the Diesel 
HAD were based on diesel engines currently in use, including nonroad 
diesel engines such as those found in bulldozers, graders, excavators, 
farm tractor drivers and heavy construction equipment. As new diesel 
engines with significantly cleaner exhaust emissions replace existing 
engines, the conclusions of the Diesel HAD will need to be reevaluated.
---------------------------------------------------------------------------

    \32\ U.S. EPA (2002). Health Assessment Document for Diesel 
Engine Exhaust. EPA/600/8-90/057F Office of Research and 
Development, Washington DC. This document is available 
electronically at http://cfpub.epa.gov/ncea/cfm/
recordisplay.cfm?deid=29060.
---------------------------------------------------------------------------

    For the EPA Diesel HAD, EPA reviewed 22 epidemiologic studies in 
detail, finding increased lung cancer risk in 8 out of 10 cohort 
studies and 10 out of 12 case-control studies. Relative risk for lung 
cancer associated with exposure range from 1.2 to 2.6. In addition, two 
meta-analyses of occupational studies of diesel exhaust and lung cancer 
have estimated the smoking-adjusted relative risk of 1.35 and 1.47, 
examining 23 and 30 studies, respectively.33 34  That is, 
these two studies show an overall increase in lung cancer for the 
exposed groups of 35 percent and 47 percent compared to the groups not 
exposed to diesel exhaust. In the EPA Diesel HAD, EPA selected 1.4

[[Page 28342]]

as a reasonable estimate of occupational relative risk for further 
analysis.
---------------------------------------------------------------------------

    \33\ Bhatia, R., Lopipero, P., Smith, A. (1998). Diesel exhaust 
exposure and lung cancer. Epidemiology 9(1):84-91.
    \34\ Lipsett, M: Campleman, S.; (1999). Occupational exposure to 
diesel exhaust and lung cancer: a meta-analysis. Am J Public Health 
80(7):1009-1017.
---------------------------------------------------------------------------

    EPA generally derives cancer unit risk estimates to calculate 
population risk more precisely from exposure to carcinogens. In the 
simplest terms, the cancer unit risk is the increased risk associated 
with average lifetime exposure of 1 ug/m\3\. EPA concluded in the 
Diesel HAD that it is not possible currently to calculate a cancer unit 
risk for diesel exhaust due to a variety of factors that limit the 
current studies, such as a lack of standard exposure metric for diesel 
exhaust and the absence of quantitative exposure characterization in 
retrospective studies.
    EPA generally derives cancer unit risk estimates to calculate 
population risk more precisely from exposure to carcinogens. In the 
simplest terms, the cancer unit risk is the increased risk associated 
with average lifetime exposure of 1 ug/m\3\. EPA concluded in the 
Diesel HAD that it is not possible currently to calculate a cancer unit 
risk for diesel exhaust due to a variety of factors that limit the 
current studies, such as lack of an adequate dose-response relationship 
between exposure and cancer incidence.
    However, in the absence of a cancer unit risk, the EPA Diesel HAD 
sought to provide additional insight into the possible ranges of risk 
that might be present in the population. Such insights, while not 
confident or definitive, nevertheless contribute to an understanding of 
the possible public health significance of the lung cancer hazard. The 
possible risk range analysis was developed by comparing a typical 
environmental exposure level to a selected range of occupational 
exposure levels and then proportionally scaling the occupationally 
observed risks according to the exposure ratio's to obtain an estimate 
of the possible environmental risk. If the occupational and 
environmental exposures are similar, the environmental risk would 
approach the risk seen in the occupational studies whereas a much 
higher occupational exposure indicates that the environmental risk is 
lower than the occupational risk. A comparison of environmental and 
occupational exposures showed that for certain occupations the 
exposures are similar to environmental exposures while, for others, 
they differ by a factor of about 200 or more.
    The first step in this process is to note that the occupational 
relative risk of 1.4, or a 40 percent from increased risk compared to 
the typical 5 percent lung cancer risk in the U.S. population, 
translates to an increased risk of 2 percent (or 10-2) for 
these diesel exhaust exposed workers. The Diesel HAD derived a typical 
nationwide average environmental exposure level of 0.8 ug./m\3\ for 
diesel PM from highway sources for 1996. Diesel PM is a surrogate for 
diesel exhaust and, as mentioned above, has been classified as a 
carcinogen by some agencies.
    This estimate was based on national exposure modeling; the 
derivation of this exposure is discussed in detail in the EPA Diesel 
HAD. The possible risk range in the environment was estimated by taking 
the relative risks in the occupational setting, EPA selected 1.4 and 
converting this to absolute risk of 2% and then ratioing this risk by 
differences in the occupational vs environmental exposures of interest. 
A number of calculations are needed to accomplish this, and these can 
be seen in the EPA Diesel HAD. The outcome was that environmental risks 
from diesel exhaust exposure could range from a low of 10-4 
to 10-5 or be as high as 10-3 this being a 
reflection of the range of occupational exposures that could be 
associated with the relative and absolute risk levels observed in the 
occupational studies.
    While these risk estimates are exploratory and not intended to 
provide a definitive characterization of cancer risk, they are useful 
in gauging the possible range of risk based on reasonable judgement. It 
is important to note that the possible risks could also be higher or 
lower and a zero risk cannot be ruled out. Some individuals in the 
population may have a high tolerance to exposure from diesel exhaust 
and low cancer susceptibility. Also, one cannot rule out the 
possibility of a threshold of exposure below which there is no cancer 
risk, although evidence has not been seen or substantiated on this 
point.
    Also, as discussed in the Diesel HAD, there is a relatively small 
difference between some occupational settings where increased lung 
cancer risk is reported and ambient environmental exposures. The 
potential for small exposure differences underscores the 
appropriateness of the extrapolation from occupational risk to ambient 
environmental exposure levels is reasonable and appropriate.
    EPA also recently completed an assessment of air toxic emissions 
(the National-Scale Air Toxics Assessment or NATA) and their associated 
risk, and we concluded that diesel exhaust ranks with other substances 
that the national-scale assessment suggests pose the greatest relative 
risk.\35\ This assessment estimates average population inhalation 
exposures to diesel PM in 1996 for nonroad as well as on-road sources. 
These are the sum of ambient levels in various locations weighted by 
the amount of time people spend in each of the locations. This analysis 
shows a somewhat higher diesel exposure level than the 0.8 ug/m\3\ used 
to develop the risk perspective in the Diesel HAD. The NATA levels are 
1.4 ug/m\3\ total with an on-road source contribution of 0.5 ug/m\3\ to 
average nationwide exposure in 1996 and a nonroad source contribution 
of 0.9 ug/m\3\. The average urban exposure concentration was 1.6 ug/
m\3\ and the average rural concentration was 0.55 ug/m\3\. In five 
percent of urban census tracts across the United States, average 
concentrations were above 4.3 ug/m\3\. The Diesel HAD states that use 
of the NATA exposure number results instead of the 0.8 ug/m\3\ results 
in a similar risk perspective.
---------------------------------------------------------------------------

    \35\ U.S. EPA (2002). National-Scale Air Toxics Assessment. This 
material is available electronically at http://www.epa.gov/ttn/atw/
nata/.
---------------------------------------------------------------------------

    In 2001, EPA completed a rulemaking on mobile source air toxics 
with a determination that diesel particulate matter and diesel exhaust 
organic gases be identified as a Mobile Source Air Toxic (MSAT).\36\ 
This determination was based on a draft of the Diesel HAD on which the 
Clean Air Scientific Advisory Committee of the Science Advisory Board 
had reached closure. The purpose of the MSAT list is to provide a 
screening tool that identifies compounds emitted from motor vehicles or 
their fuels for which further evaluation of emissions controls is 
appropriate.
---------------------------------------------------------------------------

    \36\ U.S. EPA (2001). Control of Emissions of Hazardous Air 
Pollutants from Mobile Sources; Final Rule. 66 FR 17230-17273 (March 
29, 2001).
---------------------------------------------------------------------------

    In summary, even though EPA does not have a specific carcinogenic 
potency with which to accurately estimate the carcinogenic impact of 
diesel PM, the likely hazard to humans at environmental exposure levels 
leads us to conclude that diesel exhaust emissions of PM and organic 
gases should be reduced from nonroad engines in order to protect public 
health.
ii. Other Health Effects of Diesel Exhaust
    The acute and chronic exposure-related effects of diesel exhaust 
emissions are also of concern to the Agency. The Diesel HAD established 
an inhalation Reference Concentration (RfC) specifically based on 
animal studies of diesel exhaust. An RfC is defined by EPA as ``an 
estimate of a continuous inhalation exposure to the human population, 
including sensitive subgroups, with uncertainty spanning

[[Page 28343]]

perhaps an order of magnitude, that is likely to be without appreciable 
risks of deleterious noncancer effects during a lifetime.'' EPA derived 
the RfC from consideration of four chronic rat inhalation studies 
showing adverse pulmonary effects. The diesel RfC is based on a ``no 
observable adverse effect'' level of 144 ug/m\3\ that is further 
reduced by applying uncertainty factors of 3 for interspecies 
extrapolation and 10 for human variations in sensitivity. The resulting 
RfC derived in the Diesel HAD is 5 ug/m\3\ for diesel exhaust as 
measured by diesel PM. This RfC does not consider allergenic effects 
such as those associated with asthma or immunologic effects. There is 
growing evidence that diesel exhaust can exacerbate these effects, but 
the exposure-response data is presently lacking to derive an RfC. 
Again, this RfC is based on animal studies and is meant to estimate 
exposure that is unlikely to have deleterious effects on humans based 
on those studies alone.
    The Diesel HAD also briefly summarizes health effects associated 
with ambient PM and the EPA's annual NAAQS for PM2.5 of 15 
ug/m\3\. There is a much more extensive body of human data showing a 
wide spectrum of adverse health effects associated with exposure to 
ambient PM, of which diesel exhaust is an important component due to 
its large contribution to ambient concentrations. The RfC is not meant 
to say that 5 ug/m\3\ provides adequate public health protection for 
ambient PM2.5. There may be benefits to reducing diesel PM 
below 5 ug/m\3\ since diesel PM is a major contributor to ambient 
PM2.5. Recent epidemiologic studies of ambient PM2.5 do not 
indicate a threshold of effects at low concentrations.\37\
---------------------------------------------------------------------------

    \37\ EPA-SAB-Council-ADV-99-012, 1999. The Clean Air Act 
Amendments Section 812 Prospective Study of Costs and Benefits 
(1999): Advisory by the Health and Ecological Effects Subcommittee 
on Initial Assessments of Health and Ecological Effects, Part 1. 
July 28, 1999.
---------------------------------------------------------------------------

    Also, as mentioned earlier in the health effects discussion for 
PM2.5, there are a number of other health effects associated 
with PM in general, and motor vehicle exhaust including diesels in 
particular, that provide additional evidence for the need for 
significant emission reductions from nonroad diesel sources. For 
example, the Diesel HAD notes that acute or short-term exposure to 
diesel exhaust can cause acute irritation (e.g., eye, throat, 
bronchial), neurophysiological symptoms (e.g., lightheadedness, 
nausea), and respiratory symptoms (e.g., cough, phlegm). There is also 
evidence for an immunologic effect such as the exacerbation of 
allergenic responses to know allergens and asthma-like symptoms. All of 
these health effects plus the designation of diesel exhaust as a likely 
human carcinogen provide ample health justification for control.
iii. Ambient Levels and Exposure to Diesel Exhaust PM
    Because diesel PM is part of overall ambient PM and cannot be 
easily distinguished from overall PM, we do not have direct 
measurements of diesel PM in the ambient air. Ambient diesel PM 
concentrations are estimated instead using one of three approaches: (1) 
Ambient air quality modeling based on diesel PM emission inventories; 
(2) using elemental carbon concentrations in monitored data as 
surrogates; or (3) using the chemical mass balance (CMB) model in 
conjunction with ambient PM measurements. (Also, in addition to CMB, 
UNMIX/PMF have also been used). Estimates using these three approaches 
are described below. In addition, estimates developed using the first 
two approaches above are subjected to a statistical comparison to 
evaluate overall reasonableness of estimated concentrations. It is 
important to note that, while there are inconsistencies in some of 
these studies on the relative importance of gasoline and diesel PM, the 
studies which are discussed in the Diesel HAD all show that diesel PM 
is a significant contributor to overall ambient PM. Some of the studies 
differentiate nonroad from on-road diesel PM.
(1) Air Quality Modeling
    In addition to the general ambient PM modeling conducted for this 
proposal, diesel PM concentrations specifically were recently estimated 
for 1996 as part of NATA. In this assessment, the PM inventory 
developed for the recent regulation promulgating 2007 heavy duty 
vehicle standards was used. Note that the nonroad inventory used in 
this modeling was based on an older version of the draft NONROAD Model 
which showed higher diesel PM than the current version. Ambient impacts 
of mobile source emissions were predicted using the Assessment System 
for Population Exposure Nationwide (ASPEN) dispersion model. Overall 
mean annual national levels for both on-road and nonroad diesels of 
2.06 ug/m\3\ diesel PM were calculated with a mean of 2.41 in urban 
counties and 0.74 in rural counties. These are ambient levels such as 
would be seen at monitors rather than the exposure levels discussed 
earlier. Over half of the diesel PM comes from nonroad diesels.
    Diesel PM concentrations were also recently modeled across a 
representative urban area, Houston, for 1996, using the Industrial 
Source Complex Short Term (ISCST3) model. This modeling is designed to 
more specifically account for local traffic patterns including diesel 
truck traffic along specific roadways. The modeling in Houston suggests 
strong spatial gradients for Diesel PM and indicates that ``hotspot'' 
concentrations can be very high, up to 8 ug/m\3\ at receptor versus a 3 
ug/m\3\ average in Houston. Such concentrations are above the RfC for 
diesel exhaust and indicate a potential for adverse health effects from 
chronic exposure to diesel PM. These results also suggest that PM from 
diesel vehicles makes a major contribution to total ambient PM 
concentrations. Such ``hot spot'' concentrations along certain roadways 
suggest the presence of both high localized exposures plus higher 
estimated average annual exposure levels for urban centers than what 
has been estimated in assessments such as NATA, which are designed to 
focus on regional and national scale averages. There are similar ``hot 
spot'' concentrations in the immediate vicinity of use of nonroad 
equipment such as in urban construction sites.
(2) Elemental Carbon Measurements
    As mentioned before, the carbonaceous component is significant in 
ambient PM. The carbonaceous component consists of organic carbon and 
elemental carbon. Monitoring data on elemental carbon concentrations 
can be used as a surrogate to determine ambient diesel PM 
concentrations. Elemental carbon is a major component of diesel 
exhaust, contributing to approximately 60 to 80 percent of diesel 
particulate mass, depending on engine technology, fuel type, duty 
cycle, lube oil consumption, and state of engine maintenance. In most 
areas, diesel engine emissions are major contributors to elemental 
carbon in the ambient air, with other potential sources including 
gasoline exhaust, combustion of coal, oil, or wood (including forest 
fires), charbroiling, cigarette smoke, and road dust. Because of the 
large portion of elemental carbon in diesel particulate matter, and the 
fact that diesel exhaust is one of the major contributors to elemental 
carbon in most areas, ambient diesel PM concentrations can be bounded 
using elemental carbon measurements.
    The measured mass of elemental carbon at a given site varies 
depending on the measurement technique used. Moreover, to estimate 
diesel PM concentration based on elemental

[[Page 28344]]

carbon level, one must first estimate the percentage of PM attributable 
to diesel engines and the percentage of elemental carbon in diesel PM. 
Thus, there are significant uncertainties in estimating diesel PM 
concentrations using an elemental carbon surrogate. Depending on the 
measurement technique used, and assumptions made, average nationwide 
concentrations for current years of diesel PM estimated from elemental 
carbon data range from about 1.2 to 2.2 ug/m\3\. EPA has compared these 
estimates based on elemental carbon measurements to modeled 
concentrations in NATA and concluded that the two sets of data agree 
reasonably well. This performance compares favorably with the model to 
monitor results for other pollutants assessed in NATA, with the 
exception of benzene, for which the performance of the NATA modeling 
was better. These comparisons are discussed in greater detail in the 
draft RIA.
(3) Chemical Mass Balance
    The third approach for estimating ambient diesel PM concentrations 
uses the CMB model for source apportionment in conjunction with ambient 
PM measurements and chemical source ``fingerprints'' to estimate 
ambient diesel PM concentrations. The CMB model uses a statistical 
fitting technique to determine how much mass from each source would be 
required to reproduce the chemical fingerprint of each speciated 
ambient monitor. This source apportionment technique presently does not 
distinguish between on-road and nonroad but, instead, gives diesel PM 
as a whole. This source apportionment technique can distinguish between 
diesel and gasoline PM. Caution in interpreting CMB results is 
warranted, as the use of fitting species that are not specific to the 
sources modeled can lead to misestimation of source contributions. 
Ambient concentrations using this approach are generally about 1 ug/
m\3\ annual average. UNMIX/PMF models show similar results. Results 
from various studies are discussed in the draft RIA.
iv. Diesel Exhaust PM Exposures
    Exposure of people to diesel exhaust depends on their various 
activities, the time spent in those activities, the locations where 
these activities occur, and the levels of diesel exhaust pollutants 
(such as particulate) in those locations. The major difference between 
ambient levels of diesel particulate and exposure levels for diesel 
particulate is that exposure accounts for a person moving from location 
to location, proximity to the emission source, and whether the exposure 
occurs in an enclosed environment.
(1) Occupational Exposures
    Diesel particulate exposures have been measured for a number of 
occupational groups over various years but generally for more recent 
years (1980s and later) rather than earlier years. Occupational 
exposures had a wide range varying from 2 to 1,280 ug/m3 for 
a variety of occupational groups including miners, railroad workers, 
firefighters, air port crew, public transit workers, truck mechanics, 
utility linemen, utility winch truck operators, fork lift operators, 
construction workers, truck dock workers, short-haul truck drivers, and 
long-haul truck drivers. These individual studies are discussed in the 
Diesel HAD. As discussed in the Diesel HAD, the National Institute of 
Occupational Safety and Health (NIOSH) has estimated a total of 
1,400,000 workers are occupationally exposed to diesel exhaust from on-
road and nonroad equipment.
    Many measured or estimated occupational exposures are for on-road 
diesel engines although some (especially the higher ones) are for 
occupational groups (e.g., fork lift operators, construction workers, 
or mine workers) who would be exposed to nonroad diesel exhaust. 
Sometimes, as is the case for the nonroad engines, there are only 
estimates of exposure based on the length of employment or similar 
factors rather than a ug/m3 level. Estimates for exposures 
to diesel PM for diesel fork lift operators have been made that range 
from 7 to 403 ug/m3 as reported in the Diesel HAD. In 
addition, the Northeast States for Coordinated Air Use Management 
(NESCAUM) is presently measuring occupational exposures to particulate 
and elemental carbon near the operation of various diesel non-road 
equipment. Exposure groups include agricultural farm operators, grounds 
maintenance personnel (lawn and garden equipment), heavy equipment 
operators conducting multiple job tasks at a construction site, and a 
saw mill crew at a lumber yard. Samples will be obtained in the 
breathing zone of workers. Some initial results are expected in late 
2003.
(2) General Ambient Exposures
    Currently, personal exposure monitors for PM cannot differentiate 
diesel from other PM. Thus, we use modeling to estimate exposures. 
Specifically, exposures for the general population are estimated by 
first conducting dispersion modeling of both on-road and non-road 
diesel emissions, described above, and then by conducting exposure 
modeling. The most comprehensive modeling for cumulative exposures to 
diesel PM is the NATA. This assessment calculates exposures of the 
national population as a whole to a variety of air toxics, including 
diesel PM. As discussed previously, the ambient levels are calculated 
using the ASPEN dispersion model. The preponderance of modeled diesel 
PM concentrations are within a factor of 2 of diesel PM concentrations 
estimated from elemental carbon measurements.\38\ This comparison adds 
credence to the modeled ASPEN results and associated exposure 
assessment.
---------------------------------------------------------------------------

    \38\ U.S. EPA (2002). Diesel PM model-to-measurement comparison. 
Prepared by ICF Consulting for EPA, Office of Transportation and Air 
Quality. Report No. EPA420-D-02-004.
---------------------------------------------------------------------------

    The modeled ambient concentrations are used as inputs into the 
Hazardous Air Pollution Exposure Model (HAPEM4) to calculate exposure 
levels. Average exposures calculated nationwide are 1.44 ug/
m3 with levels of 1.64 ug/m3 for urban counties 
and 0.55 ug/m3 for rural counties. Again, nonroad diesels 
account for over half of this modeled exposure.
(3) Ambient Exposures--Microenvironments
    One common microenvironment for diesel exposure is beside freeways. 
Although freeway locations are associated mostly with on-road rather 
than nonroad diesels, there are many similarities between on-road and 
nonroad diesel emissions as discussed in the Diesel HAD. The California 
Air Resources Board (CARB) measured elemental carbon near the Long 
Beach Freeway in 1993. Levels measured ranged from 0.4 to 4.0 ug/
m3 (with one value as high as 7.5 ug/m3) above 
background levels. Microenvironments associated with nonroad engines 
would include construction zones. PM and elemental carbon samples are 
being collected by NESCAUM in the immediate area of the nonroad engine 
operations (such as at the edge or fence line of the construction 
zone). Besides PM and elemental carbon levels, various toxics such as 
benzene, 1,3-butadiene, formaldehyde, and acetaldehyde will be sampled. 
Some initial results should be available in late 2003 and will be 
especially useful since they focus on those microenvironments affected 
by nonroad diesels.
    Also, EPA is funding research in Fresno to measure indoor and 
outdoor PM component concentrations in the homes of over 100 asthmatic 
children. Some of these homes are located near

[[Page 28345]]

agricultural, construction, and utility nonroad equipment operations. 
This work will measure infiltration of elemental carbon and other PM 
components to indoor environments. The project also evaluates lung 
function changes in the asthmatic children during fluctuations in 
exposure concentrations and compositions. This information may allow an 
evaluation of adverse health effects associated with exposures to 
elemental carbon and other PM components from on-road and nonroad 
sources. Some initial results may be available in late 2003.
b. Gaseous Air Toxics
    Nonroad diesel engine emissions contain several substances known or 
suspected as human or animal carcinogens, or that have noncancer health 
effects. These other compounds include benzene,1,3-butadiene, 
formaldehyde, acetaldehyde, acrolein, dioxin, and polycyclic organic 
matter (POM). For some of these pollutants, nonroad diesel engine 
emissions are believed to account for a significant proportion of total 
nation-wide emissions. All of these compounds were identified as 
national or regional ``risk'' drivers in the 1996 NATA. That is, these 
compounds pose a significant portion of the total inhalation cancer 
risk to a significant portion of the population. Mobile sources 
contribute significantly to total emissions of these air toxics. As 
discussed later in this section, this proposed rulemaking will result 
in significant reductions of these emissions.
    Benzene: Nonroad diesel engines accounted for about 3 percent of 
ambient benzene emissions in 1996. Of ambient benzene levels due to 
mobile sources, 5 percent in urban and 3 percent in rural areas came 
from nonroad diesel.
    The EPA's IRIS database lists benzene as a known human carcinogen 
(causing leukemia at high, prolonged air exposures) by all routes of 
exposure, and exposure is associated with additional health effects 
including genetic changes in humans and animals and increased 
proliferation of bone marrow cells in mice.39 40 41 42 EPA 
states in its IRIS database that the data indicate a causal 
relationship between benzene exposure and acute lymphocytic leukemia 
and suggest a relationship between benzene exposure and chronic non-
lymphocytic leukemia and chronic lymphocytic leukemia. Respiration is 
the major source of human exposure and at least half of this exposure 
is attributable to gasoline vapors and automotive emissions. A number 
of adverse noncancer health effects including blood disorders, such as 
preleukemia and aplastic anemia, have also been associated with low-
dose, long-term exposure to benzene.43 44
---------------------------------------------------------------------------

    \39\ U.S. EPA (2000). Integrated Risk Information System File 
for Benzene. This material is available electronically at http://
www.epa.gov/iris/subst/0276.htm.
    \40\ International Agency for Research on Cancer, IARC 
monographs on the evaluation of carcinogenic risk of chemicals to 
humans, Volume 29, Some industrial chemicals and dyestuffs, 
International Agency for Research on Cancer, World Health 
Organization, Lyon, France, p. 345-389, 1982.
    \41\ Irons, R.D., W.S. Stillman, D.B. Colagiovanni, and V.A. 
Henry, Synergistic action of the benzene metabolite hydroquinone on 
myelopoietic stimulating activity of granulocyte/macrophage colony-
stimulating factor in vitro, Proc. Natl. Acad. Sci. 89:3691-3695, 
1992.
    \42\ U.S. EPA (1998). Carcinogenic Effects of Benzene: An 
Update, National Center for Environmental Assessment, Washington, 
DC. 1998.
    \43\ Aksoy, M. (1989). Hematotoxicity and carcinogenicity of 
benzene. Environ. Health Perspect. 82: 193-197.
    \44\ Goldstein, B.D. (1988). Benzene toxicity. Occupational 
medicine. State of the Art Reviews. 3: 541-554.
---------------------------------------------------------------------------

    1,3-Butadiene: Nonroad diesel engines accounted for about 1.5 
percent of ambient butadiene emissions in 1996. Of ambient butadiene 
levels due to mobile sources, 4 percent in urban and 2 percent in rural 
areas came from nonroad diesel.
    EPA earlier identified 1,3-butadiene as a probable human carcinogen 
in its IRIS database and recently redesignated it as a known human 
carcinogen (but with a lower carcinogenic potency than previously 
used).\45\ The specific mechanisms of 1,3-butadiene-induced 
carcinogenesis are unknown, however, it is virtually certain that the 
carcinogenic effects are mediated by genotoxic metabolites of 1,3-
butadiene. Animal data suggest that females may be more sensitive than 
males for cancer effects; nevertheless, there are insufficient data 
from which to draw any conclusions on potentially sensitive 
subpopulations. 1,3-Butadiene also causes a variety of reproductive and 
developmental effects in mice; no human data on these effects are 
available. The most sensitive effect was ovarian atrophy observed in a 
lifetime bioassay of female mice.\46\
---------------------------------------------------------------------------

    \45\ U.S. EPA (2002). Health Assessment of 1,3-Butadiene. Office 
of Research and Development, National Center for Environmental 
Assessment, Washington Office, Washington, DC. Report No. EPA/600/P-
98/001F.
    \46\ Bevan, C; Stadler, JC; Elliot, GS; et al. (1996) Subchronic 
toxicity of 4-vinylcyclohexene in rats and mice by inhalation. 
Fundam. Appl. Toxicol. 32:1-10.
---------------------------------------------------------------------------

    Formaldehyde: Nonroad diesel engines accounted for about 22 percent 
of ambient formaldehyde emissions in 1996. Of ambient formaldehyde 
levels due to mobile sources, 37 percent in urban and 27 percent in 
rural areas came form nonroad diesel. These figures are for tailpipe 
emissions of formaldehyde. Formaldehyde in the ambient air comes not 
only from tailpipe (of direct) emissions but is also formed from 
photochemical reactions of hydrocarbons.
    EPA has classified formaldehyde as a probable human carcinogen 
based on evidence in humans and in rats, mice, hamsters, and 
monkeys.\47\ Epidemiological studies in occupationally exposed workers 
suggest that long-term inhalation of formaldehyde may be associated 
with tumors of the nasopharyngeal cavity (generally the area at the 
back of the mouth near the nose), nasal cavity, and sinus.\48\ 
Formaldehyde exposure also causes a range of noncancer health effects, 
including irritation of the eyes (tearing of the eyes and increased 
blinking) and mucous membranes. Sensitive individuals may experience 
these adverse effects at lower concentrations than the general 
population and in persons with bronchial asthma, the upper respiratory 
irritation caused by formaldehyde can precipitate an acute asthmatic 
attack. The agency is currently conducting a reassessment of risk from 
inhalation exposure to formaldehyde.
---------------------------------------------------------------------------

    \47\ U.S. EPA (1987). Assessment of Health Risks to Garment 
Workers and Certain Home Residents from Exposure to Formaldehyde, 
Office of Pesticides and Toxic Substances, April 1987.
    \48\ Blair, A., P.A. Stewart, R.N. Hoover, et al. (1986). 
Mortality among industrial workers exposed to formaldehyde. J. Natl. 
Cancer Inst. 76(6): 1071-1084.
---------------------------------------------------------------------------

    Acetaldehyde: Nonroad diesel engines accounted for about 34 percent 
of acetaldehyde emissions in 1996. Of ambient acetaldehyde levels due 
to mobile sources, 24 percent in urban and 17 percent in rural areas 
came form nonroad diesel. Also, acetaldehyde can be formed 
photochemically in the atmosphere. Counting both direct emissions and 
photochemically formed acetaldehyde, mobile sources were responsible 
for the major portion of acetaldehyde in the ambient air according to 
the National-Scale Air Toxics Assessment for 1996.
    Acetaldehyde is classified in EPA's IRIS database as a probable 
human carcinogen and is considered moderately toxic by the inhalation, 
oral, and intravenous routes.\49\ The primary acute effect of exposure 
to acetaldehyde vapors is irritation of the eyes, skin, and

[[Page 28346]]

respiratory tract. At high concentrations, irritation and pulmonary 
effects can occur, which could facilitate the uptake of other 
contaminants. Some asthmatics have been shown to be a sensitive 
subpopulation to decrements in FEV1 upon acetaldehyde inhalation.\50\ 
The agency is currently conducting a reassessment of risk from 
inhalation exposure to acetaldehyde.
---------------------------------------------------------------------------

    \49\ U.S. EPA (1988). Integrated Risk Information System File of 
Acetaldehyde. This material is available electronically at http://
www.epa.gov/iris/subst/0290.htm.
    \50\ Myou, S.; Fujimura, M.; Nishi K.; Ohka, T.; and Matsuda, T. 
(1993) Aerosolized acetaldehyde induces histamine-mediated 
bronchoconstriction in asthmatics. Am Rev Respir Dis 148(4 Pt 1): 
940-3.
---------------------------------------------------------------------------

    Acrolein: Nonroad diesel engines accounted for about 17.5 percent 
of acrolein emissions in 1996. Of ambient acrolein levels due to mobile 
sources, 28 percent in urban and 18 percent in rural areas came form 
nonroad diesel.
    Acrolein is extremely toxic to humans when inhaled, with acute 
exposure resulting in upper respiratory tract irritation and 
congestion. The Agency has developed a reference concentration for 
inhalation (RfC) of acrolein of 0.02 micrograms/m\3\.\51\
---------------------------------------------------------------------------

    \51\ U.S. EPA (1993). Environmental Protection Agency, 
Integrated Risk Information System (IRIS), National Center for 
Environmental Assessment, Cincinnati, OH.
---------------------------------------------------------------------------

    Although no information is available on its carcinogenic effects in 
humans, based on laboratory animal data, EPA considers acrolein a 
possible human carcinogen.
    Polycyclic Organic Matter (POM): POM is generally defined as a 
large class of chemicals consisting of organic compounds having 
multiple benzene rings and a boiling point greater than 100 degrees C. 
Polycyclic aromatic hydrocarbons (PAHs) are a chemical class that is a 
subset of POM. POM are naturally occurring substances that are 
byproducts of the incomplete combustion of fossil fuels and plant and 
animal biomass (e.g., forest fires). They occur as byproducts from 
steel and coke productions and waste incineration. They also are a 
component of diesel particulate emissions. Many of the compounds 
included in the class of compounds known as POM are classified by EPA 
as probable human carcinogens based on animal data. In particular, EPA 
frequently obtains data on 7 of the POM compounds, which we analyzed 
separately as a class in the 1996 NATA. Nonroad diesel engines account 
for less than 1 percent of these 7 POM compounds with total mobile 
sources responsible for only 4 percent of the total; most of the 7 POMs 
come from area sources. For total POM compounds, mobile sources as a 
whole are responsible for only 1 percent. The mobile source emission 
numbers used to derive these inventories are based on only particulate 
phase POM and do not include the semi-volatile phase POM levels. Were 
those additional POMs included (which is now being done), these 
inventory numbers would be substantially higher.
    Even though mobile sources are responsible for only a small portion 
of total POM emissions, the particulate reductions from today's action 
will reduce these emissions.
    Dioxins: Recent studies have confirmed that dioxins are formed by 
and emitted from diesels (both heavy-duty diesel trucks and non-road 
diesels although in very small amounts) and are estimated to account 
for about 1 percent of total dioxin emissions in 1995. Recently EPA 
issued a draft assessment designating one dioxin compound, 2,3,7,8-
tetrachlorodibenzo-p-dioxin as a human carcinogen and the complex 
mixtures of dioxin-like compounds as likely to be carcinogenic to 
humans using the draft 1996 carcinogen risk assessment guidelines. EPA 
is working on its final assessment for dioxin.\52\ An interagency 
review group is evaluating EPA's designation of dioxin as a likely 
human carcinogen. Reductions from today's nonroad proposal will have 
minimal impact on overall dioxin emissions.
---------------------------------------------------------------------------

    \52\ U.S. EPA (June 2000) Exposure and Human Health Reassessment 
of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds, 
External Review Draft, EPA/600/P-00/001Ag. This material is 
available electronically at http://www.epa.gov/ncea/dioxin.htm.
---------------------------------------------------------------------------

3. Ozone
a. What Are the Health Effects of Ozone Pollution?
Ground-level ozone pollution (sometimes called ``smog'') is formed by 
the reaction of volatile organic compounds (VOC) and nitrogen oxides 
(NOX) in the atmosphere in the presence of heat and 
sunlight. These two pollutants, often referred to as ozone precursors, 
are emitted by many types of pollution sources, including on-road and 
off-road motor vehicles and engines, power plants and industrial 
facilities, and smaller ``area'' sources.
    Ozone can irritate the respiratory system, causing coughing, throat 
irritation, and/or uncomfortable sensation in the 
chest.53 54 Ozone can reduce lung function and make it more 
difficult to breathe deeply, and breathing may become more rapid and 
shallow than normal, thereby limiting a person's normal activity. Ozone 
also can aggravate asthma, leading to more asthma attacks that require 
a doctor's attention and/or the use of additional medication. In 
addition, ozone can inflame and damage the lining of the lungs, which 
may lead to permanent changes in lung tissue, irreversible reductions 
in lung function, and a lower quality of life if the inflammation 
occurs repeatedly over a long time period (months, years, a lifetime). 
People who are of particular concern with respect to ozone exposures 
include children and adults who are active outdoors. Those people 
particularly susceptible to ozone effects are people with respiratory 
disease, such as asthma, and people with unusual sensitivity to ozone, 
and children. Beyond its human health effects, ozone has been shown to 
injure plants, which has the effect of reducing crop yields and 
reducing productivity in forest ecosystems.55 56
---------------------------------------------------------------------------

    \53\ U.S. EPA (1996). Air Quality Criteria for Ozone and Related 
Photochemical Oxidants, EPA/600/P-93/004aF. Docket No. A-99-06. 
Document Nos. II-A-15 to 17.
    \54\ U.S. EPA. (1996). Review of National Ambient Air Quality 
Standards for Ozone, Assessment of Scientific and Technical 
Information, OAQPS Staff Paper, EPA-452/R-96-007. Docket No. A-99-
06. Document No. II-A-22.
    \55\ U.S. EPA (1996). Air Quality Criteria for Ozone and Related 
Photochemical Oxidants, EPA/600/P-93/004aF. Docket No. A-99-06. 
Document Nos. II-A-15 to 17.
    \56\ U.S. EPA. (1996). Review of National Ambient Air Quality 
Standards for Ozone, Assessment of Scientific and Technical 
Information, OAQPS Staff Paper, EPA-452/R-96-007. Docket No. A-99-
06. Document No. II-A-22.
---------------------------------------------------------------------------

    The 8-hour ozone standard, established by EPA in 1997, is based on 
well-documented science demonstrating that more people are experiencing 
adverse health effects at lower levels of exertion, over longer 
periods, and at lower ozone concentrations than addressed by the one-
hour ozone standard. (See, e.g., 62 FR 38861-62, July 18, 1997). The 8-
hour standard addresses ozone exposures of concern for the general 
population and populations most at risk, including children active 
outdoors, outdoor workers, and individuals with pre-existing 
respiratory disease, such as asthma.
    There has been new research that suggests additional serious health 
effects beyond those that had been known when the 8-hour ozone health 
standard was set. Since 1997, over 1,700 new health and welfare studies 
relating to ozone have been published in peer-reviewed journals.\57\ 
Many of these studies have investigated the impact of ozone exposure on 
such health effects as changes in lung structure and biochemistry, 
inflammation of the

[[Page 28347]]

lungs, exacerbation and causation of asthma, respiratory illness-
related school absence, hospital and emergency room visits for asthma 
and other respiratory causes, and premature mortality. EPA is currently 
in the process of evaluating these and other studies as part of the 
ongoing review of the air quality criteria and NAAQS for ozone. A 
revised Air Quality Criteria Document for Ozone and Other Photochemical 
Oxidants will be prepared in consultation with EPA's Clean Air Science 
Advisory Committee (CASAC). Key new health information falls into four 
general areas: development of new-onset asthma, hospital admissions for 
young children, school absence rate, and premature mortality.
---------------------------------------------------------------------------

    \57\ New Ozone Health and Environmental Effects References, 
Published Since Completion of the Previous Ozone AQCD, National 
Center for Environmental Assessment, Office of Research and 
Development, U.S. Environmental Protection Agency, Research Triangle 
Park, NC 27711 (7/2002) Docket No. A-2001-11. Document No. IV-A-19.
---------------------------------------------------------------------------

    Aggravation of existing asthma resulting from short-term ambient 
ozone exposure was reported prior to the 1997 decision and has been 
observed in studies published subsequently.58 59 In 
particular, a relationship between long-term ambient ozone 
concentrations and the incidence of new-onset asthma in adult males 
(but not in females) was reported by McDonnell et al. (1999).\60\ 
Subsequently, an additional study suggests that incidence of new 
diagnoses of asthma in children is associated with heavy exercise in 
communities with high concentrations (i.e., mean 8-hour concentration 
of 59.6 ppb) of ozone.\61\ This relationship was documented in children 
who played 3 or more sports and thus had higher exposures and was not 
documented in those children who played one or two sports. The larger 
effect of high activity sports than low activity sports and an 
independent effect of time spent outdoors also in the higher ozone 
communities strengthened the inference that exposure to ozone may 
modify the effect of sports on the development of asthma in some 
children.
---------------------------------------------------------------------------

    \58\ Thurston, G.D., M.L. Lippman, M.B. Scott, and J.M. Fine. 
1997. Summertime Haze Air Pollution and Children with Asthma. 
American Journal of Respiratory Critical Care Medicine, 155: 654-
660.
    \59\ Ostro, B, M. Lipsett, J. Mann, H. Braxton-Owens, and M. 
White (2001) Air pollution and exacerbation of asthma in African-
American children in Los Angeles. Epidemiology 12(2): 200-208.
    \60\ McDonnell, W.F., D.E. Abbey, N. Nishino and M.D. Lebowitz. 
1999. ``Long-term ambient ozone concentration and the incidence of 
asthma in nonsmoking adults: the ahsmog study.'' Environmental 
Research. 80(2 Pt 1): 110-121.
    \61\ McConnell, R.; Berhane, K.; Gilliland, F.; London, S.J.; 
Islam, T.; Gauderman, W.J.; Avol, E.; Margolis, H.G.; Peters, J.M. 
(2002) Asthma in exercising children exposed to ozone: a cohort 
study. Lancet 359: 386-391.
---------------------------------------------------------------------------

    Previous studies have shown relationships between ozone and 
hospital admissions in the general population. A study in Toronto 
reported a significant relationship between 1-hour maximum ozone 
concentrations and respiratory hospital admissions in children under 
the age of two.\62\ Given the relative vulnerability of children in 
this age category, we are particularly concerned about the findings.
---------------------------------------------------------------------------

    \62\ Burnett, R.T.; Smith--Doiron, M.; Stieb, D.; Raizenne, 
M.E.; Brook, J.R.; Dales, R.E.; Leech, J.A.; Cakmak, S.; Krewski, D. 
(2001) Association between ozone and hospitalization for acute 
respiratory diseases in children less than 2 years of age. Am. J. 
Epidemiol. 153: 444-452.
---------------------------------------------------------------------------

    Increased respiratory disease that are serious enough to cause 
school absences have been associated with 1-hour daily maximum and 8-
hour average ozone concentrations in studies conducted in Nevada \63\ 
in kindergarten to 6th grade and in Southern California in grades 4 
through 6.\64\ These studies suggest that higher ambient ozone levels 
may result in increased school absenteeism.
---------------------------------------------------------------------------

    \63\ Chen, L.; Jennison, B.L.; Yang, W.; Omaye, S.T. (2000) 
Elementary school absenteeism and air pollution. Inhalation Toxicol. 
12: 997-1016.
    \64\ Gilliland, FD, K Berhane, EB Rappaport, DC Thomas, E Avol, 
WJ Gauderman, SJ London, HG Margolis, R McConnell, KT Islam, JM 
Peters (2001) The effects of ambient air pollution on school 
absenteeism due to respiratory illnesses Epidemiology 12:43-54.
---------------------------------------------------------------------------

    The air pollutant most clearly associated with premature mortality 
is PM, with dozens of studies reporting such an association. However, 
repeated ozone exposure is a possible contributing factor for premature 
mortality, causing an inflammatory response in the lungs which may 
predispose elderly and other sensitive individuals to become more 
susceptible to other stressors, such as PM.65 66 67 Although 
the findings have been mixed, the findings of three recent analyses 
suggest that ozone exposure is associated with increased mortality. 
Although the National Morbidity, Mortality, and Air Pollution Study 
(NMMAPS) did not report an effect of ozone on total mortality across 
the full year, the investigators who conducted the NMMAPS study did 
observe an effect after limiting the analysis to summer when ozone 
levels are highest.68 69 Similarly, other studies have shown 
associations between ozone and mortality.70 71 Specifically, 
Toulomi et al. (1997) found that 1-hour maximum ozone levels were 
associated with daily numbers of deaths in 4 cities (London, Athens, 
Barcelona, and Paris), and a quantitatively similar effect was found in 
a group of four additional cities (Amsterdam, Basel, Geneva, and 
Zurich).
---------------------------------------------------------------------------

    \65\ Samet JM, Zeger SL, Dominici F, Curriero F, Coursac I, 
Dockery DW, Schwartz J, Zanobetti A. 2000. The National Morbidity, 
Mortality and Air Pollution Study: Part II: Morbidity, Mortality and 
Air Pollution in the United States. Research Report No. 94, Part II. 
Health Effects Institute, Cambridge MA, June 2000. (Docket Number A-
2000-01, Document Nos. IV-A-208 and 209).
    \66\ Devlin, R.B.; Folinsbee, L.J.; Biscardi, F.; Hatch, G.; 
Becker, S.; Madden, M.C.; Robbins, M.; Koren, H. S. (1997) 
Inflammation and cell damage induced by repeated exposure of humans 
to ozone. Inhalation Toxicol. 9: 211-235.
    \67\ Koren HS, Devlin RB, Graham DE, Mann R, McGee MP, Horstman 
DH, Kozumbo WJ, Becker S, House DE, McDonnell SF, Bromberg, PA. 
1989. Ozone-induced inflammation in the lower airways of human 
subjects. Am. Rev. Respir. Dies. 139: 407-415.
    \68\ Samet JM, Zeger SL, Dominici F, Curriero F, Coursac I, 
Dockery DW, Schwartz J, Zanobetti A. 2000. The National Morbidity, 
Mortality and Air Pollution Study: Part II: Morbidity, Mortality and 
Air Pollution in the United States. Research Report No. 94, Part II. 
Health Effects Institute, Cambridge MA, June 2000. (Docket Number A-
2000-01, Documents No. IV-A-208 and 209).
    \69\ Samet JM, Zeger SL, Dominici F, Curriero F, Coursac I, 
Zeger, S. Fine Particulate Air Pollution and Mortality in 20 U.S. 
Cities, 1987--1994. The New England Journal of Medicine. Vol. 343, 
No. 24, December 14, 2000. P. 1742-1749.
    \70\ Thurston, G.D.; Ito, K. (2001) Epidemiological studies of 
acute ozone exposures and mortality. J. Exposure Anal. Environ. 
Epidemiol. 11: 286-294.
    \71\ Touloumi, G.; Katsouyanni, K.; Zmirou, D.; Schwartz, J.; 
Spix, C.; Ponce de Leon, A.; Tobias, A.; Quennel, P.; Rabczenko, D.; 
Bacharova, L.; Bisanti, L.; Vonk, J.M.; Ponka, A. (1997) Short-term 
effects of ambient oxidant exposure on mortality: a combined 
analysis within the APHEA project. Am. J. Epidemiol. 146: 177-185.
---------------------------------------------------------------------------

    In all, the new studies that have become available since the 8-hour 
ozone standard was adopted in 1997 continue to demonstrate the harmful 
effects of ozone on public health, and the need to attain and maintain 
the NAAQS.
b. Current and projected 8-hour ozone levels
    As shown earlier (Figure II-1), unhealthy ozone concentrations 
exceeding the level of the 8-hour standard (i.e., not requisite to 
protect the public health with an adequate margin of safety) occur over 
wide geographic areas, including most of the nation's major population 
centers. These monitored areas include much of the eastern half of the 
U.S. and large areas of California.
    Based upon data from 1999-2001, there are 291 counties where 111 
million people live that are measuring values that violate the 8-hour 
ozone NAAQS.\72\ An additional 37 million people live in 155 counties 
that have air quality measurements within 10 percent of the level of 
the standard. These areas, though currently not violating the standard, 
will also benefit from the additional emission reductions from this 
rule.
---------------------------------------------------------------------------

    \72\ Additional counties may have levels above the NAAQS but do 
not currently have monitors.
---------------------------------------------------------------------------

    From our air quality modeling for this proposal, we anticipate that 
without emission reductions beyond those

[[Page 28348]]

already required under promulgated regulation and approved SIPs, ozone 
nonattainment will likely persist into the future. With reductions from 
programs already in place, the number of counties violating the ozone 
8-hour standard is expected to decrease in 2020 to 30 counties where 43 
million people are projected to live. Thereafter, exposure to unhealthy 
levels of ozone is expected to begin to increase again. In 2030 the 
number of counties violating the ozone 8-hour NAAQS is projected to 
increase to 32 counties where 47 million people are projected to live. 
In addition, in 2030, 82 counties where 44 million people are projected 
to live will be within 10 percent of violating the ozone 8-hour NAAQS.
    EPA is still developing the implementation process for bringing the 
nation's air into attainment with the ozone 8-hour NAAQS. EPA's current 
plans call for designating ozone 8-hour nonattainment areas in April 
2004. EPA is planning to propose that States submit SIPs that address 
how areas will attain the 8-hour ozone standard within three years 
after nonattainment designation regardless of their classification. EPA 
is also planning to propose that certain SIP components, such as those 
related to reasonably available control technology (RACT) and 
reasonable further progress (RFP) be submitted within 2 years after 
designation. We therefore anticipate that States will submit their 
attainment demonstration SIPs by April 2007. Section 172(a)(2) of the 
Clean Air Act requires that SIP revisions for areas that may be covered 
only under subpart 1 of part D, title I of the Act demonstrate that the 
nonattainment areas will attain the ozone 8-hour standard as 
expeditiously as practicable but no later than five years from the date 
that the area was designated nonattainment. However, based on the 
severity of the air quality problem and the availability and 
feasibility of control measures, the Administrator may extend the 
attainment date ``for a period of no greater than 10 years from the 
date of designation as nonattainment.'' Based on these provisions, we 
expect that most or all areas covered under subpart 1 will attain the 
ozone standard in the 2007 to 2014 time frame. For areas covered under 
subpart 2, the maximum attainment dates provided under the Act range 
from 3 to 20 years after designation, depending on an area's 
classification. Thus, we anticipate that areas covered by subpart 2 
will attain in the 2007 to 2014 time period.
    Since the emission reductions expected from this proposal would 
begin during the same time period, the projected reductions in nonroad 
emissions would be extremely important to States in their effort to 
meet the new NAAQS. It is our expectation that States will be relying 
on such nonroad reductions in order to help them attain and maintain 
the 8-hour NAAQS. Furthermore, since the nonroad emission reductions 
will continue to grow in the years beyond 2014, they will also be 
important for maintenance of the NAAQS for areas with attainment dates 
of 2014 and earlier.
    Using air quality modeling of the impacts of emission reductions, 
we have made estimates of the change in future ozone levels that would 
result from the proposed rule.\73\ That modeling shows that this rule 
would produce nationwide air quality improvements in ozone levels. On a 
population-weighted basis, the average change in future year design 
values would be a decrease of 1.6 ppb in 2020, and 2.6 ppb in 2030. 
Within areas predicted to violate the NAAQS in the projected base case, 
the average decrease would be somewhat higher: 1.9 ppb in 2020 and 3.0 
ppb in 2030.\74\
---------------------------------------------------------------------------

    \73\ These results are ozone changes projected for the 
preliminary control option used for our modeling, as discussed in 
the Draft RIA in section 3.6. The proposal differs from the modeled 
control case based on updated information; however, we believe that 
the net results would approximate future emissions, although we 
anticipate the ozone changes might be slightly different.
    \74\ This is in spite of the fact that NOX reductions 
can at certain times in some areas cause ozone levels to increase. 
Such ``disbenefits'' are predicted in our modeling, but these 
results make clear that the overall effect of the proposed rule is 
positive. See the draft RIA for more information.
---------------------------------------------------------------------------

    The model predictions of whether specific counties will violate the 
NAAQS or not is uncertain, especially for counties with design values 
falling very close to the standard. This makes us more confident in our 
prediction of average air quality changes than in our prediction of the 
exact numbers of counties projected as exceeding the NAAQS. 
Furthermore, actions by States to meet their SIP obligations will 
change the number of counties violating the NAAQS in the time frame we 
are modeling for this rule. If State actions resulted in an increase in 
the number of areas that are very close to, but still above, the NAAQS, 
then this rule might bring many of those counties down sufficiently to 
eliminate remaining violations. In addition, if State actions brought 
several counties we project to be very close to the standard in the 
future down sufficiently to eliminate violations, then the air quality 
improvements from this proposal might serve more to assist these areas 
in maintaining the standards than in changing their status. Bearing 
this in mind, our modeling indicates that, out of 32 counties predicted 
to violate the NAAQS, the proposal would reduce the number of violating 
counties by 2 in 2020 and by 4 in 2030, without consideration of new 
State or Federal programs.

C. Other Environmental Effects

    The following section presents information on five categories of 
public welfare and environmental impacts related to nonroad heavy-duty 
vehicle emissions: visibility impairment, acid deposition, 
eutrophication of water bodies, plant damage from ozone, and water 
pollution resulting from deposition of toxic air pollutants with 
resulting effects on fish and wildlife.
1. Visibility
a. Visibility is Impaired by Fine PM and Precursor Emissions From 
Nonroad Engines Subject to this Proposed Rule
    Visibility can be defined as the degree to which the atmosphere is 
transparent to visible light.\75\ Fine particles with significant 
light-extinction efficiencies include organic matter, sulfates, 
nitrates, elemental carbon (soot), and soil. Size and chemical 
composition of particles strongly affects their ability to scatter or 
absorb light. Sulfates contribute to visibility impairment especially 
on the haziest days across the U.S., accounting in the rural Eastern 
U.S. for more than 60 percent of annual average light extinction on the 
best days and up to 86 percent of average light extinction on the 
haziest days. Nitrates and elemental carbon each typically contribute 1 
to 6 percent of average light extinction on haziest days in rural 
Eastern U.S. locations.\76\
---------------------------------------------------------------------------

    \75\ National Research Council, 1993. Protecting Visibility in 
National Parks and Wilderness Areas. National Academy of Sciences 
Committee on Haze in National Parks and Wilderness Areas. National 
Academy Press, Washington, DC. This document is available on the 
Internet at http://www.nap.edu/books/0309048443/html/. See also U.S. 
EPA Air Quality Criteria Document for Particulate Matter (1996) 
(available on the Internet at http://cfpub.epa.gov/ncea/cfm/
partmatt.cfm) and Review of the National Ambient Air Quality 
Standards for Particulate Matter: Policy Assessment of Scientific 
and Technical Information. These documents can be found in Docket A-
99-06, Documents No. II-A-23 and IV-A-130-32.
    \76\ U.S. EPA Trends Report 2001. This document is available on 
the Internet at http://www.epa.gov/airtrends/.
---------------------------------------------------------------------------

    Visibility is important because it directly affects people's 
enjoyment of daily activities in all parts of the country. Individuals 
value good visibility for the well-being it provides them directly, 
both in where they live and work, and in places where they enjoy 
recreational opportunities.

[[Page 28349]]

Visibility is also highly valued in significant natural areas such as 
national parks and wilderness areas, because of the special emphasis 
given to protecting these lands now and for future generations.
    To quantify changes in visibility, we compute a light-extinction 
coefficient, which shows the total fraction of light that is decreased 
per unit distance. Visibility can be described in terms of visual range 
or light extinction and is reported using an indicator called 
deciview.\77\ In addition to limiting the distance that one can see, 
the scattering and absorption of light caused by air pollution can also 
degrade the color, clarity, and contrast of scenes.
---------------------------------------------------------------------------

    \77\ Visual range can be defined as the maximum distance at 
which one can identify a black object against the horizon sky. It is 
typically described in miles or kilometers. Light extinction is the 
sum of light scattering and absorption by particles and gases in the 
atmosphere. It is typically expressed in terms of inverse megameters 
(Mm-1), with larger values representing worse visibility. 
The deciview metric describes perceived visual changes in a linear 
fashion over its entire range, analogous to the decibel scale for 
sound. A deciview of 0 represents pristine conditions. Under many 
scenic conditions, a change of 1 deciview is considered perceptible 
by the average person.
---------------------------------------------------------------------------

    In addition, visibility impairment can be described by its impact 
over various periods of time, by its source, and the physical 
conditions in various regions of the country. Visibility impairment can 
be said to have a time dimension in that it might relate to short-term 
excursions or to longer periods (e.g., worst 20 percent of days and 
annual average levels). Anthropogenic contributions account for about 
one-third of the average extinction coefficient in the rural West and 
more than 80 percent in the rural East. In the Eastern U.S., reduced 
visibility is mainly attributable to secondarily formed particles, 
particularly those less than a few micrometers in diameter, such as 
sulfates. While secondarily formed particles still account for a 
significant amount in the West, primary emissions contribute a larger 
percentage of the total particulate load than in the East. Because of 
significant differences related to visibility conditions in the Eastern 
and Western U.S., we present information about visibility by region.
    Furthermore, it is important to note that even in those areas with 
relatively low concentrations of anthropogenic fine particles, such as 
the Colorado Plateau, small increases in anthropogenic fine particulate 
concentrations can lead to significant decreases in visual range. This 
is one of the reasons mandatory Federal Class I areas have been given 
special consideration under the Clean Air Act.\78\
---------------------------------------------------------------------------

    \78\ The Clean Air Act designates 156 national parks and 
wilderness areas as mandatory Federal Class I areas for visibility 
protection.
---------------------------------------------------------------------------

b. Visibility Impairment Where People Live, Work and Recreate
    The secondary PM NAAQS is designed to protect against adverse 
welfare effects which includes visibility impairment. In 1997, EPA 
established the secondary PM2.5 NAAQS as equal to the 
primary (health-based) NAAQS of 15 ug/m3 (based on a 3-year average of 
the annual mean) and 65 ug/m3 (based on a 3-year average of 
the 98th percentile of the 24-hour average value) (62 FR 38669, July 
18, 1997). EPA concluded that PM2.5 causes adverse effects 
on visibility in various locations, depending on PM concentrations and 
factors such as chemical composition and average relative humidity. In 
1997, EPA demonstrated that visibility impairment is an important 
effect on public welfare and that unacceptable visibility impairment is 
experienced throughout the U.S., in multi-state regions, urban areas, 
and remote federal Class I areas. In many cities having annual mean 
PM2.5 concentrations exceeding annual standard, improvements 
in annual average visibility resulting from the attainment of the 
annual PM2.5 standard are expected to be perceptible to the 
general population. Based on annual mean monitored PM2.5 
data, many cities in the Northeast, Midwest, and Southeast as well as 
Los Angeles would be expected to experience perceptible improvements in 
visibility if the PM2.5 annual standard were attained.
    The updated monitoring data and air quality modeling, summarized 
above and presented in detail in the draft RIA, confirm that the 
visibility situation identified during the NAAQS review in 1997 is 
still likely to exist, and it will continue to persist when these 
proposed standards for nonroad diesel engines take effect. Thus, the 
determination in the NAAQS rulemaking about broad visibility impairment 
and related benefits from NAAQS compliance are still relevant.
    Furthermore, in setting the PM2.5 NAAQS, EPA 
acknowledged that levels of fine particles below the NAAQS may also 
contribute to unacceptable visibility impairment and regional haze 
problems in some areas, and section 169 of the Act provides additional 
authorities to remedy existing impairment and prevent future impairment 
in the 156 national parks, forests and wilderness areas labeled as 
mandatory Federal Class I areas (62 FR 38680-81, July 18, 1997).
    In making determinations about the level of protection afforded by 
the secondary PM NAAQS, EPA considered how the section 169 regional 
haze program and the secondary NAAQS would function together.\79\ 
Regional strategies are expected to improve visibility in many urban 
and non-Class I areas as well.
---------------------------------------------------------------------------

    \79\ U.S. EPA Review of the National Ambient Air Quality 
Standards for Particulate Matter: Policy Assessment of Scientific 
and Technical Information OAQPS Staff Paper. EPA-452/R-96-013. 1996. 
Docket Number A-99-06, Documents Nos. II-A-18, 19, 20, and 23. The 
particulate matter air quality criteria documents are also available 
at http://www.epa.gov/ncea/partmatt.htm.
---------------------------------------------------------------------------

    Fine particles may remain suspended for days or weeks and travel 
hundreds to thousands of kilometers, and thus fine particles emitted or 
created in one county may contribute to ambient concentrations in a 
neighboring region.\80\
---------------------------------------------------------------------------

    \80\ Review of the National Ambient Air Quality Standards for 
Particulate Matter: Policy Assessment for Scientific and Technical 
Information, OAQPS Staff Paper, EPA-452/R-96-013, July, 1996, at IV-
7. This document is available from Docket A-99-06, Document II-A-23.
---------------------------------------------------------------------------

    The 1999-2001 PM2.5 monitored values indicate that at 
least 74 million people live in areas where long-term ambient fine PM 
levels are at or above 15 [mu]g/m3.\81\ Thus, at least these 
populations (plus those who travel to those areas) are experiencing 
significant visibility impairment, and emissions of PM and its 
precursors from nonroad diesel engines contribute to this 
impairment.\82\
---------------------------------------------------------------------------

    \81\ U.S. EPA Air Quality Data Analysis 1999-2001. Technical 
Support Document for Regulatory Actions. March 2003.
    \82\ These populations would also be exposed to PM 
concentrations associated with the adverse health impacts discussed 
above.
---------------------------------------------------------------------------

    Because of the importance of chemical composition and size to 
visibility, we used EPA's Regional Modeling System for Aerosols and 
Deposition (REMSAD)\83\ model to project visibility conditions in 2020 
and 2030 in terms of deciview, accounting for the chemical composition 
of the particles and transport of precursors. Our projections included 
anticipated emissions from the nonroad diesel engines subject to this 
proposed rule as well as all other sources.
---------------------------------------------------------------------------

    \83\ Additional information about the Regional Modeling System 
for Aerosols and Deposition (REMSAD) and our modeling protocols can 
be found in our Regulatory Impact Analysis: Heavy-Duty Engine and 
Vehicle Standards and Highway Diesel Fuel Sulfur Control 
Requirements, document EPA420-R-00-026, December 2000. Docket No. A-
2000-01, Document No. A-II-13. This document is also available at 
http://www.epa.gov/otaq/disel.htm#documents.
---------------------------------------------------------------------------

    Based on this modeling, we predict that in 2030, 85 million people 
(25

[[Page 28350]]

percent of the future population) would be living in areas with 
visibility degradation where fine PM levels are above 15 [mu]g/m3 
annually.\84\ Thus, at least a quarter of the population would 
experience visibility impairment in areas where they live, work and 
recreate.
---------------------------------------------------------------------------

    \84\ Technical Memorandum, EPA Air Docket A-99-06, Eric O. 
Ginsburg, Senior Program Advisor, Emissions Monitoring and Analysis 
Division, OAQPS, Summary of Absolute Modeled and Model-Adjusted 
Estimates of Fine Particulate Matter for Selected Years, December 6, 
2000, Table P-2. Docket Number 2000-01, Document Number II-B-14.
---------------------------------------------------------------------------

    As shown in Table I.C-1, accounting for the different visibility 
impact of the chemical constituents of the PM2.5, in 2030 we 
expect visibility in the East to be about 20.5 deciviews (or visual 
range of 50 kilometers) on average, with poorer visibility in urban 
areas, compared to the average Eastern visibility conditions without 
man-made pollution of 9.5 deciviews (or visual range of 150 
kilometers). Likewise, we expect visibility in the West to be about 8.8 
deciviews (or visual range of 162 kilometers) on average in 2030, with 
poorer visibility in urban areas, compared to the average Western 
visibility conditions without man-made pollution of 5.3 deciviews (or 
visual range of 230 kilometers). Thus, the emissions from these nonroad 
diesel sources, especially SOx emissions that become sulfates in the 
atmosphere, contribute to future visibility impairment summarized in 
the table.
    Control of nonroad land-based engines emissions, as shown in Table 
I.C-1, will improve visibility across the nation. Taken together with 
other programs, reductions from this proposal will help to improve 
visibility. Control of these emissions in and around areas with PM 
levels above the annual PM2.5 NAAQS will likely improve 
visibility in other locations such as mandatory Federal Class I areas. 
Specifically, for a preliminary control option described in the draft 
RIA chapter 3.6 that is similar to our proposal, we expect on average 
for visibility to improve to about 0.33 deciviews in the East and 0.35 
deciviews in the West. The improvement from our proposal is likely to 
be similar but slightly smaller than what was modeled due to the 
differences in emission reductions between the proposal and the modeled 
scenario.

   Table I.C-1--Summary of Modeled 2030 National Visibility Conditions
                       [Average annual deciviews]
------------------------------------------------------------------------
                                                 Predicted
                                    Predicted       2030      Change in
                                       2030      visibility     annual
           Regions \a\              visibility   with rule     average
                                     baseline     controls    deciviews
                                                    \b\
------------------------------------------------------------------------
Eastern U.S......................        20.54        20.21         0.33
    Urban........................        21.94        21.61         0.33
    Rural........................        19.98        19.65         0.33
Western U.S......................         8.83         8.58         0.25
    Urban........................         9.78         9.43         0.35
    Rural........................         8.61         8.38        0.23
------------------------------------------------------------------------
Notes:
\a\ Eastern and Western Regions are separated by 100 degrees north
  longitude. Background visibility conditions differ by region. Natural
  background is 9.5 deciviews in the East and 5.3 in the West.
\b\ The results illustrate the type of visibility improvements for the
  preliminary control option, as discussed in the Draft RIA. The
  proposal differs based on updated information; however, we believe
  that the net results would approximate future PM emissions, although
  we anticipate the visibility improvements would be slightly smaller.

c. Visibility Impairment in Mandatory Federal Class I Areas
    The Clean Air Act establishes special goals for improving 
visibility in many national parks, wilderness areas, and international 
parks. In the 1990 Clean Air Act amendments, Congress provided 
additional emphasis on regional haze issues (see CAA section 169B). In 
1999, EPA finalized a rule that calls for States to establish goals and 
emission reduction strategies for improving visibility in all 156 
mandatory Federal Class I areas. In that rule, EPA established a 
``natural visibility'' goal, and also encouraged the States to work 
together in developing and implementing their air quality plans. The 
regional haze program is focused on long-term emissions decreases from 
the entire regional emissions inventory comprised of major and minor 
stationary sources, area sources and mobile sources. The regional haze 
program is designed to improve visibility and air quality in our most 
treasured natural areas from these broad sources. At the same time, 
control strategies designed to improve visibility in the national parks 
and wilderness areas are expected to improve visibility over broad 
geographic areas. For mobile sources, there is a need for a Federal 
role in reduction of those emissions, especially because mobile source 
engines are regulated primarily at the Federal level.
    Because of evidence that fine particles are frequently transported 
hundreds of miles, all 50 states, including those that do not have 
mandatory Federal Class I areas, participate in planning, analysis, 
and, in many cases, emission control programs under the regional haze 
regulations. Virtually all of the 156 mandatory Federal Class I areas 
experience impaired visibility, requiring all States with those areas 
to prepare emission control programs to address it. Even though a given 
State may not have any mandatory Federal Class I areas, pollution that 
occurs in that State may contribute to impairment in such Class I areas 
elsewhere. The rule encourages states to work together to determine 
whether or how much emissions from sources in a given state affect 
visibility in a downwind mandatory Federal Class I area.
    The regional haze program also calls for states to establish goals 
for improving visibility in national parks and wilderness areas to 
improve visibility on the haziest 20 percent of days and to ensure that 
no degradation occurs on the clearest 20 percent of days (64 FR 35722, 
July 1, 1999). The rule requires states to develop long-term strategies 
including enforceable measures designed to meet reasonable progress 
goals toward natural visibility conditions. Under the regional haze

[[Page 28351]]

program, States can take credit for improvements in air quality 
achieved as a result of other Clean Air Act programs, including 
national mobile source programs.\85\
---------------------------------------------------------------------------

    \85\ In a recent case, American Corn Growers Association v. EPA, 
291 F. 3d 1 (D.C. Cir 2002), the court vacated the Best Available 
Retrofit Technology (BART) provisions of the Regional Haze rule, but 
the court denied industry's challenge to EPA's requirement that 
states' SIPs provide for reasonable progress towards achieving 
natural visibility conditions in national parks and wilderness areas 
and the ``no degradation'' requirement. Industry did not challenge 
requirements to improve visibility on the haziest 20 percent of 
days. A copy of this decision can be found in Docket A-2000-01, 
Document IV-A-113.
---------------------------------------------------------------------------

    In the PM air quality modeling described above, we also modeled 
visibility conditions in the mandatory Federal Class I areas, and we 
summarize the results by region in Table I.C-2. The information shows 
that these areas also are predicted to have high annual average 
deciview levels in the future. Emissions from nonroad land-based diesel 
engines and locomotive and marine engines contributed significantly to 
these levels, because these diesel engines represent a sizeable portion 
of the total inventory of anthropogenic emissions related to 
PM2.5 (as shown in the tables above.). Furthermore, numerous 
types of nonroad engines may operate in or near mandatory Federal Class 
I areas (e.g., mining, construction, and agricultural equipment). As 
summarized in the table, we expect visibility improvements in mandatory 
Federal Class I areas from the reductions of emissions from nonroad 
diesel engines subject to this proposed rule.

 Table I.C-2--Summary of Modeled 2030 Visibility Conditions in Mandatory
                          Federal Class I Areas
                        [Annual average deciview]
------------------------------------------------------------------------
                                                Predicted
                                   Predicted       2030       Change in
            Region a                  2030      visibility     annual
                                   visibility   with rule      average
                                   baseline b   control c     deciviews
------------------------------------------------------------------------
Eastern:
    Southeast...................        21.62        21.38          0.24
    Northeast/Midwest...........        18.56        18.32          0.24
Western:
    Southwest...................         7.03         6.82          0.21
    California..................         9.56         9.26          0.3
    Rocky Mountain..............         8.55         8.34          0.21
    Northwest...................        12.18        11.94          0.24
National Class I Area Average...         11.8        11.56          0.24 
------------------------------------------------------------------------
Notes:
a Regions are depicted in Figure VI-5 in the Regulatory Support
  Document. Background visibility conditions differ by region: Eastern
  natural background is 9.5 deciviews (or visual range of 150
  kilometers) and in the West natural background is 5.3 deciviews (or
  visual range of 230 kilometers).
b The results average visibility conditions for mandatory Federal Class
  I areas in the regions.
c The results illustrate the type of visibility improvements for the
  preliminary control option, as discussed in the draft RIA. The
  proposal differs based on updated information; however, we believe
  that the net results would approximate future PM emissions, although
  we anticipate the improvements would be slightly smaller.

2. Acid Deposition
    Acid deposition, or acid rain as it is commonly known, occurs when 
SO2 and NOX react in the atmosphere with water, 
oxygen, and oxidants to form various acidic compounds that later fall 
to earth in the form of precipitation or dry deposition of acidic 
particles.\86\ It contributes to damage of trees at high elevations and 
in extreme cases may cause lakes and streams to become so acidic that 
they cannot support aquatic life. In addition, acid deposition 
accelerates the decay of building materials and paints, including 
irreplaceable buildings, statues, and sculptures that are part of our 
nation's cultural heritage. To reduce damage to automotive paint caused 
by acid rain and acidic dry deposition, some manufacturers use acid-
resistant paints, at an average cost of $5 per vehicle--a total of $80-
85 million per year when applied to all new cars and trucks sold in the 
U.S.
---------------------------------------------------------------------------

    \86\ Much of the information in this subsection was excerpted 
from the EPA document, Human Health Benefits from Sulfate Reduction, 
written under title IV of the 1990 Clean Air Act Amendments, U.S. 
EPA, Office of Air and Radiation, Acid Rain Division, Washington, DC 
20460, November 1995. Available in Docket A-2000-01, Document No. 
II-A-32.
---------------------------------------------------------------------------

    Acid deposition primarily affects bodies of water that rest atop 
soil with a limited ability to neutralize acidic compounds. The 
National Surface Water Survey (NSWS) investigated the effects of acidic 
deposition in over 1,000 lakes larger than 10 acres and in thousands of 
miles of streams. It found that acid deposition was the primary cause 
of acidity in 75 percent of the acidic lakes and about 50 percent of 
the acidic streams, and that the areas most sensitive to acid rain were 
the Adirondacks, the mid-Appalachian highlands, the upper Midwest and 
the high elevation West. The NSWS found that approximately 580 streams 
in the Mid-Atlantic Coastal Plain are acidic primarily due to acidic 
deposition. Hundreds of the lakes in the Adirondacks surveyed in the 
NSWS have acidity levels incompatible with the survival of sensitive 
fish species. Many of the over 1,350 acidic streams in the Mid-Atlantic 
Highlands (mid-Appalachia) region have already experienced trout losses 
due to increased stream acidity. Emissions from U.S. sources contribute 
to acidic deposition in eastern Canada, where the Canadian government 
has estimated that 14,000 lakes are acidic. Acid deposition also has 
been implicated in contributing to degradation of high-elevation spruce 
forests that populate the ridges of the Appalachian Mountains from 
Maine to Georgia. This area includes national parks such as the 
Shenandoah and Great Smoky Mountain National Parks.
    A study of emissions trends and acidity of water bodies in the 
Eastern U.S. by the General Accounting Office (GAO) found that from 
1992 to 1999 sulfates declined in 92 percent of a representative sample 
of lakes, and nitrate levels increased in 48 percent of the lakes 
sampled.\87\ The decrease in sulfates is consistent with emissions

[[Page 28352]]

trends, but the increase in nitrates is inconsistent with the stable 
levels of nitrogen emissions and deposition. The study suggests that 
the vegetation and land surrounding these lakes have lost some of their 
previous capacity to use nitrogen, thus allowing more of the nitrogen 
to flow into the lakes and increase their acidity. Recovery of 
acidified lakes is expected to take a number of years, even where soil 
and vegetation have not been ``nitrogen saturated,'' as EPA called the 
phenomenon in a 1995 study.\88\ This situation places a premium on 
reductions of SOx and especially NOX from all 
sources, including nonroad diesel engines, in order to reduce the 
extent and severity of nitrogen saturation and acidification of lakes 
in the Adirondacks and throughout the U.S.
---------------------------------------------------------------------------

    \87\ Acid Rain: Emissions Trends and Effects in the Eastern 
United States, U.S. General Accounting Office, March, 2000 (GOA/
RCED-00-47). Available in Docket A-99-06, Document No. IV-G-159.
    \88\ Acid Deposition Standard Feasibility Study: Report to 
Congress, EPA 430R-95-001a, October, 1995.
---------------------------------------------------------------------------

    The SOX and NOX reductions from today's 
action will help reduce acid rain and acid deposition, thereby helping 
to reduce acidity levels in lakes and streams throughout the country 
and help accelerate the recovery of acidified lakes and streams and the 
revival of ecosystems adversely affected by acid deposition. Reduced 
acid deposition levels will also help reduce stress on forests, thereby 
accelerating reforestation efforts and improving timber production. 
Deterioration of our historic buildings and monuments, and of 
buildings, vehicles, and other structures exposed to acid rain and dry 
acid deposition also will be reduced, and the costs borne to prevent 
acid-related damage may also decline. While the reduction in sulfur and 
nitrogen acid deposition will be roughly proportional to the reduction 
in SOX and NOX emissions, respectively, the 
precise impact of today's action will differ across different areas.
3. Eutrophication and Nitrification
    Eutrophication is the accelerated production of organic matter, 
particularly algae, in a water body. This increased growth can cause 
numerous adverse ecological effects and economic impacts, including 
nuisance algal blooms, dieback of underwater plants due to reduced 
light penetration, and toxic plankton blooms. Algal and plankton blooms 
can also reduce the level of dissolved oxygen, which can also adversely 
affect fish and shellfish populations.
    In 1999, NOAA published the results of a five year national 
assessment of the severity and extent of estuarine eutrophication. An 
estuary is defined as the inland arm of the sea that meets the mouth of 
a river. The 138 estuaries characterized in the study represent more 
than 90 percent of total estuarine water surface area and the total 
number of U.S. estuaries. The study found that estuaries with moderate 
to high eutrophication conditions represented 65 percent of the 
estuarine surface area. Eutrophication is of particular concern in 
coastal areas with poor or stratified circulation patterns, such as the 
Chesapeake Bay, Long Island Sound, or the Gulf of Mexico. In such 
areas, the ``overproduced'' algae tends to sink to the bottom and 
decay, using all or most of the available oxygen and thereby reducing 
or eliminating populations of bottom-feeder fish and shellfish, 
distorting the normal population balance between different aquatic 
organisms, and in extreme cases causing dramatic fish kills.
    Severe and persistent eutrophication often directly impacts human 
activities. For example, losses in the nation's fishery resources may 
be directly caused by fish kills associated with low dissolved oxygen 
and toxic blooms. Declines in tourism occur when low dissolved oxygen 
causes noxious smells and floating mats of algal blooms create 
unfavorable aesthetic conditions. Risks to human health increase when 
the toxins from algal blooms accumulate in edible fish and shellfish, 
and when toxins become airborne, causing respiratory problems due to 
inhalation. According to the NOAA report, more than half of the 
nation's estuaries have moderate to high expressions of at least one of 
these symptoms--an indication that eutrophication is well developed in 
more than half of U.S. estuaries.
    In recent decades, human activities have greatly accelerated 
nutrient inputs, such as nitrogen and phosphorous, causing excessive 
growth of algae and leading to degraded water quality and associated 
impairments of freshwater and estuarine resources for human uses.\89\ 
Since 1970, eutrophic conditions worsened in 48 estuaries and improved 
in 14. In 26 systems, there was no trend in overall eutrophication 
conditions since 1970.\90\ On the New England coast, for example, the 
number of red and brown tides and shellfish problems from nuisance and 
toxic plankton blooms have increased over the past two decades, a 
development thought to be linked to increased nitrogen loadings in 
coastal waters. Long-term monitoring in the U.S., Europe, and other 
developed regions of the world shows a substantial rise of nitrogen 
levels in surface waters, which are highly correlated with human-
generated inputs of nitrogen to their watersheds.
---------------------------------------------------------------------------

    \89\ Deposition of Air Pollutants to the Great Waters, Third 
Report to Congress, June, 2000. Available in Docket A-99-06, 
Document No. IV-A-06.
    \90\ Deposition of Air Pollutants to the Great Waters, Third 
Report to Congress, June, 2000. Great Waters are defined as the 
Great Lakes, the Chesapeake Bay, Lake Champlain, and coastal waters. 
The first report to Congress was delivered in May, 1994; the second 
report to Congress in June, 1997. Available in Docket A-99-06, 
Document No. IV-A-06.
---------------------------------------------------------------------------

    Between 1992 and 1997, experts surveyed by National Oceanic and 
Atmospheric Administration (NOAA) most frequently recommended that 
control strategies be developed for agriculture, wastewater treatment, 
urban runoff, and atmospheric deposition.\91\ In its Third Report to 
Congress on the Great Waters, EPA reported that atmospheric deposition 
contributes from 2 to 38 percent of the nitrogen load to certain 
coastal waters.\92\ A review of peer reviewed literature in 1995 on the 
subject of air deposition suggests a typical contribution of 20 percent 
or higher.\93\ Human-caused nitrogen loading to the Long Island Sound 
from the atmosphere was estimated at 14 percent by a collaboration of 
Federal and State air and water agencies in 1997.\94\ The National 
Exposure Research Laboratory, U.S. EPA, estimated based on prior 
studies that 20 to 35 percent of the nitrogen loading to the Chesapeake 
Bay is attributable to atmospheric deposition.\95\ The mobile source 
portion of atmospheric NOX contribution to the Chesapeake 
Bay was modeled at about 30 percent of total air deposition.\96\
---------------------------------------------------------------------------

    \91\ Bricker, Suzanne B., et al., National Estuarine 
Eutrophication Assessment, Effects of Nutrient Enrichment in the 
Nation's Estuaries, National Ocean Service, National Oceanic and 
Atmospheric Administration, September, 1999. Available in Docket A-
99-06, Document No. IV-G-145.
    \92\ Deposition of Air Pollutants to the Great Waters, Third 
Report to Congress, June, 2000. Available in Docket A-99-06, 
Document No. IV-A-06.
    \93\ Valigura, Richard, et al., Airsheds and Watersheds II: A 
Shared Resources Workshop, Air Subcommittee of the Chesapeake Bay 
Program, March, 1997. Available in Docket A-99-06, Document No. IV-
G-144.
    \94\ The Impact of Atmospheric Nitrogen Deposition on Long 
Island Sound, The Long Island Sound Study, September, 1997.
    \95\ Dennis, Robin L., Using the Regional Acid Deposition Model 
to Determine the Nitrogen Deposition Airshed of the Chesapeake Bay 
Watershed, SETAC Technical Publications Series, 1997.
    \96\ Dennis, Robin L., Using the Regional Acid Deposition Model 
to Determine the Nitrogen Deposition Airshed of the Chesapeake Bay 
Watershed, SETAC Technical Publications Series, 1997.
---------------------------------------------------------------------------

    Deposition of nitrogen from nonroad diesel engines contributes to 
elevated nitrogen levels in waterbodies. The proposed standards for 
nonroad diesel

[[Page 28353]]

engines will reduce total NOX emissions by 831,000 tons in 
2030. The NOX reductions will reduce the airborne nitrogen 
deposition that contributes to eutrophication of watersheds, 
particularly in aquatic systems where atmospheric deposition of 
nitrogen represents a significant portion of total nitrogen loadings.
4. Polycyclic Organic Matter Deposition
    EPA's Great Waters Program has identified 15 pollutants whose 
deposition to water bodies has contributed to the overall contamination 
loadings to the these Great Waters.\97\ One of these 15 pollutants, a 
group known as polycyclic organic matter (POM), are compounds that are 
mainly adhered to the particles emitted by mobile sources and later 
fall to earth in the form of precipitation or dry deposition of 
particles. The mobile source contribution of the 7 most toxic POM is at 
least 62 tons/year and represents only those POM that adhere to mobile 
source particulate emissions.\98\ The majority of these emissions are 
produced by diesel engines.
---------------------------------------------------------------------------

    \97\ Deposition of Air Pollutants to the Great Waters-Third 
Report to Congress, June, 2000, Office of Air Quality Planning and 
Standards Deposition of Air Pollutants to the Great Waters-Second 
Report to Congress, Office of Air Quality Planning and Standards, 
June 1997, EPA-453/R-97-011. Available in Docket A-99-06, Document 
No. IV-A-06.
    \98\ The 1996 National Toxics Inventory, Office of Air Quality 
Planning and Standards, October 1999.
---------------------------------------------------------------------------

    The PM reductions from this proposed action will help reduce not 
only the PM emissions from nonroad diesel engines but also the 
deposition of the POM adhering to the particles, thereby helping to 
reduce health effects of POM in lakes and streams, accelerate the 
recovery of affected lakes and streams, and revive the ecosystems 
adversely affected.
5. Plant Damage From Ozone
    Ground-level ozone can also cause adverse welfare effects. 
Specifically, ozone enters the leaves of plants where it interferes 
with cellular metabolic processes. This interference can be manifest 
either as visible foliar injury from cell injury or death, and/or as 
decreased plant growth and yield due to a reduced ability to produce 
food. With fewer resources, the plant reallocates existing resources 
away from root storage, growth and reproduction toward leaf repair and 
maintenance. Plants that are stressed in these ways become more 
susceptible to disease, insect attack, harsh weather and other 
environmental stresses. Because not all plants are equally sensitive to 
ozone, ozone pollution can also exert a selective pressure that leads 
to changes in plant community composition.
    Since plants are at the center of the food web in many ecosystems, 
changes to the plant community can affect associated organisms and 
ecosystems (including the suitability of habitats that support 
threatened or endangered species and below ground organisms living in 
the root zone). Given the range of plant sensitivities and the fact 
that numerous other environmental factors modify plant uptake and 
response to ozone, it is not possible to identify threshold values 
above which ozone is toxic and below which it is safe for all plants. 
However, in general, the science suggests that ozone concentrations of 
0.10 ppm or greater can be phytotoxic to a large number of plant 
species, and can produce acute foliar injury responses, crop yield loss 
and reduced biomass production. Ozone concentrations below 0.10 ppm 
(0.05 to 0.09 ppm) can produce these effects in more sensitive plant 
species, and have the potential over a longer duration of creating 
chronic stress on vegetation that can lead to effects of concern such 
as reduced plant growth and yield, shifts in competitive advantages in 
mixed populations, and decreased vigor leading to diminished resistance 
to pests, pathogens, and injury from other environmental stresses.
    Studies indicate that these effects described here are still 
occurring in the field under ambient levels of ozone. The economic 
value of some welfare losses due to ozone can be calculated, such as 
crop yield loss from both reduced seed production (e.g., soybean) and 
visible injury to some leaf crops (e.g., lettuce, spinach, tobacco) and 
visible injury to ornamental plants (i.e., grass, flowers, shrubs), 
while other types of welfare loss may not be fully quantifiable in 
economic terms (e.g., reduced aesthetic value of trees growing in Class 
I areas).
    As discussed above, nonroad diesel engine emissions of VOCs and 
NOX contribute to ozone. This proposed rule would reduce 
ozone and, therefore, help to reduce crop damage and stress from ozone 
on vegetation. See the draft RIA for a more detailed discussion of the 
science of these effects.

D. Other Criteria Pollutants Affected by This NPRM

    The standards being proposed today would also help reduce levels of 
other pollutants for which NAAQS have been established: carbon monoxide 
(CO), nitrogen dioxide (NO2), and sulfur dioxide 
(SO2). Currently every area in the United States has been 
designated to be in attainment with the NO2 NAAQS. As of 
November 4, 2002, there were 24 areas designated as non-attainment with 
the SO2 standard, and 14 designated CO non-attainment areas.
    The current primary NAAQS for CO are 35 parts per million for the 
one-hour average and 9 parts per million for the eight-hour average. 
These values are not to be exceeded more than once per year. Over 22 
million people currently live in the 14 non-attainment areas for the CO 
NAAQS. See the draft RIA for a detailed discussion of the emission 
benefits of this proposed rule.
    Carbon monoxide is a colorless, odorless gas produced through the 
incomplete combustion of carbon-based fuels. Carbon monoxide enters the 
bloodstream through the lungs and reduces the delivery of oxygen to the 
body's organs and tissues. The health threat from CO is most serious 
for those who suffer from cardiovascular disease, particularly those 
with angina or peripheral vascular disease. Healthy individuals also 
are affected, but only at higher CO levels. Exposure to elevated CO 
levels is associated with impairment of visual perception, work 
capacity, manual dexterity, learning ability and performance of complex 
tasks.
    Land-based nonroad engines contributed about one percent of CO from 
mobile sources in 1996. EPA previously determined that the category of 
nonroad diesel engines cause or contribute to ambient CO and ozone in 
more than one non-attainment area (65 FR 76790, December 7, 2000). In 
that action EPA found that nonroad engines contribute to CO non-
attainment in areas such as Los Angeles, Phoenix, Spokane, Anchorage, 
and Las Vegas. Nonroad land-based diesel engines emitted 927,500 tons 
of CO in 1996 (1% of mobile source CO).

E. Emissions From Nonroad Diesel Engines

    Emissions from nonroad diesel engines will continue to be a 
significant part of the emissions inventory in the coming years. In the 
absence of new emission standards, we expect overall emissions from 
nonroad diesel engines subject to this proposal to generally decline 
across the nation for the next 10 to 15 years, depending on the 
pollutant.\99\ Although nonroad diesel engine emissions will decline 
during this period, this trend will not be enough to adequately reduce 
the large amount of emissions that these engines contribute. For 
example, the declines are insufficient to prevent significant

[[Page 28354]]

contributions to nonattainment of PM2.5 and ozone NAAQS, or 
to prevent widespread exposure to significant concentrations of nonroad 
engine air toxics. In addition, after the 2010 to 2015 time period we 
project that this trend reverses and emissions rise into the future in 
the absence of additional regulation of these engines. (This phenomenon 
is further described later in this section.) The initial downward trend 
occurs as the nonroad fleet becomes increasingly dominated over time by 
engines that comply with existing emission regulations. The upturn in 
emissions beginning around 2015 results as growth in the nonroad sector 
overtakes the effect of the existing emission standards.
---------------------------------------------------------------------------

    \99\ As defined here, nonroad diesel engines include land-based, 
locomotive, commercial marine vessel, and recreational marine 
engines.
---------------------------------------------------------------------------

    The engine and fuel standards in this proposal will affect fine 
particulate matter (PM2.5), oxides of nitrogen 
(NOX), sulfur oxides (SO2), volatile organic 
hydrocarbons (VOC), and air toxics. For locomotive, commercial marine 
vessel (CMV), and recreational marine vessel (RMV) engines, the 
proposed fuel standards will affect PM2.5 and 
SO2. CO is not specifically targeted in this proposal but 
its reductions are discussed in the draft RIA.\100\
---------------------------------------------------------------------------

    \100\ We are proposing only a few minor adjustments of a 
technical nature to current CO standards.
---------------------------------------------------------------------------

    Each sub-section within section II discusses the emissions of a 
pollutant that the proposal addresses.\101\ This is followed by a 
discussion of the expected emission reductions associated with the 
proposed standards for land-based nonroad diesel engines.\102\ The 
tables and figures illustrate the Agency's projection of future 
emissions from nonroad diesel engines for each pollutant.\103\ The 
baseline case represents future emissions from land-based nonroad 
diesel engines with current standards. The controlled case estimates 
the future emissions of these engines based on the proposed standards 
in this notice.
---------------------------------------------------------------------------

    \101\ The estimates of baseline emissions and emissions 
reductions from the proposed rule reported here for nonroad land-
based, recreational marine, locomotive, and commercial marine vessel 
diesel engines are based on 50 state emissions inventory estimates. 
However, 50 state emissions inventory data are not available for 
other emission sources. Thus, emissions estimates for other sources 
are based on a 48 state inventory that excludes Alaska and Hawaii. 
The 48 state inventory was done for air quality modeling that EPA 
uses to analyze regional ozone transport, of which Alaska and Hawaii 
are not a part. In cases where land-based nonroad diesel engine 
emissions are summed or compared with other emissions sources, we 
use a 48 state emissions inventory.
    \102\ For the purpose of this proposal, land-based nonroad 
diesel engines include engines used in equipment modeled by the 
draft NONROAD emissions model, except for recreational marine 
engines. Recreational marine diesel engines are not subject to the 
exhaust emission standards contained in this proposal but would be 
affected by the fuel sulfur requirements applicable to locomotive 
and commercial marine vessel engines.
    \103\ The air quality modeling results described in sections 
II.B and II.C use a slightly different emissions inventory based on 
earlier, preliminary modeling assumptions. Chapter 3 of the draft 
RIA and the technical support documents fully describe this 
inventory, as well as the differences between it and the inventory 
reflecting the proposal.
---------------------------------------------------------------------------

1. PM2.5
    As described earlier in this section of the preamble, the Agency 
believes that reductions of diesel PM2.5 emissions are 
needed as part of the Nation's progress toward clean air and to reach 
attainment of the NAAQS for PM2.5. The nonroad engines 
controlled by this proposal are the major sources of nonroad diesel 
emissions. Table II.E-1 shows that the PM2.5 emissions from land-based 
nonroad diesels amount to increasingly large percentages of total 
manmade diesel PM2.5 in the years 1996, 2020 and 
2030.104 105
---------------------------------------------------------------------------

    \104\ Nitrate and sulfate secondary fine particulate as 
described in section II.B and are not included in the values 
reported here or elsewhere, but are discussed in the Regulatory 
Impact Analysis, chapter X.
    \105\ As a function of the available national inventories from 
other sources, we are only able to present a 48-state inventory. 
Wherever possible we present a 50-state inventory.

        Table II.E-1--Base-Case National (48 State) Diesel PM2.5
                              (Short tons)
------------------------------------------------------------------------
                                                                Nonroad
                                                                 land-
                                                     Nonroad     based
                                           Total      land-     percent
                  Year                     diesel     based     of total
                                           PM2.5      diesel     diesel
                                                      PM2.5      PM2.5
                                                               (percent)
------------------------------------------------------------------------
1996...................................    414,000    177,000         43
2020...................................    206,000    124,000         60
2030...................................    220,000    140,000         64
------------------------------------------------------------------------

    The contribution of land-based nonroad CI engines to PM2.5 
inventories can be significant, especially in densely populated urban 
areas.\106\ As illustrated in Table II.E.-2, our city-specific analysis 
of selected metropolitan areas for 1996 and 2020 shows that the land-
based nonroad diesel engine contribution to total PM2.5 
ranges up to 18 percent in 1996 and 19 percent in 2020.\107\
---------------------------------------------------------------------------

    \106\ Construction, industrial, and commercial nonroad diesel 
equipment comprise most of the land-based nonroad emissions 
inventory. These types of equipment are more concentrated in urban 
areas where construction projects, manufacturing, and commercial 
operations are prevalent. For more information, please refer to the 
report, ``Geographic Allocation of State Level Nonroad Engine 
Population Data to the County Level,'' NR-014b, EPA 420-P-02-009.
    \107\ We selected these cities to show a collection of typical 
cities spread across the United States in order to compare typical 
urban inventories with national average ones.

Table II.E-2--Baseline Land-Based Nonroad Diesel Percent Contribution to
       PM2.5 Inventories in Selected Urban Areas in 1996 and 2020
------------------------------------------------------------------------
                                               Land-Based    Land-Based
                                                 Nonroad       Nonroad
                                                  PM2.5         PM2.5
                 MSA, State                   Contribution  Contribution
                                                to Total      to Total
                                                PM2.5a in     PM2.5a in
                                                  1996          2020
------------------------------------------------------------------------
Atlanta, GA.................................            7             6
Boston, MA..................................           18            18
Chicago, IL.................................            8             7
Dallas-Ft. Worth, TX........................           13            10
Indianapolis, IN............................           15            13
Minneapolis-St. Paul, MN....................           10             8
New York, NY................................           13            12
Orlando, FL.................................           14            12
Sacramento, CA..............................            7             7
San Diego, CA...............................            9             7
Denver, CO..................................           11             8
El Paso, TX.................................           15            19
Las Vegas, NV...............................           15            12
Phoenix-Mesa, AZ............................           15            12
Seattle, WA.................................            7             7
National Averageb...........................            8            6
------------------------------------------------------------------------
\a\ Includes only direct exhaust diesel emissions; see Section II.C for
  a discussion of secondary fine PM levels.
\b\ This is a 48 state national average.

    Emissions of PM2.5 from land-based nonroad diesel 
engines based on a 50 state inventory are shown in Table II.E-3, along 
with our estimates of the reductions in 2020 and 2030 we expect would 
result from our proposal for a PM2.5 exhaust emission 
standard and changes in the sulfur level in nonroad diesel fuel. For 
comparison purposes, PM2.5 emissions based on lowering 
nonroad diesel fuel sulfur levels to about 340 ppm in-use \108\ (500 
ppm maximum) without any other controls are shown, along with the 
estimated emissions with the proposed PM2.5 standard and a 
sulfur level of 11 ppm in-use (15 ppm maximum). Figure II.E-1 shows our 
estimate of PM2.5 emissions between 2000 and 2030 both 
without

[[Page 28355]]

and with the proposed PM2.5 standard (along with an assumed 
sulfur level of 11 ppm in-use, 15 ppm maximum). By 2030, we estimate 
that PM2.5 emissions from this source would be reduced by 86 
percent in that year.
---------------------------------------------------------------------------

    \108\ This value (340 ppm) represents the average in-use sulfur 
concentration of fuel produced to meet a 500 ppm sulfur standard. In 
practice, off-highway equipment will sometimes be refueled with 
diesel fuel meeting the more stringent highway standard of 15 ppm. 
Therefore, the actual average in-use sulfur level of the fuel used 
by off-highway equipment will be somewhat lower than 340 ppm. The 
emission benefits shown here reflect this lower in-use sulfur level.

   Table II.E-3.--Estimated National (50 State) Reductions in PM2.5 Emissions From Nonroad Land-Based, Locomotive, Commercial Marine, and Recreational
                                                                  Marine Diesel Engines
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                  PM2.5 reductions
                                                                                PM2.5 with 500   with 500 ppm fuel   PM2.5 with rule    PM2.5 reductions
                                                             PM2.5* without    ppm fuel sulfur    sulfur (340 in-     (15 ppm sulfur   with rule (15 ppm
                           Year                               rule  [short     (340 in-use) and  use) and no other  level, 11 in-use)   sulfur level, 11
                                                                 tons]        no other controls   controls  [short     [short tons]     in-use)  [short
                                                                                 [short tons]          tons]                                 tons]
--------------------------------------------------------------------------------------------------------------------------------------------------------
2020.....................................................            186,000            163,000            100,000             23,000             86,000
2030.....................................................            205,000            178,000             77,000             27,000            127,000
--------------------------------------------------------------------------------------------------------------------------------------------------------

                                                                                                                                       [GRAPHIC] [TIFF OMITTED] TP23MY03.001
                                                                                                                                       
    Nonroad diesel engines used in locomotives, commercial marine 
vessels, and recreational marine vessels are not affected by the 
emission standards of this proposal. PM2.5 emissions from 
these engines would be reduced by the reductions in diesel fuel sulfur 
for these types of engines from an in-use average of between 2,300 and 
2,400 ppm today to an in-use average of about 340 ppm (500 ppm maximum) 
in 2007. The estimated reductions in PM2.5 emissions from 
these engines based on the proposed change in diesel fuel sulfur are 
about 6,000 tons in 2020 and 7,000 tons in 2030.\109\ For more 
information on proposed fuel sulfur reductions, please see chapter 7 of 
the draft RIA.
---------------------------------------------------------------------------

    \109\ These reductions are based on a 50 state emissions 
inventory estimate.
---------------------------------------------------------------------------

2. NOX
    Table II.E-4 shows the 50 state estimated tonnage of NOX 
emissions for 2020 and 2030 without the proposed rule and the estimated 
tonnage of emissions eliminated with the proposed rule in place. These 
results are shown graphically in Figure II.E-2. By 2030, we estimate 
that NOX emissions from these engines will be reduced by 67 
percent in that year.

     Table II.E.-4.--Estimated National (50 State) Reductions in NOX
            Emissions From Nonroad Land-Based Diesel Engines
------------------------------------------------------------------------
                                                                 NOX
                                   NOX without    NOX with    reductions
          Calendar year                rule         rule      with rule
                                      [short       [short       [short
                                      tons]        tons]        tons]
------------------------------------------------------------------------
2020.............................    1,147,000      640,000      507,000
2030.............................    1,239,000      412,000      827,000
------------------------------------------------------------------------


[[Page 28356]]

[GRAPHIC] [TIFF OMITTED] TP23MY03.002

    Table E.II-5 shows that the engines affected by the proposal emit a 
significant portion of total NOX emissions in 1996 and 2020, 
especially in cities. This is not surprising given the high density of 
these engines operating in urban areas.\110\ We selected a variety of 
cities from across the nation and found that these engines contribute 
up to 14 percent of the total NOX inventories in 1996 and as 
much as 20 percent to total NOX inventories in 2020.\111\
---------------------------------------------------------------------------

    \110\ Construction, industrial, and commercial nonroad diesel 
equipment comprise most of the land-based nonroad emissions 
inventory. These types of equipment are more concentrated in urban 
areas where construction projects, manufacturing, and commercial 
operations are prevalent. For more information, please refer to the 
report, ``Geographic Allocation of State Level Nonroad Engine 
Population Data to the County Level,'' NR-014b, EPA 420-P-02-009.
    \111\ We selected these cities to show a collection of typical 
cities spread across the United States in order to compare typical 
urban inventories with national average ones.[FEDREG][VOL]*[/
VOL][NO]*[/NO][DATE]*[/DATE][PRORULES][PRORULE][PREAMB][AGENCY]*[/
AGENCY][SUBJECT]*[/SUBJECT][/PREAMB][SUPLINF][HED]*[/HED]

Table II.E-5--Baseline Land-Based Nonroad Diesel Percent Contribution to
             NOX Inventories in Selected Urban Areas in 2020
------------------------------------------------------------------------
                                 Land-based NR NOX    Land-based NR NOX
          MSA, State              as percentage of     as percentage of
                                 total NOX in 1996    total NOX in 2020
------------------------------------------------------------------------
Atlanta, GA...................                    5                    7
Boston, MA....................                   14                   19
Chicago, IL...................                    6                    7
Dallas-Fort Worth, TX.........                   10                   13
Indianapolis, IN..............                    8                   12
Minneapolis-St. Paul, MN......                    6                    6
New York, NY..................                   11                   20
Orlando, FL...................                   10                   13
Sacramento, CA................                   10                   19
San Diego, CA.................                    9                   14
Denver, CO....................                    8                    8
El Paso, TX...................                    8                   15
Las Vegas, NV-AZ..............                   11                   12
Phoenix-Mesa, AZ..............                    9                   11
Seattle, WA...................                    8                   11
National Averagea.............                    6                   7
------------------------------------------------------------------------
a This is a 48 state national average.

3. SO2
    We estimate that land-based nonroad, CMV, RMV, and locomotive 
diesel engines emitted about 227,000 tons of SO2 in 1996, 
accounting for about 30 percent of the SO2 from mobile 
sources (based on a 48 state inventory). With no reduction in diesel 
fuel sulfur levels, we estimate that these emissions will continue to 
increase, accounting for about 60 percent of mobile source 
SO2 emissions by 2030.
    As part of this proposal, sulfur levels in fuel would be 
significantly reduced, leading to large reductions in nonroad diesel 
SO2 emissions. By 2007, the sulfur in diesel fuel used by 
all nonroad diesel engines would be reduced from the current average 
in-use level of between 2,300 and 2,400 ppm to an average in-use level 
of about 340 ppm with a maximum level of 500 ppm. By 2010, the sulfur 
in diesel fuel used by land-based nonroad engines would be

[[Page 28357]]

reduced to an average in-use level of 11 ppm with a maximum level of 15 
ppm. The sulfur in diesel fuel used by locomotives, CMVs, and RMVs 
would remain at an average in-use level of about 340 ppm. Figure II.E-3 
shows the estimated reductions from these sulfur changes. For more 
information on this topic, please see chapter 7 of the RIA.\112\
---------------------------------------------------------------------------

    \112\ Under this proposal, the introduction of 340 ppm 
(approximate average in-use level, 500 ppm maximum) sulfur diesel 
fuel for all nonroad diesel engines would take place in June of 
2007. The introduction of 11 ppm sulfur diesel fuel (average in-use, 
15 ppm maximum) for land-based nonroad engines would take place in 
June 2010.
[GRAPHIC] [TIFF OMITTED] TP23MY03.003

    Table II.E-6 shows 50 state estimates of total SO2 
emissions without the proposed rule and how SO2 emissions 
would be reduced by the diesel fuel sulfur reductions in 2020 and 2030.
    Lowering diesel fuel sulfur to a maximum of 500 ppm (340 ppm in-
use) for CMV, locomotive and land-based nonroad engines would result in 
a reduction of about 360,000 tons/year of SO2 in 2030. 
Lowering diesel fuel sulfur to a maximum of 500 ppm (340 ppm in-use) 
for CMV and locomotive engines and a maximum of 15 ppm (11 ppm in-use) 
for land-based nonroad engines would result in a reduction of about 
390,000 tons of SO2 in 2030.

   Table II.E-6--Estimated National (50 State) Emissions of Land-Based Nonroad, Locomotive, Commercial Marine
                                     Vessel, and Recreational Marine Vessel
                             [SO2 Emissions From Lowering Diesel Fuel Sulfur Levels]
----------------------------------------------------------------------------------------------------------------
                                          Total SO2
                                      emissions at 2400    500 ppm sulfur     500 ppm sulfur   15 ppm sulfur (11
                                          ppm sulfur      (340 ppm in-use)  (340 in-use) land- ppm in-use) land-
                Year                   without proposed     locomotives,      based nonroad      based nonroad
                                         rule  [short       CMVs, RMVsa        [short tons]       [short tons]
                                            tons]           [short tons]
----------------------------------------------------------------------------------------------------------------
1996................................            229,000  .................  .................  .................
2020................................            345,000              9,000             26,000              1,000
2030................................            401,000             10,000             30,000             1,000
----------------------------------------------------------------------------------------------------------------
Notes:
a CMV = commercial marine vessels, RMV = Recreational marine vessels.

4. VOC and Air Toxics
    Based on a 48 state emissions inventory, we estimate that land-
based nonroad diesel engines emitted over 221 thousand tons of VOC in 
1996. Between 1996 and 2030, we estimate that land-based nonroad diesel 
engines will contribute about 2 to 3 percent to mobile source VOC 
emissions. Without further controls, land-based nonroad diesel engines 
will emit over 97

[[Page 28358]]

thousand tons/year of VOC in 2020 and 2030 nationally.\113\
---------------------------------------------------------------------------

    \113\ VOC emissions remain about the same in 2030 as 2020 
because while nonroad diesel emission factors decrease and newer 
engines continue to be introduced into the fleet, the engine/
equipment population continues to increase. The increase in engine/
equipment population offsets the effect of decreasing emission 
factors.
---------------------------------------------------------------------------

    Tables II.E-7 shows our projection of the reductions in 2020 and 
2030 for VOC emissions that we expect from implementing the proposed 
NMHC standards. This estimate is based on a 50 state emissions 
inventory. By 2030, VOC reductions would be reduced by 30 percent.

 Table II.E-7--Estimated National (50 State) Reductions in VOC Emissions From Nonroad Land-Based Diesel Engines
----------------------------------------------------------------------------------------------------------------
                                                                                                VOC reductions
                  Calendar year                      VOC without rule      VOC with rule      with rule  [short
                                                       [short tons]         [short tons]            tons]
----------------------------------------------------------------------------------------------------------------
2020.............................................               97,000               79,000               18,000
2030.............................................               98,000               68,000               30,000
----------------------------------------------------------------------------------------------------------------

    Air toxics pollutants are in VOCs and are included in the total 
land-based nonroad diesel VOC emissions estimate. We base these numbers 
on the assumption that air toxic emissions are a constant fraction of 
hydrocarbon exhaust emissions.
    Although we are not proposing any specific gaseous air toxics 
standards, air toxics emissions would nonetheless be reduced through 
NMHC standards included in the proposed rule. By 2030, we estimate that 
emissions of air toxics pollutants, such as benzene, formaldehyde, 
acetaldehyde, 1,3-butadiene, and acrolein, would be reduced by 30 
percent from land-based nonroad diesel engines. For specific air toxics 
reductions please see chapter 3 of the RIA. In section II.B.2 we 
discuss the health effects of these pollutants.?

III. Nonroad Engine Standards

    In this section we describe the nonroad diesel emission standards 
we are proposing in order to address the serious air quality problems 
discussed in section II. Specifically, we discuss:
    [sbull] The Clean Air Act and why we are proposing new emission 
standards.
    [sbull] The technology opportunity for nonroad diesel emissions 
control.
    [sbull] Our proposed engine standards, and our proposed schedule 
for implementing them.
    [sbull] Proposals for supplemental test procedures and standards to 
help control emissions during transient operating modes and engine 
start-up.
    [sbull] Proposals to help ensure robust emissions control in use.
    [sbull] The feasibility of the proposed standards (in conjunction 
with the proposed low-sulfur nonroad diesel fuel requirement discussed 
in section IV).
    [sbull] How diesel fuel sulfur affects an engine's ability to meet 
the proposed standards.
    [sbull] Plans for a future reassessment of the technology needed to 
comply with proposed standards for engines below 75 hp.
    Additional proposed provisions for engine and equipment 
manufacturers are discussed in detail in section VII. Briefly, these 
include changes to our engine manufacturer averaging, banking, and 
trading (ABT) program, changes to our transition program for equipment 
manufacturers, special provisions to aid small businesses in 
implementing our requirements, and an incentive program to encourage 
innovative technologies and the early introduction of new technologies.
    We welcome comment on all facets of this discussion, including the 
levels and timing of the proposed emissions standards and our 
assessment of technological feasibility, as well as on the supporting 
analyses contained in the Draft Regulatory Impact Analysis (RIA). We 
also request comment on the timing of the proposed diesel fuel standard 
in conjunction with these proposed emission standards. We ask that 
commenters provide any technical information that supports the points 
made in their comments.

A. Why Are We Setting New Engine Standards?

1. The Clean Air Act and Air Quality
    We believe that Agency action is needed to address the air quality 
problems discussed in section II. We are therefore proposing new engine 
standards and related provisions under sections 213(a)(3) and (4) of 
the Clean Air Act which, among other things, direct us to establish 
(and from time to time revise) emission standards for new nonroad 
diesel engines. Because emissions from these engines contribute greatly 
to a number of serious air pollution problems, especially the health 
and welfare effects of ozone, PM, and air toxics, we believe that the 
air quality need for stringent nonroad diesel standards is well 
established. This, and our belief that a significant degree of emission 
reduction from these engines is achievable through the application of 
diesel emission control technology that will be available in the lead 
time provided (giving appropriate consideration to cost, noise, safety, 
and energy factors as required by the Act), along with coordinated 
reductions in nonroad diesel fuel sulfur levels, leads us to believe 
that these new emission standards are warranted and appropriate.
    We also believe that the proposed engine standards are consistent 
with the Clean Air Act section 213 requirements on availability of 
technology and appropriate lead time. The basis for our conclusion is 
described in this section and in the Draft RIA.
2. The Technology Opportunity for Nonroad Diesel Engines
    Substantial progress has been made in recent years in controlling 
diesel exhaust emissions through the use of robust, high-efficiency 
catalytic devices placed in the exhaust system. Particularly promising 
are the catalytic soot filter or particulate trap for PM and 
hydrocarbon control, and the NOX adsorber. These 
technologies are expected to be applied to highway heavy-duty diesel 
engines (HDDEs) beginning in 2007 to meet stringent new standards for 
these engines. The final EPA rule establishing those standards contains 
extensive discussion of how these devices work, how effective they are 
at reducing emissions, and what their limitations are, particularly 
their dependence on very-low sulfur diesel fuel to function properly 
(66 FR 5002, January 18, 2001; see especially section III of the 
preamble starting at 5035). Reviews of ongoing progress in the 
development of these technologies have recently been performed by EPA 
and by

[[Page 28359]]

an independent review panel.114 115 These reviews found that 
significant progress has been made since the final rule was published, 
reinforcing our confidence that the highway engine standards can be 
met. (Our consideration of these highway engine standards is consistent 
with the requirement in Clean Air Act section 213(a)(3) that EPA 
consider nonroad engine standards equivalent in stringency to those 
adopted for comparable highway engines regulated under section 202 of 
the Act.)
---------------------------------------------------------------------------

    \114\ ``Highway Diesel Progress Review'', U.S. EPA, June 2002. 
EPA420-R-02-016. (www.epa.gov/air/caaac/dieselreview.pdf).
    \115\ ``Meeting Technology Challenges For the 2007 Heavy-Duty 
Highway Diesel Rule'', Final Report of the Clean Diesel Independent 
Review Subcommittee, Clean Air Act Advisory Committee, October 30, 
2002. (www.epa.gov/air/caaac/diesel/finalcdirpreport103002.pdf).
---------------------------------------------------------------------------

    Although there are important differences, nonroad diesel engines 
operate fundamentally like heavy-duty highway diesel engines. In fact, 
many nonroad engine designs are derived from highway engine platforms. 
We believe that, given the availability of nonroad diesel fuel meeting 
our proposed 15 ppm maximum sulfur requirement and adequate development 
lead time, nonroad diesel engines can be designed to successfully 
employ the same high-efficiency exhaust emission control technologies 
now being developed for highway use. Indeed, some nonroad diesel 
applications, such as in underground mining, have pioneered the use of 
similar technologies for many years. These technologies, the experience 
gained with them in nonroad applications, the issues involved in 
transferring technology from highway to nonroad applications, and the 
appropriate standards and test procedures for this nonroad Tier 4 
program are discussed in detail in the remainder of this section.

B. What Engine Standards Are We Proposing?

1. Exhaust Emissions Standards
    The PM, NOX, and NMHC emissions standards being proposed 
for nonroad diesel engines are summarized in Figures III.B-1 and 2. We 
are also making minor adjustments to CO standards as discussed in 
section III.B.1.f. All of these standards would apply to covered 
nonroad engines over the useful life periods specified in our 
regulations, except where temporary in-use compliance margins would 
apply as discussed in section VII.J.\116\ We are not proposing changes 
to the current useful life periods because we do not have any relevant 
new information that would lead us to propose changes. However, we do 
ask for comment on whether or not changes are warranted and, if so, on 
what the useful life periods should be. The testing requirements by 
which compliance with the standards would be measured are discussed in 
section III.C. In addition we are proposing new ``not-to-exceed'' (NTE) 
emission standards and associated test procedures to help ensure robust 
control of emissions in use. These standards are discussed as part of a 
broader outline of proposed NTE provisions in sections III.D and VII.G.
---------------------------------------------------------------------------

    \116\ The useful life for engines =50 hp is 8,000 
hours or 10 years, whichever occurs first. For engines <25 hp, and 
for 25-50 hp engines that operate at constant speed at or above 3000 
rpm, it is 3000 hours or 5 years. For other 25-50 hp engines, it is 
5,000 hours or 7 years.

                          Figure III.B-1--Proposed PM Standards (g/bhp-hr) and Schedule
----------------------------------------------------------------------------------------------------------------
                                                                           Model Year
                 Engine Power                  -----------------------------------------------------------------
                                                   2008       2009       2010       2011       2012       2013
----------------------------------------------------------------------------------------------------------------
hp < 25 (kW < 19).............................   \a\ 0.30  .........  .........  .........  .........  .........
25 <= hp < 75 (19 <= kW < 56).................    \b\0.22  .........  .........  .........  .........       0.02
75 <= hp < 175 (56 <= kW < 130)...............  .........  .........  .........  .........       0.01  .........
175 <= hp <= 750 (130 <= kW <= 560)...........  .........  .........  .........       0.01  .........  .........
hp  750 (kW  560).......  .........  .........  .........   \c\ 0.01  .........  .........
----------------------------------------------------------------------------------------------------------------
Notes:
\a\ For air-cooled, hand-startable, direct injection engines under 11 hp, a manufacturer may instead delay
  implementation until 2010 and demonstrate compliance with a less stringent PM standard of 0.45 g/bhp-hr,
  subject also to additional provisions discussed in Section III.B.1.d.i.
\b\ A manufacturer has the option of skipping the 0.22 g/bhp-hr PM standard for all 50-75 hp engines; the 0.02 g/
  bhp-hr PM standard would then take effect one year earlier for all 50-75 hp engines (in 2012).
\c\ 50% of a manufacturer's U.S.-directed production must meet the 0.01 g/bhp-hr PM standard in this model year.
  In 2014, 100% must comply.



                          Figure III.B-2--Proposed NOX and NMHC Standards and Schedule
----------------------------------------------------------------------------------------------------------------
                                                                               Standard (g/bhp-hr)
                         Engine Power                          -------------------------------------------------
                                                                          NOX                      NMHC
----------------------------------------------------------------------------------------------------------------
25 <= hp < 75 (19 <= kW < 56).................................                  3.5 NMHC+NOX \a\
75 <= hp < 175 (56 <= kW < 130)...............................                     0.30                     0.14
175 <= hp <= 750 (130 <= kW <= 560)...........................                     0.30                     0.14
hp  750 (kW  560).......................                     0.30                     0.14
----------------------------------------------------------------------------------------------------------------


----------------------------------------------------------------------------------------------------------------
                                                                               Phase-in Schedule
                        Engine Power                         ---------------------------------------------------
                                                                  2011         2012         2013         2014
----------------------------------------------------------------------------------------------------------------
25 <= hp < 75 (19 <= kW < 56)...............................  ...........  ...........         100%  ...........
75 <= hp < 175 (56 <= kW < 130).............................  ...........      \b\ 50%      \b\ 50%     \b\ 100%
175 <= hp <= 750 (130 <= kW <= 560).........................          50%          50%          50%         100%
hp  750 (kW  560).....................          50%          50%          50%        100%
----------------------------------------------------------------------------------------------------------------
Notes:
Percentages are U.S.-directed production required to comply with the Tier 4 standards in the indicated model
  year.
\a\ This is the existing Tier 3 combined NMHC+NOX standard level for the 50-75 hp engines in this category; in
  2013 it would apply to the 25-50 hp engines as well.

[[Page 28360]]

 
\b\ Manufacturers may use banked Tier 2 NMHC+NOX credits to demonstrate compliance with the proposed 75-175 hp
  engine NOX standard in this model year. Alternatively, manufacturers may forego this special banked credit
  option and instead meet an alternative phase-in requirement in 2012, 2013, and part of 2014. See Section
  III.B.1.b.

    The proposed long-term 0.01 and 0.02 g/bhp-hr Tier 4 PM standards 
for 75 hp and 25-75 hp engines, respectively, combined with 
the fuel change and proposed new requirements to ensure robust control 
in the field, represent a reduction of over 95% from in-use levels 
expected with Tier 2/Tier 3 engines.\117\ The proposed 0.30 g/bhp-hr 
Tier 4 NOX standard for 75 hp engines represents 
a NOX reduction of about 90% from in-use levels expected 
with Tier 3 engines. The basis for the proposed standard levels is 
presented in Section III.E.
a. Standards Timing
---------------------------------------------------------------------------

    \117\ Note that we are grouping all standards proposed in this 
rule under the general designation of ``Tier 4 standards'', 
including those proposed to take effect in 2008. As a result, there 
are no ``Tier 3'' standards in the multi-tier nonroad program for 
engines below 50 hp or above 750 hp.
---------------------------------------------------------------------------

    The timing of the Tier 4 NOX, PM, and NMHC standards is 
closely tied to the proposed timing of fuel quality changes discussed 
in section IV, in keeping with the systems approach we are taking for 
this program. The earliest Tier 4 standards would take effect in model 
year 2008, in conjunction with the introduction of 500 ppm maximum 
sulfur nonroad diesel fuel in mid-2007. This fuel change serves a dual 
environmental purpose. First, it provides a large immediate reduction 
in PM emissions for the existing fleet of engines in the field. Second, 
its widespread availability by the end of 2007 aids engine designers in 
employing emission controls capable of achieving the proposed standards 
for model year 2008 and later engines; this is because the performance 
and durability of such technologies as exhaust gas recirculation (EGR) 
and diesel oxidation catalysts is improved by lower sulfur fuel.\118\ 
The reduction of sulfur in nonroad diesel fuel will also provide 
sizeable economic benefits to machine operators as it will extend oil 
change intervals and reduce wear and corrosion (see section V).
---------------------------------------------------------------------------

    \118\ ``Nonroad Diesel Emissions Standards Staff Technical 
Paper'', EPA420-R-01-052, October 2001.
---------------------------------------------------------------------------

    We are not, however, proposing new 2008 standards for engines at or 
above 100 hp because these engines are subject to existing Tier 3 
NMHC+NOX standards (Tier 2 for engines above 750 hp) in 2006 
or 2007. Setting new 2008 standards would provide only one or two years 
before another round of design changes would have to be made for Tier 
4. Engines between 50-100 hp also have a Tier 3 NMHC+NOX 
standard, but it takes effect in 2008, providing an opportunity to 
coordinate with Tier 4 to provide the desired pull-ahead of PM control. 
We believe that we can accomplish this PM pull-ahead without hampering 
manufacturers' Tier 3 compliance efforts by providing two Tier 4 
compliance options for 50-75 hp engines. This reflects the splitting of 
the current 50-100 hp category of engines to match the new rated power 
\119\ categories shown in Figures III.B-1 and 2. We are proposing to 
provide manufacturers with the option to skip the Tier 4 2008 PM 
standard (see Figure III-B.1) and instead to focus design efforts on 
introducing PM filters for these engines one year earlier, in 2012. 
This option would ensure that a manufacturer's Tier 3 
NMHC+NOX compliance plans are not complicated by having to 
meet a new Tier 4 PM standard in the same timeframe, if that were to 
become a concern for a manufacturer.
---------------------------------------------------------------------------

    \119\ The term rated power is used in this document to mean the 
maximum power of an engine. See section VII.L for more information 
about how the maximum power of an engine is determined.
---------------------------------------------------------------------------

    We are concerned that this optional approach for 50-75 hp engines 
might be abused by equipment manufacturers whose engine suppliers opt 
not to meet the PM pull-ahead standard in 2008, but who then switch 
engine suppliers to avoid PM filter-equipped engines in 2012. We are 
therefore proposing that an equipment manufacturer making a product 
with engines not meeting the pull-ahead standard in any of the years 
2008-2011, must use engines in that product in 2012 meeting the 0.02 g/
bhp-hr PM standard; that is, from the same engine manufacturer or from 
another engine manufacturer choosing the same compliance option. This 
restriction would not apply if the 2008-2011 engines at issue are being 
produced under the equipment manufacturer flexibility provisions 
discussed in section VII.B. Also, we would not prohibit an equipment 
manufacturer who is using non-pull-ahead engines in 2008-2011 from 
making use of available equipment manufacturer flexibility provisions 
in 2012 or later. That is, they could continue to use Tier 3 engines in 
2012 that are purchased under these provisions; they would, however, 
still be subject to the above-described restriction on switching 
manufacturers. We solicit comment on whether this restriction should 
have a numerical basis (e.g., the ``no switch'' restriction in 2012 
applies to the same percentage of 50-75 hp machines produced with non-
pull-ahead engines in 2008-2011) to avoid further abuse by equipment 
manufacturers who redefine their product models to dodge the 
requirement, and on other suggestions for dealing with this concern.
    Note that we are not proposing the optional 2008 PM standard for 
engines between 75 and 100 hp, even though they, like the 50-75 hp 
engines, are subject to a 2008 Tier 3 standard. This is because we 
believe that these larger engines, proposed to be grouped into a new 
75-175 hp category, would be subject to stringent new PM and 
NOX standards beginning in 2012, and adding a 2008 PM 
component to this program for a quarter of this 75-175 hp range would 
complicate manufacturers' efforts to comply in 2012 for the overall 
category.
    We view the 2008 portion of the Tier 4 program as highly important 
because it provides substantial PM and NOX emissions 
reductions during the several years prior to 2011. Initiating Tier 4 in 
2008 also fits well with the lead time, stability, cost, and technology 
availability considerations of the overall program.\120\ Initiating the 
Tier 4 standards in 2008 would provide three to four years of stability 
after the start of Tier 2 for engines under 50 hp. As mentioned above, 
it also coincides with the start date of Tier 3 NOX+NMHC 
standards for engines between 50 and 75 hp and so introduces no 
stability issues for these engines. As the Agency expects to finalize 
this rule in early 2004, the 2008 start date provides almost 4 years of 
lead time to accomplish redesign and testing. The evolutionary 
character of the 2008 standards, based as they are on proven 
technologies, and the fact that some certified engines already meet 
these standards as discussed in Section

[[Page 28361]]

III.E leads us to conclude that this will provide adequate lead time.
---------------------------------------------------------------------------

    \120\ Section 213(b) of the Clean Air Act does not specify a 
minimum lead time period, nor does it mandate a set minimum period 
of stability for the standards (differing in these respects from the 
comparable provision section (202(a)(3)(C)) applicable to highway 
engines). However, in considering the amount of lead time and 
stability provided, EPA takes into consideration the need to avoid 
disruptions in the engine and equipment manufacturing industries 
caused by redesign mandates that are too frequent or too soon after 
a final rulemaking. These are appropriate factors to consider in 
determining ``the lead time necessary to permit the development and 
application of the requisite technology'', and are part of taking 
cost into consideration, as required under section 213 (b).
---------------------------------------------------------------------------

    The second fuel change, to 15 ppm maximum sulfur in mid-2010, and 
the related engine standards that begin to phase-in in the 2011 model 
year, provide the large majority of the environmental benefits of the 
program. These standards are also timed to provide adequate lead time 
for manufacturers, and to phase in over time to allow for the orderly 
transfer of technology from the highway sector. We believe that the 
high-efficiency exhaust emission technologies being developed to meet 
our 2007 emission standards for heavy-duty highway diesel engines can 
be adapted to nonroad diesel applications. The engines for which we 
believe this adaptation from highway applications will be most 
straightforward are those in the over 175 hp power range, and thus 
under our proposal these engines would be subject to new standards 
requiring high-efficiency exhaust emission controls as soon as the 15 
ppm sulfur diesel fuel is widely available, that is, in the 2011 model 
year. Engines between 75 and 175 hp would be subject to the new 
standards in the following model year, 2012, reflecting the greater 
effort involved in adapting highway technologies to these engines. 
Lastly, engines between 25 and 75 hp would be subject to the new PM 
standard in 2013, reflecting the even greater challenge of adapting PM 
filter technology to these engines which typically do not have highway 
counterparts. There are additional phase-in provisions discussed in 
Section III.B.1.b aimed at further drawing from the highway technology 
experience.
    In addition to addressing technology transfer, this approach 
reflects the need to distribute the workload for engine and equipment 
redesign over three model years, as was provided for in Tier 3. 
Overall, this approach provides 4 to 6 years of real world experience 
with the new technology in the highway sector, involving millions of 
engines (in addition to the several additional years provided by 
demonstration fleets already on the road), before the new standards 
take effect.
    b. Phase-In of NOX and NMHC Standards
    Because the Tier 4 NOX emissions control technology, 
like PM control technology, is expected to be derived from technology 
first introduced in highway HDDEs, we believe that the implementation 
of the Tier 4 NOX standard should follow the pattern we 
adopted for the highway program. This will help to ensure a focused, 
orderly development of robust high-efficiency NOX control in 
the nonroad sector and will also help to ensure that manufacturers are 
able to take maximum advantage of the highway engine development 
program, with resulting cost savings. The heavy-duty highway rule 
allows for a gradual phase-in of the NOX and NMHC 
requirements over multiple model years: 50 percent of each 
manufacturer's U.S.-directed production volume must meet the new 
standard in 2007-2009, and 100 percent must do so by 2010. We also 
provided flexibility for highway engine manufacturers to meet that 
program's environmental goals by allowing somewhat less-efficient 
NOX controls on more than 50% of their production before 
2010 via emissions averaging. Similarly, we are proposing to phase in 
the NOX standards for nonroad diesels over 2011-2013 as 
indicated in Figure III.B-2, based on compliance with the Tier 4 
standards for 50% of a manufacturer's U.S.-directed production in each 
power category at or above 75 hp in each phase-in model year.
    With a NOX phase-in, all manufacturers are able to 
introduce their new technologies on a limited number of engines, 
thereby gaining valuable experience with the technology prior to 
implementing it on their entire product line. In tandem with the 
equipment manufacturer transition program discussed in section VII.B, 
the phase-in ensures timely progress to the Tier 4 standards levels 
while providing a great degree of implementation flexibility for the 
industry.
    We are proposing this ``percent of production phase-in'' to take 
maximum advantage of the highway program technology development. It 
adds a new dimension of implementation flexibility to the staggered 
``phase-in by power category'' used in the nonroad program for Tiers 1, 
2 and 3 which, though structured to facilitate technology development 
and transfer, is more aimed at spreading the redesign workload. Because 
the Tier 4 program would involve substantial challenges in addressing 
both technology development and redesign workload, we believe that 
incorporating both of these phase-in mechanisms into the proposed 
program is warranted, resulting in the coordinated phase-in plan shown 
in Figure III.B-2. Note that this results in our proposing that new 
NOX requirements for 75-175 hp engines be deferred for the 
first year of the 2011-2013 general phase-in, in effect creating a 50-
50% phase-in in 2012-2013 for this category. This then staggers the 
Tier 4 start years by power category as in past tiers: 2011 for engines 
at or above 175 hp, 2012 for 75-175 hp engines, and 2013 for 25-75 hp 
engines (for which no NOX adsorber-based standard and thus 
no percentage phase-in is being proposed), while still providing a 
production-based phase-in for advanced NOX control 
technologies.
    We believe that the 75-175 hp category of engines and equipment may 
involve added workload challenges for the industry to develop and 
transfer technology. We note that this category, though spanning only 
100 hp, represents a great diversity of applications, and comprises a 
disproportionate number of the total nonroad engine and machine models. 
Some of these engines, though having characteristics comparable to many 
highway engines such as turbocharging and electronic fuel control, are 
not directly derived from highway engine platforms and so are likely to 
require more development work than larger engines to transfer emission 
control technology from the highway sector. Furthermore, the engine and 
equipment manufacturers have greatly varying market profiles in this 
category, from focused one- or two-product offerings to very diverse 
product lines with a great many models. We are interested in providing 
useful flexibility for a wide range of companies in implementing the 
Tier 4 standards, while keeping a priority on bringing PM emissions 
control into this diverse power category as quickly as possible.
    We are therefore proposing two compliance flexibility provisions 
just for this category. First, we propose to allow manufacturers to use 
NMHC+NOX credits generated by Tier 2 engines over 50 hp (in 
addition to any other allowable credits) to demonstrate compliance with 
the Tier 4 requirement for 75-175 hp engines in 2012, 2013, and 2014 
only. This would not otherwise be allowed, for reasons explained in 
section VII.A. These Tier 2 credits would be subject to the power 
rating conversion already established in our ABT program, and to the 
20% credit adjustment we are proposing for use of NMHC+NOX 
credits as NOX credits. (See section VII.A.)
    Second, we realize that some manufacturers, especially those with 
limited product offerings, may not have sufficient banked credits 
available to them to benefit from this special flexibility, and so we 
are also proposing an alternative flexibility provision. A manufacturer 
may optionally forego the Tier 2 banked credit use provision described 
above, and instead demonstrate compliance with a reduced phase-in 
requirement for NOX and NMHC. Use of credits other than 
banked Tier 2 credits would still be allowed, in

[[Page 28362]]

accordance with the other ABT program provisions. In no case could the 
phase-in compliance demonstration drop below 25% in each of 2012, 2013, 
and the first 9 months of 2014, except as allowed under the ``good 
faith projection deficit'' provision discussed in Section VII.D. Full 
compliance (100% phase-in) with the Tier 4 standards would need to be 
demonstrated in the last 3 months of 2014 and thereafter.
    In addition, a manufacturer using this reduced phase-in option 
would not be allowed to generate credits from engines in this power 
category in 2012, 2013, and the first 9 months of 2014, except for use 
in averaging within this power category only (no banking or trading, or 
averaging with engines in other power categories). This restriction 
would apply throughout this period even if the reduced phase-in option 
is exercised during only a portion of this period. We believe that this 
ABT restriction is important to avoid potential abuse of the added 
flexibility allowance, considering that larger engine categories will 
be required to demonstrate substantially greater compliance levels with 
the 0.30 g/bhp-hr NOX standard several years earlier than 
engines built under this option. The restriction should be no burden to 
manufacturers, as only those using the option would be subject to it, 
and the production of credit-generating engines would be contrary to 
the option's purpose.
    We are proposing to phase in the Tier 4 NMHC standard with the 
NOX standard, as is being done in the highway program. 
Engines certified to the new NOX requirement would be 
expected to certify to the NMHC standard as well. The ``phase-out'' 
engines (the 50 percent not certified to the new Tier 4 NOX 
and NMHC standards) would continue to be certified to the applicable 
Tier 3 NMHC+NOX standard. As discussed in section III.E, we 
believe that the NMHC standard is readily achievable through the 
application of PM traps to meet the PM standard, which for most engines 
does not involve a phase-in. However, in the highway program we chose 
to phase in the NMHC standard with the NOX standard for 
administrative reasons, to simplify the phase-in under the percent-of-
production approach taken there, thus avoiding subjecting the ``phase-
out'' engines to separate standards for NMHC and NMHC+NOX. 
The same reasoning applies here because, as in the highway program, the 
previous-tier standards are combined NMHC+NOX standards.
    Because of the tremendous variety of engine sizes represented in 
the nonroad diesel sector, we are proposing that the 50 percent phase-
in requirement be met separately in each of the three power categories 
for which a phase-in is proposed (75-175 hp, 175-750 hp, and 
750 hp).\121\ For example, a manufacturer that produces 1000 
engines for the 2011 U.S. market in the 175 to 750 hp range would have 
to demonstrate compliance to the proposed NOX and NMHC 
standards on at least 500 of these engines, regardless of how many 
complying engines the manufacturer produces in other hp categories. 
(Note, however, that we would allow averaging of emissions across these 
engine category cutpoints through the use of power-weighted ABT program 
credits, as provided for in the existing nonroad diesel engine 
program.) We believe that this restriction reflects the availability of 
emissions control technology, and is needed to avoid erosion of 
environmental benefits that might occur if a manufacturer with a 
diverse product offering were to meet the phase-in with relatively low 
cost smaller engines, thereby delaying compliance on larger engines 
with much higher lifetime emissions potential. Even so, the horsepower 
ranges for these power categories are fairly broad, so this restriction 
allows ample freedom to manufacturers to structure compliance plans in 
the most cost-effective manner. We could as well choose to handle this 
concern by weighting complying engines by horsepower, as we do in the 
ABT program, but we believe that creating a simple phase-in structure 
based simply on counting engines, as we did in the highway HDDE rule, 
avoids unnecessary complexity and functional overlap with ABT.
---------------------------------------------------------------------------

    \121\ Note proposed exceptions to the 50 percent requirements 
during the phase-in model years discussed in sections VII.D and 
VII.E. These deal with differences between a manufacturer's actual 
and projected production levels, and with incentives for early or 
very low emission engine introductions.
---------------------------------------------------------------------------

 c. Rationale for Restructured Horsepower Categories
    We are proposing to regroup the power categories in the proposed 
Tier 4 program compared to the previous tiers of standards.\122\ We are 
doing so because this will more closely match the degree of challenge 
involved in transferring advanced emissions control technology from 
highway engines to nonroad engines. For a variety of reasons, highway 
engines have in the past been equipped with new emission control 
technologies some years before nonroad engines. As a result, the 
nonroad engine platforms that are directly derived from highway engine 
designs in turn become the lead application point for the migration of 
emission control technologies into the nonroad sector. Smaller and 
larger nonroad engines, as well as similar-sized engines that cannot 
directly use a highway base engine (such as farm tractor engines that 
are structurally part of the tractor chassis), may then employ these 
technologies after additional lead time for needed adaptation. This 
progression has been reflected in EPA standards-setting activity to 
date, especially in implementation schedules, in which the earliest 
standards are applied to engines in the most ``highway-like'' power 
categories.
---------------------------------------------------------------------------

    \122\ The Tier 1 / 2 / 3 programs make use of 9 categories 
divided by horsepower: <11, 11-25, 25-50, 50-100, 100-175, 175-300, 
300-600, 600-750, and 750 hp.
---------------------------------------------------------------------------

    Although there is not an abrupt power cutpoint above and below 
which the highway-derived nonroad engine families do and do not exist, 
we believe that 75 hp is a more appropriate cutpoint for this purpose 
than either of the closest previously adopted power category cutpoints 
of 50 or 100 hp. These two cutpoints were first adopted in a 1994 final 
rule that chose them in order to establish categories for a staggered 
implementation schedule designed to spread out development costs (59 FR 
31306, June 17, 1994). Nonroad diesels produced today with rated power 
above 75 hp (up to several hundred hp) are mostly variants of nonroad 
engine platforms with four or more cylinders and per-cylinder 
displacements of one liter or more. These in turn are derived from or 
are similar to heavy-duty highway engine platforms. Even where nonroad 
engine models above 75 hp are not so directly derived from highway 
models, they typically share many common characteristics such as 
displacements of one liter per cylinder or more, direct injection 
fueling, turbocharging, and, increasingly, electronic fuel injection. 
These common features provide key building blocks in transferring high-
efficiency exhaust emission control technology from highway to similar 
nonroad diesel engines. We have discussed this matter with relevant 
engine manufacturers, and we are confident based on these discussions 
that 75 hp represents an industry consensus on the appropriate cutpoint 
for this purpose. We invite comment on the 75 hp cutpoint.
    We are therefore proposing to regroup power ratings using the 75 hp 
cutpoint. Some have expressed that this may somewhat complicate the 
transition from tier to tier and efforts to harmonize with the European 
Union's nonroad diesel program (which currently uses

[[Page 28363]]

power cutpoints corresponding to 50 and 100 hp). However, we believe 
that it provides substantial long-term benefits for the environment 
(for example, by linking NOX standard-setting to an engine 
technology-based 75 hp cutpoint). We will continue working with key 
entities to advance harmonization as this rule is developed.
    We are also proposing to consolidate some power categories that 
were created in the past to allow for variations in standards levels 
and timing appropriate for Tiers 1, 2 and 3, and that remain in effect 
for those tiers, but which under this proposal are no longer distinct 
from each other with respect to standards levels and timing. These 
consolidations are: (1) The less than 11 hp and 11-25 hp categories 
into a single category of less than 25 hp, (2) the 75-100 hp portion of 
the 50-100 hp category and the 100-175 hp category into a single 
category of 75-175 hp, and (3) the 175-300 hp, 300-600 hp, and 600-750 
hp categories into a single category of 175-750 hp. The result is the 5 
power bands shown in Figures III.B-1 and 2 instead of the former 9. 
This will also help to facilitate use of equipment manufacturer 
transition flexibility allowances which can be applied only within each 
power band, as discussed in section VII.B. We ask for comment on this 
regrouping, especially with regard to the appropriate power cutpoint 
for the engine families that are similar to highway engine families. 
Again, most useful in this regard would be information showing how 
highway and nonroad engines in this range do or do not share common 
design bases.
d. PM Standards for Smaller Engines
i. <25 hp
    We believe that standards based on the use of PM filters should not 
be proposed at this time for the very small diesel engines below 25 hp. 
Although this technology could be adapted to these engines, the cost of 
doing so with known technology could be unacceptably high, relative to 
the cost of producing the engines themselves. Based on past experience, 
we expect that advancements in reducing these costs will occur over 
time. We plan to reassess the appropriate long-term standards in a 
technology review as discussed in section III.G. For the nearer-term, 
we believe that other proven PM-reducing technologies such as diesel 
oxidation catalysts and engine optimization can be applied to engines 
under 25 hp for very cost-efficient PM control, as discussed in 
sections III.E and V.A. When implemented, the PM standard proposed in 
Figure III.B-1 for these engines, along with the proposed transient 
test cycle, will yield an in-use PM reduction of over 50% for these 
engines, and large reductions in toxic hydrocarbons as well. Achieving 
these emission reductions is very important, considering the fact that 
many of these smaller engines operate in populated areas and in 
equipment without closed cabs-- in mowers, portable electric power 
generators, small skid steer loaders, and the like. We invite comment 
on this proposed approach to controlling harmful emissions from very 
small nonroad diesel engines.
    It is our assessment that achieving low PM emission levels is 
especially challenging for one subclass of small engines: the air-
cooled, direct injection engines under 11 hp that are startable by 
hand, such as with a crank or recoil starter. These typically one-
cylinder engines find utility in applications such as plate compactors, 
where compactness and simplicity are needed, but where the ruggedness 
typical of a diesel engine is also essential. There are a number of 
considerations in the design, manufacture, and marketing of these 
engines that combine to make them difficult to optimize for low 
emissions. These include the air-cooled engine's need for relatively 
loose design fit tolerances to accommodate thermal expansion 
variability (which can lead to increased soluble organic PM), small 
cylinder displacement and bore sizes that limit use of some combustion 
chamber design strategies and increase the propensity for PM-producing 
fuel impingement on cylinder walls, the difficulty in obtaining 
components for small engines with machining tolerances tight enough to 
yield consistent emissions performance, and cost reduction pressures 
caused by competition from cheaper gasoline engines in some of the same 
applications.
    As a result, we are proposing an alternative compliance option that 
allows manufacturers of these engines to delay Tier 4 compliance until 
2010, and in that year to certify them to a PM standard of 0.45 g/hp-
hr, rather than to the 0.30 g/hp-hr PM standard applicable to the other 
engines in this power category beginning in 2008. Engines certified 
under this alternative compliance requirement would not be allowed to 
generate credits as part of the ABT program, although credit use by 
these engines would still be allowed. We believe that this ABT 
restriction is important to avoid potential abuse of this option, and 
is a reasonable means of dealing with the concern as it would apply 
only to those air-cooled, hand-startable, direct injection engines 
under 11 hp that are certified under this special compliance option, 
and the production of credit-generating engines would be contrary to 
the option's purpose. Furthermore, because the proposed 2010 Tier 4 
implementation year for these engines is the same year that 15 ppm 
sulfur nonroad diesel fuel would become available, we are also 
proposing that certification testing and any subsequent compliance 
testing on engines certified under this option may be conducted using 
the 7-15 ppm sulfur test fuel discussed in section VII.H. Although this 
is one year earlier than would be otherwise allowable, we believe it 
would have a minimal impact on the proposed program's environmental 
benefit considering the extremely small contribution these engines make 
to emissions inventories, and the fact that these engines would 
generally operate in the field on higher sulfur fuels for at most a few 
months.
ii. 25-75 hp
    We believe that the proposed 0.22 g/bhp-hr PM standard for 25-75 hp 
engines in 2008 is warranted because the Tier 2 PM standards that take 
effect in 2004 for these engines, 0.45 and 0.30 g/bhp-hr for 25-50 and 
50-75 hp engines, respectively, do not represent the maximum achievable 
reduction using technology which will be available by 2008. However, as 
discussed in section III.B.1.a, filter-based technology for these 
engines is not expected to be available on a widespread basis until the 
2013 model year. The proposed 2008 PM standard for these engines should 
maximize reduction of PM emissions based on technology available in 
that year. We believe that the 2008 standards are feasible for these 
engines, based on the same engine or oxidation catalyst technologies 
feasible for engines under 25 hp in 2008, following the proposed 
introduction of nonroad diesel fuel with sulfur levels reduced below 
500 ppm. We expect in-use PM reductions for these engines of over 50%, 
and large reductions in toxic hydrocarbons as well over the five model 
years this standard would be in effect (2008-2012). These engines will 
constitute a large portion of the in-use population of nonroad diesel 
engines for many years after 2008.
    We request comment on our proposal to implement Tier 4 PM standards 
for 25-75 hp engines in the two phases just noted: a non-PM filter 
based standard in 2008 and a filter-based standard in 2013. In 
addition, we request comment on whether it would be better not to set a 
Tier 4 PM standard in 2008 so that engine designers could instead focus

[[Page 28364]]

their efforts on meeting a PM-filter based standard for these engines 
earlier, say in 2012. (It should be noted that the proposed rule would 
provide this as an option for a subgroup of these engines (50-75 hp). 
See Figure III.B-1 note b.) We would assume that under this approach 
the proposed new NOX+NMHC standard for 25-50 hp engines in 
this category would also start in 2012, to avoid requiring two design 
changes in two years. Any comments in support of this approach should, 
if possible, include information to support a conclusion that the 
earlier start date for a PM filter-based standard would be 
technologically feasible.
    We believe that the proposed 2008 PM standards for engines under 75 
hp can be met either through engine optimization, by the use of diesel 
oxidation catalysts, or by some combination thereof, as discussed in 
section III.E. For engines that comply through the use of oxidation 
catalysts, NMHC emissions are expected to be very low because properly 
designed oxidation catalysts are effective at oxidizing gaseous 
hydrocarbons as well as the soluble organic fraction of diesel exhaust 
PM. Engines complying with the proposed 2008 PM standard without the 
use of oxidation catalysts would, on the other hand, be expected to 
emit NMHC at about the same levels as Tier 2 engines. Recognizing that 
NMHC emissions from diesel engines can include a number of toxic 
compounds, and that there are many of these small diesel engines 
operating in populated areas, we are interested in comment on the 
appropriateness of setting a more stringent NMHC standard for these 
engines in 2008 to better control these emissions. We expect that doing 
so would likely result in more widespread use of oxidation catalysts 
(rather than engine optimization) for these engines. We would not, 
however, expect this to lead to a more stringent PM standard than the 
one we are proposing, based on the feasibility discussion in section 
III.E.
e. Engines Above 750 hp
    For engines above 750 hp, additional lead time to fully implement 
Tier 4 is warranted due to the relatively long product design cycles 
typical of these high-cost, low-sales volume engines and machines. The 
long product design cycle issue is the primary reason we did not set 
Tier 3 standards for these engines in the 1998 rule and are not 
proposing to do so now. Instead, we are proposing that these engines 
move from the Tier 2 standards, which take effect in 2006, to Tier 4 
standards beginning in 2011, five years later. Moreover, we are 
proposing that the Tier 4 PM standard be phased in for these engines on 
the same 50-50-50-100% schedule as the NOX and NMHC phase-in 
schedule, rather than all at once in 2011 as for engines between 175 
and 750 hp. (See Figure III.B-1.) This would provide engine 
manufacturers with up to 8 years of design stability to address 
concerns associated with product design cycles and low sales volumes 
typical of this category. The engine manufacturer ABT program adds 
additional flexibility. Even longer stability periods could exist for 
equipment manufacturers using these engines because they have their own 
transition flexibility provisions available on top of the engine 
standard phase-in. This is especially significant because many of these 
large machines are built by manufacturers who build their own engines, 
or who work closely with their engine suppliers, and can thus create a 
long-term product plan making coordinated use of engine and equipment 
flexibility provisions.
    We think that, taken together, these provisions appropriately 
balance the need for expeditious emission reductions with issues 
relating to lead time, technology development, and cost for these 
engines and machines. Even so, some engine and equipment manufacturers 
have expressed concerns to us that, though not challenging the Tier 4 
program endpoint (high-efficiency PM and NOX exhaust 
emission controls), in their estimation our proposed program 
implementation provisions do not adequately address their timing 
concerns. In particular, they have expressed a view that they need 
until 2012 (one additional year) before they could begin to phase in 
Tier 4 standards for this category. They have also expressed the view 
that mobile machinery such as mine haul trucks and dozers (as 
differentiated from equipment such as nonroad diesel generators that 
also use engines in this hp range) present unique challenges that could 
require more time to resolve than would be afforded by the proposed 
2014 phase-in completion date.
    Although we believe that the implementation schedule and 
flexibility provisions we are proposing will enable the manufacturers 
to meet these challenges, we acknowledge the manufacturers' concerns 
and ask for comment on this issue. Specifically, we request comment on 
whether this category, or some subset of it defined by hp or 
application, should have a later phase-in start date, a later phase-in 
end date, adjusted standards, additional equipment manufacturer 
flexibility provisions, or some combination of these. Technical 
information backing the commenter's view would be most helpful in this 
regard.
    As with the NOX/NMHC phase-in for all engines at or 
above 75 hp, we are proposing that the PM phase-in for engines above 
750 hp would have to be met on the same engines as the Tier 4 
NOX and NMHC standards during the phase-in years. That is, 
engines certified to the Tier 4 NOX and NMHC requirements 
would be expected to certify to the Tier 4 PM standard as well.
f. CO Standards
    We are proposing minor changes in CO standards for some engines 
solely for the purpose of helping to consolidate power categories. 
These amount to a change for engines under 11 hp from 6.0 to 4.9 g/bhp-
hr in 2008 to match the existing Tier 2 CO standard for 11-25 hp 
engines, and a change for engines at or above 25 hp but below 50 hp 
from 4.1 to 3.7 g/bhp-hr to match the existing Tier 3 CO standard for 
50-75 hp engines, also in 2008. These minor proposed changes are not 
expected to add a notable compliance burden. Nevertheless, we expect 
that the use of high-efficiency exhaust emission controls will yield a 
substantial reduction in CO emissions, as discussed in Chapter 4 of the 
draft RIA.
    These minor adjustments to the CO standard are based solely on our 
desire to simplify the administrative process for the engine 
manufacturers which arises from the reduction in the number of the 
engine power categories we have proposed for Tier 4. We are not 
exercising our authority to revise the CO standard for nonroad diesel 
engines for the purpose of improving air quality at this time, and 
therefore the minor adjustments we have proposed today, though 
feasible, are not based on a detailed evaluation of the capabilities of 
advanced exhaust aftertreatment technology to reduce CO levels.
g. Exclusion of Marine Engines
    These proposed emission standards would apply to engines in the 
same applications covered by EPA's existing nonroad diesel engine 
standards, at 40 CFR part 89, except that they would not apply to 
marine diesel engines. Marine diesel engines below 50 hp were included 
in our 1998 rule that set nonroad diesel emission standards (63 FR 
56968, October 23, 1998). In that rule, we expected that the engine 
modifications needed to achieve those standards (e.g., in-cylinder 
controls) for marine engines would not need to be different from those 
for land-based engines of this size.

[[Page 28365]]

    The standards for diesel engines below 50 hp being proposed in this 
action are likely to require PM filters or diesel oxidation catalysts 
on many or all engines, and transferring this technology to the marine 
diesel engines of any size raises unique issues. For example, many 
marine diesel engines have water-jacketed exhaust which may result in 
different exhaust temperatures and which could affect aftertreatment 
efficiency. The modified marine engine designs would also have to meet 
Coast Guard requirements. These and other conditions may require 
separate design efforts for marine diesel engines. Therefore, we 
believe it is more appropriate to consider more stringent standards for 
marine diesel engines below 50 hp in a future action. It should be 
noted, however, that the existing Tier 2 standards will continue to 
apply to marine diesel engines under 50 hp until that future action is 
completed.
2. Crankcase Emissions Control
    Crankcase emissions are the pollutants that are emitted in the 
gases that are vented from an engine's crankcase. These gases are also 
referred to as ``blowby gases'' because they result from engine exhaust 
from the combustion chamber ``blowing by'' the piston rings into the 
crankcase. These gases are often vented to prevent high pressures from 
occurring in the crankcase. Our existing emission standards require 
control of crankcase emissions from all nonroad diesel engines except 
turbocharged engines. The most common way to eliminate crankcase 
emissions has been to vent the blowby gases into the engine air intake 
system, so that the gases can be recombusted. Following the precedent 
we set for heavy-duty highway diesel engines in an earlier rulemaking, 
we made the exception for turbocharged nonroad diesel engines because 
of concerns about fouling that could occur by routing the diesel 
particulates (including engine oil) into the turbocharger and 
aftercooler. Our concerns are now alleviated by newly developed closed 
crankcase filtration systems, specifically designed for turbocharged 
diesel engines. These new systems are already required in parts of 
Europe for new highway diesel engines under the EURO III emission 
standards, and are expected to be used in meeting new U.S. EPA 
crankcase emission control standards for heavy-duty highway diesel 
engines beginning in 2007 (see section III.C.1.c of the preamble to the 
2007 heavy-duty highway final rule).
    We are therefore proposing to eliminate the exception for 
turbocharged nonroad diesel engines starting in the same model year 
that Tier 4 exhaust emission standards first apply in each power 
category. This is 2008 for engines below 75 hp, except for 50-75 hp 
engines for which a manufacturer opts to skip the 2008 PM standard. The 
crankcase requirement applies to ``phase-in'' engines above 750 hp 
under the 50% phase-in requirement for 2011-2013, but not to the 
``phase-out'' engines in that power category during those years. This 
is an environmentally significant proposal since many nonroad machine 
models use turbocharged engines, and a single engine can emit over 100 
pounds of NOX, NMHC, and PM from the crankcase over the 
lifetime of the engine. We also note that the cost of control is small 
(see section V).
    Our existing regulatory requirement for controlling crankcase 
emissions from naturally-aspirated nonroad engines allows manufacturers 
to route the crankcase gases into the exhaust stream instead of the 
engine air intake system, provided they keep the combined total of the 
crankcase emissions and the exhaust emissions below the applicable 
exhaust emission standards. We are proposing to extend this allowance 
to the turbocharged engines as well. We are also proposing to give 
manufacturers the option to measure crankcase emissions instead of 
completely eliminating them, and adding the measured emissions to 
exhaust emissions in assessing compliance with exhaust emissions 
standards. This allowance was adopted for highway HDDEs in 2001 (see 
section VI.A.3 of the preamble to the 2007 heavy-duty highway final 
rule). As in the highway program, manufacturers choosing to use this 
allowance rather than to seal the crankcase would need to modify their 
exhaust deterioration factors or to develop separate deterioration 
factors to account for increases in crankcase emissions as the engine 
ages. Manufacturers would also be responsible for ensuring that 
crankcase emissions would be readily measurable in use.

C. What Test Procedure Changes Are Being Proposed?

    We are proposing a number of changes to the certification test 
procedures by which compliance with emission standards is determined. 
Two of these are particularly significant: The addition of a 
supplemental transient emissions test and the addition of a cold start 
testing component to the proposed transient emissions test. These are 
discussed briefly in this section, and in more detail in section VII.F. 
Other proposed changes are also discussed in section VII.F and deal 
with:
    [sbull] Adoption of an improved smoke testing procedure, with 
associated standards levels and exemptions.
    [sbull] Addition of a steady-state test cycle for transportation 
refrigeration units.
    [sbull] Test procedure changes intended to improve testing 
precision, especially with regards to sampling methods.
    [sbull] A clarification to existing EPA defeat device regulations.
1. Supplemental Transient Test
    In the 1998 final rule that set new emission standards for nonroad 
diesel engines, we expressed a concern that the steady-state test 
cycles used to demonstrate compliance with emission standards did not 
adequately reflect transient operation, and, because most nonroad 
engines are used in applications that are largely transient in nature, 
would therefore not yield adequate control in use (63 FR 56984, October 
23, 1998). Although we were not prepared to adopt a transient test at 
that time, we announced our intention in that final rule to move 
forward with the development of such a test. This development has 
progressed steadily since that time, and has resulted in the creation 
of a Nonroad Transient Composite (NRTC) test cycle, which we are now 
proposing to adopt in our nonroad diesel program, to supplement the 
existing steady-state tests. We expect that this proposed requirement 
will significantly reduce real world emissions from nonroad diesel 
equipment. Instead of sampling engine operation at the few isolated 
operating points of steady-state emission tests, proper transient 
testing can capture emissions from the broad range of engine speed and 
load combinations that the engine may attain in use, as well as 
emissions resulting from the change in speed or load itself, such as 
those induced by turbocharger lag.
    The proposed NRTC cycle will capture transient emissions over much 
of the typical nonroad engine operating range, and thus help ensure 
effective control of all regulated pollutants. In keeping with our goal 
to maximize the harmonization of emissions control programs as much as 
possible, we have developed this cycle in collaboration with nonroad 
engine manufacturers and regulatory bodies in the United States, 
Europe, and Japan over the last several years.\123\ Further, the NRTC 
cycle has been introduced as a work item for

[[Page 28366]]

possible adoption as a potential global technical regulation under the 
1998 Agreement for Working Party 29 at the United Nations.\124\
---------------------------------------------------------------------------

    \123\ Letter from Jed Mandel of the Engine Manufacturers 
Association to Chet France of U.S. EPA, Office of Transportation and 
Air Quality, Docket A-2001-28.
    \124\ Informal Document No. 2, ISO--45th GRPE, ``Proposal for a 
Charter for the Working Group on a New Test Protocol for Exhaust 
Emissions from Nonroad Mobile Machinery,'' 13-17 January 2003, 
Docket A-2001-28.
---------------------------------------------------------------------------

    The Agency is proposing that emission standards be met on both the 
current steady-state duty cycles and the new transient duty cycles. The 
transient testing would begin in the model year that the trap-based 
Tier 4 PM standards and/or adsorber-based Tier 4 NOX 
standards first apply. This would be 2011 for engines at or above 175 
hp, 2012 for 75-175 hp engines (2012 for 50-75 hp engines made by a 
manufacturer choosing the optional approach described in footnote b of 
Figure III.B-1), and 2013 for engines under 75 hp. See also Table 
VII.F.-1. In addition, any engines for which a manufacturer claims 
credit under the incentive program for early-introduction engines (see 
section VII.E) would have to be certified to that program's standards 
under the NRTC cycle and, in turn, the 2011 or later model year engines 
that use these engine count-based credits would not need to demonstrate 
compliance under the NRTC cycle.
    Although we intend that transient emissions control be an integral 
part of Tier 4 design considerations, we do not believe it appropriate 
to mandate compliance with the transient test for the engines under 75 
hp subject to proposed PM standards in 2008. We recognize that 
transient emissions testing, though routine in highway engine programs, 
involves a fair amount of new laboratory equipment and expertise in the 
nonroad engine certification process. As with the transfer of advanced 
emission control technology itself, we believe that the transient test 
requirement should be implemented first for larger engines more likely 
to be made by engine manufacturers who also have highway engine 
markets. We do not believe that the smaller engines should be the lead 
power categories in implementing the new transient test, especially 
because many manufacturers of these engines do not make highway engines 
and are not as experienced or well-equipped as their large-engine 
counterparts for conducting transient cycle testing.
    Engines below 25 hp involve an additional consideration for timing 
of the transient test requirement because we are not proposing PM-
filter based standards for them. We propose that testing on the NRTC 
cycle not be required for these engines until the 2013 model year, the 
last year in which engines in higher power categories are required to 
use this test. We are concerned that manufacturers not view this 
proposed deferral of the transient test requirement as a structured 
second level of required control for these engines. To address this 
concern and because we wish to encourage the demonstration of transient 
emission control as early as possible, we are proposing to allow 
manufacturers to optionally certify engines below 25 hp under the NRTC 
cycle beginning in the 2008 model year, and to extend this option to 
25-75 hp engines subject to engines meeting the transitional PM 
standard in 2008. (See also the discussion in section VII.F.1 on this 
issue.) We request comment on this proposed approach and on whether it 
would be better to deal with this concern by requiring compliance under 
the transient test when the Tier 4 standards begin in 2008.
    In applying the NRTC test requirement coincident with the start of 
PM filter-based standards, we do not mean to imply that control of PM 
from filter-equipped engines is the only or even the primary concern 
being addressed by transient testing. In fact, we believe that advanced 
NOX emission controls may be more sensitive to transient 
operation than PM filters. It is, however, our intent that the control 
of emissions during transient operation be an integral part of Tier 4 
engine design considerations, and we therefore have proposed that 
transient testing be applied with the PM filter-based Tier 4 PM 
standards, because these standards precede or accompany the earliest 
Tier 4 NOX or NMHC standards in every power category. Even 
so, we request comment on whether the ``phase-out'' engines above 75 hp 
(those engines for which compliance with the Tier 4 NOX 
standard is not required during the phase-in period) should be exempted 
from the requirement to meet the applicable NMHC+NOX 
standard using the transient test. Although our interest in ensuring 
transient emissions control as quickly as possible in the Tier 4 
program, and in avoiding test program complexity, would argue against 
this approach, we are also interested in not diverting engine designers 
from the challenging task of redesigning engines to meet the proposed 
0.30 g/bhp-hr Tier 4 NOX standard before and during the 
phase-in years by having to deal with transient control under an 
NMHC+NOX standard that is being phased out.
    We are in fact not proposing to apply the transient test to phase-
out engines above 750 hp that are carried over from pre-2011 Tier 2 
engine designs. Unlike phase-out engines at or below 750 hp, these 
engines are not subject to a Tier 4 PM standard in 2011. They would 
thus be Tier 2 engine designs and we do not believe that subjecting 
them to transient testing would be appropriate. On the other hand, 
engines in any power category certified to an average NOX 
standard under the ``split family'' provision described in section 
VII.A would all be subject to the transient test requirement, as they 
would clearly have to be substantially redesigned to achieve Tier 4 
compliance, regardless of whether or not they use high-efficiency 
exhaust emission controls.
    The Agency is proposing that engine manufacturers may certify 
constant-speed engines using EPA's Constant Speed Variable Load (CSVL) 
transient duty cycle \125\ as an alternative to testing these engines 
under the NRTC provisions. The CSVL transient cycle more closely 
matches the speed and load operating characteristics of many constant-
speed nonroad diesel applications than EPA's proposed NRTC cycle.\126\ 
However, the manufacturer would be obligated to ensure that such 
engines would be used only in constant-speed applications. A more 
detailed discussion of the proposed NRTC and CSVL supplemental 
transient test cycles and associated provisions is contained in section 
VII.F of this preamble and in chapter 4 of the Draft RIA.
---------------------------------------------------------------------------

    \125\ Memoranda from Kent Helmer to Cleophas Jackson, ``Speed 
and Load Operating Schedule for the Constant Speed Variable Load 
(CSVL) transient test cycle'' and ``CSVL Cycle Construction''; and 
Southwest Research Institute--Final Report, all in Docket A-2001-28.
    \126\ Memorandum from Kent Helmer to Cleophas Jackson, ``Brake-
specific Emissions Impact of Nonroad Diesel Engine Testing Over the 
NRTC, AWQ, and AW1 duty cycles'', Docket A-2001-28.
---------------------------------------------------------------------------

2. Cold Start Testing
    In the field, the typical nonroad diesel machine will be started 
and will warm to a point of heat-stable operation at least once a 
workday. Such ``cold start'' conditions may also occur at other times 
over the course of the workday, after a lunch break for example. During 
these periods of cold start operation, the engine may be emitting at a 
higher rate than when the engine is running efficiently at its 
stabilized operating temperature. This may be especially the case for 
emission control designs employing catalytic devices in the exhaust 
system, which require heating to a ``light-off'' temperature to begin 
working. EPA's highway engine and vehicle programs, which have resulted 
in increasingly widespread use of such catalytic devices, have 
recognized and dealt with this concern for several years,

[[Page 28367]]

typically by repeating transient tests in both the ``cold'' and ``hot'' 
conditions, and weighting emission results in some fashion to create a 
combined result for evaluation against emission standards.
    We believe that our proposed move to supplemental transient 
testing, combined with our proposed Tier 4 standards that will bring 
about the use of catalytic devices in nonroad diesel engines, makes it 
imperative that we also propose to include such a cold start test as 
part of the transient test procedure requirement. We propose to weight 
the cold start emission test results as one-tenth of the total with 
hot-start emissions accounting for the other nine-tenths. The one-tenth 
weighting factor is derived from a review of the present nonroad 
equipment population. For more detailed information on this proposal, 
refer to section VII.F of this preamble and chapter 4 of the Draft RIA. 
EPA requests comment on this approach to ensuring control of cold start 
emissions.

D. What Is Being Done To Help Ensure Robust Control in Use?

    EPA's goal is to ensure real-world emissions control over the broad 
range of in-use operation that can occur, rather than just controlling 
emissions over prescribed test cycles executed under restricted 
laboratory conditions. An important tool for achieving this in-use 
emissions control is the setting of Not-To-Exceed (NTE) emission 
standards, which, in this notice, the Agency is proposing to adopt for 
new nonroad engines. EPA is also considering two additional means of 
in-use emissions control that will be proposed in separate notices. 
These are (1) a manufacturer-run in-use emissions test program and (2) 
on-board diagnostics (OBD) requirements for new nonroad diesel engines. 
When implemented, all three of these will help assure that in-use 
emissions control is achieved.
1. Not-to-Exceed Requirements
    EPA proposes to adopt not-to-exceed (NTE) emission standards for 
all new nonroad diesel engines subject to the Tier 4 emissions 
standards beginning in 2011 proposed in section III. B. of this 
proposal. EPA already has similar NTE standards set for highway heavy-
duty diesel engines, compression ignition marine engines, and nonroad 
spark-ignition engines.
    To help ensure that nonroad diesel emissions are controlled over 
the wide range of speed and load combinations commonly experienced in-
use, EPA is proposing to apply NTE limits and related test procedures. 
The NTE approach establishes an area (the ``NTE zone'') under the 
torque curve of an engine where emissions must not exceed a specified 
value for any of the regulated pollutants. The NTE standard would apply 
under any conditions that could reasonably be expected to be seen by 
that engine in normal vehicle operation and use, within certain broad 
ranges of real ambient conditions. The NTE requirements would help to 
ensure emission benefits over the full range of in-use operating 
conditions. EPA believes that basing the emissions standards on a set 
of distinct steady state and transient cycles and using the NTE zone to 
help ensure in-use control creates a comprehensive program. In 
addition, the NTE requirements would also be an effective element of an 
in-use testing program. The test procedure is very flexible so it can 
represent most in-use operation and ambient conditions. Therefore, the 
NTE approach takes all of the benefits of a numerical standard and test 
procedure and expands it to cover a broad range of conditions. Also, 
with the NTE approach, in-use testing and compliance become much easier 
since emissions may be sampled during normal vehicle use. A standard 
that relies on laboratory testing over a very specific driving schedule 
makes it harder to perform in-use testing, especially for engines, 
since the engines would have to be removed from the vehicle. Testing 
during normal vehicle use, using an objective numerical standard, makes 
enforcement easier and provides more certainty of what is occurring in 
use versus a fixed laboratory procedure.
    In today's notice, we are proposing an NTE standard which is based 
on the approach taken for the 2007 highway heavy-duty diesel engines. 
In addition, we are requesting comment on an alternative NTE standard 
approach which, while different from the highway NTE standard approach, 
is designed to achieve the same environmental objectives. Both of these 
approaches are described below.
a. NTE Standards We Are Proposing
    The Agency proposes to adopt for new Tier 4 non-road diesel engines 
similar NTE specifications as those finalized as part of the heavy-duty 
highway diesel engine rulemaking (See 66 FR 5001, January 18, 2001). 
These specifications for the highway diesel engines are contained in 40 
CFR part 86.007-11 and 40 CFR part 86.1370-2007.
    Our NTE proposal for nonroad contains the same basic provisions as 
the highway NTE. The proposed nonroad NTE standard establishes an area 
(the ``NTE control area'') under the torque curve of an engine where 
emissions must not exceed a specified value for any of the regulated 
pollutants.\127\ This NTE control area is defined in the same manner as 
the highway NTE control areas, and is therefore a subset of the 
engine's possible speed and load operating range. The NTE standard 
would apply under any engine operating conditions that could reasonably 
be expected to be seen by that engine in normal vehicle/equipment 
operation and use which occurs within the NTE control zone and which 
also occurs during the wide range of real ambient conditions specified 
for the NTE. The NTE standard applies to emissions sampled during a 
time duration as small as 30 seconds. The NTE standard requirements for 
nonroad diesel engines are summarized below and specified in the 
proposed regulations at 40 CFR 1309.101 and 40 CFR 1039.515. These 
requirements would take effect as early as 2011, as shown in shown in 
Table III.D-1. The NTE standard would apply to engines at the time of 
certification as well as in use throughout the useful life of the 
engine.
---------------------------------------------------------------------------

    \127\ Torque is a measure of rotational force. The torque curve 
for an engine is determined by an engine ``mapping'' procedure 
specified in the Code of Federal Regulations. The intent of the 
mapping procedure is to determine the maximum available torque at 
all engine speeds. The torque curve is merely a graphical 
representation of the maximum torque across all engine speeds.

          Table III.D-1.--NTE Standard Implementation Schedule
------------------------------------------------------------------------
                                                               NTE
                    Power category                       Implementation
                                                         model year \a\
------------------------------------------------------------------------
<25 hp................................................              2013
25-75 hp..............................................          \b\ 2013

[[Page 28368]]

 
75-175 hp.............................................              2012
175-750 hp............................................              2011
750 hp.....................................         \c\ 2011
------------------------------------------------------------------------
Notes:
\a\ The NTE applies for each power category once Tier 4 standards were
  implemented, such that all engines in a given power category are
  required to meet NTE standards.
\b\ The NTE standard would apply in 2012 for any engines in the 50-75 hp
  range who choose not to comply with the proposed 2008 transitional PM
  standard.
\c\ The NTE standard only applies to the 50 percent of the engines in
  the 750 hp category which are complying with the proposed
  Tier 4 standard. Beginning in 2014 the NTE standard would apply to all
  nonroad engines 750 hp when the remaining 50 percent of the
  engines must comply with the Tier 4 standard.

    The NTE test procedure can be run in nonroad equipment during field 
operation or in an emissions testing laboratory using an appropriate 
dynamometer. The test itself does not involve a specific operating 
cycle of any specific length, rather it involves nonroad equipment 
operation of any type which could reasonably be expected to occur in 
normal nonroad equipment operation that could occur within the bounds 
of the NTE control area. The nonroad equipment (or engine) is operated 
under conditions that may reasonably be expected to be encountered in 
normal vehicle operation and use, including operation under steady-
state or transient conditions and under varying ambient conditions. 
Emissions are averaged over a minimum time of thirty seconds and then 
compared to the applicable emission standard. The NTE standard applies 
over a wide range of ambient conditions, including up to an altitude of 
5,500 feet above-sea level at ambient temperatures as high as 86 deg. 
F, and at sea-level up to ambient temperatures as high as 100 deg. F. 
The specific temperature and altitude conditions under which the NTE 
applies, as well as the proposed methodology for correcting emissions 
results for temperature and/or humidity are specified in the proposed 
regulations.
    In addition, as with the 2007 highway NTE standard, we are 
proposing a transition period during which a manufacturer could apply 
for an NTE deficiency for a nonroad diesel engine family. The NTE 
deficiency provisions would allow the Administrator to accept a nonroad 
diesel engine as compliant with the NTE standards even though some 
specific requirements are not fully met. We are proposing these NTE 
deficiency provisions because we believe that, despite the best efforts 
of manufacturers, for the first few model years it is possible some 
manufacturers may have technical problems that are limited in nature 
but can not be remedied in time to meet production schedules. We are 
not limiting the number of NTE deficiencies a manufacturer can apply 
for during the first 3 model years for which the NTE applies. For the 
fourth through the seventh model year after which the NTE standards are 
implemented, a manufacturer could apply for no more than three NTE 
deficiencies per engine family. No deficiency may be applied for or 
granted after the seventh model year. The NTE deficiency provision will 
only be considered for failures to meet the NTE requirements. EPA will 
not consider an application for a deficiency for failure to meet the 
FTP or supplemental transient standards.
    The NTE standards we are proposing are a function of FTP emission 
standards contained in this proposal and described in section III.B. As 
with the NTE standards we have established for the 2007 highway rule, 
we are proposing an NTE standard which is determined as a multiple of 
the engine families underlying FTP emission standard. In addition, as 
with the 2007 highway standard, the multiple is either 1.25 or 1.5, 
depending on the value of the FTP standard (or the engine families 
FEL). These multipliers are based on EPA's assessment of the 
technological feasibility of the NTE standard, and our assessment that 
as the underlying FTP standard becomes more stringent, the NTE 
multiplier should increase (from 1.25 to 1.5). The proposed standard or 
FEL thresholds for the 1.25x multiplier and the 1.5x multiplier are 
specified for each regulated emission in Table III.D-2.

         Table III.D-2.--Thresholds for Applying NTE Standard of 1.25xFTP Standard vs. 1.5x FTP Standard
----------------------------------------------------------------------------------------------------------------
             Emission              Apply 1.25xNTE when . . .               Apply 1.5xNTE when . . .
----------------------------------------------------------------------------------------------------------------
NOX..............................  NOX std or FEL >=1.5 g/    NOX std or FEL <1.5 g/bhp-hr
                                    bhp-hr.
NMHC.............................  NOX std or FEL >=1.5 g/    NOX std or FEL <1.5 g/bhp-hr
                                    bhp-hr.
NOX+NMHC.........................  NMHC+NOX std or FEL >=1.6  NMHC+NOX std or FEL <1.6 g/bhp-hr
                                    g/bhp-hr.
PM....................  PM std or FEL =0.05 g/bhp-hr.
CO...............................  All stds or FELs.........  No stds or FELs
----------------------------------------------------------------------------------------------------------------

    For example, beginning in 2011, the proposed NTE standard for 
engines meeting a FTP PM standard of 0.01 g/bhp-hr and a FTP 
NOX standard of 0.30 g/bhp-hr would be 0.02 g/bhp-hr PM and 
0.45 g/bhp-hr NOX.
    In addition, the nonroad NTE proposal specifies a number of 
additional engine operating conditions which are not subject to the NTE 
standard. Specifically: The NTE does not apply during engine start-up 
conditions; the NTE does not apply during very cold engine intake 
conditions defined in the proposed regulations for EGR equipped engines 
during which the engine may require an engine protection strategy; and, 
finally, for engines equipped with an exhaust emission control device 
(such as a CDPF or a NOX adsorber), the NTE does not apply 
during warm-up conditions for the exhaust emission control device, 
specifically the NTE does not apply

[[Page 28369]]

with the exhaust gas temperature on the outlet side of the exhaust 
emission control device is less than 250 degrees Celsius.
b. Comment Request on an Alternative NTE Approach
    In addition the Agency requests comment on the following set of NTE 
specifications as an alternative to those NTE provisions proposed. This 
alternative NTE would use the same numeric standard values as under the 
proposed NTE standards discussed in section III.D.1a, however, the test 
procedure itself is quite different, as described below. The Agency 
believes that these alternative specifications and the range of 
operation covered by the standard would provide for similar, if not 
more robust nonroad engine compliance compared to the application of 
the proposed NTE specifications to nonroad engines. These alternative 
provisions have been developed to emphasize compliance over all engine 
operation, including engine operation that would not be covered under 
the proposed NTE approach. In addition these specifications were 
developed specifically to simplify on-vehicle testing for NTE 
compliance. The NTE control area would include all engine operation. 
The averaging intervals over which NTE standards must be met are 
different than the 30-second minimum set in the proposal. They are 
variable in time but are constant as a function of work. Emissions 
would be measured over a constant averaging work interval, determined 
as ten percent (10%) of the total work performed by the engine over a 
specified period of time (e.g., a minimum of six hours of operation). 
This 10% window of work ``moves'' through data at one percent (1%) 
increments so as to always return about ninety (90) individual data 
points for direct comparison to the NTE standards.
    Comments should address the potential exclusive use of these 
alternative provisions for nonroad diesel engine NTE compliance. For 
more detailed information on these alternative NTE provisions, refer to 
Preamble section VIIG ``Not-to-Exceed Requirements'' and chapter 4 of 
the draft RIA of this proposal.
2. Plans for a Future In-Use Testing and Onboard Diagnostics
    In addition to the proposals in this notice, EPA is currently 
reviewing several related regulatory provisions concerning control of 
emissions from nonroad diesel engines. They are not included in this 
proposal, as EPA believes these aspects of an effective emission 
control program would benefit from further evaluation and development 
prior to their proposal. EPA intends to explore these provisions 
further in the coming months and publish a separate notice of proposed 
rulemaking dealing with these issues. In particular, there are two 
issues which will be discussed: (1) A manufacturer-run in-use emissions 
testing program; and (2) OBD requirements for nonroad diesel engines. 
The Agency believes that it is appropriate to proceed with the current 
rulemaking, expecting that these two issues will be proposed in the 
near future. EPA expects these programs would be adopted in advance of 
the effective date of the engine emissions standards. This will allow 
us to gather information and work with interested parties in a separate 
process regarding these issues. EPA will work with all parties 
involved, including states, environmental organizations and 
manufacturers, to develop robust, creative, environmentally protective 
and cost-effective proposals addressing these issues.
a. Plans for a Future Manufacturer-Run In-Use Test Program
    It is critical that nonroad diesel engines meet the applicable 
emission standards throughout their useful lives, to sustain those 
emission benefits over the broadest range of in-use operating 
conditions. The Agency believes that a manufacturer-run in-use testing 
program that is designed to generate data on in-use emissions of 
nonroad diesel engines can be used by EPA and the engine manufacturers 
to ensure that emissions standards are met throughout the useful life 
of the engines, under conditions normally experienced in-use. An 
effective program can be designed to monitor for NTE compliance and to 
help ensure overall compliance with emission standards.
    The Agency expects to pattern the manufacturer-run in-use testing 
requirements for nonroad diesel engines after a program that is being 
developed for heavy-duty highway vehicles. In this latter program, EPA 
is committed to incorporating a two-year pilot program. The pilot 
program will allow the Agency and manufacturers to gain the necessary 
experience with the in-use testing protocols and generation of in-use 
test data using portable emission measurement devices prior to fully 
implementing program. A similar pilot program is expected to be part of 
any manufacturer-run in-use NTE test program for nonroad engines.
    The Agency plans to promulgate the in-use testing requirements for 
heavy-duty highway vehicles in the December 2004 time frame. EPA 
anticipates proposing a manufacturer-run in-use testing program for 
nonroad diesel engines by 2005 or earlier. As mentioned above, the 
nonroad diesel engine program is expected to be patterned after the 
heavy-duty highway program.
b. Onboard Diagnostics
    Today's notice does not propose to require onboard diagnostic (OBD) 
systems for non-road diesel vehicles and engines. However, EPA has 
committed to creating OBD requirements for heavy-duty highway engines/
vehicles over 14,000 lbs GVWR and will develop OBD requirements for 
nonoad in conjunction with or following the highway OBD development. 
The Agency will propose nonroad diesel OBD requirements, along with 
heavy-duty highway OBD requirements, because OBD is necessary for 
maintaining and ensuring compliance with emission standards over the 
lifetime of engines. We will gather further information and coordinate 
with the heavy-duty highway and nonroad diesel industry and other 
stakeholders to develop proposed OBD system requirements.

E. Are the Proposed New Standards Feasible?

    Prior to 1990, diesel engines could be broadly grouped into two 
categories; indirect-injection (IDI) diesel engines that were 
relatively inexpensive while providing somewhat better fuel economy 
compared to gasoline engines, and direct-injection (DI) diesel engines 
that were substantially more expensive but which offered better fuel 
economy. The majority of diesel engines fell into the first category, 
especially in the case of passenger cars, smaller heavy-duty trucks and 
most nonroad engines below 200 horsepower.
    Diesel engine technology has changed rapidly since the early 1990s 
with the widespread use of electronics, onboard computers and the rise 
to preeminence of turbocharged direct-injection diesel engines. While 
some IDI engines remain, especially in the low horsepower portion of 
the nonroad market, most new diesel engines (including higher 
horsepower nonroad diesel engines) are turbocharged and direct-
injected. Today's diesel engine has significantly improved, compared to 
historic engines with regard to issues of most concern to the user 
including noise, vibration, visible smoke emissions, startability, and 
performance. At the same time environmental benefits have also been 
realized with lower NOX emissions, lower PM emissions, and 
improving fuel economy. These changes have been most pronounced for 
smaller

[[Page 28370]]

diesel engines applied in passenger cars and light-heavy trucks. 
Acceptance of the technology by the public, especially in Europe, has 
lead to a rapid increase in diesel use for smaller vehicles with diesel 
sales for passenger cars exceeding 50 percent in some countries.
    At the end of the 1990s continuing concern regarding the serious 
risk to public health and welfare from diesel emissions and the 
emergence of new emission control technologies enabled by low sulfur 
fuels led policy makers to set new future diesel fuel specifications 
and to set challenging new diesel emission standards for highway 
vehicles. In the United States, the EPA has set stringent new diesel 
emission standards for heavy-duty highway engines which will go into 
effect in 2007. These new standards are predicated on the use of 
Catalyzed Diesel Particulate Filters (CDPFs) which when used with less 
than 15ppm sulfur diesel fuel can reduce PM emissions by well over 90%, 
and on the use of NOX adsorber catalyst technology which 
when used with less than 15 ppm diesel fuel can reduce NOX 
emissions by more than 90%. When these technologies are fully 
implemented, the resulting diesel engine emissions will be 98% lower 
than the levels common to these diesel engines before 1990.
    EPA has been conducting an ongoing technology progress review to 
measure industry progress to develop and introduce the needed clean 
fuel and clean engine technologies by 2007. The first in what will be a 
series of reports was published by EPA in June of 2002.\128\ In the 
report, we concluded that technology developments by industry were 
progressing rapidly and that the necessary catalyzed diesel particulate 
filter and NOX adsorber technologies would be available for 
use by 2007.
---------------------------------------------------------------------------

    \128\ Highway Diesel Progress Review, United States 
Environmental Protection Agency, June 2002, EPA 420-R-02-016. Copy 
available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    Nonroad diesel engines are fundamentally similar to highway diesel 
engines. As noted above in section III.B, in many cases, virtually 
identical engines are certified and sold for use in highway vehicles 
and nonroad equipment. Thus, emission control technologies developed 
for diesel engines can in general be applied to both highway and 
nonroad engines giving appropriate considerations to unique aspects of 
each application.
    Today, we are proposing to set stringent new standards for a broad 
category of nonroad diesel engines. At the same time we are proposing 
to dramatically lower the sulfur level in nonroad diesel fuel 
ultimately to 15 ppm. We believe these standards are feasible given the 
availability of the clean 15 ppm sulfur fuel and the rapid progress to 
develop the needed emission control technologies. We acknowledge that 
these standards will be challenging for industry to meet in part due to 
differences in operating conditions and duty cycles for nonroad diesel 
engines. Also, we recognize that transferring and effectively applying 
these technologies, which have largely been developed for highway 
engines, will require additional lead time. We have given consideration 
to these issues in determining the appropriate timing and emission 
levels for the standards proposed today.
    The following sections will discuss how these technologies work, 
issues specific to the application of these technologies to new nonroad 
engines, and why we believe that the emission standards proposed here 
are feasible. A more in-depth discussion of these technologies can be 
found in the draft RIA associated with this proposal, in the final RIA 
for the HD2007 emission standards and in the recently completed 2002 
Highway Diesel Progress Review.\129\ The following discussion 
summarizes the more detailed discussion found in the Draft RIA.
---------------------------------------------------------------------------

    \129\ Highway Diesel Progress Review, United States 
Environmental Protection Agency, June 2002, EPA 420-R-02-016. Copy 
available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

1. Technologies To Control NOX and PM Emissions From Mobile 
Source Diesel Engines
    Present mobile source rules control the emissions of non-methane 
hydrocarbons (NMHC), oxides of nitrogen (NOX), carbon 
monoxide (CO), air toxics and particulate matter (PM) from diesel 
engines. Of these, PM and NOX emissions are typically the 
most difficult to control. CO and NMHC emissions are inherently low 
from diesel engines and under most conditions can be controlled to low 
levels without difficulty. NMHC emissions also serve as a proxy for 
some of the air toxic emissions from these engines, since many air 
toxics are a component of NMHC and are typically reduced in proportion 
to NMHC reductions. Most diesel engine emission control technologies 
are designed to reduce PM and NOX emissions without 
increasing CO and NMHC emissions above the already low diesel levels. 
Technologies to control PM and NOX emissions are described 
below separately. We also discuss the potential for these technologies 
to decrease CO and NMHC emissions as well as their potential to reduce 
emissions of air toxics.
a. PM Control Technologies
    Particulate matter from diesel engines is made of three components;

[sbull] Solid carbon soot,
[sbull] Volatile and semi-volatile organic matter, and
[sbull] Sulfate.

The formation of the solid carbon soot portion of PM is inherent in 
diesel engines due to the heterogenous distribution of fuel and air in 
a diesel combustion system. Diesel combustion is designed to allow for 
overall lean (excess oxygen) combustion giving good efficiencies and 
low CO and HC emissions with a small region of rich (excess fuel) 
combustion within the fuel injection plume. It is within this excess 
fuel region of the combustion that PM is formed when high temperatures 
and a lack of oxygen cause the fuel to pyrolize, forming soot. Much of 
the soot formed in the engine is burned during the combustion process 
as the soot is mixed with oxygen in the cylinder at high temperatures. 
Any soot that is not fully burned before the exhaust valve is opened 
will be emitted form the engine as diesel PM.
    The soot portion of PM emissions can be reduced by increasing the 
availability of oxygen within the cylinder for soot oxidation during 
combustion. Oxygen can be made more available by either increasing the 
oxygen content in-cylinder or by increasing the mixing of the fuel and 
oxygen in-cylinder. A number of technologies exist that can influence 
oxygen content and in-cylinder mixing including, improved fuel 
injection systems, air management systems, and combustion system 
designs.\130\ Many of these PM reducing technologies offer better 
control of combustion in general, and better utilization of fuel 
allowing for

[[Page 28371]]

improvements in fuel efficiency concurrent with reductions in PM 
emissions. Improvements in combustion technologies and refinements of 
these systems is an ongoing effort for highway engines and for some 
nonroad engines where emission standards or high fuel use encourage 
their introduction. The application of better combustion system 
technologies across the broad range of nonroad engines in order to meet 
the new emission standards proposed here offers an opportunity for 
significant reductions in engine-out PM emissions and possibly for 
reductions in fuel consumption. The soot portion of PM can be reduced 
further with aftertreatment technologies as discussed later in this 
section.
---------------------------------------------------------------------------

    \130\ The most effective means to reduce the soot portion of 
diesel PM engine-out is to operate the diesel engine with a 
homogenous method of operation rather than the typical heterogenous 
operation. In homogenous combustion, also called premixed 
combustion, the fuel is dispersed evenly with the air throughout the 
combustion system. This means there are no fuel rich/oxygen deprived 
regions of the system where fuel can be pyrolized rather than 
burned. Gasoline engines are typically premixed combustion engines. 
Homogenous combustion is possible with a diesel engine under certain 
circumstances, and is used in limited portions of engine operation 
by some engine manufacturers. Unfortunately, homogenous diesel 
combustion is not possible for most operation in today's diesel 
engine. We believe that more manufacturers will utilize this means 
to control diesel emissions within the limitations of the 
technology. A more in-depth discussion of homogenous diesel 
combustion can be found in the draft RIA.
---------------------------------------------------------------------------

    The volatile and semi-volatile organic material in diesel PM is 
often simply referred to as the soluble organic fraction (SOF) in 
reference to a test method used to measure its level. SOF is primarily 
composed of engine oil which passes through the engine with no or only 
partial oxidation and which condenses in the atmosphere to form PM. The 
SOF portion of diesel PM can be reduced through reductions in engine 
oil consumption and through oxidation of the SOF catalytically in the 
exhaust.
    The sulfate portion of diesel PM is formed from sulfur present in 
diesel fuel and engine lubricating oil that oxidizes to form sulfuric 
acid (H2SO4) and then condenses in the atmosphere 
to form sulfate PM. Approximately two percent of the sulfur that enters 
a diesel engine from the fuel is emitted directly from the engine as 
sulfate PM.\131\ The balance of the sulfur content is emitted from the 
engine as SO2. Oxidation catalyst technologies applied to 
control the SOF and soot portions of diesel PM can inadvertently 
oxidize SO2 in the exhaust to form sulfate PM. The oxidation 
of SO2 by oxidation catalysts to form sulfate PM is often 
called sulfate make. Without low sulfur diesel fuel, oxidation catalyst 
technology to control diesel PM is limited by the formation of sulfate 
PM in the exhaust as discussed in more detail in Section III.F below.
---------------------------------------------------------------------------

    \131\ Exhaust and Crankcase Emission Factors for Nonroad Engine 
Modeling--Compression-Ignition, EPA420-P-02-016, NR-009B. Copy 
available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    There are two common forms of exhaust aftertreatment designed to 
reduce diesel PM, the diesel oxidation catalyst (DOC) and the diesel 
particulate filter (DPF). DOCs reduce diesel PM by oxidizing a small 
fraction of the soot emissions and a significant portion of the SOF 
emissions. Total DOC effectiveness to reduce PM emissions is normally 
limited to approximately 30 percent because the SOF portion of diesel 
PM for modern diesel engines is typically less than 30 percent and 
because the DOC increases sulfate emissions reducing the overall 
effectiveness of the catalyst. Limiting fuel sulfur levels to 15 ppm, 
as we have proposed today, allows DOCs to be designed for maximum 
effectiveness (nearly 100% control of SOF with highly active catalyst 
technologies) since their control effectiveness is not reduced by 
sulfate make (i.e., there sulfate make rate is high but because the 
sulfur level in the fuel is low the resulting PM emissions are well 
controlled). A reduction in diesel fuel sulfur to 500 ppm as we are 
proposing today, is also directionally helpful for the application of 
DOCs. While 500 ppm sulfur fuel will not make the full range of highly 
active catalyst technologies available to manufacturers, it will 
decrease the amount of sulfate make and may allow for slightly more 
active (i.e., effective) catalysts to be used. We believe that this is 
an additional benefit of the proposed 500 ppm sulfur fuel program. DOCs 
are also very effective at reducing the air toxic emissions from diesel 
engines. Test data shows that emissions of toxics such as polycyclic 
aromatic hydrocarbons (PAHs) can be reduced by more than 80 percent 
with a DOC.\132\ DOCs also significantly reduce (by more than 80 
percent) the already low HC and CO emissions of diesel engines.\133\ 
DOCs are ineffective at controlling the solid carbon soot portion of 
PM. Therefore, even with 15 ppm sulfur fuel DOCs would not be able to 
achieve the level of PM control needed to meet the standard proposed 
today.
---------------------------------------------------------------------------

    \132\ ``Demonstration of Advanced Emission Control Technologies 
Enabling Diesel-Powered Heavy-Duty Engines to Achieve Low Emission 
Levels'', Manufacturers of Emission Controls Association, June 1999. 
Air Docket A-2001-28.
    \133\ ``Demonstration of Advanced Emission Control Technologies 
Enabling Diesel-Powered Heavy-Duty Engines to Achieve Low Emission 
Levels'', Manufacturers of Emission Controls Association, June 1999. 
Air Docket A-2001-28.
---------------------------------------------------------------------------

    DPFs control diesel PM by capturing the soot portion of PM in a 
filter media, typically a ceramic wall flow substrate, and then by 
oxidizing (burning) it in the oxygen-rich atmosphere of diesel exhaust. 
The SOF portion of diesel PM can be controlled through the addition of 
catalytic materials to the DPF to form a catalyzed diesel particulate 
filter (CDPF).\134\ The catalytic material is also very effective to 
promote soot burning. This burning off of collected PM is referred to 
as ``regeneration.'' In aggregate over an extended period of operation, 
the PM must be regenerated at a rate equal to or greater that its 
accumulation rate, or the DPF will clog. For a non-catalyzed DPF the 
soot can regenerate only at very high temperatures, in excess of 
600[deg]C, a temperature range which is infrequently realized in normal 
diesel engine operation (for many engines exhaust temperatures may 
never reach 600[deg]C). With the addition of a catalytic coating to 
make a CDPF, the temperature necessary to ensure regeneration is 
decreased significantly to approximately 250[deg]C, a temperature 
within the normal operating range for most diesel engines.\135\
---------------------------------------------------------------------------

    \134\ With regard to gaseous emissions such as NMHCs and CO, the 
CDPF works in the same manner with similar effectiveness as the DOC 
(i.e., NMHC and CO emissions are reduced by more than 80 percent).
    \135\ Engelhard DPX catalyzed diesel particulate filter retrofit 
verification, www.epa.gov/otaq/retrofit/techlist-engelhard.htm, a 
copy of this information is available in Air Docket A-2001-28.
---------------------------------------------------------------------------

    However, the catalytic materials that most effectively promote soot 
and SOF oxidation are significantly impacted by sulfur in diesel fuel. 
Sulfur both degrades catalyst oxidation efficiency (i.e. poisons the 
catalyst) and forms sulfate PM. Both catalyst poisoning by sulfur and 
increases in PM emissions due to sulfate make influence our decision to 
limit the sulfur level of diesel fuel to 15 ppm as discussed in greater 
detail in section III.F.
    Filter regeneration is affected by catalytic materials used to 
promote oxidation, sulfur in diesel fuel, engine-out soot rates, and 
exhaust temperatures. At higher exhaust temperatures soot oxidation 
occurs at a higher rate. Catalytic materials accelerate soot oxidation 
at a single exhaust temperature compared to non-catalyst DPFs, but even 
with catalytic materials increasing the exhaust temperature further 
accelerates soot oxidation.
    Having applied 15 ppm sulfur diesel fuel and the best catalyst 
technology to promote low temperature oxidation (regeneration), the 
regeneration balance of soot oxidation equal to or greater than soot 
accumulation over aggregate operation simplifies to: are the exhaust 
temperatures high enough on aggregate to oxidize the engine-out PM 
rate? \136\ The answer is yes, for most highway applications and many 
nonroad applications, as demonstrated by the widespread success of 
retrofit CDPF systems for nonroad equipment and the

[[Page 28372]]

use of both retrofit and original equipment CDPF systems for highway 
vehicles.137 138 139 However, it is possible that for some 
nonroad applications the engine-out PM rate may exceed the soot 
oxidation rate, even with low sulfur diesel fuel and the best catalyst 
technologies. Should this occur, successful regeneration requires that 
either engine-out PM rates be decreased or exhaust temperatures be 
increased, both feasible strategies. In fact, we expect both to occur 
as highway based technologies are transferred to nonroad engines. As 
discussed earlier, engine technologies to lower PM emissions while 
improving fuel consumption are continuously being developed and 
refined. As these technologies are applied to nonroad engines driven by 
both new emission standards and market pressures for better products, 
engine-out PM rates will decrease. Similarly, techniques to raise 
exhaust temperatures periodically in order to initiate soot oxidation 
in a PM filter have been developed for highway diesel vehicles as 
typified by the PSA system used on more than 400,000 vehicles in 
Europe.140 141
---------------------------------------------------------------------------

    \136\ If the question was asked, ``without 15 ppm sulfur fuel 
and the best catalyst technology, are the exhaust temperatures high 
enough on aggregate to oxidize the engine-out PM rate?'' the answer 
would be no, for all but a very few nonroad or highway diesel 
engines.
    \137\ ``Particulate Traps for Construction Machines, Properties 
and Field Experience,'' 2000, SAE 2000-01-1923.
    \138\ Letter from Dr. Barry Cooper, Johnson Matthey, to Don 
Kopinski, U.S. EPA. Copy available in EPA Air Docket A-2001-28.
    \139\ EPA Recognizes Green Diesel Technology Vehicles at 
Washington Ceremony, Press Release from International Truck and 
Engine Company, July 27, 2001. Copy available in EPA Air Docket A-
2001-28.
    \140\ There is one important distinction between the current PSA 
system and the kind of system that we project industry will use to 
comply with the Tier 4 standards. The PSA system incorporates a 
cerium fuel additive to help promote soot oxidation. The additive 
serves a similar function to a catalyst to promote soot oxidation at 
lower temperatures. Even with the use of the fuel additive passive 
regeneration is not realized on the PSA system and an active 
regeneration is conducted periodically involving late cycle fuel 
injection and oxidation of the fuel on an up-front diesel oxidation 
catalyst to raise exhaust temperatures. This form of supplemental 
heating to ensure infrequent but periodic PM filter regeneration has 
proven to be robust and reliable for more than 400,000 PSA vehicles. 
Our 2002 progress review found that other manufacturers will be 
introducing similar systems in the next few years without the use of 
a fuel additive.
    \141\ Nino, S. and Lagarrigue, M. ``French Perspective on Diesel 
Engines and Emissions,'' presentation at the 2002 Diesel Engine 
Emission Reduction workshop in San Diego, California, Air Docket A-
2001-28.
---------------------------------------------------------------------------

    During our 2002 Highway Diesel Progress Review, we investigated the 
plans of highway engine manufacturers to use CDPF systems to comply 
with the HD2007 emission standards for PM. We learned that all diesel 
engine manufacturers intend to comply through the application of CDPF 
system technology. We also learned that the manufacturers are 
developing means to raise the exhaust temperature, if necessary, to 
ensure that CDPF regeneration occurs.\142\ These technologies include 
modifications to fuel injection strategies, modifications to EGR 
strategies, and modifications to turbocharger control strategies. These 
systems are based upon the technologies used by the engine 
manufacturers to comply with the 2004 highway emission standards. In 
general, the systems anticipated to be used by highway manufacturers to 
meet the 2004 emission standards are the same technologies that engine 
manufacturers have indicated to EPA that they will use to comply with 
the Tier 3 nonroad regulations (e.g., electronic fuel systems).\143\ In 
a manner similar to highway engine manufacturers, we expect nonroad 
engine manufacturers to adapt their Tier 3 emission control 
technologies to provide back-up regeneration systems for CDPF 
technologies in order to comply with the standards we are proposing 
today. We have estimated costs for such systems in our cost analysis.
---------------------------------------------------------------------------

    \142\ Highway Diesel Progress Review, United States 
Environmental Protection Agency, June 2002, EPA 420-R-02-016. Copy 
available in EPA Air Docket A-2001-28.
    \143\ ``Nonroad Diesel Emissions Standards Staff Technical 
Paper'', EPA420-R-01-052, October 2001. Copy available in EPA Air 
Docket A-2001-28.
---------------------------------------------------------------------------

    Emission levels from CDPFs are determined by a number of factors. 
Filtering efficiencies for solid particle emissions like soot are 
determined by the characteristics of the PM filter, including wall 
thickness and pore size. Filtering efficiencies for diesel soot can be 
99 percent with the appropriate filter design.\144\ Given an 
appropriate PM filter design the contribution of the soot portion of PM 
to the total PM emissions are negligible (less than 0.001 g/bhp-hr). 
This level of soot emission control is not dependent on engine test 
cycle or operating conditions due to the mechanical filtration 
characteristics of the particulate filter.
---------------------------------------------------------------------------

    \144\ Miller, R. et. al, ``Design, Development and Performance 
of a Composite Diesel Particulate Filter,'' March 2002, SAE 2002-01-
0323.
---------------------------------------------------------------------------

    Control of the SOF portion of diesel soot is accomplished on a CDPF 
through catalytic oxidation. The SOF portion of diesel PM consists of 
primarily gas phase hydrocarbons in engine exhaust due to the high 
temperatures and only forms particulate in the environment when it 
condenses. Catalytic materials applied to CDPFs can oxidize a 
substantial fraction of the SOF in diesel PM just as the SOF portion 
would be oxidized by a DOC. However, we believe that for engines with 
very high SOF emissions the emission rate may be higher than can be 
handled by a conventionally sized catalyst resulting in higher than 
zero SOF emissions. If a manufacturer's base engine technology has high 
oil consumption rates, and therefore high engine-out SOF emissions 
(i.e., higher than 0.04 g/bhp-hr), compliance with the 0.01 g/bhp-hr 
emission standard proposed today may require additional technology 
beyond the application of a CDPF system alone.\145\
---------------------------------------------------------------------------

    \145\ SOF oxidation efficiency is typically better than 80 
percent and can be better than 90 percent. Given a base engine SOF 
rate of 0.04 g/bhp-hr and an 80 percent SOF reduction a tailpipe 
emission of 0.008 can be estimated from SOF alone. This level may be 
too high to comply with a 0.01 g/bhp-hr standard once the other 
constituents of diesel PM (soot and sulfate) are added. In this 
case, SOF emissions will need to be reduced engine-out or SOF 
control greater than 90 percent will need to be realized by the 
CDPF.
---------------------------------------------------------------------------

    Modern highway diesel engines have controlled SOF emission rates in 
order to comply with the existing 0.1 g/bhp-hr emission standards. For 
modern highway diesel engines, the SOF portion of PM is typically on a 
small fraction of the total PM emissions (less than 0.02 g/bhp-hr). 
This level of SOF control is accomplished by controlling oil 
consumption through the use of engine modifications (e.g., piston ring 
design, the use of 4-valve heads, the use of valve stem seals, 
etc.).\146\ Nonroad diesel engines may similarly need to control 
engine-out SOF emissions in order to comply with the standard proposed 
today. The means to control engine-out SOF emissions are well known and 
have additional benefits, as they decrease oil consumption reducing 
operating costs. With good engine-out SOF control (i.e., engine-out SOF 
< 0.02 g/bhp-hr) and the application of catalytic material to the DPF, 
SOF emissions from CDPF equipped nonroad engines will contribute only a 
very small fraction of the total tailpipe PM emissions (less than 0.004 
g/bhp-hr). Alternatively, it may be less expensive or more practical 
for some applications to ensure that the SOF control realized by the 
CDPF is in excess of 90 percent, thereby allowing for higher engine-out 
SOF emission levels.
---------------------------------------------------------------------------

    \146\ Hori, S. and Narusawa, K. ``Fuel Composition Effects on 
SOF and PAH Exhaust Emissions from DI Diesel Engines,'' SAE 980507.
---------------------------------------------------------------------------

    The best means to reduce sulfate emissions from diesel engines is 
by reducing the sulfur content of diesel fuel and lubricating oils. 
This is one of the reasons that we have proposed today to limit nonroad 
diesel fuel sulfur levels to be 15ppm or less. The catalytic material 
on the CDPF is crucial to

[[Page 28373]]

ensuring robust regeneration and high SOF oxidation; however, it can 
also oxidize the sulfate in the exhaust with high efficiency. The 
result is that the predominant form of PM emissions from CDPF equipped 
diesel engines is sulfate PM. Even with 15ppm sulfur diesel fuel a CDPF 
equipped diesel engine can have total PM emissions including sulfate 
emissions as high as 0.009 g/bhp-hr over some representative operating 
cycles using conventional diesel engine oils.\147\ Although this level 
of emissions will allow for compliance with our proposed PM emissions 
standard of 0.01 g/bhp-hr, we believe that there is room for reductions 
from this level in order to provide engine manufacturers with 
additional compliance margin. During our 2002 Highway Progress Review, 
we learned that a number of engine lubricating oil companies are 
working to reduce the sulfur content in engine lubricating oils. Any 
reduction in the sulfur level of engine lubricating oils will be 
beneficial. Similarly, as discussed above, we expect engine 
manufacturers to reduce engine oil consumption in order to reduce SOF 
emissions and secondarily to reduce sulfate PM emissions. While we 
believe that sulfate PM emissions will be the single largest source of 
the total PM from diesel engines, we believe with the combination of 
technology, and the appropriate control of engine-out PM, that sulfate 
and total PM emissions will be low enough to allow compliance with a 
0.01 g/bhp-hr standard, except in the case of small engines with higher 
fuel consumption rates as described later in this section.
---------------------------------------------------------------------------

    \147\ See Table III.F.1 below.
---------------------------------------------------------------------------

    CDPFs have been shown to be very effective at reducing PM mass by 
reducing dramatically the soot and SOF portions of diesel PM. In 
addition, recent data show that they are also very effective at 
reducing the overall number of emitted particles when operated on low 
sulfur fuel. Hawker, et. al., found that a CDPF reduced particle count 
by over 95 percent, including some of the smallest measurable particles 
(< 50 nm), at most of the tested conditions. The lowest observed 
efficiency in reducing particle number was 86 percent. No generation of 
particles by the CDPF was observed under any tested conditions.\148\ 
Kittelson, et al., confirmed that ultrafine particles can be reduced by 
a factor of ten by oxidizing volatile organics, and by an additional 
factor of ten by reducing sulfur in the fuel. Catalyzed PM traps 
efficiently oxidize nearly all of the volatile organic PM precursors 
(i.e. SOF), and the reduction of diesel fuel sulfur levels to 15ppm or 
less will substantially reduce the number of ultrafine PM emitted from 
diesel engines. The combination of CDPFs with low sulfur fuel is 
expected to result in very large reductions in both PM mass and the 
number of ultrafine particles.
---------------------------------------------------------------------------

    \148\ Hawker, P., et al., Effect of a Continuously Regenerating 
Diesel Particulate Filter on Non-Regulated Emissions and Particle 
Size Distribution, SAE 980189.
---------------------------------------------------------------------------

    As described here, the range of technologies available to reduce PM 
emissions is broad, extending from improvements to existing combustion 
system technologies to oxidation catalyst technologies to complete CDPF 
systems. The CDPF technology along with 15ppm or less sulfur diesel 
fuel is the system that we believe will allow engine manufacturers to 
comply with the 0.01 g/bhp-hr PM standard that we have proposed for a 
wide range of nonroad diesel engines. While it may be possible to apply 
CDPFs across the full range of nonroad diesel engine sizes, the 
complexity of full diesel particulate filter systems makes application 
to the smallest range of diesel engines difficult to accurately 
forecast at this time. As described in the following sections, the 
Agency has given consideration to the engineering complexity, cost and 
packaging of these systems in setting emission standards for various 
nonroad engine power categories.
b. NOX Control Technologies
    Oxides of nitrogen (NO and NO2, collectively called 
NOX) are formed at high temperatures during the combustion 
process from nitrogen and oxygen present in the intake air. The 
NOX formation rate is exponentially related to peak cylinder 
temperatures and is also strongly related to nitrogen and oxygen 
content (partial pressures). NOX control technologies for 
diesel engines have focused on reducing emissions by lowering the peak 
cylinder temperatures and by decreasing the oxygen content of the 
intake air. A number of technologies have been developed to accomplish 
these objectives including fuel injection timing retard, fuel injection 
rate control, charge air cooling, exhaust gas recirculation (EGR) and 
cooled EGR. The use of these technologies can result in significant 
reductions in NOX emissions, but are limited due to 
practical and physical constraints of heterogeneous diesel 
combustion.149 150
---------------------------------------------------------------------------

    \149\ Flynn, P. et al, ``Minimum Engine Flame Temperature 
Impacts on Diesel and Spark-Ignition Engine NOX 
Production,'' SAE 2000-01-1177, March 2000.
    \150\ Dickey, D. et al, ``NOX Control in Heavy-Duty 
Diesel Engines--What is the Limit?,'' SAE 980174, February 1998.
---------------------------------------------------------------------------

    EPA is investigating strategies to address these limitations of 
heterogenous diesel combustion in a research program. This concept 
consists of higher intake charge boost levels using a low-pressure loop 
cooled EGR system, combined with a proprietary fuel injection and 
combustion system to control engine-out NOX.\151\ The 
results from prototype laboratory research engines show NOX 
control consistent with the standards proposed today. The technology 
must still overcome the limitations of increased PM emissions at low 
NOX levels as well as other practical considerations of 
performance and durability. EPA intends to continue investigating this 
technology, but at this time cannot project that this technology would 
be generally available for use in compliance with the proposed 
standards.
---------------------------------------------------------------------------

    \151\ Gray, Charles ``Assessing New Diesel Technologies,'' 
November 2002, MIT Light Duty Diesel Workshop, available on MIT's 
website or in Air Docket A-2001-28. http://web.mit.edu/chrisng/www/
dieselworkshop_files/Charles%20Gray.PDF.
---------------------------------------------------------------------------

    A new form of diesel engine combustion, commonly referred to as 
homogenous diesel combustion or premixed diesel combustion, can give 
very low NOX emissions over a limited range of diesel engine 
operation. In the regions of diesel engine operation over which this 
combustion technology is feasible (light load conditions), 
NOX emissions can be reduced enough to comply with the 0.3 
g/bhp-hr NOX emission standard that we have proposed 
today.\152\ Some engine manufacturers are today producing engines which 
utilize this technology over a narrow range of engine operation.\153\ 
Unfortunately, it is not possible today to apply this technology over 
the full range of diesel engine operation. We do believe that more 
engine manufacturers will utilize this alternative combustion approach 
in the limited range over which it is effective, but will have to rely 
on conventional heterogenous diesel combustion for the bulk of engine 
operation. Therefore, we believe that catalytic NOX emission 
control technologies will be required in order to realize the 
NOX emission standards proposed today. Catalytic emission 
control technologies can extend the reduction of NOX 
emissions

[[Page 28374]]

by an additional 90 percent or more over conventional ``engine-out'' 
control technologies alone.
---------------------------------------------------------------------------

    \152\ Stanglmaier, Rudolf and Roberts, Charles ``Homogenous 
Charge Compression Ignition (HCCI): Benefits, Compromises, and 
Future Engine Applications''. SAE 1999-01-3682.
    \153\ Kimura, Shuji, et al., ``Ultra-Clean Combustion Technology 
Combining a Low-Temperature and Premixed Combustion Concept for 
Meeting Future Emission Standards'', SAE 2001-01-0200.
---------------------------------------------------------------------------

    NOX emissions from gasoline-powered vehicles are 
controlled to extremely low levels through the use of the three-way 
catalyst technology first introduced in the 1970s. Three-way-catalyst 
technology is very efficient in the stoichiometric conditions found in 
the exhaust of properly controlled gasoline-powered vehicles. Today, an 
advancement upon this well-developed three-way catalyst technology, the 
NOX adsorber, has shown that it too can make possible 
extremely low NOX emissions from lean-burn engines such as 
diesel engines.\154\ The potential of the NOX adsorber 
catalyst is limited only by its need for careful integration with the 
engine and engine control system (as was done for three-way catalyst 
equipped passenger cars in the 1980s and 1990s) and by poisoning of the 
catalyst from sulfur in the fuel. The Agency set stringent new 
NOX standards for highway diesel engines beginning in 2007 
predicated upon the use of the NOX adsorber catalyst enabled 
by significant reductions in fuel sulfur levels (15 ppm sulfur or 
less). In today's action, we are proposing similarly stringent 
NOX emission standards for nonroad engines again using 
technology enabled by a reduction in fuel sulfur levels.
---------------------------------------------------------------------------

    \154\ NOX adsorber catalysts are also called, 
NOX storage catalysts (NSCs), NOX storage and 
reduction catalysts (NSRs), and NOX traps.
---------------------------------------------------------------------------

    NOX adsorbers work to control NOX emissions 
by storing NOX on the surface of the catalyst during the 
lean engine operation typical of diesel engines. The adsorber then 
undergoes subsequent brief rich regeneration events where the 
NOX is released and reduced across precious metal catalysts. 
The NOX storage period can be as short as 15 seconds and as 
along as 10 minutes depending upon engine-out NOX emission 
rates and exhaust temperature. A number of methods have been developed 
to accomplish the necessary brief rich exhaust conditions necessary to 
regenerate the NOX adsorber technology including late-cycle 
fuel injection, also called post injection, in exhaust fuel injection, 
and dual bed technologies with off-line 
regeneration.155 156 157 This method for NOX 
control has been shown to be highly effective when applied to diesel 
engines but has a number of technical challenges associated with it. 
Primary among these is sulfur poisoning of the catalyst as described in 
section III.F below. In the HD2007 RIA we identified four issues 
related to NOX adsorber performance: performance of the 
catalyst across a broad range of exhaust temperatures, thermal 
durability of the catalyst when regenerated to remove sulfur 
(desulfated), management of sulfur poisoning, and system integration on 
a vehicle. In the HD 2007 RIA, we provided a description of the 
technology paths that we believed manufacturers would use to address 
these challenges. We are conducting an ongoing review of industry's 
progress to overcome these challenges and have updated our analysis of 
the progress to address these issues in the draft RIA associated with 
today's NPRM.
---------------------------------------------------------------------------

    \155\ Johnson, T. ``Diesel Emission Control in Review--the Last 
12 Months,'' SAE 2003-01-0039.
    \156\ Koichiro Nakatani, Shinya Hirota, Shinichi Takeshima, 
Kazuhiro Itoh, Toshiaki Tanaka, and Kazuhiko Dohmae, ``Simultaneous 
PM and NOX Reduction System for Diesel Engines.'', SAE 
2002-01-0957, SAE Congress March 2002.
    \157\ Schenk, C., McDonald, J. and Olson, B. ``High Efficiency 
NOX and PM Exhaust Emission Control for Heavy-Duty On-
Highway Diesel Engines,'' SAE 2001-01-1351.
---------------------------------------------------------------------------

    One of the areas that we have identified as needing improvement for 
the NOX adsorber catalyst is performance at low and high 
exhaust temperatures. NOX adsorber performance is limited at 
very high temperatures (due to thermal release of NOX under 
lean conditions) and very low temperatures (due to poor catalytic 
activity for NO oxidation under lean conditions and low activity for 
NOX reduction under rich conditions) as described 
extensively in the draft RIA. Our review of highway HD2007 technologies 
showed that significant progress has been made to broaden the 
temperature range of effective NOX control of the 
NOX adsorber catalysts (the temperature ``window'' of the 
catalyst). Every catalyst development company that we visited was able 
to show us new catalyst formulations with improved performance at both 
high and low temperatures. Similarly, many of the engine manufacturers 
we visited showed us data indicating that the improvements in catalyst 
formulations corresponded to improvements in emission reductions over 
the regulated test cycles. It is clear from the data presented to EPA 
that the progress with regard to NOX adsorber performance 
has been both substantial and broadly realized by most technology 
developers. The importance of this temperature window to nonroad engine 
manufacturers is discussed in more detail later in this section.
    Long term durability has been the greatest concern for the 
NOX adsorber catalyst. We have concluded as described 
briefly in III.F below and in some detail in the draft RIA, that in 
order for NOX adsorbers to effectively control 
NOX emission throughout the life of a nonroad diesel engine 
the fuel sulfur level will have to be maintained at or below 15 ppm, 
that the NOX adsorber catalyst thermal durability will need 
to improve in order to allow for sulfur regeneration events (since 
adsorber thermal degradation, ``sintering,'' is associated with each 
desulfation event, the number of desulfation events should be 
minimized), and that system improvements will have to be made in order 
to allow for appropriate management of sulfur poisoning. It is in this 
area of durability that NOX adsorbers had the greatest need 
for improvement, and it is here where some of the most impressive 
ongoing strides in technology development have been made. During our 
ongoing review, we have learned that catalyst companies are making 
significant improvements in the thermal durability of the catalyst 
materials used in NOX adsorbers. Similarly, the substrate 
manufacturers are developing new materials that address the problem of 
NOX storage material migration into the substrate.\158\ The 
net gain from these simultaneous improvements are NOX 
adsorber catalysts which can be desulfated (go through a sulfur 
regeneration process) with significantly lower levels of thermal damage 
to the catalyst function. In addition, engine manufacturers and 
emission control technology vendors are developing new strategies to 
accomplish desulfation that allow for improved sulfur management while 
minimizing the damage due to sulfur poisoning. It was clear in our 
review that the total system improvements being made when coupled with 
changes to catalytic materials and catalyst substrates are delivering 
significantly improved catalyst durability to the NOX 
adsorber technology.
---------------------------------------------------------------------------

    \158\ Some NOX storage materials can interact with 
the catalyst substrate especially at elevated temperatures making 
the storage material unavailable for NOX storage and 
weakening the substrate.
---------------------------------------------------------------------------

    Practical application of the NOX adsorber catalyst in a 
vehicle was an issue during the HD2007 rulemaking and similarly there 
are issues regarding the application of NOX adsorbers to 
nonroad equipment. Although there is considerable evidence that 
NOX adsorbers are highly effective and that durability 
issues can be addressed, some worry that the application of the 
NOX adsorber systems to vehicles and nonroad equipment will 
be impractical due to packaging constraints and the

[[Page 28375]]

potential for high fuel consumption. Our review of progress has left us 
more certain than ever that practical system solutions can be applied 
to control emissions using NOX adsorbers. We have tested a 
diesel passenger car (one of the most difficult packaging situations) 
with a complete NOX adsorber and particulate filter system 
that demonstrated both exceptional emission control and very low fuel 
consumption.\159\ Heavy-duty engine manufacturers have shared with us 
their improvements in system design and means to regenerate 
NOX while minimizing fuel consumption.\160\ Our own in-house 
testing program at the National Vehicle and Fuel Emissions Laboratory 
(NVFEL) is developing a number of novel ideas to reduce the total 
system package size while maintaining high levels of emission control 
and low fuel consumption rates as discussed more fully in the draft 
RIA. Similarly, a number of Department of Energy (DOE), Advanced 
Petroleum Based Fuel--Diesel Emission Control (APBF-DEC) program 
NOX adsorber projects are working to address the system 
integration challenges for a diesel passenger car, a large sport 
utility vehicle and for a heavy heavy-duty truck.\161\ By citing these 
numerous examples, we are not intending to imply that the challenge of 
integrating and packaging advanced emission control technologies is 
easy. Rather, we believe these examples show that even though 
significant challenges exist, they can be overcome through careful 
design and integration efforts. Nonroad equipment manufacturers have 
addressed similar challenges in the past when they have added 
additional customer features (e.g., packaged an air-conditioning 
system) or in accommodating other emission control technologies (e.g., 
charge air cooling systems).
---------------------------------------------------------------------------

    \159\ McDonald, J and Bunker, B. ``Testing of the Toyota Avensis 
DPNR at U.S. EPA-NVFEL,'' SAE 2002-01-2877.
    \160\ Hakim, N. ``NOX Adsorbers for Heavy Duty Truck 
Engines--Testing and Simulation,'' presentation at Motor Fuels: 
Effects on Energy Efficiency and Emissions in the Transportation 
Sector Joint Meeting of Research Program Sponsored by the USA Dept. 
of Energy, Clean Air for Europe and Japan Clean Air, October 9-10, 
2002. Copy available in EPA Air Docket A-2001-28.
    \161\ Details with quarterly updates on the APBF-DEC programs 
can be found on the DOE website at the following location http://
www.ott.doe.gov/apbf.shtml.
---------------------------------------------------------------------------

    All of the issues described above and highlighted first during the 
HD2007 rulemaking are likely to be concerns to nonroad engine and 
nonroad equipment manufacturers. We believe the challenge to overcome 
these issues will be significant for nonroad engines and equipment. 
Yet, we have documented substantial progress by industry in the last 
year to overcome these challenges, and we continue to believe based on 
the progress we have observed that the NOX adsorber catalyst 
technology will be mature enough for application to many diesel engines 
by 2007. In the case of NOX adsorber temperature window, 
which could be especially challenging for nonroad engines, we have 
performed an analysis summarized below in section III.E.2 and 
documented in the draft RIA, that leads us to conclude the technology 
can be successfully applied to nonroad engines provided there is some 
additional lead time for further engine and catalyst system technology 
development. Similarly, we acknowledge that the diverse nature and 
sheer number of different nonroad equipment types makes the challenge 
of packaging advanced emission control technologies more difficult. 
Therefore, we have included a number of equipment manufacturer 
flexibilities in the program proposed today in order to allow equipment 
manufacturers to manage the engineering resource challenges imposed by 
these regulations.
    Another NOX catalyst based emission control technology 
is selective catalytic reduction (SCR). SCR catalysts require a 
reductant, ammonia, to reduce NOX emissions. Because of the 
significant safety concerns with handling and storing ammonia, most SCR 
systems make ammonia within the catalyst system from urea. Such systems 
are commonly called urea SCR systems. (Throughout this document the 
term SCR and urea SCR may be used interchangeably and should be 
considered as referring to the same urea based catalyst system.) With 
the appropriate control system to meter urea in proportion to engine-
out NOX emissions, urea SCR catalysts can reduce 
NOX emissions by over 90 percent for a significant fraction 
of the diesel engine operating range.\162\ Although EPA has not done an 
extensive analysis to evaluate its effectiveness, we believe it may be 
possible to reduce NOX emissions with a urea SCR catalyst to 
levels consistent with compliance with the proposed NOX 
standards.
---------------------------------------------------------------------------

    \162\ ``Demonstration of Advanced Emission Control Technologies 
Enabling Diesel-Powered Heavy-Duty Engines to Achieve Low Emission 
Levels'', Manufacturers of Emissions Controls Association, June 1999 
Air Docket A-2001-28.
---------------------------------------------------------------------------

    However, we have significant concerns regarding a technology that 
requires extensive user intervention in order to function properly and 
the lack of the urea delivery infrastructure necessary to support this 
technology. Urea SCR systems consume urea in proportion to the engine-
out NOX rate. The urea consumption rate can be on the order 
of five percent of the engine fuel consumption rate. Therefore, unless 
the urea tank is prohibitively large, the urea must be replenished 
frequently. Most urea systems are designed to be replenished every time 
fuel is added or at most every few times that fuel is added. Today, 
there is not a system in place to deliver or dispense automotive grade 
urea to diesel fueling stations. One study conducted for the National 
Renewable Energy Laboratory (NREL), estimated that if urea were to be 
distributed to every diesel fuel station in the United States, the cost 
would be more than $30 per gallon.\163\
---------------------------------------------------------------------------

    \163\ Fable, S. et al, ``Subcontractor Report--Selective 
Catalytic Reduction Infrastructure Study,'' AD Little under contract 
to National Renewable Energy Laboratory, July 2002, NREL/SR-5040-
32689. Copy available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    We are not aware of a proven mechanism that ensures that the user 
will replenish the urea supply as necessary to maintain emissions 
performance. Further, we believe given the additional cost for urea, 
that there will be significant disincentives for the end-user to 
appropriately replenish the urea because the cost of urea could be 
avoided without equipment performance loss. See NRDC v. EPA, 655 F. 2d 
318, 332 (D.C. Cir. 1981) (referring to ``behavioral barriers to 
periodic restoration of a filter by a [vehicle] owner'' as a valid 
basis for EPA considering a technology unavailable). Due to the lack of 
an infrastructure to deliver the needed urea, and the lack of a track 
record of successful ways to ensure urea use, we have concluded that 
the urea SCR technology is not likely to be available for general use 
in the time frame of the proposed standards. Therefore, we have not 
based the feasibility or cost analysis of this emission control program 
on the use or availability of the urea SCR technology. However, we 
would not preclude its use for compliance with the emission standards 
provided that a manufacturer could demonstrate satisfactorily to the 
Agency that urea would be used under all conditions. We believe that 
only a few unique applications will be able to be controlled in a 
manner such that urea use can be assured, and therefore believe it is 
inappropriate to base a national emission control program on a 
technology which can serve effectively only in a few niche 
applications.
    This section has described a number of technologies that can reduce

[[Page 28376]]

emissions from diesel engines. The following section describes the 
challenges to applying these diesel engine technologies to engines and 
equipment designed for nonroad applications.
2. Can These Technologies Be Applied to Nonroad Engines and Equipment?
    The emission standards and the introduction dates for those 
standards, as described earlier in this section, are premised on the 
transfer of diesel engine technologies being or already developed to 
meet light-duty and heavy-duty vehicle standards that begin in 2007. 
The standards that we are proposing today for engines =75 
horsepower will begin to go into effect four years later. This time lag 
between equivalent highway and nonroad diesel engine standards is 
necessary in order to allow time for engine and equipment manufacturers 
to further develop these highway technologies for nonroad engines and 
to align this program with nonroad Tier 3 emission standards that begin 
to go into effect in 2006.
    As discussed previously, the test procedures and regulations for 
the HD2007 highway engines include a transient test procedure, a broad 
steady-state procedure, and NTE provisions that require compliant 
engines to emit at or below 1.5 times the regulated emission levels 
under virtually all conditions. An engine designed to comply with the 
2007 highway emission standards would comply with the equivalent 
nonroad emission standards proposed today if it were to be tested over 
the transient and steady-state nonroad emission test procedures 
proposed today, which cover the same regions and types of engine 
operation. Said in another way, a highway diesel engine produced in 
2007 could be certified in compliance with the transient and steady-
state standards proposed today for nonroad diesel engines several years 
in advance of the date when these standards would go into effect. 
However, that engine, while compliant with certain of the nonroad 
emission standards proposed today, would not necessarily be designed to 
address the various durability and performance requirements of many 
nonroad equipment manufacturers. We expect that the engine 
manufacturers will need additional time to further develop the 
necessary emission control systems to address some of the nonroad 
issues described below as well as to develop the appropriate 
calibrations for engine rated speed and torque characteristics required 
by the diverse range of nonroad equipment. Furthermore, not all nonroad 
engine manufacturers produce highway diesel engines or produce nonroad 
engines that are developed from highway products. Therefore, there is a 
need for lead time between the Tier 3 emission standards which go into 
effect in 2006-2008 and the Tier 4 emission standards. We believe the 
technologies developed to comply with the Tier 3 emission standards 
such as improved air handling systems and electronic fuel systems will 
form an essential technology baseline which manufacturers will need to 
initiate and control the various regeneration functions required of the 
catalyst based technologies for Tier 4. The Agency has given 
consideration to all of these issues in setting the emission standards 
and the timing of those standards as proposed today.
    This section describes some of the challenges to applying advanced 
emission control technologies to nonroad engines and equipment, and why 
we believe that technologies developed for highway diesel engines can 
be further refined to address these issues in a timely manner for 
nonroad engines consistent with the emission standards proposed today. 
This section paraphrases a more in-depth analysis in the draft RIA.
a. Nonroad Operating Conditions and Exhaust Temperatures
    Nonroad equipment is highly diverse in design, application, and 
typical operating conditions. This variety of operating conditions 
affects emission control systems through the resulting variety in the 
torque and speed demands (i.e. power demands). This wide range in what 
constitutes typical nonroad operation makes the design and 
implementation of advanced emission control technologies more 
difficult. The primary concern for catalyst based emission control 
technologies is exhaust temperature. In general, exhaust temperature 
increases with engine power and can vary dramatically as engine power 
demands vary.
    For most catalytic emission control technologies there is a minimum 
temperature below which the chemical reactions necessary for emission 
control do not occur. The temperature above which substantial catalytic 
activity is realized is often called the light-off temperature. For 
gasoline engines, the light-off temperature is typically only important 
in determining cold start emissions. Once gasoline vehicle exhaust 
temperatures exceed the light-off temperature, the catalyst is ``lit-
off'' and remains fully functional under all operating conditions. 
Diesel exhaust is significantly cooler than gasoline exhaust due to the 
diesel engine's higher thermal efficiency and its operation under 
predominantly lean conditions. Absent control action taken by an 
electronic engine control system, diesel exhaust may fall below the 
light-off temperature of catalyst technology even when the vehicle is 
fully warmed up.
    The relationship between the exhaust temperature of a nonroad 
diesel engine and light-off temperature is an important factor for both 
CDPF and NOX adsorber technologies. For the CDPF technology, 
exhaust temperature determines the rate of filter regeneration and if 
too low causes a need for supplemental means to ensure proper filter 
regeneration. In the case of the CDPF, it is the aggregate soot 
regeneration rate that is important, not the regeneration rate at any 
particular moment in time. A CDPF controls PM emissions under all 
conditions and can function properly (i.e., not plug) even when exhaust 
temperatures are low for an extended time and the regeneration rate is 
lower than the soot accumulation rate, provided that occasionally 
exhaust temperatures and thus the soot regeneration rate are increased 
enough to regenerate the CDPF. A CDPF can passively (without 
supplemental heat addition) regenerate if exhaust temperatures remain 
above 250[deg]C for more than 30 percent of engine operation.\164\ 
Similarly, there is a minimum temperature (e.g., 200[deg]C) for 
NOX adsorbers below which NOX regeneration is not 
readily possible and a maximum temperature (e.g., 500[deg]C) above 
which NOX adsorbers are unable to effectively store 
NOX. These minimum and maximum temperatures define a 
characteristic temperature window of the NOX adsorber 
catalyst. When the exhaust temperature is within the temperature window 
(above the minimum and below the maximum) the catalyst is highly 
effective. When exhaust temperatures fall outside this window of 
operation, NOX adsorber effectiveness is diminished. 
Therefore, there is a need to match diesel exhaust temperatures to 
conditions for effective catalyst operation under the various operating 
conditions of nonroad engines.
---------------------------------------------------------------------------

    \164\ Engelhard DPX catalyzed diesel particulate filter retrofit 
verification, www.epa.gov/otaq/retrofit/techlist-engelhard.htm, a 
copy of this information is available in Air Docket A-2001-28.
---------------------------------------------------------------------------

    Although the range of products for highway vehicles is not as 
diverse as for nonroad equipment, the need to match exhaust 
temperatures to catalyst characteristics is still present. This is a 
significant concern for highway engine

[[Page 28377]]

manufacturers and has been a focus of our ongoing diesel engine 
progress review. There we have learned that substantial progress is 
being made to broaden the operating temperature window of catalyst 
technologies while at the same time engine systems are being designed 
to better control exhaust temperatures. Highway diesel engine 
manufacturers are working to address this need through modifications to 
engine design, modifications to engine control strategies and 
modifications to exhaust system designs. Engine design changes, 
including the ability for multiple late fuel injections and the ability 
to control total air flow into the engine, give controls engineers 
additional flexibility to change exhaust temperature characteristics. 
Modifications to the exhaust system, including the use of insulated 
exhaust manifolds and exhaust tubing, can help to preserve the 
temperature of the exhaust gases. New engine control strategies 
designed to take advantage of engine and exhaust system modifications 
can then be used to manage exhaust temperatures across a broad range of 
engine operation. The technology solutions being developed for highway 
engines to better manage exhaust temperature are built upon the same 
emission control technologies (i.e., advanced air handling systems and 
electronic fuel injection systems) that we expect nonroad engine 
manufacturers to use in order to comply with the Tier 3 emission 
standards.
    Matching the operating temperature window of the broad range of 
nonroad equipment may be somewhat more challenging for nonroad engines 
than for many highway diesel engines simply because of the diversity in 
equipment design and equipment use. Nonetheless, the problem has been 
successfully solved in highway applications facing low temperature 
performance situations as difficult to address as any encountered by 
nonroad applications. The most challenging temperature regime for 
highway engines are encountered at very light-loads as typified by 
congested urban driving. Under congested urban driving conditions 
exhaust temperatures may be too low for effective NOX 
reduction with a NOX adsorber catalyst. Similarly, exhaust 
temperatures may be too low to ensure passive CDPF regeneration. To 
address these concerns, light-duty diesel engine manufacturers have 
developed active temperature management strategies that provide 
effective emissions control even under these difficult light-load 
conditions. Toyota has shown with their prototype DPNR vehicles that 
changes to EGR and fuel injection strategies can realize an increase in 
exhaust temperatures of more than 100[deg]F under even very light-load 
conditions allowing the NOX adsorber catalyst to function 
under these normally cold exhaust conditions.\165\ Similarly, PSA has 
demonstrated effective CDPF regeneration under demanding light-load 
taxi cab conditions with current production technologies.\166\ Both of 
these are examples of technology paths available to nonroad engine 
manufacturers to increase temperatures under light-load conditions.
---------------------------------------------------------------------------

    \165\ Sasaki, S., Ito, T., and Iguchi, S., ``Smoke-less Rich 
Combustion by Low Temperature Oxidation in Diesel Engines,'' 9th 
Aachener Kolloquim Fahrzeug--und Motorentechnik 2000. Copy available 
in EPA Air Docket A-2001-28.
    \166\ Jeuland, N., et al, ``Performances and Durability of DPF 
(Diesel Particulate Filter) Tested on a Fleet of Peugeot 607 Taxis 
First and Second Test Phases Results,'' October 2002, SAE 2002-01-
2790.
---------------------------------------------------------------------------

    We are not aware of any nonroad equipment in-use operating cycles 
which would be more demanding of low temperature performance than 
passenger car urban driving. Both the Toyota and PSA systems are 
designed to function even with extended idle operation as would be 
typified by a taxi waiting to pick up a fare. By actively managing 
exhaust temperatures engine manufacturers can ensure highly effective 
catalyst based emission control performance (i.e., compliance with the 
emission standards) and reliable filter regeneration (failsafe 
operation) across a wide range of engine operation as would be typified 
by the broad range of in-use nonroad duty cycles and the new nonroad 
transient test proposed today.
    The systems described here from Toyota and PSA are examples of 
highly integrated engine and exhaust emission control systems based 
upon active engine management designed to facilitate catalyst function. 
Because these systems are based upon the same engine control 
technologies likely to be used to comply with the Tier 3 standards and 
because they allow great flexibility to trade-off engine control and 
catalyst control approaches depending on operating mode and need, we 
believe most nonroad engine manufacturers will use similar approaches 
to comply with the emission standards proposed today. However, there 
are other technologies available that are designed to be added to 
existing engines without the need for extensive integration and engine 
management strategies. One example of such a system is an active DPF 
system developed by Deutz for use on a wide range on nonroad equipment. 
The Deutz system has been sold as an OEM retrofit technology that does 
not require changes to the base engine technology. The system is 
electronically controlled and uses supplemental in-exhaust fuel 
injection to raise exhaust temperatures periodically to regenerate the 
DPF. Deutz has sold over 2,000 of these units and reports that the 
systems have been reliable and effective. Some manufacturers may choose 
to use this approach for compliance with the PM standard proposed 
today, especially in the case of engines which may be able to comply 
with the proposed NOX standards with engine-out emission 
control technologies (i.e., engines rated between 25 and 75 
horsepower).
    High temperature operating regimes such as a heavy heavy-duty 
diesel truck at full payload driving up a grade are also challenging 
for the NOX catalyst technology. Although less common, 
similar high temperature conditions of full engine load operation can 
be imagined for nonroad equipment. However, because highway engines 
typically have higher power density (defined as rated power divided by 
engine displacement), the highest operating conditions would be 
expected to be encountered with highway vehicles. High exhaust 
temperatures (in excess of 500[deg]C) are challenging for the 
NOX adsorber catalyst technology because the stored 
NOX emissions can be released thermally without going 
through a reduction step, leading to increased NOX 
emissions. In the absence of a reductant (normally provided by the 
standard NOX regeneration function) the thermally released 
NOX is emitted from the exhaust system without treatment. To 
address this issue, NOX storage catalyst technologies with 
higher levels of thermal stability are being developed, but these 
technologies trade-off improved high temperature performance for even 
greater sensitivity to fuel sulfur. Beyond catalyst improvements, the 
exhaust temperature from the engine can be controlled prior to the 
NOX adsorber catalyst simply through heat loss in the 
exhaust system (i.e. by locating the catalyst further from the engine). 
Another approach being considered for GDI vehicle applications which 
operate at much higher temperatures than would be encountered by a 
diesel engine is to use a relatively simple exhaust layout design to 
increase heat loss at high temperatures while still providing 
acceptable low temperature

[[Page 28378]]

performance.\167\ Additionally, exhaust temperatures well in excess of 
500[deg]C are not frequently experienced by nonroad engines. Higher 
exhaust temperatures would be expected from naturally aspirated engines 
due to their lower air flow (for the same power/heat input, naturally 
aspirated engines have less air to heat up and thus the exhaust reaches 
a higher temperature). Today, less than ten percent of nonroad diesel 
engines with rated power greater than 100 horsepower are naturally 
aspirated and we have projected that an even greater percentage of 
nonroad engines meeting the Tier 3 emission standards will be 
turbocharged.
---------------------------------------------------------------------------

    \167\ Damson, B., ``Exhaust Cooling for NOX-Traps for 
Lean Spark-Ignition Engines,'' SAE 2002-01-0737.
---------------------------------------------------------------------------

    We have conducted an analysis of various nonroad equipment 
operating cycles and various nonroad engine power density levels to 
better understand the matching of nonroad engine exhaust temperatures, 
catalyst installation locations and catalyst technologies. This 
analysis, documented in the draft RIA, showed that for many engine 
power density levels and equipment operating cycles, exhaust 
temperatures are quite well matched to catalyst temperature window 
characteristics. In particular, the nonroad transient cycle (NRTC), the 
cycle we are proposing to use for certification, was shown to be well 
matched to the NOX adsorber characteristics with estimated 
performance in excess of 90 percent for a turbocharged diesel engine 
tested under a range of power density levels. The analysis also 
indicated that the exhaust temperatures experienced over the NRTC are 
better matched to the NOX adsorber catalyst temperature 
window than the temperatures that would be expected over the highway 
FTP test cycle. This suggests that compliance with the proposed NRTC 
will be somewhat easier, using similar technology, than complying with 
the highway 2007 emission standards on the FTP.
    For engines with low power density (e.g., <25 hp per liter of 
engine displacement) the analysis showed that, absent actions to 
increase exhaust temperatures (e.g., increased use of EGR a light 
loads), compliance with the NRTC cycle will be more difficult than for 
engines with higher power density levels. Specifically, the analysis 
predicted 92% control for the high power density engine and 86% control 
for the low power density engine.
    Note that this analysis approach is only effective to predict 
differences in performance, but not effective to predict absolute 
performance. The same analysis approach predicted 83% control for the 
high power density engine on the heavy-duty FTP, although testing at 
EPA has shown for this engine (a different example of this same engine) 
greater than 90% NOX control.\168\ Nevertheless, the 
analysis suggests that additional attention must be made to designing 
system for low power density applications, and that technology changes 
may be necessary to ensure adequate performance (e.g., the use of EGR 
or other control methods to raise exhaust temperatures). One change, 
which is occurring independent of EPA's regulation, is increasing power 
density for nonroad engines. EPA has documented in the draft RIA a 
clear trend of certified engine ratings that indicates manufacturers 
are increasing engine power without increasing engine displacement. 
Engine manufacturers are motivated to increase engine power density 
because engine pricing is largely done on a power basis, while the cost 
of manufacturing is more closely related to engine displacement. 
Therefore, increasing engine power levels without increasing 
displacement may increase the sale price of the engine more than it 
increases the cost of manufacturing. Increasing power density typically 
results in higher exhaust temperatures and, in this case, better 
matching to catalyst operating requirements. Alternatively, nonroad 
engine manufacturers can apply the same temperature management 
strategies previously described for highway engines.
---------------------------------------------------------------------------

    \168\ Schenk, C., McDonald, J. and Olson, B. ``High Efficiency 
NOX and PM Exhaust Emission Control for Heavy-Duty On-
Highway Diesel Engines,'' SAE 2001-01-1351.
---------------------------------------------------------------------------

    The analysis also suggests that the temperature challenge for 
nonroad equipment will be greater with regard to the NTE provisions of 
this proposal than for the nonroad transient test (NRTC) provisions. In 
fact as discussed previously, the NRTC cycle appears to be a better 
match to the characteristics of the NOX adsorber catalyst 
than the FTP cycle used for heavy-duty highway truck certification. 
This is due to the higher average engine load experienced over the NRTC 
and thus the higher average temperature. Therefore, we believe that 
complying with the NOX standard over the transient test 
cycle proposed today for nonroad engines will not be significantly more 
difficult than complying with the HD2007 NOX emission 
standard over the FTP. The analysis also shows that many nonroad 
engines may operate in-use in a way different from the NRTC (i.e. even 
the NRTC is not an all-encompassing test; no single test realistically 
could be), and that NTE standards are therefore needed to assure that 
nonroad engine emissions are controlled for the full range of possible 
in-use operating conditions.\169\ The technical challenge of 
controlling NOX emissions, even under these diverse 
conditions, is no more difficult on a per engine basis than for highway 
diesel engines which must comply with similar NTE test provisions. This 
is because both highway and nonroad engine manufacturers must address 
control at the same high load and low load conditions (minimum power 
from both are the same, 0 hp, and maximum power is typically higher for 
highway engines, due to higher power density). Also, both engine 
manufacturers must be able to respond to changes in user demanded 
torque (transient conditions) that are similarly unpredictable. 
However, given the sheer number of different nonroad equipment types 
and engine ratings, this represents a real challenge for the nonroad 
industry which is one of the primary considerations given by the Agency 
in determining the appropriate timing for the emission standards 
proposed today.
---------------------------------------------------------------------------

    \169\ The fact that developing compliant engines for the NTE 
provisions may be more difficult than developing for the transient 
test cycle does not diminish the value of the transient test as a 
means to evaluate the overall effectiveness of the emission control 
system under transient conditions. There is no doubt that 
controlling average emissions under transient conditions will be an 
important part of the emission control system and that evaluating 
overall performance under transient conditions is needed.
---------------------------------------------------------------------------

    We believe, based on our analysis of nonroad engines and equipment 
operating characteristics, that in-use some nonroad engines will 
experience conditions that require the use of temperature management 
strategies in order to effectively use the NOX adsorber and 
CDPF systems needed to meet the proposed standards. We have assumed in 
our cost analysis that all nonroad engines complying with a PM standard 
of 0.02 g/bhp-hr or lower will have an active means to control 
temperature (i.e. we have costed a backup regeneration system, although 
some applications likely may not need one). We have made this 
assumption believing that manufacturers will not be able to accurately 
predict in-use conditions for every piece of equipment and will thus 
choose to provide the technologies on a back-up basis. As explained 
earlier, the technologies necessary to accomplish this temperature 
management are enhancements of the Tier 3 emission control technologies 
that will form the

[[Page 28379]]

baseline for Tier 4 engines, and the control strategies being developed 
for highway diesel engines. We do not believe that there are any 
nonroad engine applications above 25 horsepower for which these highway 
engine approaches will not work. However, given the diversity in 
nonroad equipment design and application, we believe that additional 
time will be needed in order to match the engine performance 
characteristics to the full range of nonroad equipment.
    We believe that given the timing of the emissions standards 
proposed today, and the availability and continuing development of 
technologies to address temperature management for highway engines 
which technologies are transferrable to all nonroad engines with 
greater than 25 hp power rating, that nonroad engines can be designed 
to meet the proposed standards in the lead time provided in this 
proposal.
b. Nonroad Operating Conditions and Durability
    Nonroad equipment is designed to be used in a wide range of tasks 
in some of the harshest operating environments imaginable, from mining 
equipment to crop cultivation and harvesting to excavation and loading. 
In the normal course of equipment operation the engine and its 
associated hardware will experience levels of vibration, impacts, and 
dust that may exceed conditions typical of highway diesel vehicles.
    Specific efforts to design for the nonroad operating conditions 
will be required in order to ensure that the benefits of these new 
emission control technologies are realized for the life of nonroad 
equipment. Much of the engineering knowledge and experience to address 
these issues already exists with the nonroad equipment manufacturers. 
Vibration and impact issues are fundamentally mechanical durability 
concerns (rather than issues of technical feasibility of achieving 
emissions reductions) for any component mounted on a piece of equipment 
(e.g., an engine coolant overflow tank). Equipment manufacturers must 
design mounting hardware such as flanges, brackets, and bolts to 
support the new component without failure. Further, the catalyst 
substrate material itself must be able to withstand the conditions 
encountered on nonroad equipment without itself cracking or failing. 
There is a large body of real world testing with retrofit emission 
control technologies that demonstrates the durability of the catalyst 
components themselves even in the harshest of nonroad equipment 
applications.
    Deutz, a nonroad engine manufacturer, sold approximately 2,000 
diesel particulate filter systems for nonroad equipment in the period 
from 1994 through 2000. Many of these systems were sold for use in 
mining equipment. No other applications are likely to be more demanding 
than this. Mining equipment is exposed to extraordinarily high levels 
of vibration, experiences impacts with the mine walls and face, and 
high levels of dust. Yet in meetings with the Agency, Deutz shared 
their experience that no system had failed due to mechanical failure of 
the catalyst or catalyst housing.\170\ The Deutz system utilized a 
conventional cordierite PM filter substrate as is commonly used for 
heavy-duty highway truck CDPF systems. The canning and mounting of the 
system was a Deutz design. Deutz was able to design the catalyst 
housing and mounting in such a way as to protect the catalyst from the 
harsh environment as evidenced by its excellent record of reliable 
function.
---------------------------------------------------------------------------

    \170\ ``Summary of Conference Call between U.S. EPA and Deutz 
Corporation on September 19, 2002 regarding Deutz Diesel Particulate 
Filter System'', EPA Memorandum to Air Docket A-2001-28.
---------------------------------------------------------------------------

    Other nonroad equipment manufacturers have also offered OEM diesel 
particulate filter systems in order to comply with requirements of some 
mining and tunneling worksite standards. Liebherr, a nonroad engine and 
equipment manufacturer, offers diesel particulate filter systems as an 
OEM option on its range of construction machine models. As of January 
2000, 340 Liebherr machines have been fitted with PM filter 
systems.\171\ We believe that this experience shows that appropriate 
design considerations, as are necessary with any component on a piece 
of nonroad equipment, will be adequate to address concerns with the 
vibration and impact conditions which can occur in some nonroad 
applications. This experience applies equally well to the 
NOX adsorber catalyst technologies as the mechanical 
properties of DOCs, CDPFs, and NOX adsorbers are all 
similar. We do not believe that any new or fundamentally different 
solutions will need to be invented in order to address the vibration 
and impact constraints for nonroad equipment. Our cost analysis 
includes the hardware costs for mounting and shrouding the 
aftertreatment equipment as well as the engineering cost for equipment 
redesign.
---------------------------------------------------------------------------

    \171\ ``Particulate Traps for Construction Machines: Properties 
and Field Experience'' J. Czerwinski et. al., Society of Automotive 
Engineers Technical Paper 2000-01-1923.
---------------------------------------------------------------------------

    Certain nonroad applications, including some forms of harvesting 
equipment and mining equipment, may have specific limits on maximum 
surface temperature for equipment components in order to ensure that 
the components do not serve as ignition sources for flammable dust 
particles (e.g. coal dust or fine crop dust). Some have suggested that 
these design constraints might limit the equipment manufacturers 
ability to install advanced diesel catalyst technologies such as 
NOX adsorbers and CDPFs. This concern seems to be largely 
based upon anecdotal experience with gasoline catalyst technologies 
where under certain circumstances catalyst temperatures can exceed 
1,000[deg]C and without appropriate design considerations could 
conceivably serve as an ignition source. We do not believe that these 
concerns are justified in the case of either the NOX 
adsorber catalyst or the CDPF technology. Catalyst temperatures for 
NOX adsorbers and CDPFs should not exceed the maximum 
exhaust manifold temperatures already commonly experienced by diesel 
engines (i.e, catalyst temperatures are expected to be below 
800[deg]C).\172\ CDPF temperatures are not expected to exceed 
approximately 700[deg]C in normal use and are expected to only reach 
the 650[deg]C temperature during periods of active regeneration. 
Similarly, NOX adsorber catalyst temperatures are not 
expected to exceed 700[deg]C and again only during periods of active 
sulfur regeneration as described in Section III.F below. Under 
conditions where diesel exhaust temperatures are naturally as high as 
650[deg]C, no supplemental heat addition from the emission control 
system will be necessary and therefore exhaust temperatures will not 
exceed their natural level. When natural exhaust temperatures are too 
low for effective emission system function then supplemental heating as 
described earlier may be necessary but would not be expected to produce 
temperatures higher than the maximum levels normally encountered in 
diesel exhaust. Furthermore, even if it were necessary to raise exhaust 
temperatures to a higher level in order to promote effective emission 
control, there are technologies available to isolate the higher exhaust

[[Page 28380]]

temperatures from flammable materials such as dust. One approach would 
be the use of air-gapped exhaust systems (i.e., an exhaust pipe inside 
another concentric exhaust pipe separated by an air-gap) that serve to 
insulate the inner high temperature surface from the outer surface 
which could come into contact with the dust. The use of such a system 
may be additionally desirable in order to maintain higher exhaust 
temperatures inside the catalyst in order to promote better catalyst 
function. Another technology to control surface temperature already 
used by some nonroad equipment manufacturers is water cooled exhaust 
systems.\173\ This approach is similar to the air-gapped system but 
uses engine coolant water to actively cool the exhaust system. We do 
not believe that flammable dust concerns will prevent the use of either 
a NOX adsorber or a CDPF because catalyst temperatures are 
not expected to be unacceptably high and because remediation 
technologies exist to address these concerns. In fact, exhaust emission 
control technologies (i.e., aftertreatment) have already been applied 
on both an OEM basis and for retrofit to nonroad equipment for use in 
potentially explosive environments. Many of these applications must 
undergo Underwriters Laboratory (UL) approval before they can be 
used.\174\
---------------------------------------------------------------------------

    \172\ The hottest surface on a diesel engine is typically the 
exhaust manifold which connects the engines exhaust ports to the 
inlet of the turbocharger. The hot exhaust gases leave the engine at 
a very high temperature (800[deg]C at high power conditions) and 
then pass through the turbocharger where the gases expand driving 
the turbocharger providing work. The process of extracting work from 
the hot gases cools the exhaust gases. The exhaust leaving the 
turbocharger and entering the catalyst and the remaining pieces of 
the exhaust system is cooler (as much as 200[deg]C at very high 
loads) than in the exhaust manifold.
    \173\ ``Engine Technology and Application Aspects for 
Earthmoving Machines and Mobile Cranes, Dr. E. Brucker, Liebherr 
Machines Bulle, SA, AVL International Commercial Powertrain 
Conference, October 2001. Copy available in EPA Air Docket A-2001-
28, Docket Item  II-A-12.
    \174\ Phone conversation with Manufacturers of Emission Control 
Association (MECA), 9 April, 2003 confirming the use of emission 
control technologies on nonroad equipment used in coal mines, 
refineries, and other locations where explosion proofing may be 
required.
---------------------------------------------------------------------------

    Nonroad engines greater than 750 hp are unique in that they do not 
have direct highway equivalents. However, this does not mean that 
unique catalyst based emission control technologies need to be 
developed separately for these larger applications. Rather, larger 
engines can, and do in retrofit applications today, use multiple 
catalyst systems in a parallel configuration. As an example, a highway 
12 liter displacement in-line six cylinder engine might use a single 18 
liter CDPF, while a nonroad 24 liter displacement V12 cylinder (a vee 
engine has two rows of cylinders set at an angle to each other) engine 
would use two 18 liter CDPFs, one for each bank of the vee engine. 
Using two smaller catalysts in place of one larger catalyst can be 
easier to package and may allow for close coupling of the catalyst 
technology to the turbocharger exhaust outlet to improve temperature 
management in some applications. Today, many passenger cars and light-
duty trucks with V6 or V8 engines use individual catalysts for each 
engine bank to improve packaging and better manage temperatures.
    We agree that nonroad equipment must be designed to address durable 
performance for a wide range of operating conditions and applications 
that would not commonly be experienced by highway vehicles. We believe 
further as demonstrated by retrofit experiences around the world that 
technical solutions exist which allow catalyst-based emission control 
technologies to be applied to nonroad equipment.
3. Are the Standards Proposed for Engines of 75 hp or Higher Feasible?
    There are three primary test provisions and associated standards in 
the Tier 4 program we are proposing today. These are the proposed 
Nonroad Transient Cycle (NRTC), the existing ISO C1 steady-state cycle, 
and the proposed highway based Not-To-Exceed (NTE) provisions. A 
nonroad diesel engine meeting the proposed standards for each of these 
three test cycles would be lawful for use in any kind of nonroad 
equipment. Additionally, we have alternative optional test cycles 
including the proposed Constant Speed Variable Load (CSVL) cycle, the 
existing ISO-D2 steady-state cycle and the proposed Transportation 
Refrigeration Unit (TRU) cycle which a manufacturer can choose to use 
for certification provided that the manufacturer can demonstrate to the 
Agency that the engine will only be used in a limited range of nonroad 
equipment with specifically defined operating conditions. Compliance on 
the proposed transient test cycles includes weighting the results from 
a cold start and hot start test with the cold start emissions weighted 
at 1/10 and hot start emissions weighted at 9/10. A complete discussion 
of these various test cycles can be found in chapter 4.2 and 4.3 of the 
draft RIA.
    The standards proposed today for nonroad engines with rated power 
greater than or equal to 75 horsepower are based upon the technologies 
and standards for highway diesel engines which go into effect in 2007. 
As explained above, we believe these technologies, namely 
NOX adsorbers and catalyzed diesel particulate filters 
enabled by 15 ppm sulfur diesel fuel, can be applied to nonroad diesel 
engines in a similar manner as for highway diesel engines. We 
acknowledge that there are additional constraints on nonroad diesel 
engines which must be considered in setting these standards, and we 
have addressed those issues by allowing for additional lead time or 
slightly less stringent standards for nonroad diesel engines in 
comparison to highway diesel engines (and likewise have made 
appropriate cost estimates to account for the technology and 
engineering needed to address these constraints).
    We have proposed a PM standard for engines in this category of 0.01 
g/bhp-hr based upon the emissions reductions possible through the 
application of a CDPF and 15 ppm sulfur diesel fuel. This is the same 
emissions level as for highway diesel engines in the HD2007 program. 
While baseline soot (the solid carbon fraction of PM) emission levels 
may be somewhat higher for some nonroad engines when compared to 
highway engines, these emissions are virtually eliminated (reduced by 
99 percent) by the CDPF technology. As discussed previously, the 
baseline (engine-out) SOF emissions levels may also need to be reduced 
through the application of modern piston ring pack designs and valve 
stem seals. With application of the CDPF technology, the SOF portion of 
diesel PM is predicted to be all but eliminated. The primary emissions 
from a CDPF equipped engine are sulfate PM emissions formed from sulfur 
in diesel fuel. The emissions rate for sulfate PM is determined 
primarily by the sulfur level of the diesel fuel and the rate of fuel 
consumption. With the 15 ppm sulfur diesel fuel the PM emissions level 
from a CDPF equipped nonroad diesel engine will be similar to the 
emissions rate of a comparable highway diesel engine. Therefore, the 
0.01 g/bhp-hr emission level is feasible for nonroad engines tested on 
the NRTC cycle and on the steady-state cycles, C1 and D2. Put another 
way, control of PM using CDPF technology is essentially independent of 
duty cycle given active catalyst technology (for reliable regeneration 
and SOF oxidation), adequate control of temperature (for reliable 
regeneration) and low sulfur diesel fuel (for reliable regeneration and 
low PM emissions).
    The most challenging PM emissions control conditions for a CDPF are 
encountered under high engine load operation where high exhaust 
temperatures promote conversion of sulfur in diesel fuel to sulfate PM 
emissions. Under these high load conditions, soot and SOF oxidation 
rates will be very high and control of those portions of PM emissions 
will be highly effective. Sulfate PM emissions, however, will be higher 
than for other operating conditions. In a worst case scenario, where 
all of the sulfur is

[[Page 28381]]

converted to sulfate, it could be perhaps as high as 0.02 g/bhp-
hr.\175\ This level of PM emissions would comply with our proposed NTE 
provisions once consideration is given to the 1.5 times multiplier on 
the emission standard for NTE test conditions.\176\ Since this estimate 
is made at a worst case condition (assuming 100% conversion of sulfur 
to sulfate), we feel confident that the PM NTE provisions of this 
proposal can be met.
---------------------------------------------------------------------------

    \175\ An estimate of the maximum sulfate PM emissions rate can 
be made by assuming a fuel consumption rate (e.g., 0.5 lbm/bhp-hr), 
the fuel sulfur level (e.g., 15 ppm) and the sulfur to sulfate 
conversion (e.g., 100% maximum) resulting in a calculated sulfate PM 
emissions rate of 0.02 g/bhp-hr. This represents a worst case 
analysis (100% sulfur conversion with 15 ppm sulfur fuel). In-use 
emissions would be significantly lower.
    \176\ The PM standard is expressed to two significant digits 
0.01 g/bhp-hr, so when the 1.5 NTE multiplier is applied, the NTE 
limit becomes 0.015 which is rounded to two significant figures as 
0.02 g/bhp-hr.
---------------------------------------------------------------------------

    Under contract from the California Air Resources Board, two nonroad 
diesel engines were recently tested for PM emissions performance with 
the application of a CDPF over a number of transient and steady-state 
test cycles.\177\ The first engine is a 1999 Caterpillar 3408 (480 hp, 
18 liter displacement) nonroad diesel engine certified to the Tier 1 
standards. The engine was tested with and without a CDPF on 12 ppm 
sulfur diesel fuel. The transient emission results for this engine are 
summarized in Table III.E-1 below. The steady-state emission results 
are summarized in Table III.1-2. The test results confirm the excellent 
PM control performance realized by a CDPF with low sulfur diesel fuel 
across a wide range of nonroad operating cycles in spite of the 
relatively high engine-out PM emissions from this Tier 1 engine. We 
would expect engine-out PM emissions to be lower for production Tier 3 
compliant diesel engines that will form the technology baseline for 
Tier 4 engines meeting the proposed standard. The engine demonstrated 
PM emissions of 0.009 g/bhp-hr on the proposed Nonroad Transient Cycle 
(NRTC) from an engine-out level of 0.256 g/bhp-hr, a reduction of 0.247 
g/bhp-hr. The engine also demonstrated excellent PM performance on the 
existing steady-state ISO C1 cycle with PM emissions of 0.010 g/bhp-hr 
from an engine-out level of 0.127, a reduction of 0.107 g/bhp-hr. Thus 
this engine would be compliant with the proposed PM emission standard 
for =75 hp variable speed nonroad engines.
---------------------------------------------------------------------------

    \177\ Application of Diesel Particulate Filters to Three Nonroad 
Engines--Interim Report, January 2003. Copy available in EPA Air 
Docket A-2001-28.
---------------------------------------------------------------------------

    When tested on the proposed optional constant speed variable load 
cycle (CSVL) (a test to which this engine would not be subject to under 
this proposal) the engine-out PM emission levels were 0.407 g/bhp-hr 
and were reduced to 0.016 g/bhp-hr (a reduction of 0.391 g/bhp-hr) with 
the addition of the PM filter. As tested this engine would not be 
compliant with the proposed optional CSVL standard, but this is not 
surprising given that this Tier 1 engine was designed for variable 
speed engine operation and not for single speed operation. We have 
great confidence given the substantial PM reduction realized in this 
testing over the proposed CSVL cycle with a CDPF that a properly 
designed nonroad diesel engine will be able to meet the standard of 
0.01 g/bhp-hr.
[GRAPHIC] [TIFF OMITTED] TP23MY03.004

    Table III.E-1 also shows results over a large number of additional 
test cycles developed from real world in-use test data to represent 
typical operating cycles for different nonroad equipment applications 
(see chapter 4.2 of the draft RIA for information on these test 
cycles). These test cycles are not used for regulatory purposes, 
although the information from these cycles was used in developing the 
proposed NRTC. The results show that the CDPF technology is highly 
effective to control in-use PM emissions over any number of disparate 
operating conditions. Remembering that the base Tier 1 engine was not 
designed to meet a transient PM standard, the CDPF emissions 
demonstrated here

[[Page 28382]]

show that very low emission levels are possible even when engine-out 
emissions are exceedingly high (e.g., a reduction of 0.558 g/bhp-hr is 
demonstrated on the AW2 cycle).
    The results summarized in the two tables are also indicative of the 
feasibility of the proposed NTE provisions of this rulemaking. In spite 
of the Tier 1 baseline of this engine, there are only three test 
results with emissions higher than the permissible limit for the 
proposed NTE. The first in Table III.E-1 shows PM emissions of 0.031 
over the AW2 cycle but from a very high baseline level of nearly 0.6 g/
bhp-hr. We believe that simple improvements to the engine-out PM 
emissions as needed to comply with the Tier 2 emission standard would 
reduce these emission below the 0.02 level required by the standard. 
There are two other test points in Table III.E-2 which are above the 
proposed NTE emission level, both at 10 percent engine load. However, 
both are outside the NTE zone which excludes emissions for engine loads 
below 30 percent. It is important to note that although the engine 
would not be constrained to meet the NTE under these conditions, the 
resulting reductions at both points are still substantial in excess of 
96 percent.

                 Table III.E-2--Steady-State PM Emissions from a Tier 1 NR Diesel Engine w/CDPF
----------------------------------------------------------------------------------------------------------------
                                   1999 (Tier 1) Caterpillar 3408 (480hp, 181)
-----------------------------------------------------------------------------------------------------------------
                                                             PM ([g/bhp-hr]
   Engine speed  %         Engine load  %    ----------------------------------------------     Reduction  %
                                                    Engine out               w/CDPF
----------------------------------------------------------------------------------------------------------------
               100                    100                  0.059                   0.10                    83
               100                     75                  0.103                  0.009                    91
               100                     50                  0.247                  0.012                    95
               100                     25                  0.247                  0.000                   100
               100                     10                  0.925                  0.031                    97
                60                    100                  0.028                  0.011                    61
                60                     75                  0.138                  0.009                    93
                60                     50                  0.180                  0.010                    95
                60                     25                  0.370                  0.007                    98
                60                     10                  0.801                  0.018                    98
                91                     82                  0.091                  0.006                    93
                80                     63                  0.195                  0.008                    96
                63                     40                  0.240                  0.008                    97
                 0                      0     .....................  .....................  ....................
                                    (\1\)                  0.127                  0.011                   91
----------------------------------------------------------------------------------------------------------------
ISO C1 Composite.

    The second engine tested was a prototype engine developed at 
Southwest Research Institute (SwRI) under contract to EPA.\178\ The 
engine, dubbed Deere Development Engine 4045 (DDE-4045) because the 
prototype engine was based on a John Deere 4045 production engine, was 
also tested with a CDPF from a different manufacturer on the same 12 
ppm diesel fuel. The engine is very much a prototype and experienced a 
number of part failures during testing, including to the turbocharger 
actuator. Nevertheless, the transient emission results summarized in 
Table III.E-3 and the steady-state results summarized in Table III.E-4 
show that substantial PM reductions are realized on this engine as 
well. The emission levels on the NRTC and the ISO C1 cycle would be 
compliant with the proposed PM standard of 0.01 g/bhp-hr once the 
appropriate rounding convention was applied.\179\ It is also 
interesting to note that the highway FTP transient emissions are higher 
than for either of the proposed nonroad transient tests. This suggests 
that developing PM compliant engines on the proposed nonroad transient 
cycles may not be substantially different from developing compliant 
technologies for highway engines. Our analysis of exhaust temperature 
characteristics for NOX adsorber catalysts discussed in the 
preceding section, noted a similar trend for NOX 
technologies (i.e., that the exhaust temperature characteristics of the 
NRTC may be better matched catalyst technologies than the HD FTP).
---------------------------------------------------------------------------

    \178\ ``Nonroad Diesel Emission Standards--Staff Technical 
Paper'', EPA Publication EPA420-R-01-052, October 2001. Copy 
available in EPA Air Docket A-2001-28.
    \179\ The rounding procedures in ASTM E29-90 are applied to the 
emission standard, therefore, the emission results are rounded to 
the same number of significant digits as the specified standard, 
i.e., 0.014 g/bhp-hr is rounded to 0.01 g/bhp-hr, while 0.015 g/bhp-
hr would be rounded to 0.02 g/bhp-hr.

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

[[Page 28383]]

[GRAPHIC] [TIFF OMITTED] TP23MY03.005

    As with the results from the Caterpillar engine, the two low-load 
(10 percent load) steady-state emissions points have some of the 
highest brake specific emission rates. These rates are not high enough, 
however, to preclude compliance with the steady-state emission cycle, 
are not within the proposed NTE zone, and still show substantial PM 
reduction levels.
[GRAPHIC] [TIFF OMITTED] TP23MY03.006


[[Page 28384]]


    While the resulting PM emission levels for nonroad diesel engines 
are similar to the levels for highway diesel engines, the challenge of 
ensuring soot regeneration of the CDPF may be more difficult for some 
nonroad equipment types. As explained earlier, effective regeneration 
occurs when the aggregate soot rate into the CDPF over an extended 
period is less than or equal to the soot oxidation rate over the same 
period. Because the baseline PM soot rate into the CDPF level may be 
higher for some nonroad engines and because the average exhaust 
temperature may be lower for some operating cycles, additional engine 
and aftertreatment system development will be needed for some nonroad 
engines. These additional developments include improved thermal 
management and improved active back-up systems which can periodically 
raise exhaust temperatures in order to initiate regeneration. We expect 
these systems to be evolutionary advancements based primarily on the 
core technologies used by nonroad manufacturers to comply with the Tier 
3 emission standards with enhancements from the highway technologies 
developed to comply with the HD2007 standards. The implementation dates 
for the standards proposed today were selected in part based upon the 
time we believe will be necessary to transfer and further develop these 
highway technologies to nonroad diesel engines and equipment.
    We are proposing a NOX standard of 0.3 g/bhp-hr for 
engines in this category based upon the emission reductions possible 
from the application of NOX adsorber catalysts and the 
expected emission levels for Tier 3 compliant engines which form the 
baseline technology for Tier 4 engines. The Tier 3 emission standards 
are a combined NOX+NMHC standard of 3.0 g/bhp-hr for engines 
greater than 100 hp and less than 750 horsepower. For engines less than 
100 hp but greater than 50 horsepower the Tier 3 NOX+NMHC 
emission standard is 3.5 g/bhp-hr. For engines greater than 750 
horsepower there is no Tier 3 NOX+NMHC standard. We believe 
that in the time-frame of the Tier 4 emission standards proposed today, 
all engines of 75 horsepower or higher can be developed to control 
NOX emissions to engine-out levels of 3.0 g/bhp-hr or lower. 
This means that all engines will need to apply Tier 3 emission control 
technologies (i.e., turbochargers, charge-air-coolers, electronic fuel 
systems, and for some manufacturers EGR systems) to get to this 
baseline level, even those engines without a Tier 3 standard (i.e., 
750hp engines). As discussed in more detail in the draft 
RIA, our analysis of the NRTC and the ISO C1 cycles indicates that the 
NOX adsorber catalyst can provide a 90 percent or greater 
NOX reduction level on the cycles. The proposed standard of 
0.3 g/bhp-hr reflects a baseline emissions level of 3.0 g/bhp-hr and a 
90 percent or greater reduction of NOX emissions through the 
application of the NOX adsorber catalyst. The additional 
lead time available to nonroad engine manufacturers and the substantial 
learning that will be realized from the introduction of these same 
technologies to highway diesel engines, plus the lack of any 
fundamental technical impediment, makes us confident that the proposed 
NOX standards can be met.
    The proposed standard is 50 percent higher than the corresponding 
HD2007 standard of 0.2 g/bhp-hr because of the higher baseline 
NOX emissions for Tier 3 engines. The higher baseline 
(engine-out) NOX level is due primarily to a lack of ram-air 
for improved charge-air cooling for nonroad diesel engines when 
compared to highway diesel engines compliant with the 2004 highway 
emission standards. Although nonroad engine manufacturers may be able 
to lower engine-out NOX emissions below the levels required 
for Tier 3, we continue to expect that the lack of ram air will limit 
nonroad engine-out NOX performance, and therefore we have 
accounted for that difference by proposing this higher NOX 
emissions level.
    We believe that the NOX adsorber technology developed 
for highway engines can be applied with equal effectiveness to nonroad 
diesel engines with additional developments in engine thermal 
management (as discussed in section III.E.2 above) to address the more 
widely varied nonroad operating cycles. In fact, as discussed 
previously, the NOX adsorber catalyst temperature window is 
particularly well matched to transient operating conditions as typified 
by the NRTC.
    Compliance with the NTE provisions proposed today will be 
challenging for the nonroad engine industry due to the diversity of 
nonroad products and operating cycles. However, the technical challenge 
is reduced somewhat by the 1.5 multiplier used to calculate the NTE 
standard. Controlling NOX emissions under NTE conditions is 
fundamentally similar for both highway and nonroad engines. The range 
of control is the same and the amount of reduction required is also the 
same. We know of no technical impediment that would prevent achieving 
the NTE standard under the full range of operating conditions.
    The proposed NOX standard is phased in over a number of 
years in a manner similar to the HD2007 NOX phase-in. In the 
early years of the program half of the engines produced by a 
manufacturer must be certified to the new emission standard while the 
remaining engines can continue to be sold at the previous standard. We 
provided this phase-in period for highway engines in the HD2007 
rulemaking to allow manufacturers to focus resources on the portion of 
their products best suited to NOX catalysts first and then 
to apply the learning to the remainder of their products three years 
later.\180\ Provisions of the averaging program in the HD2007 
rulemaking allow manufacturers to alternatively comply with some engine 
families at an ``averaged'' standard that is approximately halfway 
between the old and new NOX standards. In fact, we have 
learned from a number of engine manufacturers that they are likely to 
employ this strategy for some fraction of their new highway engines in 
2007. The averaging provisions that we have proposed today for Tier 4 
would also allow for compliance with the proposed Tier 4 NOX 
standard with a single engine product during the transitional 
NOX phase-in period. This provision allows manufacturers to 
transfer the same highway NOX technologies to nonroad 
engines and to comply with an appropriately stringent standard. We 
believe as with the HD2007 rule that this provision is necessary in 
order to manage resource requirements to develop the necessary 
technologies and that this provision provides significant additional 
flexibility for manufacturers to comply with the proposed 
NOX standards. Similarly, we have proposed a modified phase-
in schedule for the greater than 750 horsepower engines in part because 
of the lack of a Tier 3 standard for those engine and the extra work 
required to develop a full Tier 4 emission control system from a Tier 2 
baseline.
---------------------------------------------------------------------------

    \180\ Control of Air Pollution from New Motor Vehicles: Heavy-
duty Engine and Vehicle Standards and Highway Diesel Sulfur Control 
Requirements; Final Rule, 66 FR 5002, January 18, 2001.
---------------------------------------------------------------------------

    Meeting the proposed NMHC standard under the lean operating 
conditions typical of the biggest portion of NOX adsorber 
operation should not present any special challenges to nonroad diesel 
engine manufacturers. Since CDPFs and NOX adsorbers contain 
platinum and other precious metals to oxidize NO to NO2, 
they are also very efficient oxidizers of hydrocarbons. NMHC reductions 
of greater than 95 percent have been shown over transient

[[Page 28385]]

and steady-state test procedures.\181\ Given that typical engine-out 
NMHC is expected to be in the 0.40 g/bhp-hr range or lower for engines 
meeting the Tier 3 standards, this level of NMHC reduction will mean 
that under lean conditions emission levels will be well below the 
standard.
---------------------------------------------------------------------------

    \181\ ``The Impact of Sulfur in Diesel Fuel on Catalyst Emission 
Control Technology,'' report by the Manufacturers of Emission 
Controls Association, March 15, 1999, pp. 9 & 11. Copy available in 
EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    The NOX regeneration strategies for the NOX 
adsorber technology may prove difficult to control precisely, leading 
to a possible increase in NMHC emissions under the rich operating 
conditions required for NOX regeneration. Even with precise 
control of the regeneration cycle, NMHC slip may prove to be a 
difficult problem due to the need to regenerate the NOX 
adsorber under net rich conditions (excess fuel) rather than the 
stoichiometric (fuel and air precisely balanced) operating conditions 
typical of a gasoline three-way catalyst. It seems possible therefore, 
that in order to meet the NMHC standards we have proposed, an 
additional clean up catalyst may be required. A diesel oxidation 
catalyst, like those applied historically for NMHC and partial PM 
control, can reduce NMHC emissions (including toxic HCs) by more than 
90 percent.\182\ This amount of additional control along with optimized 
NOX regeneration strategies will ensure very low NMHC 
emissions. Our cost analysis described in section V includes the cost 
for the application of a clean-up DOC catalyst for all engines which 
must comply with the 0.3 g/bhp-hr NOX standard.
---------------------------------------------------------------------------

    \182\ ``Demonstration of Advanced Emission Control Technologies 
Enabling Diesel-Powered Heavy-Duty Engines to Achieve Low Emission 
Levels'', Manufacturers of Emissions Controls Association, June 
1999. Copy available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    Test results from a prototype integrated NOX/PM and NMHC 
control system for diesel engines documented in the draft RIA show that 
NMHC emissions can be controlled below 0.14 g/bhp-hr under transient 
and steady-state test conditions for highway diesel engines while 
simultaneously controlling NOX emissions below 0.2 g/bhp-hr 
and PM emissions below 0.01 g/bhp-hr. Since the slip of hydrocarbon 
emissions are predominantly a function of the NOX 
regeneration event and not engine transient events, the level of 
control demonstrated in this testing is expected to be the same for 
other operating conditions as represented by the proposed NRTC cycle 
and the NTE provisions of this rulemaking. Based on our engineering 
judgement and experience testing integrated NOX adsorber and 
PM filter systems with DOC clean-up catalyst technologies, we can 
conclude that the 0.14 g/bhp-hr NMHC standard will be feasible in the 
Tier 4 time frame.
    The proposed standards include a cold start provision with the 
transient test procedures. This means that the results of a cold start 
transient test will be weighted with the emissions of a hot start test 
in order to calculate the emissions for compliance against the proposed 
standards. The proposed weightings are 1/10 cold start and 9/10 for the 
hot start as described more fully in chapter 4.2 of the draft RIA. 
Because exhaust temperatures are so important to catalyst performance 
the cold start provision is an important tool to ensure that the 
emissions realized in use are consistent with the expectations of this 
program and represents an additional technical challenge for 
NOX control and to a lesser extent CO and NMHC control. PM 
control with a CDPF is not expected to be significantly impacted by 
cold-start provisions. NOX control in the period before 
temperatures exceed the catalyst light-off temperature are reduced 
significantly. As a result, exhaust stack NOX emissions will 
be higher over the cold start portion of the test. However, we believe 
that this increase in NOX emissions will not be high enough 
to preclude compliance with the proposed NOX standard once 
the 1/10 weighting is applied.
    There are a number of technologies available to the engine 
manufacturer to promote rapid warmup of the exhaust and emission 
control system. These include retarding injection timing, increasing 
EGR, and potentially late cycle injection all of which are technologies 
we expect manufacturers to apply as part of the normal operation of the 
NOX adsorber catalyst system. These are the same 
technologies we expect highway engine manufacturers to use in order to 
comply with the highway cold start FTP provision which weights cold 
start emissions more heavily with a 1/7 weighting. As a result, we 
expect the transfer of highway technology to be well matched to 
accomplish this control need for nonroad engines as well. Using these 
technologies we expect nonroad engine manufacturers to be able to 
comply with the proposed NOX, NMHC and CO emissions 
including the cold start provisions of the transient test procedure.
    We did not set new Tier 3 emission standards for 750 hp 
nonroad engines in the 1998 Tier \2/3\ rulemaking because of the long 
lead time we believed appropriate, given the long product redesign 
cycles typical of these large engines and their low sales volumes. The 
Tier 2 standards set in that rulemaking for 750 hp engines 
do not go into effect until 2006. We reasoned in the Tier \2/3\ rule 
that the uncertainties involved in setting a Tier 3 standard for 
750hp nonroad engines that wouldn't go into effect before 
2010 would be too large. Therefore, we deferred setting new standards 
for these engines at that time. Given new technology enabled by low 
sulfur diesel fuel, we believe that it is now appropriate to project 
the technologies which will be available for these engines in the 
future (i.e., CDPFs and NOX adsorbers) and to set new 
standards accordingly.
    Although we have proposed a unique phase-in schedule for 
750hp engines as explained in section III.B, we do not doubt 
that these engines, like engines <750hp, can be developed to meet the 
standards proposed today. These large engines are fundamentally similar 
to other nonroad engines. The project emissions control mechanisms are 
the same. Retrofits of PM filter systems have been applied to large 
locomotives and other similar size engines. We are unaware of any 
fundamental difference in technology function that would lead us to 
conclude that the proposed standards are inappropriate for engines 
750hp. However, given the need to apply both new engine-out 
control technologies (i.e., Tier 3 type technologies) in addition to 
the new catalyst based technologies in order to comply with the 
proposed standards, and given the low sales volumes for these engines, 
we do believe it is appropriate to have a different phase-in structure 
for these engines. We invite comment supported by data on this issue, 
particularly if a commenter believes there are fundamental technology 
differences which would make alternate standards more appropriate for 
750hp nonroad engines.
    The standards that we have proposed today for nonroad engines with 
rated horsepower levels =75 horsepower are based upon the 
same emission control technologies, clean 15ppm or lower sulfur diesel 
fuel, and relative levels of emission control effectiveness as the HD 
2007 emission standards. We have given consideration to the diversity 
of nonroad equipment for which these technologies must be developed and 
the timing of the Tier 3 emissions standards in determining the 
appropriate timing for the Tier 4 standards we have proposed today. 
Based upon the availability of the emission control technologies, the 
proven effectiveness of the technologies to control diesel emissions to 
these levels, the technology

[[Page 28386]]

paths identified here to address constraints specific to nonroad 
equipment, and the additional lead time afforded by the timing of the 
standards, we have concluded that the proposed standards are feasible.
4. Are the Standards Proposed for Engines =25 hp and <75 hp 
Feasible?
    As discussed in section III.B, our proposal for standards for 
engines between 25 and 75 hp consists of a 2008 transitional standard 
and long-term 2013 standards. The proposed transitional standard is a 
0.22 g/bhp-hr PM standard. The 2013 standards consist of a 0.02 g/bhp-
hr PM standard and a 3.5 g/bhp-hr NMHC+NOX standard. As 
discussed in section III.B, the transitional standard is optional for 
50-75 hp engines, as the proposed 2008 implementation date is the same 
as the effective date of the Tier 3 standards. Manufacturers may 
decide, at their option, not to undertake the 2008 transitional PM 
standard, in which case their implementation date for the 0.02 g/bhp-hr 
PM standard begins in 2012.
    In addition, we have proposed a minor revision to the CO standard 
for the 25-50 hp engines beginning in 2008 to align these engines with 
the 50-75 hp engines. This proposed CO standard is 3.7 g/bhp-hr.
    The remainder of this section discusses:
    [sbull] What makes the 25-75 hp category unique;
    [sbull] What engine technology is used today, and will be used for 
applicable Tier 2 and Tier 3 standards;
    [sbull] Why the proposed standards are technologically feasible; 
and,
    [sbull] Why EPA has not proposed more stringent NOX 
standards at this time for these engines.
    a. What makes the 25--75 hp category unique?
    As discussed in section III.B.1.d, many of the nonroad diesel 
engines =75 hp are either a direct derivative of highway 
heavy-duty diesel engines, or share a number of common traits with 
highway diesel engines. These include similarities in displacement, 
aspiration, fuel systems, and electronic controls. Table III.E-3 
contains a summary of a number of key engine parameters from the 2001 
engines certified for sale in the U.S.\183\
---------------------------------------------------------------------------

    \183\ Data in Table III.E-3 is derived from a combination of the 
publically available certification data for model year 2001 engines, 
as well as the manufacturers reported estimates of 2001 production 
targets, which is not public information.

                Table III.E-3: Summary of Model Year 2001 Key Engine Parameters by Power Category
----------------------------------------------------------------------------------------------------------------
                                                                Percent of 2001 U.S. Production \a\
                                                 ---------------------------------------------------------------
                Engine Parameter                                                                  100
                                                      0-25 hp        25-75 hp        75-100 hp          hp
----------------------------------------------------------------------------------------------------------------
IDI Fuel System.................................             83%             47%              4%           <0.1%
DI Fuel System..................................             17%             53%             96%  99%
Turbocharged....................................              0%              7%             62%             91%
1 or 2 Cylinder Engines.........................             47%              3%              0%              0%
Electronic fuel systems (estimated).............   not available         limited    availability        commonly
                                                           today       available           today       available
                                                                           today                          today
----------------------------------------------------------------------------------------------------------------
Notes:
\a\ Based on sales weighting of 2001 engine certification data.

    As can be seen in Table III.E-3, the engines in the 25-75 hp 
category have a number of technology differences from the larger 
engines. These include a higher percentage of indirect-injection fuel 
systems, and a low fraction of turbocharged engines. (The distinction 
in the <25 hp category is quite different, with no turbocharged 
engines, nearly one-half of the engines have two cylinders or less, and 
a significant majority of the engines have indirect-injection fuel 
systems.)
    The distinction is particularly marked with respect to 
electronically controlled fuel systems. These are commonly available in 
the = 75 hp power categories, but, based on the available 
certification data as well as our discussions with engine 
manufacturers, we believe there are very limited numbers, if any, in 
the 25-75 hp category (and no electronic fuel systems in the less than 
25 hp category). The research and development work being performed 
today for the heavy-duty highway market is targeted at engines which 
are 4-cylinders or more, direct-injection, electronically controlled, 
turbocharged, and with per-cylinder displacements greater than 0.5 
liters. As discussed in more detail below, as well as in section 
III.E.5 (regarding the <25 hp category), these engine distinctions are 
important from a technology perspective and warrant a different set of 
standards for the 25-75 hp category (as well as for the <25 hp 
category).
b. What Engine Technology Is Used Today, and Will Be Used for the 
Applicable Tier 2 and Tier 3 Standards?
    In the 1998 nonroad diesel rulemaking, we established Tier 1 and 
Tier 2 standards for engines in the 25-50 hp category. Tier 1 standards 
were implemented in 1999, and the Tier 2 standards take effect in 2004. 
The 1998 rule also established Tier 2 and Tier 3 standards for engines 
between 50 and 75 hp. The Tier 2 standards take effect in 2004, and the 
Tier 3 standards take effect in 2008. The Tier 1 standards for engines 
between 50 and 75 hp took effect in 1998. Therefore, all engines in the 
25-75 hp range have been meeting Tier 1 standards for the past several 
years, and the data presented in Table III.E-3 represent performance of 
Tier 1 technology for this power range.
    As discussed in section III.E.4.a, engines in the 25-75 hp category 
use either indirect injection (IDI) or direct injection (DI) fuel 
systems. The IDI system injects fuel into a pre-chamber rather than 
directly into the combustion chamber as in the DI system.\184\ This 
difference in fuel systems results in substantially different emission 
characteristics, as well as differences in several important operating 
parameters. In general, the IDI engine has lower engine-out PM and 
NOX emissions, while the DI engine has better fuel 
efficiency and lower heat rejection.\185\
---------------------------------------------------------------------------

    \184\ See for example ``Diesel-engine Management'' published by 
Robert Bosch GmbH, 1999, second edition, pages 6-8 for a more 
detailed discussion of the differences between IDI and DI engines.
    \185\ See chapter 14, section 4 of ``Turbocharging the Internal 
Combustion Engine'', N. Watson and M.S. Janota, published by John 
Wiley and Sons, 1982.
---------------------------------------------------------------------------

    We expect a significant shift in the engine technology which will 
be used in this power category as a result of the upcoming Tier 2 and 
Tier 3 standards, in particular for the 50-75 hp engines. In the 50-75 
hp category, the 2008 Tier

[[Page 28387]]

3 standards will likely result in the significant use of turbocharging 
and electronic fuel systems, as well as the introduction of both cooled 
and uncooled exhaust gas recirculation by some engine manufacturers and 
possibly the use of charge-air-cooling.\186\ In addition, we have heard 
from some engine manufacturers that the engine technology used to meet 
Tier 3 for engines in the 50-75 hp range will also be made available on 
those engines in the 25-50 hp range which are built on the same engine 
platform. For the Tier 2 standards for the 25-50 hp products, a large 
number of engines meet these standards today, and therefore we expect 
to see only moderate changes in these engines, including the potential 
additional use of turbocharging on some models.\187\
---------------------------------------------------------------------------

    \186\ See section 2.2 through 2.3 in ``Nonroad Diesel Emission 
Standards--Staff Technical Paper'', EPA Publication EPA420-R-01-052, 
October 2001. Copy available in EPA Air Docket A-2001-28.
    \187\ See Table 3-2 in ``Nonroad Diesel Emission Standards--
Staff Technical Paper'', EPA Publication EPA420-R-01-052, October 
2001. Copy available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

c. Are the Proposed Standards for 25-75 hp Engines Technologically 
Feasible?
    This section will discuss the technical feasibility of both the 
proposed 2008 PM standard and the 2013 standards. For an explanation 
and discussion of the proposed implementation dates, please refer to 
section III.B of this this proposal.
    i. 2008 PM Standards.\188\ As just discussed in section III.E.4.b, 
engines in the 25-50 hp category must meet Tier 1 NMHC+NOX 
and PM standards today. We have examined the model year 2002 engine 
certification data for engines in the 25-50 hp category. These data 
indicate that over 10 percent of the engine families meet the proposed 
2008 0.22 g/bhp-hr PM standard and 5.6 g/bhp-hr NMHC+NOX 
standard (unchanged from Tier 2 in 2008) today. These include a variety 
of engine families using a mix of engine technologies (IDI and DI, 
turbocharged and naturally aspirated) tested on a variety of 
certification test cycles.\189\ Five engine families are more than 20 
percent below the proposed 0.22 g/bhp-hr PM standard, and an additional 
24 engine families are within 30 percent of the proposed 2008 PM 
standards while meeting the NMHC+NOX standard. A detailed 
discussion of these data is contained in the draft RIA. Unfortunately, 
similar data do not exist for engines between 50 and 75 hp. There is no 
Tier 1 PM standard for engines in this power range, and therefore 
engine manufacturers are not required to report PM emission levels 
until Tier 2 starts in 2004. However, in general, the 50-75 hp engines 
are more technologically advanced than the smaller horsepower engines 
and would be expected to perform as well as, if not better than, the 
engines in the 25-50 hp range.
---------------------------------------------------------------------------

    \188\ As discussed in section III.B., manufacturers can choose, 
at their option, to pull-ahead the 2013 PM standard for the 50-75 hp 
engines to 2012, in which case they do not need to comply with the 
transitional 2008 PM standard.
    \189\ The Tier 1 standards for this power category must be 
demonstrated on one of a variety of different engine test cycles. 
The appropriate test cycle is selected by the engine manufacturer 
based on the intended in-use application of the engine.
---------------------------------------------------------------------------

    The model year 2002 engines in this power range use well known 
engine-out emission control technologies, such as optimized combustion 
chamber design and fuel injection timing control strategies, to comply 
with the existing standards. These data have a two-fold significance. 
First, they indicate that a number of engines in this power range can 
already achieve the proposed 2008 standard for PM using only engine-out 
technology, and that other engines should be able to achieve the 
standard making improvements just to engine-out performance. Despite 
being certified to the same emission standards with similar engine 
technology, the emission levels from these engines vary widely. Figure 
III.E-1 is a graph of the model year 2002 HC+NOX and PM data 
for engines in the 25-50 hp range. As can be seen in the figure, the 
emission levels cover a wide range. Figure III.E-1 highlights a 
specific example of this wide range: engines using naturally aspirated 
DI technology and tested on the 8-mode test cycle. Even for this subset 
of DI engines achieving approximately the same HC+NOX level 
of [sim]6.5 g/bhp-hr, the PM rates vary from approximately 0.2 to more 
than 0.5 g/bhp-hr. There is limited information available to indicate 
why for these small diesel engines with similar technology operating at 
approximately the same HC+NOX level the PM emission rates 
cover such a broad range. We are therefore not predicating the proposed 
2008 PM standard on the combination of diesel oxidation catalysts and 
the lowest engine-out emissions being achieved today, because it is 
uncertain whether or not additional engine-out improvements would lower 
all engines to the proposed 2008 PM standard. Instead, we believe there 
are two likely means by which companies can comply with the proposed 
2008 PM standard. First, some engine manufacturers can comply with this 
standard using known engine-out techniques (e.g., optimizing combustion 
chamber designs, fuel-injection strategies). However, based on the 
available data it is unclear whether engine-out techniques will work in 
all cases. Therefore, we believe some engine companies will choose to 
use a combination of engine-out techniques and diesel oxidation 
catalysts, as discussed below.

[[Page 28388]]

[GRAPHIC] [TIFF OMITTED] TP23MY03.007

    For those engines which do not already meet the proposed 2008 Tier 
4 PM standard, a number of engine-out technologies are available to 
achieve the standards by 2008. In our recent Staff Technical Paper on 
the feasibility of the Tier 2 and Tier 3 standards, we projected that 
in order to comply with the Tier 3 standards, engines greater than 50 
hp would rely on some combination of a number of technologies, 
including electronic fuel systems such as electronic rotary pumps or 
common-rail fuel systems.\190\ In addition to enabling the Tier 3 
NMHC+NOX standards, electronic fuel systems with high 
injection pressure and the capability to perform pilot-injection and 
rate-shaping, have the potential to substantially reduce PM 
emissions.\191\ Even for mechanical fuel systems, increased injection 
pressures can reduce PM emissions substantially.\192\ As discussed 
above, we are projecting that the Tier 3 engine technologies used in 
engines between 50 and 75 hp, such as turbocharging and electronic fuel 
systems, will make their way into engines in the 25-50 hp range. 
However, we do not believe this technology will be required to achieve 
the proposed 2008 PM standard. As demonstrated by the 2002 
certification data, engine-out techniques such as optimized combustion 
chamber design, fuel injection pressure increases and fuel injection 
timing can be used to achieve the proposed standards for many of the 
engines in the 25-75 hp category without the need to add turbocharging 
or electronic fuel systems.
---------------------------------------------------------------------------

    \190\ See section 2.2 through 2.3 in ``Nonroad Diesel Emission 
Standards--Staff Technical Paper'', EPA Publication EPA420-R-01-052, 
October 2001. Copy available in EPA Air Docket A-2001-28.
    \191\ Ikegami, M., K. Nakatani, S. Tanaka, K. Yamane: ``Fuel 
Injection Rate Shaping and Its Effect on Exhaust Emissions in a 
Direct-Injection Diesel Engine Using a Spool Acceleration Type 
Injection System'', SAE paper 970347, 1997. Dickey D.W., T.W. Ryan 
III, A.C. Matheaus: ``NOX Control in Heavy-Duty Engines--
What is the Limit?'', SAE paper 980174, 1998. Uchida N, K. 
Shimokawa, Y. Kudo, M. Shimoda: ``Combustion Optimization by Means 
of Common Rail Injection System for Heavy-Duty Diesel Engines'', SAE 
paper 982679, 1998.
    \192\ ``Effects of Injection Pressure and Nozzle Geometry on DI 
Diesel Emissions and Performance,'' Pierpont, D., and Reitz, R., SAE 
Paper 950604, 1995.
---------------------------------------------------------------------------

    For those engines which are not able to achieve the proposed 
standards with known engine-out techniques, we project that diesel 
oxidation catalysts can be used to achieve the proposed standards. DOCs 
are passive flow-through emission control devices which are typically 
coated with a precious metal or a base-metal washcoat. DOCs have been 
proven to be durable in use on both light-duty and heavy-duty diesel 
applications. In addition, DOCs have already been used to control 
carbon monoxide on some nonroad applications.\193\
---------------------------------------------------------------------------

    \193\ EPA Memorandum ``Documentation of the Availability of 
Diesel Oxidation Catalysts on Current Production Nonroad Diesel 
Equipment'', William Charmley. Copy available in EPA Air Docket A-
2001-28.
---------------------------------------------------------------------------

    Certain DOC formulations can be sensitive to diesel fuel sulfur 
level, and depending on the level of emission reduction necessary, 
sulfur in diesel fuel can be an impediment to PM reductions. As 
discussed in section III.E.1.a, precious metal oxidation catalysts can 
oxidize the sulfur in the fuel and form particulate sulfates. However, 
even with today's high sulfur nonroad fuel, some manufacturers have 
demonstrated that a properly formulated DOC can be used to achieve the 
existing Tier 2 PM standards for larger engines (i.e., the 0.15 g/bhp-
hr standard).\194\ However, given the high level of sulfur in nonroad 
fuel today, the use of DOCs

[[Page 28389]]

as a PM reduction technology is severely limited. Data presented by one 
engine manufacturer regarding the existing Tier 2 PM standard shows 
that while a DOC can be used to meet the current standard even when 
tested on 2,000 ppm sulfur fuel, lowering the fuel sulfur level to 380 
ppm enabled the DOC to reduce PM by 50 percent from the 2,000 ppm 
sulfur fuel.\195\ Without the availability of 500 ppm sulfur fuel in 
2008, DOCs would be of limited use for nonroad engine manufacturers and 
would not provide the emissions necessary to meet the proposed 
standards for most engine manufacturers. With the availability of 500 
ppm sulfur fuel, DOC's can be designed to provide PM reductions on the 
order of 20 to 50%, while suppressing particulate sulfate reduction. 
These levels of reductions have been seen on transient duty cycles as 
well as highway and nonroad steady-state duty cycles.\196\ As discussed 
in section VII of this preamble, the 2008 PM standard must be met on 
the existing nonroad steady-state cycle, the supplemental standards 
(nonroad transient cycle and NTE) are not implemented until 2013 for 
this power category. As discussed above, 24 engine families in the 25-
50 hp range are within 30 percent of the proposed 2008 PM standard and 
are at or below the 2008 NMHC+NOX standard for this power 
range, indicating that use of DOCs should readily achieve the 
incremental improvement necessary to meet the proposed 2008 PM 
standard.
---------------------------------------------------------------------------

    \194\ See Table 2-4 in ``Nonroad Diesel Emission Standards--
Staff Technical Paper'', EPA Publication EPA420-R-01-052, October 
2001. Copy available in EPA Air Docket A-2001-28.
    \195\ See Table 2-4 in ``Nonroad Diesel Emission Standards--
Staff Technical Paper'', EPA Publication EPA420-R-01-052, October 
2001. Copy available in EPA Air Docket A-2001-28.
    \196\ ``Demonstration of Advanced Emission Control Technologies 
Enabling Diesel-Powered Heavy-duty Engines to Achieve Low Emission 
Levels: Interim Report Number 1--Oxidation Catalyst Technology, copy 
available in EPA Air Docket A-2001-28. ``Reduction of Diesel Exhaust 
Emissions by Using Oxidation Catalysts,'' Zelenka et al., SAE Paper 
90211, 1990. See Table 2-4 in ``Nonroad Diesel Emission Standards--
Staff Technical Paper'', EPA Publication EPA420-R-01-052, October 
2001, copy available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    Based on the existence of a number of engine families which already 
comply with the proposed 0.22 g/bhp-hr PM standard (and the 2008 
NMHC+NOX standard), and the availability of well known PM 
reduction technologies such as engine-out improvements and diesel 
oxidation catalysts, we project the proposed 0.22 g/bhp-hr PM standards 
is technologically feasible by model year 2008. All of these are 
conventional technologies which have been used on both highway and 
nonroad diesel engines in the past. As such, we do not expect there to 
be any negative impacts with respect to noise or safety. In addition, 
PM reduction technologies such as improved combustion through the use 
of higher pressure fuel injection systems have the potential to improve 
fuel efficiency. DOCs are not predicted to have any substantial impact 
on fuel efficiency.
    As discussed in section III.B, we have also proposed a minor change 
in the CO standard for the 25-50 hp engines, in order to align it with 
the standard for the 50-75 hp engines. As discussed in section III.B., 
this small change in the CO standard is intended to simplify EPA's 
regulations as part of our decision to propose a reduction in the 
number of engine power categories for Tier 4. The current CO standard 
for this category is 4.1 g/bhp-hr, and the proposed standard is 3.7 g/
bhp-hr (i.e., the current standard for engines in the 50-75 hp range). 
The model year 2002 certification data shows that more than 95 percent 
of the engine families in the 25-50 hp engine range meet the proposed 
CO standard today. In addition, a recent EPA test program run by a 
contractor on two nonroad diesel engines in this power range showed 
that CO emissions were well below the proposed standards not only when 
tested on the existing steady-state 8-mode test procedure, but also 
when tested on the nonroad transient duty cycle we are proposing in 
today's action.\197\ Finally, DOCs typically reduce CO emissions on the 
order of 50 percent or more, on both transient and steady-state 
conditions.\198\ Given that more than 95 percent of the engines in this 
category meet the proposed standard today, and the ready availability 
of technology which can easily achieve the proposed standard, we 
project this CO standard will be achievable by model year 2008.
---------------------------------------------------------------------------

    \197\ See Tables 6, 8, and 14 of ``Nonroad Emission Study of 
Catalyzed Particulate Filter Equipped Small Diesel Engines' 
Southwest Research Institute, September 2001. Copy available in EPA 
Air Docket A-2001-28.
    \198\ ``Demonstration of Advanced Emission Control Technologies 
Enabling Diesel-Powered Heavy-duty Engines to Achieve Low Emission 
Levels: Interim Report Number 1--Oxidation Catalyst Technology and 
``Reduction of Diesel Exhaust Emissions by Using Oxidation 
Catalysts'', P. Zelenka et al., Society of Automotive Engineers 
paper 902111, October 1990.
---------------------------------------------------------------------------

ii. 2013 Standards
    For engines in the 25-50 range, we are proposing standards 
commencing in 2013 of 3.5 g/bhp-hr for NMHC+NOX and 0.02 g/
bhp-hr for PM. For the 50-75 hp engines, we are proposing a 0.02 g/bhp-
hr PM standard which will be implemented in 2013, and for those 
manufacturers who choose to pull-ahead the standard one-year, 2012 
(manufacturers who choose to pull-ahead the 2013 standard for engine in 
the 50-75 range do not need to comply with the transitional 2008 PM 
standard).

PM Standard

    Sections III.E.1 through III.E.3 have already discussed catalyzed 
diesel particulate filters, including explanations of how CDPFs reduce 
PM emissions, and how to apply CDPFs to nonroad engines. We concluded 
there that CDPFs can be used to achieve the proposed PM standard for 
engines =75 hp. As also discussed in section III.E.2.a, PM 
filters will require active back-up regeneration systems for many 
nonroad applications above and below 75 hp because low temperature 
operation is an issue across allpower categories. A number of secondary 
technologies are likely required to enable proper regeneration, 
including possibly electronic fuel systems such as common rail systems 
which are capable of multiple post-injections which can be used to 
raise exhaust gas temperatures to aid in filter regeneration.
    Particulate filter technology, with the requisite trap regeneration 
technology, can also be applied to engines in the 25 to 75 hp range. 
The fundamentals of how a filter is able to reduce PM emissions as 
described in section III.E.1. are not a function of engine power, and 
CDPF's are just as effective at capturing soot emissions and oxidizing 
SOF on smaller engines as on larger engines. As discussed in more 
detail below, particulate sulfate generation rates are slightly higher 
for the smaller engines, however, we have addressed this issue in our 
proposal. The PM filter regeneration systems described in section 
III.E.1 and 2 are also applicable to engines in this size range and are 
therefore likewise feasible. There are specific trap regeneration 
technologies which we believe engine manufacturers in the 25-75 hp 
category may prefer over others. Specifically, an electronically-
controlled secondary fuel injection system (i.e., a system which 
injects fuel into the exhaust upstream of a PM filter). Such a system 
has been commercially used successfully by at least one nonroad engine 
manufacturer, and other systems have been tested by technology 
companies.\199\
---------------------------------------------------------------------------

    \199\ ``The Optimized Deutz Service Diesel Particulate Filter 
System II'', H. Houben et al., SAE Technical Paper 942264, 1994 and 
``Development of a Full-Flow Burner DPF System for Heavy Duty Diesel 
Engines, P. Zelenka et al., SAE Technical Paper 2002-01-2787, 2002.
---------------------------------------------------------------------------

    We are, however, proposing a slightly higher PM standard (0.02 g/
bhp-hr rather than 0.01) for these engines. As discussed in section 
III.E.1.a, with the

[[Page 28390]]

use of a CDPF, the PM emissions emitted by the filter are primarily 
derived from the fuel sulfur. The smaller power category engines tend 
to have higher fuel consumption than larger engines. This occurs for a 
number of reasons. First, the lower power categories include a high 
fraction of IDI engines which by their nature consume approximately 15 
percent more fuel than a DI engine. Second, as engine displacements get 
smaller, the engine's combustion chamber surface-to-volume ratio 
increases. This leads to higher heat-transfer losses and therefor lower 
efficiency and higher fuel consumption. In addition, frictional losses 
are a higher percentage of total power for the smaller displacement 
engines which also results in higher fuel consumption. Because of the 
higher fuel consumption rate, we expect a higher particulate sulfate 
level, and therefore we have proposed a 0.02 g/bhp-hr standard.
    Test data confirm that this proposed standard is achievable. In 
2001, EPA completed a test program run by a contractor on two small 
nonroad diesel engines (a 25 hp IDI engine and a 50 hp IDI engine) 
which demonstrated the proposed 0.02 g/bhp-hr standard can be achieved 
with the use of a CDPF.\200\ This test program included testing on the 
existing 8-mode steady-state test cycle as well as the nonroad 
transient cycle proposed in today's action. The 0.02g/bhp-hr level was 
achieved on each engine over both test cycles. One of the engines was 
also tested on the proposed constant speed, variable load transient 
cycle with a particulate filter, and this engine also met the proposed 
0.02 g/bhp-hr PM standard.\201\ This test program also demonstrates why 
EPA has proposed a slightly higher PM standard for the 25-75 hp 
category (0.02 g/bhp-hr vs 0.01). The data from the test program 
described above showed fuel consumption rates over the 8-mode test 
procedure between 0.4 and 0.5 lbs/bhp-hr, while typical values for a 
modern turbocharged DI engine with 4-valves per cylinder in the 
=75 hp categories are on the order of 0.3 to 0.35 lbs/hp-hr. 
However, the data is less conclusive with respect to the proposed NTE 
standard. The test program at SwRI included a number of individual 
steady-state emission points which are within the proposed NTE control 
zone for nonroad diesel engines. For most of these points, the 
emissions were well below the proposed NTE standard for both engines. 
However, both engines included as a test point the maximum torque test 
point, and in each case the emissions were above the proposed NTE 
standard. For one engine, the engine-out emissions were 1.2 g/bhp-hr PM 
and when equipped with a CDPF the emissions were 0.05 g/bhp-hr. While 
this is more than a 95 percent reduction in PM, 0.05 is above our 
proposed NTE standard of 0.03 g/bhp-hr. The second test engine at the 
maximum torque mode produced an engine-out PM value of 0.35 g/bhp-hr, 
and when equipped with a CDPF the results were 0.04g/bhp-hr. While this 
is nearly a 90 percent reduction in PM, the engines do not meet the 
proposed NTE standard. We believe these results are a combination of 
high engine-out PM emissions as well as high exhaust gas temperature. 
While a CDPF is very effective at reducing PM emissions, it is not 100 
percent effective. These engines would likely require additional 
engine-out PM reductions at the maximum torque mode in order to comply 
with the proposed NTE standard. In addition, the peak torque mode is 
one of the highest exhaust gas temperature mode, and therefore one of 
the highest particulate-sulfate generating modes when equipped with a 
CDPF. More careful management of the engine-out temperature at this 
mode, such as by altering the engines air-fuel ratio, may be necessary 
to lower the engine-out temperature and comply with the proposed NTE 
standard.
---------------------------------------------------------------------------

    \200\ See Tables 6, 8, and 14 of ``Nonroad Emission Study of 
Catalyzed Particulate Filter Equipped Small Diesel Engines'' 
Southwest Research Institute, September 2001. Copy available in EPA 
Air Docket A-2001-28.
    \201\ See Tables 8 of ``Nonroad Emission Study of Catalyzed 
Particulate Filter Equipped Small Diesel Engines' Southwest Research 
Institute, September 2001. Copy available in EPA Air Docket A-2001-
28. Note that the ``AWQ'' cycle specified in Table 8 is the same as 
the proposed constant speed, variable load cycle.
---------------------------------------------------------------------------

NMHC+NOX Standard
    We have proposed a 3.5 g/bhp-hr NMHC+NOX standard for 
engines in the 25-50 hp range for 2013. This will align the 
NMHC+NOX standard for engines in this power range with the 
Tier 3 standard for engines in the 50-75 hp range which are implemented 
in 2008. EPA's recent Staff Technical paper which reviewed the 
technological feasibility of the Tier 3 standards contains a detailed 
discussion of a number of technologies which are capable of achieving a 
3.5 g/bhp-hr standard. These include cooled EGR, uncooled EGR, as well 
as advanced in-cylinder technologies relying on electronic fuel systems 
and turbocharging.\202\ These technologies are capable of reducing 
NOX emission by as much as 50 percent. Given the Tier 2 
NMHC+NOX standard of 5.6 g/bhp-hr, a 50 percent reduction 
would allow a Tier 2 engine to comply with the 3.5 g/bhp-hr 
NMHC+NOX standard proposed in this action. In addition, 
because this NMHC+NOX standard is concurrent with the 0.02 
g/bhp-hr PM standards which we project will be achievable with the use 
of particulate filters, engine designers will have significant 
additional flexibility in reducing NOX because the PM filter 
will eliminate the traditional concerns with the engine-out 
NOX vs. PM trade-off. Our recent highway 2004 standard 
review rulemaking (see 65 FR 59896) demonstrated that a diesel engine 
with advanced electronic fuel injection technology as well as 
NOX control technology such as cooled EGR is capable of 
complying with an NTE standard set at 1.25 times the laboratory based-
standard FTP standard. We project that the same technology (electronic 
fuel systems and cooled EGR) are also capable for engine in the 25-75 
hp range of complying with the proposed NTE standard of 4.4 g/bhp-hr 
NMHC+NOX (1.25 x 3.5) in 2013. This is based on the broad 
NOX reduction capability of cooled EGR technology, which is 
capable of reducing NOX emissions across the engine 
operating map by at least 30 percent even under high load 
conditions.\203\
---------------------------------------------------------------------------

    \202\ See section 2.2 through 2.3 in ``Nonroad Diesel Emission 
Standards--Staff Technical Paper'', EPA Publication EPA420-R-01-052, 
October 2001. Copy available in EPA Air Docket A-2001-28.
    \203\ See section 8 of ``Control of Emissions of Air Pollution 
from 2004 and Later Model Year Heavy-Duty Highway Engines and 
Vehicles: Response to Comments'', EPA document EPA420-R-00-011, July 
2000, and Chapter 3 of ``Regulatory Impact Analysis: Control of 
Emissions of Air Pollution from Highway Heavy-duty Engines'', EPA 
document EPA420-R-00-010, July 2000. Copies of both documents 
available in EPA docket A-2001-28.
---------------------------------------------------------------------------

    Based on the information available to EPA and presented here, and 
giving appropriate consideration to the lead time necessary to apply 
the technology as well, we have concluded the proposed 0.02 g/bhp-hr PM 
standard for engines in the 25-75 hp category and the 3.5 g/bhp-hr 
NMHC+NOX standards for the 25-50 hp engines are achievable.
d. Why EPA has not Proposed More Stringent Tier 4 NOX 
Standards
    Today's notice proposes to revise the NMHC+NOX standard 
for engines between 25 and 50 hp to a level of 3.5 g/bhp-hr beginning 
in 2013 (the same numeric level as the Tier 3 standards for engines in 
the 50-75 hp range). As discussed below, we believe this standard can 
be met using a variety of technologies, including but not limited to 
cooled EGR. Similar technologies will be used on engines in the 50-100 
hp

[[Page 28391]]

range beginning in 2008. At this time, we are not proposing further 
reductions in the NOX standards for engines between 25 and 
75 hp.
    As discussed in section III.B.1.d, engines =75 hp are 
similar to, or are direct derivatives of, highway HDDEs. As discussed 
in section III.E.1-III.E.3, NOX adsorber technology is being 
developed today in order to comply with the 2007 highway heavy-duty 
standards. However, NOX adsorber technologies will require 
additional development beyond what has occurred at this time in order 
to achieve the 2007 highway standards. Section III.E.1-III.E.3 also 
discuss the high degree of complexity and engine/aftertreatment 
integration which will be required in order for NOX 
adsorbers to be applied successfully to nonroad diesel engines.
    As discussed above, and as illustrated in Table III.E-3, engines 
<75 hp include a significant fraction of naturally aspirated engines 
and engines with indirect-injection fuel systems, and we are not 
predicting a significant shift away from IDI technology engines. Given 
the relatively unsophisticated level of technology used in this power 
category today, as well as our prediction that even in the 2011-13 time 
frame these engines will lag significantly behind the =75 hp 
engines, we believe it is appropriate not to propose NOX 
adsorber based standards at this time. Rather, as discussed in section 
III.H, we have proposed to undertake a technology assessment in the 
2007 time frame which would evaluate the status of emission control 
technologies for engines less than 75 hp, and such a review would 
revisit this issue. In addition, section VI of this proposal contains 
additional discussion regarding our analysis of applying NOX 
adsorbers to engines in the 25-75 hp category. EPA invites further 
comment on the above discussion, and also solicits comment on the cost 
impacts of NOX aftertreatment devices, including unit costs, 
on these engines.
5. Are the Standards Proposed for Engines <25 hp Feasible?
    As discussed in section III.B, our proposal for standards for 
engines less than 25 hp is a new PM standard of 0.30 g/bhp-hr beginning 
in 2008. As discussed below, we are not proposing to set a new standard 
more stringent than the existing Tier 2 NMHC+NOX standard 
for this power category at this time. This section describes:
    [sbull] What makes the <25 hp category unique;
    [sbull] Engine technology currently used in the <25 hp category;
    [sbull] Why the proposed standards are technologically feasible; 
and,
    [sbull] Why EPA has not proposed more stringent standards at this 
time.
a. What Makes the <25 hp Category Unique?
    Nonroad engines less than 25 hp are the least sophisticated nonroad 
diesel engines from a technological perspective. All of the engines 
currently sold in this power category lack electronic fuel systems and 
turbochargers (see Table III.E-3). Nearly 50 percent of the products 
have two-cylinders or less, and 14 percent of the engines sold in this 
category are single-cylinder products, a number of these have no 
batteries and are crank-start machines, much like today's simple walk 
behind lawnmower engines. In addition, given what we know today and 
taking into account the Tier 2 standards which have not yet been 
implemented, we are not projecting any significant penetration of 
advanced engine technology, such as electronically controlled fuel 
systems, into this category in the next 5 to 10 years.
    We have proposed a PM standard for engines in the <25 hp category 
which is higher than the standard proposed for engines in the 25-75 hp 
category (0.30 g/bhp-hr vs. 0.22 g/bhp-hr). We have done this for a 
number of reasons. First, the existing Tier 2 PM standards specifies 
standards which become numerically higher for the smaller power 
categories. Specifically, for engines 175 hp, the Tier 2 PM 
standard is 0.15 g/bhp-hr, which increases to 0.30 g/bhp-hr for engines 
in the 50-100hp range, 0.45 g/bhp-hr for engines in the 25-50hp range, 
and finally 0.60 g/bhp-hr for engines <25 hp. In the Tier 2 time frame, 
engines in the higher power categories are expected to use more 
sophisticated technologies such as turbocharging and high pressure 
electronically controlled fuel systems. These technologies are more 
capable of reducing PM emissions as compared to naturally aspirated 
engines with lower pressure mechanical fuel systems. To some extent 
this same trend is expected to continue in the 2008 time frame. As 
discussed above, we expect that many engines in the 25-75hp engine 
category will use turbocharging, and some engines will have electronic 
fuel systems. However, we are not predicting that any engines in the 
<25hp category will use either of these technologies. In addition, very 
small diesel engines present a number of unique challenges for reducing 
PM emissions. First, the smaller engines inherently have high 
combustion chamber surface-to-volume ratios. This results in higher 
heat loss, which results in a quenching of the oxidation process 
earlier than for larger engines, and therefore higher PM emission 
rates. In addition, the small diesel engines are more limited in the PM 
reduction which can be achieved by higher fuel injection pressures. Due 
to the very small size of the combustion chamber, high pressure 
injection (which is intended to improve fuel atomization and mixing, 
both of which lower PM emissions) will result in fuel impaction on the 
combustion chamber, which will not improve fuel atomization. The 
benefits of higher pressure fuel injection as a PM reduction technology 
therefore reaches a point of diminishing returns with higher and higher 
pressures, and this point of diminishing returns is reached much 
quicker for the smaller engines than for the larger engines. For these 
reasons we have proposed a 2008 PM standard for engines <25 hp which is 
higher than the proposed 2008 PM standard for engines in the 25-75 hp 
category.
b. What Engine Technology is Currently Used in the <25 hp category?
    In the 1998 nonroad diesel rulemaking we established Tier 1 and 
Tier 2 standards for these products. Tier 1 was implemented in model 
year 2000, and Tier 2 will be implemented in model year 2005. As 
discussed in EPA's recent Staff Technical Paper, we project the Tier 2 
standards will be met by basic engine-out emission optimization 
strategies.\204\ We are not predicting that Tier 2 will require 
electronic fuel systems, EGR, or turbocharging. As discussed in the 
Staff Technical Paper, a large number of engines in this power category 
already meet the Tier 2 standards by a wide margin.\205\
---------------------------------------------------------------------------

    \204\ See section 3 of ``Nonroad Diesel Emission Standards--
Staff Technical Paper'', EPA Publication EPA420-R-01-052, October 
2001. Copy available in EPA Air Docket A-2001-28.
    \205\ See Table 3-2 in ``Nonroad Diesel Emission Standards--
Staff Technical Paper'', EPA Publication EPA420-R-01-052, October 
2001. Copy available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    Two basic types of engine fuel injection technologies are currently 
present in the less than 25 hp category, mechanical indirect injection 
(IDI) and mechanical direct injection (DI). As discussed in section 
III.D.4, the IDI system injects fuel into a pre-chamber rather than 
directly into the combustion chamber as in the DI system. This 
difference in fuel systems results in substantially different emission 
characteristics, as well as several important operating parameters. In 
general, as noted earlier, the IDI engine has lower engine-out PM and 
NOX

[[Page 28392]]

emissions, while the DI engine has better fuel efficiency and lower 
heat rejection.
c. What Data Indicates That the Proposed Standards Are Feasible?
    We project the proposed Tier 4 PM standard can be met by 2008 based 
on:
    [sbull] The existence of a large number of engine families which 
meet the proposed standards today;
    [sbull] The use of engine-out reduction techniques; and
    [sbull] The use of diesel oxidation catalysts.
    We have examined the recent model year (2002) engine certification 
data for nonroad diesel engines less than 25 hp. These data indicate 
that a number of engine families meet the proposed Tier 4 PM standard 
(and the 2008 NMHC+NOX standard, unchanged from Tier 2) 
today. The current data indicates approximately 28% of the engine 
families are at or below the proposed PM standard today, while meeting 
the 2008 NMHC+NOX standard. These include both IDI and DI 
engines, as well as a range of certification test cycles.\206\ Many of 
the engine families are certified well below the proposed Tier 4 
standard while meeting the 2008 NMHC+NOX level. 
Specifically, 15 percent of the engine families exceed the proposed 
Tier 4 PM standard by more than 20 percent. The public certification 
data indicate that these engines do not use turbocharging, electronic 
fuel systems, exhaust gas recirculation, or aftertreatment 
technologies.
---------------------------------------------------------------------------

    \206\ The Tier 1 and Tier 2 standards for this power category 
must be demonstrated on one of a variety of different engine test 
cycles. The appropriate test cycle is selected by the engine 
manufacturer based on the intended in-use applications(s) of the 
engine.
---------------------------------------------------------------------------

    These model year 2002 engines use well known engine-out emission 
control technologies, such as combustion chamber design and fuel 
injection timing control strategies, to comply with the existing 
standards. As with 25-75 hp engines, these data have a two-fold 
significance. First, they indicate that a number of engines in this 
power category can already achieve the proposed 2008 standard for PM 
using only engine-out technology, and that other engines should be able 
to achieve the standard making improvements just to engine-out 
performance. Second, despite being certified to the same emission 
standards with similar engine technology, the emission levels from 
these engines vary widely. Figure III.E-2 is a graph of the model year 
2002 HC+NOX and PM data. As can be seen in the figure, the 
emission levels cover a wide range. Figure III.E-2 highlights a 
specific example of this wide range: engines using naturally aspirated 
IDI technology and tested on the 6-mode test cycle. Even for this 
subset of IDI engines achieving approximately the same 
HC+NOX level of[sim]4.5 g/bhp-hr, the PM rates vary from 
approximately 0.15 to 0.5 g/bhp-hr. (A more detailed discussion of this 
data is contained in the draft RIA.) There is limited information 
available to indicate why for these small diesel engines with similar 
technology operating at approximately the same HC+NOX level 
the PM emission rates cover such a broad range. We are therefore not 
predicating the proposed 2008 PM standard on the combination of diesel 
oxidation catalysts and the lowest engine-out emissions being achieved 
today, because it is uncertain whether or not additional engine-out 
improvements would lower all engines to the proposed 2008 PM standard. 
Instead, we believe there are two likely means by which companies can 
comply with the proposed 2008 PM standard. First, some engine 
manufacturers can comply with this standard using known engine-out 
techniques (e.g., optimizing combustion chamber designs, fuel-injection 
strategies). However, based on the available data it is unclear whether 
engine-out techniques will work in all cases. Therefore, we believe 
some engine companies will choose to use a combination of engine-out 
techniques and diesel oxidation catalysts, as discussed below.

[[Page 28393]]

[GRAPHIC] [TIFF OMITTED] TP23MY03.008

    PM emissions can be reduced through in-cylinder techniques for 
small nonroad diesel engines using similar techniques as used in larger 
nonroad and highway engines. As discussed in section III.E.1.a, there 
are a number of technologies which exist that can influence oxygen 
content and in-cylinder mixing (and thus lower PM emissions) including 
improved fuel injection systems and combustion system designs. For 
example, increased injection pressure can reduce PM emissions 
substantially.\207\ The wide-range of emission characteristics present 
in the existing engine certification data is likely a result of 
differences in fuel systems and combustion chamber designs. For many of 
the engines which have higher emission levels, further optimization of 
the fuel system and combustion chamber can provide additional PM 
reductions.
---------------------------------------------------------------------------

    \207\ ``Effects of Injection Pressure and Nozzle Geometry on DI 
Diesel Emissions and Performance,'' Pierpont, D., and Reitz, R., SAE 
Paper 950604, 1995.
---------------------------------------------------------------------------

    Diesel oxidation catalysts (DOC) also offer the opportunity to 
reduce PM emissions from the engines in this power category. DOCs are 
passive flow through emission control devices which are typically 
coated with a precious metal or a base-metal wash-coat. DOCs have been 
proven to be durable in-use on both light-duty and heavy-duty diesel 
applications. In addition, DOCs have already been used to control 
carbon monoxide on some nonroad applications.\208\ However, as 
discussed in section III.E.1.a., certain DOC formulations can be 
sensitive to diesel fuel sulfur level. Specifically, precious-metal 
based oxidation catalysts (which have the greatest potential for 
reducing PM) can oxidize the sulfur in the fuel and form particulate 
sulfates. Given the high level of sulfur in nonroad fuel today, the use 
of DOCs as a PM reduction technology is severely limited. Data 
presented by one engine manufacturer regarding the existing Tier 2 PM 
standard shows that while a DOC can be used to meet the current 
standard when tested on 2,000 ppm sulfur fuel, lowering the fuel sulfur 
level to 380 ppm enabled the DOC to reduce PM by 50 percent from the 
2,000 ppm sulfur fuel.\209\ Without the availability of 500 ppm sulfur 
fuel in 2008, DOCs would be of limited use for nonroad engine 
manufacturers and would not provide the emissions necessary to meet the 
proposed standards for most engine manufacturers. With the availability 
of 500 ppm sulfur fuel, DOC's can be designed to provide PM reductions 
on the order of 20 to 50%, while suppressing particulate sulfate 
reduction. These levels of reductions have been seen on transient duty 
cycles as well as highway and nonroad steady-state duty cycles.\210\ As 
discussed in section III.D, we are proposing to apply supplemental test 
procedures and standards (nonroad transient test cycle

[[Page 28394]]

and not-to-exceed requirements) to engines in the <25 hp category 
beginning in 2013. The supplemental test procedures and standards will 
apply not only to PM, but also to NMHC+NOX. While we believe 
the engine technology necessary to comply with the supplemental test 
procedures and standards is the same as the technology necessary to 
comply with the 2008 standard, we are delaying the implementation of 
the supplemental test procedures and standards until 2013 in order to 
implement the supplemental requirements on the larger powered nonroad 
engines before the smallest power category (see section III.C. above). 
This will also provide engine manufacturers with additional time to 
install any emission testing equipment upgrades they may need in order 
to implement the new nonroad transient test cycle. Nevertheless, the 
technologies described above are capable of complying with both the 
proposed nonroad transient test cycle and the NTE standard. As just 
described, DOCs are capable of reducing PM emissions up to 50 percent 
during transient testing. With respect to feasibility under NTE 
testing, it has been demonstrated, as a result of a recent Agency 
action, that engines which rely on retarded injection timing as a 
primary NOX control technology, which is also the primary 
technology that engines in the <25 hp category will likely use to 
comply with the Tier 2 NMHC+NOX standard, are capable of 
complying with an NMHC+NOX NTE standard of 1.25 x the FTP 
for engines with emission levels on the order of 4 g/bhp-hr 
NOX. Specifically, as a result of federal consent decrees 
with a number of highway heavy-duty diesel engine manufactures, many 
highway engines certified to an FTP standard of 4 g/bhp-hr 
NOX were also designed to comply with an NTE limit of 5 g/
bhp-hr (i.e., 1.25 x FTP standard).\211\ The Tier 2 NMHC+NOX 
standard for engines <25hp is 5.6 g/bhp-hr, therefore, in 2013 the 
proposed NTE standard is 7.0 g/bhp-hr NMHC+NOX. Based on the 
experience which a number of highway diesel engine companies, we 
project that the proposed NTE standard for engines <25 hp can be 
achieved by 2013.
---------------------------------------------------------------------------

    \208\ EPA Memorandum ``Documentation of the Availability of 
Diesel Oxidation Catalysts on Current Production Nonroad Diesel 
Equipment'', William Charmley. Copy available in EPA Air Docket A-
2001-28.
    \209\ See Table 2-4 in ``Nonroad Diesel Emission Standards--
Staff Technical Paper'', EPA Publication EPA420-R-01-052, October 
2001. Copy available in EPA Air Docket A-2001-28.
    \210\ ``Demonstration of Advanced Emission Control Technologies 
Enabling Diesel-Powered Heavy-duty Engines to Achieve Low Emission 
Levels: Interim Report Number 1--Oxidation Catalyst Technology, copy 
available in EPA Air Docket A-2001-28. ``Reduction of Diesel Exhaust 
Emissions by Using Oxidation Catalysts,'' Zelenka et. al., SAE Paper 
90211, 1990. See Table 2-4 in ``Nonroad Diesel Emission Standards--
Staff Technical Paper'', EPA Publication EPA420-R-01-052, October 
2001, copy available in EPA Air Docket A-2001-28.
    \211\ EPA Memorandum ``Summary of Model Year 1999 and 2000 
Federal On-highway Heavy-duty Diesel Engine Families Certified as 
Compliant with Not-to-Exceed Requirements, Euro-3 Steady State 
Requirements, and Maximum Allowable Emission Limits Requirements'', 
copy available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    As discussed in section III.B, we have also proposed a minor change 
in the CO standard for the <11 hp engines, in order to align those 
standards with the standards for the 11-25 hp engines. As discussed in 
section III.B., the small change in the CO standard is intended to 
simplify EPA's regulations as part of our decision to propose a 
reduction in the number of engine power categories for Tier 4. The 
current CO standard for this category is 6.0 g/bhp-hr, and the proposed 
standard is 4.9 g/bhp-hr (i.e., the current standard for engines in the 
11-25 hp range). The model year 2002 certification data shows that more 
than 90 percent of the engine families in this power category meet the 
proposed standards today. In addition, DOCs typically reduce CO 
emissions on the order of 50 percent or more during both transient and 
steady-state operation.\212\ Given that more than 90 percent of the 
engines in this category meet the proposed standard today, and the 
ready availability of technology which can easily achieve the proposed 
standard, we project this CO standard will be achievable by model year 
2008.
---------------------------------------------------------------------------

    \212\ ``Demonstration of Advanced Emission Control Technologies 
Enabling Diesel-Powered Heavy-duty Engines to Achieve Low Emission 
Levels: Interim Report Number 1--Oxidation Catalyst Technology, and 
``Reduction of Diesel Exhaust Emissions by Using Oxidation 
Catalysts'', P. Zelenka et. al., Society of Automotive Engineers 
paper 902111, October 1990.
---------------------------------------------------------------------------

    Based on the existence of a number of engine families which already 
comply with the proposed Tier 4 PM standard (and the 2008 
NMHC+NOX standard), and the availability of PM reduction 
technologies such as improved fuel systems, combustion chamber 
improvements, and in particular diesel oxidation catalysts, we project 
the proposed 0.30 g/bhp-hr PM standards is technologically feasible by 
model year 2008. All of these are conventional technologies which have 
been used on both highway and nonroad diesel engines in the past. As 
such, we do not expect there to be any negative impacts with respect to 
noise or safety. In addition, PM reduction technologies such as 
improved combustion through the use of higher pressure fuel injection 
systems as well as DOCs are not predicted to have any substantial 
impact on fuel efficiency.
d. Why has EPA not Proposed More Stringent PM or NOX 
Standards for Engines <25 hp?
    Section III.E.4 contains a detailed discussion of why we don't 
believe it is appropriate at this time to revise the NOX 
standards based on NOX absorber technology for engines 
between 25 and 75 hp. These same arguments apply for engines below 25 
hp. In addition, we have not proposed to revise the NOX 
standard for <25 hp engines in this action, nor do we believe PM 
standards based on particulate filters are appropriate for this power 
category based on a number of factors, as discussed below.
    In EPA's recent Staff Technical Paper regarding the feasibility of 
the Tier 3 NMHC+NOX standards for engines greater than 50 
hp, we projected that a number of engine technologies can be used to 
meet the Tier 3 standards, including cooled EGR or hot EGR, both with 
advanced electronic fuel systems, as well as with internal combustion 
techniques using advanced electronic fuel systems, advanced 
turbocharging systems (e.g., waste-gated or variable geometry 
turbochargers), and possibly variable valve actuation.\213\ In 
addition, we presumed the use of charge-air cooling In order to set 
more stringent NOX standards for <25 hp engines without 
increasing PM emissions, the most logical list of technologies is 
turbocharging, electronically controlled hot or cooled EGR, an 
electronic fuel system, and possibly charge-air-cooling. No nonroad 
diesel engine <25 hp uses any combination of these technologies today. 
While we are able to postulate that some of this technology could be 
applied to the <25 hp engines, the application of some of the 
technology (such as turbocharging) is technologically uncertain. It is 
the combination of these two issues (the traditional NOX-PM 
trade-off and the difficulties with turbocharging 1 and 2 cylinder 
engines) which is the primary reason we are not proposing to revise the 
NOX standard for engines in this size range. NOX 
reduction control technologies such as advancing fuel injection timing 
or using EGR will increase PM emissions. In order to reduce 
NOX emissions and reduce or maintain current PM levels 
additional technologies must be used. Fundamental among these is the 
need to increase oxygen content, which can be achieved principally with 
turbocharging. However, turbocharging systems do not lend themselves to 
1 and 2 cylinder products, which are approximately 50 percent of the 
engines in this power category. In addition, even if these technologies 
could be applied to engines in the < 25 hp category, the costs would be 
substantial relative to both the base engine cost and to the cost of 
the nonroad equipment itself . Therefore, for the reasons discussed 
above, we have not proposed to revise the NOX standard for 
these engines at

[[Page 28395]]

this time. As discussed in section III.H, we have proposed that a 
technology assessment occur in 2007 which would evaluate the status of 
emission control technologies for engines less than 75 hp, and such a 
review would revisit this issue.
---------------------------------------------------------------------------

    \213\ See section 2.3.1 through 2.3.3 of ``Nonroad Diesel 
Emission Standards--Staff Technical Paper'', EPA Publication EPA420-
R-01-052, October 2001. Copy available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    In addition, we have not proposed to apply particulate filter based 
standards for engines less than 25 hp. As discussed in sections III.E.1 
through 4, there are two basic types of particulate filter systems we 
believe could be used by engine manufacturers. The first is a CDPF 
which uses post-injection from a common-rail electronic fuel injection 
system in order to ensure filter regeneration. The second type of 
system would use a CDPF with a stand-alone (i.e., independent from the 
engine's fuel system) fuel injection system to ensure filter 
regeneration. In either case, an electronic control system is required, 
as well as the CDPF. Such systems are not being developed for engines 
of this size for either highway light-duty or heavy-duty diesel 
applications, and (as noted earlier) it is unclear whether the 
technology development which is being done for the highway market will 
transfer down to engines in this power category. In addition, based on 
currently available information, we believe the cost of these 
technologies are relatively high compared to the overall cost of the 
equipment. As discussed in section III.H, we have proposed that a 
technology assessment occur in 2007 which would evaluate the status of 
emission control technologies for engines less than 75 hp, and such a 
review would revisit this issue.
6. Meeting the Crankcase Emissions Requirements
    The most common way to eliminate crankcase emissions has been to 
vent the blow-by gases into the engine air intake system, so that the 
gases can be recombusted. Prior to the HD2007 rulemaking, we have 
required that crankcase emissions be controlled only on naturally 
aspirated diesel engines. We had made an exception for turbocharged 
diesel engines (both highway and nonroad) because of concerns in the 
past about fouling that could occur by routing the diesel particulates 
(including engine oil) into the turbocharger and aftercooler. However, 
this is an environmentally significant exception since most nonroad 
equipment over 70hp use turbocharged engines, and a single engine can 
emit over 100 pounds of NOX, NMHC, and PM from the crankcase 
over its lifetime.
    Given the available means to control crankcase emissions, we 
eliminated this exception for highway engines in 2007 and are proposing 
to eliminate the exception for nonroad diesel engines as well. We 
anticipate that the diesel engine manufacturers will be able to control 
crankcase emissions through the use of closed crankcase filtration 
systems or by routing unfiltered blow-by gases directly into the 
exhaust system upstream of the emission control equipment. However, the 
proposed provision has been written such that if adequate control can 
be had without ``closing'' the crankcase then the crankcase can remain 
``open.'' Compliance would be ensured by adding the emissions from the 
crankcase ventilation system to the emissions from the engine control 
system downstream of any emission control equipment. We propose to 
limit this provision for controlling emissions from open crankcases to 
turbocharged engines, which is the same as for heavy-duty highway 
diesel engines. We request comment on extending this provision to 
naturally aspirated engines, as we did for marine diesel engines in our 
1999 final rule (64 FR 73300, December 29, 1999).
    We expect that in order to meet the stringent tailpipe emission 
standards set here, that manufacturers will have to utilize closed 
crankcase approaches as described here. Closed crankcase filtration 
systems work by separating oil and particulate matter from the blow-by 
gases through single or dual stage filtration approaches, routing the 
blow-by gases into the engine's intake manifold and returning the 
filtered oil to the oil sump. Oil separation efficiencies in excess of 
90 percent have been demonstrated with production ready prototypes of 
two stage filtration systems.\214\ By eliminating 90 percent of the oil 
that would normally be vented to the atmosphere, the system works to 
reduce oil consumption and to eliminate concerns over fouling of the 
intake system when the gases are routed through the turbocharger. Hatz, 
a nonroad engine manufacturer, currently has closed crankcase systems 
on many of its turbocharged engines.
---------------------------------------------------------------------------

    \214\ Letter from Marty Barris, Donaldson Corporation, to Byron 
Bunker U.S. EPA, March 2000. Copy available in EPA Air Docket A-
2001-28.
---------------------------------------------------------------------------

F. Why Do We Need 15ppm Sulfur Diesel Fuel?

    As stated earlier, we strongly believe that fuel sulfur control is 
critical to ensuring the success of NOX and PM 
aftertreatment technologies. In order to evaluate the effect of sulfur 
on diesel exhaust control technologies, we used three key factors to 
categorize the impact of sulfur in fuel on emission control function. 
These factors were efficiency, reliability, and fuel economy. Taken 
together these three factors lead us to believe that diesel fuel sulfur 
levels of 15 ppm will be required for the nonroad emission standards 
proposed here to be feasible. Brief summaries of these factors are 
provided below.
    The efficiency of emission control technologies to reduce harmful 
pollutants is directly affected by sulfur in diesel fuel. Initial and 
long term conversion efficiencies for NOX, NMHC, CO and 
diesel PM emissions are significantly reduced by catalyst poisoning and 
catalyst inhibition due to sulfur. NOX conversion 
efficiencies with the NOX adsorber technology in particular 
are dramatically reduced in a very short time due to sulfur poisoning 
of the NOX storage bed. In addition, total PM control 
efficiency is negatively impacted by the formation of sulfate PM. As 
explained in the following sections, the CDPF, NOX adsorber, 
and urea SCR catalyst technologies described here have the potential to 
make significant amounts of sulfate PM under operating conditions 
typical of many nonroad engines. We believe that the formation of 
sulfate PM will be in excess of the total PM standard, unless diesel 
fuel sulfur levels are at or below 15 ppm. Based on the strong negative 
impact of sulfur on emission control efficiencies for all of the 
technologies evaluated, we believe that 15 ppm represents an upper 
threshold of acceptable diesel fuel sulfur levels.
    Reliability refers to the expectation that emission control 
technologies must continue to function as required under all operating 
conditions for the life of the engine. As discussed in the following 
sections, sulfur in diesel fuel can prevent proper operation of both 
NOX and PM control technologies. This can lead to permanent 
loss in emission control effectiveness and even catastrophic failure of 
the systems. Sulfur in diesel fuel impacts reliability by decreasing 
catalyst efficiency (poisoning of the catalyst), increasing diesel 
particulate filter loading, and negatively impacting system 
regeneration functions. Among the most serious reliability concerns 
with sulfur levels greater than 15 ppm are those associated with 
failure to properly regenerate. In the case of the NOX 
adsorber, failure to regenerate the stored sulfur (desulfate) will lead 
to rapid loss of NOX emission control as a result of sulfur 
poisoning of the NOX adsorber bed. In the case of the diesel 
particulate filter, sulfur in the fuel reduces the reliability of the 
regeneration function.

[[Page 28396]]

If regeneration does not occur, catastrophic failure of the filter 
could occur. It is only by the availability of low sulfur diesel fuels 
that these technologies become feasible.
    Fuel economy impacts due to sulfur in diesel fuel affect both 
NOX and PM control technologies. The NOX adsorber 
sulfur regeneration cycle (desulfation cycle) can consume significant 
amounts of fuel unless fuel sulfur levels are very low. The larger the 
amount of sulfur in diesel fuel, the greater the adverse effect on fuel 
economy. As sulfur levels increase above 15 ppm, the adverse effect on 
fuel economy becomes more significant, increasing above one percent and 
doubling with each doubling of fuel sulfur level. Likewise, PM trap 
regeneration is inhibited by sulfur in diesel fuel. This leads to 
increased PM loading in the diesel particulate filter and increased 
work to pump exhaust across this restriction. With low sulfur diesel 
fuel, diesel particulate filter regeneration can be optimized to give a 
lower (on average) exhaust backpressure and thus better fuel economy. 
Thus, for both NOX and PM technologies the lower the fuel 
sulfur level the lower the operating costs of the vehicle.
1. Catalyzed Diesel Particulate Filters and the Need for Low Sulfur 
Fuel
    CDPFs function to control diesel PM through mechanical filtration 
of the solid PM (soot) from the diesel exhaust stream and then 
oxidation of the stored soot (trap regeneration) and oxidation of the 
SOF. Through oxidation in the catalyzed diesel particulate filter the 
stored PM is converted to CO2 and released into the 
atmosphere. Failure to oxidize the stored PM leads to accumulation in 
the trap, eventually causing the trap to become so full that it 
severely restricts exhaust flow through the device, leading to trap or 
vehicle failure.
    Uncatalyzed diesel particulate filters require exhaust temperatures 
in excess of 650[deg]C in order for the collected PM to be oxidized by 
the oxygen available in diesel exhaust. That temperature threshold for 
oxidation of PM by exhaust oxygen can be decreased to 450[deg]C through 
the use of base metal catalytic technologies. For a broad range of 
operating conditions typical of in-use diesel engine operation, diesel 
exhaust can be significantly cooler than 400[deg]C. If oxidation of the 
trapped PM could be assured to occur at exhaust temperatures lower than 
300[deg]C, then diesel particulate filters would be expected to be more 
robust for most applications and operating regimes. Oxidation of PM 
(regeneration of the trap) at such low exhaust temperatures can occur 
by using oxidants which are more readily reduced than oxygen. One such 
oxidant is NO2.
    NO2 can be produced in diesel exhaust through the 
oxidation of the nitrogen monoxide (NO), created in the engine 
combustion process, across a catalyst. The resulting NO2-
rich exhaust is highly oxidizing in nature and can oxidize trapped 
diesel PM at temperatures as cool as 250[deg]C.\215\ Some platinum 
group metals are known to be good catalysts to promote the oxidation of 
NO to NO2. Therefore in order to promote more effective 
passive regeneration of the diesel particulate filters, significant 
amounts of platinum group metals (primarily platinum) are being used in 
the wash-coat formulations of advanced CDPFs. The use of platinum to 
promote the oxidation of NO to NO2 introduces several system 
vulnerabilities affecting both the durability and the effectiveness of 
the CDPF when sulfur is present in diesel exhaust. (In essence, diesel 
engine exhaust temperatures are in a range necessitating use of 
precious metal catalysts in order to adequately regenerate the PM 
filter, but precious metal catalysts are in turn highly sensitive to 
sulfur in diesel fuel.) The two primary mechanisms by which sulfur in 
diesel fuel limits the robustness and effectiveness of CDPFs are 
inhibition of trap regeneration, through inhibition of the oxidation of 
NO to NO2, and a dramatic loss in total PM control 
effectiveness due to the formation of sulfate PM. Unfortunately, these 
two mechanisms trade-off against one another in the design of CDPFs. 
Changes to improve the reliability of regeneration by increasing 
catalyst loadings lead to increased sulfate emissions and, thus, loss 
of PM control effectiveness. Conversely, changes to improve PM control 
by reducing the use of platinum group metals and, therefore, limiting 
``sulfate make'' leads to less reliable regeneration. Even with an 
active regeneration system, reducing catalytic loading to reduce 
sulfate make unacceptably trades off regeneration effectiveness (i.e., 
robustness). We believe the best means of achieving good PM emission 
control and reliable operation is to reduce sulfur in diesel fuel, as 
shown in the following subsections.
---------------------------------------------------------------------------

    \215\ Hawker, P. et al, ``Experience with a New Particulate Trap 
Technology in Europe,'' SAE 970182.
---------------------------------------------------------------------------

a. Inhibition of Trap Regeneration Due to Sulfur
    The CDPF technology relies on the generation of a very strong 
oxidant, NO2, to ensure that the carbon captured by the PM 
trap's filtering media is oxidized under the exhaust temperature range 
of normal operating conditions. This prevents plugging and failure of 
the PM trap. NO2 i2 produced through the oxidation of NO in 
the exhaust across a platinum catalyst. This oxidation is inhibited by 
sulfur poisoning of the catalyst surface.\216\ This inhibition limits 
the total amount of NO2 available for oxidation of the 
trapped diesel PM, thereby raising the minimum exhaust temperature 
required to ensure trap regeneration. Without sufficient 
NO2, the amount of PM trapped in the diesel particulate 
filter will continue to increase and can lead to excessive exhaust back 
pressure and low engine power.
---------------------------------------------------------------------------

    \216\ Hawker, P. et al, ``Experience with a New Particulate Trap 
Technology in Europe,'' SAE 970182.
---------------------------------------------------------------------------

    The failure mechanisms experienced by diesel particulate filters 
due to low NO2 availability vary significantly in severity 
and long term consequences. In the most fundamental sense, the failure 
is defined as an inability to oxidize the stored particulate at a rate 
fast enough to prevent net particulate accumulation over time. The 
excessive accumulation of PM over time blocks the passages through the 
filtering media, making it more restrictive to exhaust flow. In order 
to continue to force the exhaust through the now more restrictive 
filter, the exhaust pressure upstream of the filter must increase. This 
increase in exhaust pressure is commonly referred to as increasing 
``exhaust backpressure'' on the engine.
    The increase in exhaust backpressure represents increased work 
being done by the engine to force the exhaust gas through the 
increasingly restrictive particulate filter. Unless the filter is 
frequently cleansed of the trapped PM, this increased work can lead to 
reductions in engine performance and increases in fuel consumption. 
This loss in performance may be noted by the equipment operator in 
terms of sluggish engine response.
    Full field test evaluations and retrofit applications of these 
catalytic trap technologies are occurring in parts of the United States 
and Europe where low sulfur diesel fuel is already available.\217\ The 
experience gained in these field

[[Page 28397]]

tests helps to clarify the need for low sulfur diesel fuel. In Sweden 
and some European city centers where below 10 ppm diesel fuel sulfur is 
readily available, more than 3,000 catalyzed diesel particulate filters 
have been introduced into retrofit applications without a single 
failure. Given the large number of vehicles participating in these test 
programs, the diversity of the vehicle applications which included 
intercity trains, airport buses, mail trucks, city buses and garbage 
trucks, and the extended time periods of operation (some vehicles have 
been operating with traps for more than 5 years and in excess of 
300,000 miles\218\, there is a strong indication of the robustness of 
this technology on 10 ppm low sulfur diesel fuel. The field experience 
in areas where sulfur is capped at 50 ppm has been less definitive. In 
regions without extended periods of cold ambient conditions, such as 
the United Kingdom, field tests on 50 ppm cap low sulfur fuel have also 
been positive, matching the durability at 10 ppm, although sulfate PM 
emissions are much higher. However, field tests on 50 ppm fuel in 
Finland, where colder winter conditions are sometimes encountered 
(similar to many parts of the United States), showed a significant 
number of failures (10 percent) due to trap plugging. This 10 percent 
failure rate has been attributed to insufficient trap regeneration due 
to fuel sulfur in combination with low ambient temperatures.\219\ Other 
possible reasons for the high failure rate in Finland when contrasted 
with the Swedish experience appear to be unlikely. The Finnish and 
Swedish fleets were substantially similar, with both fleets consisting 
of transit buses powered by Volvo and Scania engines in the 10 to 11 
liter range. Further, the buses were operated in city areas and none of 
the vehicles were operated in northern extremes such as north of the 
Arctic Circle.\220\ Given that the fleets in Sweden and Finland were 
substantially similar, and given that ambient conditions in Sweden are 
expected to be similar to those in Finland, we believe that the 
increased failure rates noted here are due to the higher fuel sulfur 
level in a 50 ppm cap fuel versus a 10 ppm cap fuel.\221\
---------------------------------------------------------------------------

    \217\ Through tax incentives 50 ppm cap sulfur fuel is widely 
available in the United Kingdom and 10 ppm sulfur fuel is available 
in Sweden and in certain European city centers.
    \218\ Allansson, et al., ``European Experience of High Mileage 
Durability of Continuously Regenerating Filter Technology,'' SAE 
2000-01-0480.
    \219\ Letter from Dr. Barry Cooper, Johnson Matthey, to Don 
Kopinski, U.S. EPA. Copy available in EPA Air Docket A-2001-28.
    \220\ Telephone conversation between Dr. Barry Cooper, Johnson 
Matthey, and Todd Sherwood, EPA, Air Docket A-99-06.
    \221\ The average temperature in Helsinki, Finland, for the 
month of January is 21[deg]F. The average temperature in Stockholm, 
Sweden, for the month of January is 26[deg]F. The average 
temperature at the University of Michigan in Ann Arbor, Michigan, 
for the month of January is 24[deg]F. The temperatures reported here 
are from www.worldclimate.com based upon the Global Historical 
Climatology Network (GHCN) produced jointly by the National Climatic 
Data Center and Carbon Dioxide Information Analysis Center at Oak 
Ridge National Laboratory (ORNL).
---------------------------------------------------------------------------

    Testing on an even higher fuel sulfur level of 200 ppm was 
conducted in Denmark on a fleet of 9 vehicles. In less than six months 
all of the vehicles in the Danish fleet had failed due to trap 
plugging.\222\ The failure of some fraction of the traps to regenerate 
when operated on fuel with sulfur caps of 50 ppm and 200 ppm is 
believed to be primarily due to inhibition of the NO to NO2 
conversion as described here. Similarly the increasing frequency of 
failure with higher fuel sulfur levels is believed to be due to the 
further suppression of NO2 formation when higher sulfur 
level diesel fuel is used. Since this loss in regeneration 
effectiveness is due to sulfur poisoning of the catalyst this real 
world experience would be expected to apply equally well to nonroad 
engines (i.e., operation on lower sulfur diesel fuel, 15 ppm versus 50 
ppm, will increase regeneration robustness).
---------------------------------------------------------------------------

    \222\ Letter from Dr. Barry Cooper to Don Kopinski U.S. EPA. 
Copy available in EPA Air Docket A-2001-28.
---------------------------------------------------------------------------

    As shown above, sulfur in diesel fuel inhibits NO oxidation leading 
to increased exhaust backpressure and reduced fuel economy. Therefore, 
we believe that, in order to ensure reliable and economical operation 
over a wide range of expected operating conditions, nonroad diesel fuel 
sulfur levels should be at or below 15 ppm.
b. Loss of PM Control Effectiveness
    In addition to inhibiting the oxidation of NO to NO2, 
the sulfur dioxide (SO2) in the exhaust stream is itself 
oxidized to sulfur trioxide (SO3) at very high conversion 
efficiencies by the precious metals in the catalyzed particulate 
filters. The SO3 serves as a precursor to the formation of 
hydrated sulfuric acid (H2SO4+H2O), or 
sulfate PM, as the exhaust leaves the vehicle tailpipe. Virtually all 
of the SO3 is converted to sulfate under dilute exhaust 
conditions in the atmosphere as well in the dilution tunnel used in 
heavy-duty engine testing. Since virtually all sulfur present in diesel 
fuel is converted to SO2, the precursor to SO3, 
as part of the combustion process, the total sulfate PM is directly 
proportional to the amount of sulfur present in diesel fuel. Therefore, 
even though diesel particulate filters are very effective at trapping 
the carbon and the SOF portions of the total PM, the overall PM 
reduction efficiency of catalyzed diesel particulate filters drops off 
rapidly with increasing sulfur levels due to the formation of sulfate 
PM downstream of the CDPF.
    SO2 oxidation is promoted across a catalyst in a manner 
very similar to the oxidation of NO, except it is converted at higher 
rates, with peak conversion rates in excess of 50 percent. The 
SO2 oxidation rate for a platinum based oxidation catalyst 
typical of the type which might be used in conjunction with, or as a 
washcoat on, a CDPF can vary significantly with exhaust temperature. At 
the low temperatures the oxidation rate is relatively low, perhaps no 
higher than ten percent. However at the higher temperatures that might 
be more typical of agricultural tractor use pulling a plow and the 
highway Supplemental Emission Test (also called the EURO III or 13 mode 
test), the oxidation rate may increase to 50 percent or more. These 
high levels of sulfate make across the catalyst are in contrast to the 
very low SO2 oxidation rate typical of diesel exhaust 
(typically less than 2 percent). This variation in expected diesel 
exhaust temperatures means that there will be a corresponding range of 
sulfate production expected across a CDPF.
    The U.S. Department of Energy in cooperation with industry 
conducted a study entitled DECSE to provide insight into the 
relationship between advanced emission control technologies and diesel 
fuel sulfur levels. Interim report number four of this program gives 
the total particulate matter emissions from a heavy-duty diesel engine 
operated with a diesel particulate filter on several different fuel 
sulfur levels. A straight line fit through this data is presented in 
Table III.F-1 below showing the expected total direct PM emissions from 
a diesel engine on the supplemental emission test cycle.\223\ The SET 
test cycle, a 13 mode steady-state cycle, that this data was developed 
on is similar to the C1 eight mode steady-state nonroad test cycle. 
Both cycles include operation at full and intermediate load points at 
approximately rated speed conditions and torque peak speed conditions. 
As a

[[Page 28398]]

result, the sulfate make rate for the C1 cycle and the SET cycle would 
be expected to be similar. The data can be used to estimate the PM 
emissions from diesel engines operated on fuels with average fuel 
sulfur levels in this range.
---------------------------------------------------------------------------

    \223\ Note that direct emissions are those pollutants emitted 
directly from the engine or from the tailpipe depending on the 
context in which the term is used, and indirect emissions are those 
pollutants formed in the atmosphere through chemical reactions 
between direct emissions and other atmospheric constituents.

         Table III. F-1--Estimated PM Emissions From a Diesel Engine at the Indicated Fuel Sulfur Levels
----------------------------------------------------------------------------------------------------------------
                                                                          Steady state emissions performance
                                                                    --------------------------------------------
                         Fuel sulfur [ppm]                            Tailpipe PMb    PM increase relative to 3
                                                                       [g/bhp-hr]             ppm sulfur
----------------------------------------------------------------------------------------------------------------
3..................................................................           0.003  ...........................
7a.................................................................           0.006                         100%
15a................................................................           0.009                         200%
30.................................................................           0.017                         470%
150................................................................           0.071                       2300%
----------------------------------------------------------------------------------------------------------------
Notes:
a The PM emissions at these sulfur levels are based on a straight-line fit to the DECSE data; PM emissions at
  other sulfur levels are actual DECSE data. (Diesel Emission Control Sulfur Effects (DECSE) Program--Phase II
  Interim Data Report No. 4, Diesel Particulate Filters-Final Report, January 2000. Table C1.) Although DECSE
  tested diesel particulate filters at these fuel sulfur levels, they do not conclude that the technology is
  feasible at all levels, but they do note that testing at 150 ppm is a moot point as the emission levels exceed
  the engine's baseline emission level.
b Total exhaust PM (soot, SOF, sulfate).

    Table III.F-1 makes it clear that there are significant PM emission 
reductions possible with the application of catalyzed diesel 
particulate filters and low sulfur diesel fuel. At the observed sulfate 
PM conversion rates, the DECSE program results show that the 0.01 g/
bhp-hr total PM standard is feasible for CDPF equipped engines operated 
on fuel with a sulfur level at or below 15 ppm. The results also show 
that diesel particulate filter control effectiveness is rapidly 
degraded at higher diesel fuel sulfur levels due to the high sulfate PM 
make observed with this technology. It is clear that PM reduction 
efficiencies are limited by sulfur in diesel fuel and that, in order to 
realize the PM emissions benefits sought in this rule, diesel fuel 
sulfur levels must be at or below 15 ppm.
c. Increased Maintenance Cost for Diesel Particulate Filters Due to 
Sulfur
    In addition to the direct performance and durability concerns 
caused by sulfur in diesel fuel, it is also known that sulfur can lead 
to increased maintenance costs, shortened maintenance intervals, and 
poorer fuel economy for CDPFs. CDPFs are highly effective at capturing 
the inorganic ash produced from metallic additives in engine oil. This 
ash is accumulated in the filter and is not removed through oxidation, 
unlike the trapped soot PM. Periodically the ash must be removed by 
mechanical cleaning of the filter with compressed air or water. This 
maintenance step is anticipated to occur on intervals of well over 
1,500 hours (depending on engine size). However, sulfur in diesel fuel 
increases this ash accumulation rate through the formation of metallic 
sulfates in the filter, which increases both the size and mass of the 
trapped ash. By increasing the ash accumulation rate, the sulfur 
shortens the time interval between the required maintenance of the 
filter and negatively impacts fuel economy.
2. Diesel NOX Catalysts and the Need for Low Sulfur Fuel
    NOX adsorbers are damaged by sulfur in diesel fuel 
because the adsorption function itself is poisoned by the presence of 
sulfur. The resulting need to remove the stored sulfur (desulfate) 
leads to a need for extended high temperature operation which can 
deteriorate the NOX adsorber. These limitations due to 
sulfur in the fuel affect the overall performance and feasibility of 
the NOX adsorber technology.
a. Sulfur Poisoning (Sulfate Storage) on NOX Adsorbers
    The NOX adsorber technology relies on the ability of the 
catalyst to store NOX as a metallic nitrate 
(MNO3) on the surface of the catalyst, or adsorber (storage) 
bed, during lean operation. Because of the similarities in chemical 
properties of SOx and NOX, the SO3 present in the 
exhaust is also stored by the catalyst surface as a sulfate 
(MSO4). The sulfate compound that is formed is significantly 
more stable than the nitrate compound and is not released and reduced 
during the NOX release and reduction step (NOX 
regeneration step). Since the NOX adsorber is essentially 
100 percent effective at capturing SO2 in the adsorber bed, 
the sulfur build up on the adsorber bed occurs rapidly. As a result, 
sulfate compounds quickly occupy all of the NOX storage 
sites on the catalyst thereby rendering the catalyst ineffective for 
NOX storage and subsequent NOX reduction 
(poisoning the catalyst).
    The stored sulfur compounds can be removed by exposing the catalyst 
to hot (over 650 [deg]C) and rich (air-fuel ratio below the 
stoichiometric ratio of 14.5 to 1) conditions for a brief period.\224\ 
Under these conditions, the stored sulfate is released and reduced in 
the catalyst.\225\ While research to date on this procedure has been 
very favorable with regards to sulfur removal from the catalyst, it has 
revealed a related vulnerability of the NOX adsorber 
catalyst. Under the high temperatures used for desulfation, the metals 
that make up the storage bed can change in physical structure. This 
leads to lower precious metal dispersion, or ``metal sintering,'' (a 
less even distribution of the catalyst sites) reducing the 
effectiveness of the catalyst.\226\ This degradation of catalyst 
efficiency due to high temperatures is often referred to as thermal 
degradation. Thermal degradation is known to be a cumulative effect. 
That is, with each excursion to high temperature operation, some 
additional degradation of the catalyst occurs.
---------------------------------------------------------------------------

    \224\ Dou, Danan and Bailey, Owen, ``Investigation of 
NOX Adsorber Catalyst Deactivation,'' SAE 982594.
    \225\ Guyon, M. et al, ``Impact of Sulfur on NOX Trap 
Catalyst Activity--Study of the Regeneration Conditions'', SAE 
982607.
    \226\ Though it was favorable to decompose sulfate at 800 
[deg]C, performance of the NSR (NOX Storage Reduction 
catalyst, i.e. NOX Adsorber) catalyst decreased due to 
sintering of precious metal.--Asanuma, T. et al, ``Influence of 
Sulfur Concentration in Gasoline on NOX Storage--
Reduction Catalyst'', SAE 1999-01-3501.
---------------------------------------------------------------------------

    One of the best ways to limit thermal degradation is by limiting 
the accumulated number of desulfation events over the life of the 
vehicle. Since

[[Page 28399]]

the period of time between desulfation events is expected to be 
determined by the amount of sulfur accumulated on the catalyst (the 
higher the sulfur accumulation rate, the shorter the period between 
desulfation events) the desulfation frequency is expected to be 
proportional to the fuel sulfur level. In other words for each doubling 
in the average fuel sulfur level, the frequency and accumulated number 
of desulfation events are expected to double. We concluded in the 
HD2007 rulemaking, that this thermal degradation would be unacceptable 
high for fuel sulfur levels greater than 15 ppm. Some commenters to the 
HD2007 rule suggested that the NOX adsorber technology could 
meet the HD2007 NOX standard using diesel fuel with a 30 ppm 
average sulfur level. This would imply that the NOX adsorber 
could tolerate as much as a four fold increase in desulfation frequency 
(when compared to an expected seven to 10 ppm average) without any 
increase in thermal degradation. That conclusion was inconsistent with 
our understanding of the technology at the time of the HD2007 
rulemaking and remains inconsistent with our understanding of progress 
made by industry since that time. Diesel fuel sulfur levels must be at 
or below 15 ppm in order to limit the number and frequency of 
desulfation events. Limiting the number and frequency of desulfation 
events will limit thermal degradation and, thus, enable the 
NOX adsorber technology to meet the NOX standard.
    This conclusion remains true for the highway NOX 
adsorber catalyst technology that this proposal is based upon and will 
be equally true for nonroad engines applying the NOX 
adsorber technology to comply with our proposed Tier 4 standards.
    Nonroad and highway diesel engines are similarly durable and thus 
over their lifetimes consume a similar amount of diesel fuel. This 
means that both nonroad and highway diesel engines will have the same 
exposure to sulfur in diesel fuel and thus will require the same number 
of desulfation cycles over their lifetimes. This is true independent of 
the test cycle or in-use operation of the nonroad engine.
    Sulfur in diesel fuel for NOX adsorber equipped engines 
will also have an adverse effect on fuel economy. The desulfation event 
requires controlled operation under hot and net fuel rich exhaust 
conditions. These conditions, which are not part of a normal diesel 
engine operating cycle, can be created through the addition of excess 
fuel to the exhaust. This addition of excess fuel causes an increase in 
fuel consumption.
    Future improvements in the NOX adsorber technology, as 
we have observed in our ongoing diesel progress reviews, are expected 
and needed in order to meet the NOX emission standards 
proposed today. Some of these improvements are likely to include 
improvements in the means and ease of removing stored sulfur from the 
catalyst bed. However because the stored sulfate species are inherently 
more stable than the stored nitrate compounds (from stored 
NOX emissions) and so will always be stored preferentially 
to NOX on the adsorber storage sites, we expect that a 
separate release and reduction cycle (desulfation cycle) will always be 
needed in order to remove the stored sulfur. Therefore, we believe that 
fuel with a sulfur level at or below 15 ppm sulfur will be necessary in 
order to control thermal degradation of the NOX adsorber 
catalyst and to limit the fuel economy impact of sulfur in diesel fuel.
b. Sulfate Particulate Production and Sulfur Impacts on Effectiveness 
of NOX Control Technologies
    The NOX adsorber technology relies on a platinum based 
oxidation function in order to ensure high NOX control 
efficiencies. As discussed more fully in section III.F.1, platinum 
based oxidation catalysts form sulfate PM from sulfur in the exhaust 
gases significantly increasing PM emissions when sulfur is present in 
the exhaust stream. The NOX adsorber technology relies on 
the oxidation function to convert NO to NO2 over the 
catalyst bed. For the NOX adsorber this is a fundamental 
step prior to the storage of NO2 in the catalyst bed as a 
nitrate. Without this oxidation function the catalyst will only trap 
that small portion of NOX emissions from a diesel engine 
which is NO2. This would reduce the NOX adsorber 
effectiveness for NOX reduction from in excess of 90 percent 
to something well below 20 percent. The NOX adsorber relies 
on platinum to provide this oxidation function due to the need for high 
NO oxidation rates under the relatively cool exhaust temperatures 
typical of diesel engines. Because of this fundamental need for a 
precious metal catalytic oxidation function, the NOX 
adsorber inherently forms sulfate PM when sulfur is present in diesel 
fuel, since sulfur in fuel invariably leads to sulfur in the exhaust 
stream.
    The Compact-SCR technology, like the NOX adsorber 
technology, uses an oxidation catalyst to promote the oxidation of NO 
to NO2 at the low temperatures typical of much of diesel 
engine operation. By converting a portion of the NOX 
emissions to NO2 upstream of the ammonia SCR reduction 
catalyst, the overall NOX reductions are improved 
significantly at low temperatures. Without this oxidation function, low 
temperature SCR NOX effectiveness is dramatically reduced 
making compliance with the NOX standard impossible. 
Therefore, future Compact-SCR systems would need to rely on a platinum 
oxidation catalyst in order to provide the required NOX 
emission control. This use of an oxidation catalyst in order to enable 
good NOX control means that Compact SCR systems will produce 
significant amounts of sulfate PM when operated on anything but the 
lowest fuel sulfur levels due to the oxidation of SO2 to 
sulfate PM promoted by the oxidation catalyst.
    Without the oxidation catalyst promoted conversion of NO to 
NO2, neither of these NOX control technologies 
can meet the proposed NOX standard. Therefore, each of these 
technologies will require low sulfur diesel fuel to control the sulfate 
PM emissions inherent in the use of highly active oxidation catalysts. 
The NOX adsorber technology may be able to limit its impact 
on sulfate PM emissions by releasing stored sulfur as SO2 
under rich operating conditions. The Compact-SCR technology, on the 
other hand, has no means to limit sulfate emissions other than through 
lower catalytic function or lowering sulfur in diesel fuel. The degree 
to which the NOX emission control technologies increase the 
production of sulfate PM through oxidation of SO2 to 
SO3 varies somewhat from technology to technology, but it is 
expected to be similar in magnitude and environmental impact to that 
for the PM control technologies discussed previously, since both the 
NOX and the PM control catalysts rely on precious metals to 
achieve the required NO to NO2 oxidation reaction.
    At fuel sulfur levels below 15 ppm this sulfate PM concern is 
greatly diminished. Without this low sulfur fuel, the NOX 
control technologies are expected to create PM emissions well in excess 
of the PM standard regardless of the engine-out PM levels. Thus, we 
believe that diesel fuel sulfur levels will need to be at or below 15 
ppm in order to apply the NOX control technology.

G. Reassessment of Control Technology for Engines Less Than 75 hp in 
2007

    By structuring our program to benefit extensively from prior 
experience with core technologies in the highway sector, we believe 
that a nonroad diesel technology review of the extent being pursued for 
the heavy-duty highway

[[Page 28400]]

engine program will not be needed.\227\ Indeed the results of that 
ongoing review have already had a very helpful impact in shaping this 
proposal. Nevertheless, there are some technology issues that will not 
be addressed in the highway program review. In particular we believe 
that a future review of particulate filter technology for engines under 
75 hp may be warranted. Under our proposed schedule presented in 
section III.B, standards based on the performance of this technology 
will take effect in the 2013 model year for 25-75 hp engines (or in the 
2012 model year for manufacturers opting to skip the transitional 
standards for 50-75 hp engines).
---------------------------------------------------------------------------

    \227\ See ``Highway Diesel Progress Review'', U.S. EPA, June 
2002. EPA420-R-02-016. (www.epa.gov/air/caaac/dieselreview.pdf).
---------------------------------------------------------------------------

    At this time we have not decided what the long-term PM standards 
should be for engines under 25 hp. No PM filter-based standards are 
being proposed for engines under 25 hp as part of this Tier 4 proposal. 
Likewise, we have not decided what the long-term NOX 
standards should be for engines under 75 hp, and no NOX 
adsorber-based standards are being proposed for engines under 75 hp. As 
part of the technology review, we plan to thoroughly evaluate progress 
made toward applying advanced PM and NOX control 
technologies to these smaller engines.
    We propose to conduct the technology review in 2007, and to 
conclude it by the end of that year, to give manufacturers lead time 
should an adjustment in the program be considered appropriate. We do 
not intend to include in the technology review a reassessment of PM 
filter technology needed to meet the optional 0.02 g/hp-hr PM standard 
for 50-75 hp engines in 2012. We assume that manufacturers would only 
choose this option if they had confidence that they could meet the 0.02 
g/hp-hr standard in 2012, a year earlier than otherwise required.
    We recognize the importance of harmonization of international 
standards and have worked diligently with our colleagues in Europe and 
Japan to achieve that objective. Harmonization of these standards will 
allow manufacturers continued access to world markets and lower the 
required research and development and tooling costs needed to meet 
different standards. We will continue to work with both governments and 
the manufacturers abroad and within the United States. We have 
incorporated feedback from the on-going dialogue and have continued to 
work through the international process as we have developed this 
proposal. The Commission has proposed amendments in December 2002 to EC 
Directive 97/68 which are currently being addressed in the European 
Council and Parliament. We believe that today's proposal and the 
European approach together provide the framework for additional 
harmonization. While not identical, manufacturers have expressed 
appreciation for the similarities which do exist and they represent a 
significant step toward mitigating the differences in design challenges 
that would otherwise exist. The limit values and test procedures 
provide a basis for common development which manufacturers can use on a 
global basis. The amendments would control fuel sulfur levels to enable 
aftertreatment, set nonroad mobile machine emissions limits that would 
be based on performance of diesel particulate traps. NOX 
limits are being set to match the Agency's Tier 3 NOX 
program. There are a few differences in approaches that we will 
continue to discuss with the EU. One difference is that the EC has 
chosen a leadtime for trap-based PM standards for engines in the 50-100 
hp range which is one year earlier than we are proposing today. Another 
difference is the inclusion of a review of the availability of 
NOX emission control technology for larger engines. The EC 
has also chosen not to set performance requirements that would require 
the use of PM traps for engines under 50 hp, while we are proposing 
performance-based standards that would likely require the use of PM 
traps for engines between 25-75 hp. The EC has again chosen not to set 
standards for engines below 19 kW (25 hp) and greater than 560 kW (750 
hp). With respect to long term NOX control, the Commission 
has chosen to have a technology review (which would also reassess 
issues related to PM) to address implementing potentially more 
stringent NOX standards in the same timeframe as potential 
EPA standards.\228\ For additional information about the harmonization 
effort and the results to date, please see chapter 2.4.2 of the SBREFA 
panel report. We request comment on opportunities to further enhance 
harmonization.
---------------------------------------------------------------------------

    \228\ Commission of the European Communities, ``Proposal for a 
Directive of the European Parliament and of the Council amending 
Directive 97/68/EC'', section 3.9.
---------------------------------------------------------------------------

    We expect that any changes to the level or timing of emission 
standards found appropriate in the 2007 review would be made as part of 
a rulemaking process, and that process would take additional time after 
the review is completed. If the 2007 review should determine that PM 
trap technology is feasible for engine under 25 hp, or that advanced 
NOX control technology is feasible for engines under 75 hp, 
or that Tier 4 standards should be made more stringent in some other 
way, we would expect the rulemaking implementing such changes to 
provide for adequate lead time. Therefore, it would be premature for us 
to target 2013 or any specific model year for implementing such 
standards changes at this time. We solicit comment on the scope, 
timing, and need for a future reassessment of emissions control 
technology for nonroad diesel engines.

IV. Our Proposed Program for Controlling Nonroad, Locomotive and Marine 
Diesel Fuel Sulfur

    We are proposing to reduce the sulfur content of nonroad, 
locomotive and marine (NRLM) diesel fuel to no more than 500 ppm 
beginning in 2007. We are also proposing to reduce the sulfur content 
of nonroad diesel fuel to no more than 15 ppm beginning in 2010. These 
provisions mirror controls on highway diesel fuel to 500 ppm in 1993 
\229\ and 15 ppm in 2006.\230\
---------------------------------------------------------------------------

    \229\ 55 FR 34120 (August 21, 1990).
    \230\ 66 FR 5002 (January 18, 2001).
---------------------------------------------------------------------------

    There are two reasons that we are proposing these standards. First, 
fuel sulfur significantly inhibits or impairs the function of the 
diesel exhaust emission control devices, which would generally be 
necessary to meet the proposed nonroad diesel engine emission 
standards. In conjunction with the proposed 15 ppm sulfur standard for 
nonroad diesel fuel we have concluded that this emission control 
technology will be available to achieve the reductions required by the 
stringent NOX and PM emission standards proposed for model 
year 2011 and later nonroad diesel engines. Second, sulfur in diesel 
fuel is emitted from the engine as sulfate PM and sulfur dioxide, both 
of which cause adverse health and welfare impacts, as described in 
section II. above. Reducing the level of sulfur in diesel fuel to 500 
ppm beginning in 2007 would achieve important emission reductions of 
these pollutants and provide significant public health and welfare 
benefits. The further reduction to 15 ppm in 2010 will expand upon 
these benefits.
    In developing the proposed diesel fuel program, we identified 
several principles that we wanted the program to achieve:

[[Page 28401]]

    (1) Maintain the benefits and program integrity of the highway 
diesel fuel program;
    (2) Achieve the greatest reduction in sulfate PM and sulfur dioxide 
emissions from nonroad, locomotive, and marine diesel engines as early 
as practicable;
    (3) Provide for a smooth transition of the nonroad diesel fuel pool 
to 15 ppm sulfur;
    (4) Ensure that 15 ppm sulfur diesel fuel is produced and 
distributed widely for use in all 2011 and later model year nonroad 
engines;
    (5) Enable the efficient distribution of all diesel fuels; and
    (6) Ensure that the program's requirements are enforceable and 
verifiable.
    As described below, we believe the proposed fuel program achieves 
these principles.
    The remainder of this section is organized as follows:
    (A) The fuel standards proposed today,
    (B) The design and structure of the fuel program,
    (C) Special hardship provisions proposed for small refiners and 
refiners facing particularly difficult circumstances,
    (D) Special provisions proposed for fuel sold in the State of 
Alaska and U.S. Territories,
    (E) The affect of the proposed program on state diesel fuel control 
programs,
    (F) The technological feasibility of the production and 
distribution of 500 ppm and 15 ppm sulfur nonroad, locomotive and 
marine diesel fuel,
    (G) The impact of the program on other fuel properties and 
specialty fuels, and
    (H) The need for some refiners to obtain air permits for their 
desulfurization equipment.
    Analyses supporting the design of these provisions can be found in 
chapter V and VII of the Draft RIA for today's action. Section VIII of 
this preamble provides a discussion of the compliance and enforcement 
provisions affecting diesel fuel and additional explanation of various 
elements of the proposed program.

A. Proposed Nonroad, Locomotive and Marine Diesel Fuel Quality 
Standards

    The following paragraphs describe the requirements, standards, and 
deadlines that apply to refiners, importers, and distributors of 
nonroad, locomotive and marine (NRLM) diesel fuel and the options 
available to all refiners.
1. What Fuel Is Covered by This Proposal?
    The proposed standards generally cover all the diesel fuel that is 
used in mobile applications but is not already covered by the previous 
standards for highway diesel fuel. This fuel is defined primarily by 
the type of engine which it is used to power: nonroad, locomotive, and 
marine diesel engines. These fuels typically include:
    (1) Any number 1 and 2 distillate fuels used, intended for use, or 
made available for use in nonroad, locomotive or marine diesel engines,
    (2) Any number 1 distillate fuel (e.g., kerosene) added to such 
number 2 diesel fuel, e.g., to improve its cold flow properties, and
    (3) Any other fuel used in or blended with diesel fuel for use in 
nonroad, locomotive, or marine diesel engines that has comparable 
chemical and physical characteristics.
    Primary examples of fuels under (1) would be those meeting ASTM 
D975 or D396 specifications for grades number 1-D and number 2-D or 
ASTM DMX and DMA specifications, if used in the engines mentioned 
above. Primary examples under (3) would be certain specialty fuels 
grades such as JP-5, JP-8, and F76 if used in nonroad, locomotive, or 
marine equipment for which a national security exemption has not been 
approved (See section VIII.A.2) and non-distillate fuels such as 
biodiesel.
    This proposal would not apply to:
    (1) Number 1 distillate fuel used to power jet aircraft,
    (2) Number 1 or number 2 distillate fuel used for other purposes, 
such as to power stationary diesel engines or for heating,
    (3) Number 4 and 6 fuels (e.g., bunker or residual fuels, IFO Heavy 
Fuel Oil Grades 30 and higher, ASTM DMB and DMC fuels), and
    (4) Any fuel used to power equipment for which a national security 
exemption has been approved (see section VIII.A.2).
    The proposed program would reduce the sulfur in all diesel fuel 
likely used in mobile off-highway equipment and achieve very 
significant short and long-term environmental benefits. States, not the 
Agency, have responsibility for any fuel sulfur specifications for 
heating oil, so this fuel would not be covered by this proposal.\231\ 
However, we do propose a number of provisions, as described below, that 
would ensure that heating oil would not be used in nonroad, locomotive, 
or marine applications.
---------------------------------------------------------------------------

    \231\ For the purposes of this proposal, the term heating oil 
refers to any number 1 or number 2 distillate other than jet fuel 
and diesel fuel used in highway, nonroad, locomotive, or marine 
applications. For example, heating oil includes fuel which is 
suitable for use in furnaces, boilers, stationary diesel engines and 
similar applications and is commonly or commercially known or sold 
as heating oil, fuel oil, and other similar trade names.
---------------------------------------------------------------------------

    As in the recent highway diesel rule, in those cases where the same 
batch of kerosene is distributed for two purposes (e.g., as kerosene to 
be used for heating and to improve the cold flow of number 2 nonroad 
diesel fuel), that batch of kerosene would have to meet the standards 
being proposed today for nonroad diesel fuel. However, an alternative 
compliance approach would be to produce and distribute two distinct 
kerosene fuels. In our example above, one batch would meet the proposed 
sulfur standards and could be blended into number 2 NRLM diesel fuel. 
The other batch would only have to meet any applicable specifications 
for heating oil.
2. Standards and Deadlines for Refiners, Importers, and Fuel 
Distributors
    The proposed fuel program consists of a two-step program to reduce 
the sulfur content of nonroad diesel fuel. By doing so, the program 
would allow the refining industry to smoothly transition the sulfur 
content from its current uncontrolled levels down to the very stringent 
15 ppm level. By beginning with an initial step down to 500 ppm, we can 
start to achieve significant emission reductions and associated health 
and welfare benefits from the current fleet of equipment as soon as 
possible. While we considered and are seeking comment on a one-step 
approach of going directly to 15 ppm in 2008, as discussed in section 
VI, we believe that on balance the advantages of the proposed two-step 
approach outweigh those of a single step.
    The specific proposed deadlines for meeting the 500 and 15 ppm 
sulfur standards would not apply to refineries covered by special 
hardship provisions for small refiners. In addition, a different 
schedule would apply for any refineries approved under the proposed 
general hardship provisions. All of these hardship provisions are 
described below in section IV.C.
a. The First Step to 500 ppm
    Under this proposal NRLM diesel fuel produced by refiners or 
imported into the U.S. would be required to meet a 500 ppm sulfur 
standard beginning June 1, 2007. Refiners and importers could comply by 
either producing such fuel at or below 500 ppm, or could comply by 
obtaining credits as discussed in Section B below.
    We believe that the proposed level of 500 ppm is appropriate for 
several reasons. This 500 ppm level is consistent with current highway 
diesel fuel, a grade which may remain for

[[Page 28402]]

highway purposes until 2010. As such, adopting the same 500 ppm level 
for NRLM helps to avoid any issues and costs associated with more 
grades of fuel in the distribution system during this initial step of 
the program. The reduction to 500 ppm is also significant 
environmentally. The 500 ppm level achieves approximately 90 percent of 
the sulfate PM and SO2 benefits otherwise achievable by 
going all the way to 15 ppm. Yet, the costs would be roughly half that 
associated with full control down to 15 ppm. Because this first step is 
only to 500 ppm, it also allows for a short lead time for 
implementation, enabling the environmental benefits to begin accruing 
as soon as possible. After careful analysis of feasibility as discussed 
in section IV.F.5, we believe that the proposed start date of June 1, 
2007, is the earliest that the 500 ppm step could take effect.
    To allow for the enforcement of the proposed fuel standards while 
at the same time allowing for a smooth and orderly transition of diesel 
fuel in the distribution system to 500 ppm, we are proposing that 
parties downstream of the refineries be allowed time to turnover their 
NRLM tanks to 500 ppm. We are proposing that at the terminal level, 
NRLM diesel fuel would be required to meet the 500 ppm sulfur standard 
beginning August 1, 2007. At bulk plants, wholesale purchaser-
consumers, and any retail stations carrying NRLM diesel, this fuel 
would have to meet the 500 ppm sulfur standard by October 1, 2007.\232\ 
The only exceptions to these dates would be for high sulfur NRLM 
produced under the hardship and fuel credit provisions discussed below 
in sections IV.B. and C.\233\
---------------------------------------------------------------------------

    \232\ A bulk plant is a secondary distributor of refined 
petroleum products. They typically receive fuel from terminals and 
distribute fuel in bulk by truck to end users. Consequently, while 
for highway fuel, bulk plants often serve the role of a fuel 
distributor, delivering fuel to retail stations, for nonroad fuel, 
they often serve the role of the retailer, delivering fuel directly 
to the end-user.
    \233\ Furthermore, as discussed in subsection B, we propose that 
high sulfur nonroad diesel fuel which is produced after June 1, 2007 
due to the small refiner and fuel credit provisions could be 
commingled with 500 ppm nonroad diesel fuel after it has been dyed 
to the IRS specifications. Thus, at some points in the distribution 
system, nonroad fuel higher than the 500 ppm standard would remain 
until it is precluded from production beginning June 1, 2010.
---------------------------------------------------------------------------

    This downstream turnover schedule is slightly more relaxed than for 
the second step to 15 ppm discussed below. This first step down to 500 
ppm is designed to achieve the public health and welfare benefits from 
reduced emissions in the current fleet of engines. Since the sulfate PM 
and SO3 benefits accrue as the fuel is desulfurized to any 
degree, mixing in the distribution system during the transition to 500 
ppm would not reduce this benefit or cause any adverse consequences. 
Mixing in the distribution system would also not reduce the engine 
performance and durability benefits from the reduction in sulfur. As a 
result, the immediate turnover of the fuel pool downstream of the 
refinery gate is of less concern and a more relaxed schedule than 
described below for the second step is possible. We seek comment on 
this proposed schedule.
b. The Second Step to 15 ppm
    In order to enable the application of high efficiency exhaust 
emission control technologies to nonroad diesel engines beginning with 
the 2011 model year, we are proposing that all nonroad diesel fuel 
produced or imported after June 1, 2010, would have to meet a 15 ppm 
sulfur cap. We are proposing that diesel fuel used for locomotive and 
marine diesel engines could continue to the meet the 500 ppm cap first 
applicable in 2007.
    In order to allow for a smooth and orderly transition of diesel 
fuel in the distribution system to 15 ppm, we are proposing that 
parties downstream of the refineries be allowed some additional time to 
turnover their tanks to 15 ppm. We are proposing that at the terminal 
level, nonroad diesel fuel would be required to meet the 15 ppm sulfur 
standard beginning July 15, 2010. At bulk plants, wholesale purchaser-
consumers, and any retail stations carrying nonroad diesel, this fuel 
would have to meet the 15 ppm sulfur standard by September 1, 2010. The 
proposed transition schedule for compliance with the 15 ppm standard at 
refineries, terminals, and secondary distributors is the same as that 
allowed under the recently promulgated highway diesel fuel program.
    As with the 500 ppm standard, refiners and importers could comply 
with this standard by either physically producing 15 ppm fuel or by 
obtaining sulfur credits, as described below.
    We are seriously considering bringing the sulfur level of 
locomotive and marine diesel fuel to 15 ppm as early as June 1, 2010, 
along with nonroad diesel fuel. As discussed in more detail in section 
VI and in chapter 12 of the draft RIA, there are several advantages 
associated with this alternative. First, it would provide important 
sulfate PM and SO3 emission reductions and the estimated 
benefits from these reductions would outweigh the costs by a 
considerable margin. Second, it would simplify the fuel distribution 
system and the design of the fuel program proposed today. Third, it 
would help reduce the potential opportunity for misfueling of 2007 and 
later model year highway vehicles and 2011 and later model year nonroad 
equipment with higher sulfur fuel. Finally, it would allow refiners to 
coordinate plans to reduce the sulfur content of all of their nonroad 
diesel fuel at one time.
    However, discussions with refiners have suggested there are 
advantages to leaving locomotive and marine diesel fuel at 500 ppm, at 
least in the near-term and until we set more stringent standards for 
those engines. The locomotive and marine diesel fuel markets could 
provide a market for off-spec product which is important for refiners, 
particularly during the transition to 15 ppm for highway and nonroad 
diesel fuel in 2010. Waiting just a year or two beyond 2010 would 
address the critical near term needs during the transition. Second, 
waiting just another year or two beyond 2010 is also projected to allow 
virtually all refiners to take advantage of the new lower cost 
technology.
    In addition to seeking comment on whether to apply the 15 ppm 
standard to locomotive and marine diesel fuel in 2010, we also seek 
comment on other timing for doing so, and especially on how the Agency 
should coordinate a 15 ppm standard for locomotive and marine with the 
nonroad diesel fuel standard being proposed today. It is the Agency's 
intention to propose in the near future new emission standards for 
locomotive and marine engines that could require the use of high 
efficiency exhaust emission control technology, and thus, also require 
the use of 15 ppm sulfur diesel fuel. We anticipate that such engine 
standards would likely take effect in the 2011-13 time frame, requiring 
15 ppm locomotive and marine diesel fuel in the 2010-12 time frame. We 
intend to publish an advanced notice of proposed rulemaking (ANPRM) for 
such a rule in the Spring of 2004 and complete action on a final rule 
by 2007.
c. Other Standard Provisions
    We are proposing that the 500 ppm NRLM and 15 ppm nonroad diesel 
fuel standards would apply to the areas of Alaska served by the Federal 
Aid Highway System (FAHS). Rural areas, those outside the FAHS, would 
not be subject to either the 15 or 500 ppm standards. Market forces in 
these areas would be relied upon to provide 15 ppm diesel fuel for 2011 
and later nonroad diesel engines used in these areas. This is 
consistent with the approach which is

[[Page 28403]]

in the process of being developed by the State of Alaska for 
implementing the 2007 highway diesel fuel program. EPA can revisit this 
issue when it takes action on Alaska's plan for implementation of the 
highway sulfur requirements, allowing for coordination of the nonroad 
and highway fuel requirements. The specifics of our proposal for diesel 
fuel sold in Alaska are described in more detail in section IV.D.1. 
below. In addition, these proposed 500 and 15 ppm sulfur caps would not 
apply to diesel fuel sold in three Pacific U.S. territories, as 
described in more detail in section IV.D.2. below.
    The early credits and other special provisions create the 
probability that high sulfur NRLM diesel fuel would be produced and 
sold after June 1, 2007, and that 500 ppm nonroad diesel fuel would be 
produced and sold after June 1, 2010. Under the proposal, fuel 
distributors would be responsible for ensuring the necessary product 
segregations and that statements on product transfer documents and fuel 
product labels are consistent with the corresponding fuel quality. The 
specific requirements for both fuel distributors and end-users are 
described in detail in section VIII.
d. Cetane Index or Aromatics Standard
    Currently, in addition to containing no more than 500 ppm sulfur, 
EPA requires that highway diesel fuel meet a minimum cetane index level 
of 40 or, as an alternative contain no more than 35 volume percent 
aromatics. We are proposing today to extend this cetane index/aromatics 
content specification to NRLM diesel fuel. Extension of these content 
specifications would reduce NOX and PM emissions from the 
current nonroad equipment fleet slightly, providing associated public 
health and welfare benefits.
    Low diesel fuel cetane levels are associated with increases in 
NOX and PM emissions in current nonroad diesel engines. 
Thus, we expect that this cetane index specification would lead to a 
reduction in these emissions from the existing fleet. Because the vast 
majority of current NRLM diesel fuel already meets this specification, 
the NOX and PM emission reductions would be small. Also, the 
impact of cetane on NOX and PM emissions appears to be very 
weak or nonexistent for diesel engines equipped with EGR. Thus, the 
positive emission impact of this specification would likely decrease 
over time as these engines gradually dominate the in-use fleet.
    ASTM already applies a cetane number specification of 40 to NRLM 
diesel fuel, which in general is more stringent than the similar 40 
cetane index specification. Because of this, the vast majority of 
current NRLM diesel fuel already meets the EPA cetane index/aromatics 
specification for highway diesel fuel. Thus, the proposed requirement 
would have an actual impact only on a limited number of refiners and 
there would be little overall cost associated with producing fuel to 
meet the proposed cetane/aromatic requirement. In fact, as discussed in 
section 5.9 of the draft RIA, complying with the sulfur standards 
proposed today is expected to result in a small cetane increase, 
leaving little or no further control to meet the standard.
    In addition, we expect that if all NRLM fuel met the cetane index 
or aromatics specification as proposed, refiners would benefit from the 
ability to fungibly (mixed together) distribute highway and NRLM diesel 
fuels of like sulfur content. For that fraction of fuel that today does 
not meet this specification, the proposed requirement would eliminate 
the need to separately distribute fuels of different cetane/aromatics 
specifications that would otherwise need to occur. Requiring NRLM 
diesel fuel to meet this cetane index specification would thus give 
fuel distributors certainty in being able to combine shipments of 
highway and NRLM diesel fuels. Overall, we believe that the economic 
benefits from more efficient fuel distribution would likely exceed the 
cost of refining the small volume of NRLM diesel fuel that might not 
currently meet the cetane index or aromatics content specification.
    We request comment on the costs and benefits of our proposal to 
extend the cetane index and alternative aromatics standard applicable 
to highway diesel fuel to NRLM diesel fuel.

B. Program Design and Structure

    In addition to the proposed content standards and their timing, the 
program must be designed and structured carefully to achieve the 
overall principles of this proposed nonroad diesel fuel program. The 
health and welfare benefits and the need for widespread availability of 
15 ppm highway diesel fuel must be maintained. This will only happen if 
the program is designed such that the amount of low sulfur fuel 
expected to be produced under the highway diesel program is in fact 
produced. Likewise, the benefits of the low sulfur diesel program 
proposed today will only be achieved if the program is designed such 
that the volume of diesel fuel consumed by NRLM engines is matched by 
the production and distribution of at least the same volume of diesel 
fuel produced to the appropriate low sulfur levels. At the same time, 
promoting the efficiency of the distribution system calls for fungible 
distribution of physically similar products, and minimizing the need 
for segregation of products in the distribution system.
1. Background
    Prior to the highway diesel sulfur standard that took effect in 
1993, most number 2 distillate fuel was produced to essentially the 
same specifications, shipped fungibly, and used interchangeably for 
highway diesel engines, nonroad diesel engines, locomotive and marine 
diesel engines and heating oil applications. Beginning in 1993, highway 
diesel fuel was required to meet a 500 ppm sulfur cap and was 
segregated from other distillate fuels as it left the refinery by the 
use of a visible level of dye solvent red 164 in all non-highway 
distillate.\234\ At about the same time, the IRS similarly required 
non-highway diesel fuel to be dyed red to a much higher concentration 
prior to retail sale to distinguish it from highway diesel fuel for 
excise tax purposes. Dyed non-highway fuel is exempt from this tax. 
This splitting of the distillate pool necessitated changes in the 
distribution system to ship and store the now distinct products 
separately. In some parts of the country where the costs to segregate 
non-highway diesel fuel from highway diesel fuel could not be 
justified, both fuels have been produced to the highway 
specifications.\235\
---------------------------------------------------------------------------

    \234\ Non-highway distillate for the purposes of this proposal 
refers to all diesel fuel and distillate used for nonroad, 
locomotive, marine and heating oil purposes; in other words, all 
number 1 or number 2 distillate other than that used for highway 
purposes, and excluding jet fuels.
    \235\ Diesel fuel produced to highway specifications but used 
for non-highway purposes is referred to as ``spill-over.'' It leaves 
the refinery gate and is fungibly distributed as if it were highway 
diesel fuel, and is typically dyed at a point later in the 
distribution system. Once it is dyed it is no longer available for 
use in highway vehicles, and is not part of the supply of highway 
fuel. Based on the most recent EIA data, roughly 15 percent of fuel 
produced to highway specifications is spillover, representing nearly 
a third of non-highway consumption.
---------------------------------------------------------------------------

    This proposal would set new specifications for nonroad, locomotive, 
and marine diesel fuel. However, currently there is no grade of diesel 
fuel which is produced and marketed as a distinguishable grade for NRLM 
uses. It is typically produced and shipped fungibly with other 
distillate used for heating oil purposes, and it is all dyed red in 
accordance with EPA and IRS regulations. Therefore, in order to control 
the sulfur content of NRLM, but

[[Page 28404]]

not heating oil, this proposal requires some means of distinguishing 
fuel used for the two purposes. This is similar to the situation faced 
in 1993 in the case of highway diesel fuel. The solution in 1993 for 
highway diesel fuel was to dye the non-highway distillate. As discussed 
below, a similar approach is proposed today to identify and distinguish 
heating oil from NRLM.
    This proposal would control the sulfur level of NRLM diesel fuel to 
500 ppm in 2007, the same level currently applicable to highway diesel 
fuel, and the same level as up to 20 percent of the highway diesel fuel 
pool from June 1, 2006, through December 31, 2009. Under the current 
provisions of the highway diesel rule, this 500 ppm nonroad diesel fuel 
would have to be dyed red at the refinery gate and distributed 
separately from 500 ppm highway diesel fuel.
    Continuing to implement this dye provision would allow for simple 
enforcement of both the proposed NRLM standard and the more stringent 
highway standards during this timeframe. Clear, undyed diesel fuel 
would have to meet the 80/20 ratio of 15 ppm and 500 ppm applicable to 
highway fuel, and diesel fuel (dyed red) would have to meet the 500 ppm 
standard applicable to NRLM. Continuing the current dye provisions 
would therefore ensure that the intended benefits of both programs were 
achieved. However, maintaining this dye distinction would also require 
segregation of a new grade of diesel fuel, 500 ppm NRLM, throughout the 
entire distribution system. The costs of requiring segregation of two 
otherwise identical fuels throughout the entire distribution system 
could be quite substantial.\236\
---------------------------------------------------------------------------

    \236\ Under the highway program the potential exists to add a 
third grade of diesel fuel in an estimated 40% of the country, and 
we projected one-time tankage and distribution system costs of $1.05 
billion to accomplish this. Using similar assumptions, to add a 
second 500 ppm grade nationwide would cost in excess of $2 billion. 
This assumes that the capability exists to add such new tankage.
---------------------------------------------------------------------------

    In order to avoid adding unnecessary cost to the fuel distribution 
system, we are proposing that the current requirement that non-highway 
distillate fuels be dyed at the refinery gate be made voluntary 
effective June 1, 2006.\237\ However, in its place we are proposing an 
alternate means for refiners to differentiate their highway diesel fuel 
from NRLM diesel fuel (see IV.B.3 below). Where it is feasible and cost 
effective to continue to dye and segregate their nonroad fuel, we 
propose that refiners and importers may continue this option.
---------------------------------------------------------------------------

    \237\ The IRS requirements concerning dyeing of non-highway fuel 
prior to sale to consumers are not changed by this rulemaking.
---------------------------------------------------------------------------

    Since 500 ppm highway and NRLM diesel fuel would physically be the 
same, without some means of differentiating highway diesel fuel from 
NRLM diesel fuel, it would be impossible to maintain the benefits and 
program integrity of the 2006 highway diesel fuel program. Pre-2007 
model year highway vehicles are free to continue using 500 ppm fuel 
until 2010 as long as it is available. However, if a refiner produced 
all 500 ppm fuel, designating it as nonroad fuel, that refiner would 
have no obligation to produce any 15 ppm highway diesel fuel. Without 
an effective way of limiting the use in the highway market of 500 ppm 
diesel fuel produced as NRLM fuel (provided currently by the refinery 
gate dye requirement), much more 500 ppm fuel could, and likely would 
find its way into the highway market than would otherwise happen under 
the current highway program, displacing 15 ppm that would have 
otherwise been produced. This likely series of events would circumvent 
the 80/20 intent of the highway rule and sacrifice some of the 
resulting PM and SO3 emission benefits of that program. 
Perhaps more importantly, if this occurred to any significant degree, 
it could also undermine the integrity of the highway program by failing 
to ensure adequate availability of 15 ppm fuel nationwide for the 
vehicles that need it.
2. Proposed Fuel Program Design and Structure
a. Program Beginning June 1, 2007
    To avoid the costs associated with segregating 500 ppm NRLM diesel 
fuel from 500 ppm highway fuel, we propose that the existing 
requirement that NRLM diesel fuel be dyed leaving the refinery would be 
made voluntary. We propose that this change could occur as early as 
June 1, 2006. In its place we propose that a baseline volume percentage 
of non-highway diesel fuel would be established and enforced for each 
refinery and importer. The baseline percentage would be based on a 
historical average for a refinery or importer. The baseline percentage 
of non-highway diesel fuel would then be used to identify the amount of 
500 ppm diesel fuel produced by that refinery or importer that is 
subject to the NRLM requirements and the amount of 500 ppm fuel is 
subject to the highway requirements. As detailed below, in conjunction 
with a marker to prevent the use of heating oil in nonroad equipment, 
the baseline percentage would effectively protect the benefits and 
integrity of the highway program, ensure that the benefits of the first 
step of NRLM diesel fuel to 500 ppm sulfur would be obtained, and would 
enable the efficient, fungible distribution of like grades of fuel. A 
discussion of this proposal follows, beginning with the introduction of 
a fuel marker for heating oil.
i. Use of A Marker to Differentiate Heating Oil From NRLM
    If all NRLM diesel fuel were required to meet the 500 ppm standard 
beginning June 1, 2007, then heating oil and NRLM diesel fuel could be 
differentiated merely on the basis of their sulfur levels. However, 
this proposal would allow the limited production of high-sulfur NRLM 
fuel by small refiners, and by other refiners through the use of 
credits between 2007 and 2010 (see section IV.B.2.b). To ensure that 
the only high sulfur diesel fuel used in nonroad, locomotive, and 
marine diesel engines is high sulfur NRLM and not heating oil, it would 
be necessary for parties in the distribution system, and for EPA, to be 
able to distinguish heating oil from high-sulfur NRLM diesel fuel. One 
way of ensuring that these fuels remain segregated in the distribution 
system would be to require that either a dye or a marker be added to 
heating oil to distinguish it from NRLM diesel fuel during the period 
of 2007 through 2010.\238\ There is no differentiation today between 
fuel used for NRLM uses and heating oil. Both are typically produced to 
the same sulfur specification today, and both are required to have the 
same red dye added prior to distribution and sale.\239\ As a result, 
the dye or marker would have to be different from the current red dye 
requirement.
---------------------------------------------------------------------------

    \238\ A marker is an additive which is phosphorescent or has 
some other property which allows it to be easily detected, though 
not necessarily visible to the naked eye. A dye is intended to be 
visibly identified by the naked eye.
    \239\ There may be some exceptions where a refiner produces a 
unique grade of distillate fuel solely for heating oil purposes.
---------------------------------------------------------------------------

    There are a number of types of dyes and markers. Visible dyes are 
most common, are inexpensive, and are easily detected. Invisible 
markers are beginning to see more use in branded fuels and are somewhat 
more expensive than visible markers. Such markers are detected either 
by the addition of a chemical reagent or by their fluorescence when 
subjected to near-infra-red or ultraviolet light. Some chemical-based 
detection methods are suitable for use in the field. Others must

[[Page 28405]]

be conducted in the laboratory due to the complexity of the detection 
process or concerns regarding the toxicity of the reagents used to 
reveal the presence of the marker. Near-infra-red and ultra-violet 
flourescent markers can be easily detected in the field using a small 
device and after brief training of the operator. There are also more 
exotic markers available such as those based on immunoassay, and 
isotopic or molecular enhancement. Such markers typically need to be 
detected by laboratory analysis.
    Using a second dye for segregation of heating oil based on visual 
identification raises certain challenges. Most dye colors that provide 
a strong visible trace in fuels are already in use for different fuel 
applications. More importantly, mixing two fuels containing different 
strong dyes can result in interference between the two dyes rendering 
identification of the presence of either dye difficult. Yet, the mixing 
of NRLM diesel fuel into heating oil for eventual sale as heating oil 
would be an acceptable and often an economically desirable practice. 
Furthermore, to avoid interfering with the IRS tax code, it would be 
advantageous to maintain the current red color. Based on these 
considerations, the best approach to prevent the use of heating oil as 
NRLM diesel fuel would appear to be requiring the addition to heating 
oil of either a dye that does not impart a significant color to diesel 
fuel or a marker that imparts no color at all. The dye or marker would 
be added at the refinery gate, just as visible evidence of the red dye 
is required today. Fuel containing the marker would be segregated from 
highway and NRLM diesel fuel and would be prohibited from use in 
highway, nonroad, locomotive, or marine application.
    Effective in August 2002, the European Union (EU) enacted a marker 
requirement for diesel fuel that is taxed at a lower rate (which 
applies in all of the EU member states).\240\ The marker selected by 
the EU is N-ethyl-N-[2-[1-(2-methylpropoxy)ethoxyl]-4-phenylazo]-
benzeneamine.\241\ This compound is also referred to as solvent yellow 
124 or the Euromarker. We propose that beginning June 1, 2007, solvent 
yellow 124 must be added to heating oil in the U.S. We propose that it 
be added in a concentration of 6 milligrams per liter, the same 
treatment rate as required by the EU. This would ensure adequate 
detection in the distribution system even if diluted by a factor of 50. 
A level of 0.1 milligrams per liter would therefore be used as a 
threshold level to identify heating oil--below this level incidental 
contamination would be assumed to have occurred and the prohibition on 
use in highway, nonroad, locomotive, or marine applications would not 
apply. Despite its name, solvent yellow 124 does not impart a strong 
color to diesel fuel when used at the proposed concentration. 
Therefore, we do not expect that its use in diesel fuel containing the 
IRS-specified red dye would interfere with the use of the red dye by 
IRS to identify non-taxed fuels. We request comment on this assessment.
---------------------------------------------------------------------------

    \240\ The European Union marker legislation, 2001/574/EC, 
document C(2001) 1728, was published in the European Council 
Official Journal, L203 28.072001.
    \241\ Opinion on Selection of a Community-wide Mineral Oils 
Marking System, (``Euromarker''), European Union Scientific 
Committee for Toxicity, Ecotoxicity and the Environment plenary 
meeting, September 28, 1999.
---------------------------------------------------------------------------

    Solvent yellow 124 is chemically similar to other additives used in 
gasoline and diesel fuel, and has been registered by EPA as a fuel 
additive under 40 CFR part 79. Its products of combustion would not be 
anticipated to have an adverse impact on emission control devices, such 
as a catalytic converter. In addition, extensive evaluation and testing 
of solvent yellow 124 was conducted by the EC. This included combustion 
testing which showed no detectable difference between the emissions 
from marked and unmarked fuel. We understand that Norway specifically 
evaluated the use of distillate fuel containing solvent yellow 124 for 
heating purposes and determined that the presence of the Eurmarker did 
not cause an increase in harmful emissions from heating equipment. 
Based on the European experience with solvent yellow 124, we do not 
expect that there would be concerns regarding the compatibility of 
solvent yellow 124 in the U.S. fuel distribution system or for use in 
motor vehicle engines and other equipment such as in residential 
furnaces. We request comment on whether there are unique public health 
concern regarding the use of distillate fuel containing solvent yellow 
124. The European Union intends to review the use of Solvent yellow 124 
after December 2005, or earlier if any health and safety or 
environmental concerns about its use are raised. We intend to keep 
abreast of such activities and may initiate our own review of the use 
of solvent yellow 124 depending on the European Union's findings.
    We also request comment on the extent to which jet fuel might 
become contaminated with solvent yellow 124 due to the presence of 
solvent yellow 124-containing fuels and jet fuel in the U.S. common 
carrier pipeline distribution system, and whether such contamination 
would raise concerns for the operation of jet engines. Due to safety 
concerns, jet fuel is held to very strict standards regarding the 
allowable presence of contaminants and additives. For example, the 
Department of Defense maintains a zero-tolerance for any contamination 
of jet fuel with the red dye required by the IRS (and EPA) which is 
chemically similar to solvent yellow 124. We are not aware that any 
testing has been done to date to assess whether solvent yellow 124 does 
raise similar concerns, and we request comment with any supporting data 
on this issue.
    We do not believe that there any significant pathways for such 
contamination to take place other than by potential human error. In 
addition, the fact that the fuel distribution industry in the U.S. has 
been successful in managing contamination of jet fuel with red dye 
indicates that the potential contamination of jet fuel with solvent 
yellow 124 can also be successfully managed in the U.S. fuel 
distribution system. Therefore, we believe that our proposed use of 
solvent yellow 124 should not pose a significant risk to the 
maintenance of jet fuel purity. We request comment on this assessment.
    Solvent yellow 124 is marketed by several manufactures and is in 
current wide-scale use in the European community. We anticipate that 
these manufactures would have sufficient lead-time to increase their 
production of solvent yellow 124 to supply the need for fuel marker 
that would result from this proposal. We request comment on whether 
there are product licencing or other concerns regarding the manufacture 
of solvent yellow 124 for use under this proposed rule.
    We request comment on other potential markers that might be used to 
identify and segregate heating oil from NRLM fuel. In particular, we 
ask that as commenters raise potential concerns with the use of solvent 
yellow 124 that they also identify other possible markers that could 
overcome their concerns without raising others. One potential 
alternative we have identified is the Clir-Code[reg] marker system 
manufactured by ISOTAG Technologies Inc. The Clir-Code[reg] marker 
system has been used extensively in U.S. fuel and includes a field test 
that employs a hand-held near infra-red detector which does not require 
the use of any reagents. EPA deferred proposing the use of the Clir-
Code[reg] marker because we believe that the advantage of a simpler 
field test would not compensate for the increased

[[Page 28406]]

treatment cost relative to the use of solvent yellow 124. We 
furthermore seek comment on whether more than one marker could be 
selected, but which could all be detected using the same detection 
method. In this manner refiners would not be dependent on a sole 
supplier for the marker. Additional discussion of the rationale for our 
selection of solvent yellow 124 and the feasibility of its use is 
contained in Chapter 5 of the Draft RIA.
    Since marked heating oil would be a relatively small volume product 
in many parts of the country, we anticipate that it will not be carried 
everywhere as a separate fungible product. In places where it is not 
carried as a separate fungible grade we anticipate that most shipments 
of marked heating oil will be from refinery racks or other segregated 
shipments directly into end-user tankage. In these areas any distillate 
supplied from the fungible supply system for heating oil purposes will 
therefore likely be spillover from 500 ppm NRLM supply. Clearly, in 
those parts of the country with high demand for heating oil, 
particularly the Northeast and Pacific Northwest, we anticipate that 
marked heating oil will in fact be carried by the distribution system 
as a separate fungible product. To the extent this is the case, it is 
entirely possible that heating oil will no longer be produced to diesel 
fuel cetane or aromatic specifications, reducing production costs. The 
most difficult to desulfurize streams in a refinery are in fact those 
that are low in cetane and high in aromatics. Shifting these streams to 
a unique heating oil product can therefore reduce desulfurization 
costs, while still producing a high quality heating oil (though we have 
not reflected this in our cost analysis in section V.)
ii. Non-highway Distillate Baseline Cap
    As discussed above, we are proposing use of a marker in heating oil 
to effectively distinguish uncontrolled heating oil from NRLM fuel, so 
that the NRLM standards can be enforced throughout the distribution 
system and at the end-user. However, in order to allow for the fungible 
distribution of highway diesel fuel and NRLM, and continue to have 
enforceable highway diesel fuel standards in the absence of a NRLM dye 
requirement, we are proposing that a non-highway distillate baseline 
percentage be established for each refinery and importer in the 
country. This non-highway baseline would be defined as the volume 
percentage of all diesel fuel and heating oil (number 1 and number 2) 
that a refinery or importer produced or imported during the specified 
baseline period that was dyed for non-highway purposes.
    We propose that if a refiner chooses to fungibly distribute its 
NRLM and highway fuels, then under the first step of the nonroad 
program (June 1, 2007--June 1, 2010), the volume of diesel fuel 
represented by its non-highway baseline percentage would have to either 
meet the 500 ppm NRLM standard or be marked as heating oil. All the 
remaining production would have to meet the requirements of the highway 
fuel program (i.e., 80 percent of this fuel would have to meet a 15 ppm 
sulfur cap). As we recognized in the highway rule, some variation in 
the production of highway and non-highway diesel fuel is normal from 
year to year. As a result, in any given year it may be possible that a 
refiner is unable to produce the amount of 15 ppm diesel fuel required 
to meet its highway requirement (80% of 100% minus the non-highway 
baseline) simply because of this normal variation. The provisions of 
the highway diesel rule already allow for a 5% shortfall in the 
production of 15 ppm fuel in a year as long as it is made up in the 
following year. We seek comment on whether any additional flexibility 
beyond that provided in the highway rule is appropriate to account for 
normal fluctuations in refinery output.
    An example will help to explain the use of the baseline. Assume the 
baseline non-highway percentage has been established as discussed below 
and is 40%. That means 40% of the total diesel fuel production in the 
baseline years was non-highway fuel, dyed at the refinery gate. If the 
refinery then produced a total of 100,000,000 gallons of diesel fuel in 
2008, 40,000,000 gallons would be its applicable non-highway baseline. 
If it then produced and marked 10,000,000 gallons as heating oil, 
30,000,000 gallons of the remaining diesel fuel (dyed or undyed) would 
be subject to the NRLM standard of 500 ppm, and all the remaining 
diesel fuel, 60,000,000 gallons, would be considered highway diesel 
fuel and would have to meet the applicable 80/20 requirements.
    We propose that a refiner, for each of its refineries, would need 
to choose either to continue to dye all of its NRLM fuel at the 
refinery gate, or to apply the non-highway baseline approach to all of 
its production. If a refinery's production could be split between these 
two options, the refiner could avoid the cap on NRLM imposed by the 
baseline percentage by dyeing additional volumes over its baseline, for 
example at their refinery rack or co-located terminal. The result could 
be a diversion of extra 500 ppm fuel to the highway market while the 
dyed 500 ppm fuel was used to serve the nonroad market, resulting in 
little or no production of 15 ppm highway diesel fuel. Therefore, the 
choice of whether to dye all of their 500 ppm NRLM fuel at the refinery 
gate, or comply with the non-highway distillate baseline would have to 
be made in advance. We propose that compliance with the baseline be 
determined on an annual basis. We therefore also propose that the 
decision of whether to dye NRLM 500 ppm fuel at the refinery gate or 
comply with the baseline could also be made on an annual basis.
    This approach allows a refinery's production of 500 ppm NRLM fuel 
and heating oil to remain flexible in response to market demand, while 
ensuring that the proportion of fuel they produce in the future to 
highway and non-highway requirements remains consistent with their 
historical baseline production. Since the non-highway baseline is set 
as a percentage of production, the actual volume needed for compliance 
with this baseline would rise and fall with the refinery's total 
production of diesel fuel. In this way, it would provide refineries 
with flexibility similar to that under the 80/20 volume percentage 
provisions of the highway rule. If total production of diesel fuel 
decreased, the absolute volume of diesel fuel which had to be produced 
to highway or NRLM specifications would decrease. If total production 
increased, the amount of diesel fuel subject to the 80/20 highway and 
the NRLM standards would also increase. A refiner wishing not to be 
limited to this non-highway distillate baseline percentage of 
production could elect to segregate and dye its NRLM diesel fuel at the 
refinery gate.
    Like the current dye requirement, this approach would focus 
compliance with the highway and NRLM requirements on the refinery or 
importer. Once undyed 500 ppm or 15 ppm diesel fuel was produced or 
imported and accounted for under the baseline percentage approach, it 
could be mixed and shipped fungibly, and sold to either the highway or 
the NRLM diesel fuel market by anyone further down the distribution 
system. This would provide a significant degree of market flexibility 
to refiners and distributors and enable the efficient distribution of 
diesel fuel. Compliance with the non-highway baseline would be enforced 
at the refinery gate in the same manner as the current 2006 highway 
provisions. With the marker for heating oil, compliance with the 15 ppm 
and 500 ppm standards could also be enforced through to the

[[Page 28407]]

end-user. But most importantly, this approach would maintain the health 
benefits and fuel availability needs of the highway diesel fuel 
program, because the overall volume of highway diesel fuel produced to 
the 15 ppm cap would be maintained.
iii. Setting the Non-highway Distillate Baseline
    The purpose of the non-highway baseline is to identify a historical 
level of non-highway production occurring prior to implementation of 
the provisions of this proposal, for use as a baseline after such 
implementation. We propose to determine the non-highway baseline 
percentage for each refinery by averaging the volume of dyed diesel 
fuel and heating oil (number 1 and number 2, excluding jet fuel and 
exported fuel) that it produced or imported annually over the three 
year period from January 1, 2003, through December 31, 2005, and 
dividing that volume by the average of all diesel fuel and heating oil 
(number 1 and number 2, excluding jet fuel and exported fuel) it 
produced or imported annually over the same period (and then multiplied 
by 100).\242\ By using a multi-year average, variations that might 
otherwise occur from year to year in a refinery's production will get 
averaged out. Importers would establish a separate baseline for each 
area of importation.\243\
---------------------------------------------------------------------------

    \242\ Specialty fuels such as JP-5, JP-8 and F76 are in some 
instances also used in nonroad diesel equipment today. However, our 
expectation is that the majority of this fuel is today and will be 
in the future continue to be used in tactical military equipment 
that would be exempted from the provisions of this proposal. 
Consequently, we propose that these fuels would not be counted in 
either setting the baseline or in determining compliance with the 
baseline.
    \243\ The areas would be defined as the credit trading areas 
(CTAs) as defined in the highway rule.
---------------------------------------------------------------------------

    Selecting a baseline period prior to finalization of the final rule 
would help to prevent the possibility of entities inappropriately 
adjusting their operations solely for the purpose of modifying their 
baseline. At the same time, setting a baseline period as close to the 
implementation date as possible helps to capture the most recent 
changes in the industry's production patterns. The proposed period of 
January 1, 2003, through December 31, 2005, is split roughly equally 
between production prior to the final rule and production after the 
final rule to appropriately balance these competing objectives. One 
advantage of ending the baseline period on December 31, 2005, is that 
it allows the opportunity for refiners to generate credit for the early 
production of 500 ppm NRLM fuel after that date, and at the same time 
avoid having to dye it at the refinery gate. The three year period 
serves to limit any potential actions to inappropriately adjust the 
baseline that a refinery might otherwise attempt. A refiner or importer 
would have to dye and sell a greater fraction of its fuel to the non-
highway market over an extended period of time to significantly modify 
its baseline. The potential financial loss associated with this, 
particularly if other refineries or importers tried to do the same 
thing, would likely be prohibitive.
    At the same time, we anticipate that a number of refiners may be 
changing their highway diesel production volumes as they comply with 
the highway diesel fuel standards in 2006. To the extent that a refiner 
planned to lower its highway production in 2006, a non-highway baseline 
set based on 2003-5 data would penalize them by forcing them to 
continue to meet the highway requirements for a greater volume, based 
on their pre-2006 production pattern. To avoid this situation, we 
propose that refiners would be allowed to set their non-highway 
baseline percentage using June 1, 2006, through May 31, 2007, as the 
baseline time period. By doing so the refinery's baseline would 
automatically take into account changes made for compliance with the 
2006 highway standard. It would, however, preclude that refinery from 
participating in the early NRLM credit program prior to June 1, 2007, 
using the baseline approach, and would require them to continue dyeing 
their NRLM at the refinery gate until June 1, 2007, since that is the 
period during which the baseline was being established. Since the 
purpose of this option is to provide an opportunity to account for the 
physical changes refineries make in complying with the highway rule, we 
propose that this option would only apply to refiners and not 
importers.
    Each refinery and importer would have to submit its application for 
a non-highway baseline to EPA by February 28, 2006, along with the 
supporting information. If the refinery elected to use the optional 
baseline period, we propose that the refinery would have to submit its 
application for a non-highway baseline to EPA by August 1, 2007. EPA 
would then approve these baselines by June 1, 2006, and any optional 
baselines by December 1, 2007. We propose that any refinery or importer 
which was not in operation for the full period of January 1, 2003, 
through December 31, 2005, would establish their baseline using data 
from the period they were in operation, as long as that period was 
greater than or equal to 12 months. The 12 months need not be 
continuous. Any refinery or importer unable to establish a baseline 
during this period would have to comply using the dye alternative. In 
the case of a new or restarted refinery or new importer, we propose to 
assign a non-highway baseline percentage reflecting the projected 
average production of non-highway fuel in 2004 for their region of the 
country. We propose to use the credit trading areas (CTAs) as defined 
in the highway Based on data from the Department of Energy's Energy 
Information Agency (EIA) on the current production of low and high 
sulfur diesel fuel and heating oil, and EIA and EPA projections of 
future fuel use, these PADD average non-highway baseline would be as 
shown in Table IV-1.
---------------------------------------------------------------------------

    \244\ A value of zero is proposed for California, since we 
anticipate that all non-highway diesel fuel in California will be 
covered by the same State standards applicable to highway diesel 
fuel during this time period.

                                                   Table IV-1--Non-highway Baseline for New Refineries
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                            Oregon and                                      California
                 PADD 1                       PADD 2          PADD 3          PADD 4        Washington        Alaska          Hawaii           \244\
--------------------------------------------------------------------------------------------------------------------------------------------------------
41%.....................................             20%             26%             13%             21%             68%             40%              0%
--------------------------------------------------------------------------------------------------------------------------------------------------------

    In discussions with various refiners, there was a strong interest 
in allowing refiners with multiple refineries and importers with 
multiple points of import to aggregate the baselines across all of 
their facilities nationwide. However, since the baselines determine how 
much of a refineries production must comply with the highway standards, 
allowing nationwide aggregation of the baselines would have the same 
impact as allowing nationwide

[[Page 28408]]

averaging, banking, and trading of credits under the highway rule. That 
approach was rejected in the highway rule due to the negative impact it 
would have on the nationwide availability of 15 ppm highway diesel 
fuel. For the same reason we are not proposing to allow nationwide 
aggregation of the non-highway baselines. However, in the highway rule, 
we do allow credit trading within certain credit trading areas (CTAs). 
We seek comment on allowing the aggregation of non-highway baselines 
within the same CTA and how such aggregation could be accomplished. We 
also seek comment on whether a trading program could be established 
that allowed for refiners with only one refinery within a CTA to 
benefit from similar flexibility, and whether some reasonable 
restrictions on refiners who aggregate baselines are needed to protect 
the integrity of the highway program.
    EPA requests comments on the provisions described above for 
establishing the non-highway baseline percentage for each refinery and 
importer. We also request comment on any alternative provisions that 
could be used to accomplish the objectives discussed above.
iv. Diesel Sulfur Credit Banking, and Trading Provisions for 2007
    This proposal includes provisions for refiners and importers to 
generate early credits for production of 500 ppm NRLM fuel prior to 
June 1, 2007. This will provide implementation flexibility at the start 
of the 500 ppm NRLM standard in 2007. These credits would be tradeable 
and could be used to delay compliance with either the 500 ppm NRLM 
standard in 2007 or the 15 ppm nonroad standard in 2010. The proposed 
banking and trading provisions would allow an individual refinery to 
purchase credits and delay compliance. This would allow for a somewhat 
smoother transition at the start of the program, with some refineries 
complying early, others on time, and others a little later. 
Nevertheless, on average the overall benefits of the program would be 
obtained or perhaps increased, and some environmental benefits could be 
achieved earlier than expected. Perhaps the most advantageous use of 
these credit provisions, however, might be for individual refineries to 
utilize available credits to permit the continued sale of otherwise 
off-spec product during the start up of the program when they are still 
adjusting their operations for consistent production to the new sulfur 
standards.

Credit Generation

    We propose two ways to generate credits that can be used to allow 
for high sulfur NRLM fuel to be produced after June 1, 2007. First, we 
propose that a refinery or importer can generate credit for early 
production of NRLM diesel fuel to the 500 standard from June 1, 2006, 
through May 31, 2007. Credits would be calculated either using the non-
highway baseline approach or by counting 500 ppm NRLM dyed at the 
refinery gate. Refiners that chose to establish their non-highway 
baseline using the June 1, 2006--May 31, 2007, baseline period would be 
precluded from generating any early credits using the non-highway 
baseline approach. Second, under the small refiner hardship provisions 
described below in subsection C, small refiners could generate credits 
for any production of NRLM fuel to the 500 ppm standard from June 1, 
2007, through May 31, 2010. In either case, credits could be banked for 
future use, or traded to any other refinery or importer nationwide. For 
early credits and small refinery credits generated using the non-
highway baseline approach, these credits would be calculated according 
to the following formula:
    High-Sulfur NRLM credits \245\ = (15 ppm production volume + 500 
ppm production volume )--(100%-non-highway baseline percentage) * 
(total 1 and 2 distillate production excluding jet 
fuel and exported fuel).
---------------------------------------------------------------------------

    \245\ For the purposes of this proposal, the credits are labeled 
on the basis of their use in order to follow the convention used in 
the highway rule. A high-sulfur credit is generated through the 
production of one gallon of 500 ppm NRLM fuel and allows the 
production of one gallon of high sulfur NRLM fuel.
---------------------------------------------------------------------------

    Early credits or small refinery credits generated using the dye 
option would be calculated using the following formula: High-Sulfur 
NRLM credits = 500 ppm production volume dyed at the refinery gate.
    If the excess production was 15 ppm fuel instead of 500 ppm fuel, 
the refiner would of course still have the option of using it to 
generate 500 ppm highway credits under the existing highway diesel 
provisions. Credit could not be earned under both programs.

Credit Use

    There would be two ways in which refiners could use high-sulfur 
NRLM credits. First, we propose that these credits could be used during 
the period from June 1, 2007--May 31, 2010, to continue to produce high 
sulfur NRLM diesel fuel. Any high sulfur NRLM fuel produced, however, 
would have to be dyed red at the refinery gate, kept segregated from 
other fuels in the distribution system, and tracked through the use of 
unique codes on product transfer documents.
    Only at the point in the distribution system where NRLM fuel has 
been dyed to IRS specifications for excise tax purposes (e.g., after a 
terminal or bulk plant) do we propose that high sulfur and 500 ppm 
sulfur NRLM fuels could be commingled. Such commingling will not 
diminish the PM and SO3 emission reductions or other 
benefits associated with the 500 ppm sulfur standard. However, in order 
to ensure that owners of nonroad equipment can be confident in knowing 
whether the fuel being purchased meets the 500 ppm cap, the PTD and 
labels for any commingled fuel will have to indicate that the sulfur 
level exceeds 500 ppm. This is particularly a concern for some 2008 and 
later model year equipment that may need to run on 500 ppm or lower 
sulfur fuel in order to achieve the emission benefits in-use of the 
standards proposed today, as discussed in section III.
    In most cases we anticipate that the distribution costs associated 
with segregating such a small volume product will prevent high-sulfur 
NRLM from being carried in the fungible distribution system. As a 
result, we anticipate that only those refineries that have their own 
segregated distribution system could continue to produce solely high 
sulfur NRLM fuel after June 1, 2007. Since there are few refineries set 
up to accomplish this, our expectation is that the most likely manner 
in which refiners will be able to use high-sulfur NRLM credits will be 
through sales made directly from their on-site fuel rack or co-located 
terminal. Nevertheless, in order to have confidence that refiners are 
making the transition to 500 ppm for NRLM uses, we seek comment on 
whether caps on the use of credits would be necessary. In particular, 
we seek comment on placing a cap on the use of credits at 25 percent of 
its non-highway baseline, less marked heating oil, beginning June 1, 
2008.
    The second way in which refiners and importer could use high-sulfur 
NRLM credits is by banking them for use during the June 1, 2010--May 
31, 2012, period. During this period they could continue producing 500 
ppm fuel subject to the usage restrictions that apply during that 
period, as discussed in subsection B.2.b.ii below. This use of high-
sulfur credits would provide a cost-effective environmental benefit, 
since credits generated from the reduction of sulfur levels from high 
sulfur to 500 ppm would be used to

[[Page 28409]]

offset the much smaller increment of sulfur control from 500 ppm down 
to 15 ppm.
b. 2010
    After June 1, 2010, the fuel standards situation is simplified 
considerably and the fuel program structure can therefore also be 
simplified. The need for the non-highway baseline percentage 
disappears, since all highway and nonroad diesel fuel must meet the 
same 15 ppm cap. Furthermore, the only high sulfur distillate remaining 
in the market should be heating oil, since we are proposing that high 
sulfur diesel fuel no longer be permitted to be used in any NRLM 
equipment. Heating oil would have to be kept segregated. Preventing its 
use in NRLM equipment could be enforced on the basis of sulfur level, 
avoiding the need for a unique marker to be added to heating oil.
    After June 1, 2010, under this proposal locomotive and marine 
diesel fuel would be allowed to remain at the 500 ppm level. In 
addition, assuming we allowed the continued production and use of 500 
ppm nonroad diesel fuel through the small refiner hardship provisions 
discussed in subsection C and fuel credit provisions, 500 ppm nonoad 
fuel would continue to exist in the distribution system as late as May 
31, 2014. A refiner could produce 500 ppm diesel fuel without the use 
of credits for the intended use in locomotive and marine applications, 
but if this 500 ppm fuel later made its way into nonroad equipment, 
less 15 ppm nonroad fuel would be produced and the full benefits of the 
15 ppm nonroad standard would not be achieved. If this happened to a 
large enough extent it could call into question the adequate supply of 
15 ppm for nonroad purposes beginning in 2010. Thus, some method is 
needed to differentiate locomotive and marine 500 ppm diesel fuel from 
nonroad 500 ppm diesel fuel after June 1, 2010. EPA is proposing to use 
a marker for this purpose.
i. A Marker To Differentiate Locomotive and Marine Diesel From Nonroad 
Diesel
    This proposal would allow the limited production of 500 ppm nonroad 
diesel fuel by small refiners and by other refiners through the use of 
credits between 2010 and 2014 (see section IV.B.3.b). This 500 ppm fuel 
could only be used in pre-2011 model year nonroad diesel engines, and 
would have to be segregated from 15 ppm nonroad diesel fuel and 500 ppm 
locomotive and marine diesel fuel. To ensure compliance with the 
proposed segregation requirements for such fuel, it would be necessary 
for parties in the distribution system, and for EPA, to be able to 
distinguish such 500 ppm nonroad diesel fuel from 500 ppm locomotive 
and marine diesel fuel. Differentiating locomotive and marine diesel 
fuel from nonroad diesel fuel presents a very analogous situation, 
though perhaps on a smaller scale, to that described above for heating 
oil prior to June 1, 2010.\246\ As a result, we propose to use a marker 
to segregate locomotive and marine diesel fuel from 500 ppm nonroad 
diesel fuel beginning June 1, 2010. Since both fuels need to be dyed 
red for tax purposes prior to sale, for the reasons discussed above 
with respect to heating oil, we propose that solvent yellow 124 be used 
as the marker for locomotive and marine diesel fuel beginning June 1, 
2010. We propose that the marker would be required to be added at the 
refinery gate just as visible evidence of the red dye is required 
today, and fuel containing more than the trace concentration of 0.1 mg/
l of the marker would be prohibited from use in any nonroad 
application.
---------------------------------------------------------------------------

    \246\ Without the proposed marker requirement for locomotive and 
marine diesel fuel discussed in this section, we expect that there 
would be no physical difference between 500 ppm nonroad diesel fuel 
and 500 ppm locomotive and marine diesel fuel.
---------------------------------------------------------------------------

    Since marked locomotive and marine diesel fuel would be a 
relatively small volume product, we anticipate that in most parts of 
the distribution system it would not be carried as a separate product 
in the fungible distribution system. Therefore we anticipate that most 
shipments of marked locomotive and marine fuel would be from refinery 
racks or other segregated shipments directly into end-user tankage. Any 
diesel fuel supplied off the fungible supply system for locomotive and 
marine uses would therefore likely be spillover from 15 ppm nonroad or 
highway diesel supply.
    Since we anticipate that 500 ppm locomotive and marine diesel fuel 
will be a small volume product, not carried in the fungible 
distribution system, and only available in limited locations, we also 
seek comment on whether the approach of using a marker for locomotive 
and marine diesel fuel could be replaced with an alternative approach. 
Specifically, we seek comment on whether to just limit supply of 500 
ppm locomotive and marine diesel fuel to segregated shipments, with 
refineries being liable to ensure and keep records demonstrating that 
500 ppm fuel produced for locomotive and marine purposes was 
distributed solely for these purposes.
ii. Diesel Sulfur Credit Banking and Trading Provisions for 2010
    For the reasons described above for 2007, we are proposing a 
similar diesel sulfur credit banking and trading program for 2010. We 
propose that refiners and importers could generate early credit for 
production of 15 ppm nonroad diesel fuel prior to June 1, 2010. These 
credits could be used to delay compliance with the 15 ppm nonroad 
diesel standard in 2010. As in 2007, while it is possible that a 
refinery could entirely delay compliance with the 15 ppm standard in 
2010 through the use of credits, the most advantageous use of these 
credit provisions is likely to be the continued sale by individual 
refineries of otherwise off-spec product during the startup of the 2010 
program, when they are still adjusting their operations for consistent 
production to the 15 ppm sulfur standard.

Credit Generation

    Under this proposal, highway and NRLM fuels of like sulfur level 
would be allowed to be distributed fungibly, and as such would be 
indistinguishable. For example, prior to June 1, 2010, undyed 15 ppm 
diesel fuel would be distributed together whether or not it was later 
dyed for nonroad purposes. Consequently, we are proposing that credits 
for production of early 15 ppm nonroad diesel fuel prior to June 1, 
2010, be determined using the non-highway baseline. Any volume up to a 
refinery's total highway requirement (100 percent minus the non-highway 
baseline) would continue to be counted under the provisions of 2007 
highway diesel fuel program.\247\ Any production of 15 ppm fuel greater 
than this amount (100% minus the non-highway baseline) beginning June 
1, 2009 could be used to generate early nonroad credits.
---------------------------------------------------------------------------

    \247\ Under the highway program four gallons of excess 15 ppm 
diesel fuel produced or imported would generate one 500 ppm diesel 
fuel credit. This credit grants the refiner or importer the right to 
produce one additional gallon of undyed 500 ppm diesel fuel between 
June 1, 2006 and May 31, 2010. These credits can be used (or traded 
within the PADD in which they were generated) to produce or import 
less than 80% of its highway volume as 15 ppm fuel. This would 
continue under this proposal for any production up to (100% minus 
the non-highway baseline). For any volume of 15 ppm fuel greater 
than 100% minus the non-highway baseline a refiner could either 
receive gallon-for-gallon nonroad credit under this proposal, or 
treat it as highway fuel and receive 1:4 credit under the provisions 
of the highway rule.
---------------------------------------------------------------------------

    An example will help to explain the use of these credits. Assume 
the baseline non-highway percentage has been established at 40% and the 
refinery produces a total of 100,000,000 gallons of diesel fuel from 
June 1,

[[Page 28410]]

2009--December 31, 2009. Its applicable non-highway baseline would be 
40,000,000 gallons. If it then produced and marked 10,000,000 gallons 
of heating oil, 30,000,000 gallons of the remaining diesel fuel (dyed 
or undyed) would be subject to the NRLM standard of 500 ppm, and the 
remaining 60,000,000 gallons of diesel fuel would be considered highway 
diesel fuel and would have to meet the applicable 80/20 requirements 
(48,000,000 at 15 ppm and 12,000,000 at 500 ppm). If the refiner 
instead produced only 20,000,000 gallons of fuel to the 500 ppm NRLM 
standard and produced 70,000,000 gallons to the 15 ppm standard, then 
it would receive credit for the 10,000,000 gallons excess 15 ppm NRLM 
fuel that it produced. In this example the refiner could also earn 
3,000,000 highway credits for the excess production of 15 ppm highway 
fuel (1:4 ratio).
    In addition to this source of credits, we propose two other sources 
of credits to allow production of 500 ppm nonroad diesel fuel after 
June 1, 2010. First, as discussed in subsection B.3.a.iv above, high-
sulfur NRLM credits generated prior to June 1, 2010, could be converted 
into 500 ppm nonroad credits and carried over for use beginning June 1, 
2010. Second, under the small refiner hardship provisions described 
below in subsection C, small refiners could generate credits for any 
production of NRLM fuel to the 15 ppm standard from June 1, 2010, 
through May 31, 2012. These credits could be traded to any other 
refinery or importer nationwide.
    We seek comment on whether credits should be permitted to be 
generated prior to June 1, 2009. Our proposal would restrict the early 
credit period to just one year for two main reasons. First, any 15 ppm 
fuel produced prior to June 1, 2009, can be treated as highway diesel 
fuel and any credits generated on the fuel under the highway program 
can be traded under the highway credit program. We do not want the 
early nonroad credit provisions to detract from the smooth functioning 
of the highway diesel credit program. Second, we do not want the early 
credit provisions to undermine the availability of 15 ppm diesel fuel 
for nonroad applications in 2010. Allowing more than a years worth of 
credits to be generated, plus up to a years worth of high sulfur 
credits to be generated and carried over for use in 2010 would raise 
concerns that insufficient 15 ppm nonroad diesel fuel might be produced 
in 2010 to ensure availability everywhere nationwide. Nevertheless, we 
seek comment on extending the period for early credit generation and on 
this assessment.

Credit Use

    We propose that 500 ppm nonroad credits could be used on a gallon 
for gallon basis during the period from June 1, 2010-May 31, 2012, 
allowing continued production of 500 ppm nonroad diesel fuel. Small 
refiners could continue to produce 500 ppm nonroad diesel until June 1, 
2014, without credits. Any 500 ppm nonroad fuel produced would have to 
be dyed red at the refinery gate, kept segregated from other fuels in 
the distribution system, and tracked through the use of unique codes on 
product transfer documents all the way through to the end-user. 
Refiners wishing to produce 500 ppm fuel and sell it as nonroad would 
have to get EPA approval in advance demonstrating how they will ensure 
such segregation.
    Given the cost and burden associated with segregating 500 ppm 
nonroad diesel fuel as a separate product in the distribution system, 
we anticipate that the most likely manner in which refiners will be 
able to use 500 ppm nonroad credits will be through sales made directly 
from their on-site fuel rack.
    We request comment on all aspects of the proposed credit trading 
system.
c. 2014
    Beginning June 1, 2014, after all small refiner and credit 
provisions have ended, both the 15 ppm nonroad diesel fuel standard and 
the 500 ppm locomotive and marine diesel fuel standard could be 
enforced based on sulfur level throughout the distribution system and 
at the end-user. There would no longer be a need for a baseline, a 
marker, or a dye. Consequently, we are proposing that after May 31, 
2014, the different grades of diesel fuel, 15 ppm, 500 ppm, and high-
sulfur would merely have to be kept segregated in the distribution 
system.
3. Other Options Considered
    In developing the proposed program structure described above, we 
also evaluated a number of other possible approaches. Some of the 
alternatives discussed below would allow for even greater fuel 
fungibility, for example, extending to smaller volume products such as 
those produced through the use of credits. However, these alternative 
approaches would either place more restrictions on refinery operations, 
or raise significant enforcement and program integrity concerns. As a 
result, we are not proposing the following alternatives but seek 
comment on them, including ways to minimize or alleviate the concerns 
associated with them.
a. Highway Baseline and a NRLM Baseline for 2007
    The proposed program described above relies on a non-highway 
baseline percentage to distinguish highway fuel from NRLM fuel, and a 
marker to distinguish heating oil from NRLM fuel. In lieu of using a 
marker for heating oil, another approach would be to use a second 
baseline aimed at identifying the NRLM portion of non-highway diesel 
fuel. In this case a highway baseline would be established consistent 
with the non-highway baseline proposed above (100 percent minus the 
proposed non-highway baseline). The highway 80/20 standards would apply 
to this baseline. A second NRLM baseline would be established to which 
the 500 ppm NRLM standard would apply. The remaining diesel fuel 
percentage would be uncontrolled (i.e., it could be high sulfur). This 
approach would allow for greater fungibility of fuels with the same 
sulfur level. Not only could 500 ppm highway and 500 ppm NRLM fuel be 
distributed together, but high sulfur NRLM fuel produced through the 
credit and hardship provisions could be fungibly distributed with 
heating oil. Heating oil would not need to contain a marker. As a 
result, this approach would allow for greater flexibility in using the 
fuel credit and hardship provisions. The disadvantage, however, is that 
refiners would face additional burden when shifting into the heating 
oil market. An explanation of this approach follows.
i. Highway Baseline
    The highway baseline would be very analogous to the non-highway 
baseline proposed above. It would be calculated in the same way, except 
that it would be 100 percent minus the proposed non-highway baseline. 
The requirement that NRLM fuel be dyed at the refinery gate would 
become voluntary. From June 1, 2007, through May 31, 2010, any volume 
of 500 ppm fuel not dyed at the refinery gate would have to meet the 
80/20 highway provisions up to the refinery specific highway baseline 
percentage. The highway baseline percentage would be determined for 
each refinery and importer in the same manner as described above for 
the non-highway baseline.
ii. Nonroad, Locomotive, and Marine Baseline
    Under this approach, a refiner or importer would be assigned a NRLM 
baseline percentage. This baseline

[[Page 28411]]

percentage of a refinery's or importer's current high-sulfur diesel 
fuel and heating oil (number 1 and number 2) production would be deemed 
to be NRLM diesel fuel and thus, subject to the proposed 500 ppm cap 
beginning June 1, 2007. The remaining percentage would remain 
uncontrolled and would not need to contain a marker. A refiner's NRLM 
baseline percentage would be applied to the percentage of distillate 
not included in the highway baseline (i.e., the proposed non-highway 
baseline). For example, if a refiner's highway baseline was 50% and its 
NRLM baseline was also 50%, then 25% of its production would have to 
meet the 500 ppm NRLM standard.
    If a refiner chose not to use the NRLM baseline percentage, a 
refinery or importer would have to add the proposed marker and 
segregate their heating oil from any NRLM diesel fuel throughout the 
distribution system, including high sulfur NRLM diesel fuel (produced 
through the use of credits or by small refiners or refiners utilizing 
hardship provisions). The refinery would have to demonstrate that the 
fuel was segregated all the way through to the end-user and that the 
end-user used the fuel for legitimate heating oil purposes only. NRLM 
end-users would be prohibited from using any fuel with a marker.
    There are, however, certain difficulties in establishing an NRLM 
baseline percentage. Unlike the situation today where highway diesel 
fuel and non-highway distillates are accounted for based upon their 
different sulfur levels and the presence of red dye, there is no easy 
way to measure a given refinery's current production of NRLM diesel 
fuel as compared to their production of heating oil, in order to 
accurately establish an individual refinery baseline percentage. 
Generally the two fuels are produced and shipped as a single fuel. We 
considered whether refiners and importers could reliably track their 
high sulfur fuel through the distribution system and estimate the 
volumes used as diesel fuel and heating oil to establish individual 
refinery baselines. However, most high sulfur diesel fuel and heating 
oil is shipped by fungible carriers and we do not believe that 
sufficient data exist to accurately determine which refiner's fuel was 
actually consumed in either end-use. Discussion with several refiners 
have supported this belief. Therefore, we developed an approach that 
would assign each refinery a percentage of their current high-sulfur 
distillate production, based on the PADD they reside in, as their NRLM 
baseline. PADDs 1 and 3 would be combined due to the large amount of 
high sulfur non-highway diesel fuel shipped from PADD 3 to PADD 1 
today.
    Under this approach we would project consumption of NRLM diesel 
fuel and heating oil to determine the relative consumption of these two 
fuels by PADD. This would be the NRLM baseline assigned to refiners and 
importers in that PADD. This volume percentage of non-highway diesel 
fuel would then be considered NRLM and have to meet the proposed 500 
ppm cap beginning June 1, 2007. The remainder of the non-highway diesel 
fuel would remain uncontrolled by EPA and would only have to meet any 
applicable state sulfur standards for heating oil. If a refinery 
desired to only produce heating oil, then they could either purchase 
credits from other refineries or segregate and mark their heating oil.
    Using EIA estimated fuel consumption data for the year 2000, grown 
to 2008 using EPA NONROAD emission model growth rates for nonroad and 
EIA growth rates for other fuels, produces the NRLM baseline 
percentages shown in Table IV-2.

                                Table IV-2--NRLM Diesel Fuel Baseline Percentages
----------------------------------------------------------------------------------------------------------------
                                                                     Breakdown of High Sulfur Distillate Fuel
                                                                             Production  (In percent)
                              PADD                               -----------------------------------------------
                                                                                     Loco and
                                                                      Nonroad         marine         Combined
----------------------------------------------------------------------------------------------------------------
1 and 3.........................................................              26              16              42
2...............................................................              57              27              84
4...............................................................              67              29              96
5 (excluding Alaska)............................................              59              18              77
Alaska..........................................................              22              28              50
----------------------------------------------------------------------------------------------------------------

    One particular concern with this NRLM baseline approach is whether 
refiners can easily respond to above average demand for heating oil 
(e.g., in unusually cold winter). As today, any short-term, unexpected 
increases in demand will be made up from existing inventories of fuel. 
Today, if there are insufficient inventories of high sulfur fuel, 500 
ppm inventories are tapped as well. The same situation will continue to 
occur in the future. As a result, the issue is not one of being able to 
supply the market with sufficient fuel to meet demand, but rather what 
quality of fuel must be produced to build inventories back up after 
high demand has brought them down. This could be addressed in a number 
of ways. First, in setting the NRLM baseline itself we could make sure 
it is not too high and allows for sufficient volumes of high sulfur 
heating oil to be produced even in the event of an unusually cold 
winter. Second, we could allow credits to flow across the country 
through a nationwide credit trading program. This would allow the 
production of high sulfur fuel to likewise flow across the country to 
the places experiencing higher than normal demand. Third, provisions 
could be made for deficit carry over of credits. If demand for high 
sulfur fuel is unusually high in one year, a refiner could increase 
production to respond to that demand as long as it is made up the 
following year.
    Another concern raised by this baseline approach is the inability 
to accurately tailor it to each refinery's actual historical production 
of NRLM. This NRLM baseline approach does reflect the historical 
practice for the industry as a whole--refineries produced fungible high 
sulfur fuel for distribution as a common pool of fuel that was later 
sold as either NRLM or heating oil. However, it does not allow for 
refinery specific customization. The proposed non-highway baseline 
approach determines the specific non-highway percentage for each 
refinery, and the actual volume of marked and dyed heating oil is 
allowed to vary. The lack of individual specificity for the NRLM 
baseline approach, however, avoids the need to add a marker to heating 
oil.

[[Page 28412]]

iii. Combined Impact of Highway and NRLM Baselines
    These baselines, as with the proposed non-highway baseline, are set 
on the basis of a percentage of production. Therefore, as a refinery's 
overall production of diesel fuel rises and falls, the required volume 
of each grade of fuel will also rise and fall. Thus, the baselines are 
flexible enough to respond to changes in a refinery's market or 
situation. Furthermore, a nationwide credit trading program for 500 ppm 
NRLM fuel could be put in place, allowing refineries further 
flexibility to change production in response to consumer demand. To add 
additional flexibility we could allow for some deficit carry-over of 
NRLM credits. Finally, a refinery could always avoid use of the 
baselines entirely by dyeing or marking their fuel and ensuring that it 
is only used in appropriate end-uses.
    The combined effect of the highway baseline and NRLM baseline is 
shown in Table IV-3.

          Table IV-3--Combined Impact of the Highway and NRLM Baselines for June 1, 2007--May 31, 2010
----------------------------------------------------------------------------------------------------------------
            Sulfur level                                        Percentage requirement
----------------------------------------------------------------------------------------------------------------
15 ppm..............................   or = 80% x (highway baseline) or;
                                       or = 80% x All undyed diesel fuel (whichever is less)
15+500 ppm..........................   or = (highway baseline) + (NRLM baseline)(100% highway
                                       baseline) or;
                                      = All fuel without a marker and segregated through to the end-user
----------------------------------------------------------------------------------------------------------------

    An example will help to explain the use of these baselines. Assume 
a refinery in PADD 3 produces 100,000,000 gallons of diesel fuel and 
heating oil per year from 2003-5, 60 percent of which is undyed. Its 
highway baseline would thus be 60 percent of its total diesel fuel and 
heating oil production. Its NRLM baseline, assigned by EPA from Table 
IV-2, would be 42 percent applied to the remaining 40 percent of total 
distillate, or 16.8 percent of total distillate. If the refinery then 
continues to produce a total of 100,000,000 gallons of diesel fuel in 
2008, 60,000,000 gallons would be required to meet the highway 80/20 
standards, i.e., 48,000,000 at 15 ppm and 12,000,000 at 500 ppm. An 
additional 16.8 percent, or 16,800,000 gallons would be required to 
meet the 500 ppm NRLM standard, for a total required 500 ppm production 
of 28,800,000 gallons. Its remaining 23,200,000 gallons of production 
could remain uncontrolled and could be sold as heating oil or high 
sulfur NRLM. If the refiner reduced this 23,200,000 gallons to 500 ppm 
it would then earn credits that could be sold to another refiner.
b. Locomotive and Marine Baseline for 2010
    The proposed non-highway baseline percentage approach described 
above relies on a marker to distinguish locomotive and marine diesel 
fuel from nonroad diesel fuel after June 1, 2010. Just as in the 
alternative above, a baseline for locomotive and marine fuel could be 
used in lieu of a marker. The 2010 locomotive and marine baseline would 
be established by EPA and used in the same manner as described above 
for the NRLM baseline in 2007. Possible locomotive and marine baselines 
are shown in Table IV-2. The advantage of this baseline approach over 
the proposed approach is that it allows for the fungible distribution 
of 500 ppm locomotive and marine fuel with 500 ppm nonroad fuel 
produced through the credit and hardship provisions. As a result, this 
approach would allow for greater flexibility in using the diesel fuel 
credit and hardship provisions. The disadvantage, however, is that 
refiners wishing to produce locomotive and marine fuel in quantities 
larger than their baseline would have to purchase credits from other 
refiners.
    It may be possible for each refiner and importer to track the use 
of its diesel fuel to determine what percentage was used by railroads 
and marine vessels. This information could then be used in lieu of the 
PADD average values shown in Table IV-2. However, this approach would 
have to be taken by every refinery and importer to avoid double 
counting. Any new refineries or importers would still be assigned a 
locomotive and marine baseline from Table IV-2. Tracking diesel fuel 
use in this instance could be feasible, since the number of railroads 
and marine terminals is relatively small. We request comment on this 
alternative approach and details of how such an approach could be 
implemented.
c. Designate and Track Volumes in 2007
    One main benefit of the proposed non-highway baseline approach is 
to allow 500 ppm highway and 500 ppm NRLM diesel fuel to be fungibly 
distributed while still ensuring achievement of the benefits of the 
highway program. In developing the proposal, several refiners 
recommended another possible approach, referred to here as the 
``designate and track'' approach. It was suggested as a replacement for 
the proposed non-highway baseline approach. After further discussion, a 
modified designate and track approach was also described as an 
alternative for refiners to choose from, in addition to the baseline 
and dye alternatives. We discuss both of these designate and track 
approaches below.
    We invite comment on these designate and track approaches. However, 
we are not proposing them for a number of reasons as discussed in more 
detail below. We are concerned that such an approach could reduce the 
volume of 15 ppm fuel required to be produced under the highway 
program, eroding environmental benefits and calling into question 
availability of 15 ppm highway fuel. This concern is compounded by 
serious concerns with respect to the workability and enforceability of 
such a program, particularly if it is a replacement for the baseline 
approach. We are also concerned that such an approach would place too 
much burden on the many entities, including many small entities, in the 
distribution system. Unlike the situation with the existing highway 
diesel program, the downstream parties, not the refiners, would be 
liable if insufficient 15 ppm highway diesel fuel was produced and 
distributed. Finally, these concerns would appear to be reduced if the 
designate and track approach were to be allowed as a choice for 
refiners. However, it may then be of such limited usefulness that it is 
of little value and only adds program complexity. We are interested in 
comments describing how these concerns could be addressed in order to 
implement such an approach.

[[Page 28413]]

i. Designate and Track as a Replacement for the Non-Highway Baseline 
Approach
    Under the designate and track approach, a refiner or importer would 
designate its 500 ppm diesel fuel as highway diesel fuel or NRLM diesel 
fuel and this refiner designation would be used to differentiate 
highway fuel and NRLM fuel instead of the non-highway baseline. For 
example, the highway 80/20 requirement would only apply to the amount 
of diesel fuel designated by the refinery or importer as highway diesel 
fuel. A marker would still be used to segregate heating oil, but the 
dye requirement for NRLM at the refinery gate would be removed. As with 
the baseline approach, undyed 500 ppm highway and 500 ppm NRLM could be 
fungibly distributed up until the point the NRLM diesel fuel is dyed. 
These refiner designations would have to follow the fuels through the 
distribution system. Under this designate and track approach, fuel 
distributors would be required to ensure that they did not sell more 
diesel fuel to the highway market than they took in as highway fuel. 
For example, if 60% of the fuel they took in was originally designated 
by the refineries as NRLM, they could not sell more than 40% to the 
highway market. The refiner or importer would have no obligation to 
ensure this occurred and no liability if it did not occur.
    This approach shifts the focus from monitoring and enforcement of 
production at the refinery gate to monitoring and enforcement of the 
volumes of fuel handled by each party in the distribution system. Under 
the designation and track approach, refiners and importers would have 
complete flexibility to designate individual batches of diesel fuel or 
even portions of batches as either highway fuel or NRLM fuel. A 
pipeline could mix undyed highway 500 ppm and NRLM diesel fuels and 
ship them fungibly as a single physical batch as in the non-highway 
baseline approach. However, two sets of records would be kept, one 
applicable to the highway fuel portion and one applicable to the NRLM 
fuel portion. Whenever all or a portion of the fungible batch was split 
off or sold, that portion would have to carry one of the two 
designations, highway or NRLM. The sum of the volumes designated as 
either fuel would always be required to add up to the volumes 
designated in the original batch. A combination of fungibly mixed 
batches would be handled similarly, with the total volumes of each 
designation of volume split off or sold equaling the sum of the volumes 
of each designation of the original batches, respectively.
    Each party in the distribution system beyond the refinery gate 
would be required to reconcile the volumes taken in and the volumes 
discharged, based on the designations of the diesel fuel, annually. For 
example, assume that over a year a pipeline received a total of 
100,000,000 gallons of undyed 500 ppm diesel fuel from various 
refineries, with 70% of what it received being designated by the 
refiners as highway and 30% designated as NRLM. Over the year the 
pipeline would also designate what it discharged at various terminals 
or other points as either highway or NRLM. The pipeline would have to 
ensure that over a year's time it did not discharge more than 70% of 
the volume of this entire pool of 500 ppm diesel fuel as highway diesel 
fuel, to ensure that fuel designated as NRLM was not inappropriately 
converted to highway use. It could not discharge more 500 ppm fuel as 
highway than it took in as highway, and it would have to discharge at 
least as much 500 ppm diesel fuel designated as NRLM as it took in. 
This same reconciliation process would apply to every party in the 
distribution system.
    A primary advantage of this designate and track approach for 
refiners is that it would allow them complete flexibility in deciding 
how much 15 ppm highway diesel fuel to produce, allowing them to react 
to changing market conditions. As long as 80 percent of whatever volume 
they designated as highway was 15 ppm, they would be in compliance. 
However, in order to maintain the integrity of the highway program, EPA 
would have to ensure that all diesel fuel designated as NRLM eventually 
was dyed and sold to the NRLM market. Otherwise, for example, refiners 
and importers could simply designate diesel fuel under the more lenient 
NRLM diesel fuel program while downstream in the distribution system 
the fuel was shifted to the highway diesel fuel market. Such shifting 
would compromise the required 80/20 split between 15 ppm and 500 ppm 
highway diesel fuel and undermine the benefits and integrity of the 
highway program. Various refiners proposed that EPA compare the volume 
of all diesel fuel designated as NRLM by the refineries and importers 
nationwide and compare that with the volume dyed nationwide to 
determine whether the approach was working. Unfortunately, this 
approach is not feasible, since EPA could not determine and take 
corrective action against refiners, importers, or distributors if the 
designated and dyed volumes did not reconcile. To locate the cause of a 
discrepancy between the designated and dyed volumes, EPA would have to 
audit the records of every party in the distribution system nationwide. 
The refiners and importers would not face any liability under this 
approach for any downstream discrepancy unless there was evidence of 
collusion with downstream entities.
    Thus, under this designate and track approach, EPA would need to 
require that all parties handling undyed diesel fuel designated as NRLM 
maintain records for each batch of fuel shipped and received and submit 
reports periodically demonstrating that the volume of undyed NRLM 
designated fuel that they dyed plus that transferred undyed to another 
fuel distributor equaled or exceeded the volume of undyed NRLM 
designated fuel that they received.\248\ We would also need to require 
that all parties handling dyed or undyed NRLM diesel fuel maintain 
records and submit reports demonstrating that the volume of NRLM 
designated fuel that they received was sold for use in nonroad, 
locomotive or marine diesel engines or transferred with the same 
designation to another fuel distributor. These requirements would be 
applied on an annual basis, providing fuel distributors with 
flexibility to shift fuel designated for one use to the other market 
and vice versa to address short term supply fluctuations of each fuel 
but still maintain overall program integrity.
---------------------------------------------------------------------------

    \248\ If the volume of dyed NRLM fuel exceeded the designated 
volume, this would imply that some highway 500 ppm fuel was dyed. 
This would not compromise the required 80/20 split between 15 ppm 
and 500 ppm fuel under the highway program, although the total 
social cost of producing the fuel would be higher.
---------------------------------------------------------------------------

    Given the large number of entities involved in distributing diesel 
fuel and the number of transactions, there are a number of serious 
practical concerns regarding the enforceability of such an approach. 
Under the baseline approach described above, enforcement is focused on 
the roughly 128 refineries producing either highway or NRLM diesel 
fuel. This designation and track approach would add the various 
entities in the distribution system. In order to improve the chances of 
effectively enforcing the program, we would at a minimum have to limit 
the scope of the entities involved to bulk terminals and entities 
upstream. Thus, all NRLM diesel fuel would have to exhibit visible 
evidence of dye after leaving a large bulk terminal. Even with this 
limitation, there would be as many as 100 pipelines and 1000 terminals 
reporting. Enforcement of such an approach would be difficult because 
to determine whether inappropriate changes in

[[Page 28414]]

designation occurred by a given entity, the records of each entity from 
which it received fuel and to which it sent fuel over the course of an 
entire year would also have to be compared. An electronic reporting 
mechanism would likely have to be set up to facilitate reporting and to 
track the volumes of fuel received and shipped out by each entity in 
the distribution system down to the terminal. If any entity in the 
distribution system were unable to verify through their records that 
they distributed the same amount or more of NRLM fuel as they took in 
with this designation, then they, not the refiners, would be presumed 
liable for violating the provisions of the highway rule. Therefore, in 
addition to our concerns of ensuring compliance, we invite comment on 
the appropriateness of shifting the compliance burden for tracking fuel 
volumes, maintaining records, reporting to the Agency, and responding 
to enforcement audits from the refiners to the downstream parties, 
particularly since many of these entities are small businesses.
    In addition to the number of entities involved and transactions 
needing to be tracked, there are a number of complications which would 
make such an approach difficult to implement. First, due to 
contamination in the distribution system that results in some product 
being downgraded from one grade to another in the distribution system, 
in actuality the volumes of fuel designated at the refinery and those 
downstream will likely never match. Some means of addressing this 
situation would have to be developed which did not allow fuel produced 
as NRLM fuel to be subsequently sold as highway fuel. Second, kerosene 
will be blended into NRLM diesel fuel in northern areas during the 
winter months. It is difficult to understand how refiners would be able 
to designate portions of this fuel as NRLM fuel or highway fuel at the 
refinery gate given its many other uses. Therefore, this would further 
disrupt the volume reconciliation. Third, it would not always be 
entirely clear who should be the entity responsible for compliance, 
recordkeeping, and reporting. In many cases in the distribution system 
there are entities who have custody of the fuel while a variety of 
other entities maintain ownership. A means of sorting out who the 
responsible party was under such circumstances would have to be 
determined.
    One of the advantages of the proposed baseline approach is that 
once 500 ppm fuel leaves the refinery gate, the distribution system has 
complete flexibility to shift it to either the highway or the NRLM 
markets to respond to changing market conditions. Conversely, as 
discussed above, one of the main advantages of the designate and track 
approach is that it allows refiners complete flexibility to modify 
their relative production of 15 ppm and 500 ppm fuel by their choice of 
designations (highway or NRLM). However, the market will demand a 
certain volume of highway fuel and NRLM fuel, and these decisions will 
be made downstream. If the market demands more highway diesel fuel than 
what the refiners designated as highway on an annual basis, then under 
the designate and track approach the terminals will be restricted from 
responding to this market change. They could shift NRLM fuel into the 
highway market on a temporary basis, but by the end of the year, they 
would have to be able to reconcile their highway and NRLM volumes. 
Given the refiner's inability to predict future demand precisely, and 
their economic incentive to produce as little 15 ppm diesel fuel as 
possible, there is a real possibility that some terminals could find 
themselves in a noncomplying situation. Were this to occur, a terminal 
would be faced with two difficult choices. They could stop shipping 
highway diesel fuel, in which case they would not only fail to deliver 
on their contracts to their customers, but would also constrain highway 
diesel fuel supply, raising market prices. Or, they could continue to 
respond to market pressure and sell additional volumes of fuel 
designated as NRLM into the highway market. In this case, they would 
risk significant non-compliance penalties from EPA, were we able to 
detect the violation. Thus, we are concerned that the designate and 
track approach could result in either widespread noncompliance or 
disruption of the fuel distribution system.
    We are also concerned that the designate and track approach would 
not maintain the benefits and integrity of the highway program. Nearly 
a third of all non-highway distillate today is produced to the highway 
specifications due primarily to limitations in the distribution system. 
The sulfate PM and SO2 emission benefits predicted from the 
highway rule, and the assumptions with respect to program cost and fuel 
availability, were all based on the assumption that 80% of this 
spillover volume would comply with the 15 ppm highway standard and 
would be available for highway use if needed. Under the proposed dye 
approach, in the future this ``spillover'' from the highway market 
could technically be dyed at the refinery gate to avoid compliance with 
the 2006 highway standards. However, our expectation is that the 
majority of the spillover today would continue into the future as it 
would be costly to significantly change the current distribution 
practices. While the dye approach would not ensure this and spillover 
could decline, it would be unlikely to drop significantly. Similarly, 
the proposed baseline approach would maintain spillover at historical 
rates (either 2003-5 the average level or June 1, 2006--May 31, 2007, 
level). However, under the designate and track approach, wherever 
undyed 500 ppm was distributed as a grade of fuel, the prior spillover 
volume could instead be designated as NRLM fuel, and would no longer be 
subject to the highway program standards (i.e., 80 percent of it would 
no longer have to meet the 15 ppm sulfur standard.). The segregation 
and associated cost that previously led to spillover would be gone. As 
a result, the benefits projected from this fuel volume under the 
highway rule would be reduced. Furthermore, with the reduced volume of 
15 ppm fuel produced, we would need to reevaluate whether sufficient 15 
ppm fuel would still be available in all parts of the country for the 
vehicles that would need it. The areas where availability of 15 ppm 
fuel would be of greatest concern would be those areas where 500 ppm 
fuel would be distributed and spillover would decline under the 
designate and track approach. The enforcement concerns cited in the 
paragraphs above only serve to heighten this concern.
    EPA requests comments on the practical viability of this approach. 
In addition to the issues noted above, we specifically request comments 
on the following:
    (1) What would be the impacts of this approach on fuel 
distributors?
    (2) What information would need to be kept and/or reported?
    (3) How might the required reports be automated in a common, 
electronic format?
    (4) How often should reports be required (e.g., annually, 
quarterly, each batch if electronically)?
    (5) How might`the record keeping requirements be combined with 
those already required by the U.S. Internal Revenue Service?
    (6) How would the record keeping requirements work for pipelines 
and certain terminals that handle fuel without taking ownership and 
that do not control the decision to dye certain diesel fuel prior to 
sale?
    (7) How might the IRS records for refiners, importers and 
distributors be used as an independent check on the

[[Page 28415]]

volumes of undyed diesel fuel handled which are eventually dyed and 
which are sold undyed?
    (8) What would be the cost associated with the tracking, record 
keeping and reporting?
    (9) Could the industry utilize independent auditors to simplify 
EPA's enforcement oversight?
    (10) Could refiners feasibly be responsible to ensure the necessary 
volumes are dyed downstream at the terminal rather than placing the 
responsibility and liability with the fuel distributors?
    (11) What changes could be made to the program to avoid losing the 
benefits of the highway program (e.g., avoid loss in production of 15 
ppm attributable to the spillover volume)?
ii. Designate and Track as a Refiner's Option in Addition to the 
Baseline Approach
    Several refiners indicated that the designate and track approach 
should be considered as an option in addition to the baseline approach. 
Including the designate and track approach as a refiner's option, 
however, would significantly alter the design and implications of the 
approach.
    With such an approach, no longer could compliance be determined 
simply on the basis of whether a terminal dyed at least as much volume 
of diesel fuel as the volume received designated as NRLM 500 ppm fuel, 
since the dyed diesel fuel could have been produced under either the 
non-highway baseline approach or the designate and track approach. In a 
situation where volumes produced under the designate and track approach 
are fungibly distributed with volumes produced under the baseline 
approach, there is no clear way to identify whether dyed volumes have 
been accurately reconciled under the designate and track approach, 
risking significant loss in the benefits expected from the highway 
program.
    For example, assume a terminal receives a certain volume of undyed 
diesel fuel and 30% of it was originally designated by the refinery as 
NRLM under the designate and track approach. The remaining 70% would 
have been produced by refineries using the non-highway baseline 
approach. Some significant portion of the 70% produced by refineries 
under the baseline approach would have been produced subject to the 500 
ppm standard for the NRLM market, not the standards for highway market, 
and produced with the expectation that it could later be dyed at the 
terminal. If the terminal dyes only 30% of the entire volume it 
receives, there is every expectation that some or even all of that 30% 
could have been produced by refineries using the baseline approach, and 
should not be counted towards the volume reconciliation under the 
designate and track approach. If all of the 30% of dyed diesel fuel was 
produced by refineries using the baseline approach, then the terminal 
would have effectively sold into the highway market all of the fuel 
received as NRLM under the designate and track approach.
    Thus, in order to allow for volumes to be reconciled using such an 
approach, we concluded that fuel distributors would have to track which 
refinery or importer the fuel came from and how they disposed of the 
fuel for that refinery or importer, in addition to whether it was NRLM 
or highway. Thus, allowing the designate and track approach as a 
refiner's option would add one more layer of complexity to the 
tracking, recordkeeping, and reporting.
    The following example explains how the approach could work in 
theory. Over the course of a year, a terminal receives 6 million 
gallons of 500 ppm diesel fuel identified as baseline fuel from 
refinery A, 2 million gallons of 500 ppm diesel fuel designated as 
``designate and track'' NRLM fuel from refinery B, and 2 million 
gallons of 500 ppm diesel fuel designated as ``designate and track'' 
highway fuel from refinery B. At the end of the year, the terminal 
would have had to have dyed at least 2 million gallons of the fuel it 
received from refinery B and delivered it to or on behalf of that 
refinery as dyed NRLM. (If they do not deliver the fuel back to the 
entity that designated the fuel, then the dyed fuel could have been 
baseline fuel from refinery A, and we could not enforce the dyeing of 
the designate and track fuel volume from refinery B.) The terminal 
would need to do this separately for each refinery or importer from 
which it received designate and track diesel fuel.
    Based on the above discussion, we believe that in order to have an 
enforceable program, only those refineries and importers who maintain 
ownership of the fuel all the way through the pipeline and terminal 
could take advantage of the option to designate and track their fuel. 
This could be a very small subset of refiners since they would have to 
maintain ownership of all of their NRLM diesel fuel distributed through 
all of its distribution pathways to the point where the fuel is dyed. 
If this were a very small subset, then it would raise questions as to 
whether the flexibility of this approach would be worth the added 
program and enforcement complexity.
    Since the pipelines and terminals in this situation are basically 
providing a service to these refineries and importers, transporting 
their fuel and dyeing it for them, a different responsibility and 
liability scheme could be considered. Instead of the fuel distributors 
being solely responsible for recordkeeping and reporting to the Agency 
and liable for any violations, it might be possible to leave this 
burden with the refiner. The refiner could be responsible for ensuring 
that they took delivery from a terminal of at least as much dyed NRLM 
diesel fuel as they sent undyed NRLM to that terminal from their 
refinery gate. The refiner would be responsible for collecting and 
maintaining the records from the various points in the distribution 
system to demonstrate compliance and to submit an annual report 
demonstrating compliance. At the same time EPA would have to be able to 
verify the refiner's report and as a result, fuel distributors may 
still have to maintain records.
    For the baseline approach to exist simultaneously with the 
designate and track approach, a refinery or importer would have to 
choose which approach to utilize and maintain that approach. We could 
consider allowing the refinery to change approaches on a year to year 
basis, as with the baseline and dye alternatives.
    EPA requests comment on the designate and track approach as a 
refinery's option and whether it could be enforced as described above. 
EPA specifically requests comment on:
    (1) The advantages and disadvantages of placing the recordkeeping, 
reporting, and liability burden on the refinery of the designate and 
track approach if it is an option along with baseline approach;
    (2) If this responsibility were not place on the refiners, what 
level of voluntary participation would occur among fuel distributors 
(e.g., pipelines and terminals) and how might EPA structure a viable 
enforcement oversight program;
    (3) What level of voluntary refinery participation would occur and 
whether it warrants the added program complexity;
    (4) The extent to which this approach might reduce 15 ppm highway 
diesel production (i.e., reduced spillover to non-highway markets)
    (5) What would be the cost associated with the tracking, record 
keeping and reporting?

[[Page 28416]]

C. Hardship Provisions for Qualifying Refiners

1. Hardship Provisions for Qualifying Small Refiners
    In developing our proposed off-highway diesel sulfur program, we 
evaluated the need and the ability of refiners to meet the 500 and 15 
ppm standards as expeditiously as possible. We believe it is feasible 
and necessary for the vast majority of the program to be implemented in 
the proposed time frame to achieve the air quality benefits as soon as 
possible. Based on information available from small refiners and 
others, we believe that refineries owned by small businesses generally 
face unique hardship circumstances, compared to larger refiners. Thus, 
as discussed below, we are proposing several special provisions for 
refiners that qualify as ``small refiners'' to reduce the 
disproportionate burden that nonroad diesel sulfur requirements would 
have on these refiners.\249\
---------------------------------------------------------------------------

    \249\ The proposed small refiner provisions would not apply to 
importers, as the burden from capital expenditures for physical 
refinery improvements are not imposed on importers.
---------------------------------------------------------------------------

a. Qualifying Small Refiners
    EPA is proposing several special provisions that would be available 
to companies approved as small refiners. The primary reason for these 
provisions is that small refiners generally lack the resources 
available to large companies that help large companies, including those 
large companies that own small-capacity refineries, to raise capital 
for investing in desulfurization equipment, such as shifting of 
internal funds, securing of financing, or selling of assets. Small 
refiners are also likely to have more difficulty in competing for 
engineering resources and completing construction of the needed 
desulfurization equipment in time to meet the standards proposed today.
    Since small refiners are more likely to face hardship circumstances 
than larger refiners, we are proposing temporary provisions that would 
provide additional time to meet the sulfur standards for refineries 
owned by small businesses. This approach would allow the overall 
program to begin as early as possible, avoiding the need for delay in 
order to address the ability of small refiners to comply.
i. Regulatory Flexibility for Small Refiners
    As explained in the discussion of our compliance with the 
Regulatory Flexibility Act in section X.C and in the Initial Regulatory 
Flexibility Analysis in chapter 11 of the Draft RIA, we considered the 
impacts of the proposed regulations on small businesses. Most of our 
analysis of small business impacts was performed as a part of the work 
of the Small Business Advocacy Review (SBAR) Panel convened by EPA, 
pursuant to the Regulatory Flexibility Act as amended by the Small 
Business Regulatory Enforcement Fairness Act of 1996 (SBREFA). The 
final report of the Panel is available in the docket for this proposed 
rule.
    For the SBREFA process, EPA conducted outreach, fact-finding, and 
analysis of the potential impacts of our regulations on small 
businesses. Based on these discussions and analyses by all panel 
members, the Panel concluded that small refiners in general would 
likely experience a significant and disproportionate financial hardship 
in reaching the objectives of the proposed nonroad diesel fuel sulfur 
program.
    One indication of this disproportionate hardship for small refiners 
is the relatively high cost per gallon projected for producing nonroad 
diesel fuel under the proposed program. Refinery modeling of refineries 
owned by refiners likely to qualify as small refiners, and of non-small 
refineries, indicates significantly higher refining costs for small 
refiners. Specifically, we project that without special provisions, 
refining costs for small refiners on average would be about 5.5 cents 
per gallon compared to about 4.0 cents per gallon for non-small 
refiners.
    The Panel also noted that the burden imposed on the small refiners 
by the proposed sulfur standards may vary from refiner to refiner. 
Thus, the Panel recommended more than one type of burden reduction 
measure so that most if not all small refiners could benefit. We have 
continued to consider the issues raised during the SBREFA process and 
have decided to propose each of the provisions recommended by the 
Panel.
ii. Rationale for Small Refiner Provisions
    Generally, we structured these proposed provisions to reduce the 
burden on small refiners while expeditiously achieving air quality 
benefits and ensuring that the availability of 15 ppm nonroad diesel 
fuel would coincide with the introduction of 2011 model year nonroad 
diesel engines and equipment. We believe the proposed special 
provisions for small refiners are necessary and appropriate.
    First, the proposed compliance schedule for the nonroad diesel 
program, combined with flexibility for small refiners, would achieve 
the air quality benefits of the program as soon as possible, while 
helping to ensure that small refiners will have adequate time to raise 
capital for new or upgraded fuel desulfurization equipment. Most small 
refiners have limited additional sources of income beyond refinery 
earnings for financing and typically do not have the financial backing 
that larger and generally more integrated companies have. Therefore, 
they can benefit from additional time to accumulate capital internally 
or to secure capital financing from lenders.
    Second, we recognize that while the sulfur levels in this proposed 
program can be achieved using conventional refining technologies, new 
technologies are also being developed that may reduce the capital and/
or operational costs of sulfur removal. Thus, we believe that allowing 
small refiners some additional time for newer technologies to be proven 
out by other refiners would have the added benefit of reducing the 
risks faced by small refiners. The added time would likely allow for 
small refiners to benefit from the lower costs of these improvements in 
desulfurization technology (e.g., better catalyst technology or lower-
pressure hydrotreater technology). This would help to offset the 
financial burden facing small refiners.
    Third, providing small refiners more time to comply would increase 
the availability of engineering and construction resources. Most 
refiners would need to install additional processing equipment to meet 
the nonroad diesel sulfur requirements. We anticipate that there may be 
significant competition for technology services, engineering resources, 
and construction management and labor. In addition, vendors will be 
more likely to contract their services with the larger refiners first, 
as their projects will offer larger profits for the vendors. 
Temporarily delaying compliance for small refiners would spread out the 
demand for these resources and probably reduce any cost premiums caused 
by limited supply.
    We discuss below the provisions we are proposing to minimize the 
degree of hardship for small refiners. With these provisions we are 
confident about going forward with the 500 ppm sulfur standard for NRLM 
diesel fuel in 2007 and the 15 ppm sulfur standard for nonroad diesel 
fuel in 2010 for the rest of the industry. Without small refiner 
flexibility, EPA would have to consider delaying the overall program 
until the burden of the program on many small refiners were diminished, 
which would delay the air quality benefits of the overall program. By 
providing

[[Page 28417]]

temporary relief to small refiners, we are able to adopt a program that 
expeditiously reduces off-highway diesel sulfur levels in a feasible 
manner for the industry as a whole.
iii. Limited Impact of Small Refiner Options on Program Emissions 
Benefits
    Small refiners that choose to make use of the delayed nonroad 
diesel sulfur requirements would also delay to some extent the emission 
reductions that would otherwise have been achieved. However, the 
overall impact of these postponed emission reductions would be small, 
for several reasons.
    First, small refiners represent only a fraction of national non-
highway diesel production. Today, refiners that we expect would qualify 
as small refiners represent only about 6 percent of all high-sulfur 
diesel production. Second, the proposed delayed compliance provisions 
described below would affect only engines without new emission 
controls. During the first step to 500 ppm NRLM fuel, small refiner 
nonroad fuel could be well above 500 ppm, but the new advanced engine 
controls would not yet be required. During the second step to 15 ppm 
nonroad diesel fuel, equipment with the new controls would be entering 
the market, but use of the 500 ppm small refiner fuel would be 
restricted to older engines without the new controls. There would be 
some loss of sulfate PM control in the older engines that operated on 
higher sulfur small refiner fuel, but no effect on the major emission 
reductions that the proposed new engine standards would achieve 
starting in 2011. Finally, because small diesel refiners are generally 
dispersed geographically across the country, the limited loss of 
sulfate PM control would also be dispersed.
    One proposed small refiner option would allow a modest 20% 
relaxation in the gasoline sulfur interim standards for small refiners 
that produce all nonroad diesel fuel at 15 ppm by June 1, 2006. To the 
extent that small refiners elected this option, a small loss of 
emission control from Tier 2 gasoline vehicles that used the higher 
sulfur gasoline could occur. We believe that such a loss of control 
would be very small. A very few small refiners would be in a position 
to use this provision. Further, the relatively small production of 
gasoline with slightly higher sulfur levels should have no measurable 
impact on the emission of new Tier 2 vehicles, even if the likely 
``blending down'' of sulfur levels did not occur as this fuel mixed 
with lower sulfur fuel during distribution. This provision would also 
maintain the maximum 450 ppm gasoline sulfur per-gallon cap standard in 
all cases, providing a reasonable sulfur ceiling for any small refiners 
making use of this provision.
b. How Do We Define Small Refiners for Purposes of the Hardship 
Provisions?
    The definition of small refiner for the proposed nonroad diesel 
program is basically the same as our small refiner definitions in the 
Tier 2/Gasoline Sulfur and Highway Diesel rules. A small refiner must 
demonstrate that it meets both of the following criteria:
    [sbull] No more than 1,500 employees corporate-wide, based on the 
average number of employees for all pay periods from January 1, 2002 to 
January 1, 2003.
    [sbull] A corporate crude oil capacity less than or equal to 
155,000 barrels per calendar day (bpcd) for 2002.
    As with the earlier fuel sulfur programs, the dates for the 
employee count and for calculation of the crude capacity represent the 
latest complete years prior to the issuing of the proposed rule.
    In determining the total number of employees and crude oil 
capacity, a refiner must include the number of employees and crude oil 
capacity of any subsidiary companies, any parent company and 
subsidiaries of the parent company, and any joint venture partners. We 
define a subsidiary of a company to mean any subsidiary in which the 
company has a 50 percent or greater ownership interest. However, we are 
proposing that a refiner be eligible for small refiner status if it is 
owned and controlled by an Alaska Regional or Village Corporation 
organized under the Alaska Native Claims Settlement Act (43 U.S.C. 
1626), regardless of number of employees and crude oil capacity. Such 
an exclusion would be consistent with our desire to grant relief from 
the regulatory burden to that part of the industry that can least 
afford compliance. We believe that very few refiners, probably only 
one, would qualify under this provision. Similarly, we are proposing to 
incorporate this exclusion into the small refiner provisions of the 
highway diesel and gasoline sulfur rules, which did not address this 
issue.
    As with the earlier fuel sulfur rules, we are proposing that a 
refiner that restarts a refinery in the future may be eligible for 
small refiner status. Thus, a refiner restarting a refinery that was 
shut down or non-operational between January 1, 2002, and January 1, 
2003, could apply for small refiner status. In such cases, we would 
judge eligibility under the employment and crude oil capacity criteria 
based on the most recent 12 consecutive months unless we conclude from 
data provided by the refiner that another period of time is more 
appropriate. Companies with refineries built after January 1, 2002, 
would not eligible for the small refiner hardship provisions.
2. The Effect of Financial Transactions on Small Refiner Status and 
Small Refiner Relief Provisions
    During the implementation of the gasoline sulfur and highway diesel 
sulfur programs, several refiners have raised concerns about how 
various kinds of financial transactions would affect implementation of 
the small refiner fuel sulfur provisions. The kind of transactions 
typically involve refiners with approved small refiner status that are 
involved in potential or actual sales of the small refiner's refinery, 
or involve the purchase by the small refiner of another refinery or 
other non-refining asset. We believe that these concerns are also 
relevant to the small refiner provisions proposed below for the nonroad 
diesel sulfur program.
a. Large Refiner Purchasing a Small Refiner's Refinery
    One situation involves a ``non-small'' refiner that wishes to 
purchase a refinery owned by an approved small refiner. The small 
refiner may not have completed or even begun refinery upgrades to meet 
the long-term fuel sulfur standards, since it is making use of the 
special small refiner relief provisions. This situation is of most 
concern where the purchase is to take place near or after the beginning 
of the gasoline or highway diesel sulfur programs. Under the existing 
gasoline sulfur and highway diesel sulfur programs, once such a 
purchase is completed, the ``non-small'' purchaser would not have the 
benefit of the small refiner relief provisions that had applied to the 
previous owner.
    The purchasing refiner would have to perform the necessary upgrades 
to meet the ``non-small'' sulfur standards. As the gasoline sulfur and 
highway diesel sulfur provisions exist today, such a refiner would be 
left with very little or (if the respective fuel sulfur control program 
has already begun) no lead time for compliance. The refiners that have 
raised this issue have claimed that refiners in this situation would 
not be able to comply with the ``non-small refiner'' standards upon 
acquisition of the new refinery. These refiners claim that this could 
prevent them from purchasing a refinery from a small refiner and, as a 
result, this would severely limit the ability of small refiners to sell 
such an asset. The refiners that have raised this issue have

[[Page 28418]]

said that some sort of ``grace period'' of additional lead time before 
the non-small refiner sulfur standards take effect would address this 
issue.
    We believe these concerns are valid and are proposing that an 
appropriate period of lead time for compliance with the nonroad diesel 
sulfur requirements be provided where a refiner purchases any refinery 
owned by a small refiner, whether by purchase of the refinery or 
purchase of the small refiner entity. We propose that a refiner that 
acquires a refinery from an approved small refiner be provided 24 
additional months from the date of the completion of the purchase 
transaction (or until the end of the applicable small refiner relief 
interim period if it is within 24 months--June 1, 2010, for 500 ppm 
fuel and June 1, 2014, for 15 ppm fuel). During this interim period, 
production at the newly-acquired refinery could remain at the interim 
sulfur levels that applied to that refinery for the previous small 
refiner owner under the small refiner options discussed below. At the 
end of this period, the refiner would need to comply with the ``non-
small refinery'' sulfur standards.
    We expect that in most if not all cases, the proposed 24 months of 
additional lead time would be sufficient for the new refiner-owner to 
accomplish the necessary engineering, permitting, construction, and 
start-up of the necessary desulfurization project, since planning for 
this could be expected to be a part of any purchase decision. If a 
refiner nonetheless believed that the technical characteristics of its 
planned desulfurization project would require additional lead time, the 
refiner could apply for additional time and EPA would consider such 
requests on a case-by-case basis. Such an application would be based on 
the technical factors supporting the need for more time and include 
detailed technical information and projected schedules for engineering, 
permitting, construction, and startup. Based on information provided in 
such an application and other relevant information, EPA would decide 
whether additional time was technically necessary and, if so, how much 
additional time would be appropriate. As discussed above, in no case 
would compliance dates be extended beyond the time frame of the 
applicable small refiner relief provisions (June 1, 2010, for 500 ppm 
fuel and June 1, 2014, for 15 ppm fuel).\250\
---------------------------------------------------------------------------

    \250\ This process would be similar to the general hardship 
provisions of the existing gasoline sulfur and highway diesel sulfur 
programs and proposed today for nonroad diesel fuel. However, the 
focus here would be simply on the lead time needed for the technical 
upgrades and would not consider any claimed financial hardship.
---------------------------------------------------------------------------

    During the 24 months additional lead time (and any further lead 
time approved by EPA for the purchasing refiner), all existing small 
refiner provisions and restrictions, as described below, would also 
remain in place for that refinery. This would include the per-refinery 
volume limitation on the amount of nonroad diesel that could be 
produced at the small refiner standards. There would be no adverse 
environmental impact of this provision, since the small refiner would 
already have been provided relief prior to the purchase and this 
provision would be no more generous.
b. Small Refiner Losing Its Small Refiner Status
    A second situation involves a refiner with approved small refiner 
status that later loses its small refiner status because it exceeds the 
small refiner criteria. In the existing gasoline sulfur and highway 
diesel sulfur programs, an approved small refiner that exceeds 1,500 
employees due to merger or acquisition would lose its small refiner 
status. (We also intended for refiners that exceeded the 155,000 barrel 
per calendar day crude capacity limit due to merger or acquisition to 
lose its small refiner status and we are proposing below to amend the 
regulations to reflect that criterion as well.) This includes 
exceedences of the criteria caused by acquisitions of assets such as 
plant and equipment, as well as acquisitions of business entities.
    Our intent in the gasoline and highway diesel sulfur programs, as 
well as the proposed nonroad diesel sulfur program, has been and 
continues to be to reserve the small refiner relief provisions for a 
small subset of refiners that generally tend to face the kinds of 
special challenges discussed above. At the same time, it is also our 
intent to avoid stifling normal business growth among small refiners. 
Therefore, we designed our existing regulations, as well as the 
proposed regulations, to disqualify a refiner from small refiner status 
when it exceeds the small refiner criteria through its involvement in 
transactions such as being acquired by or merging with another entity 
or through the small refiner itself purchasing another entity or assets 
from another entity. However, as in the existing regulations, we are 
proposing that if an approved small refiner were to exceed the criteria 
without merger or acquisition, it would keep its small refiner status.
    Consistent with our intent in the earlier fuel sulfur programs to 
limit the use of the small refiner hardship provisions, we also 
intended in the gasoline sulfur and highway diesel sulfur programs for 
an exceedence of the other small refiner criterion--a limit of 155,000 
barrels per calendar day of crude capacity--due to merger or 
acquisition to be grounds for disqualifying a refiner's small refiner 
status. However, we inadvertently failed to include this second 
criterion as grounds for disqualification. In today's action, we 
propose to resolve this error by adding the crude capacity limit to the 
employee limit in this context for both the gasoline sulfur and highway 
diesel sulfur programs, to begin January 1, 2004. Thus, a refiner 
exceeding either criterion due to merger or acquisition would lose its 
small refiner status.
    We recognize that a small refiner that loses its small refiner 
status because of a merger or acquisition would face the same type of 
lead time concerns in complying with the non-small refiner standards as 
would a non-small refiner that acquired a small refiner's refinery, as 
discussed above. Therefore, we propose that the additional lead time 
proposed above for non-small refiners purchasing a small refiner's 
refinery also apply to this situation. Thus, this additional lead time 
would apply to any refineries, existing or newly-purchased, that had 
previously been subject to the small refiner program, but would not 
apply to a newly-purchased refinery that is subject to the non-small 
refiner standards. Again, there would be no adverse environmental 
impact because of the newly-purchased small refiner's pre-existing 
relief provisions.
    The issues discussed in this subsection apply equally to the 
gasoline sulfur and highway diesel sulfur programs. Thus, we are also 
proposing that the same provisions relating to additional lead time in 
cases of financial transaction be applied to the small refiner programs 
in the earlier fuel sulfur programs. Because these proposed provisions 
for the existing fuel sulfur programs are independent of today's 
nonroad diesel fuel program, we may choose to finalize them separately 
from and earlier than the identical provisions proposed for today's 
nonroad rule. If this occurs, we will seek to finalize nonroad diesel 
fuel provisions that are identical or as similar as appropriate to 
those finalized for the gasoline sulfur and highway diesel program.
    In addition, we are inviting comment on several other related 
provisions we are considering:

[[Page 28419]]

    (1) We propose above that a small refiner that loses its small 
refiner status be granted 24 months of lead time at its existing 
refineries. Should such a small refiner instead be allowed to 
``grandfather in'' its existing small refiner relief program for its 
existing refinery or refineries? An argument can be made that in 
purchasing a new refinery or other assets, the small refiner would no 
longer demonstrate the kind of financial hardship that was the basis 
for general small refiner relief. However, we also do not intend to 
stifle normal growth of small refiners, and ``grandfathering in'' the 
small refiner interim relief program would have no environmental 
impact, since it would merely continue an existing program at that 
refinery.
    (2) If a small refiner exceeds the small refiner criteria due to 
the purchases of a non-small refiner, should the proposed additional 
lead time apply to that refinery? Or should the refiner be required to 
meet the non-small refiner standards on schedule at the ``new'' 
refinery, since the previous owner could be assumed to have anticipated 
the new standards and taken steps to accomplish this prior to the 
purchase?
c. What Options Are Available for Small Refiners?
    We propose several provisions intended to reduce the burdens on 
small refiners discussed above as well as to encourage their early 
compliance whenever possible. As described below, these proposed small 
refiner provisions consist of additional time for compliance and, for 
small refiners that choose to comply earlier than required, the option 
of either generating diesel sulfur credits or receiving a limited 
relaxation of gasoline sulfur requirements.
i. Delays in Nonroad Fuel Sulfur Standards for Small Refiners
    We propose that small refiners be allowed to postpone reducing 
sulfur in nonroad locomotive and marine diesel fuel until June 1, 2010. 
As described earlier, we are proposing that all refiners producing 
nonroad diesel fuel be provided significant lead time for making the 
capital and operational investments to produce 15 ppm fuel, including 
about three years before the 500 ppm requirement would become 
effective, and three additional years before 15 ppm was required--June 
1, 2007, through May 31, 2010, when 500 ppm fuel could be produced. 
While this lead time would be useful for small and non-small refiners 
alike, we believe that in general small refiners would still face 
disproportionate challenges, and the proposed delay in the first step 
of control for small refiners would help mitigate these challenges.
    Then, beginning June 1, 2010, when the second step of the proposed 
base program would require 15 ppm fuel for other refiners for nonroad 
diesel fuel, we propose that small refiners be required to meet a 500 
ppm sulfur standard for NR diesel fuel. We propose that this interim 
standard be effective for four years (until June 1, 2014), after which 
small refiners would meet the 15 ppm sulfur standard for nonroad diesel 
fuel. As for other refiners, the small refiner standard for locomotive 
and marine diesel fuel would remain at 500 ppm. Since new engines with 
sulfur sensitive emission controls would begin to become widespread 
during this time, small refiners would need to segregate the 500 ppm NR 
fuel and supply it only for use in pre-2011 nonroad equipment or in 
locomotives or marine engines. Section VIII below discusses the 
requirements for product transfer documents (PTDs) associated with the 
production of 500 ppm NR fuel by small refiners during this period.
    The following table illustrates the proposed small refiner NRLM and 
NRdiesel standards as compared to the standards proposed in the base 
nonroad diesel program. (For simplicity, the proposed locomotive and 
marine diesel standards for small and non-small refiners described 
above do not appear in the table.)

                                        Table IV-4--Proposed Small Refiner Nonroad Diesel Sulfur Standards, ppm a
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                  2006     2007     2008     2009     2010     2011     2012     2013     2014    2015+
--------------------------------------------------------------------------------------------------------------------------------------------------------
Non-small refiners............................................  .......      500      500      500       15       15       15       15       15       15
Small Refiners................................................  .......  .......  .......  .......      500      500      500      500       15      15
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes:
a New standards would take effect in June of the applicable year.

    We also request comment on a slightly different compliance schedule 
that would require small refiners to produce 15 ppm nonroad diesel fuel 
beginning June 1, 2013, one year earlier than proposed above. Such a 
schedule would align the end of the interim small refiner provisions 
with the end of the proposed phase-in for nonroad engines and equipment 
and eliminate higher sulfur nonroad fuel from the distribution system 
by the time all new nonroad diesel engines required 15 ppm fuel.
    The proposed delayed compliance schedule for small refiners is 
intended to compensate for the relatively higher compliance burdens on 
these refiners. It is not intended as an opportunity for those refiners 
to greatly expand their production of uncontrolled diesel fuel (2007-
2010) or 500 ppm sulfur fuel (2010-2014). To help ensure that any 
significant expansion of refining capacity that a small refiner might 
undertake in the future would be accompanied by an expansion of 
desulfurization capacity, we are proposing that small refiners 
producing higher sulfur fuel limit that production to baseline volume 
levels.
    Specifically, during the first step of the diesel program to 500 
ppm (June 2007-June 2010), a small refiner could produce uncontrolled 
NRLM diesel fuel up to the proposed non-highway baseline for that 
refiner less any marked heating oil it produces, refer to sub-section B 
above for an explanation of this baseline. Any diesel fuel produced 
over its non-highway baseline would be subject to the 500 ppm standard 
applying to other refiners. Similarly, from June 1, 2010, through May 
31, 2014, a small refiner could produce nonroad diesel fuel at 500 ppm 
up to the non-highway baseline less any volume of heating oil and 
marked locomotive and marine diesel fuel it produced. Fuel produced in 
excess of this volume would be subject to the 15 ppm nonroad diesel 
standard.
ii. Options To Encourage Earlier Compliance by Small Refiners
    Some small refiners have indicated that they might find it 
necessary to produce fuel meeting the nonroad diesel sulfur standards 
earlier than required by the small refiner program described above, for 
a variety of reasons. For some small refiners, the distribution systems 
might limit the number of grades of diesel fuel that will be carried. 
Others might find it economically advantageous to make 500 ppm or 15

[[Page 28420]]

ppm fuel earlier so as not to lose market share. At least one small 
refiner has indicated that it might decide to desulfurize its NR pool 
at the same time as it desulfurized its highway diesel fuel, in June of 
2006, due to limitations in its distribution system and to take 
advantage of economies of scale. Given these situations, we propose 
that small refiners be able to choose between two mutually exclusive 
options, as an incentive for early compliance.
    The first proposed option would make the diesel sulfur credit 
banking and trading program discussed earlier in this section fully 
applicable to small refiners. A small refiner could generate diesel 
sulfur credits for production of 500 ppm NRLM diesel fuel prior to June 
1, 2010, and for production of 15 ppm nonroad fuel from June 1, 2010, 
through May 31, 2012. The specifics of the credit program are described 
above in subsection B.2, including how they would be applicable to 
small refiners. Generating and selling credits could provide funds to 
defray the costs of early nonroad compliance.
    The second proposed option would apply to a small refiner that 
produced all of its NRLM diesel production at 15 ppm by June 1, 2006, 
and elected not to use the provision described above to earn NRLM 
sulfur credits for this early compliance. (As for other refiners, 
locomotive and marine fuel sulfur would not be controlled in 2006 and 
could meet the 500 ppm standard beginning June 1, 2007.) Such a refiner 
would receive a modest revision in its interim small refiner gasoline 
sulfur standards, starting January 1, 2004. Specifically, the 
applicable small refiner annual average and per-gallon cap gasoline 
standards would be revised upward by 20 percent for the duration of the 
small refiner gasoline sulfur interim program (i.e., through either 
2007 or 2010, depending on whether the refiner had extended its 
participation in the gasoline sulfur interim program by complying with 
the highway diesel standard at the beginning of that program (June, 
2006, as provided in 40 CFR 80.552(c))). However, in no case could the 
per-gallon cap exceed 450 ppm, the highest level allowed under the 
gasoline sulfur program.
    We believe it is very important to link any such temporary 
relaxation of a small refiner gasoline sulfur interim sulfur standards 
with environmental benefit of early desulfurization of a significant 
volume of NRLM diesel fuel. Thus, we propose that a small refiner 
wishing to use this option must produce a minimum volume of NRLM diesel 
fuel at 15 ppm by June 1, 2006. Each participating small refiner would 
need to produce a volume of 15 ppm fuel that was at least 85% of the 
volume represented by its non-highway distillate baseline percentage. 
If the refiner began to produce gasoline in 2004 at the higher interim 
standard of this provision but then either failed to meet the 15 ppm 
standard for its NRLM fuel by June 1, 2006, or failed to meet the 85% 
minimum volume requirement, the original small refiner interim gasoline 
sulfur standard applicable to that refiner would be reinstated. In 
addition, the refiner would need to compensate for the higher gasoline 
levels that it had enjoyed by purchasing gasoline sulfur credits or 
producing an equivalent volume of gasoline below the required sulfur 
levels.
    Under this option, a small refiner could in effect shift some funds 
from its gasoline sulfur program to accelerate desulfurization of 
nonroad diesel fuel. Given the environmental benefit that would result 
from the production of 15 ppm diesel fuel earlier than necessary, and 
the small potential loss of emission reduction under the gasoline 
sulfur program from fuel produced by the very few small refiners that 
we believe would qualify under this second option, we believe the 
environmental impact of this option would be neutral or positive.
d. How Do Refiners Apply for Small Refiner Status?
    A refiner applying for status as a small refiner would provide EPA 
with several types of information by December 31, 2004. The detailed 
application requirements are summarized in section VII.E.2 below. In 
general, a refiner would need to provide information about the 
following for the parent company and all subsidiaries at all locations: 
(1) The average number of employees for all pay periods from January 1, 
2002, through January 1, 2003; (2) total corporate crude refining 
capacity; and (3) an indication of which small refiner option the 
refiner is likely to use (see subsection c. above). As with 
applications for relief under other rules, applications for small 
refiner status under this proposed diesel rule that were later found to 
contain false or inaccurate information would be void ab initio.
2. General Hardship Provisions
a. Temporary Waivers from Non-highway Diesel Sulfur Requirements in 
Extreme Unforseen Circumstances
    We are proposing a provision which, at our discretion, would permit 
any domestic or foreign refiner to seek a temporary waiver from the 
nonroad, locomotive, or marine diesel sulfur standards under certain 
rare circumstances. This waiver provision is similar to provisions in 
the reformulated gasoline (RFG), low sulfur gasoline, and highway 
diesel sulfur regulations. It is intended to provide refiners short-
term relief in unanticipated circumstances--such as a refinery fire or 
a natural disaster--that cannot be reasonably foreseen now or in the 
near future.
    Under this provision, a refiner may seek permission to distribute 
nonroad, locomotive, or marine diesel fuel that does not meet the 
applicable 500 or 15 ppm sulfur standards for a brief time period. An 
approved waiver of this type could, for example, allow a refiner to 
produce and distribute diesel fuel with higher than allowed sulfur 
levels, so long as the other conditions described below were met. Such 
a request would be based on the refiner's inability to produce 
complying nonroad, locomotive or marine diesel fuel because of extreme 
and unusual circumstances outside the refiner's control that could not 
have been avoided through the exercise of due diligence. The request 
would also need to show that other avenues for mitigating the problem, 
such as purchase of credits toward compliance under the proposed credit 
provisions, had been pursued and yet were insufficient. As with other 
types of relief established in this rule, this type of temporary waiver 
would have to be designed to prevent fuel exceeding the 15 ppm standard 
from being used in 2011 and later model year nonroad engines.
    The conditions for obtaining a nonroad diesel waiver are similar to 
those in the RFG, Tier 2 gasoline sulfur, and highway diesel 
regulations. These conditions are necessary and appropriate to ensure 
that any waivers that are granted are limited in scope, and that 
refiners do not gain economic benefits from a waiver. Therefore, 
refiners seeking a waiver would need to show that the waiver is in the 
public interest, that the refiner was not able to avoid the 
nonconformity, that it would make up the air quality detriment 
associated with the waiver, that it would make up any economic benefit 
from the waiver, and that it would meet the applicable diesel sulfur 
standards as expeditiously as possible.
b. Temporary Waivers Based on Extreme Hardship Circumstances
    In addition to the provision for short-term relief in extreme 
unforseen circumstances, we are proposing a provision for relief based 
on extreme hardship circumstances that is very similar to those 
established in the

[[Page 28421]]

gasoline sulfur and highway diesel sulfur programs. Under the gasoline 
sulfur program, we granted waivers to four refiners. Each waiver was 
designed for the specific situation of that refiner. Under the highway 
diesel program, we have received two applications for which the 
decisions are still pending.
    As in the earlier rules, we have considered whether any refiners 
would face particular difficulty in complying with the standards in the 
lead time provided. As described earlier in this section, we concluded 
that in general small refiners would experience more difficulty in 
complying with the standards on time because they have less ability to 
raise the capital necessary for refinery investments, face 
proportionately higher costs because of poorer economies of scale, and 
are less able to successfully compete for limited engineering and 
construction resources. However, it is possible that other refiners 
that are not small refiners would also face particular difficulty in 
complying with the sulfur standards on time. Therefore, we are 
including in this proposed rule a provision which allows us, at our 
discretion, to grant temporary waivers from the proposed nonroad diesel 
sulfur standards based on a showing of extreme hardship circumstances.
    The extreme hardship provision allows any domestic or foreign 
refiner to request a waiver from the sulfur standards based on a 
showing of unusual circumstances that result in extreme hardship and 
significantly affect a refiner's ability to comply with either the 500 
ppm or 15 ppm sulfur diesel standards by either June 1, 2007, or June 
1, 2010, respectively. EPA would evaluate each application on a case-
by-case basis, considering the factors described below. If EPA approved 
a hardship application, we could provide refiners with relief similar 
to the provision for small refiners. That is, we might provide an 
allowance for producing high sulfur fuel during the 2007-2010 period 
when the 500 ppm cap is in effect, or an allowance for producing 500 
ppm fuel for a period of time after June 1, 2010. Depending on the 
situation of the refiner, such approved delays in meeting the sulfur 
requirements might be shorter than those allowed for small refiners 
i.e., 3 years for high sulfur fuel beginning June 1, 2007, and 4 years 
for 500 ppm fuel beginning June 1, 2010, but would not be longer. In 
such an approval, we would expect to impose appropriate conditions to 
assure the refiner is making its best effort and to minimize any loss 
of emission control from the program. As with other relief provisions 
established in this rule, any waiver under this provision would be 
designed to prevent fuel exceeding the 15 ppm standard from being used 
in 2011 and later model year nonroad engines.
    Providing short-term relief to those refiners that need additional 
time because they face hardship circumstances facilitates adoption of 
an overall program that reduces NRLM diesel fuel sulfur to 500 ppm 
beginning in 2007, and nonroad diesel fuel sulfur to 15 ppm in 2010, 
for the majority of the industry. However, we do not intend for this 
waiver provision to encourage refiners to delay planning and 
investments they would otherwise make. We do not expect to grant 
temporary waivers that apply to more than approximately one percent of 
the national NRLM diesel fuel pool in any given year.
    The regulatory language for today's action includes a list of the 
information that must be included in a refiner's application for an 
extreme hardship waiver. If a refiner fails to provide all the 
information, as specified in the regulations, as part of its hardship 
application, we can deem the application void. EPA may request 
additional information as needed. The following are some examples of 
the types of information that must be contained in an application:
    [sbull] The crude oil refining capacity and fuel sulfur level(s) of 
each diesel fuel product at each of the refiner's refineries.
    [sbull] Technical plan for capital equipment and operating changes 
to achieve future diesel fuel sulfur levels.
    [sbull] The anticipated timing for the overall project the refiner 
is proposing and key milestones to ultimately produce 100 percent of 
NRLM diesel fuel at 500 ppm sulfur and 100 percent of its nonroad 
diesel fuel at 15 ppm sulfur.
    [sbull] The refiner's capital requirements for each step of the 
proposed projects.
    [sbull] Detailed plans for financing the project and financial 
statements demonstrating the nature of and degree of financial hardship 
and how the requested relief would mitigate this hardship. This would 
include a description of the overall financial situation of the company 
and its plans to secure financing for the desulfurization project 
(e.g., internal cash flow, bank loans, issuing of bonds, sale of 
assets, or sale of stock).
    [sbull] Description of the market area for the refiner's diesel 
fuel products.
    [sbull] A plan demonstrating how they would achieve the standards 
as quickly as possible, including a timetable for obtaining the 
necessary capital, contracting for engineering and construction 
resources, obtaining any necessary permits, and beginning and 
completing construction.
    We would consider several factors in our evaluation of the hardship 
waiver applications. Such factors would include whether a refinery's 
configuration is unique or atypical; the proportion of non-highway 
diesel fuel production relative to other refinery products; whether the 
refiner, its parent company, and its subsidiaries are faced with severe 
economic limitations (for example, a demonstrated inability to raise 
necessary capital or an unfavorable bond rating); and steps the refiner 
has taken to attempt to comply with the standards, including efforts to 
obtain credits towards compliance. In addition, we would consider the 
total crude oil capacity of the refinery and its parent or subsidiary 
corporations, if any, in assessing the degree of hardship and the 
refiner's role in the diesel market. Finally, we would consider where 
the diesel fuel would be sold in evaluating the environmental impacts 
of granting a waiver.
    This extreme hardship provision is intended to address unusual 
circumstances that should be apparent now or would emerge in the near 
future. Thus, refiners seeking additional time under this provision 
would have to apply for relief by June 1, 2005. We request comment on 
this date and whether a separate date would be appropriate for the 
second (15 ppm) step of the nonroad diesel program to 15 ppm. We would 
review and act on applications and, if a waiver is granted, would 
specify a detailed desulfurization schedule under the waiver. 
Typically, because of EPA's comprehensive evaluation both financial and 
technical information, action on hardship applications can take six or 
more months.

D. Should Any Individual States or Territories Be Excluded From This 
Rule?

1. Alaska
    We propose that the diesel fuel sulfur standards--the 500 ppm cap 
for NRLM diesel fuel beginning June 1, 2007, and the 15 ppm cap for 
nonroad diesel fuel beginning June 1, 2010--and the aromatics and 
cetane standards proposed today apply to the portion of Alaska served 
by the Federal Aid Highway System. However, we propose that Alaska's 
rural areas be excluded from these proposed fuel content standards. The 
engine standards proposed today would apply to all nonroad engines 
throughout Alaska.

[[Page 28422]]

Consequently, even in rural Alaska we would still require 2011 and 
later model year nonroad diesel engines and equipment to be fueled with 
15 ppm diesel fuel. The rationale supporting this proposal follows.
a. How Was Alaska Treated Under the Highway Diesel Standards?
    Unlike the rest of the nation, Alaska is currently exempt from the 
500 ppm sulfur standard for highway diesel fuel and the dye provisions 
for diesel fuel not subject to this standard. Since the beginning of 
the 500 ppm highway diesel fuel program, we have granted Alaska 
exemptions from both the sulfur standard and dye provisions because of 
its unique geographical, meteorological, air quality, and economic 
factors.\251\
---------------------------------------------------------------------------

    \251\ Copies of information regarding Alaska's petition for 
exemption, subsequent requests by Alaska, public comments received, 
and actions by EPA are available in public docket A-96-26.
---------------------------------------------------------------------------

    On December 12, 1995, Alaska submitted a petition for a permanent 
exemption for all areas of the state served by the Federal Aid Highway 
System, that is, those areas previously covered only by a temporary 
exemption. While considering that petition, we started work on a 
nationwide rule to consider more stringent highway diesel fuel 
requirements for sulfur content. In the subsequent January 18, 2001, 
highway diesel sulfur rule (66 FR 5002) the highway engine emission 
standards were applied fully in Alaska. Based on factors unique to 
Alaska, we provided the State with: (1) an extension of the exemption 
from the 500 ppm sulfur highway diesel fuel standard until the 
effective date of the new 15 ppm sulfur standard for highway diesel 
fuel in 2006, (2) an opportunity to request an alternative 
implementation plan for the 15 ppm sulfur diesel fuel program, and (3) 
a permanent exemption from the diesel fuel dye provisions.
    In response to these provisions in our January 18, 2001, highway 
rule, Alaska informed us that areas served by the Federal Aid Highway 
System, i.e., communities on the connected road system or served by the 
Alaska State ferry system, would follow the nationwide requirements. 
Diesel fuel produced for use in areas of Alaska served by the Federal 
Aid Highway System will therefore be required to meet the same 
requirements for highway diesel fuel as diesel fuel produced for the 
rest of the nation. For the rural parts of the State, areas not served 
by the Federal Aid Highway System, Alaska informed us that it would 
submit by mid-2003 the details for an alternative implementation 
approach.\252\ EPA will consider their alternative implementation 
approach when it is received, and if appropriate will initiate 
rulemaking to finalize its adoption.
---------------------------------------------------------------------------

    \252\ Letter and attached document to Jeffrey Holmstead of EPA 
from Michele Brown of the Alaska Department of Environmental 
Conservation, dated April 1, 2002. The communities on the connected 
road system or served by the Alaska State ferry system are listed in 
the attached document.
---------------------------------------------------------------------------

b. What Nonroad Standards Do We Propose for Urban Areas of Alaska?
    Since Alaska is currently exempt from the 500 ppm sulfur standard 
for highway diesel fuel, we also considered exempting Alaska from the 
500 ppm step of the proposed NRLM standards. However, despite the 
exemption, officials from the State of Alaska have informed us that 500 
ppm highway diesel fuel is nevertheless being marketed in many parts of 
Alaska. Market forces have brought the prices for 500 ppm diesel fuel 
down such that it is now becoming competitive with higher sulfur, 
uncontrolled diesel fuel. Assuming this trend continues, requiring that 
NRLM diesel fuel be produced to 500 ppm beginning June 1, 2007 would 
not appear to be unduly burdensome and for this reason, we propose that 
this standard apply.
    At the same time, our expectation is that compliance with the 
highway program described above may result in the transition of all of 
the highway diesel fuel distribution system to 15 ppm beginning in 
2006. It could prove very challenging for the distribution system in 
some of the areas to segregate a 500 ppm grade of NRLM from a 15 ppm 
grade of highway and an uncontrolled grade for other purposes. We 
believe economics would determine whether the distribution system would 
handle the new grade of fuel or substitute 15 ppm sulfur highway diesel 
fuel for NRLM applications. Thus, in the 2007 to 2010 time frame, the 
NRLM market in some urban areas might be supplied with 500 ppm sulfur 
diesel, and in other areas might be supplied with 15 ppm sulfur diesel.
    Regardless of what takes place prior to 2010, we anticipate that 15 
ppm highway diesel fuel will be made available in Alaska by this time 
frame. The 2007 and later model year highway fleet will be growing, 
demanding more and more supply of 15 ppm diesel fuel. Adding nonroad 
volume to this would not appear to create any undue burden. Thus, we 
also propose that the 15 ppm standard for nonroad diesel fuel would 
apply in areas of Alaska served by the FAHS, along with the rest of the 
Nation beginning June 1, 2010. We seek comment on whether the 500 ppm 
NRLM diesel standard should apply to these areas of Alaska beginning 
June 1, 2007, and whether the 15 ppm nonroad standard should apply 
beginning June 1, 2010.
    During the development of the original 500 ppm highway diesel fuel 
standards in the early 1990's refiners and distributors in Alaska 
expressed concern that if Alaska were required to dye its non-highway 
diesel fuel red along with the rest of the country, residual dye in 
tanks or other equipment would be enough to contaminate and disqualify 
Jet-A kerosene used as aviation fuel. Since much of the diesel fuel in 
Alaska is number 1 and indistinguishable from Jet A kerosene, not only 
would tanks and transfer equipment have to be cleaned, but separate 
tankage would be needed. Consequently, we granted Alaska temporary 
exemptions from the dye requirement and in the January 18, 2001, 
highway diesel rule granted them a permanent exemption. The proposed 
marker for heating oil in the 2007-10 time period and for locomotive 
and marine diesel fuel in the 2010-14 time period could present similar 
concerns in Alaska's distribution system. Consequently, we seek comment 
on whether to extend the current exemption from the red dye requirement 
to the proposed marker requirement. If we were to, we then also seek 
comment on what mechanism could be used in Alaska to ensure that 500 
ppm diesel fuel was used in NRLM equipment from 2007-10 and 15 ppm in 
nonroad equipment after 2010. One possible approach would be to 
preclude refineries and importers from using credits to comply with the 
sulfur standards and prohibit end-users in Alaska from using anything 
but 500 ppm in NRLM equipment from 2007-10 and 15 ppm in nonroad 
equipment after 2010.
c. What Do We Propose for Rural Areas of Alaska?
    Rural Alaska represents a rather unique situation. In the rural 
areas, the state estimates that the heating oil represent approximately 
95% of all distillate consumption (about 50% for heating and 45% for 
electricity generation). Highway vehicles account for about 1 percent, 
and marine engines about 4 percent.\253\ Consequently, nonroad and 
locomotive engines and equipment consume a negligible amount of diesel 
fuel in the rural areas. The fuel

[[Page 28423]]

storage infrastructure in the villages generally consists of a limited 
number of small community storage tanks. The fuel must last during the 
entire winter season when fuel deliveries may not be possible. There is 
currently only one distillate fuel that is delivered and stored for all 
distillate purposes in the villages, including home heating, power 
generation, vehicles, marine engines and possibly some nonroad engines 
and equipment. Modifications to permit the segregation of small amounts 
of low sulfur or ultra low-sulfur distillate fuel for highway and/or 
NRLM use or switching to low sulfur or ultra low-sulfur fuel for all 
purposes would be an economic hardship for the villages.
---------------------------------------------------------------------------

    \253\ E mail from the Alaska Department of Environmental 
Conservation, dated July 2, 2002.
---------------------------------------------------------------------------

    Furthermore, as discussed above, for areas not served by the 
Federal Aid Highway System, the State of Alaska is considering an 
alternative implementation plan for the 15 ppm and 500 ppm highway 
standards. One option under consideration by the State would be to not 
apply these standards in these areas. Rather, the 15 ppm fuel would be 
provided based on demand to 2007 and later model year vehicles that 
must be operated on 15 ppm fuel as they enter the fleet. Since the 
vehicle turnover rate in rural villages is typically very low, and many 
of the replacement vehicles are pre-owned vehicles themselves, some 
villages may not obtain their first 2007 or later model year diesel 
highway vehicle until long after 2010. If such a highway plan would be 
finalized and EPA subsequently incorporated it into the regulations, 
the proposed NRLM low-sulfur diesel fuel program, without similar 
provisions, would require 500 ppm diesel fuel solely for the NRLM 
market in rural areas beginning June 1, 2007, and 15 ppm sulfur solely 
for the nonroad market beginning June 1, 2010. Since the demand for new 
nonroad engines and equipment with aftertreatment (model year 2011 and 
later) is expected to be nonexistent or very low in the early years in 
rural Alaska, we believe the best approach is to propose no sulfur or 
other content requirements for areas of Alaska not served by the FAHS. 
EPA can revisit this when it receives and takes action on Alaska's 
highway implementation plan. This will allow for coordination between 
the highway and NRLM fuel requirements. As proposed, this would allow 
rural Alaska to limit the volume of 15 ppm sulfur diesel fuel to that 
which is sufficient to meet the demand from the small number of new 
nonroad diesel engines and equipment that would be certified to the 
Tier 4 nonroad standards proposed today beginning with the 2011 model 
year.
    Our goal in proposing this approach is to allow rural Alaska to 
transition to the low sulfur fuel program in a manner that minimizes 
costs while still ensuring that the model year 2011 and later nonroad 
engines and equipment with aftertreatment receive the 15 ppm diesel 
fuel they need. Similar to the flexibility being considered under the 
highway program, the flexibility offered by this proposal would likely 
result in a delay of some sulfate emission reduction benefits in the 
rural areas of Alaska. The sulfate emissions of NRLM engines and 
equipment in Alaska would remain at current levels for as long as high-
sulfur diesel fuel is used.
2. American Samoa, Guam, and the Commonwealth of Northern Mariana 
Islands
a. What Provisions Apply in American Samoa, Guam, and the Commonwealth 
of Northern Mariana Islands?
    We are proposing to exclude American Samoa, Guam and the 
Commonwealth of the Northern Mariana Islands from the proposed NRLM 
diesel fuel sulfur standard of 500 ppm sulfur in 2007 and 15 ppm sulfur 
nonroad standard in 2010, as well as the cetane index and aromatics 
requirements. We also propose to exclude these territories from the 
Tier 4 nonroad vehicle, engine and equipment emissions standards, and 
other requirements associated with those emission standards. The 
territories will continue to have access to new nonroad diesel engines 
and equipment using pre-Tier 4 technologies, at least as long as 
manufacturers choose to market those technologies. We will not allow 
the emissions control technology in the territories to backslide from 
those available in 2010. If, in the future, manufacturers choose to 
market only nonroad diesel engines and equipment with Tier 4 emission 
control technologies, we believe the market will determine if and when 
the territories will make the investment needed to obtain and 
distribute the diesel fuel necessary to support these technologies.
    We are also proposing to require that all nonroad diesel engines 
and equipment for these territories be certified and labeled to the 
applicable requirements--either to the 2010 model year standards and 
associated requirements under this proposed exclusion, or to the 2011 
and later standards and associated requirements applicable for the 
model year of production under the nationwide requirements of this 
proposal--and warranted, as otherwise required under the Clean Air Act 
and EPA regulations. Special recall and warranty considerations due to 
the use of excluded high sulfur fuel would be the same as those for 
Alaska during its exemption and transition periods for highway diesel 
fuel and for these territories for highway diesel fuel (see 66 FR 5086, 
5088, January 18, 2001).
    To protect against this exclusion being used to circumvent the 
emission requirements applicable to the rest of the United States, we 
are restricting the importation of nonroad engines and equipment from 
these territories into the rest of the United States. After the 2010 
model year, nonroad diesel engines and equipment certified under this 
exclusion to meet the 2010 model year emission standards for sale in 
American Samoa, Guam and the Commonwealth of the Northern Mariana 
Islands will not be permitted entry into the rest of the United States.
b. Why Are We Treating These Territories Uniquely?
    Like Alaska, these territories are currently exempt from the 500 
ppm sulfur standard for highway diesel fuel. Unlike Alaska and the rest 
of the nation, they are also exempt from the new highway diesel fuel 
standard effective in 2006 and the new highway vehicle and engine 
emission standards effective beginning in 2007 (see 66 FR 5088, January 
18, 2001).
    Section 325 of the CAA provides that upon request of Guam, American 
Samoa, the Virgin Islands, or the Commonwealth of the Northern Mariana 
Islands, we may exempt any person or source, or class of persons or 
sources, in that territory from any requirement of the CAA, with some 
specific exceptions. The requested exemption could be granted if we 
determine that compliance with such requirement is not feasible or is 
unreasonable due to unique geographical, meteorological, or economic 
factors of the territory, or other local factors as we consider 
significant. Prior to the effective date of the current highway diesel 
sulfur standard of 500 ppm, the territories of American Samoa, Guam and 
the Commonwealth of Northern Mariana Islands petitioned us for an 
exemption under section 325 of the CAA from the sulfur requirement 
under section 211(i) of the CAA and associated regulations at 40 CFR 
80.29. We subsequently granted the petitions.\254\ We recently 
determined that the 2007 heavy-duty emission standards and 2006 diesel 
fuel sulfur

[[Page 28424]]

standard of our January 18, 2001 highway rule (66 FR 5088) would not 
apply to these territories.
---------------------------------------------------------------------------

    \254\ See 57 FR 32010, July 20, 1992 for American Samoa; 57 FR 
32010, July 30, 1992 for Guam; and 59 FR 26129, May 19, 1994 for 
CNMI.
---------------------------------------------------------------------------

    Compliance with this proposal would result in major economic 
burden. All three of these territories lack internal petroleum supplies 
and refining capabilities and rely on long distance imports. Given 
their remote location from Hawaii and the U.S. mainland, most petroleum 
products are imported from East rim nations, particularly Singapore. 
Although Australia, the Philippines, and certain other Asian countries 
have or will soon require low sulfur diesel fuel, their sulfur limit is 
500 ppm, not the new 15 ppm sulfur limit established for highway diesel 
fuel by the January 18, 2001, highway rule or this proposal for nonroad 
diesel fuel beginning in 2010 for the United States. Compliance with 
new 15 ppm sulfur requirements for highway diesel fuel beginning in 
2006 and the proposed 15 ppm sulfur requirements for nonroad diesel 
fuel beginning in 2010 (or the proposed 500 ppm sulfur requirements for 
NRLM diesel fuel beginning 2007) would require construction of separate 
storage and handling facilities for a unique grade of diesel fuel for 
highway and nonroad purposes, or use of 15 ppm diesel fuel for all 
purposes to avoid segregation. Either of these alternatives would 
require importation of 500 and 15 ppm sulfur diesel fuel from Hawaii or 
the U.S. mainland, and would significantly add to the already high cost 
of diesel fuel in these territories, which rely heavily on United 
States support for their economies. At the same time, it is not clear 
that the environmental benefits in these areas would warrant this cost. 
Therefore, we are not proposing to apply the fuel and engine standards 
to these territories, but seek comment on this.

E. How Are State Diesel Fuel Programs Affected by the Sulfur Diesel 
Program?

    Section 211(c)(4)(A) of the CAA prohibits states and political 
subdivisions of states from prescribing or attempting to enforce, for 
purposes of motor vehicle emission control, ``any control or 
prohibition respecting any characteristic or component of a fuel or 
fuel additive in a motor vehicle or motor vehicle engine,'' if EPA has 
prescribed ``a control or prohibition applicable to such characteristic 
or component of the fuel or fuel additive'' under section 211(c)(1). 
This prohibition applies to all states except California, as explained 
in section 211(c)(4)(B). This express preemption provision in section 
211(c)(4)(A) applies only to controls or prohibitions respecting any 
characteristics or components of fuels or fuel additives for motor 
vehicles or motor vehicle engines, that is, highway vehicles. It does 
not apply to controls or prohibitions respecting any characteristics or 
components of fuels or fuel additives for nonroad engines or nonroad 
vehicles.\255\
---------------------------------------------------------------------------

    \255\ See 66 FR 36543 (July 12, 2001) (Notice proposing approval 
of Houston SIP revisions). See also letter from Carl Edlund, 
Director, Multimedia Planning and Permitting Division, U.S. 
Environmental Protection Agency, Region VI, to Jeffrey Saitas, 
Executive Director, Texas Natural Resources Conservation Commission, 
dated September 25, 2000, providing comments on proposed revisions 
to the Texas State Implementation Plan for the control of ozone, 
specifically the Post 99 Rate of Progress Plan and Attainment 
Demonstration for the Houston/Galveston area. This letter noted that 
preemption under section 211(c)(4) did not apply to controls on 
nonroad diesel fuel.
---------------------------------------------------------------------------

    Section 211(c)(4)(A) specifically mentions only controls respecting 
characteristics or components of fuel or fuel additives in a ``motor 
vehicle or motor vehicle engine,'' adopted ``for purposes of motor 
vehicle emissions control,'' and the definitions of motor vehicle and 
nonroad engines and vehicles in CAA section 216 are mutually exclusive. 
This is in contrast to section 211(a) and (b), which specifically 
mention application to fuels or fuel additives used in nonroad engines 
or nonroad vehicles, and with section 211(c)(1) which refers to fuel 
used in motor vehicles or engines or nonroad engines or vehicles.
    Thus, this proposal would not preempt state controls or 
prohibitions respecting characteristics or components of fuel or fuel 
additives used in nonroad engines or nonroad vehicles under the 
provisions of section 211(c)(4)(A). At the same time, a state control 
that regulates both highway fuel and nonroad fuel is preempted to the 
extent the state control respects a characteristic or component of 
highway fuel regulated by EPA under section 211(c)(1).
    A court could consider whether a state control for fuels or fuel 
additives used in nonroad engines or nonroad vehicles is implicitly 
preempted under the Supremacy Clause of the U.S. Constitution. Courts 
have determined that a state law is preempted by federal law where the 
state requirement actually conflicts with federal law by preventing 
compliance with the federal requirement, or by standing as an obstacle 
to accomplishment of Congressional objectives. A court could thus 
consider whether a given state standard for sulfur in nonroad, 
locomotive or marine diesel fuel is preempted if it places such 
significant cost and investment burdens on refiners that refiners 
cannot meet both state and federal requirements in time, or if the 
state control would otherwise meet the criteria for conflict 
preemption.

F. Technological Feasibility of the 500 and 15 ppm sulfur Diesel Fuel 
Program

    This section describes the nonroad, locomotive and marine diesel 
fuel market and how these fuels differ from current highway diesel 
fuel, whose sulfur content is already controlled to no more than 500 
ppm sulfur. This section then summarizes our assessment of the 
feasibility of refining and distributing NRLM diesel fuel with a sulfur 
content of no more than 500 ppm and, for nonroad fuel only, of 15 ppm. 
Based on this evaluation, we believe it is technologically feasible for 
refiners and distributors to meet both sulfur standards in the lead 
time provided. We are only summarizing our analysis here and we refer 
the reader to the Draft RIA for more details.
1. What is the Nonroad, Locomotive and Marine Diesel Fuel Market Today
    Nonroad, locomotive and marine diesel fuel comprise part of what is 
generally called the distillate fuel market. Other fuels in this market 
are highway diesel fuel and heating oil, which is used in furnaces and 
boilers as well as in stationary diesel engines to generate power. 
Nonroad diesel fuel comprises about 15% of all number 2 distillate 
fuel, while locomotive and marine diesel fuel comprise about 9% of all 
number 2 distillate fuel (see Draft RIA).
    ASTM defines three number 2 distillate fuels: (1) low sulfur No. 2-
D (which includes the 500 ppm sulfur cap for fuel used in highway 
diesel vehicles), (2) high sulfur No. 2-D, and (3) No. 2 fuel oil 
(commonly referred to as heating oil).\256\ Low sulfur No. 2-D fuel 
must contain no more than 500 ppm sulfur, have a minimum cetane number 
of 40, and have a minimum cetane index limit of 40 (or a maximum 
aromatic content of 35 volume percent). This fuel meets EPA's 
requirements for current highway diesel vehicle fuel. Both high sulfur 
No. 2-D and No. 2 fuel oil must contain no more than 5000 ppm 
sulfur.\257\ The ASTM standards for high sulfur No. 2-D fuel also 
include a minimum cetane number specification of 40. Practically, since 
most No. 2 fuel oil meets the minimum cetane number specification, 
pipelines which ship fuel fungibly need only carry one high sulfur

[[Page 28425]]

number 2 distillate fuel which meets both sets of specifications. 
Nonroad, locomotive and marine engines can be and are fueled with both 
low and high sulfur No. 2-D fuels.
---------------------------------------------------------------------------

    \256\ ``Standard Specification for Diesel Fuel Oils,'' ASTM D 
975-98b and ``Standard Specification for Fuel Oils,'' ASTM D 396-98.
    \257\ Some states, particularly those in the Northeast, limit 
the sulfur content of No. 2 fuel oil to 2000-3000 ppm.
---------------------------------------------------------------------------

    During winter months in the northern U.S., No. 1 distillate, such 
as kerosene, is sometimes added to No. 2 distillate fuel to prevent 
gelling. Any No. 1 distillate added to No. 2 NRLM diesel fuel would 
become NRLM diesel fuel.
    Highway diesel fuel, comprises about 57% of all number 2 distillate 
fuel. Eighty percent of highway diesel fuel will be capped at 15 ppm 
sulfur starting in 2006. However, because of limitations in the fuel 
distribution system and other factors, about one-third of non-highway, 
No. 2 distillate currently meets the 500 ppm highway diesel fuel cap. 
Thus, about 69 percent of number 2 distillate pool currently meets the 
500 ppm sulfur cap, not just the 57 percent used in highway vehicles. 
The result is that about one-third of the 24% of the distillate market 
comprised by NRLM diesel fuel currently meets a 500 ppm specification 
and is also expected to meet the future highway diesel fuel 
requirements even without this proposed rule. Thus, while this proposed 
rule would apply to all NRLM diesel fuel, the rule should only 
materially affect about two-thirds of all NRLM diesel fuel, or 16% of 
today's distillate market. EPA is not considering any national sulfur 
standards applicable to home heating fuel or power generation fuel at 
this time.
2. How Do Nonroad, Locomotive and Marine Diesel Fuel Differ From 
Highway Diesel Fuel?
    Refiners blend together a variety of distillate blendstocks to 
produce both highway and non-highway diesel fuels. These distillate 
blendstocks always include straight run material contained in crude 
oil, plus they often include light cycle oil from a fluidized catalytic 
cracker, light coker gas oil from a coker and hydrocrackate from a 
hydrocracker. The actual mix of these blendstocks in highway and non-
highway diesel fuel at refineries producing both fuels can differ. 
However, in general, significant quantities of all of these blendstocks 
find their way into both low sulfur and high sulfur diesel fuel today. 
A survey of distillate fuel quality conducted by API and NPRA in 1996 
indicated the following feedstock composition for low sulfur diesel 
fuel and high sulfur diesel fuel and heating oil.

Table IV-5--Composition of Low Sulfur Diesel Fuel and High Sulfur Diesel
   Fuel and Heating Oil: 1996 U.S. Non-California Average of Surveyed
                       Refiners (Volume Percent)a
------------------------------------------------------------------------
                                                      High Sulfur No. 2
         Feedstocks             Low Sulfur No. 2       Diesel Fuel and
                                   Diesel Fuel           Heating Oil
------------------------------------------------------------------------
                              Hydrotreated
------------------------------------------------------------------------
Straight Run Material.......                    52                    18
Light Cycle Oil.............                    20                    11
Light Coker Gas Oil.........                     8                     5
Hydrocrackate...............                     4                     9
-----------------------------
                            Non-Hydrotreated
------------------------------------------------------------------------
Straight Run Material.......                    12                    45
Light Cycle Oil.............                     3                    11
Light Coker Gas Oil.........                     1                    1
------------------------------------------------------------------------
Notes:
a We plan to update these compositions to reflect greater use of heavier
  crude oils in future analyses.

    The primary difference between low and high sulfur number 2 
distillate fuels today is the fact that a greater volume percentage of 
low sulfur fuel feedstocks have been hydrotreated to meet the 500 ppm 
sulfur cap applicable to highway diesel fuel. As shown in the table 
above, high sulfur distillate fuels may contain significant amounts of 
hydrotreated material, but the final sulfur level of the blend is 
usually well above 500 ppm and currently averages 3400 ppm (see Draft 
RIA). Hydrotreating today typically involves combining diesel fuel with 
hydrogen and a catalyst under pressures of 400-1200 pounds per square 
inch and temperatures of roughly 600 degrees Fahrenheit. In general, 
the existence of the 500 ppm sulfur cap gives refiners an incentive to 
use low sulfur blendstocks, such as hydrocrackate and straight run, in 
their low sulfur diesel fuel. However, some high sulfur blendstocks, 
such as light cycle oil and light gas coker oil, require hydrotreating 
to remove other undesirable compounds, such as olefins and metals. Once 
hydrotreated, they are suitable for use in low sulfur diesel fuel. 
Also, some light cycle oils and light gas coker oils contain so much 
sulfur and olefins and have such a low cetane number that they are 
unsuitable for direct blending into even high sulfur diesel fuel, since 
most high sulfur diesel fuel meets the ASTM sulfur cap of 5000 ppm and 
cetane number minimum of 40.\258\ Where material is hydrotreated in 
order to blend into a high sulfur fuel, it is often easier to 
hydrotreat the material further to meet a 500 ppm cap and blend 
straight run material directly into the high sulfur diesel pool. Thus, 
there is no bright line separating the blendstocks used to produce low 
and high sulfur diesel fuel today.
---------------------------------------------------------------------------

    \258\ Non-highway diesel fuel often meets sulfur standards of 
2000-3000 ppm in some states, particularly those in the Northeast. 
These states have limited the sulfur content of home heating oil to 
these levels. To ease fuel distribution, refiners and distributors 
sell the same fuel into the home heating fuel and non-highway diesel 
fuel markets.
---------------------------------------------------------------------------

3. What Technology Would Refiners Use to Meet the Proposed 500 ppm 
Sulfur Cap?
    Refiners currently hydrotreat some or all of their distillate 
blendstocks to meet the 500 ppm sulfur cap applicable to highway diesel 
fuel. Refiners would be able to meet the proposed 500 ppm sulfur cap 
for NRLM diesel fuel using this same technology. As will be discussed 
further in the next section, several alternative desulfurization 
technologies are being developed. However, these alternative 
technologies promise the greatest cost savings at very low sulfur 
levels, such as 15 ppm. Also, their ongoing development makes it

[[Page 28426]]

unlikely that they would be selected by most refiners for production as 
early as 2007. Finally, the use of conventional hydrotreating 
technology to meet a 500 ppm standard can readily be combined later 
with these alternative technologies to meet the subsequent 15 ppm 
standard in 2010. Thus, we expect that the vast majority of refiners 
would use conventional hydrotreating to meet the 500 ppm standard in 
2007 applicable to NRLM diesel fuel.
    Refiners would also likely need to install or modify several 
existing ancillary units related to sulfur removal (e.g., hydrogen 
production and purification, sulfur recovery, amine scrubbing and sour 
water scrubbing facilities). All of these units currently exist at the 
vast majority of refineries, but may have to be expanded or enlarged.
4. Has Technology to Meet a 500 ppm Cap Been Commercially Demonstrated?
    Conventional diesel desulfurization technologies have been 
available and in use for many years. U.S. refiners have nearly ten 
years of experience with this technology in producing diesel fuel with 
less than 500 ppm sulfur for highway use. Thus, the technology to 
produce 500 ppm NRLM diesel fuel has clearly been demonstrated and 
optimized over the last decade.
5. Availability of Leadtime To Meet the 2007 500 ppm Sulfur Cap
    About 105 refineries in the U.S. currently produce high sulfur 
distillate fuel. Under the fuel-related provisions of this proposal, we 
project that roughly 42 of these refineries would likely need to 
produce 500 ppm NRLM diesel fuel to satisfy the demand for this fuel. 
The remaining 63 or so refineries would continue to produce high sulfur 
distillate fuel, either as heating oil or as high sulfur NRLM diesel 
fuel.
    If we promulgate this proposal one year from today, this would 
provide refiners and importers with approximately 38 months before they 
would have to begin complying with the 500 ppm cap for NRLM diesel fuel 
on June 1, 2007. Our leadtime analysis, which is presented in the draft 
RIA, projects that 27-39 months are typically needed to design and 
construct a diesel fuel hydrotreater.\259\ Thus, the leadtime available 
for the 500 ppm cap in mid-2007 should be sufficient.
---------------------------------------------------------------------------

    \259\ ``Highway Diesel Progress Review,'' USEPA, EPA420-R-02-
016, June 2002.
---------------------------------------------------------------------------

    Easing the task is the fact that we project that essentially all 
refiners would use conventional hydrotreating to comply with the 500 
ppm NRLM diesel fuel cap. This technology has been used extensively for 
more than 10 years and its capabilities to process a wide range of 
diesel fuel blendstocks are well understood. Thus, the time necessary 
to optimize this technology for a specific refiner's situation should 
be relatively short.
    While conventional hydrotreating would likely be used to meet the 
500 ppm cap in 2007, most refiners would have to plan to be able 
process this fuel further to meet the 15 ppm nonroad diesel fuel cap in 
2010. Even those refiners planning on producing 500 ppm locomotive and 
marine diesel fuel starting in 2010 would likely have to plan for the 
potential that this fuel could be controlled to 15 ppm sulfur at some 
time in the future. Thus, the conventional hydrotreater built in 2007 
would have to be able to be compatible with the technology eventually 
chosen to produce 15 ppm fuel in 2010 or later. This could affect the 
hydrotreater's design pressure, physical location and layout and 
peripherals, such as hydrogen supply and utilities. However, we project 
that 34 out of the 42 refineries which we project would produce this 
fuel also produce highway diesel fuel. Thus, over 80 percent of the 
refiners likely to produce 500 ppm NRLM fuel in 2007 are already well 
into their planning for meeting the 15 ppm highway diesel fuel 
standard, effective June 1, 2006. It is likely that these refiners have 
already chemically characterized their high sulfur diesel fuel 
blendstocks, as well as their highway diesel fuel, for potential 
desulfurization. They will also have already assessed the various 
technologies for producing 15 ppm diesel fuel and have a good idea of 
what technology they might use to meet the 15 ppm nonroad diesel fuel 
cap starting in 2010. Those refiners which only produce high sulfur 
distillate fuel today would still be able to take advantage of the 
significant experience that technology vendors have obtained in helping 
refiners of highway diesel fuel plan for producing 15 ppm diesel fuel 
in 2006.
    Also, of the 34 refineries producing highway diesel fuel today, we 
project that three will likely build a new hydrotreater to produce 15 
ppm highway diesel fuel in 2006. This would allow them to produce 500 
ppm NRLM diesel fuel using their existing highway diesel fuel 
hydrotreater. Another 10 of these 34 refineries produce relatively 
small volumes of high sulfur distillate compared to highway diesel fuel 
today. Thus, we project that they should be able to produce 500 ppm 
NRLM fuel from their high sulfur distillate with minor modification to 
their existing hydrotreater.
    Refiners may also need some time to assess what diesel fuel and 
heating oil markets they plan on participating in starting 2010. While 
heating oil may not be widely distributed in PADDs 2, 3 and 4, refiners 
in PADDs 1 and 3 would still be able to produce heating oil for the 
Northeast fuel market. Likewise, heating oil may still be distributed 
in the Pacific Northwest. Under this proposal, locomotive and marine 
diesel fuel would remain at 500 ppm for some time. Thus, many refiners 
would require some time to decide what market to participate in after 
2010. This strategic planning should be able to coincide with refiners' 
evaluation of 15 ppm technologies and not add to the overall lead time 
required.
    In all, we project that the task of producing 500 ppm NRLM fuel in 
2007 would be less difficult than the task refiners faced with the 
implementation of the 500 ppm highway diesel fuel cap in 1993. Refiners 
had just over three years of leadtime for the highway diesel fuel cap, 
as is the case here and this proved sufficient.
6. What Technology Would Refiners Use to Meet the Proposed 15 ppm 
Sulfur Cap for Nonroad Diesel Fuel?
    We project that refiners would be able to use a variety of 
desulfurization technologies to meet the proposed 15 ppm sulfur cap for 
nonroad fuel. One approach would be to use an extension of conventional 
hydrotreating technology. We expect that refiners would utilize 
hydrotreating to meet the proposed 500 ppm standard. We expect that 
refiners would design this hydrotreater to facilitate the addition of a 
second reactor or hydrotreating stage to further desulfurize their 
distillate blendstocks from 500 ppm to 15 ppm. Refiners might also 
shift to the use of an improved catalyst even in the first reactor 
(i.e., that producing roughly 500 ppm sulfur product), as well as add 
equipment to further purify the hydrogen used.
    This is the same technology which EPA projected would be used by 
most refiners to meet the 15 ppm sulfur cap for highway diesel fuel. 
EPA just recently reviewed the progress being made by refining 
technology vendors and refiners in meeting the 2006 highway diesel 
sulfur cap.\260\ All evidence available confirms EPA's projection that 
conventional hydrotreating will be capable of producing diesel fuel 
containing less

[[Page 28427]]

than 10 ppm sulfur. Refiners producing only high sulfur distillate 
today should have an added advantage in meeting a 15 ppm sulfur cap for 
nonroad fuel over that for highway fuel. They would be able to design 
their hydrotreater from the ground up, while most refiners producing 15 
ppm diesel fuel for highway use will be trying to utilize their 
existing 500 ppm hydrotreaters, which may not be designed to be 
revamped to produce 15 ppm fuel in the most efficient manner.
---------------------------------------------------------------------------

    \260\ ``Highway Diesel Progress Review,'' EPA, June 2002, 
EPA420-R-02-016.
---------------------------------------------------------------------------

    Based on our review of the limited catalyst performance data in the 
published literature and the one set of confidential data submitted, we 
believe that the projections of the more optimistic vendors are the 
most accurate for the 2010 timeframe given this additional leadtime. 
For example, the confidential commercial data indicated that five ppm 
sulfur levels could be achieved with two-stage hydrotreating at 
moderate hydrogen pressure despite the presence of a significant amount 
of light cycle oil (LCO). The key factor was the inclusion of a 
hydrogenation catalyst in the second stage, which saturated many of the 
poly-nuclear, aromatic rings in the diesel fuel, allowing the removal 
of sulfur from the most sterically hindered compounds. In addition, 
refiners that are able to defer production of 15 ppm highway diesel 
fuel through the purchase of credits, as well as refiners producing 15 
ppm nonroad in 2010, would have the added benefit of being able to 
observe the operation of those hydrotreating units starting up in 2006. 
This should allow these refiners to be able to select from the best 
technologies which are employed in the highway program.
    In addition, a number of alternative technologies are presently 
being developed which could produce 15 ppm fuel at lower cost. 
ConocoPhillips, for example, has developed a version of their S-Zorb 
technology for diesel fuel desulfurization. This technology utilizes a 
catalytic adsorbent to remove the sulfur atom from hydrocarbon 
molecules. It then sends the sulfur-laden catalyst to a separate 
reactor, where the sulfur is removed and the catalyst is restored. 
Unipure is developing a process which selectively oxidizes the sulfur 
contained in diesel fuel. This process have the advantage that the 
sulfur containing compounds which are most difficult to desulfurize via 
hydrotreating are quite easily desulfurized via oxidation. Finally, 
Linde has developed a method which greatly improves the concentration 
of hydrogen on hydrotreating catalysts. This process promises to 
greatly reduce the reactor volume necessary to produce 15 ppm diesel 
fuel.
    These three new technologies are at various stages of development. 
This is discussed in more detail in the next section. Due to the 
projected ability of these technologies to reduce the cost of meeting a 
15 ppm sulfur cap and the leadtime available between now and 2010, we 
project that 80% of the new volume of 15 ppm nonroad diesel fuel would 
be produced using advanced technologies.
7. Has Technology to Meet a 15 ppm Cap Been Commercially Demonstrated?
    EPA just completed a review of refiners' progress in preparing to 
produce 15 ppm highway diesel fuel.\261\ The information we obtained 
during that review confirm the projections we made in the HD 2007 
program--refiners are technically capable of producing 15 ppm sulfur 
diesel fuel using extensions of conventional technology and, in fact, 
they are moving forward with their plans to comply with the program. 
Thus, we believe there are no technological hurdles to producing 15 ppm 
diesel fuel.
---------------------------------------------------------------------------

    \261\ Ibid.
---------------------------------------------------------------------------

    The European Union has also determined that diesel fuel can be 
desulfurized to meet a sulfur cap in the range of 10-15 ppm. Europe has 
established a 10 ppm sulfur cap on highway diesel fuel, effective in 
2009, with plans underway for a 10 ppm sulfur cap for nonroad diesel 
fuel soon thereafter. As with our standards, Europe's 10 ppm cap 
applies throughout the distribution system. However, fuel tends to be 
transported much shorter distances in Europe. Therefore, we believe 
that both the 10 and 15 ppm sulfur caps will require refiners to meet 
the same 7-8 ppm sulfur target at the refinery gate. Given this, the 
European standard will require the same technology as that required in 
the U.S. Most European diesel fuel must meet a higher cetane number 
specification than U.S. diesel fuel, which causes it to be 
predominantly comprised of straight run material. This material is 
easier to desulfurize to sub-15 ppm levels using conventional 
hydtrotreating technology. In some European countries, nonroad diesel 
fuel is the same as heating oil and contains significant amounts of 
cracked material. Thus, on average, it should be easier for European 
refiners to meet a 10 ppm sulfur cap with their highway diesel fuel 
than in the U.S. As the 10 ppm cap is extended to nonroad diesel fuel, 
the stringency of the European standard will be much closer to that of 
a 15 ppm cap here in the U.S.
    We have met with a number of diesel fuel refiners to learn about 
their plans to produce 15 ppm highway diesel fuel by the June 2006 
program compliance date. Since the 15 ppm diesel fuel sulfur standard 
was established based on the use of extensions of conventional diesel 
desulfurization technologies, diesel fuel refineries are well 
positioned to make firm plans for implementation by 2006. Our review 
has found that this is exactly what refiners are doing. We are very 
encouraged by the actions some refiners have already taken in terms of 
announcing specific plans for low sulfur diesel fuel production. It may 
still be early in the process, but virtually all refiners are already 
in the stage of planning their approach for compliance. Thus, the 
refining industry is where we anticipated it would be at this point in 
time. Moreover, some refining companies are ahead of schedule and will 
be capable of producing significant quantities of 15 ppm sulfur diesel 
fuel as early as next year. Thus, we expect that the capability of 
conventional hydrotreating to produce 15 ppm diesel fuel in refinery-
scale quantities will be demonstrated in the U.S. by the end of 2003.
    Phillips Petroleum is currently in the process of designing and 
constructing a commercial sized S-Zorb unit to produce sub-15 ppm 
diesel fuel at their Sweeney, Texas refinery. This plant is scheduled 
to begin commercial operation in 2004. This would provide refiners with 
roughly 3 years of operating data before they would have to decide 
which technology to use to meet the 15 ppm nonroad sulfur cap in 2010. 
This should be enough operating experience for most refiners to have 
sufficient confidence in this advanced process to include it in their 
options for 2010 compliance. Based on information received from 
Phillips Petroleum, we estimate that this technology could reduce the 
cost of meeting the 15 ppm cap for many refiners by 25 percent.
    Linde has also developed a new approach for improving the contact 
between hydrogen, diesel fuel and conventional desulfurization 
catalysts. Linde projects that their Iso-Therming process could reduce 
the hydrotreater volume required to achieve sub-15 ppm sulfur levels by 
roughly a factor of 2. Linde has already built a commercial-sized 
demonstration unit at a refinery in New Mexico and has been operating 
the equipment since September 2002. Thus, refiners would have 4-5 years 
of operating data available on this process before they would have to 
decide which technology to use to meet the 15 ppm nonroad sulfur cap in 
2010. This should be ample operating experience for

[[Page 28428]]

essentially all refiners to include this process in their options for 
2010. Based on information received from Linde, we estimate that this 
technology could reduce the cost of meeting the 15 ppm cap for many 
refiners by 40 percent.
    Finally, Unipure Corporation is developing a desulfurization 
process which oxidizes the sulfur atom in diesel fuel molecules, 
facilitating its removal. This process operates at low temperatures and 
ambient pressure, so it avoids the need for costly, thick walled, 
pressure vessels and compressors. It also consumes no hydrogen. Thus, 
it could be particularly advantageous for refiners who lack an 
inexpensive supply of hydrogen (e.g., isolated or smaller refineries 
who cannot construct a world scale hydrogen plant based on inexpensive 
natural gas). However, the oxidant is very powerful, so specialized, 
oxidation resistant materials are needed. Unipure has demonstrated its 
process at the pilot plant level, but has yet to build a commercial 
sized demonstration unit. However, time still remains for this to be 
done before refiners need to make final decisions for their 2010 
compliance plans. Thus, while more uncertain than the other two 
advanced processes, the Unipure oxidation process could be selected by 
a number of refiners to meet the 2010 15 ppm cap. Based on inputs from 
Unipure, we estimate that their process could reduce the cost of 
meeting the 15 ppm cap for roughly one-fourth of all refineries by 25-
35 percent.
    The savings associated with each technology varies with the size, 
location and complexity of the refinery. However, on average the Linde 
process appears to have the potential reduce the cost of desulfurizing 
500 ppm diesel fuel to 15 ppm by 35-40 percent. The savings associated 
with the Phillips and Unipure processes appear to be more refinery 
specific. For about 25 refineries, the Phillips process appears to have 
the potential to reduce these desulfurization costs by 20-40 percent. 
The primary advantage of the Unipure process is its lower capital 
costs. For about 30 refineries, the Unipure process appears to have the 
potential to reduce the capital investment related to produce 15 ppm 
fuel from 500 ppm diesel fuel by an average of 40 percent.
8. Availability of Leadtime To Meet the 2010 15 ppm Sulfur Cap
    If we promulgate this proposal one year from today, this would 
provide refiners and importers with more than six years before they 
would have to begin complying with the 15 ppm cap for nonroad diesel 
fuel on June 1, 2010. Our leadtime analysis, which is presented in the 
draft RIA, projects that 30-39 months are typically needed to design 
and construct a diesel fuel hydrotreater.\262\ Thus, refiners would 
have about 3 years before they would have to begin detailed design and 
construction. This would allow them time to observe the performance of 
the hydrotreaters being used to produce 15 ppm highway diesel fuel for 
at least one year. While not a full catalyst cycle, any unusual 
degradation in catalyst performance over time should be apparent within 
the first year. Thus, we project that the 2010 start date would allow 
refiners to be quite certain that the designs they select in mid-2007 
will perform adequately in 2010.
---------------------------------------------------------------------------

    \262\ ``Highway Diesel Progress Review,'' USEPA, EPA420-R-02-
016, June 2002.
---------------------------------------------------------------------------

    In addition, we expect that most of the advanced technologies will 
be demonstrated on a commercial scale by the end of 2004. Thus, 
refiners would have at least two and a half years to observe the 
performance of these technologies before having to select a technology 
to meet the 2010 15 ppm cap. This should be more than adequate to fully 
access the costs and capabilities of these technologies for all but the 
most cautious refiners.
9. Feasibility of Distributing Nonroad, Locomotive and Marine Diesel 
Fuels That Meet the Proposed Sulfur Standards
    There are two considerations with respect to the feasibility of 
distributing non-highway diesel fuels meeting the proposed sulfur 
standards. The first pertains to whether sulfur contamination can be 
adequately managed throughout the distribution system so that fuel 
delivered to the end-user does not exceed the specified maximum sulfur 
concentration. The second pertains to the physical limitations of the 
system to accommodate any additional segregation of product grades.
a. Limiting Sulfur Contamination
    With respect to limiting sulfur contamination during distribution, 
the physical hardware and distribution practices for non-highway diesel 
fuel do not differ significantly from those for highway diesel fuel. 
Therefore, we do not anticipate any new issues with respect to limiting 
sulfur contamination during the distribution of non-highway fuel that 
would not have already been accounted for in distributing highway 
diesel fuel. Highway diesel fuel has been required to meet a 500 ppm 
sulfur standard since 1993. Thus, we expect that limiting contamination 
during the distribution of 500 ppm non-highway diesel engine fuel can 
be readily accomplished by industry.
    In the highway diesel rule, EPA acknowledged that meeting a 15 ppm 
sulfur specification would pose a substantial new challenge to the 
distribution system. Refiners, pipelines and terminals would have to 
pay careful attention to and eliminate any potential sources of 
contamination in the system (e.g., tank bottoms, deal legs in 
pipelines, leaking valves, interface cuts, etc.) In addition, bulk 
plant operators and delivery truck operators would have to carefully 
observe recommended industry practices to limit contamination, 
including practices as simple as cleaning out transfer hoses, proper 
sequencing of fuel deliveries, and parking on a level surface. Due to 
the need to prepare for compliance with the highway diesel program, we 
anticipate that issues related to limiting sulfur contamination during 
the distribution of 15 ppm nonroad diesel fuel will be resolved well in 
advance of the proposed 2010 implementation date for nonroad fuel. We 
are not aware of any additional issues that might be raised unique to 
nonroad fuel. If anything we anticipate limiting contamination will 
become easier as batch sizes are allowed to increase and potential 
sources of contamination decrease. We request comment on whether there 
are unique considerations regarding the transition to a 15 ppm standard 
for nonroad diesel fuel and what actions we should take beyond those 
that are already underway in preparation for the 15 ppm highway diesel 
program.
b. Potential Need for Additional Product Segregation
    As discussed in sub-section B, we have designed the proposed 
program to minimize the need for additional product segregation and the 
associated feasibility and cost issues associated with it. This 
proposal would allow for the fungible distribution of 500 ppm highway 
and 500 ppm NRLM diesel fuel in 2007, and 15 ppm highway and 15 ppm 
nonroad diesel fuel in 2010, up until the point where NRLM or nonroad 
fuel must be dyed for IRS excise tax purposes. Heating oil would be 
required to be segregated as a separate pool beginning in 2007 through 
the use of a new marker, and locomotive and marine fuel by use of the 
same marker beginning in 2010. With this program design, we believe we 
have eliminated any potential feasibility issues associated with the 
need for product segregation. This is not to say that steps will not 
have to be taken. We have

[[Page 28429]]

identified only a single instance where it seems likely that the 
adoption of this proposal would result in entities in the distribution 
system choosing to add new tankage due to new product segregation. Bulk 
plants in areas of the country where heating oil is expected to remain 
in the market will have to decide whether to add tankage to distribute 
both heating oil and 500 ppm NRLM fuel. In all other cases we 
anticipate segments of the distribution system will choose to avoid any 
fuel segregation costs by limiting the range of sulfur grades they 
choose to carry, just as they do today. Regardless, however, the costs 
and impacts of these choices are small. We request comment on this 
assessment. A more detailed explanation of this assessment can be found 
in Chapter 5.6 of the draft RIA.

G. What Are the Potential Impacts of the 15 ppm Sulfur Diesel Program 
on Lubricity and Other Fuel Properties?

1. What Is Lubricity and Why Might it Be a Concern?
    Engine manufacturers and owner/operators depend on diesel fuel 
lubricity properties to lubricate and protect moving parts within fuel 
pumps and injection systems for reliable performance. Unit injector 
systems and in-line pumps, commonly used in diesel engines, are 
actuated by cams lubricated with crankcase oil, and have minimal 
sensitivity to fuel lubricity. However, rotary and distributor type 
pumps, commonly used in light and medium-duty diesel engines, are 
completely fuel lubricated, resulting in high sensitivity to fuel 
lubricity. The types of fuel pumps and injection systems used in 
nonroad diesel engines are the same as those used in highway diesel 
vehicles. Consequently, nonroad and highway diesel engines share the 
same need for adequate fuel lubricity to maintain fuel pump and 
injection system durability.
    Diesel fuel lubricity concerns were first highlighted for private 
and commercial vehicles during the initial implementation of the 
Federal 500 ppm sulfur highway diesel program and the state of 
California's diesel program. The Department of Defense (DoD) also has a 
longstanding concern regarding the lubricity of distillate fuels used 
in its equipment as evidenced by the implementation of its own fuel 
lubricity improver performance specification in 1989.\263\ The diesel 
fuel requirements in the state of California differed from the federal 
requirements by substantially restricting the content of diesel fuel 
requires more severe hydrotreating than reducing the sulfur content to 
meet a 500 ppm standard.\264\ Consequently, concerns regarding diesel 
fuel lubricity have primarily been associated with California diesel 
fuel and some California refiners treat their diesel fuel with a 
lubricity additive as needed. Outside of California, hydrotreating to 
meet the current 500 ppm sulfur specification does not typically result 
in a substantial reduction of lubricity. Diesel fuels outside of 
California seldom require the use of a lubricity additive. Therefore, 
we anticipate only a marginal increase in the use of lubricity 
additives in NRLM diesel fuel meeting the proposed 500 ppm sulfur 
standard for 2007.\265\ This proposal would require diesel fuel used in 
nonroad engines to meet a 15 ppm sulfur standard in 2010. Based on the 
following discussion, we believe that the increase in the use of 
lubricity additives in 15 ppm nonroad diesel fuel would be the same as 
that estimated for 15 ppm highway diesel fuel.
---------------------------------------------------------------------------

    \263\ DoD Performance Specification, Inhibitor, Corrosion/
Lubricity Improver, Fuel Soluble, , MIL-PRF-25017F, 10 November 
1997, Superseding MIL-I-25017E, 15 June 1989.
    \264\ Chevron Products Diesel Fuel Technical Review provides a 
discussion of the impacts on fuel lubricity of current diesel fuel 
compositional requirements in California versus the rest of the 
nation. http://www.chevron.com/prodserv/fuels/bulletin/diesel/
l2%5F7%5F2%5Frf.htm.
    \265\ The cost from the increased use of lubricity additives in 
500 ppm NRLM diesel fuel in 2007 and in 15 ppm nonroad diesel fuel 
in 2010 is discussed in section V of today's preamble.
---------------------------------------------------------------------------

    The state of California currently requires the same standards for 
diesel fuel used in nonroad equipment as in highway equipment. Outside 
of California, highway diesel fuel is often used in nonroad equipment 
when logistical constraints or market influences in the fuel 
distribution system limit the availability of high sulfur fuel. Thus, 
for nearly a decade nonroad equipment has been using federal 500 ppm 
sulfur diesel fuel and California diesel fuel, some of which may have 
been treated with lubricity additives. During this time, there has been 
no indication that the level of diesel lubricity needed for fuel used 
in nonroad engines differs substantially from the level needed for fuel 
used in highway diesel engines.
    Blending small amounts of lubricity-enhancing additives increases 
the lubricity of poor-lubricity fuels to acceptable levels. These 
additives are available in today's market, are effective, and are in 
widespread use around the world. Among the available additives, 
biodiesel has been suggested as one potential means for increasing the 
lubricity of conventional diesel fuel. Indications are that low 
concentrations of biodiesel would be sufficient to raise the lubricity 
to acceptable levels.
    Considerable research remains to be performed to better understand 
which fuel components are most responsible for lubricity. Consequently, 
it is unclear whether and to what degree the proposed sulfur standards 
for non-highway diesel engine fuel will impact fuel lubricity. 
Nevertheless, there is evidence that the typical process used to remove 
sulfur from diesel fuel--hydrotreating--can impact lubricity depending 
on the severity of the treatment process and characteristics of the 
crude. We expect that hydrotreating will be the predominant process 
used to reduce the sulfur content of non-highway diesel engine fuel to 
meet the 500 ppm sulfur standard during the first step of the proposed 
program. The highway diesel program projected that hydrotreating would 
be the process most frequently used to meet the 15 ppm sulfur standard 
for highway diesel fuel. The 2010 implementation date for the proposed 
15 ppm standard for nonroad diesel fuel would allow the use of new 
technologies to remove sulfur from fuel.\266\ These new technologies 
have less of a tendency to affect other fuel properties than does 
hydrotreating.
---------------------------------------------------------------------------

    \266\ See section IV.F for a discussion of which desulfurization 
processes we expect will be used to meet the 15 ppm standard for 
nonroad diesel fuel.
---------------------------------------------------------------------------

    Based on our comparison of the blendstocks and processes used to 
manufacture non-highway diesel fuels, we believe that the potential 
decrease in the lubricity of these fuels from hydrotreating that might 
result from the proposed sulfur standards should be approximately the 
same as that experienced in desulfurizing highway diesel fuel.\267\ To 
provide a conservative, high cost estimate, we assumed that the 
potential impact on fuel lubricity from the use of the new 
desulfurization processes would be the same as that experienced when 
hydrotreating diesel fuel to meet a 15 ppm sulfur standard. We request 
comment on the potential impact of these new desulfurization 
technologies on lubricity (as well as other fuel properties) that might 
help us to improve our estimate of the potential impacts of this 
proposal on fuel properties other than sulfur. Given that the 
requirements for fuel lubricity in highway and non-highway engines are 
the same, and the potential decrease in lubricity from desulfurization 
of non-highway diesel engine would be no greater than that experienced 
in desulfurizing highway diesel fuel, we

[[Page 28430]]

estimate that the potential need for lubricity additives in non-highway 
diesel engine fuel under this proposal would be the same as that for 
highway diesel fuel meeting the same sulfur standard.
---------------------------------------------------------------------------

    \267\ See chapter 5 of the RIA for a discussion of the potential 
impacts on fuel lubricity of this proposal.
---------------------------------------------------------------------------

2. A Voluntary Approach on Lubricity
    In the United States, there is no government or industry standard 
for diesel fuel lubricity. Therefore, specifications for lubricity are 
determined by the market. Since the beginning of the 500 ppm sulfur 
highway diesel program in 1993, refiners, engine manufacturers, engine 
component manufacturers, and the military have been working with the 
American Society for Testing and Materials (ASTM) to develop protocols 
and standards for diesel fuel lubricity in its D-975 specifications for 
diesel fuel. ASTM is working towards a single lubricity specification 
that would be applicable to all diesel fuel used in any type of engine. 
Although ASTM has not yet adopted specific protocols and standards, 
refiners that supply the U.S. market have been treating diesel fuel 
with lubricity additives on a batch to batch basis, when poor lubricity 
fuel is expected. Other examples include the U.S. military, Sweden, and 
Canada. The U.S. military has found that the traditional corrosion 
inhibitor additives used in its fuels have been highly effective in 
reducing fuel system component wear. Since 1991, the use of lubricity 
additives in Sweden's 10 ppm sulfur Class I fuel and 50 ppm sulfur 
Class II fuel has resulted in acceptable equipment durability.\268\ 
Since 1997, Canada has required that its 500 ppm sulfur diesel fuel not 
meeting a minimum lubricity be treated with lubricity additives.
---------------------------------------------------------------------------

    \268\ Letter from L. Erlandsson, MTC AB, to Michael P. Walsh, 
dated October 16, 2000. EPA air docket A-99-06, docket item IV-G-42.
---------------------------------------------------------------------------

    The potential need for lubricity additives in diesel fuel meeting a 
15 ppm sulfur specification was evaluated during the development of 
EPA's highway diesel rule. In response to the proposed highway diesel 
rule, all comments submitted regarding lubricity either stated or 
implied that the proposed sulfur standard of 15 ppm would likely cause 
the refined fuel to have lubricity characteristics that would be 
inadequate to protect fuel injection equipment, and that mitigation 
measures such as lubricity additives would be necessary. However, the 
commenters suggested varied approaches for addressing lubricity. For 
example, some suggested that we need to establish a lubricity 
requirement by regulation while others suggested that the current 
voluntary, market based system would be adequate. The Department of 
Defense recommended that we encourage the industry (ASTM) to adopt 
lubricity protocols and standards before the 2006 implementation date 
of the 15 ppm sulfur standard for highway diesel fuel.
    The final highway diesel rule did not establish a lubricity 
standard for highway diesel fuel. We believe the issues related to the 
need for diesel lubricity in fuel used in non-highway diesel engines 
are substantially the same as those related to the need for diesel 
lubricity for highway engines. Consequently, we expect the same 
industry-based voluntary approach to ensuring adequate lubricity in 
non-highway diesel fuels that we recognized for highway diesel fuel. We 
believe the best approach is to allow the market to address the 
lubricity issue in the most economical manner, while avoiding an 
additional regulatory scheme. A voluntary approach should provide 
adequate customer protection from engine failures due to low lubricity, 
while providing the maximum flexibility for the industry. This approach 
would be a continuation of current industry practices for diesel fuel 
produced to meet the current federal and California 500 ppm sulfur 
highway diesel fuel specifications, and benefits from the considerable 
experience gained since 1993. It would also include any new 
specifications and test procedures that we expect would be adopted by 
the American Society for Testing and Materials (ASTM) regarding 
lubricity of NRLM diesel fuel quality.
    Regardless, this is an issue that will be resolved to meet the 
demands of the highway diesel market, and whatever resolution is 
reached for highway diesel fuel could be applied to non-highway diesel 
engine fuel with sufficient advance notice. We are continuing to 
participate in the ASTM Diesel Fuel Lubricity Task Force \269\ and will 
assist their efforts to finalize a lubricity standard in whatever means 
possible. We are hopeful that ASTM can reach a consensus early this 
summer at the next meeting of the ASTM's Lubricity Task Force. We 
request comment on what actio ns EPA should take to ensure adequate 
lubricity of non-highway diesel engine fuel beyond those already 
underway for highway diesel fuel.
---------------------------------------------------------------------------

    \269\ ASTM sub committee D02.E0.
---------------------------------------------------------------------------

3. What Other Impact Would Today's Actions Have on the Performance of 
Diesel and Other Fuels?
    We do not expect that the proposed fuel program would have any 
negative impacts on the performance of diesel engines in the existing 
fleet which would use the fuels regulated today. In the early 1990's, 
California lowered the maximum allowable level of sulfur content of 
highway and nonroad diesel fuel to 500 ppm, and at the same time 
California significantly lowered the aromatic content of diesel fuel. 
California required a cap on total aromatics of 10 percent by volume, 
while the in-use average at the time was on the order of 35 percent. 
The lowering of the total aromatic content resulted in some problems 
with leaks from the fuel pump O-ring seals in some diesel engines due 
to a change specifically in the polynuclear aromatics content (PNA). In 
the process of meeting California's 10 percent total aromatic content 
requirement, the end result typically lowered PNA's from approximately 
10-15 percent by volume to near-zero. In the early 1990's, some diesel 
engine manufacturers used a certain material (Nitrile) for O-rings in 
diesel fuel pumps. The Nitrile seals were found to be susceptible to 
leakage with the use of diesel fuel with very low PNA content. 
Normally, the PNA in the fuel penetrated the Nitrile material and cause 
it to swell, thereby providing a seal with the throttle shaft. When 
very low PNA fuel is used after conventional fuel has been used, the 
PNA already in the swelled O-ring would leach out into the very low PNA 
fuel. Subsequently, the Nitrile O-ring would shrink and pull away, thus 
causing leaks, or the stress on the O-ring during the leaching process 
would cause it to crack and leak. Not all 500 ppm sulfur fuels caused 
this problem, because the amount and type of aromatics varied, and the 
in-use seal problems were focused in California due to the 10 percent 
aromatic requirements and the resulting very low PNA content. This was 
not a wide-spread issue for the rest of the U.S. where highway diesel 
fuel also had a 500ppm sulfur cap because the federal requirements did 
not include a lower aromatic cap. While the process of lowering sulfur 
levels to 500ppm does lower PNA, it does not achieve the near-zero 
levels seen in California. Since the 1990's, diesel engine 
manufacturers have switched to alternative materials (such as Viton), 
which do not experience leakage. We believe that no issues with leaking 
fuel pump O-rings would occur with the changes in diesel fuel sulfur 
levels

[[Page 28431]]

contained in this proposal (both the 500 ppm requirement in 2008 and 
the 15 ppm requirement in 2010) because while we do believe PNA content 
will be reduced, we are not predicting it will achieve the near-zero 
level experienced in California.
    We expect that this proposal would have no negative impacts on 
other fuels, such as jet fuel or heating oil. We do expect that the 
sulfur levels of heating oil would decrease because of this proposal. 
Beginning in mid-2007, we expect that controlling NRLM diesel fuel to 
500 ppm would lead many pipelines to discontinue carrying high sulfur 
heating oil as a separate grade. In areas served by these pipelines, 
heating oil users would likely switch to 500 ppm diesel fuel. This 
would reduce emissions of sulfur dioxide and sulfate PM from furnaces 
and boilers fueled with heating oil. The primary exception to this 
would likely be the Northeast and some areas of the Pacific Northwest, 
where a distinct higher sulfur heating oil would still be distributed 
as a separate fuel. Also, we expect that a small volume of high sulfur 
distillate fuel would be created during distribution from the mixing of 
low sulfur diesel fuels and higher sulfur fuels, such as jet fuel in 
the pipeline interface. Such high sulfur distillate would likely be 
sold by the terminal as high sulfur heating oil or reprocessed by 
transmix processors.

H. Refinery Air Permitting

    Prior to making diesel desulfurization changes, some refineries may 
be required to obtain a preconstruction permit, under the New Source 
Review (NSR) program, from the applicable state/local air pollution 
control agency.\270\ We believe that the proposed program provides 
sufficient lead time for refiners to obtain any necessary NSR permits 
well in advance of the compliance date.
---------------------------------------------------------------------------

    \270\ Hydrotreating diesel fuel involves the use of process 
heaters, which have the potential to emit pollutants associated with 
combustion, such as NOX, PM, CO and SO3. In 
addition, reconfiguring refinery processes to add desulfurization 
equipment could increase fugitive VOC emissions. The emissions 
increases associated with diesel desulfurization would vary widely 
from refinery to refinery, depending on many source-specific 
factors, such as crude oil supply, refinery configuration, type of 
desulfurization technology, amount of diesel fuel produced, and type 
of fuel used to fire the process heaters.
---------------------------------------------------------------------------

    Given that today's diesel sulfur program would provide roughly 
three years of lead time before the 500 ppm standard would take effect, 
we believe refiners would have time to obtain any necessary 
preconstruction permits. Nevertheless, we believe it is reasonable to 
continue our efforts under the Tier 2 and highway diesel fuel programs, 
to help states in facilitating the issuance of permits under the NRLM 
diesel sulfur program. For example, the guidance on Best Available 
Control Technology (BACT) and Lowest Achievable Emission Rate (LAER) 
control technology that was developed for the gasoline sulfur program 
should have application for diesel desulfurization (highway and NRLM) 
projects as well. Similarly, we believe the concept of EPA permit teams 
for gasoline sulfur projects could readily be extended to permits 
related to diesel projects as well. These teams, as needed, would track 
the overall progress of permit issuance and would be available to 
assist state/local permitting authorities, refineries and the public 
upon request to resolve site-specific permitting questions. In 
addition, these teams would be available, as necessary, to assist in 
resolving case specific issues to ensure timely issuance of permits. 
Finally, to facilitate the processing of permits, we encourage 
refineries to begin discussions with permitting agencies and to submit 
permit applications as early as possible.

V. Program Costs and Benefits

    In this section, we present the projected cost impacts and cost 
effectiveness of the proposed nonroad Tier 4 emission standards and 
low-sulfur fuel requirement. We also present a benefit-cost analysis 
and an economic impact analysis. The benefit-cost analysis explores the 
net yearly economic benefits to society of the reduction in mobile 
source emissions likely to be achieved by this rulemaking. The economic 
impact analysis explores how the costs of the rule will likely be 
shared across the manufacturers and users of the engines, equipment and 
fuel that would be affected by the standards.
    The results detailed below show that this rule would be highly 
beneficial to society, with net present value benefits through 2030 of 
$550 billion, compared to a net present value of social cost of only 
about $16.5 billion (net present values in the year 2004). The impact 
of these costs on society should be minimal, with the prices of goods 
and services produced using equipment and fuel affected by the proposal 
being expected to increase about 0.02 percent.
    Further information on these and other aspects of the economic 
impacts of our proposal are summarized in the following sections and 
are presented in more detail in the Draft RIA for this rulemaking. We 
invite the reader to comment on all aspects of these analyses, 
including our methodology and the assumptions and data that underlie 
our analysis.

A. Refining and Distribution Costs

    As described above, the fuel-related requirements associated with 
this proposed rule would be implemented in two steps. Nonroad, 
locomotive and marine diesel fuel would be subject to a 500 ppm sulfur 
cap beginning June 1, 2007, while nonroad diesel fuel would be subject 
to a 15 ppm sulfur cap beginning June 1, 2010. Meeting these standards 
would generally require refiners adding hydrotreating equipment and 
possibly new or expanded hydrogen and sulfur plants in their refineries 
for desulfurizing their nonroad diesel fuel and dispensing of the 
removed sulfur. Using information provided by vendors of 
desulfurization equipment and through discussions with distributors of 
nonroad diesel fuel, we estimated the desulfurization and associated 
distribution and additive cost for complying with this two step 
desulfurization program. Except for the costs presented at the end of 
this section, the costs below reflect a fully phased in fuels program 
without the proposed small refiner exemption. Costs are in 2002 
dollars. We request comment on the cost estimates presented below and 
the methodologies used to develop them. You can refer to the Draft RIA 
for details.
    The cost to provide nonroad, locomotive and marine diesel fuel 
under the proposed fuel program is summarized in Table V-A-1 below. The 
costs shown (and all of the costs described in the rest of this 
section) only apply to the roughly 65 percent of current nonroad, 
locomotive and marine diesel fuel that contains more than 500 ppm 
sulfur (hereafter referred to as the affected volume). We estimate that 
the other 35 percent of this fuel is actually fuel certified to the 
highway diesel fuel standards and project that this will continue. 
Thus, the proposed fuel program would not affect this fuel and no 
additional costs would be incurred by its refiners or distributors. The 
costs and benefits of desulfurizing this highway fuel which spills over 
into the non-highway markets was already included in EPA's 2007 highway 
diesel fuel rule.

[[Page 28432]]



              Table V-A-1.--Increased Cost of Providing Nonroad, Locomotive and Marine Diesel Fuel
----------------------------------------------------------------------------------------------------------------
                                                         Cents per gallon of affected fuel         Affected fuel
                                                 ------------------------------------------------     volume
                                                                                                     (million
                                                     Refining      Lubricity and       Total       gallons/year)
                                                                   distribution                          a
----------------------------------------------------------------------------------------------------------------
Step One--500 ppm NRLM diesel fuel..............             2.2             0.3             2.5           9,504
Step Two--5 ppm Nonroad diesel fuel.............             4.4             0.4             4.8           7,803
Step Two--500 ppm Locomotive and Marine diesel               2.2           b 0.2             2.4          4,093
 fuel...........................................
----------------------------------------------------------------------------------------------------------------
Notes:
a 2008 for Step One (without consideration of small refiner provisions), 2015 for Step Two.
b 0.4 cent per gallon from mid-2010 to mid-2014 due to need for marker.

    The majority of the fuel-related cost of the proposal is refining-
related. These costs include required capital investments amortized at 
7 percent per annum before taxes. The derivation of these costs is 
discussed in more detail below and in the Draft RIA. We request comment 
on the estimated cost of meeting the 15 ppm and 500 ppm sulfur caps.
    We also project that the increased cost of refining and 
distributing 15 ppm and 500 ppm fuel would be substantially offset by 
reductions in maintenance costs. These savings would apply to all 
diesel engines in the field, not just new engines. Refer to section V. 
B for a more complete discussion on the projected maintenance savings 
associated with lower sulfur fuels.
1. Refining Costs
    Our process for estimating the refining costs associated with the 
proposed fuel program consisted of four steps. One, we estimated the 
volume of 500 and 15 ppm nonroad, locomotive and marine diesel fuel 
which had to be produced in each PADD \271\ in each phase of the 
program. This step utilized diesel fuel and heating oil use estimates 
from the Energy Information Administration's (EIA) Fuel Oil and 
Kerosene Survey for 2000, shipments of diesel fuel between PADDs, 
projected loss of 15 and 500 ppm volume due to contamination during 
distribution and small refiner provisions. This nonroad diesel fuel 
consumption in 2000 is lower than that inherent in the emission 
estimates described above, which are based directly on the results of 
EPA's NONROAD emission model. We are investigating ways to make the two 
estimates more consistent.
---------------------------------------------------------------------------

    \271\ Petroleum Administrative for Defense Districts.
---------------------------------------------------------------------------

    Growth in distillate fuel use off this year 2000 base was estimated 
using projections from EIA's Annual Energy Outlook, with one exception. 
This exception was that the growth in nonroad diesel fuel use was taken 
from EPA's NONROAD emission model (roughly three percent per year), as 
opposed to EIA's projected growth of roughly one percent per year. The 
higher growth rate is consistent with that inherent in the emission 
estimates described above.
    Refinery production of low and high sulfur distillate fuel in the 
year 2000 was based on actual reports provided to EIA by all U.S. 
refiners and importers. Refinery production of low and high sulfur 
distillate fuel was assumed to grow at the same rate as consumption of 
the two types of fuel, respectively. These rates were roughly three 
percent and one and a half percent for low and high sulfur distillate 
fuel production, respectively. The specific volumes of highway, 
nonroad, locomotive, and marine diesel fuel by calendar year are 
presented in chapter 7 of the Draft RIA.
    Two, we estimated the cost for each refinery to desulfurize its 
high sulfur fuel to 500 and 15 ppm. This was based on their historical 
production volume of high sulfur diesel fuel and estimates of the 
composition of this fuel (straight run, light cycle oil, etc.).\272\ We 
also considered whether these refineries would be modifying or building 
hydrotreating capacity in order to meet the 15 ppm highway cap.
---------------------------------------------------------------------------

    \272\ The composition of nonroad diesel fuel in each PADD was 
based on a survey conducted by API and NPRA in 1996. Crude oils 
processed by domestic refiners have been becoming heavier over time, 
necessitating greater use of coking and hydrocracking to convert the 
heavy material into lighter, saleable products. Thus, the 
contributions of coker and hydrocracked distillate to the overall 
distillate pool are rising. Coker distillate is somewhat more 
difficult to desulfurize than average distillate, but hydrocracked 
distillate is much easier to desulfurize. Overall, this trend could 
increase projected desulfurization costs slightly. We plan to update 
these compositions to reflect trends in crude oil quality and 
refinery configuration in our analysis for the final rule to the 
extent that more recent data allow.
---------------------------------------------------------------------------

    Three, we estimated which refineries would find it difficult to 
market all of their current high sulfur diesel fuel as heating oil, due 
to their location relative to major pipelines and the size of the 
heating oil market in their area. Those not located in major heating 
oil markets and not connected to pipelines serving these areas were 
projected to have to meet the 500 ppm cap in 2007.
    Four, we determined the additional refineries which would produce 
500 ppm and 15 ppm fuel to satisfy demand during each phase of the fuel 
program. Refineries projected to have the lowest compliance costs in 
each PADD were projected to produce the lower sulfur fuels until demand 
was met. PADD 3 refineries were allowed to ship low sulfur fuel to the 
Northeast, but no other inter-PADD transfers were assumed. Imports of 
500 ppm highway diesel fuel were assumed to increase at the rate of 
highway diesel fuel consumption and be converted to 15 ppm diesel fuel, 
80 percent in 2006 and 100 percent in 2010. Imports of high sulfur 
distillate fuel were assumed to increase at the rate of high sulfur 
distillate fuel consumption, but were assumed to remain entirely high 
sulfur heating oil even after today's NRLM fuel proposal. In other 
words, all 15 ppm and 500 ppm NRLM fuel produced under this proposal 
was assumed to be produced by domestic refineries. This assumption 
increased the projected costs of the proposal described above more than 
would have been the case had we assumed that domestic production and 
imports of high sulfur distillate fuel would each keep their respective 
shares of the NRLM diesel fuel and heating oil markets in response to 
this proposal. The relative costs of producing 15 ppm nonroad diesel 
fuel by domestic and overseas refiners is discussed further in section 
V.A.6. below.
    With the onset of a 2007 500 ppm sulfur cap for nonroad, locomotive 
and marine diesel fuel, we project that the market for high sulfur 
diesel fuel and heating oil would become so small that high sulfur fuel 
would no longer be shipped through common carrier pipelines in most 
areas. The prime exception to this would be the Northeast, where the 
heating oil market is very large. Thus, refiners located in the 
Northeast and those along the major pipelines serving the Northeast, 
namely the Colonial and Plantation pipelines, could continue to produce 
high sulfur

[[Page 28433]]

heating oil. Other refineries would shift the production of high sulfur 
diesel fuel and heating oil to the 500 ppm NRLM market. The second 
exception would be refiners granted special provisions due to the small 
size of their business (i.e., SBREFA refiners) or economic hardship, as 
discussed in section IV above. The high sulfur distillate production 
levels of these refineries is small enough that they can sell into more 
local nonroad, locomotive and marine markets or the heating oil market 
without using pipelines and so they could continue to produce high 
sulfur distillate.
    Based on refinery distillate production data from the Energy 
Information Administration (EIA), there are 122 refineries currently 
producing highway diesel fuel and 105 refineries producing high sulfur 
diesel fuel or heating oil. Using the methodology described above, 
absent this proposal, we project that roughly 114 refineries will 
invest in additional desulfurization equipment to produce 15 ppm 
highway diesel fuel; 74 refineries in 2006 and 40 in 2010.\273\ These 
114 refineries include 109 of the 122 refineries which currently 
produce highway diesel fuel, plus 5 refineries which currently only 
produce high sulfur distillate fuel today. Again absent the proposed 
NRLM diesel fuel program, we project that roughly 13 refineries 
currently producing highway diesel fuel will shift to producing high 
sulfur distillate fuel. This would leave a total of 113 refineries 
still producing high sulfur distillate after full implementation of the 
2007 highway diesel fuel program.
---------------------------------------------------------------------------

    \273\ These (and the subsequent) estimates of the number of 
refineries investing in new equipment to produce diesel fuels of 
various sulfur levels should be understood as rough estimates which 
assist us in projecting costs and other impacts related to this 
proposal. They are most reasonable when evaluating the total number 
of refineries investing in a particular year or region. We are not 
indicating that we believe that we can predict which specific 
refineries would invest in desulfurization equipment in response to 
this proposal.
---------------------------------------------------------------------------

    The number of these 113 domestic refineries expected to produce 
either 500 ppm of 15 ppm NRLM diesel fuel in response to this proposal 
is summarized in Table V-A-2.

                Table V-A-2 Refineries Projected to Produce NRLM Diesel Fuel Under This Proposal
----------------------------------------------------------------------------------------------------------------
                                                        500 ppm diesel fuel             15 ppm diesel fuel
                                                 ---------------------------------------------------------------
                 Year of Program                                       Small                           Small
                                                  All refineries    refineries    All refineries    refineries
----------------------------------------------------------------------------------------------------------------
2007-2010.......................................              42               0               0               0
2010-2014.......................................              37              19              25               0
2014+...........................................              25              12              37               7
----------------------------------------------------------------------------------------------------------------

    As shown in this table, we project that 42 of the 113 refineries 
currently producing some high sulfur distillate would desulfurize their 
high sulfur diesel fuel in response to the proposed 500 ppm standard in 
2007. The remainder would continue producing either high sulfur NRLM 
diesel fuel under the proposed small refiner provisions, or high sulfur 
heating oil. As explained in section IV.F, we project that these 
refiners would use conventional hydrotreating technology to meet this 
standard. Of these 42 refineries, we project that 32 would build new 
hydrotreaters to meet the 500 ppm sulfur cap. We project that three of 
the remaining ten refineries would be able to meet the 500 ppm cap with 
their existing hydrotreater which is currently being used to produce 
highway diesel fuel. These three refineries are projected to build a 
new hydrotreater to produce 15 ppm highway diesel fuel in 2006, so 
their existing highway fuel hydrotreater could process their current 
high sulfur diesel fuel. The remaining seven refineries currently 
produce relatively small amounts of high sulfur diesel fuel compared to 
their highway diesel fuel production. We project that these refiners 
would be able to economically revamp their existing highway 
hydrotreater to process their non-highway diesel fuel.
    We project that the capital cost involved to meet the 2007 500 ppm 
sulfur cap would be $600 million, or $9.7 million per refinery building 
a new hydrotreater. The bulk of this capital would be invested in 2007 
($500 million), with the remainder being invested in 2010.\274\ 
Operating costs would be about $3 million per year for the average 
refinery. We request comment on the number of refiners who would need 
to build new equipment to meet the 500 ppm sulfur cap, the capital cost 
for this new equipment and the cost of operating this equipment.
---------------------------------------------------------------------------

    \274\ Some refineries would be able to delay production of 500 
ppm NRLM fuel until 2010 due to the proposed small refiner 
provisions. Likewise, some refineries would be able to delay 
production of 15 ppm nonroad diesel fuel until 2014.
---------------------------------------------------------------------------

    Starting in mid-2010, we project that 25 refineries would add or 
revamp equipment to meet the 15 ppm cap on nonroad diesel fuel, while 
20 refineries (nearly all of them small refiners) would add or revamp 
equipment to produce 500 ppm nonroad or locomotive and marine diesel 
fuel. Finally, an additional 12 refineries (again nearly all of them 
small refiners) would begin producing 15 ppm nonroad diesel fuel in 
2014.
    We project that 80 percent of the 15 ppm nonroad diesel fuel volume 
would be desulfurized by advanced technologies, while the remaining 20 
percent would be desulfurized by conventional hydrotreaters. Since the 
bulk of the hydrotreating capacity being used to meet the 2007 500 ppm 
standard for NRLM diesel fuel would have just been built in 2007 or 
2010, we expect that it would have been designed to facilitate further 
processing to 15 ppm sulfur and the added 15 ppm facilities would be 
revamps. However, those refiners who used their existing highway diesel 
fuel hydrotreaters to meet the proposed 500 ppm cap in 2007 would 
likely have to construct new equipment in 2010 or 2014 to meet the 15 
ppm cap on nonroad diesel fuel, since these hydrotreaters could not be 
revamped in 2006 to produce 15 ppm highway diesel fuel. When the 
proposed NRLM diesel fuel program would be fully implemented in 2014, 
roughly 51 refineries are still projected to produce high sulfur 
heating oil and thus, would not face any refining costs related to this 
proposal.
    Our projection that 80 percent of refineries would utilize some 
form of advanced technology to meet the proposed 15 ppm nonroad fuel 
sulfur cap is based on the fact that this 15 ppm cap would follow the 
production of 15 ppm highway diesel fuel by four years. Several firms 
are expending significant research and development resources to bring 
such advanced technologies to the market for the highway diesel fuel

[[Page 28434]]

program. We developed cost estimates for two such technologies: Linde 
Iso-Therming and Phillips S-Zorb. The development of cost estimates for 
these two advanced technologies, as well as conventional hydrotreating, 
is described in detail in Chapter 7 of the Draft RIA. We request 
comment on the potential viability and cost savings associated with 
advanced desulfurization technologies, particularly in the 2010 
timeframe.
    The total capital cost of new equipment and revamps related to the 
proposed 2010 sulfur standard would be $640 million, or $17 million per 
refinery adding or revamping equipment. Total operating costs would be 
about $5 million per year for the average refinery. The total refining 
cost, including the amortized cost of capital, would be 4.4 cents per 
gallon of new 15 ppm nonroad fuel. This cost is relative to the cost of 
producing high sulfur fuel today, and includes the cost of meeting the 
500 ppm standard beginning in 2007. We request comment on the number of 
refiners who would need to build new equipment to meet the 15 ppm 
sulfur cap, the capital cost for this new equipment and the cost of 
operating this equipment. The average cost of continuing to meet the 
500 ppm standard for locomotive and marine fuel would continue at 2.2 
cents per gallon.
    The above costs reflect national averages for the fully phased in 
program for each control step. Some refiners would face lower costs 
while others would face higher costs. Excluding small refiners because 
they are able to take advantage of the proposed small refiner 
provisions, the average refining costs by refining region are shown in 
the table below. Combined costs are shown for PADDs 1 and 3 because of 
the large volume of diesel fuel which is shipped from PADD 3 to PADD 1.

    Table V-A-3.--Average Refining Costs by Region (cents per gallon)
------------------------------------------------------------------------
                                2007 500 ppm Cap       2010 15 ppm Cap
------------------------------------------------------------------------
PADDs 1 and 3...............                   1.4                   2.6
PADD 2......................                   2.9                   5.7
PADD 4......................                   4.0                   8.5
PADD 5......................                   2.6                   5.4
Nationwide..................                   2.2                   4.4
------------------------------------------------------------------------

    We request comment on the range of estimated refining costs for the 
various regions for both the proposed 500 and 15 ppm sulfur caps.
2. Cost of Lubricity Additives
    Hydrotreating diesel fuel tends to reduce the natural lubricating 
quality of diesel fuel, which is necessary for the proper functioning 
of certain fuel system components. There are a variety of fuel 
additives which can be used to restore diesel fuel's lubricating 
quality. These additives are currently used to some extent in highway 
diesel fuel. We expect that the need for lubricity additives that would 
result from the proposed 500 ppm sulfur standard for off-highway diesel 
engine fuel would be similar to that for highway diesel fuel meeting 
the current 500 ppm sulfur cap standard.\275\ Industry experience 
indicates that the vast majority of highway diesel fuel meeting the 
current 500 ppm sulfur cap does not need lubricity additives. 
Therefore, we expect that the great majority of off-highway diesel 
engine fuel meeting the proposed 500 ppm sulfur standard would also not 
need lubricity additives. In estimating lubricity additive costs for 
500 ppm diesel fuel, we assumed that fuel suppliers would use the same 
additives at the same concentration as we projected would be used in 15 
ppm highway diesel fuel. Based on our analysis of this issue for the 
2007 highway diesel fuel program, the cost per gallon of the lubricity 
additive is about 0.2 cent. This level of use is likely conservative, 
as the amount of lubricity additive needed increases substantially as 
diesel fuel is desulfurized to lower levels. We also project that only 
5 percent of all 500 ppm NRLM diesel fuel would require the use of a 
lubricity additive. Thus, we project that the cost of additional 
lubricity additives for the affected 500 ppm NRLM diesel fuel would be 
0.01 cent per gallon. See the Draft RIA for more details on the issue 
of lubricity additives.
---------------------------------------------------------------------------

    \275\ Please refer to section IV in today's preamble for 
additional discussion regarding our projections of the potential 
impact on fuel lubricity of this proposed rule.
---------------------------------------------------------------------------

    We project that all nonroad diesel fuel meeting a 15 ppm cap would 
require treatment with lubricity additives. Thus, the projected cost 
would be 0.2 cent per affected gallon of 15 ppm nonroad diesel fuel.
3. Distribution Costs
    The proposed fuel program is projected to impact distribution costs 
in three ways. One, we project that more diesel fuel would have to be 
distributed under the proposal than without it. This is due to the fact 
that some of the desulfurization processes reduce the fuel's volumetric 
energy density during processing. Total energy is not lost during 
processing, as the total volume of fuel is increased. However, a 
greater volume of fuel must be consumed in the engine to produce the 
same amount of power. We assumed that the current cost of distributing 
diesel fuel of 10 cents per gallon (see Draft RIA for further details) 
would stay constant (i.e., a 1 percent increase in the amount of fuel 
distributed would increase total distribution costs by 1 percent).
    We project that desulfurizing diesel fuel to 500 ppm would reduce 
volumetric energy content by 0.7 percent. This would increase the cost 
of distributing fuel by 0.07 cent per gallon. We project that 
desulfurizing diesel fuel to 15 ppm would reduce volumetric energy 
content by an additional 0.35 percent. This would increase the cost of 
distributing fuel by an additional 0.04 cent per gallon, or a total 
cost of 0.11 cent per gallon of affected 15 ppm nonroad diesel fuel.
    Two, while this proposal minimizes the segregation of similar 
fuels, some additional segregation of products in the distribution 
system would still be required. The proposed allowance that highway and 
off-highway diesel engine fuel meeting the same sulfur specification 
can be shipped fungibly until it leaves the terminal obviates the need 
for additional storage tankage in this segment of the distribution 
system.\276\ This proposal would also allow 500 ppm NRLM diesel fuel to 
be mixed with high-sulfur NRLM diesel fuel once the fuels are dyed to 
meet IRS requirements. This provision would ease the last part of the 
distribution of high-sulfur NRLM diesel fuel.
---------------------------------------------------------------------------

    \276\ Including the refinery, pipeline, marine tanker, and barge 
segments of the distribution system.
---------------------------------------------------------------------------

    However, we expect that the implementation of the proposed 500 ppm 
standard for NRLM diesel fuel in 2007 would compel some bulk plants in 
those parts of the country still

[[Page 28435]]

distributing heating oil as a separate fuel grade to install a second 
diesel storage tank to handle this 500 ppm nonroad fuel. These bulk 
plants currently handle only high-sulfur fuel and hence would need a 
second tank to continue their current practice of selling fuel into the 
heating oil market in the winter and into the nonroad market in the 
summer.\277\ We believe that some of these bulk plants would convert 
their existing diesel tank to 500 ppm fuel in order to avoid the 
expense of installing an additional tank. However, to provide a 
conservatively high estimate we assumed that 10 percent of the 
approximately 10,000 bulk plants in the U.S. (1,000) would install a 
second tank in order to handle both 500 ppm NRLM diesel fuel and 
heating oil. The cost of an additional storage tank at a bulk plant is 
estimated at $90,000 and the cost of de-manifolding their delivery 
truck at $10,000.\278\ If all 1,000 bulk plants were to install a new 
tank, the total one-time capitol cost would be $100,000,000. Amortizing 
the capital costs over 20 years, results in a estimated cost for 
tankage at such bulk plants of 0.1 cent per gallon of affected NRLM 
diesel fuel supplied. Although the impact on the overall cost of the 
proposed program is small, the cost to those bulk plant operators who 
need to put in a separate storage tank may represent a substantial 
investment. Thus, as discussed in section IV.F., we believe many of 
these bulk plants could make other arrangements to continue servicing 
both heating oil and NRLM markets.
---------------------------------------------------------------------------

    \277\ See section IV.E.9. of this proposal and chapter 5 of the 
RIA for additional discussion of the potential impacts of the 
proposed sulfur standards on the distribution system.
    \278\ This estimated cost includes the addition of a separate 
delivery system on the tank truck.
---------------------------------------------------------------------------

    Due to the end of the highway program temporary compliance option 
(TCO) in 2010 and the disappearance of high-sulfur diesel fuel from 
much of the fuel distribution system due to the implementation of this 
proposed rule, we expect that storage tanks at many bulk plants which 
were previously devoted to 500 ppm TCO highway fuel and high-sulfur 
fuel would become available for dyed 15 ppm nonroad diesel service. 
Based on this assessment, we do not expect that a significant number of 
bulk plants would need to install an additional storage tank in order 
to provide dyed and undyed 15 ppm diesel fuel to their customers 
beginning in 2010 (the proposed implementation date for the 15 ppm 
nonroad standard).\279\ There could potentially be some additional 
costs related to the need for new tankage in some areas not already 
carrying 500 ppm fuel under the temporary compliance option of the 
highway diesel program and which continue to carry high sulfur fuel. 
However, we expect them to minimal relative to the above 0.1 cent per 
gallon cost. Thus, we estimate that the total cost of additional 
storage tanks that would result from the adoption of this proposal 
would be 0.1 cent per gallon of affected off-highway diesel engine fuel 
supplied.
---------------------------------------------------------------------------

    \279\ See section IV of today's preamble for additional 
discussion of our rational for this conclusion.
---------------------------------------------------------------------------

    Three, the proposed requirement that high sulfur heating oil be 
marked between 2007 and 2010 and that locomotive and marine diesel fuel 
be marked from 2010 until 2014 would increase the cost of distributing 
these fuels slightly. Based on input from marker manufacturers, we 
estimate that marking these fuels would cost no more than 0.2 cent per 
gallon and could cost considerably less. There should be no capital 
cost associated with this requirement, as we are proposing to remove 
the current requirement that refiners dye all high sulfur distillate at 
the refinery. The current dyeing equipment should work equally well for 
the marker. Because heating oil is being marked to prevent its use in 
NRLM engines, we have spread the cost for this marker over NRLM diesel 
fuel. Thus, from a regulatory point of view, the heating oil marker 
would increase the cost of NRLM diesel fuel between 2007 and 2010 by 
0.16 cent per gallon. We attribute the cost of marking 500 ppm 
locomotive and marine diesel fuel directly to this fuel, so the marker 
cost is simply 0.2 cent per gallon of locomotive and marine diesel fuel 
between 2010 and 2014.
    We do not project any additional downgrade of 15 ppm diesel fuel 
would result from the proposed fuel program. In our analysis of the 15 
ppm highway fuel program, we also projected additional distribution 
costs due to the need to downgrade more volume of highway diesel fuel 
to a lower value product. This is a consequence of the large difference 
between the sulfur content of 15 ppm fuel and other distillate 
products, like high sulfur diesel fuel, heating oil and jet fuel.\280\ 
We do not project that these costs would increase with this proposed 
rule. Highway diesel fuel meeting a 15 ppm cap will already be being 
distributed in all major pipeline and terminal networks. Thus, we 
expect that 15 ppm nonroad fuel would be added to batches of 15 ppm 
already being distributed. In this situation, the total interface 
volume needing to be downgraded would not increase. At the same time, 
we are not projecting that interface volume would decrease, as high 
sulfur fuels, such as jet fuel, would still be in the system.
---------------------------------------------------------------------------

    \280\ Off-highway diesel fuel sulfur content is currently 
unregulated and is approximately 3,400 ppm on average. The maximum 
allowed sulfur content of heating oil is 5,000 ppm. The maximum 
allowed sulfur content of kerosene (and jet fuel) is 3,000 ppm.
---------------------------------------------------------------------------

    Thus, overall, we estimate that the total additional distribution 
would be 0.3 cent per gallon of nonroad, locomotive and marine fuel 
during the first step of the proposed program (from 2007 through 2010). 
We project that distribution costs would increase to 0.4 cent gallon 
for 500 ppm locomotive and marine diesel fuel from 2010 to 2014, but 
decrease to 0.2 cent per gallon thereafter. Finally, we project that 
distribution costs for 15 ppm nonroad diesel fuel would be 0.2 cent 
gallon.
4. How EPA's Projected Costs Compare to Other Available Estimates
    We used two different methods for evaluating how well our cost 
estimates reflect the true costs for complying with the two step 
nonroad fuel program. The first method compared our costs with the 
incremental market price of diesel fuel meeting a 15 or 500 ppm 
standard. The second method compared our cost estimate to that from an 
engineering analysis analogous to the one we performed.
    Beginning with market prices, highway diesel fuel meeting a 500 ppm 
sulfur cap has been marketed in the U.S. for almost ten years. Over the 
five year period from 1995-1999, its national average price has 
exceeded that of high sulfur diesel fuel by about 2.4 cent per gallon 
(see chapter 7 of the Draft RIA). While fuel prices are a often a 
function of market forces which might not reflect the cost of producing 
the fuel, the comparison of the price difference over a fairly long 
period such as 5 years would tend to reduce the effect of the market on 
the prices and more closely reflect the cost of complying with the 500 
ppm cap standard. Thus, we feel that this is a sound basis for 
evaluating our cost estimate. This price difference is essentially the 
same as our estimated cost for refining and distributing 500 ppm non-
highway diesel fuel, thus the price difference for producing and 
distributing 500 ppm highway fuel corroborates our cost analysis.
    Some 15 ppm diesel fuel is marketed today. However, it is either 
being produced in very limited quantities using equipment designed to 
meet less

[[Page 28436]]

stringent sulfur standards or with other properties which make it 
unrepresentative of typical U.S. NRLM diesel fuel. Thus, current market 
prices are not a good indication of the long term price impact of the 
proposed 15 ppm cap.
    Regarding engineering studies, the Engine Manufactures Association 
(EMA) commissioned a study by Mathpro to estimate the cost of 
controlling the sulfur content of highway and nonroad diesel fuel to 
levels consistent with both 500 ppm and 15 ppm cap standards.\281\ 
Mathpro used a higher rate of return on new capital so we adjusted 
their per-gallon costs to reflect our own amortization methodology. 
Also, the Mathpro study was completed in 1999 so we adjusted their 
costs for inflation to year 2002 dollars. After these two adjustments, 
Mathpro's cost to desulfurize the high sulfur non-highway pool to 500 
ppm is 2.5 cents per gallon, while that for a 15 ppm cap is 5.8 cents 
per gallon.\282\ The 500 ppm cost estimate compares quite favorably 
with our own estimate of 2.2 cents per gallon cost. One reason for our 
somewhat lower estimate for complying with the 500 ppm standard is that 
our refinery-specific analysis has only the lowest cost refineries 
complying as many more expensive refineries can continue to produce 
heating oil. It is likely that the refineries which our analysis show 
would comply are more optimized for desulfurizating diesel fuel than 
the average refinery used by Mathpro. This reason applies even more for 
15 ppm cap standard as fewer, more optimized refineries need to comply 
to produce nonroad diesel fuel which complies with a 15 ppm sulfur cap 
standard. Furthermore, we considered the use of advanced 
desulfurization technologies for complying with the 15 ppm standard, 
while Mathpro did not. Since the Mathpro study was performed in 1999, 
cost estimates were not available for either of the two technologies 
which we included. The adjustment of the Mathpro costs and the 
comparison with our own cost estimates are discussed in detail in the 
Draft RIA. We request comment on the degree that the results of the 
Mathpro study for EMA and the comparison with real-world prices support 
our own cost estimates.
---------------------------------------------------------------------------

    \281\ Hirshfeld, David, MathPro, Inc., ``Refining economics of 
diesel fuel sulfur standards,'' performed for the Engine 
Manufactuers Association, October 5, 1999.
    \282\ The Mathpro costs cited reflect their case where current 
diesel fuel hydrotreaters are revamped with a new reactor in series, 
which is the most consistent with our technology projection.
---------------------------------------------------------------------------

5. Supply of Nonroad, Locomotive and Marine Diesel Fuel
    EPA has developed the proposed fuel program to minimize its impact 
on the supply of distillate fuel. For example: we have proposed to 
transition the fuel sulfur level down to 15 ppm in two steps, providing 
an estimated 6 years of leadtime for the final step; we are proposing 
to provide flexibility to refiners through the availability of banking 
and trading provisions; and we have provided relief for small refiners 
and hardship relief for any qualifying refiner. In order to evaluate 
the effect of this proposal on supply, EPA evaluated four possible 
cases: (1) whether the proposed standards could cause refiners to 
remove certain blendstocks from the fuel pool, (2) whether the proposed 
standards could require chemical processing which loses fuel in the 
process, (3) whether the cost of meeting the proposed standards could 
lead some refiners to leave that market, and (4) whether the cost of 
meeting the proposed standards could lead some refiners to stop 
operations altogether (i.e., shut down). In all cases, as discussed 
below, we have concluded that the answer is no. Therefore, consistent 
with our findings made during the 2007 highway diesel rule, we do not 
expect this proposed rule to cause any supply shortages of nonroad, 
locomotive and marine diesel fuel. The reader is referred to the draft 
RIA for a more detailed discussion of the potential supply impact of 
this proposed rule.
    Blendstock Shift: There should be no long term reduction in the 
amount of material derived from crude oil available for blending into 
diesel fuel or heating oil as a result of this proposal. Technology 
exists to desulfurize any commercial diesel fuel to less than 10 ppm 
sulfur. This technology is just now being proven on a commercial scale 
with a range of no. 2 diesel fuel blendstocks, as a number of refiners 
are producing 15 ppm fuel for diesel fleets which have been retro-
fitted with PM traps or for pipeline testing. Therefore, there is no 
technical necessity to remove certain blendstocks from the diesel fuel 
pool. It costs more to process certain blendstocks, such as light cycle 
oil, than others. Therefore, there may be economic incentives to move 
certain blendstocks out of the diesel fuel market to reduce compliance 
costs. However, that is an economic issue, not a technical issue and 
will be addressed below when we consider whether refiners might choose 
to exit the NRLM diesel fuel market.
    Processing Losses: The impact of the proposed rule on the total 
output of liquid fuel from refineries would be negligible. Conventional 
desulfurization processes do not reduce the energy content of the input 
material. However, the form of the material is affected slightly. With 
conventional hydrotreating, about 98 percent of the diesel fuel fed to 
a hydrotreater producing 15 ppm sulfur product leaves as diesel fuel. 
Of the 2 percent loss, three-fourths, or about 1.5 percent leaves the 
unit as naphtha (i.e., gasoline feedstock). The remainder is split 
evenly between liquified petroleum gas (LPG) and refinery fuel gas. 
Both naphtha and LPG have higher valuable uses as liquid fuels. Naphtha 
can be used to produce gasoline. Refiners can adjust the relative 
amounts of gasoline and diesel fuel which they produce, especially to 
this small degree. This additional naphtha can displace other gasoline 
blendstocks, which can then be shifted to the diesel fuel pool. LPG, on 
the other hand, is primarily used in heating, where it competes with 
heating oil. Thus, additional LPG can be used to displace gasoline and 
heating oil, which in turn can be shifted to the diesel fuel pool. 
Thus, there should be little or no direct impact of desulfurization on 
refinery fuel production. The shift from diesel fuel to fuel gas is 
very small (0.25 percent) and this fuel gas can be used to reduce 
consumption of natural gas within the refinery. These figures apply to 
the full effect of the proposed standards (i.e., the reduction in 
sulfur content from 3400 ppm to 15 ppm). For the first step of the 
proposed fuel program and that portion of the diesel fuel pool which 
would remain at the 500 ppm level indefinitely, the impacts would only 
be about 40 percent of those described above.
    The use of advanced desulfurization technologies would further 
reduce these impacts. These technologies are projected to be used in 
the second step of reducing 500 ppm diesel fuel to 15 ppm sulfur. We 
project that the Linde process would reduce the above losses for the 
second step by 55 percent, while the Phillips SZorb process would have 
no loss in diesel fuel production.
    Exit the NRLM Diesel Fuel Market: While the cost of meeting the 
proposed standards might cause some individual refiners to consider 
reducing their production of NRLM fuel or leave the market entirely, we 
do not believe that across the entire industry such a shift is possible 
or likely. As mentioned above, all diesel fuels and heating oil are 
essentially identical both chemically and physically, except for sulfur 
level. Thus, if a refiner could shift his high

[[Page 28437]]

sulfur distillate material from the nonroad, locomotive and marine 
diesel fuel markets to the heating oil market starting in mid-2007, it 
would avoid the need to invest in new desulfurization equipment. 
Likewise, starting in mid-2010, a refiner could focus his 500 ppm 
diesel fuel in the locomotive and marine diesel fuel markets or shift 
this material to the heating oil market. The problem would be a 
potential oversupply of heating oil starting in 2007 and locomotive and 
marine diesel fuel and heating oil starting in 2010. An oversupply 
could lead to a substantial drop in market price, significantly 
increasing the cost of leaving the nonroad, locomotive and marine 
diesel fuel markets. Or, it may be necessary to export the higher 
sulfur fuel in order to sell it. This could entail transportation costs 
and overseas prices no higher than existed in the U.S. before the 
oversupply (and possibly lower due to these imports now entering these 
overseas markets).
    We addressed this same issue during the development of 2007 highway 
diesel fuel program. There, the issue was whether refiners would shift 
some or all of their current highway diesel fuel production to either 
domestic or overseas markets for high sulfur diesel fuel or heating oil 
in order to avoid investing to meet the 15 ppm cap for highway diesel 
fuel. A study by Charles River Associates, et al., sponsored by API 
projected that there could be a near-term shortfall in highway diesel 
fuel supply of as much as 12 percent.\283\ However, supported by a 
study by Muse, Stancil, we concluded that refiners would incur greater 
economic loss in trying to avoid meeting the 15 ppm highway diesel fuel 
cap than they would by complying at current production levels even if 
the market did not allow them to recover their capital investment. A 
study by Mathpro, Inc. for AAM and EMA also criticized the conclusions 
of the Charles River study, particularly their assumption that 
compliance costs alone would drive investment decisions and that there 
was essentially a single highway diesel fuel market nationwide.\284\ 
Mathpro demonstrated that smaller refineries located, for example, in 
the Rocky Mountain region, likely faced higher per gallon compliance 
costs, but also had been more profitable over the past 15 years than 
larger refiners in other areas with lower overall costs. This was due 
to their market niches and the inability for lower cost refiners to 
ship large volumes of fuel economically to their market.
---------------------------------------------------------------------------

    \283\ ``An Assessment of the Potential Impacts of Proposed 
Environmental Regulations on U.S. Refinery Supply of Diesel Fuel,'' 
Charles River Associates and Baker and O'Brien, for API, August 
2000.
    \284\ ``Prospects for Adequate Supply of Ultra Low Sulfur Diesel 
Fuel in the Transition Period (2006-2007), An Analysis of Technical 
and Economic Driving Forces for Investment in ULSD Capacity in the 
U.S. Refining Sector,'' MathPro, Inc., for AAM and EMA, December 7, 
2001.
---------------------------------------------------------------------------

    We believe that the same conclusions apply to the proposed fuel 
program for six reasons. One, the alternative markets for high sulfur 
diesel fuel and heating oil would be even more limited after the 
proposed sulfur caps on nonroad, locomotive and marine diesel fuel than 
they will be in 2006, as half of the current U.S. market for high 
sulfur, no. 2 distillate would disappear. We expect that high sulfur 
heating oil would not even by carried be common carrier pipelines 
except those serving the Northeast. Therefore, refiners' sale of high 
sulfur distillate may be limited to markets serviceable by truck. Two, 
the desulfurization technology to meet a 500 ppm cap has been 
commercially demonstrated for over a decade. The desulfurization 
technology to meet a 15 ppm cap will have been commercially 
demonstrated in mid-2006, a full four years prior to the implementation 
of the 15 ppm cap on nonroad diesel fuel. Three, the volume of fuel 
affected by the 15 ppm nonroad diesel fuel standard would be only one-
seventh of that affected by the highway diesel fuel program. This 
dramatically reduces the required capital investment. Four, both Europe 
and Japan are implementing sulfur caps for highway and nonroad diesel 
fuel in the range of 10-15 ppm, eliminating these markets as a sink for 
high sulfur diesel fuel. Five, refineries outside of the U.S. and 
Europe are operating at a lower percentage of their capacity than U.S. 
refineries. Thus, U.S. refineries would not be able to obtain 
attractive prices for high sulfur diesel fuel overseas. Finally, 
refinery profit margins were much higher during the last part of 2000 
and most of 2001 than over the past ten years, indicating a potential 
long-term improvement in profitability. Margins decreased again during 
most in 2002, but recovered during the last few months of that year and 
in early 2003.
    Once refiners have made their investments to meet the proposed NRLM 
diesel fuel standards, or have decided to produce high sulfur heating 
oil, we expect that the various distillate markets would operate very 
similar to today's markets. When fully implemented in 2014, there will 
be three distillate fuels in the market, 15 ppm highway and nonroad 
diesel fuel, 500 ppm locomotive and marine diesel fuel and high sulfur 
heating oil. The market for 500 ppm locomotive and marine diesel fuel 
is much smaller than the other two, particularly considering that it is 
nationwide and the heating oil market is geographically concentrated. 
Therefore, the vast majority of refiners are expected to focus on 
producing either 15 ppm or high sulfur distillate, which is similar to 
today, where there are two fuels, 500 ppm and high sulfur distillate. 
In this case, refiners with the capability of producing 15 ppm diesel 
fuel have the most flexibility, since they can sell their fuel to any 
of the three markets. Refiners with only 500 ppm desulfurization 
capability can supply two markets. Those refiners only capable of 
producing high sulfur distillate would not be able to participate in 
either the 15 or 500 ppm markets. However, this is not different from 
today. Generally, we do not expect one market to provide vastly 
different profit margins than the others, as high profit margins in one 
market will attract refiners from another via investment in 
desulfurization equipment.
    Refinery Closure: There are a number of reasons why we do not 
believe that refineries would completely close down under this proposed 
rule. One reason is that we have included provisions to provide relief 
for small refiners, as well as any refiner facing unusual financial 
hardship. Another reason is that nonroad, locomotive and marine diesel 
fuel is usually the third or fourth most important product produced by 
the refinery from a financial perspective. A total shutdown would mean 
losing all the revenue and profit from these other products. Gasoline 
is usually the most important product, followed by highway diesel fuel 
and jet fuel. A few refineries do not produce either gasoline or 
highway diesel fuel, so jet fuel and high sulfur diesel fuel and 
heating oil are their most important products. The few refiners in this 
category likely face the biggest financial challenge in meeting the 
proposed requirements. However, those refiners would also presumably be 
in the best position to apply for special hardship provisions, 
presuming that they do not have readily available source of investment 
capital. The additional time afforded by these provisions should allow 
the refiner to generate sufficient cash flow to invest in the required 
desulfurization equipment. Investment here could also provide them the 
opportunity to expand into more profitable (e.g., highway diesel) 
markets.
    A quantitative evaluation of whether the cost of the proposed fuel 
program could cause some refineries to cease operations completely 
would be very difficult, if not impossible to perform. A

[[Page 28438]]

major factor in any decision to shut down is the refiner's current 
financial situation. It is very difficult to assess an individual 
refinery's current financial situation. This includes a refiner's debt, 
as well as its profitability in producing fuels other than those 
affected by a particular regulation. It can also include the 
profitability of other operations and businesses owned by the refiner.
    Such an intensive analysis can be done to some degree in the 
context of an application for special hardship provisions, as discussed 
above. However, in this case, EPA can request detailed financial 
documents not normally available. Prior to such application, as is the 
case now, this financial information is usually confidential. Even when 
it is published, the data usually apply to more than just the operation 
of a single refinery.
    Another factor is the need for capital investments other than for 
this proposed rule. EPA can roughly project the capital needed to meet 
other new fuel quality specifications, such as the Tier 2 or highway 
diesel sulfur standards. However, we cannot predict investments to meet 
local environmental and safety regulations, nor other investments 
needed to compete economically with other refiners.
    Finally, any decision to close in the future must be based on some 
assumption of future fuel prices. Fuel prices are very difficult to 
project in absolute terms. The response of prices to changes in fuel 
quality specifications, such as sulfur content, as is discussed in the 
next section, are also very difficult to predict. Thus, even if we had 
complete knowledge of a refiner's financial status and its need for 
future investments, the decision to stay in business or close would 
still depend on future earnings, which are highly dependent on the 
prices of all products produced by that refinery.
    Some studies in this area point to fuel pricing over the past 15 
years or so and conclude that prices will only increase to reflect 
increased operating costs and will not reflect the cost of capital. In 
fact, the rate of return on refining assets has been poor over the past 
15 years and until recently, there has been a steady decline in the 
number of refineries operating in the U.S. However, this may have been 
due to a couple of circumstances specific to that time period. One, 
refinery capacity utilization was less than 80 percent in 1985. Two, at 
least regarding gasoline, the oxygen mandate for reformulated gasoline 
caused an increase in gasoline supply despite low refinery utilization 
rates. While this led to healthy financial returns for oxygenate 
production, it did not help refining profit margins.
    Today, refinery capacity utilization in the U.S. is generally 
considered to be at its maximum sustainable rate. There are no 
regulatory mandates on the horizon which will increase production 
capacity significantly, even if ethanol use in gasoline increases 
substantially.\285\ Consistent with this, refining margins have been 
much better over the past two and a half years than during the previous 
15 years and the refining industry itself is projecting good returns 
for the foreseeable future.
---------------------------------------------------------------------------

    \285\ Both houses of the U.S. Congress are considering bills 
which would require the increased use of renewables, like ethanol, 
in gasoline and diesel fuel. While the amount of renewables could be 
considerable, it is well below the annual growth in transportation 
fuel use.
---------------------------------------------------------------------------

6. Fuel Prices
    It is well known that it is difficult to predict fuel prices in 
absolute terms with any accuracy. The price of crude oil dominates the 
cost of producing gasoline and diesel fuel. Crude oil prices have 
varied by more than a factor of two in the past year. In addition, 
unexpectedly warm or cold winters can significantly affect heating oil 
consumption, which affects the amount of gasoline produced and the 
amount of distillate material available for diesel fuel production. 
Economic growth, or its lack, affects fuel demand, particularly for 
diesel fuel. Finally, both planned and unplanned shutdowns of 
refineries for maintenance and repairs can significantly affect total 
fuel production, inventory levels and resulting fuel prices.
    Predicting the impact of any individual factor on fuel price is 
also difficult. The overall volatility in fuel prices limits the 
ability to determine the effect of a factor which changed at a specific 
point in time which might have led to the price change, as other 
factors continue to change over time. Occasionally, a fuel quality 
change, such as reformulated gasoline or a 500 ppm cap on diesel fuel 
sulfur content, only affects a portion of the fuel pool. In this case, 
an indication of the impact on price can be inferred by comparing the 
prices of the two fuels at the same general location over time. 
However, this is still only possible after the fact, and cannot be done 
before the fuel quality change takes place.
    Because of these difficulties, EPA has generally not attempted to 
project the impact of its rules on fuel prices. However, in response to 
Executive Order 13211, we are doing so for this proposed rule. To 
reflect the inherent uncertainty in making such projections, we 
developed three projections for the potential impact of the proposed 
fuel program on fuel prices. The range of potential long-term price 
increases are shown in Table V-A-4. Short-term price impacts are highly 
volatile, as are short-term swings in absolute fuel prices, and much 
too dependent on individual refiners' decisions, unexpected shutdowns, 
etc. to be predicted even with broad ranges.

             Table V-A-4.--Range of Possible Total Diesel Fuel Price Increases (cents per gallon) a
----------------------------------------------------------------------------------------------------------------
                                                                    Lower Limit      Mid-Point        Maximum
----------------------------------------------------------------------------------------------------------------
               2007 500 ppm Sulfur Cap: Nonroad, Locomotive and Marine Diesel Fuel
-------------------------------------------------------------------------------------------------
PADDs 1 and 3...................................................             0.9             1.5             3.4
PADD 2..........................................................             2.3             3.0             4.8
PADD 4..........................................................             1.7             4.1             5.8
PADD 5..........................................................             1.0             2.8             4.3
-----------------------------------------------------------------
                           2010 15 ppm Sulfur Cap: Nonroad Diesel Fuel
-------------------------------------------------------------------------------------------------
PADDs 1 and 3...................................................             1.8             3.0             5.4
PADD 2..........................................................             2.9             6.1             7.4
PADD 4..........................................................             3.0             8.9             9.3
PADD 5..........................................................             1.7             5.9            8.4
----------------------------------------------------------------------------------------------------------------
Notes:
a At the current wholesale price of approximately $1.00 per gallon, these values also represent the percentage
  increase in diesel fuel price.


[[Page 28439]]

    The lower end of the range assumes that prices within a PADD 
increased to reflect the highest operating cost increase faced by any 
refiner in that PADD. In this case, this refiner with the highest 
operating cost would not recover any of his invested capital, but all 
other refiners would recover some or all of their investment. In this 
case, the price of nonroad, locomotive and marine diesel fuel would 
increase in 2007 by 1-2 cents per gallon, depending on the area of the 
country. In 2010, the price of nonroad diesel fuel would increase a 
total of 2-3 cents per gallon. Locomotive and marine diesel fuel prices 
would continue to increase by 1-2 cents per gallon.
    The mid-range estimate of price impacts assumes that prices within 
a PADD increase by the average refining and distribution cost within 
that PADD, including full recovery of capital (at 7 percent per annum 
before taxes). Lower cost refiners would recover more than their 
capital investment, while those with higher than average costs recover 
less. Under this assumption, the price of nonroad, locomotive and 
marine diesel fuel would increase in 2007 by 2-4 cents per gallon, 
depending on the area of the country. In 2010, the price of nonroad 
diesel fuel would increase a total of 3-9 cents per gallon. Locomotive 
and marine diesel fuel prices would continue to increase by 2-4 cents 
per gallon.
    The upper end estimate of price impacts assumes that prices within 
a PADD increase by the maximum total refining and distribution cost of 
any refinery within that PADD, including full recovery of capital (at 7 
percent per annum before taxes). All other refiners would recover more 
than their capital investment. Under this assumption, the price of 
nonroad, locomotive and marine diesel fuel would increase in 2007 by 3-
6 cents per gallon, depending on the area of the country. In 2010, the 
price of nonroad diesel fuel would increase a total of 5-9 cents per 
gallon. Locomotive and marine diesel fuel prices would continue to 
increase by 3-6 cents per gallon.
    In addition to the differences noted above, there are a number of 
assumptions inherent in all three of the above price projections. 
First, both the lower and upper limits of the projected price impacts 
described above assume that the refinery facing the highest compliance 
costs is currently the price setter in their market. This is a worse 
case assumption which is impossible to validate. Many factors affect a 
refinery's total costs of fuel production. Most of these factors, such 
as crude oil cost, labor costs, age of equipment, etc., are not 
considered in projecting the incremental costs associated with lower 
NRLM diesel fuel sulfur levels. Thus, current prices may very well be 
set in any specific market by a refinery facing lower incremental 
compliance costs than other refineries. This point was highlighted in a 
study by the National Economic Research Associates (NERA) for AAM of 
the potential price impacts of EPA's 2007 highway diesel fuel 
program.\286\ In that study, NERA criticized the above referenced study 
performed by Charles River Associates, et al. for API, which projected 
that prices would increase nationwide to reflect the total cost faced 
by the U.S. refinery with the maximum total compliance cost of all the 
refineries in the U.S. producing highway diesel fuel. To reflect the 
potential that the refinery with the highest projected compliance costs 
under the maximum price scenario is not the current price setter, we 
included the mid-point price impacts above. It is possible that even 
the lower limit price impacts are too high, if the conditions exist 
where prices are set based on operating costs alone. However, these 
price impacts are sufficiently low that considering even lower price 
impacts was not considered critical to estimating the potential 
economic impact of this rule.
---------------------------------------------------------------------------

    \286\ ``Potential Impacts of Environmental Regulations on Diesel 
Fuel Prices,'' NERA, for AAM, December 2000.
---------------------------------------------------------------------------

    Second, we assumed that a single refinery's costs could affect fuel 
prices throughout an entire PADD. While this is a definite improvement 
over analyses which assume that a single refinery's costs could affect 
fuel prices throughout the entire nation, it is still conservative. 
High cost refineries are more likely to have a more limited 
geographical impact on market pricing than an entire PADD.
    Third, by focusing solely on the cost of desulfurizing NRLM diesel 
fuel, we assume that the production of NRLM diesel fuel is independent 
of the production of other refining products, such as gasoline, jet 
fuel and highway diesel fuel. However, this is clearly not the case. 
Refiners have some flexibility to increase the production of one 
product without significantly affecting the others, but this 
flexibility is quite limited. It is possible that the relative 
economics of producing other products could influence a refiner's 
decision to increase or decrease the production of NRLM diesel fuel 
under the proposed standards. This in turn could increase or decrease 
the price impact relative to those projected above.
    Fourth, all three of the above price projections are based on the 
projected cost for U.S. refineries of meeting the proposed NRLM diesel 
fuel sulfur caps. Thus, these price projections assume that imports of 
NRLM fuel, which are currently significant in the Northeast, are 
available at roughly the same cost as those for U.S. refineries in 
PADDs 1 and 3. We have not performed any analysis of the cost of lower 
sulfur caps on diesel fuel produced by foreign refiners. However, there 
are reasons to believe that imports of 500 and 15 ppm NRLM diesel fuel 
would be available at prices in the ranges of those projected for U.S. 
refiners.
    One recent study analyzed the relative cost of lower sulfur caps 
for Asian refiners relative to those in the U.S., Europe and 
Japan.\287\ It concluded that costs for Asian refiners would be 
comparatively higher, due to the lack of current hydrotreating capacity 
at Asian refineries. This conclusion is certainly valid when evaluating 
lower sulfur levels for highway diesel fuels which are already at low 
levels in the U.S., Europe and Japan and for which refineries in these 
areas have already invested in hydrotreating capacity. It would appear 
to be less valid when assessing the relative cost of meeting lower 
sulfur standards for nonroad diesel fuels and heating oils which are 
currently at much higher sulfur levels in the U.S., Europe and Japan. 
All refineries face additional investments to remove sulfur from these 
fuels and so face roughly comparable control costs on a per gallon 
basis.
---------------------------------------------------------------------------

    \287\ ``Cost of Diesel Fuel Desulfurization In Asian 
Refineries,'' Estrada International Ltd., for the Asian Development 
Bank, December 17, 2002.
---------------------------------------------------------------------------

    One factor arguing for competitively priced imports is the fact 
that refinery utilization rates are currently higher in the U.S. and 
Europe than in the rest of the world. The primary issue is whether 
overseas refiners will invest to meet tight sulfur standards for U.S., 
European and Japanese markets. Many overseas refiners will not invest, 
instead focusing on local, higher sulfur markets. However, many 
overseas refiners focus on exports. Both Europe and the U.S. are moving 
towards highway and nonroad diesel fuel sulfur caps in the 10-15 ppm 
range. Europe is currently and projected to continue to need to import 
large volumes of highway diesel fuel. Thus, it seems reasonable to 
expect that a number of overseas refiners would invest in the capacity 
to produce some or all of their diesel fuel at these levels. Overseas 
refiners also have the flexibility to produce 10-15 ppm diesel fuel 
from their cleanest blendstocks, as

[[Page 28440]]

most of their available markets have less stringent sulfur standards. 
Thus, there are reasons to believe that some capacity to produce 10-15 
ppm diesel fuel would be available overseas at competitive prices. If 
these refineries were operating well below capacity, they might be 
willing to supply complying product at prices which only reflect 
incremental operating costs. This could hold prices down in areas where 
importing fuel is economical. However, it is unlikely that these 
refiners could supply sufficient volumes to hold prices down 
nationwide. Despite this expectation, to be conservative, in the 
refining cost analysis conducted earlier in this chapter, we assumed no 
imports of 500 ppm or 15 ppm NRLM diesel fuel. All 500 ppm and 15 ppm 
nonroad diesel fuel was produced by domestic refineries. This raised 
the average and maximum costs of 500 ppm and 15 ppm NRLM diesel fuel 
and increased the potential price impacts projected above beyond what 
would have been projected had we projected that 5-10 percent of NRLM 
diesel fuel would be imported at competitive prices.

B. Cost Savings to the Existing Fleet from the Use of Low Sulfur Fuel

    We estimate that reducing fuel sulfur to 500 ppm would reduce 
engine wear and oil degradation to the existing nonroad diesel 
equipment fleet and that a further reduction to 15 ppm sulfur would 
result in even greater reductions. This reduction in wear and oil 
degradation would provide a dollar savings to users of nonroad 
equipment. The cost savings would also be realized by the owners of 
future nonroad engines that are subject to the standards in this 
proposal. As discussed below, these maintenance savings have been 
conservatively estimated to be greater than 3 cents per gallon for the 
use of 15 ppm sulfur fuel when compared to the use of today's 
unregulated nonroad diesel fuel. A summary of the benefits of low-
sulfur fuel is presented in Table V.B-1.\288\
---------------------------------------------------------------------------

    \288\ See Heavy-duty 2007 Highway Final RIA, Chapter V.C.5, and 
``Study of the Effects of Reduced Diesel Fuel Sulfur Content on 
Engine Wear'', EPA report  460/3-87-002, June 1987.

   Table V.B-1--Engine Components Potentially Affected by Lower Sulfur
                          Levels in Diesel Fuel
------------------------------------------------------------------------
                                 Effect of Lower    Potential Impact on
     1Affected Components             Sulfur           Engine System
------------------------------------------------------------------------
Piston Rings..................  Reduced corrosion  Extended engine life
                                 wear.              and less frequent
                                                    rebuilds.
Cylinder Liners...............  Reduced corrosion  Extended engine life
                                 wear.              and less frequent
                                                    rebuilds.
Oil Quality...................  Reduced deposits,  Reduce wear on piston
                                 reduced acid       ring and cylinder
                                 build-up, and      liner and less
                                 less need for      frequent oil
                                 alkaline           changes.
                                 additives.
Exhaust System (tailpipe).....  Reduced corrosion  Less frequent part
                                 wear.              replacement.
Exhaust Gas Recirculation       Reduced corrosion  Less frequent part
 System.                         wear.              replacement.
------------------------------------------------------------------------

    The monetary value of these benefits over the life of the equipment 
will depend upon the length of time that the equipment operates on low-
sulfur diesel fuel and the degree to which engine and equipment 
manufacturers specify new maintenance practices and the degree to which 
equipment operators change engine maintenance patterns to take 
advantage of these benefits. For equipment near the end of its life in 
the 2008 time frame, the benefits will be quite small. However, for 
equipment produced in the years immediately preceding the introduction 
of 500 ppm sulfur fuel, the savings would be substantial. Additional 
savings would be realized in 2010 when the 15 ppm sulfur fuel would be 
introduced.
    We estimate the single largest savings would be the impact of lower 
sulfur fuel on oil change intervals. The draft RIA presents our 
analysis for the oil change interval extension which would be realized 
by the introduction of 500 ppm sulfur fuel in 2007, as well as the 
additional oil extension which would be realized with the introduction 
of 15 ppm sulfur nonroad diesel fuel in 2010. As explained in the draft 
RIA, these estimates are based on our analysis of publically available 
information from nonroad engine manufacturers. Due to the wide range of 
diesel fuel sulfur which today's nonroad engines may see around the 
world, engine manufacturers specify different oil change intervals as a 
function of diesel sulfur levels. We have used this data as the basis 
for our analysis. Taken together, when compared to today's relatively 
high nonroad diesel fuel sulfur levels, we estimate the use of 15 ppm 
sulfur fuel will enable an oil change interval extension of 35 percent 
from today's products.
    We present here a fuel cost savings attributed to the oil change 
interval extension in terms of a cents per gallon operating cost. We 
estimate that an oil change interval extension of 31 percent, as would 
be enabled by the use of 500 ppm sulfur fuel in 2007, results in a fuel 
operating costs savings of 3.0 cents per gallon for the nonroad fleet. 
We project an additional cost savings of 0.3 cents per gallon for the 
oil change interval extension which would be enabled by the use of 15 
ppm sulfur beginning in 2010. Thus, for the nonroad fleet as a whole, 
beginning in 2010 nonroad equipment users can realize an operating cost 
savings of 3.3 cents per gallon compared to today's engine. This means 
that the end cost to the typical user for 15ppm sulfur fuel is 
approximately 1.5 cents per gallon (4.8 cent per gallon cost for fuel 
minus 3.3 cent per gallon maintenance savings). For a typical 100 
horsepower nonroad engine this represents a net present value lifetime 
savings of more than $500.
    These savings will occur without additional new cost to the 
equipment owner beyond the incremental cost of the low-sulfur diesel 
fuel, although these savings are dependent on changes to existing 
maintenance schedules. Such changes seem likely given the magnitude of 
the savings. We have not estimated the value of the savings from the 
other benefits listed in Table V.B-1, and therefore we believe the 3.3 
cents per gallon savings is conservative as it only accounts for the 
impact of low sulfur fuel on oil change intervals.

C. Engine and Equipment Cost Impacts

    The following sections briefly discuss the various engine and 
equipment cost elements considered for this proposal and present the 
total costs we have estimated; the reader is referred to the draft RIA 
for a complete discussion. Estimated engine and equipment costs depend 
largely on both the size of the piece of equipment and its engine, and 
on the technology package being added to the engine to ensure 
compliance with the proposed standards. The wide size variation (e.g., 
<4 horsepower engines through 2500 horsepower engines) and

[[Page 28441]]

the broad application variation (e.g., lawn equipment through large 
mining trucks) that exists in the nonroad industry makes it difficult 
to present here an estimated cost for every possible engine and/or 
piece of equipment. Nonetheless, for illustrative purposes, we present 
some example per engine/equipment cost impacts throughout this 
discussion. This analysis is presented in detail in Chapter 6 of the 
draft RIA. We are also considering doing a sensitivity analysis on 
cost/engine data, which would be put into the docket for comment.
    It is important to note that the costs presented here do not 
reflect any savings that are expected to occur because of the engine 
ABT program and the equipment manufacturer transition program, both of 
which are discussed in Section VII. As discussed in the draft RIA, 
these optional programs have the potential to provide significant 
savings for both engine and equipment manufacturers. We request comment 
with supporting data and/or analysis on the cost estimates presented 
here and the underlying analysis presented in chapter 6 of the draft 
RIA.
1. Engine Cost Impacts
    Estimated engine costs are broken into fixed costs (for research 
and development, retooling, and certification), variable costs (for new 
hardware and assembly time), and life-cycle operating costs. Total 
operating costs include the estimated incremental cost for low-sulfur 
diesel fuel, any expected increases in maintenance costs associated 
with new emission control devices, any costs associated with increased 
fuel consumption, and any decreases in operating cost (i.e., 
maintenance savings) expected due to low-sulfur fuel. Cost estimates 
presented here represent an expected incremental cost of engines in the 
model year of their introduction. Costs in subsequent years would be 
reduced by several factors, as described below. All engine and 
equipment costs are presented in 2001 dollars.
a. Engine Fixed Costs
i. Engine and Emission Control Device R&D
    The technologies described in section III represent those 
technologies we believe will be used to comply with the proposed Tier 4 
emission standards. These technologies are part of an ongoing research 
and development effort geared toward compliance with the 2007 heavy-
duty diesel highway emission standards. The engine manufacturers making 
R&D expenditures toward compliance with highway emission standards will 
have to undergo some additional R&D effort to transfer emission control 
technologies to engines they wish to sell into the nonroad market. 
These R&D efforts will allow engine manufacturers to develop and 
optimize these new technologies for maximum emission-control 
effectiveness with minimum negative impacts on engine performance, 
durability, and fuel consumption. Many nonroad engine manufacturers are 
not part of the ongoing R&D effort toward compliance with highway 
emissions standards because they do not sell engines into the highway 
market. These manufacturers are expected to benefit from the R&D work 
that has already occurred and will continue through the coming years 
through their contact with highway manufacturers, emission control 
device manufacturers, and the independent engine research laboratories 
conducting relevant R&D.
    Several technologies are projected for complying with the proposed 
Tier 4 emission standards. We are projecting that NOX 
adsorbers and catalyzed diesel particulate filters (CDPFs) would be the 
most likely technologies applied by industry to meet our proposed 
emissions standards for 75 horsepower engines. The fact that 
these technologies are being developed for implementation in the 
highway market prior to the implementation dates in this proposal, and 
the fact that engine manufacturers would have several years before 
implementation of the proposed Tier 4 standards, ensures that the 
technologies used to comply with the nonroad standards would undergo 
significant development before reaching production. This ongoing 
development could lead to reduced costs in three ways. First, we expect 
research will lead to enhanced effectiveness for individual 
technologies, allowing manufacturers to use simpler packages of 
emission control technologies than we would predict given the current 
state of development. Similarly, we anticipate that the continuing 
effort to improve the emission control technologies will include 
innovations that allow lower-cost production. Finally, we believe that 
manufacturers would focus research efforts on any drawbacks, such as 
fuel economy impacts or maintenance costs, in an effort to minimize or 
overcome any potential negative effects.
    We anticipate that, in order to meet the proposed standards, 
industry would introduce a combination of primary technology upgrades. 
Achieving very low NOX emissions would require basic 
research on NOX emission control technologies and 
improvements in engine management to take advantage of the exhaust 
emission control system capabilities. The manufacturers are expected to 
take a systems approach to the problem of optimizing the engine and 
exhaust emission control system to realize the best overall 
performance. Since most research to date with exhaust emission control 
technologies for nonroad applications has focused on retrofit programs, 
there remains room for significant improvements by taking such a 
systems approach. The NOX adsorber technology in particular 
is expected to benefit from re-optimization of the engine management 
system to better match the NOX adsorber's performance 
characteristics. The majority of the dollars we have estimated for 
research is expected to be spent on developing this synergy between the 
engine and NOX exhaust emission control systems. Therefore, 
for engines requiring both a CDPF and a NOX adsorber (i.e., 
75 horsepower), we have attributed two-thirds of the R&D 
expenditures to NOX control, and one-third to PM control.
    In the 2007 HD highway rule, we estimated that each engine 
manufacturer would expend $35 million for R&D to redesign their engines 
and apply catalyzed diesel particulate filters (CDPF) and 
NOX adsorbers. For their nonroad R&D efforts on engines 
requiring CDPFs and NOX adsorbers (i.e., 75 
horsepower), engine manufacturers selling into the highway market would 
incur some level of R&D effort but not at the level incurred for the 
highway rule. In many cases, the engines used by highway manufacturers 
in nonroad products are based on the same engine platform as those used 
in highway products. However, horsepower and torque characteristics are 
often different so some effort will have to be expended to accommodate 
those differences. For these manufacturers, we have estimated that they 
would incur an R&D expense of $3.5 million. This $3.5 million R&D 
expense would allow for the transfer of R&D knowledge from their 
highway experience to their nonroad engine product line. Two-thirds of 
this R&D is attributed to NOX control and one-third to PM 
control.
    For those manufacturers that sell engines only into the nonroad 
market, and where those engines require a CDPF and a NOX 
adsorber, we believe that they will incur an R&D expense nearing that 
incurred by highway manufacturers for the highway rule, although not at 
the level incurred by highway manufacturers for the highway rule. 
Nonroad manufacturers would be able to learn from the R&D efforts 
already

[[Page 28442]]

under way for both the highway rule and for the Tier 2 light-duty 
highway rule (65 FR 6698). This learning could be done via seminars, 
conferences, and contact with highway manufacturers, emission control 
device manufacturers, and the independent engine research laboratories 
conducting relevant R&D. Therefore, for these manufacturers, we have 
estimated an expenditure of $24.5 million. This lower number--$24.5 
million versus $35 million in the highway rule--reflects the transfer 
of knowledge to nonroad manufacturers that would occur from the many 
stakeholders in the diesel industry. Two-thirds of this R&D is 
attributed to NOX control and one-third to PM control.
    Note that the $3.5 million and $24.5 million estimates represent 
our estimate of the average R&D expected by manufacturers. These 
estimates would be different for each manufacturer--some higher, some 
lower--depending on product mix and the ability to transfer knowledge 
from one product to another.
    For those engine manufacturers selling engines that would require 
CDPF-only R&D (i.e., 25 to 75 horsepower engines in 2013), we have 
estimated that the R&D they would incur would be roughly one-third that 
incurred by manufacturers conducting CDPF/NOX adsorber R&D. 
We believe this is a good estimate because CDPF technology is further 
along in its development than is NOX adsorber technology 
and, therefore, a 50/50 split would not be appropriate. Using this 
estimate, the R&D incurred by manufacturers that have already done 
selling any engines into both the highway and the nonroad markets would 
be $1.2 million, and the R&D for manufacturers selling engines into 
only the nonroad market would be roughly $8 million. All of this R&D is 
attributed to PM control.
    For those engine manufacturers selling engines that would require 
DOC-only or some engine-out modification R&D (i.e., <75 horsepower 
engines in 2008), we have estimated that the R&D they would incur would 
be roughly one-half the amount estimated for their CDPF-only R&D. Using 
this estimate, the R&D incurred by manufacturers selling any engines 
into both the highway and nonroad markets would be roughly $600,000, 
and the R&D for manufacturers selling engines into only the nonroad 
market would be roughly $4 million. All of this R&D is attributed to PM 
control.
    Some manufacturers of engines produce engines to specifications 
developed by other manufacturers. Such joint venture manufacturers do 
not conduct engine-related R&D but simply manufacture an engine 
designed and developed by another manufacturer. For such manufacturers, 
we have assumed no R&D expenditures given that we believe they will 
conduct no R&D themselves and will rely on their joint venture partner. 
This is true unless the parent company has no engine sales in the 
horsepower categories covered by the partner company. Under such a 
situation, we have accounted for the necessary R&D by attributing it to 
the parent company. We have also estimated that some manufacturers will 
choose not to invest in R&D for the U.S. nonroad market due to low 
volume sales that probably cannot justify the expense. More detail on 
these assumptions and the number of manufacturers assumed not to expend 
R&D is presented in Chapter 6 of the draft RIA. We welcome comments and 
supporting documentation.
    We have assumed that all R&D expenditures occur over a five year 
span preceding the first year any emission control device is introduced 
into the market. Where a phase-in exists (e.g., for NOX 
standards on 75 horsepower engines), expenditures are 
assumed to occur over the five year span preceding the first year 
NOX adsorbers would be introduced, and then to continue 
during the phase-in years; the expenditures would be incurred in a 
manner consistent with the phase-in of the standard. All R&D 
expenditures are then recovered by the engine manufacturer over an 
identical time span following the introduction of the technology. We 
assume a seven percent rate of return for all R&D. We have apportioned 
these R&D costs across all engines that are expected to use these 
technologies, including those sold in other countries or regions that 
are expected to have similar standards. We have estimated the fraction 
of the U.S. sales to this total sales at 42 percent. Therefore, we have 
attributed this amount to U.S. sales.
    Using this methodology, we have estimated the total R&D 
expenditures attributable to the proposed standards at $199 million.
ii. Engine-Related Tooling Costs
    Once engines are ready for production, new tooling will be required 
to accommodate the assembly of the new engines. In the 2007 highway 
rule, we estimated approximately $1.6 million per engine line for 
tooling costs associated with CDPF/NOX adsorber systems. For 
the proposed nonroad Tier 4 standards, we have estimated that nonroad-
only manufacturers would incur the same $1.6 million per engine line 
requiring a CDPF/NOX adsorber system and that these costs 
would be split evenly between NOX control and PM control. 
For those systems requiring only a CDPF, we have estimated one-half 
that amount, or $800,000 per engine line. For those systems requiring 
only a DOC or some engine-out modifications, we have applied a one-half 
factor again, or $400,000 per engine line. Tooling costs for CDPF-only 
and for DOC engines are attributed solely to PM control.
    For those manufacturers selling into both the highway and nonroad 
markets, we have estimated one-half the baseline tooling cost, or 
$800,000, for those engine lines requiring a CDPF/NOX 
adsorber system. We believe this is reasonable since many nonroad 
engines are produced on the same engine line with their highway 
counterparts. For such lines, we believe very little to no tooling 
costs would be incurred. For engine lines without a highway 
counterpart, something approaching the $1.6 million tooling cost would 
be applicable. For this analysis, we have assumed a 50/50 split of 
engine product lines for highway manufacturers and, therefore, a 50 
percent factor applied to the $1.6 million baseline. These tooling 
costs would be split evenly between NOX control and PM 
control. For engine lines <75 horsepower, we have used the same tooling 
costs as the nonroad-only manufacturers because these engines tend not 
to have a highway counterpart. Therefore, for those engine lines 
requiring only a CDPF (i.e., those between 25 and 75 horsepower), we 
have estimated a tooling cost of $800,000. Similarly, the tooling costs 
for DOC and/or engine-out engine lines has been estimated to be 
$400,000. Tooling costs for CDPF-only and for DOC engines are 
attributed solely to PM control.
    We expect engines in the 25 to 50 horsepower range to apply EGR 
systems to meet the proposed NOX standards for 2013. For 
these engines, we have included an additional tooling cost of $40,000 
per engine line, consistent with the EGR-related tooling cost estimated 
for 50-100 horsepower engines in our Tier 2/3 rulemaking. This tooling 
cost is applied equally to all engine lines in that horsepower range 
regardless of the markets into which the manufacturer sells. We have 
applied this tooling cost equally because engines in this horsepower 
range do not tend to have highway counterparts. Tooling costs for EGR 
systems are attributed solely to NOX control.
    We have applied all the above tooling costs to all manufacturers 
that appear to actually make engines. We have not

[[Page 28443]]

eliminated joint venture manufacturers because these manufacturers 
would still need to invest in tooling to make the engines even if they 
do not conduct any R&D. We have assumed that all tooling costs are 
incurred one year in advance of the new standard and are recovered over 
a five year period following implementation of the new standard; all 
tooling costs are marked up seven percent to reflect the time value of 
money. As done for R&D costs, we have attributed a portion of the 
tooling costs to U.S. sales and a portion to sales in other countries 
expected to have similar levels of emission control. More information 
is contained in Chapter 6 of the draft RIA and we request comment on 
how we have applied our tooling cost estimates and to whom we have 
applied them.
    Using this methodology, we estimate the total tooling expenditures 
attributable to the proposed standards at $67 million.
iii. Engine Certification Costs
    Manufacturers will incur more than the normal level of 
certification costs during the first few years of implementation 
because engines will need to be certified to the new emission 
standards. Consistent with our recent standard setting regulations, we 
have estimated engine certification costs at $60,000 per new engine 
certification to cover testing and administrative costs. To this we 
have added the proposed certification fee of $2,156 per new engine 
family. This cost, $62,156 per engine family was used for <75 
horsepower engines certifying to the 2008 standards. For 25 to 75 
horsepower engines certifying to the 2013 standards, and for 
75 horsepower engines certifying to their proposed 
standards, we have added costs to cover the proposed test procedures 
for nonroad diesel engines (i.e., the transient test and the NTE); 
these costs were estimated at $10,500 per engine family. These 
certification costs--whether it be the $62,156 or the $72,656 per 
engine family--apply equally to all engine families for all 
manufacturers regardless of into what markets the manufacturer sells. 
We have applied these certification costs to only the US sold engines 
because the certification conducted for US sales is not presumed to 
fulfill the certification requirements of other countries.
    Applying these costs to each of the 665 engine families as they are 
certified to a new emissions standard results in total costs of $72 
million expended during implementation of the proposed standards. These 
costs are attributed to NOX and PM control consistent with 
the phase-in of the new emissions standards--where new NOX 
and PM standards are introduced together, the certification costs are 
split evenly; where only a new PM standard is introduced, the 
certification costs are attributed to PM only; where a NOX 
phase-in becomes 100% in a year after full implementation of a PM 
standard, the certification costs are attributed to NOX 
only. All certification costs are assumed to occur one year prior to 
the new emission standard and are then recovered over a five year 
period following compliance with the new standard; all certification 
costs are marked up seven percent to reflect the time value of money.
b. Engine Variable Costs
    This section summarizes the detailed analysis presented in the 
draft RIA for this proposed rule. We encourage the reader to refer to 
chapter 6 of that draft RIA for the details of what is presented here 
and encourage comments and supporting data and/or analysis regarding 
those details. Of particular interest are comments regarding the costs 
of precious metals, or platinum group metals (PGM). The PGM costs are a 
significant fraction of the total costs for aftertreatment devices. For 
our analysis, we have used the 2002 annual average costs for platinum 
and rhodium (the two PGMs we expect will be used) because we believe 
they represent a better estimate of the cost for PGM than other 
metrics. We request comment on this approach and whether an alternative 
approach would be more appropriate. Specifically, we request comment 
regarding the use of a five year average in place of the one year 
average we have used. Additionally, EPA invites comment on the impacts, 
if any, that this rulemaking would have in the context of a variety of 
rulemakings on the market impacts on precious metals.
i. NOX Adsorber System Costs
    The NOX adsorber system that we are anticipating would 
be applied for Tier 4 would be the same as that used for highway 
applications. In order for the NOX adsorber to function 
properly, a systems approach that includes a reductant metering system 
and control of engine A/F ratio is also necessary. Many of the new air 
handling and electronic system technologies developed in order to meet 
the Tier 2/3 nonroad engine standards can be applied to accomplish the 
NOX adsorber control functions as well. Some additional 
hardware for exhaust NOX or O2 sensing and for 
fuel metering will likely be required. The cost estimates include a DOC 
for clean-up of hydrocarbon emissions that occur during NOX 
adsorber regeneration events. We have also assumed that warranty costs 
would increase due to the application of this new hardware. Chapter 6 
of the draft RIA contains the details for how we estimated costs 
associated with the new NOX control technologies required to 
meet the proposed Tier 4 emission standards. These costs are estimated 
to increase engine costs by roughly $670 in the near-term for a 150 
horsepower engine, and $2,070 in the near-term for a 500 horsepower 
engine. In the long-term, we estimate these costs to be $550 and $1,670 
for the 150 horsepower and 500 horsepower engines, respectively. Note 
that we have estimated costs for all engines in all horsepower ranges, 
and these estimates are presented in detail in the draft RIA. 
Throughout this discussion of engine and equipment costs, we present 
costs for a 150 and a 500 horsepower engine for illustrative purposes.
ii. Catalyzed Diesel Particulate Filter (CDPF) Costs
    CDPFs can be made from a wide range of filter materials including 
wire mesh, sintered metals, fibrous media, or ceramic extrusions. The 
most common material used for CDPFs for heavy-duty diesel engines is 
cordierite. We have based our cost estimates on the use of silicon 
carbide (SiC) even though it is more expensive than other filter 
materials. We request comment on our assumption that SiC will be used 
in favor of cordierite. We estimate that the CDPF systems will add $780 
to engine costs in the near-team for a 150 horsepower engine and $2,770 
in the near-term for a 500 horsepower engine. In the long-term, we 
estimate these CDPF system costs to be $590 and $2,110 for the 150 
horsepower and the 500 horsepower engines, respectively.
iii. CDPF Regeneration System Costs
    Application of CDPFs in nonroad applications is expected to present 
challenges beyond those of highway applications. For this reason, we 
anticipate that some additional hardware beyond the diesel particulate 
filter itself may be required to ensure that CDPF regeneration occurs. 
For some engines this may be new fuel control strategies that force 
regeneration under some circumstances, while in other engines it might 
involve an exhaust system fuel injector to inject fuel upstream of the 
CDPF to provide necessary heat for regeneration under some operating 
conditions. We estimate the near-term costs of a CDPF regeneration 
system to be $190 for a 150

[[Page 28444]]

horsepower engine and $320 for a 500 horsepower engine. In the long-
term, we estimate these costs at $140 and $240, respectively.
iv. Closed-Crankcase Ventilation System (CCV) Costs
    We are proposing to eliminate the exemption that allows turbo-
charged nonroad diesel engines to vent crankcase gases directly to the 
environment. Such engines are said to have an open crankcase system. We 
project that this requirement to close the crankcase on turbo-charged 
engines would force manufacturers to rely on engineered closed 
crankcase ventilation systems that filter oil from the blow-by gases 
prior to routing them into either the engine intake or the exhaust 
system upstream of the CDPF. We have estimated the initial cost of 
these systems to be roughly $40 for low horsepower engines and up to 
$100 for very high horsepower engines. These costs are incurred only by 
turbo-charged engines because today's naturally aspirated engines 
already have CCV systems.
v. Variable Costs for Engines Below 75 Horsepower and Above 750 
Horsepower
    This proposal includes standards for engines <25 horsepower that 
begin in 2008, and two sets of standards for 25 to 75 horsepower 
engines--one set that begins in 2008 and another that begins in 2013. 
The 2008 standards for all engines <75 horsepower are of similar 
stringency and are expected to result in similar technologies (i.e., 
the addition of a DOC). The 2013 standards for 25 to 75 horsepower 
engines are considerably more stringent than the 2008 standards and are 
expected to force the addition of a CDPF along with some other engine 
hardware to enable the proper functioning of that new technology. More 
detail on the mix of technologies expected for all engines <75 
horsepower is presented in section III. As discussed there, if changes 
are needed to comply, we expect manufacturers to comply with the 2008 
standards through either engine improvements or through the addition of 
a DOC. From a cost perspective, we have projected that engines would 
comply by either adding a DOC or by making some engine modifications 
resulting in engine-out emission reductions. Presumably, the 
manufacturer would choose the least costly approach that provided the 
necessary reduction. If engine-out modifications are less costly than a 
DOC, our estimate here is conservative. If the DOC proves to be less 
costly, then our estimate is representative of what most manufacturers 
would do. Therefore, we have assumed that, beginning in 2008, all 
engines below 75 horsepower add a DOC. Note that this is a conservative 
estimate in that we have assume this cost for all engines when, as 
discussed in section IV, some engines <75 horsepower already meet the 
proposed PM standards. We have estimated this added hardware to result 
in an increased engine cost of $150 in the near-term and $140 in the 
long-term for a 30 horsepower engine.
    We have also projected that some engines in the 25 to 75 horsepower 
range would have to upgrade their fuel systems to accommodate the CDPF. 
We have estimated the incremental costs for these fuel systems at 
roughly $740 in the 25-50 horsepower range, and around $430 in the 50-
75 horsepower range. This difference reflects a different base fuel 
system, with the smaller engines assumed to have mechanical fuel 
systems and the larger engines assumed to already be electronic. The 
electronic systems will incur lower costs because they already have the 
control unit and electronic fuel pump. Also, we have assumed these fuel 
changes would occur for only direct injection (DI) engines; indirect 
injection engines (IDI) are assumed to remain IDI but to add more 
hardware as part of their CDPF regeneration system to ensure proper 
regeneration under all operating conditions. Such a regeneration 
system, described above, is expected to cost roughly twice that 
expected for DI engines, or around $320 for a 30 horsepower IDI engine 
versus $160 for a DI engine.
    We have also projected that engines in the 25-50 horsepower range 
would add cooled EGR to comply with their new NOX standard. 
We have estimated that this would add $90 in the near-term and $70 in 
the long-term to the cost of a 30 horsepower engine.
    We believe there are factors that would cause variable hardware 
costs to decrease over time, making it appropriate to distinguish 
between near-term and long-term costs. Research in the costs of 
manufacturing has consistently shown that as manufacturers gain 
experience in production, they are able to apply innovations to 
simplify machining and assembly operations, use lower cost materials, 
and reduce the number or complexity of component parts.\289\ Our 
analysis, as described in more detail in the draft RIA, incorporates 
the effects of this learning curve by projecting that the variable 
costs of producing the low-emitting engines decreases by 20 percent 
starting with the third year of production. For this analysis, we have 
assumed a baseline that represents such learning already having 
occurred once due to the 2007 highway rule (i.e., a 20 percent 
reduction in emission control device costs is reflected in our near-
term costs). We have then applied a single learning step from that 
point in this analysis. We invite comment on this methodology to 
account for the learning curve phenomenon and also request comment on 
whether learning is likely to reduce costs even further in this 
industry (e.g., should a second learning step be applied to our near-
term costs?). Additionally, manufacturers are expected to apply ongoing 
research to make emission controls more effective and to have lower 
operating costs over time. However, because of the uncertainty involved 
in forecasting the results of this research, we conservatively have not 
accounted for it in this analysis.
---------------------------------------------------------------------------

    \289\ ``Learning Curves in Manufacturing,'' Linda Argote and 
Dennis Epple, Science, February 23, 1990, Vol. 247, pp. 920-924.
---------------------------------------------------------------------------

c. Engine Operating Costs
    We are projecting that a variety of new technologies will be 
introduced to enable nonroad engines to meet the proposed Tier 4 
emissions standards. Primary among these are advanced emission control 
technologies and low-sulfur diesel fuel. The technology enabling 
benefits of low-sulfur diesel fuel are described in section III, and 
the incremental cost for low-sulfur fuel is described in section V.A. 
The new emission control technologies are themselves expected to 
introduce additional operating costs in the form of increased fuel 
consumption and increased maintenance demands. Operating costs are 
estimated in the draft RIA over the life of the engine and are 
expressed in terms of cents/gallon of fuel consumed. In section V.C.3, 
we present these lifetime operating costs as a net present value (NPV) 
in 2001 dollars for several example pieces of equipment.
    Total operating cost estimates include the following elements: the 
change in maintenance costs associated with applying new emission 
controls to the engines; the change in maintenance costs associated 
with low sulfur fuel such as extended oil change intervals; the change 
in fuel costs associated with the incrementally higher costs for low 
sulfur fuel, and the change in fuel costs due to any fuel consumption 
impacts associated with applying new emission controls to the engines. 
This latter cost is attributed to the CDPF and its need for periodic 
regeneration which we estimate may result in a one percent fuel 
consumption increase where a NOX

[[Page 28445]]

adsorber is also applied, or a two percent fuel consumption increase 
where no NOX adsorber is applied (refer to chapter 6, 
section 6.2.3.3). Maintenance costs associated with the new emission 
controls on the engines are expected to increase since these devices 
represent new hardware and, therefore, new maintenance demands. For 
CDPF maintenance, we have used a maintenance interval of 3,000 hours 
for smaller engines and 4,500 hours for larger engines and a cost of 
$65 through $260 for each maintenance event. For closed-crankcase 
ventilation (CCV) systems, we have used a maintenance interval of 675 
hours for all engines and a cost per maintenance event of $8 to $48 for 
small to large engines. Offsetting these maintenance cost increases 
would be a savings due to an expected increase in oil change intervals 
because low sulfur fuel would be far less corrosive than is current 
nonroad diesel fuel. Less corrosion would mean a slower acidification 
rate (i.e., less degradation) of the engine lubricating oil and, 
therefore, more operating hours between needed oil changes. As 
discussed in section V.B, the use of 15 ppm sulfur fuel can extend oil 
change intervals by as much as 35 percent for both new and existing 
nonroad engines and equipment. We have used a 35 percent increase in 
oil change interval along with costs per oil change of $70 through $400 
to arrive at estimated savings associated with increased oil change 
intervals.
    These operating costs are expressed as a cent/gallon cost (or 
savings). As a result, operating costs are directly proportional to the 
amount of fuel consumed by the engine. We have estimated these 
operating costs, inclusive of fuel-related costs, to be 3.4 cents/
gallon for a 150 horsepower engine and 4.2 cents/gallon for a 500 
horsepower engine. More detail on operating costs can be found in 
chapter 6 of the draft RIA.
    The existing fleet will also benefit from lower maintenance costs 
due to the use of low sulfur diesel fuel. The operating costs for the 
existing fleet are discussed in Section V.B.
2. Equipment Cost Impacts
    In addition to the costs directly associated with engines that 
incorporate new emission controls to meet new standards, we expect cost 
increases due to the need to redesign the nonroad equipment in which 
these engines are used. Such redesigns would probably be necessary due 
to the expected addition of new emission control systems, but could 
also occur if the engine has a different shape or heat rejection rate, 
or is no longer made available in the configuration previously used. 
Based on their past experiences, equipment manufacturers have told EPA 
that a major concern with a new standard is their ability to redesign a 
large number of applications in a short period of time. Therefore, we 
have provided equipment manufacturers transition flexibility provisions 
to help them avoid business disruptions resulting from the changes 
associated with new emission standards. These flexibility provisions 
are presented in detail in Section III.E.4.
    In assessing the economic impact of the new emission standards, EPA 
has made a best estimate of the modifications to equipment that relate 
to packaging (installing engines in equipment engine compartments). The 
incremental costs for new equipment would be comprised of fixed costs 
(for redesign to accommodate new emission control devices) and variable 
costs (for new equipment hardware and for labor to install new emission 
control devices). Note that the fixed costs do not include 
certification costs, as did the engine fixed costs, because equipment 
is not certified to emission standards. We have attributed all changes 
in operating costs (e.g., additional maintenance) to the cost estimates 
for engines. Included in section V.C.3 is a discussion of several 
example pieces of equipment (e.g., skid/steer loader, dozer, etc.) and 
the costs we have estimated for these specific example pieces of 
equipment. Full details of our equipment cost analysis can be found in 
chapter 6 of the draft RIA. All costs are presented in 2001 dollars.
a. Equipment Fixed Costs
    The most significant changes anticipated for equipment redesign are 
changes to accommodate the physical changes to engines, especially for 
those engines that add PM traps and NOX adsorbers. The costs 
for engine development and the emission control devices are included as 
costs to the engines, as described above. What remains to be quantified 
for equipment manufacturers is the effort to integrate the engine and 
emissions control devices into the overall functioning of the 
equipment. What remains to be quantified for equipment manufacturers is 
the effort to integrate the engine and emissions control devices into 
the overall functioning of the equipment. We have allocated extensive 
engineering time for this effort.
    The costs we have estimated are based on engine power and whether 
an application is non-motive (e.g., a generator set) or motive (e.g., a 
skid steer loader). The designs we have considered to be non-motive are 
those that lack a propulsion system. In addition, the proposed emission 
standards for engines rated under 25 horsepower and the proposed 2008 
standards for 25-75 horsepower engines are projected to require no 
significant equipment redesign beyond that done to accommodate the Tier 
2 standards. We expect that these engines would comply with the 
proposesd Tier 4 standards through either engine modifications to 
reduce engine-out emissions or through the addition of a DOC. We have 
projected that engine modifications would not affect the outer 
dimensions of the engine and that a DOC would replace the existing 
muffler. Therefore, either approach taken by the engine manufacturer 
should have minimal to no impact on the equipment design. Nonetheless, 
we have conservatively estimated their redesign costs at $50,000 per 
model.
    A number of equipment manufacturers have shared detailed 
information with us regarding the investments made for Nonroad Tier 2 
equipment redesign efforts, as well as redesign estimates for 
significant changes such as installing a new engine design. These 
estimates range from approximately $50,000 for some lower powered 
equipment models to well over $1 million dollars for high horsepower 
equipment with very challenging design constraints. Based on that 
input, for the proposed Tier 4 standards, we have estimated that 
equipment redesign costs would range from $50,000 per model for 25 
horsepower equipment up to $750,000 per model for 300 horsepower 
equipment and above. We have attributed only a portion of the equipment 
redesign costs to U.S. sales in a manner consistent with that taken for 
engine R&D costs and engine tooling costs. In addition, we expect 
manufacturers to incur some fixed costs to update service and operation 
manuals to address the maintenance demands of new emission control 
technologies and the new oil service intervals which we estimate to be 
between $2,500 and $10,000 per equipment model.
    These equipment fixed costs (redesign and manual updates) were then 
allocated appropriately to each new model to arrive at a total 
equipment fixed cost of $697 million. We have assumed that these costs 
would be recovered over a ten year period at a seven percent interest 
rate.
b. Equipment Variable Costs
    Equipment variable cost estimates are based on costs for additional 
materials to mount the new hardware (i.e., brackets and bolts required 
to secure the

[[Page 28446]]

aftertreatment devices) and additional sheet metal assuming that the 
body cladding of a piece of equipment (i.e., the hood) might change to 
accommodate the aftertreatment system. Variable costs also include the 
labor required to install these new pieces of hardware. For engines 
75 horsepower--those expected to incorporate CDPF and 
NOX adsorber technology--the amount of sheet metal is based 
on the size of the aftertreatment devices.
    For equipment of 150 horsepower and 500 horsepower, respectively, 
we have estimated the costs to be roughly $60 to $140. Note that we 
have estimated costs for equipment in all horsepower ranges, and these 
estimates are presented in detail in the draft RIA. Throughout this 
discussion of engine and equipment costs, we present costs for a 150 
and a 500 horsepower engine for illustrative purposes.
3. Overall Engine and Equipment Cost Impacts
    To illustrate the engine and equipment cost impacts we are 
estimating for the proposed standards, we have chosen several example 
pieces of equipment and presented the estimated costs for them. Using 
these examples, we can calculate the costs for a specific piece of 
equipment in several horsepower ranges and better illustrate the cost 
impacts of the proposed standards. These costs along with information 
about each example piece of equipment are shown in Table V.C-1. Costs 
presented are near-term and long-term costs for the final standards to 
which each piece of equipment would comply. Long-term costs are only 
variable costs and, therefore, represent costs after all fixed costs 
have been recovered and all projected learning has taken place. 
Included in the table are estimated prices for each piece of equipment 
to provide some perspective on how our estimated control costs relate 
to existing equipment prices.

                                   Table V.C-1--Near-Term and Long-Term Costs for Several Example Pieces of Equipmenta
                                      ($2001, for the final emission standards to which the equipment must comply)
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                            Skid/steer                                                                      Off-highway
                                              GenSet          loader          Backhoe          Dozer        Ag tractor         Dozer           truck
--------------------------------------------------------------------------------------------------------------------------------------------------------
Horsepower                                          9 hp           33 hp           76 hp          175 hp          250 hp          503 hp        1,000 hp
Incremental engine & equipment cost
  Long-term                                         $120            $760          $1,210          $2,590          $2,000          $4,210          $6,780
  Near-term                                         $170          $1,100          $1,680           3,710          $2,950          $6,120         $10,100
Estimated equipment price when new b              $3,500         $13,500         $50,000        $235,000        $130,000        $575,000        $700,000
Incremental operating costs c                       -$90             $40            $370          $1,550          $1,320          $4,950         $12,550
Baseline operating costs (fuel & oil                $940          $2,680          $7,960         $77,850         $23,750         $77,850       $179,530
 only) c
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes:
a Near-term costs include both variable costs and fixed costs; long-term costs include only variable costs and represent those costs that remain
  following recovery of all fixed costs.
b ``Estimated Price of New Nonroad Example Equipment,'' memorandum from Zuimdie Guerra to docket A-2001-28.
c Present value of lifetime costs.

    More detail and discussion regarding what these costs and prices 
mean from an economic impact perspective can be found in section V.E.

D. Annual Costs and Cost Per Ton

    One tool that can be used to assess the value of the proposed 
standards for nonroad fuel and engines is the costs incurred per ton of 
emissions reduced. This analysis involves a comparison of our proposed 
program to other measures that have been or could be implemented.
    We have calculated the cost per ton of our proposed program based 
on the net present value of all costs incurred and all emission 
reductions generated over a 30 year time window following 
implementation of the program. This approach captures all of the costs 
and emissions reductions from our proposed program including those 
costs incurred and emissions reductions generated by the existing 
fleet. The baseline (i.e., the point of comparison) for this evaluation 
is the existing set of fuel and engine standards (i.e., unregulated 
fuel and the Tier 2/Tier 3 program). The 30 year time window chosen is 
meant to capture both the early period of the program when very few new 
engines that meet the proposed standards would be in the fleet, and the 
later period when essentially all engines would meet the proposed 
standards.
    As discussed in section IV, the proposal contains two separate fuel 
programs. We are proposing a 500 ppm sulfur cap on nonroad, locomotive, 
and marine fuels beginning in 2007. This fuel program, the first step 
in our two step fuel program, provides significant air quality benefits 
through reduced SO2 and PM emissions from both new and 
existing nonroad, locomotive, and marine engines. In sections V.D.1 and 
2, we summarize the cost for this program as if it remained in place 
for 30 years, even though it would be supplanted by the second step of 
our fuel program in 2010. We also provide an analysis of the cost per 
ton for the SO2 reductions that would be realized by the 500 
ppm fuel program for the same 30 year time window. In this way, the 
cost per ton of the SO2 reductions realized by the 500 ppm 
fuel program can be compared to other available means to control 
SO2 emissions. The significant PM reductions are not 
accounted for in the relative cost per ton estimate, but are accounted 
for in our inventory analysis presented in section II and in the 
benefits analysis presented later in this section. Additional detail 
regarding all of the estimates presented here are available in the 
draft RIA.
    We are proposing a second step in the fuel program that would cap 
nonroad fuel sulfur levels at 15 ppm beginning in 2010. This fuel 
program enables the introduction of advanced emission control 
technologies including CDPFs and NOX adsorbers. The 
combination of the two-step fuel program and the new diesel engine 
standards represents the total Tier 4 program for nonroad diesel 
engines and fuel proposed today. In sections V.D.3 and 4, we present 
our estimate of the annual and total costs for

[[Page 28447]]

this complete program beginning in 2007 and continuing for 30 years. 
Also included is an estimate of the cost per ton of emissions 
reductions realized by this program for NMHC+NOX, PM, and 
SO2.
1. Annual Costs for the 500 ppm Fuel Program
    Cent per gallon costs for the proposed 500 ppm fuel program (i.e., 
the reduction to a 500 ppm sulfur cap) were presented in section V.A. 
Having this fuel would result in maintenance savings associated with 
increased oil change intervals for both the new and the existing fleet 
of nonroad, locomotive, and marine engines. These maintenance savings 
were discussed in section V.B. There are no engine and equipment costs 
associated with the 500 ppm fuel program because new emission standards 
are not part of that proposed program. Figure V.D-1 shows the annual 
costs associated with the 500 ppm fuel program.
    As can be seen in Figure V.D-1, the costs for refining and 
distributing the 500 ppm fuel range from $250 million in 2008 to nearly 
$400 million in 2036. These control costs are largely offset by the 
maintenance savings that range from $200 million in 2008 to $380 
million in 2036. Despite the fact that the costs of the 500 ppm fuel 
for nonroad diesel fuel is 2.5 cents/gallon and the maintenance savings 
are 3 cents per gallon, the net costs are positive because of the costs 
for the locomotive and marine fuel is not off-set by the maintenance 
savings. As a whole, the net cost of the program in each year is 
essentially zero, ranging from $50 million in the early years to only 
$18 million in 2036. The net present value of the net costs and savings 
associated with the proposed 500 ppm fuel program during the years 2007 
to 2036 is estimated at $510 million.
[GRAPHIC] [TIFF OMITTED] TP23MY03.009

2. Cost Per Ton for the 500 ppm Fuel Program
    The 2007 fuel program would result in large reductions of both 
SO2 and PM emissions. Roughly 98 percent of fuel sulfur is 
converted to SO2 in the engine with the remaining two 
percent being exhausted as sulfate PM. Because the majority of the 
emissions reductions associated with this program would be 
SOX, we have attributed all the control costs to 
SOX in calculating the cost per ton associated with this 
program. However, we have modeled both the SOX and PM 
reductions so that our inventory and benefits analysis fully account 
for them.
    As noted above, we have calculated both the costs and emission 
reductions of the 500 ppm fuel program as if it were to remain in place 
indefinitely. Figure V.D-1 shows the costs in each year of the program, 
the net present value of which is estimated at $510 million. We have 
estimated the 30 year net present value of the SOX emission 
reductions at 5.6 million tons.
    Table V.D-1 shows the cost per ton of emissions reduced as a result 
of the proposed 500 ppm fuel program. The cost per ton numbers include 
costs and emission reductions that would occur from both the new and 
the existing fleet (i.e., those pieces of nonroad equipment that were 
sold into the market prior to the proposed emission standards) of

[[Page 28448]]

nonroad, locomotive, and marine engines.

 Table V.D-1--500 ppm Fuel Program Aggregate Cost per Ton and Long-Term
                       Annual Cost per Ton ($2001)
------------------------------------------------------------------------
                                                 2004-2036
                                                 Discounted   Long-term
                   Pollutant                      lifetime     cost per
                                                  cost per   ton in 2036
                                                    ton
------------------------------------------------------------------------
SOX...........................................          $90          $50
------------------------------------------------------------------------

    We also considered the cost per ton of the 500 ppm fuel program 
without taking credit for the expected maintenance savings associated 
with low sulfur fuel. Without the maintenance savings, the cost per ton 
of SOX reduced would be $990 per ton for each year of the 
program. More detail on how the costs and cost per ton numbers 
associated with the 500 ppm fuel program were calculated can be found 
in the draft RIA.
3. Annual Costs for the Proposed Two-Step Fuel Program and Engine 
Program
    The costs of the total proposed engine and fuel program include 
costs associated with both steps in the fuel program--the reduction to 
500 ppm sulfur in 2007 and the reduction to 15 ppm sulfur in 2010. Also 
included are costs for the proposed 2008 engine standards for <75 
horsepower engines, the proposed 2013 standards for 25 to 75 horsepower 
engines, and costs for the proposed engine standards for 75 
horsepower engines. Included are all maintenance costs and savings 
realized by both the existing fleet (nonroad, locomotive, and marine) 
and the new fleet of engines complying with the proposed standards.
    Figure V.D-2 presents these results. All capital costs for fuel 
production and engine and equipment fixed costs have been amortized. 
The figure shows that total annual costs are estimated to be $120 
million in the first year the new engine standards apply, increasing to 
a peak of $1.7 billion in 2036 as increasing numbers of engines become 
subject to the new standards and an ever increasing amount of fuel is 
consumed. The net present value of the annualized costs over the period 
from 2007 to 2036 is $20.7 billion.
[GRAPHIC] [TIFF OMITTED] TP23MY03.010

4. Cost per Ton of Emissions Reduced for the Total Program
    We have calculated the cost per ton of emissions reduced associated 
with the proposed engine and fuel program. We have done this using the 
net present value of the annualized costs of the program through 2036 
and the net present value of the annual emission reductions through 
2036. We have also calculated the cost per ton of emissions in the year 
2036 using the annual costs

[[Page 28449]]

and emission reductions in that year alone. This number represents the 
long-term cost per ton of emissions reduced after all fixed costs of 
the program have been recovered by industry leaving only the variable 
costs of control. The cost per ton numbers include costs and emission 
reductions that would occur from the existing fleet (i.e., those pieces 
of nonroad equipment that were sold into the market prior to the 
proposed emission standards). These results are shown in Table V.D-2. 
We did the cost analysis using a 3% discount rate. We will also be 
conducting a similar analysis using a 7% discount rate and including 
this information in the docket.

 Table V.D-2--Total Proposed Fuel and Engine Program Aggregate Cost per
              Ton and Long-Term Annual Cost Per Ton ($2001)
------------------------------------------------------------------------
                                                 2004-2036
                                                 Discounted   Long-term
                   Pollutant                      lifetime     cost per
                                                  cost per   ton in 2036
                                                    ton
------------------------------------------------------------------------
NOX+NMHC......................................         $810         $530
PM............................................        8,700        6,900
SOX...........................................      \a\ 200         170
------------------------------------------------------------------------
Notes:
\a\ This result does not match that in Table 8.4-2 because the nonroad
  portion of the fuel is reduced to 15 ppm and does not stay at 500
  (locomotive and marine portions are kept at 500ppm). The costs to
  reduce fuel sulfur from uncontrolled to 15ppm were assigned 50/50 to
  NOX+NMHC and PM for the reduction to 15 ppm is to enable
  aftertreatment technology.

5. Comparison With Other Means of Reducing Emissions
    In comparison with other programs to control these pollutants, we 
believe that the proposed programs represent a cost effective strategy 
for generating substantial NOX+NMHC, PM, and SO2 
reductions. This can be seen by comparing the 2007 fuel program (i.e., 
a sulfur cap of 500 ppm) cost per ton and the total program cost per 
ton with a number of standards that EPA has adopted in the past. Table 
V.D-3 summarizes the cost per ton of several past EPA actions for 
NOX+NMHC. Table V.D-4 summarizes the cost per ton of several 
past EPA actions for PM.

 Table V.D-3--Cost per Ton of Previous Mobile Source Programs for NOX +
                                  NMHC
------------------------------------------------------------------------
                         Program                               $/ton
------------------------------------------------------------------------
Tier 2 Nonroad Diesel...................................       630
Tier 3 Nonroad Diesel...................................       430
Tier 2 vehicle/gasoline sulfur..........................  1,410-2,370
2007 Highway HD.........................................     2,260
2004 Highway HD.........................................   220-430
Off-highway diesel engine...............................   450-710
Tier 1 vehicle..........................................  2,160-2,930
NLEV....................................................      2030
Marine SI engines.......................................  1,230-1,940
On-board diagnostics....................................     2,430
Marine CI engines.......................................   30-190
------------------------------------------------------------------------
Note: Costs adjusted to 2001 dollars using the Producer Price Index for
  Total Manufacturing Industries.


  Table V.D-4.--Cost per Ton of Previous Mobile Source Programs for PM
------------------------------------------------------------------------
                        Program                               $/ton
------------------------------------------------------------------------
Tier 1/Tier 2 Nonroad Diesel...........................     2,410
2007 Highway HD........................................    14,280
Marine CI engines......................................  5,480-4,070
1996 urban bus.........................................  12,870-20,590
Urban bus retrofit/rebuild.............................    31,740
1994 highway HD diesel.................................  21,930-25,670
------------------------------------------------------------------------
Note: Costs adjusted to 2001 dollars using the Producer Price Index for
  Total Manufacturing Industries.

    To compare the cost per ton of SO2 emissions reduced, we 
looked at the cost per ton for the Title IV SO2 trading 
programs. This information is found in EPA report 430/R-02-004, 
``Documentation of EPA Modeling Applications (V.2.1) Using the 
Integrated Planning Model'', in Figure 9.11 on page 9-14 (www.epa.gov/
airmarkets/epa-ipm/index.html#documentation). The SO2 cost 
per ton results of the proposed program presented in Table V.D-2 
compare very favorably with the program shown in Table V.D-5.

 Table V.D-5--Cost per Ton of SO2 From EPA Base Case 2000 for the Title
                         IV SO2 Trading Programs
------------------------------------------------------------------------
                  Program                               $/ton
------------------------------------------------------------------------
Title IV SO2 Trading Programs.............  $490 in 2010 to $610 in
                                             2020.
------------------------------------------------------------------------
Note: Costs adjusted to 2001 dollars using the Producer Price Index for
  Total Manufacturing Industries.

E. Do the Benefits Outweigh the Costs of the Standards?

    Our analysis of the health and welfare benefits to be expected from 
this proposal are presented in this section. Briefly, the analysis 
projects major benefits throughout the period from initial 
implementation of the rule through 2030, the last year analyzed. As 
described below, thousands of deaths and other serious health effects 
would be prevented, yielding a net present value in 2004 of those 
benefits we could monetize of approximately $550 billion dollars. These 
benefits exceed the net present value of the social cost of the 
proposal ($17 billion) by a factor of over 30 to one.
1. What Were the Results of the Benefit-Cost Analysis?
    Table V.E-1 presents the primary estimate of reduced incidence of 
PM-related health effects for the years 2020 and 2030. In interpreting 
the results, it is important to keep in mind the limited set of effects 
we are able to monetize. Specifically, the table lists the PM-related 
benefits associated with the reduction of several health effects.\290\ 
In 2030, we estimate that there will be 9,600 fewer fatalities per year 
associated with fine PM, and the rule will result in about 5,700 fewer 
cases of chronic bronchitis, 8,300 fewer hospitalizations (for 
respiratory and cardiovascular disease combined), and result in 
significant reductions in days of restricted activity due to 
respiratory illness (with an estimated 5.7 million fewer cases). We 
also estimate substantial health improvements for children from reduced 
upper and lower respiratory illness, acute bronchitis, and asthma 
attacks.\291\
---------------------------------------------------------------------------

    \290\ Based upon recent preliminary findings by the Health 
Effects Institute, the concentration-response functions used to 
estimate reductions in hospital admissions may over or underestimate 
the true concentration-response relationship. See letter from Dan 
Greenberg, President, Health Effects Institute, May 30, 2002, 
attached to letter from Dr. Hopke, dated August 8, 2002. Docket A-
2000-01, Document IV-A-145.
    \291\ Our estimate incorporates significant reductions of 
150,000 fewer cases of lower respiratory symptoms in children ages 7 
to 14 each year, 110,000 fewer cases of upper respiratory symptoms 
(similar to cold symptoms) in asthmatic children each year, and 
14,000 fewer cases of acute bronchitis in children ages 8 to 12 each 
year. In addition, we estimate that this rule will reduce almost 
6,000 emergency room visits for asthma attacks in children each year 
from reduced exposure to particles. Additional incidents would be 
avoided from reduced ozone exposures. Asthma is the most prevalent 
chronic disease among children and currently affects over seven 
percent of children under 18 years of age.
---------------------------------------------------------------------------

    Table V.E-2 presents the total monetized benefits for the years 
2020 and 2030. This table also indicates with a ``B'' those additional 
health and environmental effects which we were unable to quantify or 
monetize. These effects are additive to estimate of total benefits, and 
EPA believes there is

[[Page 28450]]

considerable value to the public of the benefits that could not be 
monetized. A full listing of the benefit categories that could not be 
quantified or monetized in our estimate are provided in Table V.E-5.
    In summary, EPA's primary estimate of the benefits of the rule are 
approximately $81 + B billion in 2030. In 2020, total monetized 
benefits are approximately $43 + B billion. These estimates account for 
growth in real gross domestic product (GDP) per capita between the 
present and the years 2020 and 2030. As the table indicates, total 
benefits are driven primarily by the reduction in premature fatalities 
each year, which account for over 90 percent of total benefits.

   Table V.E-1.--Reductions in Incidence of PM-Related Adverse Health
   Effects Associated With the Proposed Nonroad Diesel Engine and Fuel
                                Standards
------------------------------------------------------------------------
                                          Avoided incidence \a\  (cases/
                                                       year)
                Endpoint                 -------------------------------
                                               2020            2030
------------------------------------------------------------------------
Premature mortality \b\--Base estimate:            5,200           9,600
 Long-term exposure (adults, 30 and
 over)..................................
Chronic bronchitis (adults, 26 and over)           3,600           5,700
Non-fatal myocardial infarctions                   9,200          16,000
 (adults, 18 and older).................
Hospital admissions--Respiratory                   2,400           4,500
 (adults, 20 and older) \c\.............
Hospital admissions--Cardiovascular                1,900           3,800
 (adults, 20 and older) \d\.............
Emergency Room Visits for Asthma (18 and           3,600           5,700
 younger)...............................
Acute bronchitis (children, 8-12).......           8,400          14,000
Lower respiratory symptoms (children, 7-          92,000         150,000
 14)....................................
Upper respiratory symptoms (asthmatic             77,000         110,000
 children, 9-11)........................
Work loss days (adults, 18-65)..........         650,000         960,000
Minor restricted activity days (adults,        3,900,000      5,700,000
 age 18-65).............................
------------------------------------------------------------------------
Notes:
\a\ Incidences are rounded to two significant digits.
\b\ Premature mortality associated with ozone is not separately included
  in this analysis
\c\ Respiratory hospital admissions for PM includes admissions for COPD,
  pneumonia, and asthma.
\d\ Cardiovascular hospital admissions for PM includes total
  cardiovascular and subcategories for ischemic heart disease,
  dysrhythmias, and heart failure.


     Table V.E-2.--EPA Primary Estimate of the Annual Quantified and
  Monetized Benefits Associated With Improved PM Air Quality Resulting
       From the Proposed Nonroad Diesel Engine and Fuel Standards
------------------------------------------------------------------------
                                             Monetary Benefits\a,\ \b\
                                           (millions 2000$, adjusted for
                Endpoint                          income growth)
                                         -------------------------------
                                               2020            2030
------------------------------------------------------------------------
Premature mortality \c\ Long-term                $39,000         $74,000
 exposure (adults, 30 and over).........
Chronic bronchitis (WTP valuation;                 1,600           2,600
 adults, 26 and over)...................
Non-fatal myocardial infarctions........             750           1,300
Hospital Admissions from Respiratory                  38              74
 Causes \d\.............................
Hospital Admissions from Cardiovascular               40              80
 Causes \e\.............................
Emergency Room Visits for Asthma........               1               2
Acute bronchitis (children, 8-12).......               3               5
Lower respiratory symptoms (children, 7-               2               3
 14)....................................
Upper respiratory symptoms (asthmatic                  2               3
 children, 9-11)........................
Work loss days (adults, 18-65)..........              90             130
Minor restricted activity days (adults,              210             320
 age 18-65).............................
Recreational visibility (86 Class I                1,200           1,900
 Areas).................................
                                         -----------------
    Total Monetized Benefits \f\........      43,000 + B     81,000 + B
------------------------------------------------------------------------
Notes:
\a\ Monetary benefits are rounded to two significant digits.
\b\ Monetary benefits are adjusted to account for growth in real GDP per
  capita between 1990 and the analysis year (2020 or 2030).
\c\ Valuation assumes the 5 year distributed lag structure described
  earlier. Results reflect the use of two different discount rates; a 3%
  rate which is recommended by EPA's Guidelines for Preparing Economic
  Analyses (US EPA, 2000a), and 7% which is recommended by OMB Circular
  A-94 (OMB, 1992).
\d\ Respiratory hospital admissions for PM includes admissions for COPD,
  pneumonia, and asthma.
\e\ Cardiovascular hospital admissions for PM includes total
  cardiovascular and subcategories for ischemic heart disease,
  dysrhythmias, and heart failure.
\f\ B represents the monetary value of the unmonetized health and
  welfare benefits. A detailed listing of unquantified PM, ozone, CO,
  and NMHC related health effects is provided in Table V.E-5.

    The estimated social cost (measured as changes in consumer and 
producer surplus) in 2030 to implement the final rule from Table V.F-2 
is $1.5 billion (2000$). Thus, the net benefit (social benefits minus 
social costs) of the program at full implementation is approximately 
$79 + B billion. In 2020, partial implementation of the program yields 
net benefits of $42 + B billion. Therefore, implementation of the final 
rule is expected to provide society with a net gain in social welfare 
based on economic efficiency criteria. Table V.E-3 presents a summary 
of the benefits,

[[Page 28451]]

costs, and net benefits of the proposed rule. Figure VE.1 displays the 
stream of benefits, costs, and net benefits of the Nonroad Land-based 
Diesel Vehicle Rule from 2007 to 2030. In addition, Table V-E.4 
presents the net present value of the stream of benefits, costs, and 
net benefits associated with the rule for this 23 year period (using a 
three percent discount rate). The total net present value in 2004 of 
the stream of net benefits (benefits minus costs) is $530 billion.

    Table V.E-3.--Summary of Benefits, Costs, and Net Benefits of the Proposed Nonroad Diesel Engine and Fuel
                                                    Standards
----------------------------------------------------------------------------------------------------------------
                                            2020 \a\  (billions of 2000          2030 \a\  (billions of 2000
                                                      dollars)                             dollars)
----------------------------------------------------------------------------------------------------------------
Social Costs \b\......................  $1.4...............................  $1.5.
Social Benefits \b,\ \c,\ \d\:
    CO, VOC, Air Toxic-related          Not monetized......................  Not monetized.
     benefits.
    Ozone-related benefits............  Not monetized......................  Not monetized.
    PM-related Welfare benefits.......  $1.2...............................  $1.9.
    PM-related Health benefits........  $42+ B.............................  $79 + B.
    Net Benefits (Benefits-Costs) \c\.  $42 + B............................  $79 + B.
----------------------------------------------------------------------------------------------------------------
Notes:
\a\ All costs and benefits are rounded to two significant digits.
\b\ Note that costs are the total costs of reducing all pollutants, including CO, VOCs and air toxics, as well
  as NOX and PM. Benefits in this table are associated only with PM, NOX and SO3 reductions.
\c\ Not all possible benefits or disbenefits are quantified and monetized in this analysis. Potential benefit
  categories that have not been quantified and monetized are listed in Table V.E-5. B is the sum of all
  unquantified benefits and disbenefits.


[[Page 28452]]

[GRAPHIC] [TIFF OMITTED] TP23MY03.011


   Table V.E-4.--Net Present Value in 2004 of the Stream of Benefits,
 Costs, and Net Benefits for the Proposed Nonroad Diesel Engine and Fuel
                                Standards
                           [Billions of 2000$]
------------------------------------------------------------------------
 
------------------------------------------------------------------------
Social Costs............................................             $17
Social Benefits.........................................             550
Net Benefits............................................         \a\ 530
------------------------------------------------------------------------
Notes:
\a\ Numbers do not add due to rounding.

2. What Was Our Overall Approach to the Benefit-Cost Analysis?
    The basic question we sought to answer in the benefit-cost analysis 
was, ``What are the net yearly economic benefits to society of the 
reduction in mobile source emissions likely to be achieved by this 
proposed rulemaking?'' In designing an analysis to address this 
question, we selected two future years for analysis (2020 and 2030) 
that are representative of the stream of benefits and costs at partial 
and full-implementation of the program.
    To quantify benefits, we evaluated PM-related health effects 
(including directly emitted PM, SO3, and NOX 
contributions to fine particulate matter). Our approach requires the 
estimation of changes in air quality expected from the rule and then 
estimating the resulting impact on health. In order to characterize the 
benefits of today's action, given the constraints on time and resources 
available for the analysis, we adopted a benefits transfer technique 
that relies on air quality and benefits modeling for a preliminary 
control option for nonroad diesel engines and fuels. Results from the 
modeled preliminary control option in 2020 and 2030 are then scaled and 
transferred to the emission reductions expected from the proposed rule. 
We also transferred modeled results by using scaling factors associated 
with time to examine the stream of benefits in years other than 2020 
and 2030.
    More specifically, our health benefits assessment is conducted in 
two phases. Due to the time requirements for running the sophisticated 
emissions and air quality models needed to obtain estimates of the 
benefits expected to result from implementation of the rule, it is 
often necessary to select an example set of emission reductions to use 
for the purposes of emissions and air quality modeling. In phase one, 
we evaluate the PM and ozone related health effects associated with a 
modeled preliminary control option that was a close approximation of 
the proposed standards in the years 2020 and 2030. Using information 
from the modeled preliminary control option on the changes in ambient 
concentrations of PM and ozone, we then conduct a

[[Page 28453]]

health assessment to estimate the number of reduced incidences of 
illnesses, hospitalizations, and premature fatalities associated with 
this scenario and estimate the total economic value of these health 
benefits. The standards we are proposing in this rulemaking, however, 
are slightly different in the amount of emission reductions expected to 
be achieved in 2020 and 2030 relative to the modeled scenario. Thus, in 
phase two of the analysis we apportion the results of the phase one 
analysis to the underlying NOX, SO3, and PM 
emission reductions and scale the apportioned benefits to reflect 
differences in emissions reductions between the modeled preliminary 
control option and the proposed standards. The sum of the scaled 
benefits for the PM, SO3, and NOX emission 
reductions provide us with the total benefits of the rule.
    The benefit estimates derived from the modeled preliminary control 
option in phase one of our analysis uses an analytical structure and 
sequence similar to that used in the benefits analyses for the Heavy 
Duty Engine/Diesel Fuel final rule and in the ``section 812 studies'' 
to estimate the total benefits and costs of the full Clean Air 
Act.\292\ We used many of the same models and assumptions used in the 
Heavy Duty Engine/Diesel Fuel analysis as well as other Regulatory 
Impact Analyses (RIAs) prepared by the Office of Air and Radiation. By 
adopting the major design elements, models, and assumptions developed 
for the section 812 studies and other RIAs, we have largely relied on 
methods which have already received extensive review by the independent 
Science Advisory Board (SAB), by the public, and by other federal 
agencies. In addition, we will be working through the next section 812 
study process to enhance our methods.\293\ Interested parties will 
therefore be able to obtain further information from the section 812 
study on the kinds of methods we are likely to use for estimating 
benefits and costs in the final nonroad diesel rule.
---------------------------------------------------------------------------

    \292\ The section 812 studies include: (1) US EPA, Report to 
Congress: The Benefits and Costs of the Clean Air Act, 1970 to 1990, 
October 1997 (also known as the ``Section 812 Retrospective 
Report''); and (2) the first in the ongoing series of prospective 
studies estimating the total costs and benefits of the Clean Air Act 
(see EPA report number: EPA-410-R-99-001, November 1999). See Docket 
A-99-06, Document II-A-21.
    \293\ We anticipate a public SAB meeting June 11-13, 2003, in 
Washington, DC, regarding components of our analytical blueprint. 
Interested parties may want to consult the Web page: http://
www.epa.gov/science1.
---------------------------------------------------------------------------

    The benefits transfer method used in phase two of the analysis is 
similar to that used to estimate benefits in the recent analysis of the 
Nonroad Large Spark-Ignition Engines and Recreational Engines standards 
(67 FR 68241, November 8, 2002). A similar method has also been used in 
recent benefits analyses for the proposed Industrial Boilers and 
Process Heaters NESHAP and the Reciprocating Internal Combustion 
Engines NESHAP.
    On September 26, 2002, the National Academy of Sciences (NAS) 
released a report on its review of the Agency's methodology for 
analyzing the health benefits of measures taken to reduce air 
pollution. The report focused on EPA's approach for estimating the 
health benefits of regulations designed to reduce concentrations of 
airborne particulate matter (PM).
    In its report, the NAS said that EPA has generally used a 
reasonable framework for analyzing the health benefits of PM-control 
measures. It recommended, however, that the Agency take a number of 
steps to improve its benefits analysis. In particular, the NAS stated 
that the Agency should:
    [sbull] Include benefits estimates for a range of regulatory 
options;
    [sbull] Estimate benefits for intervals, such as every five years, 
rather than a single year;
    [sbull] Clearly state the projected baseline statistics used in 
estimating health benefits, including those for air emissions, air 
quality, and health outcomes;
    [sbull] Examine whether implementation of proposed regulations 
might cause unintended impacts on human health or the environment;
    [sbull] When appropriate, use data from non-U.S. studies to broaden 
age ranges to which current estimates apply and to include more types 
of relevant health outcomes;
    [sbull] Begin to move the assessment of uncertainties from its 
ancillary analyses into its Base analyses by conducting probabilistic, 
multiple-source uncertainty analyses. This assessment should be based 
on available data and expert judgment.
    Although the NAS made a number of recommendations for improvement 
in EPA's approach, it found that the studies selected by EPA for use in 
its benefits analysis were generally reasonable choices. In particular, 
the NAS agreed with EPA's decision to use cohort studies to derive 
benefits estimates. It also concluded that the Agency's selection of 
the American Cancer Society (ACS) study for the evaluation of PM-
related premature mortality was reasonable, although it noted the 
publication of new cohort studies that should be evaluated by the 
Agency.
    EPA has addressed many of the NAS comments in our analysis of the 
proposed rule. We provide benefits estimates for each year over the 
rule implementation period for a wide range of regulatory alternatives, 
in addition to our proposed emission control program. We use the 
estimated time path of benefits and costs to calculate the net present 
value of benefits of the rule. In the RIA, we provide baseline 
statistics for air emissions, air quality, population, and health 
outcomes. We have examined how our benefits estimates might be impacted 
by expanding the age ranges to which epidemiological studies are 
applied, and we have added several new health endpoints, including non-
fatal heart attacks, which are supported by both U.S. studies and 
studies conducted in Europe. We have also improved the documentation of 
our methods and provided additional details about model assumptions.
    Several of the NAS recommendations addressed the issue of 
uncertainty and how the Agency can better analyze and communicate the 
uncertainties associated with its benefits assessments. In particular, 
the Committee expressed concern about the Agency's reliance on a single 
value from its analysis and suggested that EPA develop a probabilistic 
approach for analyzing the health benefits of proposed regulatory 
actions. The Agency agrees with this suggestion and is working to 
develop such an approach for use in future rulemakings. EPA plans to 
hold a meeting of its Science Advisory Board (SAB) in early Summer 2003 
to review its plans for addressing uncertainty in its analyses. Our 
likely approach will incorporate short-term elements intended to 
provide interim methods in time for the final Nonroad rule to address 
uncertainty in important analytical parameters such as the 
concentration-response relationship for PM-related premature mortality. 
Our approach will also include longer-term elements intended to provide 
scientifically sound, peer-reviewed characterizations of the 
uncertainty surrounding a broader set of analytical parameters and 
assumptions, including but not limited to emissions and air quality 
modeling, demographic projections, population health status, 
concentration-response functions, and valuation estimates.
3. What Are the Significant Limitations of the Benefit-Cost Analysis?
    Every benefit-cost analysis examining the potential effects of a 
change in

[[Page 28454]]

environmental protection requirements is limited to some extent by data 
gaps, limitations in model capabilities (such as geographic coverage), 
and uncertainties in the underlying scientific and economic studies 
used to configure the benefit and cost models. Deficiencies in the 
scientific literature often result in the inability to estimate 
quantitative changes in health and environmental effects, such as 
potential increases in premature mortality associated with increased 
exposure to carbon monoxide. Deficiencies in the economics literature 
often result in the inability to assign economic values even to those 
health and environmental outcomes which can be quantified. While these 
general uncertainties in the underlying scientific and economics 
literatures, which can cause the valuations to be higher or lower, are 
discussed in detail in the Regulatory Support Document and its 
supporting documents and references, the key uncertainties which have a 
bearing on the results of the benefit-cost analysis of this final rule 
include the following:
    [sbull] The exclusion of potentially significant benefit categories 
(such as health and ecological benefits of reduction in CO, VOCs, air 
toxics, and ozone);
    [sbull] Errors in measurement and projection for variables such as 
population growth;
    [sbull] Uncertainties in the estimation of future year emissions 
inventories and air quality;
    [sbull] Uncertainties associated with the scaling of the results of 
the modeled benefits analysis to the proposed standards, especially 
regarding the assumption of similarity in geographic distribution 
between emissions and human populations and years of analysis;
    [sbull] Variability in the estimated relationships of health and 
welfare effects to changes in pollutant concentrations;
    [sbull] Uncertainties in exposure estimation;
    [sbull] Uncertainties associated with the effect of potential 
future actions to limit emissions.
    Despite these uncertainties, we believe the benefit-cost analysis 
provides a reasonable indication of the expected economic benefits of 
the proposed rulemaking in future years under a set of assumptions.
    One significant limitation to the benefit transfer method applied 
in this analysis is the inability to scale ozone-related benefits. 
Because ozone is a homogeneous gaseous pollutant, it is not possible to 
apportion ozone benefits to the precursor emissions of NOX 
and VOC. Coupled with the potential for NOX reductions to 
either increase or decrease ambient ozone levels, this prevents us from 
scaling the benefits associated with a particular combination of VOC 
and NOX emissions reductions to another. Because of our 
inability to scale ozone benefits, we do not include ozone benefits as 
part of the monetized benefits of the proposed standards. For the most 
part, ozone benefits contribute substantially less to the monetized 
benefits than do benefits from PM, thus their omission will not 
materially affect the conclusions of the benefits analysis. Although we 
expect economic benefits to exist, we were unable to quantify or to 
value specific changes in ozone, CO or air toxics because we did not 
perform additional air quality modeling.
    There are also a number of health and environmental effects which 
we were unable to quantify or monetize. A full appreciation of the 
overall economic consequences of the proposed rule requires 
consideration of all benefits and costs expected to result from the new 
standards, not just those benefits and costs which could be expressed 
here in dollar terms. A complete listing of the benefit categories that 
could not be quantified or monetized in our estimate are provided in 
Table V.E-5. These effects are denoted by ``B'' in Table V.E-3 above, 
and are additive to the estimates of benefits.

Table V.E-5.--Additional, Non-monetized Benefits of the Proposed Nonroad
                    Diesel Engine and Fuel Standards
------------------------------------------------------------------------
          Pollutant                       Unquantified effects
------------------------------------------------------------------------
Ozone Health.................  Premature mortality.\a\
                               Increased airway responsiveness to
                                stimuli.
                               Inflammation in the lung.
                               Chronic respiratory damage.
                               Premature aging of the lungs.
                               Acute inflammation and respiratory cell
                                damage.
                               Increased susceptibility to respiratory
                                infection.
                               Non-asthma respiratory emergency room
                                visits.
                               Increased school absence rates.
Ozone Welfare................  Decreased yields for commercial forests
                                (for example, Western US).
                               Decreased yields for fruits and
                                vegetables.
                               Decreased yields for non-commercial
                                crops.
                               Damage to urban ornamental plants.
                               Impacts on recreational demand from
                                damaged forest aesthetics.
                               Damage to ecosystem functions.
PM Health....................  Infant mortality.
                               Low birth weight.
                               Changes in pulmonary function.
                               Chronic respiratory diseases other than
                                chronic bronchitis.
                               Morphological changes.
                               Altered host defense mechanisms.
                               Cancer.
                               Non-asthma respiratory emergency room
                                visits.
PM Welfare...................  Visibility in many Class I areas.
                               Residential and recreational visibility
                                in non-Class I areas.
                               Soiling and materials damage.
                               Damage to ecosystem functions.

[[Page 28455]]

 
Nitrogen and Sulfate           Impacts of acidic sulfate and nitrate
 Deposition Welfare.            deposition on commercial forests.
                               Impacts of acidic deposition to
                                commercial freshwater fishing.
                               Impacts of acidic deposition to
                                recreation in terrestrial ecosystems.
                               Reduced existence values for currently
                                healthy ecosystems.
                               Impacts of nitrogen deposition on
                                commercial fishing, agriculture, and
                                forests.
                               Impacts of nitrogen deposition on
                                recreation in estuarine ecosystems.
                               Damage to ecosystem functions.
CO Health....................  Premature mortality.\a\
                               Behavioral effects.
HC Health \b\................  Cancer (benzene, 1,3-butadiene,
                                formaldehyde, acetaldehyde).
HC Welfare...................  Direct toxic effects to animals.
                               Bioaccumulation in the food chain.
                               Damage to ecosystem function.
                               Odor.
------------------------------------------------------------------------
Notes:
\a\ Premature mortality associated with ozone and carbon monoxide is not
  separately included in this analysis. In this analysis, we assume that
  the ACS/Krewski, et al. C-R function for premature mortality captures
  both PM mortality benefits and any mortality benefits associated with
  other air pollutants. A copy of Krewski, et al., can be found in
  Docket A-99-06, Document No. IV-G-75.
\b\ Many of the key hydrocarbons related to this rule are also hazardous
  air pollutants listed in the Clean Air Act.

F. Economic Impact Analysis

    An Economic Impact Analysis (EIA) was prepared to estimate the 
economic impacts of this proposal on producers and consumers of nonroad 
engines and equipment and related industries. The Nonroad Diesel 
Economic Impact Model (NDEIM), developed for this analysis, was used to 
estimate market-level changes in price and outputs for affected engine, 
equipment, fuel, and application markets as well as the social costs 
and their distribution across economic sectors affected by the program. 
This section presents the results of the economic impact analysis. A 
detailed description of the NDEIM, the model inputs, and several 
sensitivity analyses can be found in chapter 10 of the Draft Regulatory 
Impact Analysis prepared for this proposal.
1. What Is an Economic Impact Analysis?
    Regulatory agencies conduct economic impact analyses of potential 
regulatory actions to inform decision makers about the effects of a 
proposed regulation on society's current and future well-being. In 
addition to informing decision makers within the Agency, economic 
impact analyses are conducted to meet the statutory and administrative 
requirements imposed by Congress and the Executive office. The Clean 
Air Act requires an economic impact analysis under section 317, while 
Executive Order 12866--Regulatory Planning and Review requires 
Executive Branch agencies to perform benefit-costs analyses of all 
rules it deems to be ``significant'' (typically over $100 million 
annual social costs) and submit these analyses to the Office of 
Management and Budget (OMB) for review. This economic impact analysis 
estimates the potential market impacts of the proposed rule's 
compliance costs and provides the associated social costs and their 
distribution across stakeholders for comparison with social benefits 
(as presented in Section V.E).
2. What Is EPA's Economic Analysis Approach for This Proposal?
    The underlying objective of an EIA is to evaluate the effect of a 
proposed regulation on the welfare of affected stakeholders and society 
in general. Using information on the expected compliance costs of the 
proposed program as presented in the preceding discussion, this EIA 
explores how the companies that produce nonroad diesel engines, 
equipment, or fuel may change their production behavior in response to 
the costs of complying with the standards. It also explores how the 
consumers who use the affected products may change their purchasing 
decisions. For example, the construction industry may reduce purchases 
if the prices of nonroad diesel equipment increase, thereby reducing 
the volume of equipment sold (or market demand) for such equipment. 
Alternatively, the construction industry may pass along these 
additional costs to the consumers of their final goods and services by 
increasing prices, which would mitigate the potential impacts on the 
purchases of nonroad diesel equipment.
    The conceptual approach of the NDEIM is to link significantly 
affected markets to mimic how compliance costs will potentially ripple 
through the economy. The compliance costs will be directly borne by 
engine manufacturers, equipment manufacturers, and petroleum 
refineries. Depending on market characteristics, some or all of these 
compliance costs will be passed on through the supply chain in the form 
of higher prices extending to producers and consumers in the 
application markets (i.e., construction, agriculture, and 
manufacturing). The NDEIM explicitly models these linkages and 
estimates behavioral responses that lead to new equilibrium prices and 
output for all related markets and the resulting distribution of costs 
across stakeholders.
    The NDEIM uses a multi-market partial equilibrium approach to track 
changes in price and quantity for 60 integrated product markets, as 
follows:
    [sbull] 7 diesel engine markets (less than 25 hp, 26 to 50 hp, 51 
to 75 hp, 76 to 100 hp, 101 to 175 hp, 176 to 600 hp, and greater than 
600 hp; the EIA includes more horsepower categories than the standards, 
allowing more efficient use of the engine compliance cost estimates 
developed for this proposal).
    [sbull] 42 diesel equipment markets (7 horsepower categories within 
7 application categories: agricultural, construction, general 
industrial, pumps and compressors, generator and welder sets, 
refrigeration and air conditioning, and lawn and garden; there are 7 
horsepower/application categories that did not have sales in 2000 and 
are not included in the model, so the total number of diesel equipment 
markets is 42 rather than 49).
    [sbull] 3 application markets (agricultural, construction, and 
manufacturing).
    [sbull] 8 nonroad diesel fuel markets (2 sulfur content levels of 
15 ppm and 500 ppm for each of 4 PADDs; PADDs 1 and

[[Page 28456]]

3 are combined for the purpose of this analysis). It should be noted 
that PADD 5 includes Alaska and Hawaii. Because those two states are 
geographically separate from the rest of PADD 5, we seek comment on 
whether they should be considered as separate fuel markets.
    The NDEIM uses an intermediate run time frame and assumes perfect 
competition in the market sectors. It is a computer model comprised of 
a series of spreadsheet modules that define the baseline 
characteristics of the supply and demand for the relevant markets and 
the relationships between them. A detailed description of the model 
methodology, inputs, and parameters is provided in chapter 10 of the 
draft RIA prepared for this proposal. The model methodology is firmly 
rooted in applied microeconomic theory and was developed following the 
OAQPS Economic Analysis Resource Document.\294\ Based on the specified 
market linkages, the model is shocked by applying the engineering 
compliance cost estimates to the appropriate market suppliers and then 
numerically solved using an iterative auctioneer approach by ``calling 
out'' new prices until a new equilibrium is reached in all markets 
simultaneously.
---------------------------------------------------------------------------

    \294\ U.S. Environmental Protection Agency, Office of Air 
Quality Planning and Standards, Innovative Strategies and Economics 
Group, OAQPS Economic Analysis Resource Document, April 1999. A copy 
of this document can be found in Docket A-2001-28, Document No. II-
A-14.
---------------------------------------------------------------------------

    The actual economic impacts of the proposed rule will be determined 
by the ways in which producers and consumers of the engines, equipment, 
and fuels affected by the proposal change their behavior in response to 
the costs incurred in complying with the standards. In the NDEIM, these 
behaviors are modeled by the demand and supply elasticities. The supply 
elasticities for the engine and equipment markets and the demand 
elasticities for the application markets were estimated using 
econometric methods. The procedures and results are reported in 
Appendix 10.1 of the draft RIA. Literature-based estimates were used 
for the supply elasticities in the application and fuel markets.
    There are two ways to handle the demand elasticities for the 
engine, equipment, and fuel markets. In the approach used in NDEIM, 
these demand elasticities are internally derived based on the specified 
market linkages, i.e., the demand for engines, equipment, and fuel are 
modeled as directly related to the supply and demand of goods and 
services supplied by the final application markets. In other words, the 
supply of those goods and services determines the demand for equipment 
and fuel, and the supply of equipment determines the demand for 
engines. Using this approach, the NDEIM predicts that engine and 
equipment production will decrease by only a small amount: 0.013% and 
0.014% respectively (see Table V.F-1). Also, please see draft RIA 
Appendices 10A and 10B for more detailed estimates on the price 
increase estimates. Because the application markets are modeled with 
inelastic or unit elastic demand and supply elasticities (quantity 
supplied/demanded is expected to be fairly insensitive to price changes 
or they will vary directly with price changes), the model predicts that 
engine and equipment manufacturers will pass along virtually all of 
their costs to end users.
    An alternative approach could be used in which the demand 
elasticities for the equipment, engine, and fuel markets are not 
derived as part of the model. They could be estimated separately or a 
sensitivity analysis could be conducted that assumes more elastic 
values than those generated by the NDEIM. We are continuing to 
investigate this matter and will be placing additional information 
about elasticities in the docket during the comment period for this 
rule. We request comment on that information as well as on the 
methodology and other aspects of this EIA.
    The estimated engine and equipment market impacts are based solely 
on the expected increase in variable costs associated with the proposed 
standards. Fixed costs associated with the engine emission standards 
are not included in the market analysis reported in Table IV-F-1. This 
is because in an analysis of competitive markets the industry supply 
curve is based on its marginal cost curve, and fixed costs are not 
reflected in changes in the marginal cost curve. In addition, fixed 
costs are primarily R&D costs associated with design and engineering 
changes, and firms in the affected industries currently allocate funds 
for these costs. Therefore, fixed costs are not likely to affect the 
prices of engines or equipment. This assumption is described in greater 
detail in section 10.2 of the draft RIA. R&D costs are a long-run 
concern and decisions to invest or not invest in R&D are made in the 
long run. If funds have to be diverted from some other activity into 
R&D needed to meet the environmental regulations, then these costs 
represent a component of the social costs of the rule. Therefore, fixed 
costs are included in the welfare impact estimates reported in Table 
V.F-2 as additional costs on producers. We also performed a sensitivity 
analysis, included in chapter 10 of the draft RIA for this proposal, 
that includes fixed costs as part of the model. This results in a 
transfer of welfare losses from engine and equipment markets to the 
application markets, but does not change the overall welfare losses 
associated with the proposal.
    Economic theory indicates that, in the long run, prices are 
expected to reflect the average total costs of the marginal producer in 
a market and not just variable costs. This suggests that it may be 
necessary to treat fixed costs differently for a long-run analysis. We 
will continue to investigate this effect and intend to place additional 
information in the docket during the comment period for this rule. We 
request comment on that information as well as on how fixed costs and 
R&D expenditures are handled in the NDEIM.
    In addition to the variable and fixed costs described above, there 
are three additional costs components that are included in the total 
social cost estimates of the proposed regulation but that are not 
explicitly included in the NDEIM. These are operating savings (costs), 
fuel marker costs, and spillover from 15 ppm fuel to higher sulfur 
fuel. We request comment on how best to incorporate each of these costs 
in the analysis.
    Operating savings (costs) refers to changes in operating costs that 
are expected to be realized by users of both existing and new nonroad 
diesel equipment as a result of the reduced sulfur content of nonroad 
diesel fuel. These include operating savings (cost reductions) due to 
fewer oil changes, which accrue to nonroad engines, and marine and 
locomotive engines, that are already in use as well as new nonroad 
engines that will comply with the proposed standards (see section 
V.B.). These savings (costs) also include any extra operating costs 
associated with the new PM emission control technology which may accrue 
to new engines that use this new technology. These savings (costs) are 
not included directly in the model because some of the savings accrue 
to existing engines and because these savings (costs) are not expected 
to affect consumer decisions with respect to new engines. Instead, they 
are added into the estimated welfare impacts as additional costs to the 
application markets, since it is the users of these engines that will 
see these savings (costs). Nevertheless, a sensitivity analysis was 
also performed in which these savings (costs) are included as inputs to 
the NDEIM, where they are modeled as benefits accruing to the 
application producers. The results of

[[Page 28457]]

this analysis are presented in Chapter 10 of the draft RIA.
    Fuel marker costs refers to costs associated with marking high 
sulfur diesel fuel in the locomotive, marine, and heating oil markets 
between 2007 and 2014. Marker costs are not included in the market 
analysis because locomotive, marine, and heating oil markets are not 
explicitly modeled in the NDEIM. Similar to the operating savings 
(costs), marker costs are added into the estimated welfare impacts 
separately.
    The costs of fuel that spills over from the 15 ppm market to higher 
grade sulfur fuel are also not included in the NDEIM but, instead, are 
added into the estimated welfare impacts separately. As described in 
section IV above, refiners are expected to produce more 15 ppm fuel 
than is required for the nonroad diesel fuel market. This excess 15 ppm 
fuel will be sold into markets that allow fuel with a higher sulfur 
level (e.g., locomotive, marine diesel, or home heating fuel). Because 
this spillover fuel will meet the 15 ppm limit, it is necessary to 
count the costs of sulfur reduction processes against those fuels.
    Consistent with the engine and equipment cost discussion in section 
V.C. of this preamble, the EIA does not include any cost savings 
associated with the proposed equipment transition flexibility program 
or the proposed nonroad engine ABT program. As a result, the results of 
this EIA can be viewed as somewhat conservative, in this respect.
3. What Are the Results of this Analysis?
    The economic analysis consists of two parts: a market analysis and 
welfare analysis. The market analysis looks at expected changes in 
prices and quantities for directly and indirectly affected market 
commodities. The welfare analysis looks at economic impacts in terms of 
annual and present value changes in social costs. For this proposed 
rule, the social costs are computed as the sum of market surplus offset 
by operating cost savings. Market surplus is equal to the aggregate 
change in consumer and producer surplus based on the estimated market 
impacts associated with the proposed rule. Operating cost savings are 
associated with the decreased sulfur content of diesel fuel. These 
include maintenance savings (cost reductions) and changes in fuel 
efficiency. Increased maintenance costs may also be incurred for some 
technologies. Operating costs are not included in the market analysis 
but are instead listed as a separate category in the social cost 
results tables.
    Economic impact results for 2013, 2020, and 2030 are presented in 
this section. The first of these years, 2013, corresponds to the first 
year in which the standards affect all engines, equipment, and fuels. 
It should be noted that, as illustrated in Table V.D-2, above, 
aggregate program costs peak in 2014; increases in costs after that 
year are due to increases in the population of engines over time. The 
other years, 2020 and 2030, correspond to years analyzed in our 
benefits analysis. Detailed results for all years are included in 
Appendix 10.E. for this chapter.
a. Expected Market Impacts
    The market impacts of this rule suggest that the overall economic 
impact of the proposed emission control program on society is expected 
to be small, on average. According to this analysis, the average prices 
of goods and services produced using equipment and fuel affected by the 
proposal are expected to increase by about 0.02 percent. The estimated 
price increases and quantity reductions for engines and equipment vary 
depending on compliance costs. In general, we would expect for price 
increases to be higher (lower) as a result of a high (low) relative 
level of compliance costs to market price. We would also expect the 
change in price to be highest when compliance costs are highest.
    The estimated market impacts for 2013, 2020, and 2030 are presented 
in Table V.F-1. The market-level impacts presented in this table 
represent production-weighted averages of the individual market-level 
impact estimates generated by the model: the average expected price 
increase and quantity decrease across all of the units in each of the 
engine, equipment, fuel, and final application markets. For example, 
the model includes seven individual engine markets that reflect the 
different horsepower size categories. The 23 percent price change for 
engines shown in Table V.F-1 for 2013 is an average price change across 
all engine markets weighted by the number of production units. 
Similarly, equipment impacts presented in Table V.F-1 are weighted 
averages of 42 equipment-application markets, such as small (< 25hp) 
agricultural equipment and large (600hp) industrial 
equipment. It should be noted that price increases and quantity 
decreases for specific types of engines, equipment, application 
sectors, or diesel fuel markets are likely to be different. But the 
data in this table provide a broad overview of the expected market 
impacts that is useful when considering the impacts of the proposal on 
the economy as a whole. The individual market-level impacts are 
presented in Chapter 10 of the draft RIA for this proposal.
    Engine Market Results: Most of the variable costs associated with 
the proposed rule are passed along in the form of higher prices. The 
average price increase in 2013 for engines is estimated to be about 23 
percent. This percentage is expected to decrease to about 19.5 percent 
for 2020 and later. This expected price increase varies by engine size 
because compliance costs are a larger share of total production costs 
for smaller engines. In 2013, the year of greatest compliance costs 
overall, the largest expected percent price increase is for engines 
between 25 and 50 hp: 34 percent or $852; the average price for an 
engine in this category is about $2,500. However, this price increase 
is expected to drop to 26 percent, or about $647, for 2016 and later. 
The smallest expected percent price increase in 2013 is for engines in 
the greater than 600 hp category. These engines are expected to see 
price increases of about 3 percent increase in 2013, increasing to 
about 5.6 percent in 2014 and beyond. The expected price increase for 
these engines is about $4,211 in 2013, increasing to about $6,950 in 
2014 and later, for engines that cost on average about $125,000.
    The market impact model predicts that even with these increases in 
engine prices, total demand is not expected to change very much. The 
expected average change in quantity is only about 69 engines per year 
in 2013, out of total sales of more than 500,000 engines. The estimated 
change in market quantity is small because as compliance costs are 
passed along the supply chain they become a smaller share of total 
production costs. In other words, firms that use these engines and 
equipment will continue to purchase them even at the higher cost 
because the increase in costs will not have a large impact on their 
total production costs. Diesel equipment is only one factor of 
production for their output of construction, agricultural, or 
manufactured goods. The average decrease in the quantity of all engines 
produced as a result of the regulation is estimated to be about 0.013 
percent. This decrease ranges from 0.010 percent for engines less than 
25 hp to 0.016 percent for engines 175 to 600 hp.
    Equipment Market Results: Estimated price changes for the equipment 
markets reflect both the direct costs of the proposed standards on 
equipment production and the indirect cost through increased engine 
prices. In 2013, the average price increase for nonroad diesel 
equipment is estimated

[[Page 28458]]

to be about 5.2 percent. This percentage is expected to decrease to 
about 4.5 percent for 2020 and beyond. The range of estimated price 
increases across equipment types parallels the share of engine costs 
relative to total equipment price, so the estimated percentage price 
increase among equipment types also varies. The market price in 2013 
for agricultural equipment between 175 and 600 hp is estimated to 
increase about 1.4 percent, or $1,835 for equipment with an average 
cost of $130,000. This compares with an estimated engine price increase 
of about $1,754 for engines of that size. The largest expected price 
increase in 2013 for equipment is $4,335, or 4.9 percent, for pumps and 
compressors over 600 hp. This compares with an estimated engine price 
increase of about $4,211 for engines of that size. The smallest 
expected price increase in 2013 for equipment is $125, or 3.6 percent, 
for construction equipment less than 25 hp. This compares with an 
estimated engine price increase of about $124 for engines of that size. 
The price changes for the equipment are less than that for engines 
because the engine is only one input in the production of equipment.
    The output reduction for nonroad diesel equipment is estimated to 
be very small and to average about 0.014 percent for all years. This 
decrease ranges from 0.005 percent for general manufacturing equipment 
to 0.019 percent for construction equipment. The largest expected 
decrease in quantity in 2013 is 13 units of construction equipment per 
year for construction equipment between 100 and 175 hp, out of about 
62,800 units. The smallest expected decrease in quantity in 2013 is 
less than one unit per year in all hp categories of pumps and 
compressors.

                                 Table V.F-1.--Summary of Market Impacts ($2001)
----------------------------------------------------------------------------------------------------------------
                                    Engineering           Change in price               Change in quantity
                                       cost      ---------------------------------------------------------------
             Market              ----------------    Absolute
                                     Per unit       ($million)        Percent        Absolute         Percent
----------------------------------------------------------------------------------------------------------------
                                                      2013
----------------------------------------------------------------------------------------------------------------
Engines.........................          $1,087            $840            22.9           -69 a          -0.013
Equipment.......................           1,021           1,017             5.2            -118          -0.014
Application Markets b...........  ..............  ..............            0.02  ..............          -0.010
No. 2 Distillate Nonroad........           0.039           0.038             4.1         -1.38 c          -0.013
---------------------------------
                                                      2020
----------------------------------------------------------------------------------------------------------------
Engines.........................          $1,028            $779            19.5           -79 a          -0.013
Equipment.......................           1,018           1,013             4.4            -135          -0.014
Application Markets b...........  ..............  ..............            0.02  ..............          -0.010
No. 2 Distillate Nonroad........           0.039           0.039             4.1         -1.58 c          -0.014
---------------------------------
                                                      2030
----------------------------------------------------------------------------------------------------------------
Engines.........................          $1,027            $768            19.4           -92 a          -0.013
Equipment.......................           1,004             999             4.5            -156          -0.014
Application Markets b...........  ..............  ..............            0.02  ..............          -0.010
No. 2 Distillate Nonroad........           0.039           0.039             4.1         -1.84 c         -0.014
----------------------------------------------------------------------------------------------------------------
Notes:
a The absolute change in the quantity of engines represents only engines sold on the market. Reductions in
  engines consumed internally by integrated engine/equipment manufacturers are not reflected in this number but
  are captured in the cost analysis. For this reason, the absolute change in the number of engines and equipment
  does not match.
b The model uses normalized commodities in the application markets because of the great heterogeneity of
  products. Thus, only percentage changes are presented.
c Units are in million of gallons.

    Application Market Results: The estimated price increase associated 
with the proposed standards in all three of the application markets is 
very small and averages about 0.02 percent for all years. In other 
words, on average, the prices of goods and services produced using the 
engines, equipment, and fuel affected by this proposal are expected to 
increase only negligibly. This is because in all of the application 
markets the compliance costs passed on through price increases 
represent a very small share of total production costs. For example, 
the construction industry realizes an increase in production costs of 
approximately $468 million in 2013 because of the price increases for 
diesel equipment and fuel. However, this represents only 0.03 percent 
of the $1,392 billion value of shipments in the construction industry 
in 2001. The estimated average commodity price increase in 2013 ranges 
from 0.06 percent in the agricultural application market to about 0.01 
percent in the manufacturing application market. The percentage change 
in output is also estimated to be very small and averages about 0.01 
percent. This reduction ranges from less than a 0.01 percent decrease 
in manufacturing to about a 0.02 percent decrease in construction. Note 
that these estimated price increases and quantity decreases are average 
for these sectors and may vary for specific subsectors. Also, note that 
absolute changes in price and quantity are not provided for the 
application markets in Table V.F-1 because normalized commodity values 
are used in the market model. Because of the great heterogeneity of 
manufactured or agriculture products, a normalized commodity ($1 unit) 
is used in the application markets. This has no impact on the estimated 
percentage change impacts but makes interpretation of the absolute 
changes less informative.
    Fuel Markets Results: The estimated average price increase across 
all nonroad diesel fuel is about 4 percent for all years. For 15 ppm 
fuel, the estimated price increase for 2013 ranges from 3.2 percent in 
the East Coast region (PADD 1&3) to 9.3 percent in the mountain region 
(PADD 4). The average

[[Page 28459]]

national output decrease for all fuel is estimated to be about 0.01 
percent for all years, and is relatively constant across all four 
regional fuel markets.
b. Expected Welfare Impacts
    Social cost impact estimates are presented in Table V.F-2. A time 
series of social costs from 2007 through 2030 is presented in Table 
IV.F-3. As described above, the total social cost of the regulation is 
the sum of the changes in producer and consumer surplus estimated by 
the model plus engine maintenance savings (negative costs) resulting 
from using fuel with a lower sulfur content. Total social costs in 2013 
are projected to be 1,202.4 million ($2001). About 82 percent of the 
total social costs is expected to be borne by producers and consumers 
in the application markets, indicating that the majority of the costs 
are expected to be passed on in the form of higher prices. When these 
estimated impacts are broken down, 58 percent are expected to be borne 
by consumers in the application markets and 42 percent are expected to 
be borne by producers in the application markets. Equipment 
manufacturers are expected to bear about 10 percent of the total social 
costs. Engine manufacturers and diesel fuel refineries are expected to 
bear 2.5 percent and 0.5 percent, respectively. The remaining 5.0 
percent is accounted for by fuel marker costs and the additional costs 
of 15 ppm fuel being sold in to markets such as marine diesel, 
locomotive, and home heating fuel that do not require it.
    In 2030, the total social costs are projected to be about $1,509.6 
million ($2001). The increase is due to the projected annual growth in 
the engine and equipment populations. As in earlier years, producers 
and consumers in the application markets are expected to bear the large 
majority of the costs, approximately 94 percent. This is consistent 
with economic theory, which states that, in the long run, all costs are 
passed on to the consumers of goods and services.
    The present value of total social costs through 2030 is estimated 
to be $16.5 billion ($2001). This present value is calculated using a 
social discount rate of 3 percent from 2004 through 2030. We also 
performed an analysis using an alternative 7 percent social discount 
rate. Using that discount rate, the present value of the social costs 
through 2030 is estimated to be $9.9 billion ($2001).

                          Table V.F-2.--Summary of Social Costs Estimates Associated With Primary Program: 2013, 2020, and 2030
                                                                      [$million]a,b
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                   Maximum cost year (2013)                    Year 2020                       Final year (2030)
                                             -----------------------------------------------------------------------------------------------------------
                                                Market     Operating                Market     Operating                Market     Operating
                                               surplus      savings      Total     surplus      savings      Total     surplus      savings      Total
                                               ($10\6\)    ($10\6\)                ($10\6\)    ($10\6\)                ($10\6\)    ($10\6\)
--------------------------------------------------------------------------------------------------------------------------------------------------------
Engine Producers Total......................       30.2  ............       30.2        0.1  ............        0.1        0.1  ............        0.1
Equipment Producers Total...................      116.1  ............      116.1      102.6  ............      102.6        5.3  ............        5.3
    Agricultural Equipment..................       39.9  ............       39.9       33.2  ............       33.2        1.3  ............        1.3
    Construction Equipment..................       53.0  ............       53.0       48.2  ............       48.2        3.8  ............        3.8
    Industrial Equipment....................       23.2  ............       23.2       21.2  ............       21.2        0.2  ............        0.2
Application Producers and Consumers Total...    1,231.8       (241.9)      989.8    1,386.5       (190.1)    1,196.3    1,598.9       (174.5)    1,424.5
    Total Producer..........................      515.7  ............  .........      583.4  ............  .........      672.9  ............  .........
    Total Consumer..........................      716.1  ............  .........      803.1  ............  .........      926.0  ............  .........
    Agriculture.............................      348.7        (44.7)      304.0      339.2        (35.2)      364.0      416.5        (32.3)      429.2
    Construction............................      468.3        (77.9)      390.4      550.4        (61.2)      489.3      635.7        (56.1)      579.5
    Manufacturing...........................      414.8       (119.3)      295.5      436.8        (93.8)      343.0      501.8        (86.0)      415.7
Fuel Producers Total........................        7.8  ............        7.8        9.0  ............        9.0       10.5  ............       10.5
    PADD I&III..............................        3.6  ............        3.6        4.1  ............        4.1        4.8  ............        4.8
    PADD II.................................        2.9  ............        2.9        3.3  ............        3.3        3.9  ............        3.9
    PADD IV.................................        0.8  ............        0.8        0.9  ............        0.9        1.0  ............        1.0
    PADD V..................................        0.5  ............        0.5        0.6  ............        0.6        0.8  ............        0.8
Nonroad Spillover...........................  .........         51.2   .........  .........         58.6   .........  .........         69.2
Marker Costs................................  .........          7.3   .........  .........  ............  .........  .........  ............  .........
                                             ------------
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes:
\a\ Figures are in 2001 dollars.
\b\ Operating savings are shown as negative costs.


[[Page 28460]]


  Table IV.F-3--National Engineering Compliance Costs and Social Costs
               Estimates for the Proposed Rule: 2004-2030
                               [$10 \6\] a
------------------------------------------------------------------------
                                               Engineering      Total
                    Year                       compliance       social
                                                  costs         costsb
------------------------------------------------------------------------
2004.......................................            0.00         0.00
2005.......................................            0.00         0.00
2006.......................................            0.00         0.00
2007.......................................           39.61        39.61
2008.......................................          130.41       130.40
2009.......................................          132.25       132.25
2010.......................................          262.02       262.01
2011.......................................          641.12       641.07
2012.......................................        1,010.37     1,010.27
2013.......................................        1,202.52     1,202.40
2014.......................................        1,329.14     1,329.01
2015.......................................        1,260.74     1,260.62
2016.......................................        1,298.40     1,298.27
2017.......................................        1,318.75     1,318.62
2018.......................................        1,325.02     1,324.89
2019.......................................        1,339.30     1,339.16
2020.......................................        1,366.79     1,366.66
2021.......................................        1,351.08     1,350.94
2022.......................................        1,349.58     1,349.44
2023.......................................        1,365.53     1,365.38
2024.......................................        1,371.60     1,371.45
2025.......................................        1,395.98     1,395.83
2026.......................................        1,419.79     1,419.64
2027.......................................        1,442.91     1,442.76
2028.......................................        1,465.41     1,465.26
2029.......................................        1,487.68     1,487.53
2030.......................................        1,509.77     1,509.61
--------------------------------------------
NPV at 3%..................................       16,524.29    16,522.66
NPV at 7%..................................        9,894.02    9,893.06
------------------------------------------------------------------------
Notes:
a Figures are in 2001 dollars.
b Figures in this column do not include the human health and
  environmental benefits of the proposal.

VI. Alternative Program Options

    Our proposed emission control program consists of a two-step 
program to reduce the sulfur content of nonroad diesel fuel in 
conjunction with the proposed Tier 4 engine standards. As we developed 
this proposal, we evaluated a number of alternative options with regard 
to the scope, level, and timing of the standards. This section presents 
a summary of our analysis of several alternative control scenarios. A 
complete discussion of all the alternatives, their feasibility, and 
their inventory, benefits, and cost impacts can be found in Chapter 12 
of the draft Regulatory Impact Analysis for this proposal.
    While we are interested in comments on all of the alternatives 
presented, we are especially interested in comments on two alternative 
scenarios which EPA believes merit further consideration in developing 
the final rule: a program in which sulfur levels are required to be 
reduced to 15 ppm in essentially a single step, and a variation on the 
proposed two-step fuel control program, in which the second step of 
sulfur control to 15 ppm in 2010 would apply to locomotive and marine 
diesel fuel in addition to nonroad diesel fuel. This section describes 
these two options in greater detail; additional information can be 
found in Chapter 12 of the draft Regulatory Impact Analysis for this 
proposal.

A. Summary of Alternatives

    We developed emissions, benefits, and cost analyses for a number of 
alternatives. The alternatives we considered can be categorized 
according to the structure of their fuel requirements: whether the 15 
ppm fuel sulfur limit is reached in two-steps, like the proposed 
program, or one-step.
    One-step alternatives are those in which the fuel sulfur standard 
is applied in a single step: there are no fuel-based phase-ins. We 
evaluated three one-step alternatives. Option 1 is described in detail 
in Section VI.B, below. We considered two other one-step alternatives 
which differ from Option 1 in the timing of the fuel option (2006 or 
2008) and the engines standards (level of the standards and when they 
are introduced). As described in Table IV-1, Option 1b differs from 
Option 1 regarding the timing of the fuel standards, while Option 1a 
differs from Option 1 in terms of the engine standards. Both Option 1a 
and 1b would also extend the 15 ppm fuel sulfur limit to locomotive and 
marine diesel fuel as well.
    Two-step alternatives are those in which the fuel sulfur standard 
is set first at 500 ppm and then is reduced to 15 ppm. The two-step 
alternatives vary from the proposal in terms of both the timing and 
levels of the engine standards and the timing of the fuel standards. 
Option 2a is the same as the proposed program except the 500 ppm fuel 
standard is introduced a year earlier, in 2006. Option 2b is the same 
as the proposed program except the 15 ppm fuel standard is introduced a 
year earlier in 2009 and the trap-based PM standards begin earlier for 
all engines. Option 2c is the same as the proposed program except the 
15 ppm fuel standard is introduced a year earlier in 2009 and the trap-
based PM standards begin earlier for engines 175-750 hp. Option 2d is 
the same as the proposed program except the NOX standard is 
reduced to 0.30 g/bhp-hr for engines 25-75 hp, and this standard is 
phased in. Finally, Option 2e is the same as the proposed program 
except there are no new Tier 4 NOX limits.
    Options 3 and 4 are identical to the proposed program, except 
Option 3 would exempt mining equipment over 750 hp from the Tier 4 
standards, and Option 4 would include applying the 15 ppm sulfur limit 
to both locomotive and marine diesel fuel. Option 4 is discussed in 
detail in Section IV.C, below.
    Option 5a and 5b are identical to the proposal except for the 
treatment of engines less than 75 hp. Option 5a is identical to the 
proposal except that no new program requirements would be set in Tier 4 
for engines under 75 hp. Instead Tier 2 standards and testing 
requirements for engines under 50 hp, and Tier 3 standards and testing 
requirements for 50-75 hp engines, would continue indefinitely. The 
Option 5b program is identical to the proposal except that for engines 
under 75 hp only the 2008 engine standards would be set. There would be 
no additional PM filter-based standard in 2013 for 25-75 hp engines, 
and no additional NOX+NMHC standard in 2013 for 25-50 hp 
engines.
    Table VI-1 contains a summary of a number of these alternatives and 
the expected emission reductions, costs, and monetized benefits 
associated with them in comparison to the proposal. These alternatives 
cover a broad range of possible approaches and serve to provide insight 
into the many other program design alternatives not expressly evaluated 
further. The analysis was done using a 3% discount rate. If we were to 
use another rate, the values would change but not to such a degree as 
to change our conclusions regarding the various options. A complete 
discussion of all the alternatives, their feasibility, and their 
inventory, benefits, and cost impacts can be found in Chapter 12 of the 
draft Regulatory Impact Analysis for this proposal.

[[Page 28461]]

[GRAPHIC] [TIFF OMITTED] TP23MY03.012


[[Page 28462]]


[GRAPHIC] [TIFF OMITTED] TP23MY03.013


[[Page 28463]]



B. Introduction of 15 ppm Nonroad Diesel Sulfur Fuel in One Step

    EPA carefully evaluated and is seeking comment on alternative 
regulatory approaches. Instead of the proposed two-step reduction in 
nonroad diesel sulfur, one alternative would require that the nonroad 
diesel sulfur level be reduced to 15ppm beginning June 1, 2008. This 
alternative would have the advantage of enabling use of high efficiency 
exhaust emission control technology for nonroad engines as early as the 
2009 model year. It also would have several disadvantages which have 
prompted us not to propose it. The disadvantages in comparison to the 
proposal include inadequate lead-time for engine and equipment 
manufacturers and refiners, leading to increased costs and potential 
market disruptions. In this section, we describe this alternative in 
greater detail and discuss potential engine and fuel impacts. We also 
present our estimated emission and benefit impacts. Two other one-step 
fuel options which are variations of the alternative discussed in this 
section, Options 1a and 1b in Table VI-1, are presented in Chapter 12 
of the draft RIA for this proposal.
1. Description of the One-Step Alternative
    While numerous engine standards and phase-in schedules are 
possible, we considered the standards shown in Tables VI-2 and VI-3 as 
being the most stringent one-step program that could be considered 
potentially feasible considering cost, lead-time, and other factors. 
These standards are similar to those in our proposed option, the 
primary difference being the generally earlier phase-in dates for the 
PM standards.

                               Table VI-2.--PM Standards for 1-Step Fuel Scenario
                                                   [g/bhp-hr]
----------------------------------------------------------------------------------------------------------------
                                                                           Model year
                 Engine power                  -----------------------------------------------------------------
                                                   2009       2010       2011       2012       2013       2014
----------------------------------------------------------------------------------------------------------------
hp < 25.......................................       0.30  .........  .........  .........  .........  .........
25 <= hp <50..................................      10.22  .........  .........  .........       0.02  .........
50 <= hp <75..................................  .........  .........  .........  .........       0.02  .........
75 <= hp <175.................................  .........  .........       0.01  .........  .........  .........
                                                .........    \a\ 50%    \a\ 50%   \a\ 100%  .........  .........
175 <= hp <750................................  .........       0.01  .........  .........  .........  .........
                                                  \a\ 50%    \a\ 50%   \a\ 100%  .........  .........  .........
hp = 750...........................  .........  .........  .........  .........       0.01  .........
                                                .........  .........    \a\ 50%    \a\ 50%    \a\ 50%  \a\ 100%
----------------------------------------------------------------------------------------------------------------
Notes:
\a\ Percentages are the model year sales required to comply with the indicated standard.


                          Table VI-3.--NOX and NMHC Standards for 1-Step Fuel Scenario
                                                   [g/bhp-hr]
----------------------------------------------------------------------------------------------------------------
                                                                                     Model year
                           Engine power                            ---------------------------------------------
                                                                       2011       2012       2013        2014
----------------------------------------------------------------------------------------------------------------
25 <= hp < 75.....................................................  .........  .........      a 3.5  ...........
                                                                              ------------
                                                                                            0.30 NOX
75 <= hp <175.....................................................                         0.14 NMHC
                                                                              ------------
                                                                                   b 50%      b 50%       b 100%
                                                                   ------------
                                                                                      0.30 NOX
175 <= hp <750....................................................                    0.14 NMHC
                                                                   ------------
                                                                        b 50%      b 50%      b 50%       b 100%
                                                                   ------------
                                                                                       0.30 NOX
hp =750................................................                    0.14 NMHC
                                                                   ------------
                                                                        b 50%      b 50%      b 50%      b 100%
----------------------------------------------------------------------------------------------------------------
Notes:
a A 3.5 NMHC + NOX standard would apply to the 25-50 hp engines. Engines greater than 50hp are already subject
  to this standard in 2008 under the existing Tier 3 program.
b Percentages are the model year sales required to comply with the indicated standards.

2. Engine Emission Impacts
    The main advantage associated with this one-step approach is 
pulling ahead the long-term PM engine standards. By making 15 ppm 
sulfur fuel widely available by late 2008, we could accelerate the 
long-term PM engine standards, leading to the introduction of precious 
metal catalyzed PM traps as early as 2009, two years earlier than 
possible under the two-step sulfur reduction approach. Some 
stakeholders have expressed the concern that a two-step approach leads 
to later than desired introduction of high-efficiency exhaust emissions 
controls on nonroad diesels because this cannot happen until the 15 ppm 
fuel standard goes into effect. As shown in Table VI-1, there would be 
additional public health benefits associated with this one-step 
approach. However, in comparison to the proposal, the additional 
benefits are

[[Page 28464]]

relatively small, less than one percent or about $3 billion more than 
the proposed program.\295\
---------------------------------------------------------------------------

    \295\ A variation on this one-step approach would be to also 
require the sulfur content of locomotive and marine fuel to meet the 
15 ppm standard in 2008. The decision of whether or not to require 
the sulfur content of locomotive and marine fuel to also be reduced 
to 15 ppm, however, is not unique to the one step approach, and, as 
discussed below is an alternative also being evaluated under our 
proposed 2-step program. Were we to require locomotive and marine 
diesel fuel to also meet the 15 ppm standard in 2008 under a one-
step approach, there would be additional inventory reductions of 
about 10,000 tons of PM and 128,000 tons of SO3 (NPV 3% 
through 2030).
---------------------------------------------------------------------------

    Even though 15 ppm fuel would be available beginning June 1, 2008 
under this one-step approach, we do not believe it would be feasible to 
propose an aggressive turnover of new engines to trap-equipped versions 
in 2009. Nor would it be possible to introduce NOX controls 
any earlier than we are already proposing, model year 2011. The 
proposed standards need to be coordinated with Tier 3 standards, and 
with the heavy duty highway diesel standards. The coordination of Tier 
4 standards with Tier 3 standards and with the development of emissions 
control technology for highway diesel engines is of critical importance 
to successful implementation of the Tier 4 standards. Even those 
manufacturers who do not make highway engines are expected to gain 
substantially from the highway PM and NOX control 
development work, provided they can plan for standards set at a similar 
level of stringency and timed in a way to allow for the orderly 
migration of highway engine technology to nonroad applications.
    Thus, although the application of high-efficiency exhaust PM 
emission controls to nonroad diesels would be enabled with the 
introduction of 15 ppm sulfur nonroad fuel in 2008 under a one-step 
program, we believe that to require the application of PM controls 
across the wide spectrum of nonroad engines shortly thereafter would 
raise serious feasibility concerns that could only be resolved, if at 
all, through a very large additional R&D effort undertaken roughly in 
parallel with the similarly large highway R&D effort, a duplication of 
effort we wish to avoid for reasons discussed in Section III. Nonroad 
engine designers would need to accomplish much of this development well 
before the diesel experience begins to accumulate in earnest in 2007, 
in order to be ready for a 2009 first introduction date. Waiting until 
2007 before initiating 2009 model year design work would risk the 
possibility of product failures, limited product availability and major 
market disruptions. At the same time, for those engine manufacturers 
who participate in both the highway and nonroad diesel engine markets, 
attempting to have concurrent engine product developments for highway 
and nonroad, could result in the possibility of product failures, 
limited product availability and major disruptions for the highway 
market as well. Thus, in balancing their costs and burden, many 
manufacturers may be forced to choose which products would be available 
for 2009 and which products would be delayed for release. Manufacturers 
would also incur large additional costs to redesign hundreds of engine 
models and thousand of machine types to meet Tier 4 standards only one 
to three years after Tier 3 standards take effect in 2006-2008. These 
cost impacts are reflected in Table VI-1 and their derivation is 
explained in chapter 12 of the draft RIA. This extra expenditure could 
only be modestly mitigated by phasing in the standards, since a crash 
R&D effort with limited benefit from highway experience would still be 
necessary.
    Moreover, with respect to NOX, it would be impractical 
or simply infeasible to pull the standards ahead on the same schedule. 
This is because EPA's highway diesel program allows manufacturers to 
phase in NOX technology over 2007-2010. As a result, we do 
not expect that the high-efficiency NOX control technology 
could reasonably be applied to nonroad engines any earlier under a one-
step program than under a two-step program (i.e., beginning in 2011).
    In summary, this option would lead us to apply PM and 
NOX standards in two different model years, or else forgo 
any opportunity to apply PM traps in 2009. Redesigning engines and 
emission controls for early PM control and then again a couple of years 
later for NOX control, on top of shortened Tier 3 stability 
periods, would likely add substantial costs to the program. As 
manufacturers attempt to avoid these costs and optimize their 
development they may simply have to restrict product offerings for some 
period, leading to price spikes and shortages due to lack of product 
availability. Having the NOX and PM standards phase in 
simultaneously under our proposed approach avoids cost and design 
stability issues for both engine and equipment manufacturers. In 
addition, the longer leadtime for the engine standards under our 
proposed program will allow greater economic efficiencies for engine 
manufacturers as they transfer highway emission reduction technology to 
nonroad engines.
3. Fuel Impacts
    In addition to the challenges associated with pulling ahead the PM 
standards described above, there are also some concerns regarding the 
practicality of an early 15 ppm nonroad diesel sulfur standard. A one-
step approach may result in several economic inefficiencies that would 
increase the cost of the program. For example, refiners will have 
little opportunity to take advantage of the newer desulfurization 
technologies currently being developed. As described in sections IV and 
V, refiners will only begin to be able to take advantage of these new 
technologies in 2008. By 2010, the ability to incorporate them into 
their refinery modifications is expected to double. If refiners have to 
take steps to reduce the sulfur content of nonroad diesel fuel earlier, 
they will likely have to use more expensive current technology. The 
cost impacts of this decision will persist, since the choice of 
technology is a long term decision. If a refiner is forced by the 
effective date of the standards to employ a more expensive technology, 
that choice will affect that refiner's output indefinitely, since the 
cost of upgrading to the new technologies will be prohibitive. As 
presented in section 5.2 of the Draft RIA, we estimate that the costs 
of achieving a 15 ppm standard in 2008 is approximately 0.4 c/gal 
greater than for the proposal. While difficult to quantify there are 
also considerable advantages to allowing refiners some operating time 
in producing 15 ppm diesel fuel for the highway program prior to 
requiring them to solidify their designs for producing nonroad diesel 
fuel to 15 ppm. The primary advantage is that the design of 
desulfurization equipment used to produce 15 ppm nonroad diesel fuel 
can reflect the operating experience of the equipment used to produce 
15 ppm highway diesel fuel starting in 2006. This extra time would also 
provide current refiners of high sulfur diesel fuel with highly 
confident estimates of the cost of producing 15 ppm diesel fuel, 
reducing uncertainty and increasing their likelihood of investing to 
produce this fuel. With a start date of June 1, 2008 refiners would 
have to solidify their designs and start construction prior to getting 
any data on the performance of their highway technology. This would 
increase the cost of producing 15 ppm nonroad diesel fuel for the life 
of the new desulfurization equipment, as well as potentially delaying 
some refiners' decision to invest in new desulfurization equipment due 
to uncertainties in cost, performance, etc.

[[Page 28465]]

4. Emission and Benefit Impacts
    We used the nonroad model to estimate the emission inventory 
impacts associated with this one-step option, as well as the other 
options listed in Table VI-1. As for all the alternatives, we then used 
the benefits transfer method to estimate the monetized benefits of the 
alternative.\296\ The results are shown in Table VI-1. As is evidenced 
by the values in Table VI-1, the one-step alternative would achieve 
slightly greater PM and NOX emission reductions through 2030 
than the proposed 2-step program, with 6,000 and 11,000 additional tons 
reduced, respectively (or less than 0.5 percent). Unlike the proposed 
2-step program, however, there would be no SO2 emission 
reductions in 2007 due to the delay in fuel sulfur control, although 
2009 and later emission are slightly greater due primarily to the 
earlier introduction of engines using PM filters. Nevertheless, the 
SO2 benefits of the one-step program are slightly less than 
the proposed 2-step program in the long run, by about 191,000 tons 
(about 4 percent) through 2030.
---------------------------------------------------------------------------

    \296\ The results that were obtained for Option 1a were 
extrapolated based on the emission inventory changes to the proposed 
program and were obtained for the other alternatives by assuming the 
air quality changes between the alternative and the actual case run 
were small enough to allow for such extrapolation. An explanation of 
the benefits transfer method is contained in Chapter 9 of the draft 
RIA.
---------------------------------------------------------------------------

    After careful consideration of these matters, we have decided to 
propose the two-step approach in today's notice. The two-step program 
avoids adverse risks to the smooth implementation of the entire Tier 4 
nonroad program that could be caused by the significantly shortened 
lead-time and stability of the one-step program. There are also 
concerns about the potential negative impacts the one-step option may 
have on the 2007 highway program, including the implications of the 
overlap of implementation schedules (see above and Chapter 12 of the 
draft RIA). Nevertheless, we believe that the one-step approach is a 
regulatory alternative worth considering. In addition to seeking 
comment on our proposed program, we also seek comment on the relative 
merits and shortcomings of a one-step approach to regulating nonroad 
diesel fuel and the associated schedule for implementing the engine 
standards.

C. Applying 15 ppm Requirement to Locomotive and Marine Diesel Fuel

    To enable the high efficiency exhaust emission control technology 
to begin to be applied to nonroad diesel engines beginning with the 
2011 model year, we are proposing that all nonroad diesel fuel produced 
or imported after June 1, 2010 would have to meet a 15 ppm sulfur cap. 
Although locomotive and marine diesel engines are similar in size to 
some of the diesel engines covered in this proposal, there are many 
differences that have caused us to treat them separately in past EPA 
programs.\297\ These include differences in duty cycles and exhaust 
system design configurations, size, and rebuild and maintenance 
practices. Because of these differences, we are not proposing new 
engine standards today for these engine categories. Since we are not 
proposing more stringent emission standards, we are also not proposing 
that the second step of sulfur control to 15 ppm in 2010 be applied to 
locomotive and marine diesel fuel. Instead, we are proposing to set a 
sulfur fuel content standard of 500 ppm for diesel fuel used in 
locomotive and marine applications. This fuel standard is expected to 
provide considerable sulfate PM and SO2 benefits even 
without establishing more stringent emission standards for these 
engines. We estimate that, cumulatively through 2030, reducing the 
sulfur content of locomotive and marine diesel fuel would eliminate 
about 102,000 tons of sulfate PM (net present value, based on a 3 
percent discount rate).
---------------------------------------------------------------------------

    \297\ Locomotives, in fact, are treated separately from other 
nonroad engines and vehicles in the Clean Air Act, which contains 
provisions regarding them in section 213(a)(5). Less than 50 hp 
marine engines were included in the 1998 final rule for nonroad 
diesel engines, albeit with some special provisions to deal with 
marine-specific engine characteristics and operating cycles.
---------------------------------------------------------------------------

    As discussed in section IV, we are seriously considering the option 
of extending the 15 ppm sulfur standard to locomotive and marine fuel 
as early as June 1, 2010, including them in the second step of the 
proposed two-step program. There are several advantages associated with 
this alternative. First, as reflected in Table VI-1, it would provide 
important additional sulfate PM and SO2 emission reductions 
and the estimated benefits from these reductions would outweigh the 
costs by a considerable margin. Second, in some ways it would simplify 
the fuel distribution system and the design of the fuel program 
proposed today since a marker would not be required for locomotive and 
marine diesel fuel. Furthermore, the prices for locomotive and marine 
diesel fuel may be virtually unaffected. Under the proposal, we expect 
that a certain amount of marine fuel will be 15 ppm sulfur fuel 
regardless of the standard due to limitations in the production and 
distribution of unique fuel grades. Where 500 ppm fuel is available, 
the possible suppliers of fuel will likely be more constrained, 
limiting competition and allowing prices to approach that of 15 ppm 
fuel. If we were to bring locomotive and marine fuel to 15 ppm, the 
pool of possible suppliers could expand beyond those today, since 
highway diesel fuel will also be at the same standard. Third, it would 
help reduce the potential opportunity for misfueling of 2007 and later 
model year highway vehicles and 2011 and later model year nonroad 
equipment with higher sulfur fuel. Finally, it would allow refiners to 
coordinate plans to reduce the sulfur content of all of their nonroad, 
locomotive, and marine diesel fuel at one time. While in many cases 
this may not be a significant advantage, it may be a more important 
consideration here since it is probably not a question of whether 
locomotive and marine fuel must meet a 15 ppm cap, but merely when. As 
discussed in section IV, it is the Agency's intention to propose action 
in the near future to set new emission standards for locomotive and 
marine engines that could require the use of high efficiency exhaust 
emission control technology, and thus, also require the use of 15 ppm 
sulfur diesel fuel.\298\ We anticipate that such engine standards would 
likely take effect in the 2011-13 timeframe, requiring 15 ppm 
locomotive and marine diesel fuel in the 2010-12 timeframe. We intend 
to publish an advance notice of proposed rulemaking for such standards 
by the Spring of 2004 and finalize those standards by 2007.
---------------------------------------------------------------------------

    \298\ EPA established the most recent new standards for 
locomotives and marine diesel engines (including those under 50 hp) 
in separate actions (63 FR 18977, April 16, 1998, and 67 FR 68241, 
November 8, 2002).
---------------------------------------------------------------------------

    However, discussions with refiners have suggested there are 
significant advantages to leaving locomotive and marine diesel fuel at 
500 ppm, at least in the near-term and until we set more stringent 
standards for those engines. The locomotive and marine diesel fuel 
markets could provide an important market for off-specification 
product, particularly during the transition to 15 ppm for highway and 
nonroad diesel fuel in 2010. Waiting just a year or two beyond 2010 
would address the critical near-term needs during the transition. In 
addition, waiting just another year or two beyond 2010 is also 
projected to allow virtually all refiners to take advantage of the new 
lower cost technology.
    After careful consideration of these matters, we have decided not 
to propose

[[Page 28466]]

to apply the second step of sulfur control of 15 ppm to locomotive and 
marine diesel fuel at this time. Nevertheless, for the reasons 
described above, we are carefully weighing whether it would be 
appropriate to do so. Therefore, we seek comment on this alternative 
and the various advantages, disadvantages, and implications of it.

D. Other Alternatives

    We have also analyzed a number of other alternatives, as summarized 
in Table VI-1. Some of these focus on control options more stringent 
than our proposal while others reflect modified engine requirements 
that result in less stringent control. EPA has evaluated these options 
in terms of the feasibility, emissions reductions, costs, and other 
relevant factors. EPA believes the proposed approach is the proper one 
with respect to these factors, and believes the options discussed above 
while having possible merit in some areas, raise what we believe are 
different and significant concerns with respect to these factors 
compared to the proposed approach. Hence we did not include these 
options. These concerns are discussed in more detail in Chapter 12. 
These concerns are discussed in more detail in Chapter 12 of the draft 
RIA. Hence, we did not include these options as part of our proposal 
for nonroad fuel and engine controls. We are interested in comment on 
these alternatives, especially information regarding their feasibility, 
costs, and other relevant concerns.

VII. Requirements for Engine and Equipment Manufacturers

    This section describes the regulatory changes proposed for the 
engine and equipment compliance program. First, the proposed 
regulations for Tier 4 engines have been written in plain language. 
They are structured to contain the provisions that are specific to 
nonroad CI engines in a new proposed part 1039, and to apply the 
general provisions of existing parts 1065 and 1068. The proposed plain 
language regulations, however, are not intended to significantly change 
the compliance program, except as specifically noted in today's notice 
(and we are not soliciting comment on any part of the rule that remains 
unchanged substantively). As proposed, these plain language regulations 
would only apply for Tier 4 engines. The changes from the existing 
nonroad program are described below along with other notable aspects of 
the compliance program.

A. Averaging, Banking, and Trading

1. Are We Proposing To Keep the ABT Program for Nonroad Diesel Engines?
    EPA has included averaging, banking, and trading (ABT) programs in 
most mobile source emission control programs adopted in recent years. 
Our existing regulations for nonroad diesel engines include an ABT 
program (Sec.  89.201 through Sec.  89.212). We are proposing to retain 
the basic structure of the existing nonroad diesel ABT program with 
today's notice, though we are proposing a number of changes to 
accommodate implementation of the proposed emission standards. Behind 
these changes is the recognition that the proposed standards represent 
a major technological challenge to the industry. The proposed ABT 
program is intended to enhance the ability of engine manufacturers to 
meet the stringent standards proposed today. The proposed program is 
also structured to limit production of very high-emitting engines and 
to avoid unnecessary delay of the transition to the new exhaust 
emission control technology.
    We view the proposed ABT program as an important element in setting 
emission standards that are appropriate under CAA section 213 with 
regard to technological feasibility, lead time, and cost. The ABT 
program helps to ensure that the stringent standards we are proposing 
are appropriate under section 213(a) given the wide breadth and variety 
of engines covered by the standards. For example, if there are engine 
families that will be particularly costly or have a particularly hard 
time coming into compliance with the standard, this flexibility allows 
the manufacturer to adjust the compliance schedule accordingly, without 
special delays or exceptions having to be written into the rule. 
Emission-credit programs also create an incentive (for example, to 
generate credits in early years to create compliance flexibility for 
later engines) for the early introduction of new technology, which 
allows certain engine families to act as trailblazers for new 
technology. This can help provide valuable information to manufacturers 
on the technology before they apply the technology throughout their 
product line. This early introduction of clean technology improves the 
feasibility of achieving the standards and can provide valuable 
information for use in other regulatory programs that may benefit from 
similar technologies. Early introduction of such engines also secures 
earlier emission benefits.
    In an effort to make information on the ABT program more available 
to the public, we intend to issue periodic reports summarizing use of 
the proposed ABT program by engine manufacturers. The information 
contained in the periodic reports would be based on the information 
submitted to us by engine manufacturers, and summarized in a way that 
protects the confidentiality of individual engine manufacturers. We 
believe this information will also be helpful to engine manufacturers 
by giving them a better indication of the availability of credits. 
Again, our periodic reports would not contain any confidential 
information submitted by individual engine manufacturers, such as sales 
figures. Also, the information would be presented in a format that 
would not allow such confidential information to be determined from the 
reports.
2. What Are the Provisions of the Proposed ABT Program?
    The following section describes the changes proposed to the 
existing ABT program. In addition to those areas specifically 
highlighted, we are soliciting comments on all aspects of the proposed 
ABT changes, including comments on the need for and benefit of these 
changes to manufacturers in meeting the proposed emission standards.
    The ABT program has three main components. Averaging means the 
exchange of emission credits between engine families within a given 
engine manufacturer's product line. (Engine manufacturers divide their 
product line into ``engine families'' that are comprised of engines 
expected to have similar emission characteristics throughout their 
useful life.) Averaging allows a manufacturer to certify one or more 
engine families at levels above the applicable emission standard, but 
below a set upper limit. However, the increased emissions must be 
offset by one or more engine families within that manufacturer's 
product line that are certified below the same emission standard, such 
that the average emissions from all the manufacturer's engine families, 
weighted by engine power, regulatory useful life, and production 
volume, are at or below the level of the emission standard. (The 
inclusion of engine power, useful life, and production volume in the 
averaging calculations is designed to reflect differences in the in-use 
emissions from the engines.) Averaging results are calculated for each 
specific model year. The mechanism by which this is accomplished is 
certification of the engine family to a ``family emission limit'' (FEL) 
set by the manufacturer, which may be above or below the standard. An 
FEL that is established

[[Page 28467]]

above the standard may not exceed an upper limit specified in the ABT 
regulations. Once an engine family is certified to an FEL, that FEL 
becomes the enforceable emissions limit for all the engines in that 
family for purposes of compliance testing. Averaging is allowed only 
between engine families in the same averaging set, as defined in the 
regulations.
    Banking means the retention of emission credits by the engine 
manufacturer for use in future model year averaging or trading. Trading 
means the exchange of emission credits between nonroad diesel engine 
manufacturers which can then be used for averaging purposes, banked for 
future use, or traded to another engine manufacturer.
    The existing ABT program for nonroad diesel engines covers 
NMHC+NOX emissions as well as PM emissions. With today's 
notice we are proposing to make the ABT program available for the 
proposed NOX standards and proposed PM standards. (For 
engines less than 75 horsepower where we are proposing combined 
NMHC+NOX standards, the ABT program would continue to be 
available for the proposed NMHC+NOX standards as well as the 
proposed PM standards.) ABT would not be available for the proposed 
NMHC standards for engines above 75 horsepower or for the proposed CO 
standards for any engines.
    As noted earlier, the existing ABT program for nonroad diesel 
engines includes FEL caps--limits on how high the emissions from 
credit-using engine families can be. No engine family may be certified 
above these FEL caps. These limits provide the manufacturers compliance 
flexibility while protecting against the introduction of unnecessarily 
high-emitting engines. When we propose new standards, we typically 
propose new FEL caps for the new standards. In the past, we have 
generally set the FEL caps at the emission levels allowed by the 
previous standard, unless there was some specific reason to do 
otherwise. We are proposing to do otherwise here because the proposed 
standard levels in today's notice are so much lower than the current 
standards levels, especially the Tier 4 standards for engines above 75 
horsepower. The transfer to new technology is feasible and appropriate. 
Thus, to ensure that the ABT provisions are not used to continue 
producing old-technology high-emitting engines under the new program, 
the proposed FEL caps would not, in general, be set at the previous 
standards. An exception is for the proposed NMHC+NOX 
standard for engines between 25 and 50 horsepower effective in model 
year 2013, where we are proposing to use the previously applicable 
NMHC+NOX standard for the FEL cap since the gap between the 
previous and proposed standards is approximately 40 percent (rather 
than 90 percent for engines above 75 horsepower).
    For engines above 75 horsepower certified during the phase-in 
period, there would be two separate sets of engines with different FEL 
caps. For engines certified to the existing (Tier 3) 
NMHC+NOX standards during the phase-in, the FEL cap would 
necessarily continue to be the existing FEL caps as adopted in the 
October 1998 rule. For engines certified to the proposed Tier 4 
NOX standard during the phase-in, the FEL cap would be 3.3 
g/bhp-hr for engines between 75 and 100 horsepower, 2.8 g/bhp-hr for 
engines between 100 and 750 horsepower, and 4.6 g/bhp-hr for engines 
above 750 horsepower. These proposed NOX FEL caps represent 
an estimate of the NOX emission level that is expected under 
the combined NMHC+NOX standards that apply with the existing 
previous tier standards. Beginning in model year 2014 when the proposed 
Tier 4 NOX standard for engines above 75 horsepower take 
full effect, we are proposing a NOX FEL cap of 0.60 g/bhp-hr 
for engines above 75 horsepower. (As described below, we are proposing 
to allow a small number of engines greater than 75 horsepower to have 
NOX FELs above the 0.60 g/bhp-hr cap beginning in model year 
2014.) Given the fact that the proposed Tier 4 NOX standard 
is approximately a 90 percent reduction from the existing standards for 
engines above 75 horsepower, we do not believe the previous standard 
would be appropriate as the FEL cap for all engines once the Tier 4 
standards are fully phased-in. We believe that the proposed 
NOX FEL caps will ensure that manufacturers adopt 
NOX aftertreatment technology across all of their engine 
designs (with the exception of a limited number) but will also allow 
for some meaningful use of averaging during the phase-in period. When 
compared to the proposed 0.30 g/bhp-hr NOX standard, the 
proposed NOX FEL cap of 0.60 g/bhp-hr (effective when the 
Tier 4 standards are fully phased-in) is consistent with FEL caps set 
in previous rulemakings.
    For the transitional PM standards being proposed for engines 
between 25 and 75 horsepower effective in model year 2008 and for the 
Tier 4 PM standards for engines below 25 horsepower, we are proposing 
the previously applicable Tier 2 PM standards (which do vary within the 
25 to 75 horsepower category) for the FEL caps since the gap between 
the previous and proposed standards is approximately 50 percent (rather 
than in excess of 90 percent for engines above 75 horsepower). For the 
proposed Tier 4 PM standard effective in model year 2013 for engines 
between 25 and 75 horsepower, we are proposing a PM FEL cap of 0.04 g/
bhp-hr, and for the proposed Tier 4 PM standard effective in model 
years 2011 and 2012 for engines between 75 and 750 horsepower, we are 
proposing a PM FEL cap of 0.03 g/bhp-hr. (As described below, we are 
proposing to allow a small number of Tier 4 engines greater than 25 
horsepower to have PM FELs above these caps.) Given the fact that the 
proposed Tier 4 PM standards for engines above 25 horsepower are less 
than 10 percent of the previous standards, we do not believe the 
previous standards would be appropriate as FEL caps once the Tier 4 
standards take effect. We believe that the proposed PM FEL caps will 
ensure that manufacturers adopt PM aftertreatment technology across all 
of their engine designs (except for a limited number of engines), yet 
will still provide substantial flexibility in meeting the standards.
    For the proposed Tier 4 PM standards for engines above 750 
horsepower there is a phase-in period during model years 2011 through 
2013. During the phase-in period, there would be two separate sets of 
engines with different FEL caps. For engines certified to the existing 
Tier 2 PM standard, the FEL cap would continue to be the existing PM 
FEL cap adopted in the October 1998 rule. For engines certified to the 
proposed Tier 4 PM standard during the phase-in, the FEL cap would be 
0.15 g/bhp-hr (the PM standard for the previous tier). Beginning in 
model year 2014, when the proposed Tier 4 PM standard for engines above 
750 horsepower takes full effect, consistent with the proposed caps for 
lower horsepower categories, we are proposing a PM FEL cap of 0.03 g/
bhp-hr. (As described below, we are proposing to allow a small number 
of engines greater than 750 horsepower to have PM FELs above the 0.03 
g/bhp-hr cap beginning in model year 2014.) We believe that the 
proposed PM FEL caps for engines above 750 horsepower will ensure that 
manufacturers adopt PM aftertreatment technology across all of their 
engine designs once the standard is fully phased-in (with the exception 
of a limited number) while allowing for some meaningful use of 
averaging during the phase-in period.
    Table VII.A-1 contains the proposed FEL caps and the effective 
model year

[[Page 28468]]

for the FEL caps (along with the associated standards proposed for Tier 
4). We request comment on the need for and the levels of these proposed 
FEL caps. It should be noted that for Tier 4, where we are proposing a 
new transient test, as well as retaining the current steady-state test, 
the FEL established by the engine manufacturer would be used as the 
enforceable limit for the purpose of compliance testing under both test 
cycles. In addition, under the NTE requirements, the FEL times the 
appropriate multiplier would be used as the enforceable limit for the 
purpose of such compliance testing.

                                 Table VII.A-1.--Proposed FEL Caps for the Proposed Tier 4 Standards in the ABT Program
                                                                       [g/bhp-hr]
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                              NOX
            Power category                    Effective model year          standard                 NOX FEL cap                PM standard   PM FEL cap
--------------------------------------------------------------------------------------------------------------------------------------------------------
hp < 25 (kW < 19).....................  2008+...........................        (\a\)  (\a\)..................................     \b\ 0.30         0.60
25 <= hp < 50 (19 <= kW < 37).........  2008-2012.......................        (\a\)  (\a\)..................................         0.22         0.45
25 <= hp < 50 (19 <= kW < 37).........  2013+\d\........................      \e\ 3.5  5.6 \e\................................         0.02     \f\ 0.04
50 <= hp < 75 (37 <= kW < 56).........  2008-2012.......................        (\a\)  (\a\)..................................         0.22         0.30
50 <= hp < 75 (37 <= kW < 56).........  2013+...........................        (\a\)  (\a\)..................................         0.02     \f\ 0.04
75 <= hp <175 (56 <= kW <130).........  2012-2013 \g\...................         0.30  3.3 for hp < 100 2.8 for hp = 100.
75 <= hp <175 (56 <= kW <130).........  2014+...........................         0.30  0.60 \f\...............................         0.01     \f\ 0.03
175 <= hp <=750 (130 <= kW <=560).....  2011-2013.......................         0.30  2.8....................................         0.01     \f\ 0.03
175 <= hp <=750 (130 <= kW <=560).....  2014+...........................         0.30  0.60 \f\...............................         0.01     \f\ 0.03
hp 750 (kW 560).  2011-2013.......................         0.30  4.6....................................         0.01         0.15
hp 750 (kW 560).  2014+...........................         0.30  0.60 \f\...............................         0.01     \f\ 0.03
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes:
\a\ The existing NMHC+NOX standard and FEL cap apply (see CFR Title 40, section 89.112).
\b\ A PM standard of 0.45 g/bhp-hr would apply to air-cooled, hand-startable, direct injection engines under 11 horsepower, effective in 2010.
\c\ The proposed FEL caps do not apply if the manufacturer elects to comply with the optional standards. The existing FEL caps continue to apply.
\d\ FEL caps apply in model year 2012 if the manufacturer elects to comply with the optional standards.
\e\ These are a combined NMHC+NOX standard and FEL cap.
\f\ As described in this section, a small number of engines are allowed to exceed these FEL caps.
\g\ This period would extend through the first nine months of 2014 under the alternative, reduced phase-in requirement (see Section III.B.1. for a
  description of the proposed alternative).

    As noted above, we are proposing to allow a limited number of 
engines to have a higher FEL than the caps noted in Table VII.A-1 in 
certain instances. Under this proposal, the allowance to certify up to 
these higher FEL caps would apply to Tier 4 engines at or above 25 
horsepower. The provisions are intended to provide some limited 
flexibility for engine manufacturers as they transition to the 
stringent standards while ensuring that the vast majority of engines 
are converted to the advanced low-emission technologies expected under 
the Tier 4 program. This additional lead time appears appropriate, 
given the potential that a limited set of nonroad engines may face 
especially challenging difficulties in complying, and considering 
further that the same amount of overall emission reductions would be 
achieved through the need for credit-generating nonroad engines.
    Beginning the first year Tier 4 standards apply in each power 
category above 25 horsepower, an engine manufacturer would be allowed 
to certify up to ten percent of its engines in each power category with 
PM FELs above the caps shown in Table VII.A-1. The PM FEL cap for such 
engines would instead be the applicable previous tier PM standard. The 
ten percent allowance would be available for the first four years the 
Tier 4 standards apply. For the power categories in which we are 
proposing a phase-in requirement for the Tier 4 NOX 
standards, the allowance to use a higher FEL cap would apply only to PM 
during the phase-in years. Once the phase-in period is complete, the 
allowance would apply to NOX as well. (For engines above 750 
horsepower, where we are proposing a phase-in for both NOX 
and PM, the allowance to use a higher FEL cap would not take effect 
until model year 2014 when the phase-in was complete.)
    After the fourth year the Tier 4 standards apply, the allowance to 
certify engines using the higher FEL caps would still be available but 
for no more than five percent of a manufacturer's engines in each power 
category. (For the power category between 25 and 75 horsepower, this 
allowance would apply beginning with the 2013 model year and would 
apply to PM. The allowance to use the higher FEL caps is not necessary 
for the 2008 proposed standards or the 2013 proposed 
NMHC+NOX standards because the FEL caps for those standards 
are set at the previously applicable tier standards.)
    Table VII.A-2 presents the model years, percent of engines, and 
higher FEL caps that would apply under this allowance. Because the 
engines certified with the higher FEL caps are certified to the Tier 4 
standards (albeit through the use of credits), they would be considered 
Tier 4 engines and all other requirements for Tier 4 engines would also 
apply, including the Tier 4 NMHC standard. We invite comment on whether 
additional provisions may be necessary for the limited number of 
engines certified to the higher FELs, including whether an averaging 
program for NMHC would be needed.

[[Page 28469]]



                                 Table VII.A-2.--Allowance for Limited Use of an FEL Cap Higher Than the Tier 4 FEL Caps
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                             Engines
                                                                           allowed to
            Power category                        Model years              have higher    NOX FEL cap (g/bhp-hr)           PM FEL cap (g/bhp-hr)
                                                                              FELs
--------------------------------------------------------------------------------------------------------------------------------------------------------
25 <= hp <75 (19 <= kW < 56).........  2013-2016.......................              10  Not applicable.........  0.22.
                                       2017+...........................               5                           ......................................
--------------------------------------
75 <= hp <175 (56 <= kW <130)........  2012-2013a......................              10  Not applicable.........  0.30 for hp <100.
                                      ---------------------------------------------------------------------------
                                       2014-2015.......................              10  3.3 for hp <100........  0.22 for hp =100.
                                      --------------------------------------------------
                                       2016+...........................               5  2.8 for hp =100.
--------------------------------------
175 <= hp <=750 (130 <= kW <= 560)...  2011-2013.......................              10  Not applicable.........  0.15.
                                      --------------------------------------------------
                                       2015+...........................               5                           ......................................
--------------------------------------
hp 750 (kW       2014-2017.......................              10  4.6....................  0.15.
 560).
                                      --------------------------------------------------
                                       2018+...........................               5                           ......................................
--------------------------------------------------------------------------------------------------------------------------------------------------------
a This period would extend through the first nine months of 2014 under the alternative, reduced phase-in requirement (see Section III.B.1. for a
  description of the proposed alternative).

    We request comment on the proposed provisions to allow higher FELs 
on a limited number of Tier 4 engines, including whether the proposed 
allowance limits of 10 percent and 5 percent have been set at the right 
levels and whether the allowance to use a higher FEL cap is appropriate 
for the Tier 4 program. We also request comment on allowing 
manufacturers to use the allowances in a slightly different manner over 
the first four years. Instead of allowing manufacturers to certify up 
to ten percent for each of the first four years, manufacturers could 
certify up to 40 percent of one year's production but spread it out 
over four years in an unequal manner (e.g., 15 percent in the first and 
second years, and 5 percent in the third and fourth years). Last of 
all, we request comment on whether the allowance should be available 
for NOX during the years we a proposing a phase-in for the 
Tier 4 NOX standards. As proposed, we would not cover 
NOX during the phase-in years because manufacturers already 
can certify up to 50 percent of their engines to the Tier 3 
NMHC+NOX standards.
    Under the proposed Tier 4 program, for engines above 75 horsepower 
there will be two different groups of engines during the phase-in 
period. In one group, engines would certify to the applicable Tier 3 
NMHC+NOX standard (or Tier 2 standard for engines above 750 
horsepower), and would be subject to the ABT restrictions and 
allowances previously established for those tiers. In the other group, 
engines would certify to the 0.30 g/bhp-hr NOX standard, and 
would be subject to the restrictions and allowances in this proposed 
program. While engines in each group are certified to different 
standards, we are proposing to allow manufacturers to transfer credits 
across these two groups of engines with the following adjustment. As 
proposed, manufacturers could use credits generated during the phase-
out of engines subject to the Tier 3 NMHC+NOX standard (or 
Tier 2 NMHC+NOX standard for engines above 750 horsepower) 
to average with engines subject to the 0.30 g/bhp-hr NOX 
standard, but these credits will be subject to a 20 percent discount. 
In other words, each gram of NMHC+NOX credits from the 
phase-out engines would be worth 0.8 grams of NOX credits in 
the new ABT program. The ability to average credits between the two 
groups of engines will give manufacturers a greater opportunity to gain 
experience with the low-NOX technologies before they are 
required to meet the final Tier 4 standards across their full 
production. (The 20 percent discount would also apply to 
NMHC+NOX credits generated on less than 75 horsepower 
engines and used for averaging purposes with the NOX 
standards for engines greater than 75 horsepower.)
    We are proposing the 20 percent discount for two main reasons. 
First, the discounting addresses the fact that NMHC reductions can 
provide substantial NMHC+NOX credits, which are then treated 
as though they were NOX credits. For example, a 2010 model 
year engine (between 175 and 750 horsepower) emitting at 2.7 g/bhp-hr 
NOX and 0.3 g/bhp-hr NMHC meets the 3.0 g/bhp-hr 
NMHC+NOX standard in that year, but gains no credits. In 
2011, that engine, equipped with a PM trap to meet the new PM standard, 
will have very low NMHC emissions because of the trap, an emission 
reduction already accounted for in our assessment of the air quality 
benefit of this program. As a result, without substantially redesigning 
the engine to reduce NOX or NMHC, the manufacturer could 
garner a windfall of nearly 0.3 g/bhp-hr of NMHC+NOX credit 
for each of these engines produced. (Engines designed at lower 
NOX levels than this in 2010 can gain even more credits.) 
Allowing these NMHC-derived credits to be used undiscounted to offset 
NOX emissions on the phase-in engines in 2011 (for which 
each 0.1 g/bhp-hr of margin can make a huge difference in facilitating 
the design of engines to meet the 0.30 g/bhp-hr NOX 
standard) would be inappropriate. Second, the discounting would work 
toward providing a net environmental benefit from the ABT program, such 
that the more that manufacturers use banked and averaged credits, the 
greater the potential emission reductions overall.
    Some foreign engine manufacturers have commented that it is 
difficult for them to accurately predict the number of engines that 
eventually end up in the U.S., especially when they sell to a number of 
different equipment manufacturers who may import equipment. This would 
make it difficult for the engine manufacturer to ensure they are 
complying with the proposed NOX phase-in requirements for 
engines above 75 horsepower and the proposed PM phase-in requirements 
for engines above 750 horsepower. Therefore, we are proposing to allow 
engine

[[Page 28470]]

manufacturers to demonstrate compliance with the NOX phase 
in requirements for engines above 75 horsepower and the PM phase in 
requirements for engines above 750 horsepower by certifying ``split'' 
engine families (i.e., an engine family that is split into two equal-
sized subfamilies, one that generates a number of credits and one that 
uses an equal number of credits). In order to facilitate compliance 
with the proposed standards, we are proposing that this option be 
available to all engine manufacturers (i.e., both foreign and do