Emissions and Fuel Standards: Control of Air Pollution from Motor Vehicles

This Rule document was issued by the Environmental Protection Agency (EPA)

For related information, Open Docket Folder


ENVIRONMENTAL PROTECTION AGENCY
40 CFR Parts 79, 80, 85, 86, 600, 1036, 1037, 1039, 1042, 1048, 1054, 1065, and 1066
[EPA-HQ-OAR-2011-0135; FRL 9906-86-OAR]
RIN 2060-AQ86

Control of Air Pollution From Motor Vehicles: Tier 3 Motor Vehicle Emission and Fuel Standards

Agency

Environmental Protection Agency (EPA).

Action

Final rule.

Summary

This action establishes more stringent vehicle emissions standards and will reduce the sulfur content of gasoline beginning in 2017, as part of a systems approach to addressing the impacts of motor vehicles and fuels on air quality and public health. The gasoline sulfur standard will make emission control systems more effective for both existing and new vehicles, and will enable more stringent vehicle emissions standards. The vehicle standards will reduce both tailpipe and evaporative emissions from passenger cars, light-duty trucks, medium-duty passenger vehicles, and some heavy-duty vehicles. This will result in significant reductions in pollutants such as ozone, particulate matter, and air toxics across the country and help state and local agencies in their efforts to attain and maintain health-based National Ambient Air Quality Standards. Motor vehicles are an important source of exposure to air pollution both regionally and near roads. These vehicle standards are intended to harmonize with California's Low Emission Vehicle program, thus creating a federal vehicle emissions program that will allow automakers to sell the same vehicles in all 50 states. The vehicle standards will be implemented over the same timeframe as the greenhouse gas/fuel efficiency standards for light-duty vehicles (promulgated by EPA and the National Highway Safety Administration in 2012), as part of a comprehensive approach toward regulating emissions from motor vehicles.

Dates

This final rule is effective on June 27, 2014. The incorporation by reference of certain publications listed in this regulation is approved by the Director of the Federal Register as of June 27, 2014.

Addresses

EPA has established a docket for this action under Docket ID No. EPA-HQ-OAR-2011-0135. All documents in the docket are listed on the Web site. Although listed in the index, some information is not publicly available, e.g., CBI or other information whose disclosure is restricted by statute. Certain other material, such as copyrighted material, is not placed on the Internet and will be publicly available only in hard copy form. Publicly available docket materials are available either electronically in or in hard copy at the Air and Radiation Docket and Information Center, EPA/DC, EPA West, Room 3334, 1301 Constitution Ave. NW., Washington, DC. The Public Reading Room is open from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal holidays. The telephone number for the Public Reading Room is (202) 566-1744, and the telephone number for the Air Docket is (202) 566-1742.

For Further Information Contact

JoNell Iffland, Office of Transportation and Air Quality, Assessment and Standards Division (ASD), Environmental Protection Agency, 2000 Traverwood Drive, Ann Arbor MI 48105; Telephone number: (734) 214-4454; Fax number: (734) 214-4816; Email address: .

Supplementary Information

I. General Information

A. Does this action apply to me?

Entities potentially affected by this rule include gasoline refiners and importers, ethanol producers, ethanol denaturant producers, butane and pentane producers, gasoline additive manufacturers, transmix processors, terminals and fuel distributors, light-duty vehicle manufacturers, independent commercial importers, alternative fuel converters, and manufacturers and converters of vehicles between 8,500 and 14,000 lbs gross vehicle weight rating (GVWR).

Potentially regulated categories include:

CategoryNAICS a CodeSIC b CodeExamples of potentially affected entities
Industry3241102911Petroleum refineries (including importers).
Industry3251102869Butane and pentane manufacturers.
Industry3251932869Ethyl alcohol manufacturing.
Industry324110, 2111122911, 1321Ethanol denaturant manufacturers.
Industry2111121321Natural gas liquids extraction and fractionation.
Industry3251992869Other basic organic chemical manufacturing.
Industry4869104613Natural gas liquids pipelines, refined petroleum products pipelines.
Industry4246905169Chemical and allied products merchant wholesalers.
Industry3251992869Manufacturers of gasoline additives.
Industry4247105171Petroleum bulk stations and terminals.
Industry4931904226Other warehousing and storage-bulk petroleum storage.
Industry336111, 3361123711Light-duty vehicle and light-duty truck manufacturers.
Industry811111, 811112, 8111987538, 7533, 7534Independent commercial importers.
Industry335312, 336312, 336322, 336399, 8111983621, 3714, 3519, 3599, 7534Alternative fuel converters.
Industry333618, 336120, 336211, 3363123699, 3711, 3713, 3714On-highway heavy-duty engine & vehicle (>8,500 lbs GVWR) manufacturers.

This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely to be regulated by this action. This table lists the types of entities that EPA is now aware could potentially be regulated by this action. Other types of entities not listed in the table could also be regulated. To determine whether your activities are regulated by this action, you should carefully examine the applicability criteria in 40 CFR parts 79, 80, 85, 86, 600, 1036, 1065, and 1066 and the referenced regulations. If you have any questions regarding the applicability of this action to a particular entity, consult the person listed in the preceding FOR FURTHER INFORMATION CONTACT section.

B. Did EPA conduct a peer review before issuing this action?

This regulatory action was supported by influential scientific information. Therefore, EPA conducted peer reviews in accordance with OMB's Final Information Quality Bulletin for Peer Review. EPA conducted several peer reviews in connection with data supporting the Tier 3 program, including new research on the effects of fuel properties changes (including sulfur effects) on exhaust and evaporative emissions of Tier 2 vehicles. The refinery-by-refinery cost model was also peer reviewed. The peer review reports are located in the docket for today's action, as well as the agency's response to the peer review comments.

Table of Contents

I. Executive Summary and Program Overview

A. Introduction

B. Overview of the Tier 3 Program

1. Major Public Comments and Key Changes From the Proposal

2. Key Components of the Tier 3 Program

C. What will the impacts of the standards be?

II. Why is EPA taking this action?

A. Basis for Action Under the Clean Air Act

1. Clean Air Act Section 202

2. Clean Air Act Section 211

B. Overview of Public Health Impacts of Motor Vehicles and Fuels

1. Ozone

2. Particulate Matter

3. Oxides of Nitrogen and Sulfur

4. Carbon Monoxide

5. Mobile Source Air Toxics

6. Near-Roadway Pollution

7. Environmental Impacts of Motor Vehicles and Fuels

III. How would this rule reduce emissions and air pollution?

A. Effects of the Vehicle and Fuel Changes on Mobile Source Emissions

1. How do vehicles produce the emissions addressed in this action?

2. How will the changes to gasoline sulfur content affect vehicle emissions?

B. How will emissions be reduced?

1. NO X

2. VOC

3. CO

4. Direct PM 2.5

5. Air Toxics

6. SO 2

7. Greenhouse Gases

C. How will air pollution be reduced?

1. Ozone

2. Particulate Matter

3. Nitrogen Dioxide

4. Air Toxics

5. Visibility

6. Nitrogen and Sulfur Deposition

7. Environmental Justice

IV. Vehicle Emissions Program

A. Tier 3 Tailpipe Emission Standards for Light-Duty Vehicles, Light-Duty Trucks, and Medium-Duty Passenger Vehicles

1. How the Tier 3 Program is harmonized with the California LEV III Program

2. Summary of the Tier 3 FTP and SFTP Tailpipe Standards

3. FTP Standards

4. SFTP Standards

5. Feasibility of the NMOG+NO X and PM Standards

6. Impact of Gasoline Sulfur Control on the Effectiveness of the Vehicle Emission Standards

7. Other Provisions

B. Tailpipe Emissions Standards for Heavy-Duty Vehicles

1. Overview and Scope of Vehicles Regulated

2. HDV Exhaust Emissions Standards

3. Supplemental FTP Standards for HDVs

4. HDV Emissions Averaging, Banking, and Trading

5. Feasibility of HDV Standards

6. Other HDV Provisions

C. Evaporative Emissions Standards

1. Tier 3 Evaporative Emission Standards

2. Program Structure and Implementation Flexibilities

3. Technological Feasibility

4. Heavy-Duty Gasoline Vehicle (HDGV) Requirements

5. Evaporative Emission Requirements for FFVs

6. Test Procedures and Certification Test Fuel

D. Improvements to In-Use Performance of Fuel Vapor Control Systems

1. Reasons for Adding a Leak Test Standard

2. Nature, Scope and Timing of Leak Standard

3. Leak Standard Test Procedure

4. Certification and Compliance

a. In-Use Verification Program (IUVP) Requirements for the Leak Standard

E. Onboard Diagnostic System Requirements

1. Onboard Diagnostic (OBD) System Regulation Changes—Timing

2. Revisions to EPA OBD Regulatory Requirements

3. Provisions for Emergency Vehicles

4. Future Considerations

F. Emissions Test Fuel

1. Gasoline Emissions Test Fuel: Ethanol Content and Volatility

2. Other Gasoline Emissions Test Fuel Specifications

3. Flexible Fuel Vehicle Exhaust Emissions Test Fuel

4. Implementation Schedule

5. Implications of Emission Test Fuel Changes on CAFE Standards, GHG Standards, and Fuel Economy Labels

6. Consideration of Test Fuel for Nonroad Engines and Highway Motorcycles

7. CNG and LPG Emissions Test Fuel Specifications

G. Small Business Provisions

1. Lead Time and Relaxed Interim Standards

2. Assigned Deterioration Factors

3. Reduced Testing Burden and OBD Requirements

4. Hardship Relief

5. Eligibility for the Flexibilities

H. Compliance Provisions

1. Exhaust Emission Test Procedures

2. Reduced Test Burden

3. Miscellaneous Provisions

4. Manufacturer In-Use Verification Program (IUVP) Requirements

V. Fuel Program

A. Overview

1. Background

2. Summary of Final Tier 3 Fuel Program Standards

B. Annual Average Sulfur Standard

C. Per-Gallon Sulfur Caps

1. Standards

2. Requirements for Gasoline Additives

D. Averaging, Banking, and Trading Program

1. How will the ABT Program assist with compliance?

2. ABT Modeling

3. Eligibility

4. Credit Generation and Use

5. Credit Trading Provisions

6. ABT Provisions for Small Refiners and Small Volume Refineries

7. Deficit Carryforward

E. Additional Program Flexibilities

1. Regulatory Flexibility Provisions

2. Provisions for Refiners Facing Hardship Situations

F. Compliance Provisions

1. Registration, Reporting, and Recordkeeping Requirements

2. Sampling and Testing Requirements

3. Small Refiner Compliance

4. Small Volume Refinery Compliance

5. Attest Engagements, Violations, and Penalties

6. Special Fuel Provisions and Exemptions

G. Standards for Oxygenates (Including Denatured Fuel Ethanol) and Certified Ethanol Denaturants

H. Standards for Fuel Used in Flexible Fueled Vehicles

I. Sulfur Standards for Purity Butane and Purity Pentane Streams Blended into Gasoline

J. Standards for CNG and LPG

K. Refinery Air Permitting Interactions

1. Proposal

2. Updated Assessment of Tier 3 Refinery Changes and Permitting Implications

3. Comments and Responses

L. Refinery Feasibility

1. Comments Received

2. Is it feasible for refiners to comply with a 10 ppm average sulfur standard?

3. Can refiners meet the January 1, 2017 start date?

M. Statutory Authority for Tier 3 Fuel Controls

1. Section 211(c)(1)(A)

2. Section 211(c)(1)(B)

3. Section 211(c)(2)(B)

4. Section 211(c)(2)(C)

VI. Technical Amendments and Regulatory Streamlining

A. Fuel Program Amendments

1. Fuels Program Regulatory Streamlining

2. Performance-Based Measurement Systems (PBMS)

3. Downstream Pentane Blending

4. Acceptance of Top Tier Deposit Control Test Data

5. Potential Broader Regulatory Streamlining Through Program Restructuring

B. Engine, Vehicle and Equipment Programs Amendments

1. Fuel Economy Labeling

2. Removing Obsolete Regulatory Text

3. Motorcycle Driving Schedules

4. Updating Reference Procedures

VII. What are the cost impacts of the rule?

A. Estimated Costs of the Vehicle Standards

1. What changes have been made to vehicle program costs since proposal?

2. Summary of Vehicle Program Costs

B. Estimated Costs of the Fuel Program

1. Overview

2. Methodology

3. Fuel Program Costs

4. Other Cost Estimates

C. Summary of Program Costs

VIII. What are the estimated benefits of the rule?

A. Overview

B. Quantified Human Health Impacts

C. Monetized Benefits

D. What are the limitations of the benefits analysis?

E. Illustrative Analysis of Estimated Monetized Impacts Associated With the Rule in 2018

IX. Alternatives Analysis

A. Vehicle Emission Standards

1. Shorter NMOG+NO X Standard Phase-in

2. NMOG+NO X Standards Phase-in and Early Tier 3 Credits

3. NMOG+NO X Standards

4. PM Standards

5. Higher Ethanol Content of Emissions Test Fuel

B. Fuel Sulfur Standards

1. Annual Average Sulfur Standard

2. Refinery Gate Sulfur Cap

C. Program Start Date

X. Economic Impact Analysis

A. Introduction

B. Vehicle Sales Impacts

C. Impacts on Petroleum Refinery Sector Production

D. Employment Impacts

1. Employment Impacts in the Auto Sector

2. Refinery Employment Impacts

XI. Public Participation

XII. Statutory and Executive Order Reviews

A. Executive Order 12866: Regulatory Planning and Review and Executive Order 13563: Improving Regulation and Regulatory Review

B. Paperwork Reduction Act

C. Regulatory Flexibility Act

1. Overview

2. Background

3. Reason for Today's Rule

4. Legal Basis for Agency Action

5. Summary of Potentially Affected Small Entities

6. Reporting, Recordkeeping, and Compliance

7. Related Federal Rules

8. Steps Taken To Minimize the Economic Impact on Small Entities

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 Risks and Safety Risks

H. Executive Order 13211: Actions Concerning Regulations That Significantly Affect Energy Supply, Distribution, or Use

I. National Technology Transfer and Advancement Act

J. Executive Order 12898: Federal Actions To Address Environmental Justice in Minority Populations and Low-Income Populations

K. Congressional Review Act

XIII. Statutory Provisions and Legal Authority

I. Executive Summary and Program Overview

A. Introduction

In this action, EPA is finalizing a major program designed to reduce air pollution from passenger cars and trucks. This program includes new standards for both vehicle emissions and the sulfur content of gasoline, considering the vehicle and its fuel as an integrated system. We refer to this program as the “Tier 3” vehicle and fuel standards.

This rule is part of a comprehensive approach to address the impacts of motor vehicles on air quality and public health. Over 149 million Americans are currently experiencing unhealthy levels of air pollution, which are linked with respiratory and cardiovascular problems and other adverse health impacts that lead to increased medication use, hospital admissions, emergency department visits, and premature mortality. (1) Motor vehicles are a particularly important source of exposure to air pollution, especially in urban areas. By 2018, we project that in many areas that are not attaining health-based ambient air quality standards (i.e., “nonattainment areas”), passenger cars and light trucks will contribute 10-25 percent of total nitrogen oxides (NO X) emissions, 15-30 percent of total volatile organic compound (VOC) emissions, and 5-10 percent of total direct particulate matter (PM 2.5) emissions. (2) These compounds form ozone, PM, and other air pollutants, whose health and environmental effects are described in more detail in Section II. Cars and light trucks also continue to be a significant contributor to air pollution directly near roads, with gasoline vehicles accounting for more than 50 percent of near-road concentrations of some criteria and toxic pollutants. (3) More than 50 million people live, work, or go to school in close proximity to high-traffic roadways, and the average American spends more than one hour traveling along roads each day. (4 5) Over 80 percent of daily trips use personal vehicles. (6)

The standards set forth in this rule will significantly reduce levels of multiple air pollutants (such as ambient levels of ozone, PM, nitrogen dioxide (NO 2), and mobile source air toxics (MSATs)) across the country, with immediate benefits from the gasoline sulfur control standards starting in 2017. These reductions will help state and local agencies in their effort to attain and maintain health-based National Ambient Air Quality Standards (NAAQS). Few other national strategies exist that will deliver the same magnitude of multi-pollutant reductions and associated public health protection that is projected to result from the Tier 3 standards. Without this action to reduce nationwide motor vehicle emissions, areas would have to adopt other, less cost-effective measures to reduce emissions from other sources under their state or local authority. In the absence of additional controls, certain areas would continue to have ambient ozone concentrations exceeding the NAAQS in the future. See Section III.C for more details.

The Clean Air Act authorizes EPA to establish emissions standards for motor vehicles to address air pollution that may reasonably be anticipated to endanger public health or welfare (section 202). EPA also has authority to establish fuel controls to address such air pollution (section 211). These statutory authorities are described in Section II.A.

The vehicle and gasoline sulfur standards we are finalizing represent a “systems approach” to reducing vehicle exhaust and evaporative emissions by addressing the vehicle and fuel as a system. The systems approach enables emission reductions that are both technologically feasible and cost-effective beyond what would be possible looking at vehicle and fuel standards in isolation. We first applied such an approach with our Tier 2 vehicle/gasoline sulfur standards (finalized in 2000). (7) We believe that a similar approach for the Tier 3 standards is a cost-effective way to achieve substantial additional emissions reductions.

The Tier 3 standards include new light- and heavy-duty vehicle emission standards for exhaust emissions of VOC (specifically, non-methane organic gases, or NMOG), NO X, and PM, as well as new evaporative emissions standards. The fully phased-in standards for light-duty vehicle, light-duty truck, and medium-duty passenger vehicle tailpipe emissions are an 80 percent reduction in fleet average NMOG+NO X compared to current standards, and a 70 percent reduction in per-vehicle PM standards. The fully phased-in Tier 3 heavy-duty vehicle tailpipe emissions standards for NMOG+NO X and PM are on the order of 60 percent lower than current standards. Finally, the fully phased-in evaporative emissions standards represent a 50 percent reduction from current standards.

The vehicle emission standards, combined with the reduction of gasoline sulfur content from the current 30 parts per million (ppm) average down to a 10 ppm average, will result in dramatic emissions reductions for NO X, VOC, direct PM 2.5, carbon monoxide (CO) and air toxics. For example, in 2030, when Tier 3 vehicles will make up the majority of the fleet as well as vehicle miles traveled, NO X and VOC emissions from on-highway vehicles will be reduced by about 21 percent, and CO emissions will be reduced by about 24 percent. National emissions of many air toxics from on-highway vehicles will also be reduced by 10 to nearly 30 percent. Reductions will continue beyond 2030 as more of the fleet is composed of vehicles meeting the fully phased-in Tier 3 standards. For example, the Tier 3 program will reduce on-highway emissions of NO X and VOC nearly 31 percent by 2050, when vehicles meeting the fully phased-in Tier 3 standards will comprise almost the entire fleet.

Gasoline vehicles depend to a great degree on catalytic converters to reduce levels of pollutants in their exhaust, including NMOG and NO X, as well as PM (specifically, the volatile hydrocarbon fraction), CO, and most air toxics. The catalytic converters become significantly less efficient when exposed to sulfur. The Tier 2 rulemaking required refiners to take steps to reduce sulfur levels in gasoline by approximately 90 percent, to an average of 30 ppm. As discussed in Section IV.A.6, subsequent research provides a compelling case that even this level of sulfur not only degrades the emission performance of vehicles on the road today, but also inhibits necessary further reductions in vehicle emissions performance to reach the Tier 3 standards. Thus, the 10 ppm average sulfur standard for Tier 3 is significant in two ways: it enables vehicles designed to the Tier 3 tailpipe exhaust standards to meet these standards in-use for the duration of their useful life, and it facilitates immediate emission reductions from all the vehicles on the road at the time the fuel sulfur controls are implemented. EPA is not the first regulatory agency to recognize the need for lower-sulfur gasoline. Agencies in Europe and Japan have already imposed gasoline sulfur caps of 10 ppm, and the State of California is already averaging 10 ppm sulfur with a per gallon cap of 20 ppm. Other states are preempted by the Clean Air Act from adopting new fuel programs to meet air quality objectives. Consequently, they could not receive the air quality benefits of lower sulfur gasoline without federal action.

This action is one aspect of a comprehensive national program regulating emissions from motor vehicles. EPA's final rule for reducing greenhouse gas (GHG) emissions from light-duty (LD) vehicles starting with model year (MY) 2017 (referred to here as the “2017 LD GHG” standards) is another aspect of this comprehensive program. (8) The Tier 3 program addresses interactions with the 2017 LD GHG rule in a manner that aligns implementation of the two actions, to achieve significant criteria pollutant and GHG emissions reductions while providing regulatory certainty and compliance efficiency. As vehicle manufacturers introduce new vehicle platforms for compliance with the GHG standards, they will be able to design them for compliance with the Tier 3 standards at the same time. The Tier 3 standards are also closely coordinated with California's Low Emission Vehicle (LEV) III program to create a vehicle emissions program that will allow automakers to sell the same vehicles in all 50 states. (In December 2012 EPA approved a waiver of Clean Air Act preemption for the California Air Resources Board's (CARB's) LEV III program with compliance beginning in 2015. Twelve states adopted the LEV III program under Section 177 of the Clean Air Act. (9) ) We have worked closely with individual vehicle manufacturers and their trade associations, who have emphasized the importance of a harmonized national program. Together, the Tier 3, 2017 LD GHG, and LEV III standards will provide significant reductions in GHGs, criteria pollutants and air toxics from motor vehicles while streamlining programs and enabling manufacturers to design a single vehicle for nationwide sales, thus reducing their costs of compliance. In this way, the Tier 3 program responds to the May 21, 2010 Presidential Memorandum that requested that EPA develop a comprehensive approach toward regulating motor vehicles, including consideration of non-GHG emissions standards. (10)

As part of the systems approach to this program, we have considered the types of fuels on which vehicles will be operating in the future. In particular, the renewable fuels mandate that was revised by the Energy Independence and Security Act (EISA) and is being implemented through the Renewable Fuel Standards program (RFS2)  (11) is resulting in the use of significant amounts of ethanol-blended gasoline. We are updating the specifications of the emissions test fuel with which vehicles demonstrate compliance with emissions standards, in order to better reflect the ethanol content and other properties of gasoline that is in use today and is expected in future years.

Section I provides an overview of the vehicle and fuel standards we are finalizing as well as the impacts of the standards. The public health issues and statutory requirements that have prompted this action are described in Section II, and our discussion of how the Tier 3 standards will reduce emissions and air pollution is presented in Section III. Details of the standards and how they will be implemented can be found in Sections IV through VI. Sections VII through X contain our discussion of the standards' technological feasibility and costs, benefits, and economic impacts. Sections XI through XIII address public participation, statutory and executive orders, and statutory provisions and legal authority under the Clean Air Act covered in this rulemaking.

This final rule is based on extensive public input received in response to EPA's Tier 3 proposal. The proposal was signed and posted on the EPA Web site on March 29, 2013, and published in the Federal Register on May 21, 2013. EPA held two public hearings in Philadelphia and Chicago in April 2013. In response to stakeholder requests, EPA extended the public comment period to July 1, 2013. We received more than 200,000 public comments. A broad range of stakeholders provided comments, including state and local governments, auto manufacturers, emissions control suppliers, refiners, fuel distributors and others in the petroleum industry, renewable fuels providers, environmental organizations, consumer groups, labor groups, private citizens, and others. Some of the issues raised in comments included lead time and the program's start date, the vehicle manufacturers' support for a 50-state program harmonized with California, the need for and degree of gasoline sulfur control (including the level of the sulfur cap), the ethanol content of vehicle certification test fuel, and various details on the flexibilities and other program design features of both the vehicle and fuels standards.

B. Overview of the Tier 3 Program

In the 14 years since EPA established the Tier 2 Vehicle Program, manufacturers of light-duty vehicles and automotive technology suppliers have continued to develop a wide range of improved technologies capable of reducing vehicle emissions. The California LEV II program has been instrumental in the continuous technology improvements by requiring year after year reductions in fleet average hydrocarbon levels, in addition to requiring the introduction of advanced exhaust and evaporative emission controls in partial zero emission vehicles (PZEVs). This technological progress has made it possible for manufacturers to achieve emission reductions well beyond the requirements of the Tier 2 program if gasoline sulfur levels are lowered further.

As a result, in conjunction with lower gasoline sulfur standards, we are establishing new Tier 3 standards for exhaust emissions of NMOG, NO X, and PM, as well as for evaporative hydrocarbon emissions. These vehicle emissions standards will phase in beginning with MY 2017. The structure of the Tier 3 standards is very similar to that of the existing Tier 2 program. As with the Tier 2 program, the standards will apply to all light-duty vehicles (LDVs, or passenger cars), light-duty trucks (LDT1s, LDT2s, LDT3s, and LDT4s) and Medium-Duty Passenger Vehicles (MDPVs). We also are establishing separate but closely related standards for heavy-duty vehicles up to 14,000 lbs Gross Vehicle Weight Rating (GVWR). (12) We have concluded that the vehicle emissions standards, in conjunction with the reductions in fuel sulfur also required by this action, are feasible across the fleet in the timeframe provided.

Auto manufacturers have stressed the importance of being able to design, produce, and sell a single fleet of vehicles in all 50 states that complies with both the Tier 3 and California LEV III programs, as well as the greenhouse gas (GHG)/Corporate Average Fuel Economy (CAFE) programs in the same timeframe. To that end, we worked closely with the California Air Resources Board and vehicle manufacturers to align the two programs as closely as possible. This consistency among the federal and California programs means that manufacturers do not need to design unique versions of vehicles with different emission control hardware and calibrations for different geographic areas. This allows manufacturers to avoid the additional costs of parallel design, development, calibration, and manufacturing. We also have designed the Tier 3 program to be implemented in the same timeframe as the GHG emissions and fuel economy standards for model years 2017-2025. We expect that in response to these programs, manufacturers will be developing entirely new powertrains for most of their vehicles. Because the Tier 3 standards will phase in over the same timeframe, manufacturers are in a better position to simultaneously respond to all of these requirements.

Overall, the final Tier 3 program is very similar to the program we proposed. As discussed below and throughout this preamble, the program phases in over several years—with the primary vehicle emission standards starting in Model Year (MY) 2017 (2018 for heavier vehicles) and the gasoline sulfur control provisions beginning in 2017.

As discussed above, we received a large number and wide range of comments on the proposed rule. Several comments raise particularly significant issues concerning some fundamental components of the Tier 3 program, including when the vehicle-related and fuel-related requirements begin. We briefly discuss these key issues in this section, and in more detail later in this preamble. The Summary and Analysis of Comments document provides our responses to the comments we received; it is located in the docket for this rulemaking and also on EPA's Web site at .

1. Major Public Comments and Key Changes From the Proposal
a. Start Date and Lead Time Issues
(1) Gasoline Sulfur Control Program

Many stakeholders commented on the proposed 2017 start date of the Tier 3 program, with state and NGO organizations supporting finalizing the standards as proposed. Conversely, refiners, importers, and others in the fuel industry commented that they believed the proposed start date would not provide a sufficient amount of lead time to meet the requirements of the Tier 3 program, and that EPA has historically provided at least four years of lead time in previous fuels rulemakings. These commenters noted that five years of lead time is needed to allow for necessary refinery changes to be made during a refinery's normal turnaround/shutdown schedule (these occur every four years, on average) and to allow adequate time for the permitting process. These commenters also stated that, given the proposed flexibility provisions for vehicles, that a 2017 fuel program start date was not truly needed to enable the vehicle technology. Further, these commenters stated that they believed insufficient lead time would drive up the costs for regulated entities as they would need to do unscheduled shutdowns to install and/or revamp equipment to meet the proposed standards. Lastly, they stated that the uncertainty regarding the potential availability of credits would make meeting a 2017 start date more challenging.

As discussed in greater detail in Section V below, we are finalizing the proposed start date of January 1, 2017. We understand refiners' concerns, including their concerns over the necessary capital investments and potential off-cycle turnarounds/shutdowns to make refinery modifications for Tier 3. In light of these concerns, we are finalizing additional flexibilities beyond those already in the proposal and we are confident that the program being finalized today addresses these concerns. Considering all the flexibilities offered to regulated parties, there is, in effect, nearly 6 years of time to comply provided for those refineries that may need it. As discussed in Section V.D, we are finalizing a credit averaging, banking, and trading (ABT) program that will allow for a smooth transition from the Tier 2 to Tier 3 ABT programs (including provisions for early credit generation beginning in 2014). These early credit provisions, coupled with the ability to carry over credits from Tier 2 into Tier 3 (an additional flexibility being finalized today that was not part of the proposal), will allow for early actions to reduce sulfur levels by some refineries to be used to delay the need for actions at other refineries until 2020. This structure of the ABT program allows refiners and importers the flexibility to choose the most economical compliance strategy—investment in technology, use of credits, or both—for meeting the Tier 3 average gasoline sulfur standard. In addition, approved small refiners and small volume refineries are given an additional three years from the January 1, 2017, Tier 3 program start date to comply (January 1, 2020).

We proposed that the Tier 2 ABT program would not only be separate from the Tier 3 ABT program, but that it would also end at the start of the Tier 3 program in 2017. The implications of this meant that any Tier 2 credits generated after 2012 would run the risk of expiring before the end of their full five-year life if they were not used before January 1, 2017. Commenters requested that EPA consider allowing such Tier 2 “banked” credits to receive their full five-year life. This would eliminate any incentive refiners may have to use these credits prior to the end of the Tier 2 program to raise their in-use sulfur levels. The ABT program that we are finalizing today enables a seamless transition from Tier 2 to Tier 3, including an allowance for Tier 2 banked credits to be used for their full five-year life or through December 31, 2019, whichever is earlier. Not only does this provision effectively provide more lead time and flexibility for refiners and importers, but we believe these banked credits will help to provide certainty of the availability of credits for refiners and importers who may want to rely on them for compliance.

Finally, as discussed in Section V.E.2, we are also finalizing hardship provisions that allow refiners to petition for delayed compliance, on a case-by-case basis, for situations of extreme hardship or extreme unforeseen circumstances. These provisions, similar to those implemented in past fuel rulemakings, provide a safety valve should all the other flexibilities provided prove insufficient. As part of these hardship provisions, we are finalizing the ability for refiners to carry a deficit for up to 3 years, providing them with yet additional flexibility during the transition to Tier 3 should it prove necessary.

(2) Vehicle Emission Control Program

There were no major concerns raised for the proposed MY 2017 start date for lighter light-duty vehicles, although commenters from the auto manufacturing industry raised concerns about the lead time we proposed for heavier light-duty vehicles. Specifically, commenters pointed to Clean Air Act section 202(a)(3)(C) that, for vehicles over 6,000 lbs GVWR, requires that EPA emission standards provide at least four years of lead time and three years of regulatory stability.

In light of this statutory requirement, in addition to the primary declining fleet average standards starting in MY 2018 for heavier vehicles, EPA proposed an alternative phase-in schedule for any manufacturer that prefers a longer lead time and annual stability for these vehicles in lieu of the declining fleet average standards option. The commenters stated that the proposed alternative pathway would be too difficult to take advantage of in comparison to the primary program and thereby failed to comply with the Clean Air Act.

In considering these comments, EPA also considered that during the development of the Tier 3 program and in their comments, the same auto industry commenters consistently urged EPA to design the Tier 3 program to harmonize with the California LEV III standards as closely and as early as possible. As discussed in detail below in Section IV.A, extensive data that EPA has generated or received continue to support the conclusion that the primary fleet-average standards provide a compliance path that is feasible across the industry and that closely harmonizes with LEV III. EPA believes that we have reasonably resolved these somewhat competing concerns—early harmonization vs. additional lead time—by finalizing the primary declining fleet average standards as proposed while also finalizing revised alternative phase-in compliance schedules (see Section IV.A.2.c). In response to the comments on this topic, we have revised the alternative phase-in schedules to reduce their associated burden for manufacturers, while still maintaining environmental benefits that are equivalent to the primary program. We also include provisions in the percent-of-sales phase-in alternatives that allow manufacturers to exclude vehicle models that begin their 2019 model year production early in 2018, in order to provide four years of lead time.

b. Emissions Test Fuel

In-use gasoline has changed considerably since EPA last revised specifications for the test gasoline used in emissions testing of light- and heavy-duty vehicles. Perhaps most importantly, gasoline containing 10 percent ethanol by volume (E10) has replaced non-oxygenated gasoline (E0) across the country. As a result, we are updating federal emissions test fuel specifications to better match in-use fuel.

In the NPRM, EPA proposed that the specified gasoline for emissions testing be changed from E0 to E15 as a forward-looking approach. Since then, several factors have led EPA to reconsider that approach, including minimal proliferation on a national scale of stations offering E15 and the complexities that E15 would introduce for long-term harmonization with California's use of E10 in their LEVIII program. We received comments from a broad set of stakeholders including the auto and oil industries, states, and NGOs with a general consensus that E15 would not be appropriate as the official test fuel at this time. Ethanol industry commenters supported E15 certification fuel, but provided no timeline by which this blend level would be representative of in-use fuel. In light of the comments received and EPA's assessment of the current and projected levels of ethanol in gasoline in use, we are finalizing E10 as the new emissions test fuel.

In deciding to finalize E10 test fuel, EPA considered whether to change the volatility of the test fuel, typically expressed as pounds per square inch (psi) Reid Vapor Pressure (RVP). As discussed in detail in Section IV.F, after considering technical and policy implications as well as stakeholder comments, we have concluded that the most appropriate approach is to maintain an RVP of 9 psi for the E10 emissions test fuel at this time. EPA considered raising test fuel RVP to 10 psi, but decided to leave it unchanged at 9 psi based on what would have been the associated increase in stringency of the Tier 3 evaporative standard with 10 psi and the loss of regulatory harmony on evaporative emissions with California's LEV III program.

As a result, after reassessing market trends and considering comments, EPA concludes that the most appropriate approach is to finalize an ethanol content of 10 percent and an RVP of 9 psi for emissions test gasoline. We will continue to monitor ethanol trends in the gasoline market, as discussed later in this preamble.

c. Gasoline Sulfur Caps

As described in more detail in Section V.C. we proposed two options for the Tier 3 per-gallon sulfur caps—maintaining the Tier 2 refinery gate sulfur cap of 80 ppm (with a 95 ppm downstream sulfur cap), and lowering to a 50 ppm refinery gate sulfur cap beginning January 1, 2020 (with a 65 ppm downstream cap). We received comments supporting lower per-gallon caps which noted potential environmental benefits, greater certainty that vehicles would see lower and more uniform gasoline sulfur levels, and the ability to enable new vehicle technologies requiring very low sulfur levels. Conversely, comments received in support of maintaining the Tier 2 per-gallon caps cited concerns on cost, flexibility for turnarounds/unplanned shutdowns (due to refinery fires, natural disasters, etc.), and gasoline supply and/or price impacts.

Analysis performed since the time of the proposal found that a lower refinery gate cap would likely result in higher costs to the fuels industry and a decreased ability to handle off-spec product (potentially impacting gasoline supply and pricing), without any significant increase in the nationwide emissions reductions provided by the Tier 3 program. Thus, in today's action we are retaining the Tier 2 per-gallon sulfur caps. The 80 ppm refinery gate cap will provide refiners needed flexibility in allowing for naturally-occurring fuel batch variability, as well as more certainty that they will be able to continue producing and distributing gasoline during turnarounds/upsets to avoid a total shutdown. It will also provide more certainty for transmix processors, additive manufacturers, and other downstream parties in producing gasoline.

However, we do understand commenters' concerns that retaining the Tier 2 sulfur caps might create regional differences in the benefits of the Tier 3 program. Therefore we will continue to monitor in-use sulfur levels and their impact on vehicle emissions to ascertain whether a future reduction in the per-gallon cap may be necessary.

d. Effect of Gasoline Sulfur on Tier 3 Vehicle Emissions

The need for and level of gasoline sulfur control was a key issue raised in public comments. The petroleum industry raised concerns that there was insufficient basis for the proposed 10 ppm average sulfur level, while auto manufacturers and emissions control equipment manufacturers stressed that the feasibility of the Tier 3 vehicle standards was dependent on near-zero gasoline sulfur levels. This issue is discussed in detail below in Section IV.A.6. In sum, EPA believes that the range of studies conducted by EPA and others in recent years, along with the comments submitted by the auto industry and emissions control manufacturers during the comment period and more recently, strongly reinforce our conclusion that the impact of gasoline sulfur poisoning on exhaust catalyst performance is significant.

Sulfur is a well-known catalyst poison. The nature of sulfur's interactions with active catalytic materials is complex and varies with catalyst composition, exhaust gas composition, and exhaust temperature. Thus, even if a manufacturer were able to certify a new vehicle to the new stringent standards, the manufacturer's ability to maintain the emission performance of that vehicle in-use is greatly jeopardized if the vehicle is being operated on gasoline sulfur levels greater than 10 ppm. In fact, due to the variation in actual vehicle operation, any amount of gasoline sulfur will deteriorate catalyst efficiency. Vehicle manufacturers and suppliers, both individually and through their trade associations, stressed the need for gasoline sulfur to be reduced to near zero levels in order for them to meet the proposed standards. However, we believe that a 10 ppm average sulfur level is sufficiently low to enable compliance with the Tier 3 vehicle standards, and as described below and in Section V, reducing sulfur levels further would cause sulfur control costs to quickly escalate.

Taken together, this information provides a compelling argument that the fleetwide Tier 3 vehicle standards are achievable only with a reduction of gasoline sulfur content from the current 30 ppm average down to a 10 ppm average.

e. SFTP (US06) PM Standard for Light-Duty Vehicles

The final Tier 3 vehicle standards are largely unchanged from their proposed levels. One change from the proposal is the PM emissions standards as measured on the US06 test cycle. The US06 cycle is part of the composite Supplemental Federal Test Procedure (SFTP) and simulates aggressive driving. The US06 PM standards are part of the suite of Tier 3 tailpipe standards that limit emissions under a wide range of common vehicle driving conditions. Newer emissions test data presented in the NPRM, as well as more recent additional test data submitted in public comments, show that a numerically lower US06 PM standard is feasible and appropriately reflects the actual emissions performance achieved by many vehicles in the fleet today while preventing increased emissions in the future.

Taken together, the test results clearly show that most current light-duty vehicles—regardless of engine technology, emission control strategy, or vehicle size—are performing at much lower US06 emission levels than previously documented. Based on these newer data, we believe that it is appropriate to finalize a numerically lower US06 PM emission standard for LDVs, LDTs, and MDPVs, and to set a single standard for both lighter and heavier vehicles in this vehicle segment. In general, the final US06 PM standard for these vehicles begins to phase in at a level of 10 mg/mi in MYs 2017 and 2018, stepping down to a level of 6 mg/mi in MY2019. See Section IV.A.4.b for additional discussion of the US06 standards and how they will phase in.

2. Key Components of the Tier 3 Program
a. Tailpipe Standards for Light-Duty Vehicle, Light-Duty Truck, and Medium-Duty Passenger Vehicle Tailpipe Emissions

We are establishing a comprehensive program that includes new fleet-average standards for the sum of NMOG and NO X tailpipe emissions (presented as NMOG+NO X) as well as new per-vehicle standards for PM. (13) These standards, when applied in conjunction with reduced gasoline sulfur content, will result in very significant improvements in vehicle emissions from the levels of the Tier 2 program. For these pollutants, the standards are measured on test procedures that represent a range of vehicle operation, including the Federal Test Procedure (or FTP, simulating typical driving) and the Supplemental Federal Test Procedure (or SFTP, a composite test simulating higher ambient temperatures, higher vehicle speeds, and quicker accelerations). In addition to the standards, we are extending the regulatory useful life period during which the standards apply (see Section IV.A.7.b below) and making test fuel more representative of expected real-world fuel (see Section I.B.2.e below). The final standards are in most cases identical to those of California's LEVIII program, which provides the 50-state harmonization strongly supported by the auto industry.

As proposed, the new Tier 3 FTP and SFTP NMOG+NO X standards are fleet-average standards, meaning that a manufacturer calculates the average emissions of the vehicles it sells in each model year and compares that average to the applicable standard for that model year. The manufacturer certifies each of its vehicles to a per-vehicle “bin” standard (see Section IV.A.2) and sales-weights these values to calculate its fleet-average NMOG+NO X emissions for each model year. Table I-1 summarizes the fleet average standards for NMOG+NO X evaluated over the FTP. The standards for light-duty vehicles begin in MY 2017 at a level representing a 46 percent reduction from the Tier 2 requirements. For the light-duty fleet over 6000 lbs GVWR, and MDPVs, the standards apply beginning in MY 2018. As shown, these fleet-average standards decline during the first several years of the program, becoming increasingly stringent until ultimately reaching an 81 percent reduction when the transition is complete. The FTP NMOG+NO X program includes two separate sets of declining fleet-average standards, with LDVs and small light trucks in one grouping and heavier light trucks and MDPVs in a second grouping, that converge at 30 milligrams per mile (mg/mi) in MY 2025 and later. As mentioned above, we are also providing alternative percent phase-in schedules for this and the other light-duty standards.

Table I-1—Tier 3 LDV, LDT, and MDPV Fleet Average FTP NMOG+NO X Standards
Model year2017 a 20182019202020212022202320242025 and later
LDV/LDT1 b 867972655851443730
LDT2,3,4 and MDPV1019283746556473830

Similarly, as proposed, the NMOG+NO X standards measured over the SFTP are fleet-average standards, declining from MY 2017 until MY 2025, as shown in Table I-2. In this case, the same standards apply to both lighter and heavier vehicles in the light-duty fleet. In MY 2025, the SFTP NMOG+NO X standard reaches its final fleet average level of 50 mg/mi.

Table I-2—Tier 3 LDV, LDT, and MDPV Fleet Average SFTP NMOG+NO X Standards
Model year2017 a 20182019202020212022202320242025 and later
NMOG + NO X 1039790837770635750

As proposed, manufacturers can also earn credits if their fleet average NMOG+NO X performance is better than the applicable standard in any model year. Credits that have been previously banked or obtained from other manufacturers can be used, or credits can be traded to other manufacturers. Manufacturers would also be allowed to carry forward deficits in their credit balance. (See Sections IV.A.7.a and IV.A.7.m).

We are also establishing PM standards as part of the Tier 3 program, for both the FTP and US06 cycles (as described above, US06 is a component of the SFTP test). Research has demonstrated that the level of PM from gasoline light-duty vehicles is more significant than previously thought. (14) Although many vehicles today are performing at or near the levels of the new standards, the data indicate that improvements, especially in high-load fuel control and in the durability of engine components, are possible.

Under typical driving, as simulated by the FTP, the PM emissions of most current-technology gasoline vehicles are fairly low at certification and in use, well below the Tier 2 PM standards. At the same time we see considerable variation in PM emissions among vehicles of various makes, models, and designs. As a result, as proposed, we are setting the new FTP PM standard at a level that will ensure that all new vehicles perform at the level already being achieved by well-designed Tier 2 vehicles. The PM standards apply to each vehicle separately (i.e., not as a fleet average). Also, in contrast to the declining NMOG+NO X standards, the PM standard on the FTP for certification testing is 3 mg/mi for all vehicles and for all model years. As for the NMOG+NO X standards, for vehicles over 6000 lbs GVWR, the FTP PM standard applies beginning in MY 2018. Manufacturers can phase in their vehicle models as a percent of U.S. sales through MY 2022. Most vehicles are already performing at this stringent PM level, and the primary intent of the standard is to bring all light-duty vehicles to the typical level of PM performance being demonstrated by many of today's vehicles.

As proposed, the Tier 3 program also includes a temporary in-use FTP PM standard of 6 mg/mi for the testing of in-use vehicles that applies during the percent phase-in period only. This in-use standard will address the in-use variability and durability uncertainties that accompany the introduction of new technologies. Table I-3 presents the FTP certification and in-use PM standards and the phase-in percentages.

Table I-3—Phase-In for Tier 3 FTP PM Standards
2017 a 20182019202020212022 and later
Phase-In (percent of U.S. sales) b 20204070100100
Certification Standard (mg/mi)333333
In-Use Standard (mg/mi)666663

Finally, as discussed in Section I.B.1.e above, the Tier 3 program includes PM standards evaluated over the US06 driving cycle (the US06 is one part of the SFTP procedure) of 10 mg/mi through MY 2018 and of 6 mg/mi for 2019 and later model years, for light-duty vehicles. As in the case of the FTP PM standards, the intent of the US06 PM standard is to bring the emission performance of all vehicles to that already being demonstrated by many vehicles in the current light-duty fleet.

b. Heavy-Duty Vehicle Tailpipe Emissions Standards

As discussed in detail in Section IV.B, we are setting Tier 3 exhaust emissions standards for complete heavy-duty vehicles (HDVs) between 8,501 and 14,000 lbs GVWR. Vehicles in this GVWR range are often referred to as Class 2b (8,501-10,000 lbs) and Class 3 (10,001-14,000 lbs) vehicles, and are typically heavy-duty pickup trucks and work or shuttle vans. Most are built by companies with even larger light-duty truck markets, and as such they frequently share major design characteristics and emissions control technologies with their LDT counterparts. However, in contrast to the largely gasoline-fueled LDT fleet, roughly half of the heavy-duty pickup and van fleet in the U.S. is diesel-fueled. This is an important consideration in setting emissions standards, as diesel engine emissions control strategies differ from those of gasoline engines.

As proposed, the key elements of the Tier 3 program for HDVs parallel those being adopted for passenger cars and LDTs, with adjustments in standard levels, emission test requirements, and implementation schedules appropriate to this sector. These key elements include combined NMOG+NO X declining fleet average standards, a phase-in of PM standards, adoption of a new emissions test fuel for gasoline-fueled vehicles, extension of the regulatory useful life to 150,000 miles or 15 years (whichever occurs first), and a first-ever requirement for HDVs to meet standards over an SFTP drive cycle that addresses real-world driving modes not well-represented by the FTP cycles.

We are adopting the Class 2b and Class 3 fleet average NMOG+NO X standards shown in Table I-4, as proposed. The standards become more stringent in successive model years from 2018 to 2022, with voluntary standards made available in 2016 and 2017, all of which are set at levels that match those of California's LEV III program for these classes of vehicles. Each covered HDV sold by a manufacturer in each model year contributes to this fleet average based on the mg/mi NMOG+NO X standard level of the “bin” declared for it by the manufacturer, who chooses from a set of seven discrete Tier 3 bins specified in the regulations. These bin standards then become the compliance standards for the vehicle over its useful life, with some adjustment provided for in-use testing in the early model years of the program.

As proposed, manufacturers can also earn credits for fleet average NMOG+NO X levels below the standard in any model year. Tier 3 credits that were previously banked, obtained from other manufacturers, or transferred across the Class 2b/Class 3 categories can be used to help demonstrate compliance. Unused credits expire after 5 model years. Manufacturers will also be allowed to carry forward deficits in their credit balance for up to 3 model years.

Table I-4—Tier 3 HDV Fleet Average FTP NMOG+NO X Standards
VoluntaryRequired program     
Model Year2016201720182019202020212022 and later.
Class 2b333310278253228203178.
Class 3548508451400349298247.

We are adopting the proposed FTP PM standards of 8 mg/mi and 10 mg/mi for Class 2b and Class 3 HDVs, respectively, phasing in as an increasing percentage of a manufacturer's sales per year. We are adopting the same phase-in schedule as for the light-duty sector during model years 2018-2019-2020-2021: 20-40-70-100 percent, respectively, and a more flexible but equivalent alternative PM phase-in is also being adopted. Tier 3 HDVs will also be subject to CO and formaldehyde exhaust emissions standards that are more stringent than the existing standards.

Finally, we are setting first-ever nationwide SFTP standards for HDVs to ensure a robust overall control program that precludes high off-FTP cycle emissions by having vehicle designers consider them in their choice of compliance strategies. As for light-duty vehicles, we are requiring that SFTP compliance be based on a weighted composite of measured emissions from testing over the FTP cycle, the SC03 cycle, and an aggressive driving cycle, with the latter tailored to various HDV sub-categories: the US06 cycle for most HDVs, the highway portion of the US06 cycle for low power-to-weight Class 2b HDVs, and the LA-92 (or “Unified”) cycle for Class 3 HDVs. The SFTP standards are the same as those adopted for California LEV III vehicles, and apply to NMOG+NO X, PM, and CO emissions.

The HDV program outlined above and described in detail in Section IV.B is substantially what we proposed. Commenters generally supported the scope, stringency, and implementation phase-in of this program. However, some industry commenters requested changes to some specific provisions of the proposal, and the program we are adopting reflects improvements we have made in response. These are: (1) A limited allowance for engine certification of Class 3 complete diesel vehicles to avoid a potential need for dual chassis- and engine-based certification and to better harmonize with LEV III, (2) relaxed interim in-use testing standards to facilitate a smooth transition to the Tier 3 standards and to better harmonize with LEV III, (3) adoption of combined NMOG+NO X standards for the two highest (interim) bins, with a restriction placed on NO X levels in certification testing, to enhance the utility of these bins and to better harmonize with LEV III, and (4) a provision in the percent-of-sales phase-in alternative to allow manufacturers to exclude vehicle models that begin their 2019 model year production early in 2018, in order to provide four years of lead time. Commenters also requested relaxed standards for testing at high altitudes and changes to the credits program structure for generation of early credits and use of LEV III-based “vehicle emission credits”, but we did not adopt these for reasons explained in Section IV.B.

Overall, we expect the Tier 3 program we are adopting for HDVs to result in substantial reductions in harmful emissions from this large fleet of work trucks and vans. The fully-phased in Tier 3 standards levels for NMOG+NO X and PM are on the order of 60 percent lower than the current standards that took full effect in the 2009 model year.

c. Evaporative Emission Standards

Gasoline vapor emissions from vehicle fuel systems occur when a vehicle is in operation, when it is parked, and when it is being refueled. These evaporative emissions, which occur on a daily basis from gasoline-powered vehicles, are primarily functions of temperature, fuel vapor pressure, and activity. EPA first instituted evaporative emission standards in the early 1970s to address emissions when vehicles are parked after being driven. These are commonly referred to as hot soak plus diurnal emissions. Over the subsequent years the test procedures have been modified and improved and the standards have become more numerically stringent. We have addressed emissions which arose from new fuel system designs by putting in place new requirements such as running loss emission standards and test procedure provisions to address permeation emissions. Subsequently standards were put in place to control refueling emissions from all classes of gasoline-powered motor vehicles up to 10,000 lbs GVWR. Evaporative and refueling emission control systems have been in place for most of these vehicles for many years. These controls have led to significant reductions, but evaporative and refueling emissions still constitute 30-40 percent of the summer on-highway mobile source hydrocarbon inventory. These fuel vapor emissions are ozone and PM precursors, and also contain air toxics such as benzene.

To control evaporative emissions, EPA is establishing more stringent standards that will require covered vehicles to have essentially zero fuel vapor emissions in use. These include more stringent evaporative emissions standards, new test procedures, and a new fuel/evaporative system leak emission standard. The program also includes refueling emission standards for all complete heavy-duty gasoline vehicles (HDGVs) over 10,000 lbs GVWR. EPA is including phase-in flexibilities as well as credit and allowance programs. The standards, harmonized with California's “zero evap” standards, are designed to allow for a use of common technology in vehicle models sold throughout the U.S. The level of the standard remains above zero to account for nonfuel background emissions from the vehicle hardware.

Requirements to meet the Tier 3 evaporative emission regulations phase in over a six model year period. We are finalizing three options for the 2017 model year, but after that the sales percentage requirements are 60 percent for MYs 2018 and 2019, 80 percent for model years 2020 and 2021, and 100 percent for model years 2022 and later. In Table I-5 we present the Tier 3 evaporative hot soak plus diurnal emission standards by vehicle class. The standards are approximately a 50 percent reduction from the existing standards. To enhance flexibility and reduce costs, EPA is finalizing provisions that allow manufacturers to generate allowances through early certifications (basically before the 2017 model year) and to demonstrate compliance using averaging concepts. Manufacturers may comply on average within each of the four vehicle categories, but not across these categories. EPA is not making any changes to the existing light-duty running loss or refueling emission standards, with the exception of the certification test fuel requirement discussed in Section I.B.2 below.

Table I-5—Tier 3 Evaporative Emission Standards
Vehicle classHighest hot soak + diurnal level (over both 2-day and 3-day diurnal tests)
LDV, LDT10.300
LDT20.400
LDT3, LDT4, MDPV0.500
HDGVs0.600

Flexible Fuel Vehicles (FFVs) must meet the same evaporative emission standards as non-FFVs using Tier 3 emissions certification test fuel. However, FFVs must meet the refueling emission standards using 10 psi RVP fuel to account for emissions resulting from commingling with non-E85 blends that may be in the vehicle's fuel tank.

EPA is establishing the canister bleed emission test procedure and emission standard to help ensure fuel vapor emissions are eliminated. Under this provision, manufacturers are required to measure diurnal emissions over the 2-day diurnal test procedure from just the fuel tank and the evaporative emission canister and comply with a 0.020 gram per test (g/test) standard for all LDVs, LDTs, and MDPVs, without averaging. The corresponding canister bleed test standard for HDGVs is 0.030 g/test. The Tier 3 evaporative emission standards will be phased in over a period of six model years between MY 2017 and MY 2022, with the leak test phasing in beginning in 2018.

Data from in-use evaporative emissions testing indicates that vapor leaks from vehicle fuel/evaporative systems are found in the fleet and that even very small leaks have the potential to make significant contributions to the mobile source VOC inventory. To help address this issue, we are also adding a new standard and test procedure to control vapor leaks from vehicle fuel and vapor control systems. The standard will prohibit leaks with a cumulative equivalent diameter of 0.02 inches or greater. We are adding this simple and inexpensive test and emission standard to help ensure vehicles maintain zero fuel vapor emissions over their full useful life. New LDV, LDT, MDPV, and HDGV equal to or less than 14.000 lbs GVWR meeting the Tier 3 evaporative emission regulations are also required to meet the leak standard beginning in the 2018 model year. Manufacturers must comply with the leak standard phase-in on the same percentage of sales schedule as that for the Tier 3 evaporative emission standards. Manufacturers will comply with the leak emission standard during certification and in use. The leak emission standard does not apply to HDGVs above 14,000 lbs GVWR.

EPA is also establishing new refueling emission control requirements for all complete HDGVs equal to or less than 14,000 lbs GVWR (i.e., Class 2b/3 HDGVs), starting in the 2018 model year, and for all larger complete HDGVs by the 2022 model year. The existing refueling emission control requirements apply to complete Class 2b HDGVs, and EPA is extending those requirements to other complete HDGVs, since the fuel and evaporative control systems on these vehicles are very similar to those on their lighter-weight Class 2b counterparts.

d. Onboard Diagnostic Systems (OBD)

EPA and CARB both have OBD regulations applicable to the vehicle classes covered by the Tier 3 emission standards. In the past the requirements have been very similar, so most manufacturers have met CARB OBD requirements and, as permitted in our regulations, EPA has generally accepted compliance with CARB's OBD requirements as satisfying EPA's OBD requirements. Over the past several years CARB has upgraded its requirements to help improve the effectiveness of OBD in ensuring good in-use exhaust and evaporative system emissions performance. We have reviewed these provisions and agree with CARB that these revisions will help to improve in-use emissions performance, while at the same time harmonizing with the CARB program. Toward that end, we are adopting and incorporating by reference the current CARB OBD regulations, effective for the 2017 MY, with a few minor differences including phase-in flexibility provisions and specific additions to enhance the implementation of the leak standard. EPA is retaining the provision that certifying with CARB's program would permit manufacturers to seek a separate EPA certificate on that basis.

e. Emissions Test Fuel

As described above, after reassessing market trends and considering comments, EPA is finalizing E10 as the ethanol blend level in emissions test gasoline for Tier 3 light-duty and heavy-duty gasoline vehicles. We will continue to monitor the in-use gasoline supply and based on such review may initiate rulemaking action to revise the specifications for emissions test fuel to include a higher ethanol blend level. EPA is also making additional changes that are consistent with CARB's LEV III emissions test fuel specifications, including new specifications for octane, distillation temperatures, aromatics, olefins, sulfur and benzene. (See Section IV.F below for a detailed discussion of all the revised emission test fuel parameters.)

As discussed in Sections IV.A.7.d (tailpipe emission testing) and IV.C.5.b (evaporative emission testing), we are requiring certification of all Tier 3 light-duty and chassis-certified heavy-duty gasoline vehicles on federal E10 test fuel. The new test fuel specifications will apply to new vehicle certification, assembly line, and in-use testing.

With a change in the ethanol content of the test fuel, EPA also needed to consider whether a change is warranted in the volatility of the test fuel, typically expressed as pounds per square inch (psi) Reid Vapor Pressure (RVP). As discussed in detail in Section IV.F below, after considering several technical and policy implications as well as stakeholder comments, EPA has concluded that the most appropriate approach is to maintain an RVP of 9 psi for the E10 certification fuel at this time.

In addition to finalizing a new E10 emissions test fuel, we are also finalizing detailed specifications for the E85 emissions test fuel used for flexible fuel vehicle (FFV) certification, as discussed in Section IV.F.3. (15) This will resolve uncertainty and confusion in the certification of FFVs designed to operate on ethanol levels up to 83 percent. Furthermore, we allow vehicle manufacturers to request approval for an alternative certification fuel such as a high-octane 30 percent ethanol by volume blend (E30) for vehicles that may be optimized for such fuel.

f. Fuel Standards

Under the Tier 3 fuel program, gasoline must contain no more than 10 ppm sulfur on an annual average basis beginning January 1, 2017. Similar to the Tier 2 gasoline program, the Tier 3 program will apply to gasoline in the U.S. and the U.S. territories of Puerto Rico and the Virgin Islands, excluding California. The program will result in gasoline that contains, on average, two-thirds less sulfur than it does today. In addition, following discussions with numerous refiners and other segments of the fuel market (e.g., pipelines, terminals, marketers, ethanol industry representatives, transmix processors, additive manufacturers, etc.), the Tier 3 fuel program contains considerable flexibility to ease both initial and long-term implementation of the program. The program that we are finalizing today includes an averaging, banking, and trading (ABT) program that allows refiners and importers to spread out their investments over nearly a 6-year period through the use of an early credit program and then rely on ongoing nationwide averaging to meet the 10 ppm sulfur standard. In addition there is a three-year delay for small refiners and “small volume refineries”. As a result of the early credit program, we anticipate considerable reductions in gasoline sulfur levels prior to 2017, with a complete transition to the 10 ppm average occurring by January 1, 2020. For more information on the gasoline sulfur program flexibilities, refer to Section V.E.

Under today's Tier 3 gasoline sulfur program, we are maintaining the current 80 ppm refinery gate and 95 ppm downstream per-gallon caps. We also evaluated and sought comment on the potential of lowering the per-gallon caps. While there are advantages and disadvantages with each of the sulfur cap options that we proposed, we believe that retaining the current Tier 2 sulfur caps is prudent at this time, as explained in more detail in Section V.C. Further, the stringency of the 10 ppm annual average standard will result in reduced gasoline sulfur levels nationwide. Today's program requires that manufacturers of gasoline additives that are used downstream of the refinery at less than 1 volume percent must limit the sulfur contribution to the finished gasoline from the use of their additive to less than 3 ppm when the additive is used at the maximum recommended treatment rate (see Section V.C.2). This requirement will preclude the unnecessary use of high sulfur content additives in gasoline.

The vehicle emissions standards finalized today are fuel-neutral (i.e., they are applicable regardless of the type of fuel that the vehicle is designed to use). There currently are no sulfur standards for the fuel used in compressed natural gas (CNG) and liquid propane gas (LPG) vehicles. We requested comment on whether it is necessary for EPA to establish sulfur standards for CNG and LPG to enable them meeting more stringent vehicle emissions standards. EPA is deferring finalizing in-use sulfur requirements for CNG/LPG in this final rule to provide additional time to work with stakeholders to collect data on current CNG/LPG sulfur content, to determine whether additional control of in-use CNG/LPG sulfur content is needed, and to evaluate the feasibility and costs associated with potential additional sulfur controls (see Section V.J). Given that the information provided suggests that CNG/LPG sulfur levels tend to be low already, the vehicle emissions standards finalized today will apply to CNG/LPG vehicles in addition to vehicles fueled on gasoline, diesel fuel, or any other fuel. The sulfur content of highway diesel fuel is already required to meet a 15 ppm sulfur cap, which is sufficient for diesel fuel vehicles to meet the Tier 3 emissions standards.

As the number of flex-fuel vehicles (FFVs) in the in-use fleet increases, it is becoming increasingly important that all fuels used in FFVs, not just gasoline, meet fuel quality standards. A lack of clarity regarding the standards that apply to fuels used in FFVs could also act to impede the further expansion of ethanol blended fuels with concentrations greater than 15 volume percent, which is important to satisfying the requirements of the RFS2 program. Hence, we sought comment on appropriate regulatory mechanisms to implement in-use quality standards for E51-83 and E16-50 in the Tier 3 proposal. Additional work is needed on some issues that could not be accommodated within the timeline for this Tier 3 final rule. Therefore, we are choosing not to finalize these provisions at this time. We intend to finalize in-use fuel quality standards for E51-83 and perhaps E16-50 as well in a follow-up final rule.

g. Regulatory Streamlining and Technical Amendments

This action also includes a number of items to help streamline the in-use fuels regulations at 40 CFR parts 79 and 80. The majority of these items involve clarifying vague or inconsistent language, removal or updating of outdated provisions, and decreasing in frequency and/or volume of reporting burden where data are no longer needed or are redundant with other EPA fuels programs. In general, we believe that these changes will reduce the burden on industry and allow the standards and resulting environmental benefits to be achieved as early as possible with no expected loss in environmental control. In some cases, these regulatory streamlining items are non-substantive amendments that correct minor errors or inconsistencies in the regulations.

The regulatory streamlining items that we are finalizing for the in-use fuels regulations are changes that we believe are straightforward and should be made quickly.

This action also includes a variety of technical amendments to certification-related requirements for engine and vehicle emission standards; adjusting the fuel economy label provisions to correspond to the new Tier 3 standards, removing obsolete regulatory text, and making several minor corrections and clarifications.

Please refer to Section VI for a complete discussion of technical amendments and regulatory streamlining provisions and issues.

C. What will the impacts of the standards be?

The final Tier 3 vehicle and fuel standards together will reduce dramatically emissions of NO X, VOC, PM 2.5, and air toxics. The gasoline sulfur standards, which will take effect in 2017, will provide large immediate reductions in emissions from existing gasoline vehicles and engines. NO X emissions are projected to be reduced by about 260,000 tons, or about 10 percent of emissions from on-highway vehicles, in 2018, and these emission reductions will increase over time as newer vehicles become a larger percentage of the fleet. In 2030, when 70 percent of the miles travelled are projected to be from vehicles that meet the fully phased-in Tier 3 standards, we expect the NO X and VOC emissions to be reduced by about 330,000 tons and 170,000 tons, respectively, or 25 percent and 16 percent of emissions from on-highway vehicles compared to their 2030 levels without the Tier 3 program. Emissions of CO are projected to decrease by almost 3.5 million tons, or 24 percent of emissions from on-highway vehicles. Emissions of many air toxics will also be reduced, including benzene, 1,3-butadiene, acetaldehyde, formaldehyde, acrolein and ethanol, with reductions projected to range from 10 to nearly 30 percent of national emissions from on-highway vehicles. We expect these reductions to continue beyond 2030 as more of the fleet continues to turn over to Tier 3 vehicles; for example, by 2050, when nearly all of the fleet will have turned over to vehicles meeting the fully phased-in Tier 3 standards, we estimate the Tier 3 program will reduce on-highway emissions of NO X and VOC nearly 31 percent from the level of emissions projected without Tier 3 controls. (16)

These reductions in emissions of NO X, VOC, PM 2.5 and air toxics from the Tier 3 standards are projected to lead to significant decreases in ambient concentrations of ozone, PM 2.5 and air toxics (including notable nationwide reductions in benzene concentrations) by 2030, and will immediately reduce ozone in 2017 when the sulfur controls take effect. Additional information on the emission and air quality impacts of the final Tier 3 program is presented in Sections III.B and C.

Exposure to ambient concentrations of ozone, PM 2.5, and air toxics is linked to adverse human health impacts such as premature deaths as well as other important public health and environmental effects (see Section II.B). The final Tier 3 standards are expected to reduce these adverse impacts and yield significant benefits, including those we can monetize and those we are unable to quantify. We estimate that by 2030, the emission reductions of the Tier 3 standards will annually prevent between 660 and 1,500 PM-related premature deaths, between 110 and 500 ozone-related premature deaths, 81,000 work days lost, 210,000 school absence days, and approximately 1.1 million minor restricted-activity days. The estimated annual monetized health benefits of the Tier 3 standards in 2030 (2011$) is between $7.4 and $19 billion, assuming a 3-percent discount rate (or between $6.7 billion and $18 billion assuming a 7-percent discount rate). We project the final fuel standards to cost on average 0.65 cent (i.e., less than a penny) per gallon of gasoline, and the final vehicle standards to have an average cost that increases in proportion to the increase in stringency during the phase-in period, from $28 per vehicle in 2017 to $72 per vehicle in 2025, when the standards are fully phased in. We estimate the annual cost of the overall program in 2030 will be approximately $1.5 billion, and the 2030 benefits will be between 4.5 and 13 times the costs of the program.

The estimated benefits in Table I-6 include all of the human health impacts we are able to quantify and monetize at this time. However, the full complement of human health and welfare effects associated with PM, ozone and air toxics remain unquantified because of current limitations in methods and/or available data. As a result, the health benefits quantified in this section are likely underestimates of the total benefits attributable to the final standards. See Sections VII and VIII for detailed descriptions of the costs and benefits of this action.

Table I-6—Summary of Estimated Annual Benefits and Costs Associated With the Final Tier 3 Program
Description2030
Vehicle Program Costs$0.76
Fuels Program Costs $0.70
Total Estimated Costs b $1.5
Total Estimated Health Benefits: cdef  
3 percent discount rate $7.4-$19
7 percent discount rate$6.7-$18
Annual Net Benefits (Total Benefits−Total Costs): 
3 percent discount rate $5.9-$18
7 percent discount rate$5.2-$17

II. Why is EPA taking this action?

The Clean Air Act authorizes EPA to establish emissions standards for motor vehicles to address air pollution that may reasonably be anticipated to endanger public health or welfare. EPA also has authority to establish fuel controls to address such air pollution. These statutory requirements are described in Section II.A.

Emissions from motor vehicles and their fuels contribute to ambient levels of ozone, PM, NO 2, sulfur dioxide (SO 2) and CO, which are all pollutants for which EPA has established health-based NAAQS. These pollutants are linked with respiratory and/or cardiovascular problems and other adverse health impacts leading to increased medication use, hospital admissions, emergency department visits, and premature mortality. Over 149 million people currently live in areas designated nonattainment for one or more of the current NAAQS for ozone, PM 2.5, PM 10, and SO 2. (17)

Motor vehicles also emit air toxics, and the most recent available data indicate that the majority of Americans continue to be exposed to ambient concentrations of air toxics at levels which have the potential to cause adverse health effects, including cancer, immune system damage, and neurological, reproductive, developmental, respiratory, and other health problems. (18) A more detailed discussion of the health and environmental effects of these pollutants is included in Section II.B.

Cars and light trucks also continue to be a significant contributor to air pollution directly near roads, with gasoline vehicles accounting for more than 50 percent of near-road concentrations of some criteria and toxic pollutants. (19) More than 50 million people live, work, or go to school in close proximity to high-traffic roadways, and the average American spends more than one hour traveling each day, with over 80 percent of daily trips occurring by personal vehicle. 20212223 (24) Exposure to traffic-related pollutants has been linked with adverse health impacts such as respiratory problems (particularly in asthmatic children) and cardiovascular problems.

In the absence of additional controls such as Tier 3 standards, many areas will continue to have ambient ozone and PM 2.5 concentrations exceeding the NAAQS in the future. States and local areas are required to adopt control measures to attain the NAAQS and, once attained, to demonstrate that control measures are in place sufficient to maintain the NAAQS for ten years (and eight years later, a similar demonstration is required for another ten-year period). The Tier 3 standards will be a critical part of many areas' strategies to attain and maintain the NAAQS. Maintaining the NAAQS has been challenging for some areas in the past, particularly those where high population growth rates lead to significant annual increases in vehicle trips and vehicle miles traveled. Our air quality modeling for this final rule, which is described in more detail in Section III.C, projects that in 2018 a significant number of counties outside CA will be within 10 percent of the 2008 ozone NAAQS, in the absence of additional controls. These counties in particular will benefit from the Tier 3 standards as they work to ensure long-term maintenance of the NAAQS.

Section III provides more detail on how we expect this action will reduce motor vehicle emissions and ambient levels of pollution. We project that the Tier 3 program will meaningfully reduce ozone concentrations as early as 2017 (the first year of the program), and even more significantly in 2030. The estimated reductions are of significant enough magnitude to bring ozone levels in some counties from above the standard to below the standard, even without any additional controls. We also project that the Tier 3 standards will reduce ambient PM 2.5 concentrations.

Without this action to reduce nationwide motor vehicle emissions, areas would have to adopt other measures to reduce emissions from other sources under their state or local authority. Few other measures exist for providing multi-pollutant reductions of the same magnitude and cost-effectiveness as those expected from the Tier 3 standards. Furthermore, most states do not have the authority to lower the sulfur in gasoline, which is needed to immediately reduce emissions from the existing fleet and also enable new vehicles to meet the Tier 3 emissions standards throughout their useful life.

The projected reductions in ambient ozone and PM 2.5 that will result from the Tier 3 standards will provide significant health benefits. We estimate that by 2030, the standards will annually prevent between 660 and 1,500 PM-related premature deaths, between 110 and 500 ozone-related premature deaths, 81,000 work days lost, 210,000 school absence days, and approximately 1.1 million minor restricted-activity days (see Section VIII for more details). This action will also reduce air toxics; for example, we project that in 2030, the Tier 3 standards will decrease ambient benzene concentrations by 10-25 percent in some urban areas. Furthermore, the Tier 3 standards will reduce traffic-associated pollution near major roads.

EPA is finalizing Tier 3 vehicle and fuel standards as part of a comprehensive nationwide program for regulating all types of air pollution from motor vehicles. EPA recently finalized standards to reduce GHG emissions from light-duty vehicles, starting with model year 2017. (25) The Tier 3 standards in this final rule, which address non-GHGs, will be implemented on the same timeframe, thus allowing manufacturers to optimize their vehicle redesigns over both sets of standards. Furthermore, the Tier 3 vehicle and fuel standards are also closely aligned with California's LEV III program, in such a way that manufacturers will be able to design a single vehicle for nationwide sales. This reduces the cost of compliance for auto manufacturers.

This Tier 3 rulemaking responds to the President's request in his May 2010 memorandum for EPA to review the adequacy of its existing non-GHG standards for new motor vehicles and fuels, and to promulgate new standards, if necessary, as part of a comprehensive approach to regulating motor vehicles. (26) Based on our review, we have concluded that improved vehicle technology, combined with lower sulfur gasoline, make it feasible and cost-effective to reduce emissions well below the current Tier 2 levels. These emission reductions are necessary to reduce air pollution that is (and projected to continue to be) at levels that endanger public health and welfare.

A. Basis for Action Under the Clean Air Act

1. Clean Air Act Section 202

We are setting motor vehicle emission standards under the authority of section 202 of the Clean Air Act. Section 202(a) provides EPA with general authority to prescribe vehicle standards, subject to any specific limitations elsewhere in the Act. EPA is setting standards for larger light-duty trucks and MDPVs under the general authority of section 202(a)(1) and under section 202(a)(3), which requires that standards applicable to emissions of hydrocarbons, NO X, CO and PM from heavy-duty vehicles  (27) reflect the greatest degree of emission reduction available for the model year to which such standards apply, giving appropriate consideration to cost, energy, and safety. In addition, section 202(k) provides EPA with authority to issue and revise regulations applicable to evaporative emissions of hydrocarbons from all gasoline-fueled motor vehicles during: (1) Operation, and (2) over 2 or more days of nonuse; under ozone-prone summertime conditions. Regulations under section 202(k) shall take effect as expeditiously as possible and shall require the greatest degree of emission reduction achievable by means reasonably expected to be available for production during any model year to which the regulations apply, giving appropriate consideration to fuel volatility, and to cost, energy, and safety factors associated with the application of the appropriate technology. Further, section 206 and in particular section 206(d) of the Clean Air Act authorizes EPA to establish methods and procedures for testing whether a motor vehicle or motor vehicle engine conforms with section 202 requirements.

2. Clean Air Act Section 211

We are adopting gasoline sulfur controls pursuant to our authority under section 211(c)(1) of the CAA. This section allows EPA to establish a fuel control if at least one of the following two criteria is met: (1) The emission products of the fuel cause or contribute to air pollution which may reasonably be anticipated to endanger public health or welfare; or (2) the emission products of the fuel will impair to a significant degree the performance of any emissions control device or system which is either 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 the fuel control to be adopted. We are finalizing gasoline sulfur controls based on both of these criteria. Under the first criterion, we believe that gasoline with current levels of sulfur contributes to ambient levels of air pollution that endanger public health and welfare, as described in Section II.B. Under the second criterion, we believe that gasoline sulfur impairs the emissions control systems of vehicles, as discussed in Section III.A.2.

B. Overview of Public Health Impacts of Motor Vehicles and Fuels

Motor vehicles emit pollutants that contribute to ambient concentrations of ozone, PM, NO 2, SO 2, CO, and air toxics. Motor vehicles are significant contributors to emissions of VOC and NO X, which contribute to the formation of both ozone and PM 2.5. Over 149 million people currently live in counties designated nonattainment for one or more of the NAAQS, and this figure does not include the people living in areas with a risk of exceeding the NAAQS in the future. (28) The majority of Americans continue to be exposed to ambient concentrations of air toxics at levels which have the potential to cause adverse health effects. (29) In addition, populations who live, work, or attend school near major roads experience elevated exposure concentrations to a wide range of air pollutants. (30)

EPA has already adopted many emission control programs that are expected to reduce ambient pollution concentrations. As a result of these programs, the number of areas that continue to violate the ozone and PM 2.5 NAAQS or have high levels of air toxics is expected to continue to decrease. However, the baseline air quality modeling completed for this rule predicts that without additional controls there will continue to be a need for reductions in ozone, PM 2.5 and air toxics concentrations in some locations in the future. Section III.C of this preamble presents the air quality modeling results for this action.

1. Ozone
a. Background

Ground-level ozone pollution is typically formed through reactions involving VOC and NO X in the lower atmosphere in the presence of sunlight. These pollutants, often referred to as ozone precursors, are emitted by many types of pollution sources, such as highway and nonroad motor vehicles and engines, power plants, chemical plants, refineries, makers of consumer and commercial products, industrial facilities, and smaller area sources.

The science of ozone formation, transport, and accumulation is complex. Ground-level ozone is produced and destroyed in a cyclical set of chemical reactions, many of which are sensitive to temperature and sunlight. When ambient temperatures and sunlight levels remain high for several days and the air is relatively stagnant, ozone and its precursors can build up and result in more ozone than typically occurs on a single high-temperature day. Ozone and its precursors can be transported hundreds of miles downwind from precursor emissions, resulting in elevated ozone levels even in areas with low local VOC or NO X emissions.

b. Health Effects of Ozone

This section provides a summary of the health effects associated with exposure to ambient concentrations of ozone. (31) The information in this section is based on the information and conclusions in the February 2013 Integrated Science Assessment for Ozone (Ozone ISA) prepared by EPA's Office of Research and Development (ORD). (32) The Ozone ISA concludes that human exposures to ambient concentrations of ozone are associated with a number of adverse health effects and characterizes the weight of evidence for these health effects. (33) The discussion below highlights the Ozone ISA's conclusions pertaining to health effects associated with both short-term and long-term periods of exposure to ozone.

For short-term exposure to ozone, the Ozone ISA concludes that respiratory effects, including lung function decrements, pulmonary inflammation, exacerbation of asthma, respiratory-related hospital admissions, and mortality, are causally associated with ozone exposure. It also concludes that cardiovascular effects, including decreased cardiac function and increased vascular disease, and total mortality are likely to be causally associated with short-term exposure to ozone and that evidence is suggestive of a causal relationship between central nervous system effects and short-term exposure to ozone.

For long-term exposure to ozone, the Ozone ISA concludes that respiratory effects, including new onset asthma, pulmonary inflammation and injury, are likely to be a causally related with ozone exposure. The Ozone ISA characterizes the evidence as suggestive of a causal relationship for associations between long-term ozone exposure and cardiovascular effects, reproductive and developmental effects, central nervous system effects and total mortality. The evidence is inadequate to infer a causal relationship between chronic ozone exposure and increased risk of lung cancer.

Finally, interindividual variation in human responses to ozone exposure can result in some groups being at increased risk for detrimental effects in response to exposure. The Ozone ISA identified several groups that are at increased risk for ozone-related health effects. These groups are people with asthma, children and older adults, individuals with reduced intake of certain nutrients (i.e., Vitamins C and E), outdoor workers, and individuals having certain genetic variants related to oxidative metabolism or inflammation. Ozone exposure during childhood can have lasting effects through adulthood. Such effects include altered function of the respiratory and immune systems. Children absorb higher doses (normalized to lung surface area) of ambient ozone, compared to adults, due to their increased time spent outdoors, higher ventilation rates relative to body size, and a tendency to breathe a greater fraction of air through the mouth. Children also have a higher asthma prevalence compared to adults. Additional children's vulnerability and susceptibility factors are listed in Section XII.G.

c. Current and Projected Concentrations of Ozone

Concentrations that exceed the level of the ozone NAAQS occur in many parts of the country, including major population centers such as Atlanta, Baltimore, Chicago, Dallas, Houston, New York, Philadelphia, and Washington, DC. In addition, our modeling without the Tier 3 controls projects that in the future we will continue to have many counties that will have ambient ozone concentrations above the level of the NAAQS (see Section III.C.1). States will need to meet the standard in the 2015-2032 time frame for the 2008 ozone NAAQS. The emission reductions and significant ambient ozone improvements from this rule, which will take effect starting in 2017, will be helpful to states as they work to attain and maintain the ozone NAAQS.

The primary and secondary NAAQS for ozone are 8-hour standards with a level of 0.075 ppm. The most recent revision to the ozone standards was in 2008; the previous 8-hour ozone standards, set in 1997, had a level of 0.08 ppm. In 2004, the U.S. EPA designated nonattainment areas for the 1997 8-hour ozone NAAQS. 34 (35) As of December 5, 2013, there were 39 ozone nonattainment areas for the 1997 ozone NAAQS composed of 216 full or partial counties with a total population of over 112 million. Nonattainment designations for the 2008 ozone standard were finalized on April 30, 2012 and May 31, 2012. (36) As of December 5, 2013, there were 46 ozone nonattainment areas for the 2008 ozone NAAQS, composed of 227 full or partial counties, with a population of over 123 million. As of December 5, 2013, over 135 million people are living in ozone nonattainment areas. (37)

States with ozone nonattainment areas are required to take action to bring those areas into attainment. The attainment date assigned to an ozone nonattainment area is based on the area's classification. Most ozone nonattainment areas were required to attain the 1997 8-hour ozone NAAQS in the 2007 to 2013 time frame and then to maintain it thereafter. (38) The attainment dates for areas designated nonattainment for the 2008 8-hour ozone NAAQS are in the 2015 to 2032 timeframe, depending on the severity of the problem in each area. In addition, EPA is currently working on a review of the ozone NAAQS. If EPA revises the ozone standards pursuant to that review, the attainment dates associated with areas designated nonattainment for that NAAQS would be 5 or more years after the final rule is promulgated, depending on the severity of the problem in each area.

EPA has already adopted many emission control programs that are expected to reduce ambient ozone levels. As a result of these and other federal, state and local programs, 8-hour ozone levels are expected to improve in the future. However, even with the implementation of all current state and federal regulations, there are projected to be counties violating the ozone NAAQS well into the future. Thus additional federal control programs, such as Tier 3, can assist areas with attainment dates in 2018 and beyond in attaining the NAAQS as expeditiously as practicable and may relieve areas with already stringent local regulations from some of the burden associated with adopting additional local controls.

2. Particulate Matter
a. Background

Particulate matter is a highly complex mixture of solid particles and liquid droplets distributed among numerous atmospheric gases which interact with solid and liquid phases. Particles range in size from those smaller than 1 nanometer (10 9 meter) to over 100 micrometer (μm, or 10 6 meter) in diameter (for reference, a typical strand of human hair is 70 μm in diameter and a grain of salt is about 100 μm). Atmospheric particles can be grouped into several classes according to their aerodynamic and physical sizes, including ultrafine particles (<0.1 μm), accumulation mode or `fine' particles (<1 to 3 μm), and coarse particles (>1 to 3 μm). (39) For regulatory purposes, fine particles are measured as PM 2.5 and inhalable or thoracic coarse particles are measured as PM 10-2.5, corresponding to their size (diameter) range in micrometers. The EPA currently has standards that measure PM 2.5 and PM 10. (40)

Particles span many sizes and shapes and may consist of hundreds of different chemicals. Particles are emitted directly from sources and are also formed through atmospheric chemical reactions; the former are often referred to as “primary” particles, and the latter as “secondary” particles. Particle concentration and composition varies by time of year and location, and in addition to differences in source emissions, is affected by several weather-related factors, such as temperature, clouds, humidity, and wind. A further layer of complexity comes from particles' ability to shift between solid/liquid and gaseous phases, which is influenced by concentration and meteorology, especially temperature.

Fine particles are produced primarily by combustion processes and by transformations of gaseous emissions (e.g., sulfur oxides (SO X), oxides of nitrogen, and volatile organic compounds (VOC)) in the atmosphere. The chemical and physical properties of PM 2.5 may vary greatly with time, region, meteorology, and source category. Thus, PM 2.5 may include a complex mixture of different components including sulfates, nitrates, organic compounds, elemental carbon and metal compounds. These particles can remain in the atmosphere for days to weeks and travel hundreds to thousands of kilometers.

b. Health Effects of PM

Scientific studies show ambient PM is associated with a broad range of health effects. These health effects are discussed in detail in the December 2009 Integrated Science Assessment for Particulate Matter (PM ISA). (41) The PM ISA summarizes health effects evidence associated with both short- and long-term exposures to PM 2.5, PM 10-2.5, and ultrafine particles. The PM ISA concludes that human exposures to ambient PM 2.5 concentrations are associated with a number of adverse health effects and characterizes the weight of evidence for these health outcomes. (42) The discussion below highlights the PM ISA's conclusions pertaining to health effects associated with both short- and long-term PM exposures. Further discussion of health effects associated with PM 2.5 can also be found in the rulemaking documents for the most recent review of the PM NAAQS completed in 2012. (43 44)

The EPA concludes that a causal relationship exists between both long- and short-term exposures to PM 2.5 and premature mortality and cardiovascular effects and a likely causal relationship exists between long- and short-term PM 2.5 exposures and respiratory effects. Further, there is evidence suggestive of a causal relationship between long-term PM 2.5 exposures and other health effects, including developmental and reproductive effects (e.g., low birth weight, infant mortality) and carcinogenic, mutagenic, and genotoxic effects (e.g., lung cancer mortality). (45)

As summarized in the Final PM NAAQS rule, and discussed extensively in the 2009 PM ISA, the scientific evidence available since the completion of the 2006 PM NAAQS review significantly strengthens the link between long- and short-term exposure to PM 2.5 and premature mortality, while providing indications that the magnitude of the PM 2.5- mortality association with long-term exposures may be larger than previously estimated. (46 47) The strongest evidence comes from recent studies investigating long-term exposure to PM 2.5 and cardiovascular-related mortality. The evidence supporting a causal relationship between long-term PM 2.5 exposure and mortality also includes consideration of new studies that demonstrated an improvement in community health following reductions in ambient fine particles.

Several studies evaluated in the 2009 PM ISA have examined the association between cardiovascular effects and long-term PM 2.5 exposures in multi-city studies conducted in the U.S. and Europe. While studies were not available in the 2006 PM NAAQS review with regard to long-term exposure and cardiovascular-related morbidity, studies published since then have provided new evidence linking long-term exposure to PM 2.5 with an array of cardiovascular effects such as heart attacks, congestive heart failure, stroke, and mortality. This evidence is coherent with studies of short-term exposure to PM 2.5 that have observed associations with a continuum of effects ranging from subtle changes in indicators of cardiovascular health to serious clinical events, such as increased hospitalizations and emergency department visits due to cardiovascular disease and cardiovascular mortality. (48)

As detailed in the 2009 PM ISA, extended analyses of studies available in the 2006 PM NAAQS review as well as epidemiological studies conducted in the U.S. and abroad published since then provide stronger evidence of respiratory-related morbidity effects associated with long-term PM 2.5 exposure. The strongest evidence for respiratory-related effects is from studies that evaluated decrements in lung function growth (in children), increased respiratory symptoms, and asthma development. The strongest evidence from short-term PM 2.5 exposure studies has been observed for increased respiratory-related emergency department visits and hospital admissions for chronic obstructive pulmonary disease (COPD) and respiratory infections. (49)

The body of scientific evidence detailed in the 2009 PM ISA is still limited with respect to associations between long-term PM 2.5 exposures and developmental and reproductive effects as well as cancer, mutagenic, and genotoxic effects, but is somewhat expanded from the 2006 review. The strongest evidence for an association between PM 2.5 and developmental and reproductive effects comes from epidemiological studies of low birth weight and infant mortality, especially due to respiratory causes during the post-neonatal period (i.e., 1 month to 12 months of age). (50) With regard to cancer effects, “[m]ultiple epidemiologic studies have shown a consistent positive association between PM 2.5 and lung cancer mortality, but studies have generally not reported associations between PM 2.5 and lung cancer incidence.”  (51)

Specific groups within the general population are at increased risk for experiencing adverse health effects related to PM exposures. (52 53 54 55) The evidence detailed in the 2009 PM ISA expands our understanding of previously identified at-risk populations and lifestages (i.e., children, older adults, and individuals with pre-existing heart and lung disease) and supports the identification of additional at-risk populations (e.g., persons with lower socioeconomic status, genetic differences). Additionally, there is emerging, though still limited, evidence for additional potentially at-risk populations and lifestages, such as those with diabetes, people who are obese, pregnant women, and the developing fetus. (56)

For PM 10-2.5, the 2009 PM ISA concluded that available evidence was suggestive of a causal relationship between short-term exposures to PM 10-2.5 and cardiovascular effects (e.g., hospital admissions and ED visits, changes in cardiovascular function), respiratory effects (e.g, ED visits and hospital admissions, increase in markers of pulmonary inflammation), and premature mortality. Data were inadequate to draw conclusions regarding the relationships between long-term exposure to PM 10-2.5 and various health effects. (57 58 59)

For ultrafine particles, the 2009 PM ISA concluded that the evidence was suggestive of a causal relationship between short-term exposures and cardiovascular effects, including changes in heart rhythm and vasomotor function (the ability of blood vessels to expand and contract). It also concluded that there was evidence suggestive of a causal relationship between short-term exposure to ultrafine particles and respiratory effects, including lung function and pulmonary inflammation, with limited and inconsistent evidence for increases in ED visits and hospital admissions. Data were inadequate to draw conclusions regarding the relationship between short-term exposure to ultrafine particle and additional health effects including premature mortality as well as long-term exposure to ultrafine particles and all health outcomes evaluated. (60 61)

c. Current and Projected Concentrations of PM

There are two primary NAAQS for PM 2.5: an annual standard (12.0 micrograms per cubic meter (μg/m 3)) and a 24-hour standard (35 μg/m 3), and two secondary NAAQS for PM 2.5: an annual standard (15.0 μg/m 3) and a 24-hour standard (35 μg/m 3). The initial PM 2.5 standards were set in 1997 and revisions to the standards were finalized in 2006 and in December 2012. The December 2012 rule revised the level of the primary annual PM 2.5 standard from 15.0 μg/m 3 to 12.0 μg/m 3. (62)

There are many areas of the country that are currently in nonattainment for the annual and 24-hour PM 2.5 NAAQS. Our modeling without the Tier 3 controls projects that in the future we will continue to have many areas that will have ambient PM 2.5 concentrations above the level of the NAAQS (see Section III.C.2). States will need to meet the 2006 24-hour standards in the 2015-2019 timeframe and the 2012 primary annual standard in the 2021-2025 timeframe. The emission reductions and improvements in ambient PM 2.5 concentrations from this action, which will take effect starting in 2017, will be helpful to states as they work to attain and maintain the PM 2.5 NAAQS.

In 2005 the EPA designated 39 nonattainment areas for the 1997 PM 2.5 NAAQS. (63) As of December 5, 2013, over 68 million people lived in the 24 areas that are still designated as nonattainment for the 1997 annual PM 2.5 NAAQS. These PM 2.5 nonattainment areas are comprised of 135 full or partial counties. EPA anticipates making initial area designation decisions for the 2012 primary annual PM 2.5 NAAQS in December 2014, with those designations likely becoming effective in early 2015. (64) On November 13, 2009 and February 3, 2011, the EPA designated 32 nonattainment areas for the 2006 24-hour PM 2.5 NAAQS. (65) As of December 5, 2013, 28 of these areas remain designated as nonattainment, and they are composed of 104 full or partial counties with a population of over 65 million. In total, there are currently 39 PM 2.5 nonattainment areas with a population of over 84 million people. (66)

States with PM 2.5 nonattainment areas will be required to take action to bring those areas into attainment in the future. Designated nonattainment areas not currently attaining the 1997 annual PM 2.5 NAAQS are required to attain the NAAQS by 2015 and will be required to maintain the 1997 annual PM 2.5 NAAQS thereafter. The 2006 24-hour PM 2.5 nonattainment areas are required to attain the 2006 24-hour PM 2.5 NAAQS in the 2015 to 2019 time frame and will be required to maintain the 2006 24-hour PM 2.5 NAAQS thereafter. Areas to be designated nonattainment for the 2012 primary annual PM 2.5 NAAQS will likely be required to attain the 2012 NAAQS in the 2021 to 2025 time frame. The Tier 3 standards finalized here begin taking effect in 2017.

The EPA has already adopted many mobile source emission control programs that are expected to reduce ambient PM concentrations. As a result of these and other federal, state and local programs, the number of areas that fail to meet the PM 2.5 NAAQS in the future is expected to decrease. However, even with the implementation of all current state and federal regulations, there are projected to be counties violating the PM 2.5 NAAQS well into the future. Thus additional federal control programs, such as Tier 3, can assist areas with attainment dates in 2017 and beyond in attaining the NAAQS as expeditiously as practicable and may relieve areas with already stringent local regulations from some of the burden associated with adopting additional local controls.

d. Current Concentrations of PM

In the December 2012 action in which the EPA promulgated the revised primary annual PM 2.5 NAAQS, the EPA also retained the existing primary and secondary 24-hour PM 10 standards at 150 µg/m 3. As of December 5, 2013, over 11 million people live in the 40 areas that are designated as nonattainment for the PM 10 NAAQS. There are 33 full or partial counties that make up the PM 10 nonattainment areas.

3. Oxides of Nitrogen and Sulfur
a. Background

Nitrogen dioxide (NO 2) is a member of the NO X family of gases. Most NO 2 is formed in the air through the oxidation of nitric oxide (NO) emitted when fuel is burned at a high temperature. Sulfur dioxide (SO 2), a member of the sulfur oxide (SO X) family of gases, is formed from burning fuels containing sulfur (e.g., coal or oil derived), extracting gasoline from oil, or extracting metals from ore.

SO 2 and NO 2 and their gas phase oxidation products can dissolve in water droplets and further oxidize to form sulfuric and nitric acid which react with ammonia to form sulfates and nitrates, both of which are important components of ambient PM. The health effects of ambient PM are discussed in Section II.B.2.b of this preamble. NO X and VOC are the two major precursors of ozone. The health effects of ozone are covered in Section II.B.2.1.b.

b. Health Effects of NO

The most recent review of the health effects of oxides of nitrogen completed by the EPA can be found in the 2008 Integrated Science Assessment for Nitrogen Oxides (NO X ISA). (67) The EPA concluded that the findings of epidemiologic, controlled human exposure, and animal toxicological studies provide evidence that is sufficient to infer a likely causal relationship between respiratory effects and short-term NO 2 exposure. The 2008 NO X ISA concluded that the strongest evidence for such a relationship comes from epidemiologic studies of respiratory effects including increased respiratory symptoms, emergency department visits, and hospital admissions. Based on both short- and long-term exposure studies, the 2008 NO X ISA concluded that individuals with preexisting pulmonary conditions (e.g., asthma or COPD), children, and older adults are potentially at greater risk of NO 2-related respiratory effects. Based on findings from controlled human exposure studies, the 2008 NO X ISA also drew two broad conclusions regarding airway responsiveness following NO 2 exposure. First, the NO X ISA concluded that NO 2 exposure may enhance the sensitivity to allergen-induced decrements in lung function and increase the allergen-induced airway inflammatory response following 30-minute exposures of asthmatic adults to NO 2 concentrations as low as 260 ppb. Second, exposure to NO 2 has been found to enhance the inherent responsiveness of the airway to subsequent nonspecific challenges in controlled human exposure studies of healthy and asthmatic adults. Small but statistically significant increases in nonspecific airway hyperresponsiveness were reported for asthmatic adults following 30-minute exposures to 200-300 ppb NO 2 and following 1-hour exposures of asthmatics to 100 ppb NO 2. Enhanced airway responsiveness could have important clinical implications for asthmatics since transient increases in airway responsiveness following NO 2 exposure have the potential to increase symptoms and worsen asthma control. Together, the epidemiologic and experimental data sets form a plausible, consistent, and coherent description of a relationship between NO 2 exposures and an array of adverse health effects that range from the onset of respiratory symptoms to hospital admission.

In evaluating a broader range of health effects, the 2008 NO X ISA concluded evidence was “suggestive but not sufficient to infer a causal relationship” between short-term NO 2 exposure and premature mortality and between long-term NO 2 exposure and respiratory effects. The latter was based largely on associations observed between long-term NO 2 exposure and decreases in lung function growth in children. Furthermore, the 2008 NO X ISA concluded that evidence was “inadequate to infer the presence or absence of a causal relationship” between short-term NO 2 exposure and cardiovascular effects as well as between long-term NO 2 exposure and cardiovascular effects, reproductive and developmental effects, premature mortality, and cancer. (68) The conclusions for these health effect categories were informed by uncertainties in the evidence base such as the independent effects of NO 2 exposure within the broader mixture of traffic-related pollutants, limited evidence from experimental studies, and/or an overall limited literature base.

c. Health Effects of SO

Information on the health effects of SO 2 can be found in the 2008 Integrated Science Assessment for Sulfur Oxides (SO 2 ISA). (69) Short-term peaks of SO 2 have long been known to cause adverse respiratory health effects, particularly among individuals with asthma. In addition to those with asthma (both children and adults), potentially sensitive groups include all children and the elderly. During periods of elevated ventilation, asthmatics may experience symptomatic bronchoconstriction within minutes of exposure. Following an extensive evaluation of health evidence from epidemiologic and laboratory studies, the EPA concluded that there is a causal relationship between respiratory health effects and short-term exposure to SO 2. Separately, based on an evaluation of the epidemiologic evidence of associations between short-term exposure to SO 2 and mortality, the EPA concluded that the overall evidence is suggestive of a causal relationship between short-term exposure to SO 2 and mortality.

d. Current Concentrations of NO

The EPA most recently completed a review of the primary NAAQS for NO 2 in January 2010. There are two primary NAAQS for NO 2: an annual standard (53 ppb) and a 1-hour standard (100 ppb). The EPA promulgated area designations in the Federal Register on February 17, 2012. In this initial round of designations, all areas of the country were designated as “unclassifiable/attainment” for the 2010 NO 2 NAAQS based on data from the existing air quality monitoring network. The EPA and state agencies are working to establish an expanded network of NO 2 monitors, expected to be deployed in the 2014-2017 time frame. Once three years of air quality data have been collected from the expanded network, the EPA will be able to evaluate NO 2 air quality in additional locations. (70 71)

e. Current Concentrations of SO

The EPA most recently completed a review of the primary SO 2 NAAQS in June 2010. The current primary NAAQS for SO 2 is a 1-hour standard of 75 ppb. The EPA finalized the initial area designations for 29 nonattainment areas in 16 states in a notice published in the Federal Register on August 5, 2013. In this first round of designations, EPA only designated nonattainment areas that were violating the standard based on existing air quality monitoring data provided by the states. The Agency did not have sufficient information to designate any area as “attainment” or make final decisions about areas for which additional modeling or monitoring is needed (78 FR 47191, August 5, 2013). EPA anticipates designating areas for the revised SO 2 standard in multiple rounds.

4. Carbon Monoxide

Carbon monoxide (CO) is a colorless, odorless gas emitted from combustion processes. Nationally and, particularly in urban areas, the majority of CO emissions to ambient air come from mobile sources.

a. Health Effects of Carbon Monoxide

Information on the health effects of CO can be found in the January 2010 Integrated Science Assessment for Carbon Monoxide (CO ISA). (72) The CO ISA concludes that ambient concentrations of CO are associated with a number of adverse health effects. (73) This section provides a summary of the health effects associated with exposure to ambient concentrations of CO. (74)

Controlled human exposure studies of subjects with coronary artery disease show a decrease in the time to onset of exercise-induced angina (chest pain) and electrocardiogram changes following CO exposure. In addition, epidemiologic studies show associations between short-term CO exposure and cardiovascular morbidity, particularly increased emergency room visits and hospital admissions for coronary heart disease (including ischemic heart disease, myocardial infarction, and angina). Some epidemiologic evidence is also available for increased hospital admissions and emergency room visits for congestive heart failure and cardiovascular disease as a whole. The CO ISA concludes that a causal relationship is likely to exist between short-term exposures to CO and cardiovascular morbidity. It also concludes that available data are inadequate to conclude that a causal relationship exists between long-term exposures to CO and cardiovascular morbidity.

Animal studies show various neurological effects with in-utero CO exposure. Controlled human exposure studies report central nervous system and behavioral effects following low-level CO exposures, although the findings have not been consistent across all studies. The CO ISA concludes the evidence is suggestive of a causal relationship with both short- and long-term exposure to CO and central nervous system effects.

A number of studies cited in the CO ISA have evaluated the role of CO exposure in birth outcomes such as preterm birth or cardiac birth defects. The epidemiologic studies provide limited evidence of a CO-induced effect on preterm births and birth defects, with weak evidence for a decrease in birth weight. Animal toxicological studies have found perinatal CO exposure to affect birth weight, as well as other developmental outcomes. The CO ISA concludes the evidence is suggestive of a causal relationship between long-term exposures to CO and developmental effects and birth outcomes.

Epidemiologic studies provide evidence of associations between ambient CO concentrations and respiratory morbidity such as changes in pulmonary function, respiratory symptoms, and hospital admissions. A limited number of epidemiologic studies considered copollutants such as ozone, SO 2, and PM in two-pollutant models and found that CO risk estimates were generally robust, although this limited evidence makes it difficult to disentangle effects attributed to CO itself from those of the larger complex air pollution mixture. Controlled human exposure studies have not extensively evaluated the effect of CO on respiratory morbidity. Animal studies at levels of 50-100 ppm CO show preliminary evidence of altered pulmonary vascular remodeling and oxidative injury. The CO ISA concludes that the evidence is suggestive of a causal relationship between short-term CO exposure and respiratory morbidity, and inadequate to conclude that a causal relationship exists between long-term exposure and respiratory morbidity.

Finally, the CO ISA concludes that the epidemiologic evidence is suggestive of a causal relationship between short-term concentrations of CO and mortality. Epidemiologic studies provide evidence of an association between short-term exposure to CO and mortality, but limited evidence is available to evaluate cause-specific mortality outcomes associated with CO exposure. In addition, the attenuation of CO risk estimates which was often observed in copollutant models contributes to the uncertainty as to whether CO is acting alone or as an indicator for other combustion-related pollutants. The CO ISA also concludes that there is not likely to be a causal relationship between relevant long-term exposures to CO and mortality.

b. Current Concentrations of CO

There are two NAAQS for CO: an 8-hour standard (9 ppm) and a 1-hour standard (35 ppm). The primary NAAQS for CO were retained in August 2011. There are currently no CO nonattainment areas; as of September 27, 2010, all CO nonattainment areas were redesignated to maintenance areas. The designations were based on the existing community-wide monitoring network. EPA is making changes to the ambient air monitoring requirements for CO. The new requirements are expected to result in approximately 52 CO monitors operating near roads within 52 urban areas by January 2015 (76 FR 54294, August 31, 2011).

5. Mobile Source Air Toxics

Light-duty vehicle emissions contribute to ambient levels of air toxics known or suspected as human or animal carcinogens, or that have noncancer health effects. The population experiences an elevated risk of cancer and other noncancer health effects from exposure to the class of pollutants known collectively as “air toxics.”  (75) These compounds include, but are not limited to, benzene, 1,3-butadiene, formaldehyde, acetaldehyde, acrolein, polycyclic organic matter, and naphthalene. These compounds were identified as national or regional risk drivers or contributors in the 2005 National-scale Air Toxics Assessment and have significant inventory contributions from mobile sources. (76)

a. Health Effects of Air Toxics
i. Benzene

The EPA's Integrated Risk Information System (IRIS) database lists benzene as a known human carcinogen (causing leukemia) by all routes of exposure, and concludes that exposure is associated with additional health effects, including genetic changes in both humans and animals and increased proliferation of bone marrow cells in mice. (77 78 79) EPA states in its IRIS database that 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. EPA's IRIS documentation for benzene also lists a range of 2.2 × 10 −6 to 7.8 × 10 −6 as the unit risk estimate (URE) for benzene. (80 81) The International Agency for Research on Carcinogens (IARC) has determined that benzene is a human carcinogen and the U.S. Department of Health and Human Services (DHHS) has characterized benzene as a known human carcinogen. (82 83)

A number of adverse noncancer health effects including blood disorders, such as preleukemia and aplastic anemia, have also been associated with long-term exposure to benzene. (84 85) The most sensitive noncancer effect observed in humans, based on current data, is the depression of the absolute lymphocyte count in blood. (86 87) EPA's inhalation reference concentration (RfC) for benzene is 30 μg/m 3. The RfC is based on suppressed absolute lymphocyte counts seen in humans under occupational exposure conditions. In addition, recent work, including studies sponsored by the Health Effects Institute, provides evidence that biochemical responses are occurring at lower levels of benzene exposure than previously known. (88 89 90 91) EPA's IRIS program has not yet evaluated these new data. EPA does not currently have an acute reference concentration for benzene. The Agency for Toxic Substances and Disease Registry (ATSDR) Minimal Risk Level (MRL) for acute exposure to benzene is 29 μg/m 3 for 1-14 days exposure. (92 93)

ii. Formaldehyde

In 1991, EPA concluded that formaldehyde is a carcinogen based on nasal tumors in animal bioassays. (94) An Inhalation Unit Risk for cancer and a Reference Dose for oral noncancer effects were developed by the Agency and posted on the IRIS database. Since that time, the National Toxicology Program (NTP) and International Agency for Research on Cancer (IARC) have concluded that formaldehyde is a known human carcinogen. (95 96 97)

The conclusions by IARC and NTP reflect the results of epidemiologic research published since 1991 in combination with previous animal, human and mechanistic evidence. Research conducted by the National Cancer Institute reported an increased risk of nasopharyngeal cancer and specific lymphohematopoietic malignancies among workers exposed to formaldehyde. (98 99 100) A National Institute of Occupational Safety and Health study of garment workers also reported increased risk of death due to leukemia among workers exposed to formaldehyde. (101) Extended follow-up of a cohort of British chemical workers did not report evidence of an increase in nasopharyngeal or lymphohematopoietic cancers, but a continuing statistically significant excess in lung cancers was reported. (102) Finally, a study of embalmers reported formaldehyde exposures to be associated with an increased risk of myeloid leukemia but not brain cancer. (103)

Health effects of formaldehyde in addition to cancer were reviewed by the Agency for Toxics Substances and Disease Registry in 1999  (104) and supplemented in 2010, (105) and by the World Health Organization. (106) These organizations reviewed the literature concerning effects on the eyes and respiratory system, the primary point of contact for inhaled formaldehyde, including sensory irritation of eyes and respiratory tract, pulmonary function, nasal histopathology, and immune system effects. In addition, research on reproductive and developmental effects and neurological effects were discussed.

EPA released a draft Toxicological Review of Formaldehyde—Inhalation Assessment through the IRIS program for peer review by the National Research Council (NRC) and public comment in June 2010. (107) The draft assessment reviewed more recent research from animal and human studies on cancer and other health effects. The NRC released their review report in April 2011. (108) The EPA is currently revising the draft assessment in response to this review.

iii. Acetaldehyde

Acetaldehyde is classified in EPA's IRIS database as a probable human carcinogen, based on nasal tumors in rats, and is considered toxic by the inhalation, oral, and intravenous routes. (109) The URE in IRIS for acetaldehyde is 2.2 × 10 −6 per μg/m 3. (110) Acetaldehyde is reasonably anticipated to be a human carcinogen by the U.S. DHHS in the 12th Report on Carcinogens and is classified as possibly carcinogenic to humans (Group 2B) by the IARC. (111 112) EPA is currently conducting a reassessment of cancer risk from inhalation exposure to acetaldehyde.

The primary noncancer effects of exposure to acetaldehyde vapors include irritation of the eyes, skin, and respiratory tract. (113) In short-term (4 week) rat studies, degeneration of olfactory epithelium was observed at various concentration levels of acetaldehyde exposure. (114 115) Data from these studies were used by EPA to develop an inhalation reference concentration of 9 μg/m 3. Some asthmatics have been shown to be a sensitive subpopulation to decrements in functional expiratory volume (FEV1 test) and bronchoconstriction upon acetaldehyde inhalation. (116) The agency is currently conducting a reassessment of the health hazards from inhalation exposure to acetaldehyde.

iv. Acrolein

EPA most recently evaluated the toxicological and health effects literature related to acrolein in 2003 and concluded that the human carcinogenic potential of acrolein could not be determined because the available data were inadequate. No information was available on the carcinogenic effects of acrolein in humans and the animal data provided inadequate evidence of carcinogenicity. (117) The IARC determined in 1995 that acrolein was not classifiable as to its carcinogenicity in humans. (118)

Lesions to the lungs and upper respiratory tract of rats, rabbits, and hamsters have been observed after subchronic exposure to acrolein. (119) The Agency has developed an RfC for acrolein of 0.02 μg/m 3 and an RfD of 0.5 μg/kg-day. (120) EPA is considering updating the acrolein assessment with data that have become available since the 2003 assessment was completed.

Acrolein is extremely acrid and irritating to humans when inhaled, with acute exposure resulting in upper respiratory tract irritation, mucus hypersecretion and congestion. The intense irritancy of this carbonyl has been demonstrated during controlled tests in human subjects, who suffer intolerable eye and nasal mucosal sensory reactions within minutes of exposure. (121) These data and additional studies regarding acute effects of human exposure to acrolein are summarized in EPA's 2003 IRIS Human Health Assessment for acrolein. (122) Studies in humans indicate that levels as low as 0.09 ppm (0.21 mg/m 3) for five minutes may elicit subjective complaints of eye irritation with increasing concentrations leading to more extensive eye, nose and respiratory symptoms. Acute exposures in animal studies report bronchial hyper-responsiveness. Based on animal data (more pronounced respiratory irritancy in mice with allergic airway disease in comparison to non-diseased mice  (123) ) and demonstration of similar effects in humans (e.g., reduction in respiratory rate), individuals with compromised respiratory function (e.g., emphysema, asthma) are expected to be at increased risk of developing adverse responses to strong respiratory irritants such as acrolein. EPA does not currently have an acute reference concentration for acrolein. The available health effect reference values for acrolein have been summarized by EPA and include an ATSDR MRL for acute exposure to acrolein of 7 μg/m 3 for 1-14 days exposure; and Reference Exposure Level (REL) values from the California Office of Environmental Health Hazard Assessment (OEHHA) for one-hour and 8-hour exposures of 2.5 μg/m 3 and 0.7 μg/m 3, respectively. (124)

v. 1,3-Butadiene

EPA has characterized 1,3-butadiene as carcinogenic to humans by inhalation. (125 126) The IARC has determined that 1,3-butadiene is a human carcinogen and the U.S. DHHS has characterized 1,3-butadiene as a known human carcinogen. 127 128 129 There are numerous studies consistently demonstrating that 1,3-butadiene is metabolized into genotoxic metabolites by experimental animals and humans. The specific mechanisms of 1,3-butadiene-induced carcinogenesis are unknown; however, the scientific evidence strongly suggests that the carcinogenic effects are mediated by genotoxic metabolites. Animal data suggest that females may be more sensitive than males for cancer effects associated with 1,3-butadiene exposure; there are insufficient data in humans from which to draw conclusions about sensitive subpopulations. The URE for 1,3-butadiene is 3 × 10 −5 per μg/m 3. (130) 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. (131) Based on this critical effect and the benchmark concentration methodology, an RfC for chronic health effects was calculated at 0.9 ppb (approximately 2 μg/m 3).

vi. Ethanol

EPA is planning to develop an assessment of the health effects of exposure to ethanol, a compound which is not currently listed on EPA's IRIS database. Extensive health effects data are available for ingestion of ethanol, while data on inhalation exposure effects are sparse. In developing the assessment, EPA is evaluating pharmacokinetic models as a means of extrapolating across species (animal to human) and across exposure routes (oral to inhalation) to better characterize the health hazards and dose-response relationships for low levels of ethanol exposure in the environment.

vii. Polycyclic Organic Matter

The term polycyclic organic matter (POM) defines a broad class of compounds that includes the polycyclic aromatic hydrocarbon compounds (PAHs). One of these compounds, naphthalene, is discussed separately below. POM compounds are formed primarily from combustion and are present in the atmosphere in gas and particulate form. Cancer is the major concern from exposure to POM. Epidemiologic studies have reported an increase in lung cancer in humans exposed to diesel exhaust, coke oven emissions, roofing tar emissions, and cigarette smoke; all of these mixtures contain POM compounds. (132 133) Animal studies have reported respiratory tract tumors from inhalation exposure to benzo[a]pyrene and alimentary tract and liver tumors from oral exposure to benzo[a]pyrene. (134) In 1997 EPA classified seven PAHs (benzo[a]pyrene, benz[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, dibenz[a,h]anthracene, and indeno[1,2,3-cd]pyrene) as Group B2, probable human carcinogens. (135) Since that time, studies have found that maternal exposures to PAHs in a population of pregnant women were associated with several adverse birth outcomes, including low birth weight and reduced length at birth, as well as impaired cognitive development in preschool children (3 years of age). (136 137) These and similar studies are being evaluated as a part of the ongoing IRIS assessment of health effects associated with exposure to benzo[a]pyrene.

viii. Naphthalene

Naphthalene is found in small quantities in gasoline and diesel fuels. Naphthalene emissions have been measured in larger quantities in both gasoline and diesel exhaust compared with evaporative emissions from mobile sources, indicating it is primarily a product of combustion. Acute (short-term) exposure of humans to naphthalene by inhalation, ingestion, or dermal contact is associated with hemolytic anemia and damage to the liver and the nervous system. (138) Chronic (long term) exposure of workers and rodents to naphthalene has been reported to cause cataracts and retinal damage. (139) EPA released an external review draft of a reassessment of the inhalation carcinogenicity of naphthalene based on a number of recent animal carcinogenicity studies. (140) The draft reassessment completed external peer review. (141) Based on external peer review comments received, a revised draft assessment that considers all routes of exposure, as well as cancer and noncancer effects, is under development. The external review draft does not represent official agency opinion and was released solely for the purposes of external peer review and public comment. The National Toxicology Program listed naphthalene as “reasonably anticipated to be a human carcinogen” in 2004 on the basis of bioassays reporting clear evidence of carcinogenicity in rats and some evidence of carcinogenicity in mice. (142) California EPA has released a new risk assessment for naphthalene, and the IARC has reevaluated naphthalene and re-classified it as Group 2B: possibly carcinogenic to humans. (143)

Naphthalene also causes a number of chronic non-cancer effects in animals, including abnormal cell changes and growth in respiratory and nasal tissues. (144) The current EPA IRIS assessment includes noncancer data on hyperplasia and metaplasia in nasal tissue that form the basis of the inhalation RfC of 3 μg/m 3. (145) The ATSDR MRL for acute exposure to naphthalene is 0.6 mg/kg/day.

ix. Other Air Toxics

In addition to the compounds described above, other compounds in gaseous hydrocarbon and PM emissions from motor vehicles will be affected by this action. Mobile source air toxic compounds that will potentially be impacted include ethylbenzene, propionaldehyde, toluene, and xylene. Information regarding the health effects of these compounds can be found in EPA's IRIS database. (146)

b. Current Concentrations of Air Toxics

The most recent available data indicate that the majority of Americans continue to be exposed to ambient concentrations of air toxics at levels which have the potential to cause adverse health effects. (147) The levels of air toxics to which people are exposed vary depending on where people live and work and the kinds of activities in which they engage, as discussed in detail in U.S. EPA's most recent Mobile Source Air Toxics Rule. (148) According to the National Air Toxic Assessment (NATA) for 2005, (149) mobile sources were responsible for 43 percent of outdoor toxic emissions and over 50 percent of the cancer risk and noncancer hazard associated with primary emissions. Mobile sources are also large contributors to precursor emissions which react to form secondary concentrations of air toxics. Formaldehyde is the largest contributor to cancer risk of all 80 pollutants quantitatively assessed in the 2005 NATA. Mobile sources were responsible for over 40 percent of primary emissions of this pollutant in 2005, and are major contributors to formaldehyde precursor emissions. Benzene is also a large contributor to cancer risk, and mobile sources account for over 70 percent of ambient exposure. Over the years, EPA has implemented a number of mobile source and fuel controls which have resulted in VOC reductions, which also reduced formaldehyde, benzene and other air toxic emissions.

6. Near-Roadway Pollution

Locations in close proximity to major roadways generally have elevated concentrations of many air pollutants emitted from motor vehicles. Hundreds of such studies have been published in peer-reviewed journals, concluding that concentrations of CO, NO, NO 2, benzene, aldehydes, particulate matter, black carbon, and many other compounds are elevated in ambient air within approximately 300-600 meters (about 1,000-2,000 feet) of major roadways. Highest concentrations of most pollutants emitted directly by motor vehicles are found at locations within 50 meters (about 165 feet) of the edge of a roadway's traffic lanes.

A recent large-scale review of air quality measurements in vicinity of major roadways between 1978 and 2008 concluded that the pollutants with the steepest concentration gradients in vicinities of roadways were CO, ultrafine particles, metals, elemental carbon (EC), NO, NO X, and several VOCs. (150) These pollutants showed a large reduction in concentrations within 100 meters downwind of the roadway. Pollutants that showed more gradual reductions with distance from roadways included benzene, NO 2, PM 2.5, and PM 10. In the review article, results varied based on the method of statistical analysis used to determine the trend.

For pollutants with relatively high background concentrations relative to near-road concentrations, detecting concentration gradients can be difficult. For example, many aldehydes have high background concentrations as a result of photochemical breakdown of precursors from many different organic compounds. This can make detection of gradients around roadways and other primary emission sources difficult. However, several studies have measured aldehydes in multiple weather conditions, and found higher concentrations of many carbonyls downwind of roadways. (151, 152) These findings suggest a substantial roadway source of these carbonyls.

In the past 15 years, many studies have been published with results reporting that populations who live, work, or go to school near high-traffic roadways experience higher rates of numerous adverse health effects, compared to populations far away from major roads. (153) In addition, numerous studies have found adverse health effects associated with spending time in traffic, such as commuting or walking along high-traffic roadways. 154155156 (157) The health outcomes with the strongest evidence linking them with traffic-associated air pollutants are respiratory effects, particularly in asthmatic children, and cardiovascular effects.

Numerous reviews of this body of health literature have been published as well. In 2010, an expert panel of the Health Effects Institute (HEI) published a review of hundreds of exposure, epidemiology, and toxicology studies. (158) The panel rated how the evidence for each type of health outcome supported a conclusion of a causal association with traffic-associated air pollution as either “sufficient,” “suggestive but not sufficient,” or “inadequate and insufficient.” The panel categorized evidence of a causal association for exacerbation of childhood asthma as “sufficient.” The panel categorized evidence of a causal association for new onset asthma as between “sufficient” and as “suggestive but not sufficient.” “Suggestive of a causal association” was how the panel categorized evidence linking traffic-associated air pollutants with exacerbation of adult respiratory symptoms and lung function decrement. It categorized as “inadequate and insufficient” evidence of a causal relationship between traffic-related air pollution and health care utilization for respiratory problems, new onset adult asthma, chronic obstructive pulmonary disease (COPD), nonasthmatic respiratory allergy, and cancer in adults and children. Other literature reviews have been published with conclusions similar to the HEI panel's. 159160 (161) Health outcomes with few publications suggest the possibility of other effects still lacking sufficient evidence to draw definitive conclusions. Among these outcomes with a small number of positive studies are neurological impacts (e.g., autism and reduced cognitive function) and reproductive outcomes (e.g., preterm birth, low birth weight). 162163164 (165)

In addition to health outcomes, particularly cardiopulmonary effects, conclusions of numerous studies suggest mechanisms by which traffic-related air pollution affects health. Numerous studies indicate that near-roadway exposures may increase systemic inflammation, affecting organ systems, including blood vessels and lungs. 166167168 (169) Long-term exposures in near-road environments have been associated with inflammation-associated conditions, such as atherosclerosis and asthma. 170171 (172)

Several studies suggest that some factors may increase susceptibility to the effects of traffic-associated air pollution. Several studies have found stronger respiratory associations in children experiencing chronic social stress, such as in violent neighborhoods or in homes with high family stress. 173174 (175)

The risks associated with residence, workplace, or schools near major roads are of potentially high public health significance due to the large population in such locations. According to the 2009 American Housing Survey, over 22 million homes (17.0 percent of all U.S. housing units) were located within 300 feet of an airport, railroad, or highway with four or more lanes. This corresponds to a population of more than 50 million U.S. residents in close proximity to high-traffic roadways or other transportation sources. Based on 2010 Census data, a 2013 publication estimated that 19 percent of the U.S. population (over 59 million people) lived within 500 meters of roads with at least 25,000 annual average daily traffic (AADT), while about 3.2 percent of the population lived within 100 meters (about 300 feet) of such roads. (176) Another 2013 study estimated that 3.7 percent of the U.S. population (about 11.3 million people) lived within 150 meters (about 500 feet) of interstate highways, or other freeways and expressways. (177) As discussed in Section III, on average, populations near major roads have higher fractions of minority residents and lower socioeconomic status. Furthermore, on average, Americans spend more than an hour traveling each day, bringing nearly all residents into a high-exposure microenvironment for part of the day.

In light of these concerns, EPA has required and is working with states to ensure that air quality monitors be placed near high-traffic roadways for determining NAAQS compliance for CO, NO 2, and PM 2.5 (in addition to those existing monitors located in neighborhoods and other locations farther away from pollution sources). Near-roadway monitors for NO 2 begin operation between 2014 and 2017 in Core Based Statistical Areas (CBSAs) with population of at least 500,000. Monitors for CO and PM 2.5 begin operation between 2015 and 2017. These monitors will further our understanding of exposure in these locations.

EPA continues to research near-road air quality, including the types of pollutants found in high concentrations near major roads and health problems associated with the mixture of pollutants near roads.

7. Environmental Impacts of Motor Vehicles and Fuels
a. Plant and Ecosystem Effects of Ozone

The welfare effects of ozone can be observed across a variety of scales, i.e. subcellular, cellular, leaf, whole plant, population and ecosystem. Ozone effects that begin at small spatial scales, such as the leaf of an individual plant, when they occur at sufficient magnitudes (or to a sufficient degree) can result in effects being propagated along a continuum to larger and larger spatial scales. For example, effects at the individual plant level, such as altered rates of leaf gas exchange, growth and reproduction can, when widespread, result in broad changes in ecosystems, such as productivity, carbon storage, water cycling, nutrient cycling, and community composition.

Ozone can produce both acute and chronic injury in sensitive species depending on the concentration level and the duration of the exposure. (178) In those sensitive species,  (179) effects from repeated exposure to ozone throughout the growing season of the plant tend to accumulate, so that even low concentrations experienced for a longer duration have the potential to create chronic stress on vegetation. (180) Ozone damage to sensitive species includes impaired photosynthesis and visible injury to leaves. The impairment of photosynthesis, the process by which the plant makes carbohydrates (its source of energy and food), can lead to reduced crop yields, timber production, and plant productivity and growth. Impaired photosynthesis can also lead to a reduction in root growth and carbohydrate storage below ground, resulting in other, more subtle plant and ecosystems impacts. (181) These latter impacts include increased susceptibility of plants to insect attack, disease, harsh weather, interspecies competition and overall decreased plant vigor. The adverse effects of ozone on areas with sensitive species could potentially lead to species shifts and loss from the affected ecosystems, (182) resulting in a loss or reduction in associated ecosystem goods and services. Additionally, visible ozone injury to leaves can result in a loss of aesthetic value in areas of special scenic significance like national parks and wilderness areas and reduced use of sensitive ornamentals in landscaping. (183)

The Integrated Science Assessment (ISA) for Ozone presents more detailed information on how ozone affects vegetation and ecosystems. (184) The ISA concludes that ambient concentrations of ozone are associated with a number of adverse welfare effects and characterizes the weight of evidence for different effects associated with ozone. (185) The ISA concludes that visible foliar injury effects on vegetation, reduced vegetation growth, reduced productivity in terrestrial ecosystems, reduced yield and quality of agricultural crops, and alteration of below-ground biogeochemical cycles are causally associated with exposure to ozone. It also concludes that reduced carbon sequestration in terrestrial ecosystems, alteration of terrestrial ecosystem water cycling, and alteration of terrestrial community composition are likely to be causally associated with exposure to ozone.

b. Visibility

Visibility can be defined as the degree to which the atmosphere is transparent to visible light. (186) Visibility impairment is caused by light scattering and absorption by suspended particles and gases. Visibility is important because it has direct significance to people's enjoyment of daily activities in all parts of the country. Individuals value good visibility for the well-being it provides them directly, where they live and work, and in places where they enjoy recreational opportunities. Visibility is also highly valued in significant natural areas, such as national parks and wilderness areas, and special emphasis is given to protecting visibility in these areas. For more information on visibility see the final 2009 PM ISA. (187)

EPA is working to address visibility impairment. In 1999, EPA finalized the regional haze program to protect the visibility in Mandatory Class I Federal areas. (188) There are 156 national parks, forests and wilderness areas categorized as Mandatory Class I Federal areas. (189) These areas are defined in CAA section 162 as those national parks exceeding 6,000 acres, wilderness areas and memorial parks exceeding 5,000 acres, and all international parks which were in existence on August 7, 1977. EPA has also concluded that PM 2.5 causes adverse effects on visibility in other areas that are not protected by the Regional Haze Rule, depending on PM 2.5 concentrations and other factors that control their visibility impact effectiveness such as dry chemical composition and relative humidity (i.e., an indicator of the water composition of the particles). EPA revised the PM 2.5 standards in December 2012 and established a target level of protection that is expected to be met through attainment of the existing secondary standards for PM 2.5.

i. Current Visibility Levels

As mentioned in Section II.B.2.c, millions of people live in nonattainment areas for the PM 2.5 NAAQS. These populations, as well as large numbers of individuals who travel to these areas, are likely to experience visibility impairment. In addition, while visibility trends have improved in mandatory class I federal areas, the most recent data show that these areas continue to suffer from visibility impairment. In summary, visibility impairment is experienced throughout the U.S., in multi-state regions, urban areas, and remote mandatory class I federal areas.

c. Atmospheric Deposition

Wet and dry deposition of ambient particulate matter delivers a complex mixture of metals (e.g., mercury, zinc, lead, nickel, aluminum, cadmium), organic compounds (e.g., polycyclic organic matter, dioxins, furans) and inorganic compounds (e.g., nitrate, sulfate) to terrestrial and aquatic ecosystems. The chemical form of the compounds deposited depends on a variety of factors including ambient conditions (e.g., temperature, humidity, oxidant levels) and the sources of the material. Chemical and physical transformations of the compounds occur in the atmosphere as well as the media onto which they deposit. These transformations in turn influence the fate, bioavailability and potential toxicity of these compounds. Atmospheric deposition has been identified as a key component of the environmental and human health hazard posed by several pollutants including mercury, dioxin and PCBs. (190)

Adverse impacts on water quality can occur when atmospheric contaminants deposit to the water surface or when material deposited on the land enters a waterbody through runoff. Potential impacts of atmospheric deposition to waterbodies include those related to both nutrient and toxic inputs. Adverse effects to human health and welfare can occur from the addition of excess nitrogen via atmospheric deposition. The nitrogen-nutrient enrichment contributes to toxic algae blooms and zones of depleted oxygen, which can lead to fish kills, frequently in coastal waters. Deposition of heavy metals or other toxics may lead to the human ingestion of contaminated fish, impairment of drinking water, damage to freshwater and marine ecosystem components, and limits to recreational uses. Several studies have been conducted in U.S. coastal waters and in the Great Lakes Region in which the role of ambient PM deposition and runoff is investigated. (191 192 193 194 195)

Atmospheric deposition of nitrogen and sulfur contributes to acidification, altering biogeochemistry and affecting animal and plant life in terrestrial and aquatic ecosystems across the United States. The sensitivity of terrestrial and aquatic ecosystems to acidification from nitrogen and sulfur deposition is predominantly governed by geology. Prolonged exposure to excess nitrogen and sulfur deposition in sensitive areas acidifies lakes, rivers and soils. Increased acidity in surface waters creates inhospitable conditions for biota and affects the abundance and nutritional value of preferred prey species, threatening biodiversity and ecosystem function. Over time, acidifying deposition also removes essential nutrients from forest soils, depleting the capacity of soils to neutralize future acid loadings and negatively affecting forest sustainability. Major effects include a decline in sensitive forest tree species, such as red spruce (Picea rubens) and sugar maple (Acer saccharum), and a loss of biodiversity of fishes, zooplankton, and macro invertebrates.

In addition to the role nitrogen deposition plays in acidification, nitrogen deposition also leads to nutrient enrichment and altered biogeochemical cycling. In aquatic systems increased nitrogen can alter species assemblages and cause eutrophication. In terrestrial systems nitrogen loading can lead to loss of nitrogen sensitive lichen species, decreased biodiversity of grasslands, meadows and other sensitive habitats, and increased potential for invasive species. For a broader explanation of the topics treated here, refer to the description in Section 6.3.2 of the RIA.

Adverse impacts on soil chemistry and plant life have been observed for areas heavily influenced by atmospheric deposition of nutrients, metals and acid species, resulting in species shifts, loss of biodiversity, forest decline, damage to forest productivity and reductions in ecosystem services. Potential impacts also include adverse effects to human health through ingestion of contaminated vegetation or livestock (as in the case for dioxin deposition), reduction in crop yield, and limited use of land due to contamination.

Atmospheric deposition of pollutants can reduce the aesthetic appeal of buildings and culturally important articles through soiling, and can contribute directly (or in conjunction with other pollutants) to structural damage by means of corrosion or erosion. Atmospheric deposition may affect materials principally by promoting and accelerating the corrosion of metals, by degrading paints, and by deteriorating building materials such as concrete and limestone. Particles contribute to these effects because of their electrolytic, hygroscopic, and acidic properties, and their ability to adsorb corrosive gases (principally sulfur dioxide).

i. Current Nitrogen and Sulfur Deposition

Over the past two decades, the EPA has undertaken numerous efforts to reduce nitrogen and sulfur deposition across the U.S. Analyses of long-term monitoring data for the U.S. show that deposition of both nitrogen and sulfur compounds has decreased over the last 19 years. (196) The data show that reductions were more substantial for sulfur compounds than for nitrogen compounds. In the eastern U.S., where data are most abundant, total sulfur deposition decreased by about 44 percent between 1990 and 2007, while total nitrogen deposition decreased by 25 percent over the same time frame. (197) These numbers are generated by the U.S. national monitoring network and they likely underestimate nitrogen deposition because neither ammonia nor organic nitrogen is measured. Although total nitrogen and sulfur deposition has decreased over time, many areas continue to be negatively impacted by deposition. Deposition of inorganic nitrogen and sulfur species routinely measured in the U.S. between 2005 and 2007 were as high as 9.6 kilograms of nitrogen per hectare (kg N/ha) averaged over three years and 20.8 kilograms of sulfur per hectare (kg S/ha) averaged over three years. (198)

d. Environmental Effects of Air Toxics

Emissions from producing, transporting and combusting fuel contribute to ambient levels of pollutants that contribute to adverse effects on vegetation. Volatile organic compounds, some of which are considered air toxics, have long been suspected to play a role in vegetation damage. (199) In laboratory experiments, a wide range of tolerance to VOCs has been observed. (200) Decreases in harvested seed pod weight have been reported for the more sensitive plants, and some studies have reported effects on seed germination, flowering and fruit ripening. Effects of individual VOCs or their role in conjunction with other stressors (e.g., acidification, drought, temperature extremes) have not been well studied. In a recent study of a mixture of VOCs including ethanol and toluene on herbaceous plants, significant effects on seed production, leaf water content and photosynthetic efficiency were reported for some plant species. (201)

Research suggests an adverse impact of vehicle exhaust on plants, which has in some cases been attributed to aromatic compounds and in other cases to nitrogen oxides. (202 203 204)

III. How would this rule reduce emissions and air pollution?

A. Effects of the Vehicle and Fuel Changes on Mobile Source Emissions

The Tier 3 vehicle and fuel standards will significantly reduce the tailpipe and evaporative emissions of light- and heavy-duty vehicles in several ways, as described in this section. In addition, the gasoline sulfur standard will reduce emissions of SO 2 from existing gasoline-powered vehicles and equipment. As described in Section II, all of these emission reductions will in turn improve air quality nationwide and reduce the health effects associated with air pollution from mobile sources.

As with the Tier 2 program, EPA is implementing closely-coordinated requirements for both automakers and refiners in the same rulemaking action. The Tier 3 vehicle emission standards and gasoline sulfur standards represent a “systems approach” to reducing vehicle-related exhaust and evaporative emissions. By recognizing the relationships among the various sources of emissions addressed by this action, we have been able to integrate the provisions into a single, coordinated program.

1. How do vehicles produce the emissions addressed in this action?

The degree to which vehicles produce exhaust and evaporative emissions depends on the design and functionality of the engine and the associated exhaust and evaporative emission controls, in concert with the properties of the fuel on which the vehicle is operating. In the following paragraphs, we discuss how light- and heavy-duty vehicles produce each of these types of emissions, both from the tailpipe and from the fuel system.

a. Tailpipe (Exhaust) Emissions

The pollutants emitted at the vehicle's tailpipe and their quantities depend on how the fuel is combusted in the engine and how the resulting gases are treated in the exhaust system. Historically, much of tailpipe emission control has focused on hydrocarbon compounds (HC) and NO X. The portion of hydrocarbons that is methane is minimally reactive in forming ozone. Thus, for emission control purposes, the focus is generally on non-methane hydrocarbons (NMHC), which are also expressed as non-methane organic gases (NMOG) in order to account for oxygenates (usually ethanol) now usually present in the fuel.

Tailpipe hydrocarbon emissions also include several toxic pollutants, including benzene, acetaldehyde, and formaldehyde. To varying degrees, the mass emissions of these pollutants are reduced along with other hydrocarbons by the catalytic converter and improved engine controls.

Light- and heavy-duty gasoline vehicles also emit PM and CO. PM forms directly as a combustion product (as elemental carbon or soot) and indirectly as semi-volatile hydrocarbon compounds that form particles in the exhaust system or soon after exiting the tailpipe. CO is a product of incomplete fuel combustion.

When operating properly, modern exhaust emission controls (centering on the catalytic convertor) can reduce much of the HC (including toxics), NO X and CO exiting the engine. However, tailpipe emissions are increased during periods of vehicle startup, as catalytic convertors must warm up to be effective; during subsequent operation due to the interference of sulfur in the gasoline; during high load operating events, as the catalyst is overwhelmed or its operation is modified to protect against permanent damage; and as a vehicle ages, as the catalyst degrades in performance due to the effects of high temperature operation and contaminants in the fuel and lubricating oil.

b. Evaporative Emissions

Gasoline vehicles also produce vapors in the fuel tank and fuel system that can be released as evaporative emissions. These vapors are primarily the lighter, more volatile hydrocarbon compounds in gasoline. As discussed in Section IV below, vehicle evaporative (“evap”) control systems are designed to block or capture vapors as they are generated. Vapors are generated in the vehicle fuel tank and fuel system (and released to the atmosphere if not adequately controlled) as fuel heats up due to ambient temperature increase and/or vehicle operation. Fuel vapors are also released when they permeate through elastomers in the fuel system, when they leak at connections or due to damaged components, and during refueling events.

In general, the evap emission controls on current vehicles (and that will be improved under this action) consist of a canister filled with activated charcoal and connected by hoses to the fuel system. The hoses direct generated vapors to the canister, which collects the vapors on the carbon and stores them until the system experiences a “purge” event. During purge, the engine draws fresh air through the canister, carrying vapors released by the carbon to the engine to be combusted and restoring the capacity of the canister. Evaporative emissions occur when vapors are emitted to the atmosphere because the evap system is compromised, the carbon canister is overwhelmed, or vapors permeate or leak. As such, evaporative emission controls also involve proper material selection for fuel system components, careful design of these components, and onboard diagnostics to check the system for failure.

2. How will the changes to gasoline sulfur content affect vehicle emissions?

Gasoline vehicles rely on highly efficient aftertreatment catalysts to control tailpipe emissions of harmful pollutants like CO and NO X, as well as VOCs that include air toxics and precursor compounds to ozone and secondary PM in the atmosphere. These catalysts utilize finely-dispersed precious metals that are susceptible to deactivation by sulfur compounds in the exhaust. Studies have repeatedly demonstrated that the presence of even a tiny amount of sulfur in fuel has a measurable impact on the ability of the catalyst to control emissions, and that emission levels of most pollutants, especially NO X, are very sensitive to fuel sulfur. (205 206)

Sulfur naturally occurs in crude oil and is carried through the refining process into gasoline. EPA's Tier 2 rulemaking for light-duty vehicles, published in 2000, required refiners to reduce sulfur levels in gasoline to an average of 30 ppm, a reduction of about 90 percent from the in-use baseline. At the time, there were indications that sulfur reductions below 30 ppm may provide additional emission benefits. However, the data was insufficient to quantify the benefits to the existing fleet, and the Tier 2 vehicle standards could be achieved without lowering sulfur below 30 ppm. (207)

As discussed in Section IV.A.6, subsequent research provides a compelling case that even this level of sulfur degrades the emission performance of vehicles on the road today and inhibits necessary further reductions in vehicle emissions performance, which depend on optimum catalyst performance to reach emission targets. A study conducted by EPA and the auto industry in support of the Mobile Source Air Toxics (MSAT) rule found significant reductions in NO X, CO and total HC when nine Tier 2 vehicles were tested on ultra-low sulfur fuel. (208) In particular, the study found a 32 percent decrease in NO X when sulfur was reduced from 32 ppm to 6 ppm (equivalent to a 25 percent decrease if sulfur levels were reduced from 30 to 10 ppm, assuming a linear effect). Another recent study by Umicore showed reductions of 41 percent for NO X and 17 percent for hydrocarbons on a PZEV operating on fuel with 33 ppm and 3 ppm fuel (equivalent to reductions of 27 percent and 11 percent, respectively, if sulfur levels were reduced from 30 to 10 ppm, assuming a linear effect). (209)

A larger study of Tier 2 vehicles recently completed by EPA confirmed these results, showing significant reductions in FTP-composite NO X (14 percent), CO (10 percent) and total HC (15 percent) on the 5 ppm fuel, relative to 28 ppm fuel (equivalent to 12 percent, 9 percent, and 13 percent reduction, respectively, if sulfur levels were reduced from 30 to 10 ppm, assuming a linear effect). (210) For NO X, the majority of overall reductions were driven by large reductions on warmed-up periods of the test cycle (Bag 2), which showed a 52 percent reduction using 5 ppm fuel relative to 28 ppm fuel (equivalent to 45 percent reduction if sulfur levels were reduced from 30 to 10 ppm, assuming a linear effect), consistent with the role of sulfur in catalyst degradation discussed above. For additional details regarding these results, please see Section IV.A.6.c.

Our application of these study results assumes a linear effect of sulfur level on catalyst efficiency between the high and low sulfur test fuels. This is reasonable given that the mass flow rate of sulfur in exhaust gas changes in proportion to its concentration in the fuel, and that the chemical kinetics of adsorption of sulfur to the precious metal sites is approximately first order. Linearity of effect is also supported by past studies with multiple fuel sulfur levels such as the CRC E-60 and 2000 AAM/AIAM/Oil Industry emission test programs. (211 212)

Based on these analyses, the benefits of the Tier 3 sulfur standard are significant in two ways: They enable vehicles designed to the Tier 3 tailpipe exhaust standards to meet these standards for the duration of their useful life, and they facilitate immediate emission reductions from all the vehicles on the road at the time the sulfur controls are implemented.

B. How will emissions be reduced?

The Tier 3 standards will reduce emissions of VOC, NO X (including NO 2), direct PM 2.5, CO, SO 2, and air toxics. The sulfur standards will reduce emissions from the on-road fleet immediately upon implementation in calendar year 2017. The vehicle standards will begin to reduce emissions as the cleaner cars and trucks begin to enter the fleet in model year 2017 and model year 2018, respectively. The magnitude of reduction will grow as more Tier 3 vehicles enter the fleet. We present emission reductions in calendar year 2018 to reflect the early reductions expected from the Tier 3 standards, and in calendar year 2030, when 70 percent of the miles travelled are from vehicles that meet the fully phased-in Tier 3 standards. Although 2030 is the farthest year that is feasible for air quality modeling, the full reduction of the vehicle program will be realized after 2030, when the fleet has fully turned over to vehicles that meet the fully phased-in Tier 3 standards; thus we present emission reductions projected in 2050 as well (see Chapter 7 of the RIA).

Emission reductions are estimated on an annual basis, for all 50 U.S. states plus the District of Columbia, Puerto Rico and the U.S. Virgin Islands. The reductions were estimated using a version of EPA's MOVES model updated for this analysis, as described in detail in Chapter 7 of the RIA. This version of MOVES includes our most recent data on how vehicle emissions are affected by changes in sulfur, ethanol, RVP, and other fuel properties. We estimated emission reductions compared to a reference case that assumed renewable fuel volumes and ethanol blends based on the U.S. Energy Information Administration's Annual Energy Outlook 2013 (AEO2013). (213) As described in Chapter 7 of the RIA, the reference and control scenarios based on AEO2013 reflect a mix of E10, E15, and E85 in both 2018 and 2030. The reference case assumed an average sulfur level of 30 ppm (10 ppm in California) and continuation of the Tier 2 vehicle program indefinitely, with the exception of California and Section 177 states that have adopted the LEV III program.

The analysis described here accounts for the following national onroad rules:

  • Tier 2 Motor Vehicle Emissions Standards and Gasoline Sulfur Control Requirements (65 FR 6698, February 10, 2000)
  • Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements (66 FR 5002, January 18, 2001)
  • Mobile Source Air Toxics Rule (72 FR 8428, February 26, 2007)
  • Regulation of Fuels and Fuel Additives: Changes to Renewable Fuel Standard Program (75 FR 14670, March 26, 2010)
  • Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards for 2012-2016 (75 FR 25324, May 7, 2010)
  • Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles (76 FR 57106, September 15, 2011)
  • 2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas Emissions and Corporate Average Fuel Economy Standards (77 FR 62623, October 15, 2012)

The analysis also accounts for many other national rules and standards. In addition, the modeling accounts for state and local rules including California's most recent Low Emission Vehicle (LEV III) program adopted in California and twelve other states (also referred to as Section 177 states), (214) local fuel standards, Inspection/Maintenance programs, Stage II refueling controls, the National Low Emission Vehicle Program (NLEV), and the Section 177 states LEV and LEV II programs. See the Tier 3 emissions modeling TSD for more detail.

A summary of emission reductions projected to result from Tier 3, relative to the reference case, is shown in calendar years 2018 and 2030 for NO X, VOC, direct PM 2.5, CO, SO 2, and total air toxics in Table III-1. For many pollutants, the immediate reductions in 2018 are significant; for example, combined NO X and VOC emissions will be reduced by over 300,000 tons. By 2030, combined NO X and VOC emissions will be reduced by roughly 500,000 tons, one quarter of the onroad inventory. Many of the modeled air toxics will be significantly reduced as well, including benzene, 1,3-butadiene, acetaldehyde, acrolein and ethanol (ranging from 10 to nearly 30 percent of the national onroad inventory by 2030). The relative reduction in overall emissions will continue to increase beyond 2030 as more of the fleet continues to turn over to Tier 3 vehicles; for example, by 2050, when nearly all of the fleet will have turned over to vehicles meeting the fully phased-in Tier 3 standards, we estimate the Tier 3 program will reduce onroad emissions of NO X and VOC nearly 31 percent from the level of emissions projected without Tier 3 controls.

Table III-1—Estimated Emission Reductions From the Tier 3 Standards
2018Tons% of Onroad inventory2030Tons% of Onroad inventory
NO X 264,36910328,50925
VOC47,5043167,59116
CO278,87923,458,04124
Direct PM 2.5 1300.17,89210
Benzene1,91664,76226
SO 2 14,8135612,39956
1,3-Butadiene257567729
Formaldehyde51321,27710
Acetaldehyde60032,06721
Acrolein40312715
Ethanol2,704219,95016

Reductions for each pollutant are discussed in the following sections, focusing on the contribution of program elements to the total reductions summarized above.

1. NO

The Tier 3 sulfur standards will significantly reduce NO X emissions immediately upon implementation of the program. As discussed above, recent research on the impact of sulfur on Tier 2 technology vehicles shows the potential for significant reductions in NO X emissions from the existing fleet of Tier 2 vehicles by lowering sulfur levels to 10 ppm. Prior research shows that NO X emissions will also be expected to decrease from the fleet of older (pre-Tier 2) light-duty vehicles as well as heavy-duty gasoline vehicles, (215) although to a lesser extent than for Tier 2 vehicles.

Table III-2 shows the reduction in NO X emissions, in annual short tons, projected in calendar years 2018 and 2030. The reductions are split into those attributable to the introduction of low sulfur fuel in the pre-Tier 3 fleet (defined for this analysis as model years prior to 2017); and reductions attributable to vehicle standards enabled by low sulfur fuel (model year 2017 and later). As shown, upon implementation of the Tier 3 sulfur standards, total onroad NO X emissions are projected to drop 10 percent. This is primarily due to large reductions from Tier 2 gasoline vehicles, which contribute about one-quarter of the NO X emissions from the on-road fleet in 2018. The relative reduction grows as cleaner vehicles turn over into the fleet. By 2030, we project that the reduction in overall onroad NO X inventory will be 25 percent.

Table III-2—Projected NO X Reductions From Tier 3 Program
20182030
Total reduction264,369328,509
Reduction from pre-Tier 3 fleet due to sulfur standard 242,434 56,324
Reduction from Tier 3 fleet due to vehicle and sulfur standards 21,934 272,185
Percent reduction in onroad NO X emissions10%25%
2. VOC

Table III-3 shows the reduction in VOC emissions, in annual short tons, projected in calendar years 2018 and 2030 resulting from the Tier 3 standards. In 2018, as with NO X, we project reductions from the pre-Tier 3 fleet with the fuel standards. By 2030, the reduction in overall onroad VOC emissions will be 16 percent, the majority of this from the vehicles meeting the fully phased-in Tier 3 standards. The evaporative standards are projected to account for roughly one third of the overall vehicle program reduction in 2030.

Table III-3—Projected VOC Reductions From Tier 3 Program
20182030
Total reduction47,504167,591
Reduction from pre-Tier 3 fleet due to sulfur standard38,78611,249
Reduction from Tier 3 fleet due to vehicle and sulfur standards8,718156,343
Exhaust 43,009 105,253
Evaporative 4,495 62,339
Percent reduction in onroad VOC emissions3%16%
3. CO

Table III-4 shows the reductions for CO, broken down by pre- and post-Tier 3 in the manner described for NO X and VOC above. In contrast to NO X and VOC, the immediate CO reductions in the onroad fleet from sulfur control in 2018 are small, based on research showing that fuel sulfur level has a minimal impact on CO emissions from Tier 2 vehicles. The CO exhaust standards are projected to reduce onroad CO emissions by 24 percent in 2030.

Table III-4—Projected CO Reductions From Tier 3 Program
20182030
Total reduction278,8793,458,041
Reduction from pre-Tier 3 fleet due to sulfur standard122,17117,734
Reduction from Tier 3 fleet due to vehicle and sulfur standards156,7083,440,307
Percent reduction in onroad CO emissions2%24%
4. Direct PM

Reductions in direct emissions of PM 2.5 are projected to result solely from the vehicle tailpipe standards, so meaningful reductions are realized mainly as the fleet turns over. By 2030, we project a reduction of about 7,900 tons annually, which represents approximately 10 percent of the onroad direct PM 2.5 inventory. The relative reduction in onroad emissions is projected to grow to 28 percent in 2050, when nearly all of the fleet will have turned over to vehicles meeting the fully phased-in Tier 3 standards. Reductions in NO X and VOC emissions will also reduce secondary PM formation, which is quantified as part of the air quality analysis described in Section III.C.

5. Air Toxics

Emissions of air toxics also will be reduced by the sulfur, exhaust and evaporative standards. Air toxics are generally a subset of compounds making up VOC, so the reduction trends tend to track the VOC reductions presented above, for most air toxics. Table III-5 presents reductions for certain key air toxics, and Table III-6 presents reductions for the sum of 71 different toxic compounds.

Table III-5—Reductions for Certain Individual Compounds
Tons reduced in 2018% Reduction in onroad emissionsTons reduced in 2030% Reduction in onroad emissions
Benzene1,91664,76226
Acetaldehyde60032,06721
Formaldehyde51321,27710
1,3-Butadiene257567729
Acrolein40312715
Naphthalene99326915
Ethanol2,704219,95016

The totals shown in Table III-6 represent the sum of 71 species including the toxics in Table III-5, 15 polycyclic aromatic hydrocarbon (PAH) compounds in gas and particle phase, and additional gaseous compounds such as toluene, xylenes, styrene, hexane, 2,2,4-trimethylpentane, n-hexane, and propionaldehyde (see Appendix 7A of the RIA). As shown, in 2030, the overall onroad inventory of total toxics will be reduced by 15 percent, with nearly one half of the vehicle program reductions coming from the evaporative standards.

Table III-6—Reductions in Total Mobile Source Air Toxics
20182030
Total reduction15,58364,558
Reduction from pre-Tier 3 fleet due to sulfur standard11,9813,517
Reduction from Tier 3 fleet due to vehicle and sulfur standards3,60261,041
Exhaust 13,340 34,595
Evaporative 2,243 29,963
Percent reduction in onroad toxics emissions3%15%
6. SO

SO 2 emissions from mobile sources are a direct function of sulfur in the fuel, and reducing sulfur in gasoline will result in immediate reductions in SO 2 from the on and off-road fleet. The reductions, shown in Table III-7, are a function of the sulfur level and fuel consumption. This is reflected in the relative contribution of on-road vehicles and off-road equipment, where off-road gasoline consumption accounts for approximately 5 percent of overall gasoline use. (216)

Table III-7—Projected SO 2 Reductions From Tier 3 Program
20182030
Total reduction15,56513,261
Reduction from onroad vehicles due to sulfur standard14,81312,399
Reduction from off-road equipment due to sulfur standard752862
Percent reduction in onroad SO 2 emissions56%56%
7. Greenhouse Gases

Reductions in nitrous oxide (N 2 O) emissions and methane (CH 4) emissions, both potent greenhouse gas emissions, are projected for gasoline cars and trucks as a result of the sulfur and tailpipe standards. A study conducted by the University of California-Riverside found a 29 percent reduction in N 2 O emissions over the FTP when sulfur was reduced from 30 to 5 ppm, (217) while EPA research described in Section IV.A.6 on sulfur effects found a 26 percent reduction in CH 4 emissions when sulfur was reduced from 28 to 5 ppm. (218)

Several studies have established correlations between reductions in tailpipe NO X emissions and reductions in N 2 O from gasoline cars and trucks, (219 220 221 222) as well as correlations between reductions in tailpipe HC emissions and reductions in CH 4. (223 224) Studies by Winer, et al (2005) and Behrentz et al (2004) reported N 2 O: NO X ratios of 0.06 and 0.095, respectively, and supported the application of N 2 O: NO X ratios to NO X emissions as a reasonable method for estimating N 2 O emission inventories. CARB has also used N 2 O: NO X ratio to develop the N 2 O emissions inventories for the LEV III program, based on a regression analysis suggesting N 2 O: NO X ratio of 0.04, on average. (225)

As detailed in Chapter 7.3 of the RIA, the N 2 O reductions are estimated by employing two different methodologies, resulting in a range of reductions. The first method applies the relationship between N 2 O and NO X from a regression model  (226) to NO X inventories from both Tier 3 and pre-Tier 3 vehicles. The second method applies the regression of N 2 O and NO X only to Tier 3 vehicles and uses the UC Riverside sulfur results to estimate the N 2 O reductions from pre-Tier 3 vehicles. Using a 100-year global warming potential of 298 for N 2 O according to the 2007 IPCC AR4, (227) the estimated N 2 O reduction is 2.2 million metric tons of carbon dioxide equivalent (MMTCO 2 e) in 2018, growing to the range between 3.8 to 4.0 MMTCO 2 e in 2030. For 2018, there was an agreement between the two methodologies described above, resulting in a single estimate. MOVES can be used to directly estimate CH 4 reductions from the sulfur and vehicle standards, estimating an additional 0.1 MMTCO 2 e reduction in 2018, growing to 0.3 MMTCO 2 e in 2030. The total GHG reduction from the Tier 3 rule is 2.3 MMTCO 2 e in 2018, and between 4.1 and 4.3 MMTCO 2 e in 2030.

These reductions will be partially offset by CO 2 emissions associated with higher energy use required in the process of removing sulfur within the refinery. As an extension of our refinery-by-refinery cost modeling described in Section VII.B., we calculated the CO 2 emission impacts of Tier 3 gasoline sulfur control. We estimated refinery-specific changes in process energy and then applied emission factors that correspond to those changes, on a refinery-by-refinery basis. As described in Chapter 4.5 of the RIA, the results showed an increase of up to 1.9 MMTCO 2 e in 2018 and 1.6 MMTCO 2 e in 2030 for all U.S. refineries complying with the lower sulfur standards assuming that the sulfur standards are fully phased-in. In 2018, the combined impact of CH 4 and N 2 O emission reductions from the vehicles and CO 2 emission increases from the refineries shows a slight net decrease on a CO 2 equivalent basis. While still small, this net decrease grows to a range between 2.5 to 2.7 MMTCO 2 e by 2030.

We do not expect the Tier 3 vehicle standards to result in any discernible changes in vehicle CO 2 emissions or fuel economy. Emissions of the pollutants that are controlled by the Tier 3 program—NMOG, NO X, and PM—are not a function of the amount of fuel consumed, since manufacturers need to design their catalytic emission control systems to reduce these emissions regardless of their engine-out levels.

C. How will air pollution be reduced?

Reductions in emissions of NO X, VOC, PM 2.5 and air toxics expected as a result of the Tier 3 standards are projected to lead to significant improvements in air quality. The air quality modeling predicts significant improvements in ozone concentrations due to the Tier 3 standards. Ambient PM 2.5 and NO 2 concentrations are also expected to improve as a result of the Tier 3 program. Decreases in ambient concentrations of air toxics are projected with the Tier 3 standards, including notable nationwide reductions in benzene concentrations. Our air quality modeling also predicts improvements in visibility and sulfur deposition, as well as substantial decreases in nitrogen deposition as a result of the Tier 3 standards. The results of our air quality modeling of the impacts of the Tier 3 rule are summarized in the following section.

1. Ozone

The air quality modeling done for this action projects that in 2018, with all current and required controls in effect but excluding the emissions changes expected to occur as a result of the Tier 3 standards or any other additional controls, at least 19 counties, with a projected population of over 37 million people, would have projected design values above the level of the 2008 8-hour ozone standard of 75 ppb. In 2030 the modeling projects that in the absence of Tier 3 standards or any other additional controls there will be 6 counties with a population of over 19 million people with projected design values above the level of the 2008 8-hour ozone standard of 75 ppb. An additional 37 million people will be living in the 43 counties that will be close to (within 10 percent of) the level of the ozone standard.

Air quality modeling indicates that this action will meaningfully decrease ozone design value concentrations in many areas of the country, including those that are projected to be exceeding, or close to exceeding, the ozone standard. In 2018, the majority of the design value decreases are between 0.5 and 1.0 ppb. In 2030, the Tier 3 rule will result in larger decreases in ozone design values, with the majority of counties projecting decreases of between 0.5 and 1.0 ppb, and over 250 more counties with decreases greater than 1.0 ppb. Since the Tier 3 standards go into effect during the period when some areas are still working to attain the ozone NAAQS, the projected air quality changes will help state and local agencies in their effort to attain and maintain the ozone standard.

2. Particulate Matter

The air quality modeling conducted for this action projects that in 2018, with all current controls in effect but excluding the emissions changes expected to occur as a result of Tier 3 standards or any other additional controls, at least 14 counties, with a projected population of over 20 million people, would have projected design values above the level of the annual standard of 12 μg/m 3 and at least 24 counties, with a projected population of over 18 million people, would have projected design values above the level of the 24-hour standard of 35 μg/m 3. In 2030, the modeling projects that in the absence of Tier 3 standards or any other additional controls there will be 13 counties, with a projected population of over 21 million people, with projected design values above the level of the annual standard of 12 μg/m 3 and 18 counties, with a projected population of over 12 million people, with projected design values above the level of the 24-hour standard of 35 μg/m 3. Since the Tier 3 standards go into effect during the period when some areas are still working to attain the 2006 and 2012 PM 2.5 NAAQS, the projected air quality changes will be useful to state and local agencies in their effort to attain and maintain the PM 2.5 standards.

The Tier 3 standards will reduce 24-hour and annual PM 2.5 design values due to projected tailpipe reductions in primary PM 2.5, SO 2, NO X and VOCs from reductions in fuel sulfur and engine controls. In 2018 the standards will have a small impact on annual PM 2.5 design values in the majority of modeled counties. However, in over 200 counties annual PM 2.5 design values are projected to decrease by greater than 0.01 μg/m 3. In 2030 annual PM 2.5 design values in the majority of modeled counties will decrease by between 0.01 and 0.05 μg/m 3 and in over 140 additional counties design values are projected to decrease by greater than 0.05 μg/m 3. In addition, in 2018 24-hour PM 2.5 design values in over 200 counties are projected to decrease by between 0.05 and 0.15 μg/m 3 and in 2030 24-hour PM 2.5 design values in over 180 counties decrease by at least 0.15 μg/m 3.

3. Nitrogen Dioxide

Although our modeling indicates that by 2030 the majority of the country will experience decreases of less than 0.1 ppb in their annual NO 2 concentrations due to this rule, annual NO 2 concentrations are projected to decrease by more than 0.3 ppb in most urban areas. These emissions reductions would also likely decrease 1-hour NO 2 concentrations and help any potential nonattainment areas to attain and maintain the standard. Additional information on the emissions reductions that are projected with this rule is available in Section 7.2.1 of the RIA.

4. Air Toxics

Our modeling indicates that the impacts of final Tier 3 standards include notable nationwide reductions in benzene and generally small decreases in ambient concentrations of other air toxics, mainly in urban areas. Although reductions are greater in 2030 (when 70 percent of the miles travelled are from vehicles that meet the fully phased-in Tier 3 standards) than in 2017 (the first year of the final program), our modeling projects there will be small immediate reductions in ambient concentrations of air toxics due to the Tier 3 sulfur controls. Furthermore, the full reduction of the vehicle program will be realized after 2030, when the fleet has fully turned over to vehicles meeting the fully phased-in Tier 3 standards. Air toxics pollutants dominated by primary emissions (or a decay product of a directly emitted pollutant), such as benzene, are impacted more than air toxics that primarily result from photochemical transformation.

Specifically, in 2030, our modeling projects that the Tier 3 rule will decrease ambient benzene concentrations across much of the country on the order of 1 to 5 percent, with reductions ranging from 10 to 25 percent in some urban areas. Our modeling also shows reductions of 1,3-butadiene and acrolein concentrations in 2030 ranging between 1 and 25 percent and 1 and 10 percent respectively, with 1,3-butadiene decreases of at least 0.005 μg/m 3 in urban areas. These toxics are national risk drivers and the reductions in ambient concentrations from this rule will result in reductions in risks from cancer and noncancer health effects. In some parts of the country (mainly urban areas), ethanol and formaldehyde concentrations are projected to decrease on the order of 1 to 10 percent and 1 to 2.5 percent respectively in 2030 as a result of the Tier 3 rule. Decreases in ethanol concentrations are expected due to reductions in VOC as a result of the Tier 3 standards. Changes in ambient acetaldehyde concentrations are generally less than 1 percent across the U.S., although the Tier 3 rule may decrease acetaldehyde concentrations in some urban areas by 1 to 2.5 percent in 2030. Changes in ambient naphthalene concentrations are generally between 1 and 10 percent in 2030 with absolute decreases of up to 0.005 μg/m 3.

Although the reductions in ambient air toxics concentrations expected from the Tier 3 standards are generally small, they are projected to benefit the majority of the U.S. population. As shown in Table III-8, over 75 percent of the total U.S. population is projected to experience a decrease in ambient benzene and 1,3-butadiene concentrations of at least 1 percent. Over 60 percent of the U.S population is projected to experience at least a 1 percent decrease in ambient ethanol and acrolein concentrations, and over 35 percent would experience a similar decrease in ambient formaldehyde concentrations with the Tier 3 standards.

Table III-8—Percent of Total Population Experiencing Changes in Annual Ambient Concentrations of Toxic Pollutants in 2030 as a Result of the Tier 3 Standards
Percent change(percent)Benzene(percent)Acrolein(percent)1,3-Butadiene(percent)Formaldehyde(percent)Ethanol(percent)Acetaldehyde(percent)Naphthalene(percent)
≤−50
>−50 to ≤−25
>−25 to ≤−102.290.7519.0710.74
>−10 to ≤−520.6312.7227.295.3931.56
>−5 to ≤−2.527.5025.1715.370.6024.0820.58
>−2.5 to ≤−128.6024.6218.3335.3434.1011.7714.98
>−1 to <120.9736.7419.9364.0636.4388.2322.14
≥1 to <2.5
≥2.5 to <5
≥5 to <10
≥10 to <25
≥25 to <50
≥50

In addition, as described in Section 7.2.4.4.2 of the RIA, our modeling projects that acrolein concentrations would decrease to levels below the inhalation reference concentration for acrolein (0.02 μg/m 3) for over 5 million people in 2030, meaning that as a result of the Tier 3 standards, 5 million fewer Americans will be exposed to ambient levels of acrolein high enough to present a potential for adverse health effects.

5. Visibility

Air quality modeling conducted for this final action was used to project visibility conditions in 137 mandatory class I federal areas across the U.S. The results show that in 2030 all the modeled areas will continue to have annual average deciview levels above background and the Tier 3 rule will improve visibility in all these areas. (228) The average visibility at all modeled mandatory class I federal areas on the 20 percent worst days is projected to improve by 0.02 deciviews, or 0.16 percent, in 2030. Section 7.2.5.5 of the RIA contains more detail on the visibility portion of the air quality modeling.

6. Nitrogen and Sulfur Deposition

Our air quality modeling projects substantial decreases in nitrogen deposition as a result of the Tier 3 standards. The standards will result in annual percent decreases of greater than 2.5 percent in most major urban areas and greater than 5 percent in a few areas. In addition, smaller decreases, in the 1 to 2.5 percent range, will occur over much of the rest of the country. The impacts of the Tier 3 standards on sulfur deposition are smaller, ranging from no change to decreases of over 2.5 percent in some areas. For maps of 2030 deposition impacts and additional information on these impacts see Section 7.2.5.6 of the RIA.

7. Environmental Justice

Environmental justice (EJ) is a principle asserting that all people deserve fair treatment and meaningful involvement with respect to environmental laws, regulations, and policies. EPA seeks to provide the same degree of protection from environmental health hazards for all people. As referenced below, numerous studies have found that some environmental hazards are more prevalent in areas with high population fractions of racial/ethnic minorities and people with low socioeconomic status (SES), as would be expected on the basis of those areas' share of the general population.

As discussed in Section II of this document, concentrations of many air pollutants are elevated near high-traffic roadways. If minority populations and low-income populations disproportionately live near such roads, then an issue of EJ may be present. Such disparities may be due to multiple factors. (229)

People with low SES often live in neighborhoods with multiple stressors and health risk factors, including reduced health insurance coverage rates, higher smoking and drug use rates, limited access to fresh food, visible neighborhood violence, and elevated rates of obesity and some diseases such as asthma, diabetes, and ischemic heart disease. Although questions remain, several studies find stronger associations between air pollution and health in locations with such chronic neighborhood stress, suggesting that populations in these areas may be more susceptible to the effects of air pollution. (230 231 232 233) Household-level stressors such as parental smoking and relationship stress also may increase susceptibility to the adverse effects of air pollution. (234 235)

To address the existing conditions in areas near major roadways, in comparison with other locations, we reviewed existing scholarly literature examining the topic, and conducted our own evaluation of two national datasets: The U.S. Census Bureau's American Housing Survey for calendar year 2009 and the U.S. Department of Education's database of school locations.

Existing publications that address EJ issues generally report that populations living near major roadways (and other types of transportation infrastructure) tend to be composed of larger fractions of nonwhite residents. People living in neighborhoods near such sources of air pollution also tend to be lower in income than people living elsewhere. Numerous studies evaluating the demographics and socioeconomic status of populations or schools near roadways have found that they include a greater percentage of minority residents, as well as lower SES (indicated by variables such as median household income). Locations in these studies include Los Angeles, CA; Seattle, WA; Wayne County, MI; Orange County, FL; and the State of California (236 237 238 239 240 241)

More recently, three publications report nationwide analyses that compare the demographic patterns of people who do or do not live near major roadways. (242 243 244) All three of these studies found that people living near major roadways are more likely to be minorities or low in SES. They also found that the outcomes of their analyses varied between regions within the U.S. However, only one such study looked at whether such conclusions were confounded by living in a location with higher population density and how demographics differ between locations nationwide. In general, it found that higher density areas have higher proportions of low income and minority residents.

We analyzed two national databases that allowed us to evaluate whether homes and schools were located near a major road. One database, the American Housing Survey (AHS), includes descriptive statistics of over 70,000 housing units across the nation. The study is conducted every two years by the U.S. Census Bureau. We analyzed data from the 2009 AHS. The second database we analyzed was the U.S. Department of Education's Common Core of Data, which includes enrollment and location information for schools across the U.S.

In analyzing the 2009 AHS, we focused on whether or not a housing unit was located within 300 feet of “4-or-more lane highway, railroad, or airport.”  (245) We analyzed whether there were differences between houses and householders in such locations and those not in them. (246) We included other variables, such as land use category, region of country, and housing type. We found that homes with a nonwhite householder were 22-34 percent more likely to be located within 300 feet of these large transportation facilities, while homes with a Hispanic householder were 17-33 percent more likely. Households near large transportation facilities were, on average, lower in income and educational attainment, more likely to be a rental property and located in an urban area.

In examining schools near major roadways, we examined the Common Core of Data (CCD) from the U.S. Department of Education, which includes information on all public elementary and secondary schools and school districts nationwide. (247) To determine school proximities to major roadways, we used a geographic information system (GIS) to map each school and roadways based on the U.S. Census's TIGER roadway file. (248) We found that minority students were overrepresented at schools within 200 meters of the largest roadways, and that schools within 200 meters of the largest roadways also had higher than expected numbers of students eligible for free or reduced-price lunches. For example, Black students represent 21.57 percent of students at schools located within 200 meters of a primary road, whereas Black students represent 16.62 percent of students in all U.S. schools. Hispanic students represent 30.13 percent of students at schools located within 200 meters of a primary road, whereas Hispanic students represent 21.93 percent of students in all U.S. schools.

Overall, there is substantial evidence that people who live or attend school near major roadways are more likely to be of a minority race, Hispanic ethnicity, and/or low SES. The emission reductions from this rule are projected to result in widespread air quality improvements, but the impact on pollution levels in close proximity to roadways is expected to be most direct. Thus, this rule is likely to help in mitigating the disparity in racial, ethnic, and economically-based exposures.

IV. Vehicle Emissions Program

In the 14 years since EPA finalized the Tier 2 Vehicle Program, manufacturers of light-duty vehicles have continued to develop a wide range of improved technologies capable of reducing emissions, especially exhaust hydrocarbons, nitrogen oxides (NO X), and particulate matter (PM), and evaporative hydrocarbons. The California LEV II program has been instrumental in the auto industry's continuous technology improvements by requiring year after year reductions in fleet average exhaust hydrocarbon levels. In addition, California set performance standards that have resulted in the introduction of advanced exhaust and evaporative emission controls in partial zero emission vehicles (PZEVs). Overall, this progress in vehicle technology has made it possible for manufacturers to achieve emission reductions with a number of today's vehicles that go well beyond the requirements of the Tier 2 program.

Extensive data from existing Tier 2 (and California LEV II) vehicles presented in the NPRM and received since the proposal have demonstrated the potential for further significant reductions. For exhaust emissions, these opportunities include addressing: Emissions produced at start-up; emissions under high-speed, high-load conditions; the effects of sulfur in gasoline; the effects of increased oil consumption; and the effects of age on vehicles and control systems. In addition, technologies now exist that have inherently low evaporative emission characteristics and demonstrate improved in-use durability. Based on this body of data, we are adopting more stringent standards designed to reduce emissions, primarily exhaust non-methane organic gases (NMOG), NO X, and PM and evaporative hydrocarbon emissions from new vehicles. As discussed in detail below and in the final RIA, we have concluded that, in conjunction with the reductions in fuel sulfur also required in this action, the new vehicle emissions standards are feasible, accounting for costs, across the fleet in the timeframe of the program. We believe that simultaneous reductions in fuel sulfur will be a key factor in enabling the entire fleet of vehicles subject to Tier 3 to meet the new emission standards in-use, throughout the life of the vehicles (see Section IV.A.6 below).

We received a large number and wide range of comments on the proposed vehicle emission program, and we have carefully considered all of them. (The Summary and Analysis of Comments document addresses the comments received; it is located in the docket for this rulemaking and also on EPA's Web site at .) With very few exceptions, we are finalizing the Tier 3 vehicle emission program as proposed, including the levels of the new emission standards and the phase-in schedules. In several cases, as discussed in detail below, the comments and/or newer technical information have resulted in adjustments to the proposed program, including when the requirements begin, what fuel is used for vehicle compliance testing, and what the PM standard level is for testing under aggressive driving conditions. The final Tier 3 vehicle provisions, like the proposal, also harmonize closely with California's LEV III program.

This section describes in detail the program for reducing tailpipe and evaporative emissions from light-duty vehicles (LDVs, or passenger cars), light-duty trucks (LDT1s, 2s, 3s, and 4s), Medium-Duty Passenger Vehicles (MDPVs), and certain heavy-duty vehicles (HDVs). Sections IV.A and IV.B discuss the tailpipe emission standards and time lines, and other provisions for new LDVs, LDTs, and MDPVs and for new heavy-duty vehicles up to 14,000 lbs Gross Vehicle Weight Rating (GVWR). Section IV.C presents the new Tier 3 evaporative emissions standards and program and Section IV.D describes the new evaporative emissions leak test. Section IV.E presents improvements to the existing Onboard Diagnostics (OBD) provisions. In Section IV.F, we describe new provisions to update our federal certification fuel to better match today's in-use fuel. We also discuss in this section the compliance flexibilities for small auto manufacturing companies and small-volume manufacturers (IV.G) as well as new testing and test procedure provisions and other compliance provisions (IV.H).

A. Tier 3 Tailpipe Emission Standards for Light-Duty Vehicles, Light-Duty Trucks, and Medium-Duty Passenger Vehicles

1. How the Tier 3 Program Is Harmonized With the California LEV III Program

In describing the Tier 3 program for light- and heavy-duty vehicles in this preamble, we discuss how the provisions are consistent with the California Air Resources Board (CARB) LEV III program. (249) During the development of the proposed rule and in their comments, auto manufacturers stressed to us the importance of their being able to design and produce a single fleet of vehicles for all 50 states that simultaneously complies with requirements under the Tier 3 program and the LEV III program, as well as greenhouse gas/CAFE requirements they are facing in the same timeframe. To the extent that the federal and California programs are consistent, special versions of vehicles with different emission control hardware and calibrations for different geographic areas will be unnecessary. This will allow manufacturers to avoid the additional costs of parallel design, development, calibration, and manufacturing. Consistency among programs also eliminates the need to supply aftermarket parts for repair of multiple versions of a vehicle. We believe that the most effective and efficient national program will result from close coordination between CARB LEV III and federal Tier 3 program elements and their implementation.

To that end, we worked closely with CARB and the vehicle manufacturers, the latter both individually and through their trade associations, to align the two programs. The Tier 3 program is identical to LEV III in most major respects for light-duty vehicles (and heavy-duty vehicles, as described in sections below). The levels and the timing of the declining fleet-average NMOG+NO X standards are identical to those in LEV III. The Tier 3 emissions bins to which manufacturers will certify individual vehicle models in order to comply with the fleet-average standards, are also identical to those in LEV III. Similarly, the light-duty Tier 3 FTP PM standards and percent phase-in match those for LEV III through MY 2024.

We note there are a few light-duty Tier 3 and LEV III provisions that are different, for reasons discussed below. For example, the LEV III program and the Tier 3 program have different light-duty PM requirements late in the program (i.e., after MY 2024 (IV.A.3.b.)), and the two programs have different final NMOG+NO X standards for small volume manufacturers (IV.G.1). As also discussed below, we are finalizing a revised SFTP (US06) PM standard, and CARB has commented that it plans to take similar action in near future. CARB also indicated in their comments that they intend to consider several additional actions to further align several minor aspects of LEV III with the Tier 3 program once Tier 3 is finalized.

Beyond the provisions mentioned above, the differences between the programs are not major and most will exist only in the transitional years of the Tier 3 program. These additional differences result from the fact that the LEV III requirements begin slightly earlier and that a limited phase-in of some provisions is necessary for a smooth transition to overall aligned programs. These temporary differences include the process for how early compliance credits are generated and used (e.g., Section IV.A.7.a); how quickly manufacturers will need to move toward certifying all of their vehicle models to longer useful-life values (e.g., Section IV.A.7.c) and on the new test fuel (e.g., Section IV.A.7.d); and transitional emissions bins to facilitate the transition from Tier 2 to Tier 3 (Section IV.A.7.n).

2. Summary of the Tier 3 FTP and SFTP Tailpipe Standards
a. Major Comments on and Significant Changes to the Proposal

As mentioned above, we are finalizing most aspects of the comprehensive Tier 3 vehicle program as we proposed them. The levels of the FTP and SFTP standards for the key tailpipe pollutants of concern—the sum of NMOG and NO X emissions, expressed as NMOG+NO X, and PM—are the same as proposed (except for the numerically lower final PM SFTP (US06) standard, as discussed below). In addition, the timing of the requirements remains the same as in the NPRM, starting with MY2017 and MY2018 and phasing in according to the same declining fleet-average schedule for the NMOG+NO X standards and the same percent-of-sales phase-in schedule for the PM standards. We continue to believe that these elements form a robust framework for the Tier 3 vehicle program and closely harmonize with the respective elements of California's LEV III program.

There are several important provisions of the light-duty Tier 3 program that we have revised from the proposal, based on further consideration and information that we received from commenters. We discuss each of these in detail later in this section and summarize them here.

  • As described below in Section IV.A.2.c, each of the four primary Tier 3 emission standards has an associated alternative phase-in option for heavier light-duty vehicles that a manufacturer can choose if it prefers a later start date (to provide 4 years of lead time) and a stable standard. (250) We proposed that a manufacturer choosing these options be required to apply the alternative phase-in schedule to its entire light-duty fleet. In response to comments from automakers that this restriction would be unnecessarily burdensome, we reconsidered this provision. For the reasons discussed below, we are allowing a manufacturer to apply the alternative phase-in schedules to only their heavier light-duty vehicles, instead of their entire light-duty fleet. However, manufacturers have largely indicated that they plan on adopting the primary program which is harmonized with LEV III.
  • This Tier 3 rule provides an opportunity for EPA to reassess the degree to which the gasoline used for vehicle emissions testing and certification reflects in-use gasoline around the country. In the case of ethanol content, we proposed that the emissions test fuel contain 15 percent ethanol (E15), anticipating a significant shift to higher ethanol content in use in the near future. For several reasons described below (Section IV.F.1), this shift in in-use fuel is not materializing as quickly as expected, and E10 continues to be almost universal today. We received a near consensus among comments from stakeholders that E10 test fuel is more appropriate. We agree that E10 most appropriately reflects in-use gasoline around the country today and into the foreseeable future, and thus we are finalizing E10 for the test fuel. In addition, as discussed in Section IV.F.1, we are finalizing a fuel volatility specification for test fuel of 9 psi RVP, as proposed.
  • We are finalizing a set of standards for PM as measured on the aggressive-driving segment of the SFTP test cycle (the US06 cycle) based on US06 PM test data that we published as part of the NPRM, along with more recent test data developed by California. Our review of these data has led us to finalize numerically lower levels for the US06 PM standards than we proposed. The data presented in the NPRM as well as the data provided by California clearly show that the proposed US06 PM standards were inappropriately high, that US06 PM emissions are not closely related to vehicle weight, and that lower values for the standards would achieve the goal of the program to bring all vehicles in the light-duty fleet to the US06 PM levels that are being met by many vehicles today. Based on the body of available data, we are establishing 6 mg/mi as the long-term US06 PM standard. (This compares to the proposed standards of 10 and 20 mg/mi for lighter and heavier light-duty vehicles, respectively.) However, because there remains some uncertainty about how manufacturers will achieve this level in the early years of the program, we are setting the standard at 10 mg/mi for the early years of the program, for MYs 2017 and 2018. Similarly, we are providing a less-stringent standard of 10 mg/mi for testing of in-use vehicles in recognition of the challenges of the requirements as vehicles age.
  • In the Tier 3 program, as for vehicle emission control programs in the past, manufacturers are responsible for the emissions performance of the vehicle for a specified “useful life” of the vehicle. EPA proposed that vehicles meet the Tier 3 standards for 150,000 miles or 15 years, identical to the LEV III program's approach. We proposed an option for lighter light-duty vehicles to certify to a shorter useful life of 120,000 miles or 10 (or 11, as applicable) years, as set in the Tier 2 program. We proposed that manufacturers certifying to the shorter useful life would need to meet numerically lower NMOG+NO X standards (85 percent of the respective 150,000-mile NMOG+NO X standards). We also proposed that a manufacturer choosing the shorter useful life for one vehicle model would need to use that useful life and associated standards for all of its lighter vehicles. Auto industry commenters stated that applying the provision across a manufacturer's fleet would create an onerous compliance burden. We have reconsidered our proposed approach, and as discussed in Section IV.A.7.c below, we will allow a manufacturer to split its lighter light-duty fleet among models certified for either the 150,000 mile or 120,000 mile useful life and associated standards.
  • Another area of substantial comment, primarily from the petroleum refining industry, questioned the technological need of auto manufacturers for lower in-use sulfur levels in order to meet the Tier 3 vehicle emission standards. In contrast, auto manufacturers and emissions control system manufacturers commented that lower sulfur gasoline is critical to meet the Tier 3 standards. After careful consideration of the comments, we continue to believe that the large body of data presented in the NPRM, supplemented by newer data that consistently reinforces the earlier conclusions, strongly supports our determination of the need for average in-use gasoline sulfur levels to be at 10 ppm sulfur or lower for manufacturers to meet the Tier 3 vehicle standards across their fleets for the useful life of the vehicles. See Section IV.A.6 below for a detailed discussion of the need for gasoline sulfur control.
b. Structure of the Primary Tier 3 Tailpipe Standards

As proposed, compliance with the standards is based on vehicle testing using test procedures that represent a range of vehicle operation, including the Federal Test Procedure (FTP) and the Supplemental Federal Test Procedure (SFTP). The Tier 3 FTP and SFTP NMOG+NO X standards are fleet-average standards, meaning that the manufacturer calculates the sales-weighted average emissions of the vehicles it sells in each model year, accounting for any Tier 3 emissions credits or deficits, and compares that average to the applicable standard for that model year. The fleet average standards for NMOG+NO X evaluated over the FTP are the same values as proposed and are summarized in Table IV-2 and discussed in detail below. For lighter light-duty vehicles, the standards begin in MY 2017 at a level representing a 46 percent reduction from the current Tier 2 requirements for lighter vehicles and then become increasingly stringent, culminating in an 81 percent reduction in MY 2025. The FTP NMOG+NO X program includes separate fleet average standards for heavier vehicles that begin in MY 2018 and then converge with the standards for lighter vehicles at 30 milligrams per mile (mg/mi) in MY 2025 and later, as proposed. (251 252)

Manufacturers will determine their fleet average FTP NMOG+NO X emission values as we proposed, based on the per-vehicle “bin standards” to which they certify each vehicle model. Manufacturers will be free to certify vehicles to any of the bins, so long as the sales-weighted average of the NMOG+NO X values from the selected bins meets the fleet average standard for that model year. Table IV-1 presents the per-vehicle bin standards. Similarly, the fleet average NMOG+NO X standards measured over the SFTP are summarized in Table IV-4 and discussed in detail below. The SFTP NMOG+NO X fleet average standards decline from MY 2017 until MY 2025. In this case, the same standards apply to both lighter and heavier vehicles. In MY 2025, the SFTP NMOG+NO X standard reaches its fully phased-in fleet average level of 50 mg/mi.

Also as proposed, the new Tier 3 PM standards apply to each vehicle separately. The PM standards are per-vehicle cap standards and not fleet-average standards. Also, in contrast to the declining NMOG+NO X standards, the PM standard on the FTP is a constant 3 mg/mi for all vehicles and for all model years, phasing in to an increasing percentage of vehicle sales beginning in MY 2017 for vehicles at or below 6,000 lbs Gross Vehicle Weight Rating (GVWR) and in MY 2018 for vehicles above 6,000 lbs GVWR. As discussed in Section IV.A.3.b above, based on data generated by EPA and CARB test programs, most current light-duty vehicles are already performing at or below the 3 mg/mi level. However, some vehicles are emitting above this level, due to such factors as excessive fueling during cold start and combustion chamber and fuel system designs that are not optimized for low PM emissions. The intent of the 3 mg/mi standard is to bring all light-duty vehicles to the PM level typical of that being demonstrated by most light-duty vehicles today. To address the uncertainties that will accompany the introduction of new technologies, the program includes a separate in-use FTP PM standard of 6 mg/mi for the testing of in-use vehicles during the phase-in period, as proposed, as described in more detail below.

As presented in Table IV-3, for vehicles at or below 6000 lbs GVWR, these FTP PM certification and in-use standards phase in over several years, beginning with a requirement that at least 20 percent of a company's U.S. sales of these vehicles comply with the Tier 3 standards in MY 2017. We are also finalizing an option for a manufacturer to choose to certify 10 percent of its total light-duty fleet sales—including LDVs and LDT over 6,000 lbs GVWR and MDPVs—to the Tier 3 FTP PM standards in MY 2017. Manufacturers would reach a 100 percent compliance requirement in MY 2021.

Finally, the Tier 3 program includes PM standards evaluated over the US06 cycle (a component of the SFTP test that captures higher speeds and accelerations). Based on emissions test data presented in the NPRM and additional data submitted in public comments, and as presented in Table IV-5 and further discussed in Section IV.A.4.b below, we are establishing a single long-term US06 PM standard of 6 mg/mi for both lighter and heavier vehicles, a level that is numerically lower than what we proposed. However, because there remains some uncertainty about how manufacturers will decide to achieve this level in the early years of the program, we are setting the standard through MY 2018 at 10 mg/mi. The US06 PM standards phase in using the same 20-20-40-70-100 percent schedule, and on the same vehicles, as the new FTP PM standards. The 10 mg/mi standard applies in MYs 2017 and 2018 (at a percent-of-sales requirement of 20 percent, and the long-term 6 mg/mi standard applies in MYs 2019 and later, increasing from 40 to 100 percent of sales. This US06 standard will apply to the same vehicle models that a manufacturer chooses to certify to the FTP PM standard during the percent phase-in period. As in the case of the FTP PM standards, the intent of the standard is to bring the emission performance of all vehicles to that already being demonstrated by many vehicles in the current light-duty fleet. As proposed, we include a separate in-use US06 PM standard during in the middle years of the program, but at a different numerical level and during different years than proposed (as discussed in Section IV.A.4.b below).

We did not propose new emission requirements for any vehicle or fuel over the cold temperature test cycles (i.e., the 20 °F cold carbon monoxide (CO) and non-methane hydrocarbon (NMHC) tests), but requested comment on that decision. Only the automakers commented on this topic, agreeing with EPA's approach of not changing its cold temperature requirements. As indicated in the proposal, we are not establishing any new cold temperature requirements in this rule.

c. Alternate Phase-In Schedules

For heavier light-duty vehicles (i.e., LDVs and LDTs greater than 6,000 lbs GVWR, plus MDPVs), EPA is also finalizing alternative phase-in schedules for each of the four primary vehicle emission standards: FTP NMOG+NO X, FTP PM, SFTP NMOG+NO X, and US06 PM. (253) These alternative phase-ins are available if a manufacturer prefers stable standards and four full years of lead time, as specified in the Clean Air Act for heavier vehicles. We describe each of the alternative phase-ins in more detail below, including several ways in which we have revised the proposed provisions.

EPA received comment on the proposed alternative phase-in provisions, primarily from automakers and their trade associations. These comments questioned whether the proposed structure of and restrictions on the use of the alternative phase-ins were so onerous as to unduly restrict a manufacturer from choosing the alternative phase-ins and their lead time and stability provisions as set forth in the Clean Air Act. The commenters criticized the proposed requirement that a manufacturer using the alternative phase-ins apply the alternative schedules to its entire light-duty fleet, both below and above 6,000 lbs GVWR. EPA had proposed this provision to minimize the complexity of complying with the alternative phase-in if a manufacturer's heavier and lighter light-duty vehicles had different compliance structures.

In consideration of these concerns, we have removed from the alternative phase-in provisions the requirement that a manufacturer apply the alternative schedules to its entire light-duty fleet including vehicles below 6,000 lbs GVWR. For the practical functioning of the program, the final rule requires that any manufacturer choosing to use the alternative phase-in apply all four alternative phase-in schedules to its entire light-duty fleet above 6,000 lbs GVWR. We believe that the alternative phase-ins allow manufacturers to comply with emission standards in a time frame that is clearly feasible and fully compliant with the CAA requirements for lead time and regulatory stability. To the extent that manufacturers choose to use them, the alternative would result in overall emission reductions essentially identical to those of the primary program.

The alternative phase-in schedules would begin to apply to each vehicle for either MY 2019 or MY 2020, depending on exactly when the manufacturer begins production of the vehicle. (See Section 86.1811-17(b)(8)(i) for how we implement this provision.) For models that begin MY 2019 production after the fourth anniversary of the signing of this final rule, the alternative phase-in would provide four full years of lead time and would first apply for MY 2019. The phase-in obligation would be calculated based only on those vehicles beginning production after the fourth anniversary date. For models beginning production before that date, the alternative phase-in would first apply for MY 2020, and the phase-in percentage for MY 2020 would be based on the manufacturer's entire fleet of heavier light-duty vehicles. Based on historical certification patterns, few models begin production before mid-calendar-year, so we expect that the vast majority of MY 2019 vehicles will begin production after the 4-year anniversary and thus the alternative phase-ins, if chosen, will typically apply beginning in MY 2019.

At the time of certification for MY 2018, a manufacturer must declare whether it intends to apply the alternative phase-in schedules to its heavier light-duty vehicles. A manufacturer choosing the alternative phase-ins would be committed to this phase-in approach for the duration of the phase-ins, and could not later choose the fleet-average approach for NMOG+NO X standards. For all vehicles below 6,000 lbs GVWR, the primary program will apply, beginning in MY 2017. For a manufacture's vehicles subject to the alternative phase-ins, there would be no new tailpipe emissions requirements beyond the Tier 2 program until the beginning of the alternative phase-in schedules; that is, MY 2019 or 2020, as explained above.

As discussed above, a manufacturer choosing the alternative phase-in approach for its heavier light-duty vehicles would be required to use all four phase-ins together. The next paragraphs explain how each of the alternative phase-ins requires an increasing percent of the manufacturer's sales to comply with the alternative standards. Thus, until the end of the phase-ins, some percent of a manufacturer's affected vehicles will meet the new standard and the remainder of that year's sales will not yet comply with Tier 3. For the practical functioning of the program, a manufacturer choosing the alternative phase-ins would be required to comply with exactly the same segment of their fleet in each model year for all four alternative phase-ins. For example, a manufacturer that complies with the 70 percent MY 2020 requirement for the FTP NMOG+NO X standard with a segment of its vehicle fleet must meet the 70 percent MY 2020 requirement for the FTP PM standard with the same set of vehicles. Vehicles covered by the alternative phase-in programs would be considered “Final Tier 3” vehicles and thus would also need to comply with the Tier 3 certification fuel and full useful life provisions.

For the FTP and SFTP NMOG+NO X alternative phase-in schedules, once the phase-in is complete for a segment of a manufacturer's fleet, the standards continue for that set of vehicles through MY 2024, after which the full Tier 3 program applies regardless of the phase-in strategy. Thus, the fleet-average standards that decline through MY 2024 do not apply for these vehicles.

Although manufacturers would implement all four alternative phase-in schedules together, as discussed above, each alternative phase-in has unique characteristics. The following paragraphs explain the unique provisions of each.

(1) Alternative Phase-In Schedule for the FTP NMOG+NO

Instead of the primary FTP NMOG+NO X declining fleet average standards, a manufacturer choosing the alternative phase-ins would comply with a stable fleet average FTP NMOG+NO X standard of 30 mg/mi that would apply to an increasing percentage of a manufacturer's combined sales of LDVs and LDTs above 6,000 lbs GVWR and MDPVs. This percent phase-in would match the percentages in the primary PM percent phase-in schedule, as discussed above—specifically, 40 percent of MY 2019 heavier light-duty vehicles (excluding those vehicles with production beginning before the 4-year anniversary), 70 percent of all of its heavier light-duty vehicles in MY 2020, and 100 percent compliance in MY 2021 and later model years.

(2) Alternative Phase-In Schedule for the FTP PM Standard

Instead of the primary FTP PM percent phase-in schedule, a manufacturer choosing the alternative phase-ins would postpone the beginning of its FTP PM phase-in for its LDVs and LDTs above 6,000 lbs GVWR and MDPVs until MY 2019 or 2020 (depending on the dates production begins for its vehicle models, as discussed above). The manufacturer would then comply with the 3 mg/mi per-vehicle FTP PM standard (and the 6 mg/mi in-use standard) on an increasing percentage of these vehicles, following the 40-70-100 percentage phase-in of the primary PM program—specifically, 40 percent of MY 2019 heavier light-duty vehicles (excluding those vehicles with production beginning before the 4-year anniversary), 70 percent of all of its heavier light-duty vehicles in MY 2020, and 100 percent compliance in MY 2021 and later model years.

(3) Alternative Phase-In Schedule for the SFTP NMOG+NO

As with the other alternative phase-ins, instead of the primary SFTP NMOG+NO X declining fleet average standards, a manufacturer choosing the alternative phase-ins would comply with a stable fleet average SFTP NMOG+NO X standard of 50 mg/mi that would apply to an increasing percentage of a manufacturer's combined sales of LDVs and LDTs above 6000 lbs GVWR and MDPVs. This percent phase-in again would match the percentages in the primary PM percent phase-in schedule, as discussed above—specifically, 40 percent of MY 2019 heavier light-duty vehicles (excluding those vehicles with production beginning before the 4-year anniversary), 70 percent of all of its heavier light-duty vehicles in MY 2020, and 100 percent compliance in MY 2021 and later model years.

(4) Alternative Phase-In Schedule for the US06 PM Standard

Finally, instead of the primary US06 PM percent phase-in schedule, a manufacturer choosing the alternative phase-ins would postpone the beginning of the US06 phase-in for its LDVs and LDTs above 6,000 lbs GVWR and MDPVs until MY 2019 or 2020 (depending on the dates production begins for its vehicle models, as discussed above). The manufacturer would then comply with the 10 mg/mi US06 PM standard for 40 percent of MY 2019 heavier light-duty vehicles (excluding those vehicles with production beginning before the 4-year anniversary), 70 percent of all of its heavier light-duty vehicles in MY 2020, with 100 percent compliance in MY 2021, and then 100 percent compliance with the 6 mg/mi standard in MY 2022 and later model years.

The next sections describe in more detail the new Tier 3 standards, how they will be implemented over time, and the technological approaches that we believe are or will be available to manufacturers in order to comply.

3. FTP Standards

As summarized above, we are finalizing, largely as proposed, new standards for the primary pollutants of concern for this rule (NMOG, NO X, and PM) as measured on the FTP. The following paragraphs describe in more detail these FTP standards for NMOG+NO X and PM, as well as for carbon monoxide (CO) and formaldehyde (HCHO).

a. FTP NMOG+NO

The Tier 3 NMOG and NO X standards are expressed in terms of the sum of the two pollutants—NMOG+NO X in mg/mi. (254) We received no comments recommending a different approach. The California LEV III standards are also expressed as NMOG+NO X; aligning Tier 3 with LEV III is an important element of facilitating a national program.

EPA received a number of comments about how the proposed NMOG+NO X standards transition from the existing Tier 2 standards, but there was little comment recommending different levels of the standards themselves, especially later in the program. Based on our extensive evaluation of existing and emerging vehicle technologies (see Section IV.A.5) and the level of sulfur in gasoline that will be available during the implementation timeframe of this rule, and considering the comments we received, we continue to believe that the fully phased-in level for the fleet-average FTP NMOG+NO X standard of 30 mg/mi is the most stringent level that we can reasonably establish. As discussed in Sections IV.A.5 and IV.A.6 below, when necessary margins of compliance and the demonstrated effects of fuel sulfur on emissions performance are considered, the 30 mg/mi standard is effectively very close to zero. The 30 mg/mi Tier 3 NMOG+NO X standard is also consistent with the final LEV III standard.

A key compliance mechanism adapted from the Tier 2 program is a “bin” structure for the FTP emission standards. For these purposes, a bin is a set of several standards that must be complied with as a group. Thus, as proposed, each FTP Tier 3 bin has an NMOG+NO X standard and a PM standard, as well as CO and HCHO standards.

We intend for the Tier 3 CO and HCHO standards to prevent new engine and emission control designs that result in increases in CO and HCHO emissions, compared to levels being achieved today. The standards are based on the comparable current LEV II and Tier 2 bin standards for these pollutants, which we believe are sufficiently protective at this time. There were no comments on the proposed CO and HCHO standards. The current standards are not technology-forcing, and we believe that this will continue to be the case as Tier 3 technologies are developed.

Table IV-1 presents the bin structure for light-duty vehicle, light-duty truck, and MDPV FTP standards.

Table IV-1—Tier 3 FTP Standards for LDVs, LDTs and MDPVs
BinNMOG+NO X (mg/mi)PM a (mg/mi)CO(g/mi)HCHO(mg/mi)
Bin 16016034.24
Bin 12512532.14
Bin 707031.74
Bin 505031.74
Bin 303031.04
Bin 202031.04
Bin 00000

Consistent with the Tier 2 principle of vehicle and fuel neutrality, the same Tier 3 standards apply to all LDVs, LDTs, or MDPVs, regardless of the fuel they use, as proposed. That is, vehicles certified to operate on any fuel (e.g., gasoline, diesel fuel, E85, CNG, LNG, hydrogen, and methanol) are all subject to the same standards.

The Tier 3 NMOG+NO X standards as measured on the FTP will reduce the combined fleet-average emissions gradually from MY 2017 through 2025, as shown in Table IV-2 below. Beginning in MY 2017, there are two separate sets of fleet-average standards for, first, LDVs and LDT1s and, second, all other LDTs (LDT2s, LDT3s, and LDT4s) and MDPVs. Both fleet-average standards decline annually, converging in MY 2025. These declining average standards are identical to CARB's LEV III standards. (255)

As proposed and as discussed above (Section IV.A.2.a), the declining fleet-average NMOG+NO X FTP standards begin in MY 2017 for light-duty vehicles and light-duty trucks with a GVWR up to and including 6,000 lbs and in MY 2018 for light-duty vehicles and light-duty trucks with a GVWR greater than 6,000 lbs and MDPVs. The standards apply to the heavier vehicles a year later to facilitate the transition to a 50-state program for all manufacturers. During this transition period, as described above, there will be two fleet-average NMOG+NO X standards for each model year, one for LDVs and LDT1s and one for all other LDTs (LDT2s, LDT3s, and LDT4s) and for MDPVs that decline essentially linearly from MY 2017 through MY 2025. At that point, the two fleet-average standards converge and stabilize for all later model years at the same level, 30 mg/mi, as shown in Table IV-2.

Table IV-2—Tier 3 LDV, LDT, and MDPV Fleet Average FTP NMOG+NO X Standards
Model year2017 a 20182019202020212022202320242025 and later
LDV/LDT1 b 867972655851443730
LDT2,3,4 and MDPV10192837465564738

As discussed above (Section IV.A.2.c), for LDVs and LDTs above 6,000 lbs GVWR and MDPVs, EPA is also providing an alternative phase-in of the fleet-average 30 mg/mi FTP NMOG+NO X standard.

b. FTP PM Standards

We are establishing new FTP standards for PM emissions at the proposed levels—3 mg/mi, with a temporary standard of 6 mg/mi for in-use vehicle testing—as summarized in Table IV-3 below. These levels are intended to ensure that all new vehicles will perform at a level representing what is already being achieved by well-designed emission control technologies today.

Many commenters were either silent on or supportive of the proposed FTP PM standard levels. However, some commenters—including CARB and several NGOs and auto industry suppliers—supported a more stringent standard of 1 mg/mi, which the California LEV III program phases in beginning in MY 2025. After detailed consideration of these comments and information available at this time, we continue to believe that the PM standards that we are finalizing for the federal Tier 3 program are the most stringent technically feasible standards within the implementation timeframe of this rule. (See Section 1.5.1 of the RIA.) We will continue to work closely with CARB in this area. Specifically, our agencies will continue our parallel evaluations of how improved gravimetric PM measurement methods can reduce PM mass measurement variability at very low PM levels and how this relates to the evolving technological capabilities of automakers to reach very low PM levels with sufficient compliance margins.

PM emissions over the FTP are generally attributed to the cold start, when PM formation from combustion of the fuel is facilitated by the operating conditions, including a cold combustion chamber and fuel enrichment. During cold-start operation, PM control is less effective, especially the oxidation by the catalytic converter of semi-volatile organic compounds from the lubricating oil. We believe that for vehicles that are not already at the Tier 3 levels, the new standards can be achieved with improvements to the fuel controls during the cold start, without the need for any new technology or hardware. We also expect that manufacturers will pay close attention to maintaining low PM emissions during the implementation of newer technologies like gasoline direct injection (GDI) and turbocharged engines. Improvements in cold-start exhaust catalyst performance for NMOG+NO X control will also reduce emissions of semi-volatile organic PM. For these reasons, cold start PM levels are relatively independent of vehicle application and therefore we are finalizing a single FTP PM standard for all light-duty vehicles, as proposed.

Unlike the NMOG+NO X FTP standard, it is not necessary for the FTP PM standard to phase in on a declining curve over time, since most manufacturers are already producing vehicles that meet the new standards. We are finalizing the proposed PM FTP percent-of-sales phase-in during the first 5 years of the Tier 3 program in response to concerns expressed by automakers about logistical, facilities, and compliance challenges with a standard in the range of 3 mg/mi in the early years of the program. Beginning in MY 2017 (and in MY 2018 for LDVs and LDTs over 6,000 lbs GVWR and MDPVs), manufacturers will need to comply with the PM standard with a minimum of 20 percent of their U.S. sales. As shown in Table IV-3, the percentage of the manufacturer's sales that need to comply increases each year, reaching 100 percent in MY 2021. In addition to this percent phase-in, we are also establishing, as proposed, a separate PM standard of 6 mg/mi that will apply only for in-use testing of vehicles certified to the new standards, and only during the percent phase-in period.

Due to the MY 2018 start date for vehicles over 6,000 lbs GVWR, manufacturers that have few or no vehicle models over 6,000 lbs GVWR will be required to certify a larger percentage of their total light-duty sales in MY 2017 than full line manufacturers. While we believe that most manufacturers will likely choose a single large-volume durability group to meet the 2017 requirements, we are also including an option that a manufacturer could use to comply with the MY 2017 PM requirements. Under this option, a manufacturer may choose to certify 10 percent of its total light-duty vehicle sales in MY 2017 to the new PM standards, including light-duty vehicles over 6,000 lbs. This approach is consistent with the CARB LEV III program, which requires that 10 percent of all light-duty vehicle sales meet the new PM standards in MY 2017.

Because of the expected time and expense of performing emission tests on the improved PM test procedures, we are limiting the number of tests using the new procedures that a manufacturer needs to perform at certification and during in-use testing, as proposed. Specifically, manufacturers will only be required to test vehicles representing a minimum of 25 percent of a model's durability test groups during certification each model year (and a minimum of 2 durability groups). (256) Manufacturers may select which durability groups to test, but will need to rotate the groups tested each year to eventually cover their whole fleet. Similarly, manufacturers performing in-use testing under the In-Use Verification Program can limit their testing to 50 percent of their low- and high-mileage test vehicles. Again, manufacturers will need to rotate their vehicle models so that each model will be tested every other year. Overall, we believe that the flexibility that these provisions provide will facilitate the expeditious implementation of the Tier 3 program, with no significant impact on the benefits of the program.

Table IV-3—Summary of Tier 3 LDV, LDT, and MDPV FTP Standards
Program elementUnitsModel year2017 a 201820192020202120222023+Notes
NMOG+NO X Standard (fleet average)mg/miPer declining fleet averages (see Table IV-2) b        
PM Standards         
Phase-in % 20 c 20 40 70 100 100 100
FTP:         
Certificationmg/mi3333333Note d.
In-usemg/mi6666633Note e.

As discussed in Section IV.A.2.c above, for LDVs and LDTs above 6,000 lbs GVWR and MDPVs, EPA is providing an alternative phase-in of the 3 mg/mi FTP PM standard.

4. SFTP Standards

In addition to addressing vehicle emissions during typical driving, as addressed by the FTP standards presented above, the Tier 3 program also addresses emissions during more severe driving conditions. Thus, we are finalizing NMOG+NO X and PM standards as measured on the SFTP. The SFTP (and specifically the US06 component of the test) is designed to simulate, among other conditions, higher speeds and higher acceleration rates, and thus higher loads. As described below, most commenters were supportive of or silent on the proposed SFTP NMOG+NO X standards and the associated declining fleet-average phase-in schedule, but several commenters stated that the level of the standards should be more stringent than proposed. Based on our analysis of the stringency of the program, discussed in Section IV.A.5 below and in Chapter 1 of the RIA, we disagree that more stringent SFTP NMOG+NO X standards are necessary or appropriate at this time, and we are finalizing the standards and phase-in schedule as proposed. However, we are finalizing more stringent SFTP standards for PM, which focus on the US06 test component, based on newer data and public comments. These are also described below.

The Tier 3 SFTP standards are necessary to address emissions during high-load conditions, when engines can go into a fuel “enrichment” mode and the engine's controls may temporarily create a rich air/fuel mixture to protect exhaust components from thermal damage. Enrichment can increase emissions of NMOG+NO X and PM, primarily due to the incomplete combustion that occurs under rich conditions and the diminished effectiveness of the catalyst in these circumstances. However, enrichment can be minimized or eliminated in current and future engines, where components can be thermally protected even under high-load conditions by careful electronic management of the air/fuel mixture and the combustion process. We are finalizing these SFTP standards, as well as limitations on the amount of enrichment that drivers can command (see Section IV.A.4.c below) to address this important source of vehicle emission.

We are also finalizing an SFTP composite CO standard of 4.2 g/mi for all model years 2017 (or 2018 for LDVs and LDTs over 6000 lbs GVWR and MDPVs) and later. This standard represents no effective change from the current Tier 2 SFTP CO standard, which we believe is already at a level that is sufficiently stringent.

a. SFTP NMOG+NO

We are finalizing the Tier 3 SFTP NMOG+NO X standards and declining fleet-average phase-in schedule as proposed and as presented in Table IV-5 below. Most commenters were generally supportive of these standards or silent about them. However, several commenters stated that the proposed standards are too lenient, based on their evaluation of vehicle emission test data we presented in the NPRM. We have considered these comments and have reviewed the data from the NPRM. Our conclusion from that data continues to be that the SFTP NMOG+NO X emission levels that we are finalizing ensure that manufacturers essentially eliminate fuel enrichment events and their emissions consequences, thereby resulting in important emissions reductions. See Chapter 1 of the RIA for an analysis of this data. We do not believe that significant additional reductions would result from SFTP weighted NMOG+NO X standards more stringent than the 50 mg/mi fully phased-in level. In addition, we believe that the 50 mg/mi standard will ensure that the SFTP performance of future vehicles with future technologies continues to be comparable to that of the current fleet. The SFTP emissions value for certification of gaseous pollutants will continue to be calculated as a weighted composite value of emissions on three cycles (0.35 × FTP + 0.28 × US06 + 0.37 × SC03), as is done for the Tier 2 SFTP standards.

To provide flexibility in meeting the fleet-average standards, manufacturers will, as proposed, determine the specific SFTP composite standard for each individual vehicle family and report that self-selected standard and the measured emission performance. (These self-selected standards are analogous to “family emission limits,” or “FELs,” used in other programs (e.g., heavy-duty highway engine standards).) For each family, a manufacturer will choose any composite NMOG+NO X standard, up to 180 mg/mi, in even 10 mg/mi increments. The manufacturer will then calculate the sales-weighted average of all the selected standards of the families across its fleet and compare that emissions value to the applicable fleet-average standards for that model year. Table IV-4 presents the declining fleet-average SFTP NMOG+NO X standards.

As discussed in Section IV.A.2.c above, for LDVs and LDTs above 6,000 lbs GVWR and MDPVs, EPA is providing an alternative phase-in of the 50 mg/mi SFTP NMOG+NO X standard.

b. US06 PM Standards

We are finalizing a single short-term US06 PM standard of 10 mg/mi for MYs 2017 and 2018 (or only for MY 2018 for LDVs and LDTs over 6,000 lbs GVWR and MDPVs) and a single long-term standard of 6 mg/mi for MY 2019 and later. These standards are numerically lower than those we proposed, and less complex in their structure. As discussed below and in Chapter 1 of the RIA, a substantial body of more recent PM data from a variety of vehicles tested on the US06 cycle has given us greater understanding of the feasible level of control of these emissions, both currently and in the timeframe of the Tier 3 standards, including what level of control we may reasonably require for the light-duty fleet. The standards we are finalizing reflect this review. Much of the more recent data was developed late in the development of the NPRM and, although we made it available in the rulemaking docket to inform potential commenters, the proposed standards did not reflect consideration of the newer data. Since the NPRM, additional data from CARB have become available, and we have considered all of this information in finalizing the US06 PM standards.

We believe that the fully phased-in US06 PM standard of 6 mg/mi will achieve the goal that we presented in the NPRM—to maintain the performance being achieved by current well-performing vehicles taking into account reasonable compliance margins. Comments from stakeholders representing states, including CARB, and several NGOs urged EPA to finalize more stringent standards than those proposed, in some cases advocating for standards below 6 mg/mi. Conversely, auto industry commenters generally supported the proposed standards. We have concluded that the body of recent data clearly shows that the long-term 6 mg/mi standard, is the appropriate level to prevent any significant “backsliding” in US06 PM emissions as new vehicles and technologies enter the fleet. At the same time, the 6 mg/mi standard provides a reasonable compliance margin—about 50% above the average levels of current vehicles, which are averaging about 4 mg/mi. A long-term standard numerically lower than 6 mg/mi would run counter to our intent to bring the emissions performance of all vehicles to that already being demonstrated by many vehicles in the current light-duty fleet. We believe the long-term US06 PM standard we are finalizing is appropriate based on all of the information available at this time and will not hinder introduction of new technologies manufacturers may choose for compliance with the other Tier 3 standards or other rules.

The short-term, less-stringent US06 standard of 10 mg/mi (applicable in MYs 2017 and 2018) responds to automaker concerns about uncertainties stemming from simultaneous regulatory requirements and rapidly evolving exhaust and engine technologies in the coming years. We recognize that vehicle control technologies for both criteria and GHG emissions are evolving and will continue to do so, including an expected expansion of gasoline direct injection (GDI) technologies (see IV.A.5.c and the RIA). Also, the transition to lower sulfur in-use gasoline required by this rule may create temporary additional challenges in consistently achieving lower US06 PM emissions (see IV.A.6 and the RIA). We believe that most manufacturers will implement similar if not identical emission control strategies to comply (or, more often, to continue to comply with) with both the 10 mg/mi and the 6 mg/mi standards. In so doing, we expect them to use the temporary additional compliance margin provided by the 10 mg/mi standard to reduce uncertainties about potential variability in performance (in use and, in particular, later in vehicle life) during the early years of developing and commercializing their control technologies. (257)

The 10 mg/mi standard will expire after MY 2018, and the long-term standard of 6 mg/mi will take effect. As the implementation of the program continues, we believe a limited degree of relief for testing of in-use vehicles is appropriate. Manufacturers commented that because of the industry's general lack of experience with stringent PM standards, especially as the newly-designed vehicles age, less stringent standards for in-use testing would reduce near-term concerns about performance variability early in the program. We agree, and we are finalizing a separate standard of 10 mg/mi for in-use vehicle testing for the intermediate years of the program, MYs 2019 through MY 2023. This standard is numerically lower than the proposed in-use standards—again because of the availability of improved US06 test data as described above—but the purpose of providing an in-use standard remains the same. The in-use standard, in conjunction with the short-term 10 mg/mi standard represents a longer duration for the in-use standard than we had proposed, again based on comments from the industry about their compliance concerns with new US06 standards. For MY 2024 and later, there will be no separate in-use standard and all vehicles will need to meet the long-term standard at certification and in use.

EPA proposed that different US06 PM standards apply to lighter and heavier vehicles. The newer US06 PM test data discussed above also make clear that the US06 PM performance of current vehicles is not closely related to vehicle weight, although the earlier data had indicated that this might be the case. Several commenters urged EPA to finalize a single standard for vehicles above and below 6,000 lbs GVWR based on the newer data. At the same time, auto manufacturers generally supported the proposed vehicle weight distinction, asserting a higher degree of uncertainty about the emission performance of their larger vehicles, especially in the early years of the program and in light of simultaneous technology challenges. The newer data clearly show that larger vehicles today are generally achieving US06 PM levels very similar to smaller vehicles, and well below the proposed standards. We are not finalizing separate US06 standards for heavier and lighter vehicles because separate standards are unwarranted based on a review of the newer data. However, we believe that the short-term 10 mg/mi standard, as well as the temporary in-use vehicle testing standard, will significantly reduce manufacturer compliance uncertainties in the early years of the program for all vehicles, as discussed above.

As with the FTP PM standards, manufacturers will comply with the US06 PM standards with the same increasing minimum percentage of their vehicles, as shown in Table IV.5. Also as with the FTP PM phase-in, we are providing the option for a manufacturer to choose to certify 10 percent of its total light-duty vehicle sales in MY 2017 to the new US06 PM standards, including light-duty vehicles over 6,000 lbs GVWR.

As discussed in Section IV.A.2.c above, for LDVs and LDTs more than 6,000 lbs GVWR and MDPVs, EPA is also providing an alternative phase-in of the US06 PM standards.

All of the SFTP/US06 standards are shown in Table IV-4 and Table IV-5.

Table IV-4—Tier 3 LDV, LDT, and MDPV SFTP Composite Fleet Average Standards
Model year2017 a 20182019202020212022202320242025 and later
NMOG+NO X (mg/mi)1039790837770635750
CO (g/mi)4.2 a         
Table IV-5—Summary of LDV, LDT, and MDPV Tier 3 SFTP Standards
Program elementUnits2017 a Model year2018201920202021202220232024+Notes
NMOG+NO X Standard (fleet average)mg/miPer declining fleet average for cars and trucks (see Table IV-4) b        
PM Standards:          
Phase-in % 20c 20 40 70 100 100 100 100
US06:          
LDV, LDT, MDPV: Certificationmg/mi1010666666Note d.
LDV, LDT, MDPV: In-Use1010101010
c. Enrichment Limitation for Spark-Ignition Engines

To prevent emissions that result from excessive enrichment from auxiliary emission control devices (AECD) that are substantially present during the SFTP cycles, we are finalizing limitations on the magnitude of enrichment that can be commanded, including enrichment episodes encountered during in-use operation. During conditions where enrichment is demonstrated to be present on the SFTP, the nominal air-to-fuel ratio cannot be richer at any time than the leanest air-to-fuel ratio required to obtain maximum torque (lean best torque or LBT). An air-to-fuel ratio of LBT plus a tolerance of 4 percent additional enrichment will be allowed in actual vehicle testing to protect for any in-use variance in the air-to-fuel ratio from the nominal LBT air-to-fuel determination, for such reasons as air or fuel distribution differences from production variances or aging.

LBT is defined as the leanest air-to-fuel ratio required at a speed and load point with a fixed spark advance to make peak torque. Specifically, an increase in fuel will not result in an increase in torque while maintaining a fixed spark advance. LBT is determined by setting the spark advance to a setting that is less than or equal to the spark advance required for best torque (MBT) and maintaining that spark advance when sweeping the air-to-fuel ratio. This fixed spark advance requirement is intended to prevent torque changes related to spark changes masking true LBT. One manufacturer commented that there is no universally accepted definition or procedure to determine LBT so we should retain the Tier 2 LBT requirements. We believe that the proposed definition provides sufficient clarity and will generally agree with most manufacturers' internal definition of LBT. Additionally, we are finalizing the flexibility that manufacturers may request approval of an alternative LBT definition for a unique technology or control strategy. The Agency may determine that an enrichment amount is excessive or not necessary and therefore deem that the approach does not meet the air-to-fuel ratio requirements.

Enrichment required for thermal protection will continue to be allowed upon demonstration of necessity to the Agency, based upon temperature limitations of the engine or exhaust components. Manufacturers will be required to provide descriptions of all components requiring thermal protection, temperature limitations of the components, how the enrichment strategy will detect over-temperature conditions and correct them, and a justification regarding why the enrichment is the minimum necessary to protect the specific components. The Agency may determine that the enrichment is not justified or is not the minimum necessary based on the use of engineering judgment using industry-reported thermal protection requirements.

A manufacturer commented that this requirement to report enrichment requirements for component protection for every application is burdensome and unnecessary. EPA believes that closer review of off-cycle enrichment by the agency, including enrichment for component protection, is necessary to ensure emissions are well controlled under all operating conditions. While this requirement may in some cases require additional resources at certification, this information has generally been required to be maintained by manufactures to support use of enrichment as an auxiliary emission control device (AECD) and therefore should be an exercise of reporting existing records for most manufacturers.

The requirements described in this section apply for vehicles certified to any of the Tier 3 standards.

5. Feasibility of the NMOG+NO

In the proposal, we concluded that all of the Tier 3 emissions standards are technologically feasible in the time frame of the program. The technical conclusions we reached at that time have been further reinforced by information we received in the public comments or has otherwise become available and placed in the docket for this rulemaking. After considering the comments received and with additional supporting information in Chapter 1 of the RIA, we conclude that the Tier 3 standards are feasible and reasonable, considering lead-time provided and expected compliance costs.

For each of the emission standards, the lead time provided by the program is more than sufficient for all manufacturers to comply. First, manufacturers in many cases are already adopting complying technologies for reasons other than this rulemaking. For example, many of the technologies that manufacturers have begun to develop for model years as early as MY 2014 in response to the CARB LEV III FTP and SFTP NMOG+NO X standards for the California market will likely represent steps toward compliance with this national program. Similarly, manufacturers have been producing some limited vehicle offerings since as early as MY 2000 that comply with our final MY 2025 standards in response to the CARB PZEV requirements. In addition, as described above, our program incorporates a number of phase-in provisions that will ease the transition to compliance, including time some manufacturers may need to install PM testing capability and to ramp up production on a national scale. This feasibility assessment is based on a variety of complementary technical data, studies, and analyses. As described below, these include our analysis of the stringency of the standards as compared to current Tier 2 emission levels. We also discuss below our observation that manufacturers are currently certifying several vehicle models under the California LEV II program that could likely achieve the Tier 3 NMOG+NO X and PM standards or similar levels. EPA has assessed the emissions control challenges manufacturers will generally face (e.g., cold start NMOG reductions and running (warmed-up) NO X emissions under typical and more aggressive driving conditions) and the corresponding technologies that we expect to be available to manufacturers to meet these challenges. Our feasibility assessment accounts for the fact that the Tier 3 program will apply to all types of new vehicles, ranging from small cars to large pick-up trucks and MDPVs and representing a wide diversity in applications and in specific engine designs.

It is important to note that our primary assessment of the feasibility of engine and emission control technologies is based on the assumption that vehicles will be certified on gasoline with a fuel sulfur content of 10 ppm and operated on in-use gasoline with an average of 10 ppm sulfur. (258) Therefore, our primary assessment does not incorporate the degradation of emission control system caused by higher levels of sulfur content, as is discussed in Section IV.A.6 below and further discussed in the RIA. This assessment reinforces the critical role of gasoline sulfur control in making it possible for EPA to establish emission standards at these very stringent levels. See Section IV.6 below for a full discussion of our current knowledge of the effects of gasoline sulfur on current Tier 2 vehicle emissions as well as our projections of how we expect that sulfur will affect compliance on vehicles with standards in the range of the Tier 3 standards. The projections are based on extensive EPA testing of Tier 2 vehicles as well as targeted evaluation of passenger cars and heavier trucks performing at or near the Tier 3 Bin 30 (30 mg/mi NMOG+NO X) including manufacturer supplied data of a prototype Tier 3 light-duty truck as discussed in Section IV.6.

Since there are multiple aspects to the Tier 3 program, it is necessary to consider technical feasibility in light of the different program requirements and their interactions with each other. In many cases, manufacturers will be able to address more than one requirement with the same general technological approach (e.g., faster catalyst light-off can improve both FTP NMOG+NO X and PM emissions). At the same time, the feasibility assessment must consider that different technologies may be needed on different types of vehicle applications (e.g., cars versus trucks) and must consider the relative effectiveness of these technologies in reducing emissions for the full useful life of the vehicle while operating on expected in-use fuel. For example, certain smaller vehicles with correspondingly small engines may be less challenged to meet FTP standards than larger vehicles with larger engines. Conversely, these smaller vehicles may have more difficulty meeting the more aggressive SFTP requirements than vehicles with larger and more powerful engines. Additionally, the ability to meet the SFTP emission requirements can also be impacted by the path taken to meet the FTP requirements (e.g., larger volume catalysts for US06 emissions control vs. smaller catalysts for improved FTP cold-start emissions control). Throughout the following discussion, we address how these factors, individually and in interaction with each other, affect the feasibility of the final program.

a. FTP NMOG+NO

The Tier 3 emission requirements include stringent NMOG+NO X standards on the FTP that will require new vehicle hardware in order to achieve the 30 mg/mi fleet average level in MY 2025. The type of new hardware that will be required will vary depending on the specific application and emission challenges. Smaller vehicles with corresponding smaller engines will generally need less new hardware while larger vehicles may need additional hardware and improvements beyond what will be needed for the smaller vehicles. While some vehicles, especially larger light trucks, may face higher costs in meeting the standards, it is important to remember that not every vehicle needs to meet the standard. The program has been structured to provide higher emission standard “bins” (see Table IV-1 above) to which manufacturers may certify more challenged vehicles, so long as these vehicles are offset with vehicles certified in lower emission bins such that the fleet-wide average meets the standards. We believe that the availability of the less-stringent bins will allow for the balancing of feasibility and cost considerations of compliance strategies for all vehicles. In the Tier 2 program, manufacturers took advantage of this flexibility, especially in the early years of the program. Then, as technologies improved and/or became less expensive and the need for averaging diminished, manufacturers began certifying all or most of their fleets to the average bin (Tier 2 Bin 5). We anticipate that manufacturers will follow a similar trend with the Tier 3 standards, relying on fleet averaging more significantly in the transitional years but certifying increasing numbers of their vehicles to the final fleet average standard of 30 mg/mi in the later years of the program.

In order to assess the technical feasibility of a 30 mg/mi NMOG+NO X national fleet average FTP standard, EPA conducted two supporting analyses. The initial analyses performed were of the current Tier 2 and LEV II fleets. This provided a baseline for the current federal fleet emissions performance, as well as the emissions performance of the California LEV II fleet. The second consideration was a modal analysis of typical vehicle emissions under certain operating conditions. In this way EPA determined the specific emissions performance challenges that vehicle manufacturers will face in meeting the lower fleet average emission standards. Each of these considerations is described in greater detail below.

The current Tier 2 federal fleet is certified to an average of Tier 2 Bin 5, equivalent to 160 mg/mi NMOG+NO X. (259) As an example, for MY 2009 when the Tier 2 program was fully implemented across all vehicle types, 92 percent of LDVs and LDT1s were certified to Tier 2 Bin 5 and 91 percent of LDT2s through LDT4s were certified to Tier 2 Bin 5. This trend has generally continued through MY 2013 as the most recent certification results indicate that manufacturers are continuing to certify primarily to Tier 2 Bin 5 standards for the federal fleet however there has been a shift to more certifications using the cleaner bins as discussed in the RIA. This is not an unexpected result as there is no motivation prior to implementation of the Tier 3 rulemaking for vehicle manufacturers to produce a federal fleet that over-complies with respect to the existing Tier 2 standards. By comparison, in the California fleet where compliance with the declining fleet average NMOG requirement and the “PZEV” program requires manufacturers to certify vehicles to cleaner levels, only 30 percent of the LDVs and LDT1s are certified to Tier 2 Bin 5 and 60 percent are certified to cleaner bins such as Tier 2 Bin 3 and 4. The situation regarding the truck fleet in California is similarly stratified, with 37 percent of the LDT2s through LDT4s being certified to Tier 2 Bin 5 and 55 percent being certified to the cleaner Tier 2 Bin 3 and 4. In many cases identical vehicles are being certified to a lower standard in California and a higher standard federally simply because there is no incentive to over perform to the federal standards. We note that vehicles certified to a lower standard in California are operated on gasoline with an average sulfur content of 10 ppm and thereby are able to maintain their emissions performance in-use. Based on these patterns of federal and California certification, EPA believes that much of the existing Tier 2 fleet could currently be certified to a lower federal fleet average immediately, with no significant feasibility concerns, if lower sulfur gasoline were made available nationwide.

Regardless of the Tier 2 bin standards at which manufacturers choose to certify their vehicles, actual measured emissions performance of these vehicles is typically well below the numerical standards. This difference is referred to as “compliance margin” and is a result of manufacturers' efforts to address all the sources of variability, including:

  • Test-to-test variability (within one test site and lab-to-lab)
  • Build variation expectations
  • Manufacturing tolerances and stack-up
  • Vehicle operation (for example: driving habits, ambient temperature, etc.)
  • Fuel composition
  • The effects of fuel sulfur on exhaust catalysts and oxygen sensors
  • The effects of other fuel components, including ethanol and gasoline additives
  • Oil consumption
  • The impact of oil additives and oil ash on exhaust catalysts and oxygen sensors

For MY 2009 thru MY 2013, the compliance margin for a Tier 2 Bin 5 vehicle averaged approximately 60 percent. In other words, actual vehicle emissions performance was on average about 40 percent of a 160 mg/mi NMOG+NO X standard, or about 64 mg/mi. By comparison, for California-certified vehicles, the average Super Ultra Low Emission Vehicle (SULEV) compliance margin was somewhat less for the more stringent standards, approximately 50 percent. We believe that the recent California experience is a likely indicator of compliance margins that manufacturers will design for in order to comply with the Tier 3 FTP standards. Thus, a typical Tier 2 Bin 5 vehicle, performing at 40 percent of the current standard (i.e., at about 64 mg/mi) will need improvements sufficient to reach about 15 mg/mi (50 percent of a 30 mg/mi standard).

To understand how the several currently-used technologies described below could be used by manufacturers to reach the stringent Tier 3-NMOG+NO X standards, it is helpful to consider emissions formation in common modes of operation for gasoline engines, or modal analysis. (260) The primary challenge faced by manufacturers for producing Tier 3 compliant light-duty gasoline vehicle powertrains will be keeping warmed-up running emissions at effectively zero emissions levels while reducing the emissions during cold-start operation which, based on modal analysis of a gasoline-powered vehicle being operated on the FTP cycle, occurs during about the first 50 seconds after engine start. Thus, we believe that to comply with the Tier 3 FTP standards, manufacturers will focus on effective control of these cold-start emissions while maintaining zero running emissions; this is only possible when sulfur levels in the fuel do not degrade catalyst performance. As discussed below, light-duty manufacturers are already applying several technologies capable of significant reductions in these cold start emissions to vehicles currently on the road.

During the analysis of current vehicles certified to the cleanest emission levels (Tier 2 Bin 2 and LEV II SULEV) it was noted that no large pick-ups equipped with their application specific engines were performing at the 30 mg/mi NMOG+NO X level. We believe that these applications may be the most challenging due to the fact that the design criteria required to provide the utility aspect may have direct impact on their ability to implement some of the technologies described in section IV.A.5.d below. Since these vehicles represent a substantial and important part of the light- duty fleet, EPA performed a technical feasibility study directly targeting this class of vehicles.

In order to assess the technical feasibility of a 30 mg/mi FTP NMOG+NO X standard, EPA purchased a 2011 Chevrolet Silverado heavy-light-duty (LDT4) pickup truck with a developmental goal of modifying the truck to achieve exhaust emission levels in compliance with the Tier 3 Bin 30 emissions standards including a reasonable compliance margin. The truck was equipped with a 5.3L V8 with General Motors' “Active Fuel Management” cylinder deactivation system. This particular truck was chosen as an example of a Tier 3 prototype in part because cylinder deactivation is a key technology for light-truck compliance with future GHG standards and in part because it achieved very low emissions in the OEM, Tier 2-compliant configuration (certified to Tier 2 Bin 4). A prototype exhaust system was obtained from MECA consisting of high-cell-density (900 cpsi) thin-wall (2.5 mil), high-PGM, close-coupled Pd-Rh catalysts with an additional under-body Pd-Rh catalyst. The total catalyst volume was approximately 116 in 3 with a specific PGM loading of 125 g/ft 3 and approximate loading ratio of 0:80:5 (Pt:Pd:Rh). Third-party (non-OEM) EMS calibration tools were used to modify the powertrain calibration in an effort to improve catalyst light-off performance. The final test configuration used approximately 4 degrees of timing retard and approximately 200 rpm higher idle speed relative to the OEM configuration during and immediately following cold-start. The exhaust catalyst system and HEGO sensors were bench aged to an equivalent 150,000 miles using standard EPA accelerated catalyst bench-aging procedures. The truck was tested on California LEV III E10 certification fuel at 9 ppm gasoline sulfur levels.

The EPA Tier 3 prototype Silverado achieved NMOG+NO X emissions of 18 mg/mi on the 9 ppm S fuel. The NMOG+NO X emissions were approximately 60% of the Bin 30 standard and thus are consistent with meeting the Tier 3 Bin 30 exhaust emissions standard with a moderate compliance margin. The technologies used on the prototype Silverado to achieve these emission levels are common approaches used today on smaller vehicles. They do not compromise any of the design utility of this vehicle class and are some of the same approaches we expect manufacturers to use to meet the Tier 3 Bin 30 exhaust emissions standards.

b. SFTP NMOG+NO

The increase in the stringency of the SFTP NMOG+NO X standards, specifically across the US06 cycle, will generally only require additional focus on fuel control of the engines and diligent implementation of new technologies that manufacturers are already introducing or are likely to introduce in response to the current and 2017 LD GHG emission standards. These include downsized gasoline direct injection (GDI) and turbocharged engines, which may also include improvements to the engine and emission control hardware to tolerate higher combustion and exhaust temperatures expected in these future GHG-oriented engine designs when under higher loads. The upgraded materials or components will enable manufacturers to rely less on fuel enrichment during high-speed/high-load operation to protect components from overheating. This fuel enrichment is currently the source of elevated VOC, NO X, and PM emissions seen in a subset of the current Tier 2 fleet.

With respect to enrichment, the primary method available to manufacturers to protect the catalyst and other exhaust components from over-temperature conditions has been changes to the fuel/air mixture by increasing the fuel fraction, but this is no longer the only tool available to manufacturers for this purpose. With the application of electronic throttle controls, variable valve timing, exhaust gas recirculation and other exhaust temperature influencing technologies on nearly every light-duty vehicle, the manufacturer has the ability to systematically control the operation and combustion processes of the engine to minimize or altogether avoid areas and modes of operation where thermal issues can occur. While some of these solutions could in some cases result in a small and temporary reduction in vehicle performance (absolute power levels), we believe that it could be an effective way to reduce NMOG+NO X emissions over the SFTP test.

Additionally, some components, especially catalysts, can experience accelerated thermal deterioration that occurs when operating at higher temperatures for more time than expected under normal operation (e.g., trailer towing, mountain grades). Some upgrades of existing vehicle emission control technology, like catalyst substrates and washcoats may be required to limit thermal deterioration and ensure vehicle emissions compliance throughout the useful life of the vehicle.

In order to assess the technical feasibility of a 50 mg/mi NMOG+NO X national fleet average SFTP standard, EPA conducted an analysis of SFTP levels of Tier 2 and LEV II vehicles. The analysis was performed on the US06 results from current Tier 2 and LEV II vehicles tested in the in-use verification program (IUVP) by manufacturers and submitted to EPA. This analysis provided a baseline for the current Tier 2 and LEV II fleet emissions performance, as well as the SFTP emissions performance capability of the cleanest vehicles meeting the Tier 3 FTP standards. The analysis concluded that most vehicles in the IUVP testing program are already capable of meeting the composite SFTP standard of 50 mg/mi when the Tier 3 FTP standard levels are factored into the composite calculation. With the technological improvements already underway as discussed above, we believe all MY 2017 and later vehicles will be able to comply with the SFTP standards, either directly or through the flexibility of the averaging, banking and trading program. For further information on the analysis see Chapter 1 of the RIA.

c. FTP and SFTP PM Standards

As described above for NMOG+NO X over the SFTP, the increase in the stringency of the FTP and SFTP PM standards will generally also only require additional focus on fuel control of the engines and attention to PM emissions during the implementation of new technologies like gasoline direct injection (GDI) and turbocharged engines. Some upgrades of existing vehicle emission control technology may be required to ensure vehicle emissions performance is maintained throughout the useful life of the vehicle. These upgrades may include improvements to the engine to control wear that could result in increased PM from oil consumption and selection of GDI systems that will be capable of continuing to perform optimally even as the systems age.

We based our conclusions about the ability of manufacturers to meet the PM standards largely on the PM performance of the existing fleet, both on the FTP and SFTP. In the case of FTP testing of current vehicles, data on both low and high mileage light-duty vehicles demonstrate that the majority of vehicles are currently achieving levels at or below the Tier 3 FTP PM standards.

The testing results can be found in Chapter 1 of the RIA. A small number of vehicles are at or just over the Tier 3 FTP PM standard at low mileage and could require calibration changes and/or catalyst changes to meet the new standards. It is our expectation that the same calibration and catalyst changes required to address NMOG will also provide the necessary PM control. Vehicles that currently have higher PM emissions over the FTP or SFTP at higher mileages will likely be required to control oil consumption and combustion chamber deposits.

We also analyzed PM test data results on the US06 test cycle from Tier 2 vehicles. The data show that many vehicles are already at or below the Tier 3 standards on the US06 test cycle. Vehicles that have high PM emission rates on the US06 will likely need to control enrichment and oil consumption, particularly later in life. As described above for SFTP NMOG+NO X control, enrichment can be more accurately managed through available electronic engine controls. The strategies for reducing oil consumption are similar to those described above for controlling oil consumption on the FTP. However, given the higher engine speeds experienced on the US06 and the increase in oil consumption that can accompany this kind of operation, manufacturers will most likely focus on oil sources stemming from the piston to cylinder interface and positive crankcase ventilation (PCV).

Manufacturers have informed us that they have already reduced or are planning to reduce the oil consumption of their engines by improved sealing of the paths of oil into the combustion chamber and improved piston-to-cylinder interfaces. Auto manufacturers have stated that they are already taking or considering these actions to address issues of customer satisfaction and cost of ownership. In addition, many vehicle manufacturers acknowledge the relationship between combustion chamber deposits and PM formation and are actively pursuing design changes to mitigate fuel impingement within the combustion chamber and its commensurate PM effects. Both types of controls are being widely applied by manufacturers today.

d. Technologies Manufacturers Are Likely To Apply

Most of the technologies expected to be applied to light-duty vehicles to meet the stringent Tier 3 standards will address the emissions control system's ability to reduce emission during cold start while maintaining zero or near zero running emissions. The effectiveness of current vehicle emissions control systems at reducing cold start emissions depends in large part on the time it takes for the catalyst to light off, which is typically defined as the catalyst reaching a temperature of 250 °C. In order to improve catalyst light-off, we expect that manufacturers will add technologies that provide heat from combustion more readily to the catalyst or improve the catalyst efficiency at lower temperatures. These technologies include calibration changes, catalyst platinum group metals (PGM) loading and strategy, thermal management, close-coupled catalysts, and secondary air injection, all which generally improve emission performance of all pollutants. In some cases where the catalyst light-off and efficiency are not enough to address the cold start NMOG emissions, hydrocarbon adsorbers may be applied to trap hydrocarbons until such time that the catalyst is lit off. Note that with the exception of hydrocarbon adsorbers each of these technologies addresses NMOG, NO X, and PM performance. The technologies are described in greater detail below. Additional information on these technologies can also be found in Chapter 1 of the RIA.

  • Engine Control Calibration Changes—These include changes to retard spark and/or adjust air/fuel mixtures such that more combustion heat is created during the cold start. Control changes may include injection strategies in GDI applications, unique cold-start variable valve timing and lift, and other available engine parameters. Engine calibration changes can affect NMOG, NO X and PM emissions.
  • Catalyst PGM Loading—Additional PGM loading, increased loading of other active materials, and improved dispersion of PGM and other active materials in the catalyst provide a greater number of sites available to catalyze emissions and addresses NMOG, NO X and PM emissions. Catalyst PGM loading, when implemented in conjunction with low sulfur gasoline, will effectively eliminate NO X emissions under warmed-up conditions.
  • Thermal Management—This category of technologies includes all design attributes meant to conduct the combustion heat into the catalyst with minimal cooling. This includes insulating the exhaust piping between the engine and the catalyst, reducing the wetted area of the exhaust path, reducing the thermal mass of the exhaust system, and/or using close-coupled catalysts (i.e., the catalysts are packaged as close as possible to the engine's cylinder head to mitigate the cooling effects of longer exhaust piping). Thermal management technologies primarily address NMOG emissions, but also affect NO X and PM emissions.
  • Secondary Air Injection—By injecting air directly into the exhaust stream, close to the exhaust valve, combustion can be maintained within the exhaust, creating additional heat by which to increase the catalyst temperature. The air/fuel mixture must be adjusted to provide a richer exhaust gas for the secondary air to be effective. There can be a NO X emissions disbenefit to use of secondary air injection since it can impact the ability of oxygen storage components (OSC) within the catalyst to take up excess oxygen as necessary to promote NO X reduction reactions immediately following cold start conditions.
  • Hydrocarbon Adsorber—Traps hydrocarbons during a cold start until the catalyst lights off, and then releases the hydrocarbons to be converted by the catalyst.
  • Gasoline Sulfur—The relative effectiveness for NMOG and NO X control of the exhaust-catalyst related technologies is constrained by gasoline fuel sulfur levels. Thus, reduced sulfur in gasoline is an enabling technology to achieve the standards and maintain this performance during in-use operation. We discuss the relationship between gasoline sulfur and emissions in greater detail in Section IV.6 below and in the RIA.

Several commenters indicated that large light-duty trucks (e.g., pickups and full-size sport utility vehicles (SUVs) in the LDT3 and LDT4 categories) will be the most challenging light-duty vehicles to bring into compliance with the Tier 3 NMOG+NO X standards at the 30 mg/mi corporate average emissions level. A similar challenge was addressed when large light-duty trucks were brought into compliance with the Tier 2 standards over the past decade. Figure IV-1 provides a graphical representation of the effectiveness of Tier 3 technologies for large light-duty truck applications. A compliance margin is shown in both cases. Note that the graphical representation of the effectiveness of catalyst technologies on NO X and NMOG when going from Tier 2 to Tier 3 levels also includes a reduction in gasoline sulfur levels from 30 ppm to 10 ppm.

Image #ER28AP14.000

6. Impact of Gasoline Sulfur Control on the Effectiveness of the Vehicle Emission Standards

In this section, we discuss the impact of gasoline sulfur control on the feasibility of the Tier 3 vehicle emissions standards and on the exhaust emissions of the existing in-use vehicle fleet. Section IV.A.6.a describes the chemistry and physics of the impacts of gasoline sulfur compounds on exhaust catalysts. Sections IV.A.6.b, c and d summarize research on the impacts of gasoline sulfur on vehicles utilizing various degrees of emission control technology, with Section IV.A.6.b summarizing historical studies on the impact of gasoline sulfur on vehicle emissions, Section IV.A.6.c describing impacts on Tier 2 vehicles and the existing light-duty vehicle fleet, and Section IV.A.6.d describing impacts on vehicles using technology consistent with what we expect to see in the future Tier 3 vehicle fleet. Section IV.A.6.e provides EPA's assessment of the level of gasoline sulfur control necessary for light-duty vehicles to comply with Tier 3 exhaust emission standards.

EPA's primary findings are:

  • Reducing gasoline sulfur content to a 10 ppm average will provide immediate and significant exhaust emissions reductions to the current, in-use fleet of light-duty vehicles.
  • Reducing gasoline sulfur content to an average of 10 ppm will enable vehicle manufacturers to certify their entire product lines of new light-duty vehicles to the final Tier 3 Bin 30 fleet average standards. Without such sulfur control it would not be possible for vehicle manufacturers to reduce emissions sufficiently below Tier 2 levels to meet the new Tier 3 standards because it would require offsetting significantly higher exhaust emissions resulting from the higher sulfur levels. EPA has not identified any existing or developing technologies that would compensate for or offset the higher exhaust emissions resulting from higher fuel sulfur levels.
a. Gasoline Sulfur Impacts on Exhaust Catalysts

Modern three-way catalytic exhaust systems utilize platinum group metals (PGM), metal oxides and other active materials to selectively oxidize organic compounds and carbon monoxide in the exhaust gases. These systems simultaneously reduce NO X when air-to-fuel ratio control operates in a condition of relatively low amplitude/high frequency oscillation about the stoichiometric point. Sulfur is a well-known catalyst poison. There is a large body of work demonstrating sulfur inhibition of the emissions control performance of PGM three-way exhaust catalyst systems. 262 263 264 265 266 267 268 269 270 271 The nature of sulfur interactions with washcoat materials, active catalytic materials and catalyst substrates is complex and varies with catalyst composition, exhaust gas composition and exhaust temperature. The variation of these interactions with exhaust gas composition and temperature means that the operational history of a vehicle is an important factor; continuous light-load operation, throttle tip-in events and enrichment under high-load conditions can all impact sulfur interactions with the catalyst.

Sulfur from gasoline is oxidized during spark-ignition engine combustion primarily to SO 2 and, to a much lesser extent, SO 3 2. Sulfur oxides selectively chemically bind (chemisorb) with, and in some cases react with, active sites and coating materials within the catalyst, thus inhibiting the intended catalytic reactions. Sulfur oxides inhibit pollutant catalysis chiefly by selective poisoning of active PGM, ceria sites, and the alumina washcoating material (see Figure IV-2). (272) The amount of sulfur retained by an exhaust catalyst system is primarily a function of the concentration of sulfur oxides in the incoming exhaust gases, air-to-fuel ratio feedback and control by the engine management system, the operating temperature of the catalyst and the active materials and coatings used within the catalyst.

In their supplemental comments to the Tier 3 proposal, API criticized the use of emissions data generated using gasoline with sulfur content outside of the range of 10 ppm to 30 ppm within EPA and other analyses of the impacts of gasoline sulfur on exhaust emissions from current in-use (Tier 2) and future (Tier 3) light-duty vehicles. Specific examples include:

  • Comparisons of exhaust emissions at 5 ppm and 28 ppm gasoline sulfur levels within the recent EPA study of emissions from Tier 2 vehicles  (273)
  • Comparison of exhaust emissions of a SULEV vehicle at 8 ppm and 33 ppm gasoline sulfur levels within the Takei et al. study  (274)
  • Comparison of exhaust emissions of a PZEV vehicle at 3 ppm and 33 ppm gasoline sulfur levels within the Ball et al. study. (275)

The relationship between changes in gasoline sulfur content and NO X, HC, NMHC and NMOG emissions is typically linear. The linearity of sulfur impacts on NO X, NMHC and NMOG emissions is supported by past studies with multiple fuel sulfur levels all of which compare gasoline with differing sulfur levels that are below approximately 100 ppm (e.g., CRC E-60 and 2001 AAM/AIAM programs as well as comments on this rulemaking submitted by MECA). (276 277 278) An assumption of linearity of the effect of gasoline sulfur level on catalyst efficiency between any two test fuels with differing sulfur levels is reasonable given that the mass flow rate of sulfur in exhaust gas changes in proportion to its concentration in the fuel, and that the chemistry of adsorption of sulfur on the active catalyst sites is an approximately-first-order chemisorption until all active sites within a catalyst reach an equilibrium state relative to further input of sulfur compounds. The relative linearity of the effect of gasoline sulfur level on NMOG and NO X emissions allows exhaust emissions results generated within EPA and other studies of gasoline sulfur at levels immediately above or below either 10 ppm or 30 ppm to be normalized to either 10 ppm sulfur (Tier 3 gasoline) or to 30 ppm sulfur (Tier 2 gasoline, which are used in the analysis of the impacts of the Tier 3 gasoline standards on existing in-use vehicles and future Tier 3 vehicles.

In their supplemental comments to the Tier 3 proposal, API also commented that EPA did not show the sulfur impact on exhaust emissions at intermediate sulfur levels between 10 ppm and 30 ppm. In response, based on the relative linearity of the effect of gasoline sulfur level on NMOG and NO X emissions allowing exhaust emissions to be estimated for gasoline sulfur levels between 10 and 30 ppm, data in EPA's analysis shows increases NMOG+NO X emissions (as fuel sulfur increases) that become more severe (i.e., higher percentage increase in NMOG+NO X emissions) for vehicles with extremely low  (279) exhaust emission (SULEV, PZEV, LEV III, Tier 3) as described in further detail in Sections IV.A.6.d and e.

Image #ER28AP14.001

Selective sulfur poisoning of platinum (Pt) and rhodium (Rh) is primarily from surface-layer chemisorption. Sulfur poisoning of palladium (Pd) and ceria appears to be via chemisorption combined with formation of more stable metallic sulfur compounds, e.g. PdS and Ce 2 O 2 S, present in both surface and bulk form (i.e., below the surface layer). (281 282 283 284) Ceria, zirconia and other oxygen storage components (OSC) play an important role that is crucial to NO X reduction over Rh as the engine air-to-fuel ratio oscillates about the stoichiometric closed-loop control point. (285) Ceria sulfation interferes with OSC functionality within the catalyst and thus can have a detrimental impact on the catalyst's ability to effectively reduce NO X emissions. Water-gas-shift reactions are important for NO X reduction over catalysts combining Pd and ceria. This reaction can be blocked by sulfur poisoning and may be responsible for observations of reduced NO X activity over Pd/ceria catalysts even with exposure to fairly low levels of sulfur (equivalent to 15 ppm in gasoline). (286 287) Pd is also of increased importance for meeting Tier 3 standards due to its unique application in the close-coupled-catalyst location required for vehicles certifying to very stringent emission standards. Close-coupling means that the exhaust catalyst is moved as close as possible to the engine's exhaust ports within the packaging constraints of an engine compartment. This ensures that the catalyst reaches its minimal operational, or “light-off”, temperature as quickly as possible after the vehicle is started. It also means, however, that the exhaust catalyst(s) in the close-coupled location(s) are subject to higher exhaust temperatures during fully-warmed up operation. Pd is required in closed-coupled catalysts due to its resistance to high-temperature thermal sintering thereby maintaining sufficient durability of the emissions control system over the useful life of a vehicle. Sulfur removal from Pd requires rich operation at higher temperatures than required for sulfur removal from other PGM catalysts.

In addition to its interaction with catalyst materials, sulfur can also react with the wash-coating itself to form alumina sulfate, which in turn can block coating pores and reduce gaseous diffusion to active materials below the coating surface (see Figure IV-2). (288) This may be a significant mechanism for the observed storage of sulfur compounds at light and moderate load operation with subsequent, rapid release as sulfate particulate matter emissions when high-load, high-temperature conditions are encountered. (289)

Operating the catalyst at a sufficiently high temperature under net reducing conditions (e.g., air-to-fuel equivalence that is net fuel-rich of stoichiometry) can effectively release the sulfur oxides from catalyst components. Thus, regular operation at sufficiently high temperatures at net fuel-rich air-to-fuel ratios can minimize the effects of fuel sulfur levels on catalyst active materials and catalyst efficiency; however, it cannot completely eliminate the effects of sulfur poisoning. In current vehicles, desulfurization conditions occur typically at high loads when there is a degree of commanded enrichment (i.e., fuel enrichment commanded by the engine management system primarily for protection of engine and/or exhaust system components). A study of Tier 2 vehicles in the in-use fleet recently completed by EPA (290) shows that emission levels immediately following high speed/load operation is still a function of fuel sulfur level for the gasoline used following desulfurization. If a vehicle operates on gasoline with less than 10 ppm sulfur, exhaust emissions stabilize over repeat FTP tests at emissions near those of the first FTP that follows the high speed/load operation and catalyst desulfurization. If the vehicle continues to operate on higher sulfur gasoline following desulfurization, exhaust emissions creep upward until a new equilibrium exhaust emissions level is established. This suggests that lower fuel sulfur levels achieve emission benefits unachievable by catalyst desulfurization procedures alone. Continued operation on gasoline with a 10 ppm average sulfur content or lower is necessary after catalyst desulfurization in order to achieve emissions reductions with the current in-use fleet. (291) Furthermore, regular operation at the high exhaust temperatures and rich air-to-fuel ratios necessary for catalyst desulfurization is not desirable and may not be possible for future Tier 3 vehicles for several reasons:

  • Thermal sintering and resultant catalyst degradation: The temperatures necessary to release sulfur oxides are high enough to lead to thermal degradation of the catalyst over time via thermal sintering of active materials. Sintering reduces the surface area available to participate in reactions and thus reduces the overall effectiveness of the catalyst.
  • Operational conditions: It is not always possible to maintain fuel-rich operational conditions and exhaust catalyst temperatures that are high enough for sulfur removal because of cold weather, idle conditions and light-load operation.
  • Increased emissions: In order to achieve greater emission reductions across a fuller range of in-use driving conditions, vehicle manufacturers' use of commanded enrichment, which has been beneficial for sulfur removal, will be greatly reduced or eliminated under Tier 3. Additionally, the fuel-rich air-to-fuel ratios necessary for sulfur removal from active catalytic surfaces would result in increased PM, NMOG, CO and air toxic emissions, particularly at the high-temperature, high load conditions (e.g., US06 or comparable) necessary for sulfur removal. Previously used levels of commanded enrichment (e.g., under Tier 2) would interfere with the strategies necessary to comply with more stringent Tier 3 SFTP exhaust emissions standards. There are also additional provisions within the Tier 3 standards that further restrict the use of US06 and off-cycle commanded enrichment in an effort to reduce high-load and off-cycle PM, NMOG, CO and air toxic emissions. (292)
  • Expected changes to engine performance necessary to reduce fuel consumption and greenhouse gas emissions will improve the thermal efficiency of engines and may result in reduced exhaust temperatures.
b. Previous Studies of Gasoline Sulfur Impacts

This section summarizes studies to provide historical context regarding what is known about the direct impacts of gasoline sulfur on vehicle exhaust emissions. Reducing fuel sulfur levels has been the primary regulatory mechanism EPA has used to minimize sulfur contamination of exhaust catalysts and to ensure optimum emissions performance over the useful life of a vehicle. The impact of gasoline sulfur on exhaust catalyst systems has become even more important as vehicle emission standards have become more stringent. Studies have suggested a progressive increase in catalyst sensitivity to sulfur when standards increase in stringency and emissions levels decrease. Emission standards under the programs that preceded the Tier 2 program (Tier 0, Tier 1, and National LEV, or NLEV) were high enough that the impact of sulfur was considered of little importance. The Tier 2 program recognized the importance of sulfur and reduced the sulfur levels in the fuel from around 300 ppm to 30 ppm in conjunction with the new emission standards. (293) At that time, very little work had been done to evaluate the effect of further reductions in fuel sulfur, especially on in-use vehicles that may have some degree of catalyst deterioration due to real-world operation or on vehicles with extremely low tailpipe emissions as described earlier.

In 2005, EPA and several automakers jointly conducted a research program, the Mobile Source Air Toxics (MSAT) Study that examined the effects of sulfur and other gasoline properties such as benzene and volatility on emissions from a fleet of nine Tier 2 compliant vehicles. (294) The study found significant reductions in NO X, CO and total hydrocarbons (HC) when the vehicles were tested on low sulfur fuel, relative to 32 ppm fuel. In particular, the study found a 48 percent increase in NO X over the FTP when gasoline sulfur was increased from 6 ppm to 32 ppm. Given the preparatory procedures related to catalyst clean-out and loading used by these studies, these results may represent a “best case” scenario relative to what would be expected under more typical driving conditions. Nonetheless, these data suggested the effect of in-use sulfur loading was largely reversible for Tier 2 vehicles, and that there were likely to be significant emission reductions possible with further reductions in gasoline sulfur level. More recently, EPA completed a comprehensive study on the effects of gasoline sulfur on the exhaust emissions of Tier 2 vehicles at low to moderate mileage levels. (295) Further details of this study are summarized in Section IV.A.6.c of this preamble.

In the NPRM, we summarized the limited data available regarding the impact of gasoline sulfur on the near-zero exhaust emission vehicle technologies that will be necessary for Tier 3 compliance. Vehicles certified to California LEV II SULEV and PZEV standards and federal Tier 2 Bin 2 standards achieve levels of exhaust emissions control consistent with the levels of control that will be necessary for Tier 3 compliance. While these vehicles represent only a relatively small subset (e.g., typically small light-duty vehicles and light-duty trucks with limited GVWR or towing utility) of the broad range of vehicles that will need to comply with Tier 3 standards as part of a fleet-wide average, data on these vehicles provide an opportunity to study the impact of gasoline sulfur on near-zero exhaust emission technologies and is generally representative of technology that are expected to be used with mid-size and smaller light-duty vehicles for Tier 3 compliance. Vehicle testing by Toyota (Takei et al.) of LEV I, LEV II ULEV and prototype SULEV vehicles showed larger percentage increases in NO X and HC emissions for SULEV vehicles as gasoline sulfur increased from 8 ppm to 30 ppm, as compared to other LEV vehicles they tested. (296) Ball et al. of Umicore Autocat USA, Inc. studied the impact of gasoline fuel sulfur levels of 3 ppm and 33 ppm on the emissions of a 2009 Chevrolet Malibu PZEV. (297) Umicore's testing of the Malibu PZEV vehicle showed a pronounced and progressive trend of increasing NO X emissions (referred to as “NO X creep”) when switching from a 3 ppm sulfur gasoline to repeated, back-to-back FTP tests using 33 ppm sulfur gasoline. The PZEV Chevrolet Malibu, after being aged to an equivalent of 150,000 miles, demonstrated emissions at a level consistent with the Tier 3 Bin 30 NMOG+NO X standards when operated on 3 ppm sulfur fuel and for at least one FTP test after switching to 33 ppm certification fuel. Following operation over 2 FTP cycles on 33 ppm sulfur fuel, NO X emissions alone were more than double the Tier 3 30 mg/mi NMOG+NO X standard. 271 This represents a 70% NO X increase between 3 ppm sulfur and 33 ppm sulfur gasolines, approximately 2-3 times of what has been previously reported for similar changes in fuel sulfur level for Tier 2 and older vehicles. (298 299)

Both the Umicore and Toyota studies suggest that the emissions from vehicles using near-zero exhaust emissions control technology similar to what is expected for compliance with the Tier 3 standards are more sensitive to changes in gasoline sulfur content at low (sub-30 ppm) sulfur concentrations than technology used to meet the higher Federal Tier 2 and California LEV II standards. The Umicore and Toyota studies clearly indicate that a progressive increase in catalyst sensitivity to sulfur continues as exhaust emissions decrease from levels required by federal Tier 2 and California LEV II emissions standards to the lower levels required by Tier 3 emissions standards. In addition, although vehicles with Tier 2 technology have somewhat less sulfur sensitivity compared to future Tier 3 vehicles, there is still significant opportunity for further emissions reductions from the existing in-use fleet by reducing gasoline sulfur content from 30 ppm to 10 ppm. The results of recent testing demonstrating the potential for in-use emissions reductions from further gasoline sulfur control are summarized in Section IV.A.6.c. Recent data on the impact of gasoline sulfur on vehicles with exhaust emission control technologies that we expect to be used with Tier 3 vehicles is summarized in Sections IV.A.6.d and e.

c. EPA Testing of Gasoline Sulfur Effects on Tier 2 Vehicles and the In-Use Fleet

Both the MSAT  (300) and Umicore  (301) studies showed the emission reduction potential of lower sulfur fuel on Tier 2 and later technology vehicles over the FTP cycle. However, assessing the potential for reduction on the in-use fleet requires understanding how sulfur exposure over time impacts emissions, and what the state of catalyst sulfur loading is for the typical vehicle in the field. In response to these data needs, EPA conducted a new study to assess the emission reductions expected from the in-use Tier 2 fleet with a reduction in fuel sulfur level from current levels. (302) It was designed to take into consideration what was known from prior studies on sulfur build-up in catalysts over time and the effect of periodic regeneration events that may result from higher speed and load operation over the course of day-to-day driving.

The study sample described in this analysis consisted of 93 cars and light trucks recruited from owners in southeast Michigan, covering model years 2007-9 with approximately 20,000-40,000 odometer miles. (303) The makes and models targeted for recruitment were chosen to be representative of high sales vehicles covering a range of types and sizes. Test fuels were two non-ethanol gasolines with properties typical of certification test fuel, one at a sulfur level of 5 ppm and the other at 28 ppm. All emissions data was collected using the FTP cycle at a nominal temperature of 75 °F.

Using the 28 ppm test fuel, emissions data were collected from vehicles in their as-received state as well as following a high-speed/load “clean-out” procedure consisting of two back-to-back US06 cycles intended to reduce sulfur loading in the catalyst. A statistical analysis of this data showed highly significant reductions in several pollutants including NO X and hydrocarbons, demonstrating that sulfur loadings have a large effect on exhaust catalyst performance, and that Tier 2 vehicles can achieve significant reductions based on removing, at least in part, the negative impact of the sulfur loading on catalyst efficiency (Table IV-6). For example, Bag 2 NO X emissions dropped 31 percent between the pre- and post-cleanout tests on 28 ppm fuel.

Table IV-6—Percent Reduction in In-Use Emissions After the Clean-Out Using 28 ppm Test Fuel a
NO X (p-value)THC(p-value)CO(p-value)NMHC(p-value)CH 4 (p-value)PM(p-value)
Bag 16.0% (0.0151)15.4%(< 0.0001)
Bag 231.4% (0.0003)14.9% (0.0118)18.7% (0.0131)14.4% (0.0019)
Bag 335.4% (<0.0001)20.4% (<0.0001)21.5% (0.0001)27.7% (<0.0001)10.3% (<0.0001)24.5% (<0.0001)
FTP Composite11.4% (0.0002)3.8% (0.0249)6.8% (0.0107)3.5% (0.0498)6.0% (0.0011)13.7% (<0.0001)
Bag 1-Bag 37.2%(0.0656)

To assess the impact of lower sulfur fuel on in-use emissions, further testing was conducted on a representative subset of vehicles on 28 ppm and 5 ppm fuel with accumulated mileage. A first step in this portion of the study was to assess the differences in the effectiveness of the clean-out procedure under different fuel sulfur levels. Table IV-7 presents a comparison of emissions immediately following (<50 miles) the clean-out procedures at the low vs. high sulfur level. These results show significant emission reductions for the 5 ppm fuel relative to the 28 ppm fuel immediately after this clean-out; for example, Bag 2 NO X emissions were 34 percent lower on the 5 ppm fuel vs. the 28 ppm fuel. This indicates that the catalyst is not fully desulfurized, even after a clean out procedure, as long as there is sulfur in the fuel. This further indicates that current sulfur levels in gasoline continue to have a long-term, adverse effect on exhaust emissions control that is not fully removed by intermittent clean-out procedures that can occur in day-to-day operation of a vehicle and demonstrates that lowering sulfur levels to 10 ppm on average will significantly reduce the effects of sulfur impairment on emissions control technology.

Table IV-7—Percent Reduction in Exhaust Emissions When Going From 28 ppm to 5 ppm Sulfur Gasoline for the First Three Repeat FTP Tests Immediately Following Clean-Out
NO X (p-value)THC(p-value)CO(p-value)NMHC(p-value)CH 4 (p-value)PM a
Bag 15.3%(0.0513)6.8%(0.0053)6.2%(0.0083)5.7%(0.0276)14.0%(<0.0001)
Bag 234.4%(0.0036)33.9%(<0.0001)(a)26.4%(0.0420)49.4%(<0.0001)
Bag 342.5%(<0.0001)36.9%(<0.0001)14.7%(0.0041)51.7%(<0.0001)28.5%(<0.0001)
FTP Composite15.0%(0.0002)13.3%(<0.0001)8.5%(0.0050)10.9%(0.0012)23.6%(<0.0001)
Bag 1-Bag 3(a)(a)(a)(a)(a)

To assess the overall in-use reduction between high and low sulfur fuel, a mixed model analysis of all data as a function of fuel sulfur level and miles driven after cleanout was performed. This analysis found highly significant reductions for several pollutants, as shown in Table IV-8. Reductions for Bag 2 NO X were particularly high, estimated at 52 percent between 28 ppm and 5 ppm overall. For all pollutants, the model fitting did not find a significant miles-by-sulfur interaction, suggesting the relative differences were not dependent on miles driven after clean-out.

Table IV-8—Percent Reduction in Emissions From 28 ppm to 5 ppm Fuel Sulfur on In-Use Tier 2 Vehicles
NO X (p-value)THC(p-value)CO(p-value)NMHC(p-value)CH 4 (p-value)NO X+NMOG(p-value)PM a
Bag 17.1%(0.0216)9.2%(0.0002)6.7%(0.0131)8.1%(0.0017)16.6%(< 0.0001)N/A
Bag 251.9%(< 0.0001)43.3%(< 0.0001)(a)42.7%(0.0003)51.8%(< 0.0001)N/A
Bag 347.8%(< 0.0001)40.2%(< 0.0001)15.9%(0.0003)54.7%(< 0.0001)29.2%(< 0.0001)N/A
FTP Composite14.1%(0.0008)15.3%(< 0.0001)9.5%(< 0.0001)12.4%(< 0.0001)29.3%(< 0.0001)14.4%(< 0.0001)
Bag 1-Bag 3(a)5.9%(0.0074)(a)(b)(b)N/A

Major findings from this study include:

  • Largely reversible sulfur loading is occurring in the in-use fleet of Tier 2 vehicles and has a measureable effect on emissions of NO X, hydrocarbons, and other pollutants of interest.
  • The effectiveness of high speed/load procedures in restoring catalyst efficiency is limited when operating on higher sulfur fuel.
  • Reducing fuel sulfur levels from current levels to levels in the range of the Tier 3 gasoline sulfur standards is expected to achieve significant reductions in emissions of NO X, hydrocarbons, and other pollutants of interest in the current in-use fleet.
  • Assuming that the emissions impacts vs. gasoline sulfur content are approximately linear, changing gasoline sulfur content from 30 ppm to 10 ppm would result in NMOG+NO X emissions decreasing from 52 mg/mi to 45 mg/mi, respectively (a 13% decrease), and NO X emissions decreasing from 19 mg/mi to 16 mg/mi, respectively (a 16% decrease), for the vehicles in the study.

To evaluate the robustness of the statistical analyses assessing the overall in-use emissions reduction between operation on high and low sulfur fuel (Table IV-8), a series of sensitivity analyses were performed to assess the impacts on study results of measurements from low-emitting vehicles and influential vehicles, as documented in detail in the report. (304) The sensitivity analyses showed that the magnitude and the statistical significance of the results were not impacted and thus demonstrated that the results are statistically robust. We also subjected the design of the experiment and data analysis to a contractor-led independent peer-review process in accordance with EPA's peer review guidance. The results of the peer review 305 (306) largely supported the study design, statistical analyses, and the conclusions from the program and raised only minor concerns that have not changed the overall conclusions and have subsequently been addressed in the final version of the report. (307)

Overall, the reductions found in this study are in agreement with other low sulfur studies conducted on Tier 2 vehicles, namely MSAT and Umicore studies mentioned above, in terms of the magnitude of NO X and HC reductions when switching from 28 ppm to 5 ppm fuel. (308) We have reviewed the results of the emission effects study performed by SGS, which was included with API's comments on the Tier 3 proposal, and have concluded that these results are also consistent with the findings of EPA's Tier 2 in-use study, specifically that exhaust emissions performance is sensitive to fuel sulfur level. (309) The SGS study also suggests that negative effects of exposure to a somewhat higher sulfur level (80 ppm in this case) are largely reversible for Tier 2 vehicles, meaning that reducing fuel sulfur levels nationwide will bring significant immediate benefits by reducing emissions of the existing fleet. For further details regarding the Tier 2 In-Use Gasoline Sulfur Effects Study, see the final report. (310)

As a follow-on phase to the Tier 2 in-use study, EPA analyzed five vehicles  (311) certified to Tier 2 Bin 4, LEV II ULEV and LEV II SULEV exhaust emissions standards to assess the gasoline sulfur sensitivity of Tier 2 and California LEV II vehicles with emission levels approaching or comparable to the Tier 3 standards. The analysis found that these low-emitting Tier 2 vehicles showed similar or greater sensitivity to fuel sulfur levels compared to the original Tier 2 test fleet—for example, a 24 percent reduction in FTP composite NO X emissions when sulfur is reduced from 28 ppm to 5 ppm. (312) Test results discussed below in section IV.A.6.d also confirm that there is significantly increased sensitivity of exhaust emissions to gasoline sulfur as vehicle technologies advance towards exhaust emissions approaching near-zero emissions (e.g., Tier 3 Bin 50 and lower). The impact of fuel sulfur on vehicles with exhaust emission control technologies that we expect to be used with Tier 3 vehicles is summarized in the next two sections (Preamble IV.A.6.d and e).

EPA believes that the studies by EPA and others described in this section strongly support our conclusion that reducing gasoline sulfur content to a 10 ppm average will result in significant exhaust emissions reductions from the current in-use fleet. However, some commenters have expressed concerns about the relevance and appropriateness of the data, as well as the conclusions drawn from them. The Summary and Analysis of Comments document, available in the docket for this rulemaking, provides our responses to those comments.

d. Testing of Gasoline Sulfur Effects on Vehicles With Tier 3/LEV III Technology

The Tier 3 fleet average exhaust emissions standards of 30 mg/mi NMOG+NO X will require large reductions of emissions across a broad range of light-duty vehicles and trucks with differing degrees of utility. Previous studies of sulfur impacts on extremely low exhaust emission vehicles (e.g., Toyota, Umicore) were limited to mid-size or smaller light-duty vehicles. There are currently no LDT3 or any LDT4 vehicles certified at or below Federal Tier 2 Bin 3 or to the California LEV II SULEV exhaust emission standards with the exception of a single hybrid electric SUV. At the time of the Tier 3 NPRM, EPA was not aware of any existing data demonstrating the impact of changes in gasoline sulfur content on larger vehicles with technology comparable to what would be expected for compliance with Tier 3 exhaust emission standards. In their supplemental comments to the Tier 3 proposal, API criticized EPA's reliance on emissions data from older vehicles that were not considered to be examples of future Tier-3-like vehicles. In order to further evaluate this issue, the Agency initiated a test program at EPA's National Vehicle and Fuel Emissions Laboratory (NVFEL) in Ann Arbor, Michigan. The Agency obtained a heavy-light-duty truck and applied changes to the design and layout of the exhaust catalyst system and to the calibration of the engine management system consistent with our engineering analyses of technology necessary to meet Tier 3 Bin 30 emissions with a 20 to 40% compliance margin at 150,000 miles. EPA also requested that Umicore loan the Agency the vehicle tested in their study to undergo further evaluation of gasoline sulfur impacts on exhaust emissions. In addition, Ford Motor Company completed testing of fuel sulfur effects on a Tier 3/LEV III developmental heavy-light-duty truck and submitted a summary report of their findings as part of their supplemental comments to the Tier 3 NPRM. The results of these three test programs are summarized below.

i. Ford Motor Company Tier 3 Sulfur Test Program

Ford Motor Company recently completed testing of a heavy-light-duty truck (i.e., between 6,000 and 8,500 pounds GVWR) under development to meet the Tier 3 Bin 50 standards on two different fuel sulfur levels and submitted the resulting data to EPA as part of its supplemental comments. (313 314) The test results from this vehicle are particularly important when considering the following factors:

  • These are the first detailed emissions data submitted by a vehicle manufacturer to the Agency demonstrating emissions of a heavy-light-duty-truck consistent with Tier 3 Bin 50 or lower emissions levels.
  • The truck tested uses a version of Ford's 2.0 L GTDI engine, an engine with high BMEP (approximately 23-bar) that can allow significant engine displacement downsizing while maintaining the truck's utility. This is a key enabling GHG reduction strategy analyzed by EPA in the 2017-2025 GHG Final Rule. (315)
  • The vehicle was specifically under development by a vehicle manufacturer with an engineering target of meeting Tier 3 Bin 50 and LEV III ULEV50 exhaust emissions standards.

Turbocharged, downsized engines are key technologies within Ford's strategy to reduce GHG emissions. (316) EPA expects that trucks with configurations similar to this developmental Ford Explorer (downsized engines with reduced GHG emissions and very low emissions of NMOG+NO X) will become increasingly prevalent within the timeframe of the implementation of the Tier 3 regulations.

The developmental truck used close-coupling of both catalyst substrates and relatively high PGM loading (150 g/ft 3). Ford used accelerated aging of the catalysts and O 2 sensors to an equivalent of 150,000 miles (the Tier 3 full useful life). The developmental hardware and engine management calibration configuration of this truck was designed to meet federal Tier 3 Bin 50 and California LEV III ULEV50 standards of 50 mg/mi NMOG+NO X at 150,000 miles. The emissions data submitted by Ford included NO X and NMHC emissions during operation on E10 California LEV III certification fuel at two different sulfur levels, 10 ppm and 26.5 ppm. Ford did not provide NMOG emissions data but there was sufficient information for EPA to calculate NMOG emissions from the provided NMHC data using calculations from Title 40 CFR 1066.665.

The truck demonstrated average FTP NMOG+NO X emissions of 37 mg/mi on the 10 ppm E10 California LEV III fuel, emissions that are consistent with compliance with Bin 50 and ULEV50 standards with a reasonable margin of compliance (emissions at approximately 70% of the standard). Retesting of the same vehicle on LEV3 E10 blended  (317) to 26.5 ppm S resulted in average NMOG+NO X emissions of 53 mg/mi, 6% above the Tier 3 Bin 50 standard. Ford found a high level of statistical significance with respect to the increase of emissions with increasing fuel sulfur. Assuming a linear effect of sulfur on emissions performance, NMOG+NO X emissions would be approximately 56 mg/mi at 30 ppm sulfur, which is approximately 12% above the Bin 50 exhaust emissions standard. This also represents an increase in NMOG+NO X emissions of 53% with an approximate doubling of NO X emissions and a 13% increase in NMOG for 30 ppm sulfur gasoline vs. 10 ppm sulfur gasoline.

The advanced technology Ford truck, which was shown to be capable of complying with the Tier 3 Bin 50 standard with a reasonable margin of compliance on 10 ppm sulfur gasoline, in effect reverted to approximately LEV II ULEV exhaust emissions levels when tested on higher sulfur gasoline, equivalent to the previous level of emissions control to which earlier models of this vehicle were certified for MY 2013. The effect of increasing gasoline sulfur levels from 10 ppm to 30 ppm  (318) on this vehicle essentially negated the entire benefit of the advances in emissions control technology that were applied by the vehicle manufacturer to meet developmental goals for compliance with Tier 3 standards. This clearly indicates, for this vehicle model using technology representative of what would be expected for compliance with Tier 3 Bin 50 and post 2017 GHG standards, reducing gasoline sulfur to 10 ppm is needed for the advances in technology to achieve their intended effectiveness in reducing NMOG+NO X emissions. The advances in vehicle technology and the reduction in gasoline sulfur clearly are both needed to achieve the emissions reductions called for by Tier 3.

ii. EPA Re-Test of Umicore 2009 Chevrolet Malibu PZEV

Ball et al. of Umicore Autocat USA, Inc. previously studied the impact of gasoline fuel sulfur levels of 3 ppm and 33 ppm on the emissions of a 2009 Chevrolet Malibu PZEV. (319) In their supplemental comments, API commented that the composition of the two test fuels outside of sulfur content was not held constant and thus the exhaust emissions differences attributed to the difference in gasoline sulfur levels may have been due to other fuel property differences. For example, the 3 ppm fuel used by Ball et al. was nonoxygenated EEE Clear test fuel (essentially, Tier 2 Federal certification gasoline except with near-zero sulfur) while the 33 ppm fuel was an oxygenated California Phase 2 LEV II certification fuel. Thus it was not entirely clear if the changes in NO X emissions observed between tests with the two fuels were significantly impacted by fuel composition variables other than gasoline sulfur content. EPA obtained the same test vehicle from Umicore for retesting at the EPA NVFEL facility using the 5 ppm and 28 ppm sulfur E0 test fuels and vehicle test procedures used in EPA gasoline sulfur effects testing on Tier 2 vehicles (see Section IV.6.b).

In EPA's retest of the 2009 Chevrolet Malibu PZEV, when sulfur was the only difference between the test fuels, the gasoline with higher sulfur resulted in significantly higher increases in NO X emissions with increasing fuel sulfur content than was observed in the previous testing by Ball et al. at Umicore. Assuming emissions impacts vs. gasoline sulfur content are approximately linear, the original data from Ball et al. would have resulted in a predicted increase in NO X emissions of approximately 40% when increasing gasoline sulfur from 10 ppm to 30 ppm. The EPA re-testing of the same vehicle that controlled for other fuel composition differences resulted in a predicted increase in NO X emissions of 93% when increasing gasoline sulfur from 10 ppm to 30 ppm, with NO X emissions approximately doubling from 22 g/mi to 43 g/mi, with no statistically significant difference in NMOG emissions and with an increase in NMOG+NO X emissions of 56%. The approximate doubling in NO X emissions with the Malibu PZEV between 10 ppm and 30 ppm sulfur was nearly identical to the results found during testing of the Tier 3 Bin 50 developmental Ford Explorer discussed above. The results confirm that fuel compositional differences other than sulfur may have impacted exhaust emissions results in the Ball et al. study by underreporting a substantial portion of the effect of increased sulfur on NO X emissions. When controlling for other fuel composition differences, the resultant increase in NO X exhaust emissions due to increasing gasoline sulfur was more than double that observed in the original Ball et al. study. The observed increase in NMOG+NO X emissions during EPA testing of the Malibu PZEV was also comparable to results found with the developmental Tier 3 Bin 50 Ford Explorer. There was also a much higher increase in NO X and NMOG+NO X emissions for both the Malibu PZEV and the Tier 3 Bin 50 Explorer with increased gasoline sulfur than was observed with Tier 2 vehicles in the EPA Tier 2 in-use study. (See also Chapter 1.2.4 of the RIA)

iii. EPA Prototype Tier 3 Heavy-Light-Duty Truck Test Program

EPA purchased a 2011 Chevrolet Silverado heavy-light-duty (LDT4) pickup truck with a developmental goal of modifying the truck to achieve exhaust emissions consistent with compliance with the Tier 3 Bin 30 emissions standards. The truck was equipped with a 5.3L V8 with General Motors' “Active Fuel Management” cylinder deactivation system. This particular truck was chosen in part because cylinder deactivation is a key technology for light-truck compliance with future GHG standards and in part because it achieved very low emissions in its OEM, Tier 2-compliant configuration (certified to Tier 2 Bin 4). A prototype exhaust system was obtained from MECA consisting of high-cell-density (900 cpsi) thin-wall (2.5 mil), high-PGM, close-coupled Pd-Rh catalysts with an additional under-body Pd-Rh catalyst. The total catalyst volume was approximately 116 in 3 with a specific PGM loading of 125 g/ft 3 and approximate loading ratio of 0:80:5 (Pt:Pd:Rh). Third-party (non-OEM) EMS calibration tools were used to modify the powertrain calibration in an effort to improve catalyst light-off performance. The final test configuration used approximately 4 degrees of timing retard and approximately 200 rpm higher idle speed relative to the OEM configuration during and immediately following cold-start. The exhaust catalyst system and HEGO sensors were bench aged to an equivalent 150,000 miles using standard EPA accelerated catalyst bench-aging procedures. (320) The truck was tested on California LEV III E10 certification fuel at 9 and 29 ppm gasoline sulfur levels.

The EPA Tier 3 prototype Silverado achieved NMOG+NO X emissions of 18 mg/mi on the 9 ppm S fuel. The NMOG+NO X emissions were approximately 60% of the Bin 30 standard and thus are consistent with meeting the Tier 3 Bin 30 exhaust emissions standard with a moderate compliance margin. NMOG+NO X emissions increased to 29 mg/mi on the 29 ppm S fuel and one out of four tests exceeded the Bin 30 exhaust emissions standards. NMOG+NO X emissions would be at 19 mg/mi and 30 mg/mi with 10 ppm and 30 ppm gasoline sulfur, respectively, assuming a linear effect of sulfur on emissions performance. This represents an increase in NMOG+NO X emissions of approximately 55%, comparable to increases observed with both the EPA-tested Chevrolet Malibu PZEV and the developmental Tier 3 Bin 50 Ford Explorer. The impact of increased gasoline sulfur on NMOG+NO X emissions was due to comparable increases (on a percentage basis) in both NMOG and NO X emissions. This effect of gasoline sulfur on the Prototype Silverado truck's emissions differed from the sulfur impacts observed on the developmental Ford Explorer, which primarily affected NO X emissions, and the Malibu PZEV, where the impact was entirely on NO X emissions.

e. Gasoline Sulfur Level Necessary for New Light-Duty Vehicles To Achieve Tier 3 Exhaust Emissions Standards

Meeting Tier 3 NMOG+NO X standards will require major reductions in exhaust emissions across the entire fleet of new light-duty vehicles. As discussed in previous sections, the Tier 3 program will require reductions in fleet average NMOG+NO X emissions of over 80 percent for the entire fleet of light-duty vehicles and light-duty trucks. This significant level of fleet average emission reduction will require reductions from all parts of the fleet, including vehicles models with exhaust emissions currently at or near the level of the fully phased-in Tier 3 FTP NMOG+NO X fleet average standard of 30 mg/mi.

Compliance with the more stringent Tier 3 fleet average standards will require vehicle manufacturers to certify a significant amount of vehicles to bin standards that are below the Bin 30 fleet average standard to offset other vehicles that are certified to bin standards that remain somewhat above the Bin 30 fleet average even after significantly reducing their emissions. At the same time, the stringency of the Tier 3 standards will push almost all vehicle models to be close to or below the Bin 30 fleet average standard. There are only 2 compliance bins below Bin 30, i.e., Bin 20 and Bin 0, available to offset emissions of vehicles certifying above Bin 30. There is also very limited ability for vehicle manufacturers to certify vehicles below the stringent Tier 3 fleet average exhaust emissions standard since Bin 20 and Bin 30 standards for individual vehicle certification test groups are approaching the engineering limits of what can be achieved for vehicles using an internal combustion engine and Bin 0 can only be achieved by electric-only vehicle operation. The result is that there is a very limited ability to offset sales of vehicles certified above the 30 mg/mi fleet average emission standard. This means in general that vehicle models currently with higher emissions will have to achieve significant emissions reductions to minimize the gap, if any, between their certified bin levels under Tier 3 and the Tier 3 Bin 30 fleet average standard, and vehicle models currently at or below Bin 30 will also have to achieve further emissions reductions under Tier 3 to offset the vehicles that remain certified to bin standards somewhat above Bin 30l. The end result is a need for major reductions from all types of vehicles in the light-duty fleet, including those above as well as most vehicles that are already near, at, or below the Tier 3 Bin 30 fleet average standard.

Achieving exhaust emissions reductions of over 80% for the fleet, with major reductions across all types of light-duty vehicles and light-duty trucks, will be a major technological challenge. Vehicles already have made significant advances in controlling cold start emissions and maximizing exhaust catalyst efficiency (e.g., improving warm-up and catalyst light-off after cold starts and maintaining very high catalyst efficiency once warmed up) in order to meet Tier 2 and LEV II emissions standards. There are no “low-hanging fruit” remaining for additional NMOG+NO X reductions from light-duty vehicles from a technology perspective, meaning that vehicle manufacturers cannot merely change one aspect of emissions control and thereby achieve all of the required reductions. Instead, compliance with light-duty Tier 3 exhaust emissions standards will require significant improvements in all areas of emissions control—with further improvements in fuel-system management and mixture preparation during cold start, improvements in achieving catalyst light-off immediately after cold start, and improved catalyst efficiency during stabilized, fully-warmed-up conditions. Manufacturers will need further improvements in each of these areas with nearly every vehicle in order to comply with the fleet-average Tier 3 standards.

From a technology perspective, the most likely control strategies will involve using exhaust catalyst technologies and powertrain calibration primarily focused on reducing cold-start emissions of NMOG, and on reducing both cold-start and warmed-up (running) emissions of NO X. An important part of this strategy, particularly for larger vehicles having greater difficulty achieving cold-start NMOG emissions control, will be to reduce NO X emissions to near-zero levels. This will involve controlling engine-out NO X emissions during cold start, shortening the cold start period prior to catalyst light-off of NO X reduction reactions, and better controlling NO X emissions once the catalyst is fully warmed up. This is needed to allow a sufficient NMOG compliance margin so that vehicles can meet the combined NMOG+NO X emissions standards for their full useful life.

While significant NMOG+NO X emissions reductions can be achieved from better control of cold start NMOG emissions, there are practical engineering limits to NMOG control for larger displacement vehicles (e.g., large light-duty trucks with significant payload and trailer towing capabilities). This is based in part on the impact on NMOG emissions of the larger engine surface-to-volume ratio and resultant heat conduction from the combustion chamber during warm-up. There are also tradeoffs between some cold-start NMOG controls and cold-start NO X control. For example, secondary air injection and/or leaner fueling strategies improve catalyst light-off for NMOG after a cold-start but also place OSC components in an oxidation state that limits potential for NO X reduction and thus often result in higher cold-start NO X emissions. Some applications achieve lower NMOG+NO X emissions without the use of secondary air injection by careful calibration, changes to the catalyst formulation and balancing of catalyst HC and NO X activity. The EPA Prototype Silverado and the developmental Ford Explorer are specific examples of this approach.

Because of engineering limitations with large vehicles, heavy-light-trucks and other vehicles with significant utility, we expect many applications will need close to 100% efficiency in NO X control under fully warmed-up conditions and very fast light-off of NO X reduction reactions over the exhaust catalyst almost immediately after cold-start for those applications. This will require significant improvements in catalytic and engine-out NO X reduction compared with Tier 2 vehicles and will be especially important for heavier vehicles due to the challenges of achieving low NMOG.

These technology improvements—improving warm-up and catalyst light-off after cold starts and maintaining very high catalyst efficiency—once warmed up—all rely on 10 ppm average sulfur fuel to achieve the very significant emissions reductions required for the fleet to achieve the Tier 3 Bin 30 fleet average emissions standard. The evidence from the test results and specific vehicle examples discussed above clearly indicate that leaving the gasoline sulfur level at 30 ppm would largely negate the benefits of key technology improvements expected to be used for compliance with Tier 3 exhaust emissions standards. Without the lower 10 ppm gasoline sulfur content, the Tier 3 exhaust fleet average emissions standards would not be achievable across the broad range of vehicles that must achieve significant exhaust emissions reductions.

One aspect of the need for sulfur levels of 10 ppm average stems from the fact that achieving the Tier 3 emission standards will require very careful control of the exhaust chemistry and exhaust temperatures to ensure high catalyst efficiency. The impact of sulfur on OSC components in the catalyst makes this a challenge even at relatively low (10 ppm) gasoline sulfur levels. NO X conversion by exhaust catalysts is strongly influenced by the OSC components like ceria. Ceria sulfation may play an important role in the large degradation of NO X emission control with increased fuel sulfur levels observed in the MSAT, Umicore and EPA Tier 2 In-Use Gasoline Sulfur Effects studies and the much more severe NO X emissions degradation observed in recent test data from PZEV and prototype/developmental Tier 3/LEV III vehicles. (321)

The importance of lower sulfur gasoline is also demonstrated by the fact that vehicles certified to California SULEV are typically certified to higher bins for the federal Tier 2 program. Light-duty vehicles certified to CARB SULEV and federal Tier 2 Bin 2 exhaust emission standards accounted for approximately 3.1 percent and 0.4 percent, respectively, of vehicle sales for MY2009. Light-duty vehicles certified to SULEV under LEV II are more typically certified federally to Tier 2 Bin 3, Bin 4 or Bin 5, and vehicles certified to SULEV and Tier 2 Bins 3-5 comprised approximately 2.5 percent of sales for MY2009. In particular, nonhybrid vehicles certified in California as SULEV are not certified to federal Tier 2 Bin 2 emissions standards even though the numeric limits for NO X and NMOG are shared between the California LEV II and federal Tier 2 programs for SULEV and Bin 2. Confidential business information shared by the auto companies indicate that the primary reason is an inability to demonstrate compliance with SULEV/Bin 2 emission standards after vehicles have operated in-use on gasoline with greater than 10 ppm sulfur and with exposure to the higher sulfur gasoline sold nationwide. While vehicles certified to the LEV II SULEV and Tier 2 Bin 2 standards both demonstrate compliance using certification gasoline with 15-40 ppm sulfur content, in-use compliance of SULEV vehicles in California occurs after significant, sustained operation on gasoline with an average of 10 ppm sulfur and a maximum cap of 30 ppm sulfur while federally certified vehicles under the Tier 2 program operate on gasoline with an average of 30 ppm sulfur and a maximum cap of 80 ppm sulfur. Although the SULEV and Tier 2 Bin 2 standards are numerically equivalent, the increased sulfur exposure of in-use vehicles certified under the federal Tier 2 program results in a need for a higher emissions compliance margin to take into account the impact of in-use gasoline sulfur on full useful life vehicle emissions. As a result, vehicles certified to California SULEV typically certify to emissions standards under the federal Tier 2 program that are 1-2 certification bins higher (e.g., SULEV certified federally as Tier 2 Bin 3 or Bin 4) in order to ensure in-use compliance with emissions standards out to the full useful life of the vehicle when operating on higher-sulfur gasoline.

There are currently no LDTs larger than LDT2 with the exception of a single hybrid electric SUV certified to Tier 2 Bin 2 or SULEV emissions standards. We expect that additional catalyst technologies, for example increasing catalyst surface area (volume or substrate cell density) and/or increased PGM loading, will need to be applied to larger vehicles in order to achieve the catalyst efficiencies necessary to comply with the Tier 3 standards, and any sulfur impact on catalyst efficiency will have a larger impact on vehicles and trucks that rely more on very high catalyst efficiencies in order to achieve very low emissions. The vehicle emissions data referenced in Section IV.A.6.d represents the only known data on non-hybrid vehicles spanning a range from mid-size LDVs to heavy-light-trucks at the very low criteria pollutant emissions levels that will be needed to comply with the Tier 3 exhaust emissions standards. The developmental Ford Explorer, Chevrolet Malibu PZEV and EPA prototype Chevrolet Silverado vehicles described in section IV.A.6.c also represent a range of different technology approaches to both criteria pollution control and GHG reduction (e.g., use of secondary air vs. emphasizing cold-start NO X control, use of engine downsizing via turbocharging vs. cylinder deactivation for GHG control, etc.) and represent a broad range of vehicle applications and utility (mid-size LDV, LDT3, LDT4). All of the vehicles with Tier 3/LEV III technology demonstrated greater than 50% increases in NMOG+NO X emissions when increasing gasoline sulfur from 10 ppm to 30 ppm. Two of the vehicles showed a doubling of NO X emissions when increasing gasoline sulfur from 10 ppm to 30 ppm. Both of the heavy-light-duty trucks with specific engineering targets of meeting Tier 3 emissions were capable of meeting their targeted emission standards with a sufficient compliance margin on 10 ppm sulfur gasoline and could not meet their targeted emissions standards or could not achieve a reasonable compliance margin when tested with 30 ppm sulfur gasoline.

The negative impact of gasoline sulfur on catalytic activity and the resultant loss of exhaust catalyst effectiveness to chemically reduce NO X and oxidize NMOG occur across all vehicle categories. However, the impact of gasoline sulfur on the effectiveness of exhaust catalysts to control NO X emissions in the fully-warmed-up condition is particularly of concern for larger vehicles (the largest LDVs and LDT3s, LDT4s, and MDPVs). Manufacturers face the most significant challenges in reducing cold-start NMOG emissions for these vehicles. Because of the need to reach near-zero NO X emissions levels in order to offset engineering limitations on further NMOG exhaust emissions control with these vehicles, any significant degradation in NO X emissions control over the useful life of the vehicle would likely prevent some if not most larger vehicles from reaching a combined NMOG+NO X level low enough to comply with the 30 mg/mi fleet-average standard. Any degradation in catalyst performance due to gasoline sulfur would reduce or eliminate the margin necessary to ensure in-use compliance with the Tier 3 emissions standards. Certifying to a useful life of 150,000 miles versus the current 120,000 miles will further add to manufacturers' compliance challenge for Tier 3 large light trucks (See Section IV.A.7.c below for more on the useful life requirements.) These vehicles represent a sufficiently large segment of light-duty vehicle sales now and for the foreseeable future such that their emissions could not be sufficiently offset (and thus the fleet-average standard could not be achieved) by certifying other vehicles to bins below the fleet average standard.

As discussed above, achieving Tier 3 levels as an average across the light-duty fleet will require fleet wide reductions of approximately 80%. This will require significant reductions from all light duty vehicles, with the result that some models and types of vehicles will be at most somewhat above the Tier 3 level, and all other models will be at or somewhat below Tier 3 levels. Achieving these reductions presents a major technology challenge. The required reductions are of a magnitude that EPA expects manufacturers to employ advances in technology in all of the relevant areas of emissions control—reducing engine-out emissions, reducing the time to catalyst lightoff, improving exhaust catalyst durability at 120,000 or 150,000 miles and improving efficiency of fully warmed up exhaust catalysts. All of these areas of emissions control need to be improved, and gasoline sulfur reduction to a 10 ppm average is a critical part of achieving Tier 3 levels through these emissions control technology improvements.

The use of 10 ppm average sulfur fuel is an essential part of achieving Tier 3 levels while applying an array of advancements in emissions control technology to the light-duty fleet. The testing of Tier 2 and Tier 3 type technology vehicles, as well as other information, shows that sulfur has a very large impact on the effectiveness of the control technologies expected to be used in Tier 3 vehicles. Without the reduction in sulfur to a 10 ppm average, the major technology improvements projected under Tier 3 would only result in a limited portion of the emissions reductions needed to achieve Tier 3 levels. For example, without the reduction in sulfur from a 30 ppm to 10 ppm average, the technology improvements would not come close to achieving Tier 3 levels. In some cases this may result in the same effectiveness as the current Tier 2 technology and achieve only approximately Tier 2 levels of exhaust emissions control.

Achieving Tier 3 levels without a reduction in sulfur to 10 ppm levels would only be possible if there were technology improvements significantly above and beyond those discussed above. Theoretically, without reducing sulfur levels to 10 ppm average, emissions control technology improvements would need to provide upwards of twice as much, and in some cases significantly more than twice as much, emissions control effectiveness as the Tier 3 technology improvements discussed above in Section IV.A.6.d. EPA has not identified technology improvements that could provide such a large additional increase in emissions control effectiveness, across the light-duty fleet, above and beyond that provided by the major improvements in technology discussed above, without any additional gasoline reductions in gasoline sulfur content. The impact of sulfur reduction on the effectiveness of the available technology improvements plays such a large role in achieving the Tier 3 levels that there would be no reasonable basis to expect that technology would be available, at the 30 ppm sulfur level, to fill the emission control gap left from no sulfur reduction, and achieve the very significant fleetwide reductions needed to meet the Tier 3 fleet average standards. In effect reducing sulfur from 30 ppm to 10 ppm has such a large impact on the ability of the technology improvements to achieve Tier 3 emissions levels that absent these sulfur reductions there is not a suite of technology advancements available to fill the resulting gap in emissions reductions. We cannot identify a technology path for vehicles that would achieve the Tier 3 Bin 30 average standard, across the fleet, with sulfur at 30 ppm levels, and as a result Tier 3 levels would not be technically feasible and achievable.

This analysis also applies to gasoline sulfur levels between 10 and 30 ppm, e.g., 20 ppm. The Tier 3 required emissions reductions are so large and widespread across the fleet, and the technology challenges are sufficiently high, especially for heavier vehicles, that the large increase in emissions that would occur from a higher average sulfur level compared to a 10 ppm average would lead to an inability for vehicle technologies to widely achieve Tier 3 levels as a fleet wide average in order to meet the Bin 30 fleet average standard.

EPA acknowledges that some models in the light-duty fleet, when viewed in isolation, may be able to achieve Tier 3 levels at current sulfur levels of 30 ppm average. Under the Tier 3 fleet average standards, it is not sufficient for one or a few of a manufacturer's vehicle models to meet Tier 3 levels because the manufacturer's light-duty vehicle fleet as a whole must achieve the Tier 3 30 mg/mi exhaust emissions standard as a fleet-wide average. As discussed above, all vehicle models will need to achieve further reductions and be either below or no more than somewhat above Tier 3 levels to achieve the Tier 3 standard as a fleet wide average. Absent the reductions in sulfur levels to 10 ppm average, this is not achievable from a technology perspective.

As discussed in Section V.B, the average 10 ppm gasoline sulfur standard is feasible and is the level that appropriately balances costs with the emission reductions that it provides and enables. Not only will a 10 ppm sulfur standard enable vehicle manufacturers to certify their entire product line of vehicles to the Tier 3 fleet average standards, but reducing gasoline sulfur to 10 ppm will better enable these vehicles to maintain their emission performance in-use over their full useful life. Higher sulfur levels would make it impossible for vehicle manufacturers to meet the Tier 3 standards, and would forego the very large immediate reductions from the existing fleet. Reducing the sulfur level below 10 ppm would further reduce vehicle emissions and allow the Tier 3 vehicle standards to be achieved more easily. However, we believe that a 10 ppm average standard is sufficient to allow vehicles to meet the Tier 3 standards. Further, as discussed in Sections V.B and IX.B there are significant challenges associated with reducing sulfur below 10 ppm.

7. Other Provisions
a. Early Credits

The California LEV III program is scheduled to begin at least two model years earlier than the federal Tier 3 program. (322) The Tier 3 standards begin in MY 2017 for vehicles 6,000 lbs GVWR and less, and in MY 2018 for vehicles over 6,000 lbs GVWR. As a result, LEV III vehicles sold in California beginning in MY 2015 will be required to meet a lower fleet average NMOG+NO X level than the federal fleet will be meeting at that time. In addition, the California NMOG+NO X standards will further decline before Tier 3 begins, resulting in the gap growing between the current federal program and LEV III.

We are finalizing an early credit program that with minor revisions is as we proposed. We have designed the early credit provisions to accomplish three goals: (1) To encourage manufacturers to produce a cleaner federal fleet earlier than otherwise required; (2) to provide valuable flexibility to the manufacturers to facilitate the significant “step down” from the current Tier 2 Bin 5 fleet average required in MY 2016 to the LEV III-based declining fleet average in MY 2017; and (3) to create an overall Tier 3 program that although starts later, is equivalent in stringency to the LEV III program such that manufacturers will be able to produce a 50-state fleet at the earliest opportunity. Commenters were generally supportive of or silent on the early credits program as proposed.

The early credit program we are finalizing includes several distinct provisions. The first provision allows manufacturers to generate early federal credits against the current Tier 2 Bin 5 requirement  (323) in MYs 2015 and 2016 for vehicles under 6,000 lbs GVWR and MYs 2016 and 2017 for vehicles greater than 6,000 lbs GVWR. Early credits will only be available to manufacturers that comply under the primary program (declining fleet average), not the alternative phase-in approach (Section IV.A.2.c above). In order to generate these credits, manufacturers sum the bin specific NMOG and NO X certification standards for each federally certified Tier 2 vehicle and the bin NMOG+NO X standards for any vehicle certified under the Early Tier 3 provision described below and calculate an NMOG+NO X fleet average for the entire manufacturers fleet sold in a model year. Credits are based on how far the fleet average is below the existing Tier 2 Bin 5 requirement (160 mg/mi total of NMOG and NO X). We expect that manufacturers will be able to achieve a fleetwide average below the Tier 2 Bin 5 level by several means, such as certifying LEV III vehicles either under Tier 2 or as Early Tier 3 vehicles under Tier 3 (discussed in the next section) to bin levels lower than Tier 2 Bin 5. Our analysis, presented in Section IV.A.5 above and Chapter 1 of the RIA, shows that manufacturers could certify many vehicles currently certified to Tier 2 Bin 5 to a lower bin—e.g., to Tier 2 Bin 3 or Bin 4—by simply accepting a relatively small reduction in compliance margins. Many manufacturers certify Tier 2 vehicles to Tier 2 Bin 5 but also certify the same vehicle to a cleaner emission standard under the LEV II program (e.g. ULEV) with only a compliance margin difference.

We believe that the early credit provision will help us realize both our first and second goals presented above. For example, a manufacturer certifying their federal fleet to Tier 2 Bin 4 will earn 50 mg/mi of NMOG+NO X credits per vehicle (i.e., 160 mg/mi minus 110 mg/mi), which we believe will encourage manufacturers to certify a cleaner federal fleet and provide a reasonable opportunity for credit generation to facilitate the “step down” in stringency.

At the same time, if we allowed manufacturers to generate excessive early credits, manufacturers might thereby delay their compliance with the Tier 3 program, and thus the harmonization with LEV III, for several years. This would be in direct conflict with our third goal of creating a program of equal stringency to the California program as early as possible. In order to address this concern, we proposed and are finalizing a provision limiting the application of the early Tier 3 credits to the following conditions:

  • Early Tier 3 credits generated as described above could be used without limitation in MY 2017 on the portion of the fleet entering the Tier 3 program in that MY.
  • Credits used for compliance in MY 2018 and beyond will be capped at an amount equal to the lesser of the manufacturer's federal credits as calculated above or the manufacturer's LEV III credits scaled up by the ratio of 50-state sales to California and LEV III required states sales. This limitation accounts for the fact that some LEV III credits may have begun to expire and will no longer be eligible as a basis for Tier 3 early credits.

By capping the available federal Tier 3 early credits, we believe that the two programs, LEV III and Tier 3 will be at parity in terms of relative stringency starting in MY 2018. In addition, because the number of Tier 3 early credits that can be used is based on the number of LEV III credits that the manufacturer has generated, there may be additional motivation for manufacturers to over-perform in California during the initial model years, accelerating emission reduction benefits.

Finally, we are adopting, as proposed, a limitation on the life of Tier 3 early credits to 5 years, with no discounting, consistent with the California LEV III program.

b. Early Tier 3 Compliance

We are finalizing, as proposed, the requirement that manufacturers begin the Tier 3 program in MY 2017 for vehicles up to 6,000 lbs GVWR and MY 2018 for vehicles above 6,000 lbs GVWR under the primary phase-in. The only proposed compliance approach available prior to MY 2017 was for manufacturers to continue to certify vehicles to the existing Tier 2 standards with the opportunity to earn early credits (see previous section) that could be used in MY 2017 and later.

Several auto industry commenters suggested additional provisions that could facilitate earlier harmonization between Tier 3 and LEV III and streamlining of development and certification of vehicle models. Specifically, these commenters requested the ability to have vehicles certified to the Tier 3 standards in MYs 2015 and 2016. They commented that this would allow them to develop, certify and sell a vehicle model for all 50 states, reducing the complexity of potentially different federal and California requirements in MYs 2015 and 2016. Additionally, commenters noted that the Tier 3 program provides more flexibility in the certification bin structure compared with the existing Tier 2, providing them additional opportunities to generate early credits.

To address this concern, we are finalizing a provision to allow manufacturers to certify to Tier 3 standards starting in MY 2015 as “Early Tier 3” vehicles. Manufacturers will have the option to certify their vehicle models to meet the Tier 3 emission requirements in MY 2015 and 2016 for all LDVs, LDTs, and MDPVs, which would have been required to begin in MY 2017 under the primary program. As an example, a manufacturer choosing to certify a vehicle as Early Tier 3 can bring the same vehicle models certified to LEV III standards  (324) in MY 2015 or 2016 into the Early Tier 3 program by meeting all the same requirements under the primary Tier 3 schedule. There would not be a Tier 3 fleet average requirement for FTP or SFTP in MY 2015 or 2016 (and 2017 for vehicles over 6,000 lbs GVWR and up to 8,500 and MDPVs) if all the same vehicle models certified to LEV III are also certified as the Early Tier 3 vehicles meeting the same LEV III emission standards and also the Tier 3 additional requirements (high altitude, and cold CO and hydrocarbons). These Early Tier 3 vehicles would replace any Tier 2 offering of the vehicle model consistent with the LEV III offering replacing the LEV II models. If a manufacturer chooses to certify only a portion of their LEV III vehicle models as Early Tier 3 vehicles in a given MY, they will be required to meet the LEV III fleet average requirements in that MY for those models certified as Early Tier 3 vehicles. All vehicles models not certified as Early Tier 3 vehicles must meet all Tier 2 requirements.

c. Useful Life

The “useful life” of a vehicle is the period of time, in terms of years and miles, during which a manufacturer is responsible for the vehicle's emissions performance. For the Tier 3 program, we are finalizing several changes to the existing useful life provisions that are appropriate to the new Tier 3 standards described above.

The auto manufacturing industry has uniformly expressed the desire to produce and sell a single national vehicle fleet, including a general ability and willingness of the industry to certify their vehicles to a 150,000 mile, 15 year full useful life, as required by the LEV III program. However, the CAA, written at a time when vehicles did not last as long as they do today, precludes EPA from requiring a useful life value longer than 120,000 miles (and 10 or 11 years, depending on vehicle category and weight) for lighter light-duty vehicles (LDVs and LDTs up to 3,750 lbs loaded vehicle weight (LVW) and up to 6,000 lbs GVWR (i.e., LDT1s)).

For heavier light-duty vehicles (i.e., LDT2s, 3s, 4s, as well as MDPVs, representing a large fraction of the light-duty fleet), this statutory restriction does not apply, and we are finalizing a 150,000 mile, 15 year useful life value, as proposed. For the lighter vehicles, we are continuing to apply the 120,000 mile (and 10 or 11 year, as applicable) useful life requirement from the Tier 2 program, also as proposed. For these lighter vehicles, manufacturers are allowed to choose to certify to either useful life value in complying with the fleet average. (325) In order for the Tier 3 NMOG+NO X standards to represent the same level of stringency regardless of which useful life value manufacturers choose, we proposed and are finalizing proportionally lower numerical values (85 percent of the NMOG+NO X 150,000 mile standards based on a data analysis in Chapter 1 of the RIA) for the declining fleet average FTP NMOG+NO X standards when a manufacturer chooses the 120,000 mile useful life. A manufacturer choosing the 120,000 mile useful life for any vehicle must maintain separate 120,000 mile and 150,000 mile useful life fleet averages for purposes of FTP NMOG+NO X fleet average compliance. Credits generated towards the required fleet averages are not transferable between the two useful life fleet averages.

We proposed that a manufacturer that certifies any vehicle model under the 120,000 mile provision be required to certify all their LDVs and LDT1s to the 120,000 mile useful life and associated numerically lower FTP NMOG+NO X fleet average standard. Comments from the auto industry expressed a concern that this approach would be inflexible to manufacturers' needs and unnecessarily burdensome. We have considered these comments, and we believe that the emission benefits of Tier 3 program will not be adversely affected if manufacturers are allowed to certify these lighter vehicles to the 120,000 mile useful life standards on a test group basis, and therefore we are finalizing this approach. Standards for all other pollutants  (326) and all other test cycles such as SFTP remain the same regardless of whether manufacturers choose the 120,000 mile or the 150,000 mile useful life periods.

For emission standards other than PM standards (e.g., NMOG+NO X standards), as proposed, manufacturers will be required to certify all vehicles to the 150,000 mile useful life beginning with the first model year that a vehicle model is certified to the FTP NMOG+NO X Bin 70 or lower (other than vehicles not yet required to meet a 150,000 mile useful life during the program phase in, and vehicles for which a manufacturer has the option and chooses to apply the 120,000 mile useful life value). This useful life requirement will apply as early as MY 2017. Beginning in MY 2020, all vehicles will need to certify to the 150,000 mile useful life for all emissions, regardless of NMOG+NO X certification bin, unless they are eligible for, and the manufacturer has chosen the 120,000 mile useful life and associated standards. (Note that the timing of the requirement to certify on the new test fuel follows the same approach as for the useful life requirement for emission standards other than PM standards (i.e., based on the first year a model is certified to FTP NMOG+NO X Bin 70 or below) as described in the next section.) For FTP and SFTP PM useful life requirements, manufacturers will be required to certify to 150,000 mile useful life for PM all vehicles that are included in the manufacturer's phase-in percentage meeting the new PM standards (other than eligible vehicles for which a manufacturer chooses to apply the 120,000 miles useful life value).

d. Test Fuels for Exhaust Criteria Emissions Standards

We recognize that test fuels are an important element of a national program. Vehicle manufacturers have emphasized in their comments the desire to reduce their test burdens by producing one vehicle that is tested on a single test procedure and on a single test fuel and that meets both California and federal requirements. Although we have been able to reasonably align the Tier 3 program with the LEV III program in most key respects, we recognize that the Tier 3 and LEV III test fuels are different, and that there may still exist some differences in emissions performance between vehicles tested on the two fuels. The largest difference between the two fuels is the Reid Vapor Pressure (RVP), and other differences in distillation properties and aromatic levels also exist (largely related to differences in actual in-use fuel nationally and in California). We are finalizing as proposed the requirement that manufacturers certify vehicles on the new Tier 3 E10 test fuels  (327) beginning with the first model year that a vehicle model is certified to the FTP NMOG+NO X Bin 70 or lower. (328) This requirement may apply as early as MY 2017 for vehicles up to 6000 lbs GVWR and MY 2018 for vehicles greater than 6000 lbs GVWR. (329) This requirement also applies to vehicles certified at Bin 70 and lower that are brought into the Tier 3 program under the Early Tier 3 option described in IV.A.7.b above, with the exception of the specific provision allowing the use of LEV III fuels discussed below. Beginning in MY 2020, all gasoline-fueled models will need to certify on the Tier 3 test fuels for all exhaust emission requirements, regardless of their certification bin. (330) As discussed in Section IV.A.7.c above, manufacturers must also meet the 150,000 mile useful life requirements for NMOG+NO X standards for these same vehicles as they are certified to Bin 70 and lower.

During the transition period from Tier 2 fuel to the new Tier 3 and LEV III E10 fuels, manufacturers have indicated that they face a substantial workload challenge of developing and certifying each vehicle model to the two new fuels simultaneously. We recognize this transitional challenge and are including an additional option. We are finalizing as proposed an option that vehicles certified in MYs 2015 through 2019 to California LEV III standards using California LEV III E10 certification test fuels and test procedures can be used for certifying to EPA Tier 2 or Tier 3 exhaust emission standards, including PM. A manufacturer may submit LEV III test data on vehicles tested using the new LEV III E10 fuels for Tier 2 or Tier 3 certifications. Consistent with existing Tier 2 policy, EPA may test vehicles certified to Tier 2 standards using LEV III test results on Tier 2 fuel for confirmatory or in-use exhaust testing. For vehicles certified in MY 2017 through 2019 to Tier 3 standards using LEV III E10 fuels, EPA will only use LEV III E10 fuels for confirmatory and in-use testing (except for high altitude or cold CO and hydrocarbons testing, as described below). Vehicles certified to the provisions of Early Tier 3 (Section IV.A.7.b above) will be treated the same as Tier 3 vehicles certified in MY 2017. For example, for MY 2015 and 2016, EPA will consider Early Tier 3 vehicles to be part of the Tier 3 program for purposes of fuel-related testing obligations. We will not accept test results using LEV II fuels for Tier 3 vehicle certification, including Early Tier 3 certifications, with the exception of the PZEV exhaust carry-over provision described below.

California does not have fuel specifications for high altitude testing or cold CO and hydrocarbon testing. For this reason, we are finalizing that for vehicles that manufacturers choose to certify using LEV III fuel and test procedures, manufacturers must use program-specific federal test fuels to comply with these federal-only requirements (i.e. Tier 2 vehicles will use Tier 2 fuel and Tier 3 vehicles will use Tier 3 fuel). Similarly, high altitude and cold CO and hydrocarbon confirmatory and in-use testing for these vehicles will be performed on the federal fuel that the manufacturer is required to use at certification as specified above regardless of whether LEV III or federal fuel is used for other testing.

We proposed the requirement that after MY 2019, all Tier 3 certification, confirmatory and in-use emission testing be required to use only the proposed Tier 3 E15 test fuel because it was believed to be a worst case fuel for emissions. Because we are finalizing Tier 3 E10 test fuels which are very similar as explained above to LEV III E10 test fuels, and not considered a worst case fuel, we are not finalizing the requirement for all testing to be performed on Tier 3 E10 test fuel. Instead, for certifications after MY 2019, EPA will continue to allow LEV III test results to be submitted for certification to Tier 3 standards, consistent with protocol under the Tier 2 program. However, if a manufacturer chooses to submit certification results for compliance with Tier 3 standards using the LEV III test fuel, then for confirmatory and in-use testing we will hold vehicles to the Tier 3 standards while using the Tier 3 fuel in addition to the LEV III test fuel; we will not allow new or carry-over certifications using LEV II or Tier 2 certification test fuels after MY 2019. CARB has indicated that they will accept Tier 3 test data (on federal certification test fuels) to obtain a California certificate as early as MY 2015. In this manner manufacturers should be able to avoid compliance testing on more than one fuel, since vehicles certified to Interim or Final Tier 3 status using federal certification test fuels could also obtain LEV III certification.

Auto industry commenters noted that the LEV III program provides an allowance for manufacturers to carry over PZEV-certified vehicle exhaust data  (331) from the LEV II program into LEV III compliance in MY 2015 through MY 2019. Thus, CARB allows these PZEV vehicles to use emission testing results using LEV II fuel (i.e. California Phase II test fuel) to meet the LEV III obligations. The commenters suggested that EPA allow manufacturers to carry over such PZEV 150,000 mile useful life exhaust emission data to meet the Tier 3 standards. We agree that this approach is appropriate during the transition, and we are finalizing this provision for MY 2015 through MY 2019, including allowing Early Tier 3 compliance at the Bin 30 level as a combined NMOG+NO X standard. EPA will hold vehicles certified using this provision to the Tier 3 emission requirements when they are tested on the LEV II fuel for confirmatory and in-use. Compliance testing of these vehicles for all other Tier 3 obligations (i.e., high-altitude testing and Cold CO and hydrocarbons testing) must be performed using Tier 3 fuel, and these vehicles will be required to meet the Tier 3 standards for Bin 30.

e. High Altitude Requirements

FTP emission standards are historically designed to be applicable at all altitudes. Under Tier 2, the same FTP emission bin standards applied to vehicles tested at both low and high-altitude. However, fundamental physical challenges exist at high altitude resulting in typically higher emissions during cold starts compared with starts at lower altitudes (i.e., sea level), and these challenges become more pronounced as emission standards become more stringent. This expected increase in emissions is primarily due to the lower air density at higher altitudes. Due to the lower air density, the needed volume of the hot combustion exhaust required to quickly heat the catalyst in the first minute after a cold start is reduced. As a result, catalyst light-off is delayed and cold start emissions can increase. Vehicles under the Tier 2 program typically have had sufficient compliance margins to absorb this increase in emissions during testing under high-altitude conditions. However, given the extremely low standards we are finalizing in Tier 3, manufacturers will have less compliance margin with which to address the issue.

Under the Tier 3 program, we expect that the emission control technologies selected for low altitude performance will also provide very significant emission control at high altitude. (332) However, as explained above, unique emission challenges exist with operation at higher altitude, often requiring manufacturers to design their emission controls specifically for higher altitude.

We do not believe that the impact of the fairly small fraction of overall U.S. driving that occurs in high altitude locations warrants a requirement for additional technologies to be applied specifically for high-altitude conditions. To avoid requiring manufacturers to use special high-altitude emission control technologies, we are allowing manufacturers limited relief for certification testing at high altitude, as proposed. Specifically, for sea-level certifications to Tier 3 Bins 20, 30, and 50, a manufacturer could comply with the next less-stringent bin for testing at high altitude. For example, a manufacturer can certify to Bin 50 for testing at high altitude versus Bin 30 at sea level). For vehicles certified at sea level to Bins 70 and 125, manufacturers can comply with standards 35 mg/mi higher (e.g., 105 mg/mi and 160 mg/mi, respectively. We are providing no high altitude relief for vehicles certified to Bin 160. This high altitude relief provision applies to all Final Tier 3 vehicles for the duration of the Tier 3 program.

For intermediate altitudes that fall between the specified low and high altitude test conditions, the emission performance should continue to be representative of the controls implemented to meet standards at the required altitude test conditions, consistent with Tier 2 protocol. Any deviation in the use of these controls at the intermediate altitudes may be considered an AECD that must be reported by the manufacturer and justified as not being a defeat device. (333)

Table IV-9 presents the Tier 3 high altitude standards.

Table IV-9—Tier 3 High Altitude Standards
BinSea level FTP standard(mg/mi NMOG+NO X)Altitude FTP standard(mg/mi NMOG+NO X)
Bin 160160160
Bin 125125160
Bin 7070105
Bin 505070
Bin 303050
Bin 202030
f. Highway Test Standards

Sustained high-speed operation can result in NO X emissions that may not be represented on either the FTP or SFTP cycles. Although we are not aware of any serious issues with this mode of operation with current Tier 2 vehicles, we are interested in preventing increases in these NO X emissions as manufacturers develop new or improved engine and emission control technologies.

For this reason, we are finalizing, as proposed, a provision that the Tier 3 FTP NMOG+NO X standards above also apply on the Highway Fuel Economy Test (HFET), which is performed as a part of GHG and Fuel Economy compliance testing. Thus, the Tier 3 FTP NMOG+NO X standard for the bin at which a manufacturer has chosen to certify a vehicle will also apply on the HFET test. For example, if a manufacturer certifies a vehicle to Bin 70, the vehicle's NMOG+NO X performance over the HFET could not exceed 70 mg/mi. Manufacturers will simply need to ensure that the same emission control strategies implemented for the FTP and SFTP cycles are also effective during the highway test cycle. We believe that this requirement will not require manufacturers to take any unique technological action, will not add technology costs, and will not add significantly to the certification burden.

g. Interim 4,000 Mile SFTP Standards

During the period of the declining NMOG+NO X standards, we are finalizing the proposed requirement that interim Tier 3 vehicles meet 4,000 mile SFTP standards, consistent with the existing Tier 2 and LEV II program requirements. The 4,000 mile standards apply to each vehicle model individually and to each component of the SFTP composite cycle. This approach is designed to prevent excessive emission levels from individual vehicle models being masked by the averaging of the manufacturer's fleet emissions. Similarly, this approach also prevents poor performance on a single cycle of the SFTP. We believe it is appropriate to require any individual Interim Tier 3 vehicle to at a minimum meet the existing requirements under the Tier 2 and LEV II programs. Table IV-10 below presents the 4,000 mile SFTP standards for interim Tier 3 vehicles.

Table IV-10—4,000 Mile SFTP Exhaust Standards for Interim Tier 3 Vehicles
Vehicle categoryUS06NMOG+NO X US06COSC03NMOG+NO X SC03CO
LDV/LDT10.148.00.202.7
LDT20.2510.50.273.5
LDT30.410.50.313.5
LDT40.611.80.444.0

We believe that vehicles considered to be Final Tier 3 vehicles (i.e., they meet the Tier 3 PM requirements, specifically the stringent SFTP PM standards) will have sufficiently robust designs that the 4,000 mile SFTP standards will no longer be necessary and so will not apply to those vehicles. Additionally, once the program reaches the fully phased-in fleet average composite standard of 50 mg/mi in 2025, high SFTP emissions even on a limited portion of a manufacturer's fleet should be effectively mitigated, and the 4,000 mile SFTP standards will no longer apply.

h. Phase-In Schedule

As proposed, the major provisions of the Tier 3 program phase in based on model year and on the emission levels to which manufacturers certify their vehicles. As described in Section IV.A.3, under the Tier 3 program, manufacturers are required to certify each vehicle model to an FTP bin, which is then used to calculate the NMOG+NO X fleet average of all of its Tier 3 vehicles. Manufacturers must also determine the SFTP levels of each model and calculate the NMOG+NO X fleet average for the SFTP requirements as described in Section IV.A.4. These separate FTP and SFTP fleet average calculations satisfy one aspect of certification under the Tier 3 program, specifically the standards associated with each model year.

As described in Sections IV.A.7.c and IV.A.7.d above, the longer (150,000 mile) useful life value, as applicable, and the new Tier 3 test fuel for exhaust testing will be implemented as manufacturers certify vehicles to more stringent NMOG+NO X standards, with the threshold to implement both of these provisions being Bin 70. Beginning in MY 2017, any vehicle certified to Bin 70 or lower will be required to be certified on Tier 3 test fuel. In addition, any vehicle certified to Bin 70 or lower that is required to meet the longer 150,000 mile useful life will be required to do so at that point. Independent of the Tier 3 test fuel phase in schedule, the 150,000 mile useful life for PM standards will be required when the vehicle is certified to the new Tier 3 PM standards as described below in the PM phase-in schedules. Beginning in MY 2020, all gasoline-fueled vehicles will be required to be certified for exhaust emissions on the Tier 3 test fuel, regardless of their certification bin or applicable useful life.

Manufacturers must also comply with more stringent PM standards on a percent phase-in schedule. Compliance with the PM standards, which is consistent with the CARB LEV III program, is independent of the NMOG+NO X fleet average requirements described above. The PM emission standards for FTP and SFTP described in Section IV.A.3 and 4 respectively will be implemented as a percent phase-in requirement as described below under a primary phase-in schedule or under an optional phase-in schedule.

Vehicle models that a manufacturer certifies to a Tier 3 NMOG+NO X bin, that meet the requirements of the PM phase-in schedule, and that comply with the other Tier 3 requirements (i.e., 150,000 mile useful life and Tier 3 test fuel, as applicable) will be considered “Final Tier 3” compliant vehicles. All other vehicles certified to Tier 3 bins but not yet meeting the PM and other Tier 3 requirements will be considered “Interim Tier 3” compliant vehicles. At the completion of the percent phase-in period for PM (2021 for the primary PM phase-in schedule and 2022 for the optional PM phase-in schedule, as described below), 100 percent of vehicles will need to meet all of the Tier 3 requirements and will be considered Final Tier 3 vehicles.

As proposed, for the PM requirements, each model year manufacturers must meet either the primary PM percent phase-in or the optional PM phase-in as described in the following subsections. The primary percent PM phase-in schedule is composed of fixed annual minimum phase-in percentages that we expect most manufacturers to choose in order to comply with the Tier 3 requirements. The optional PM phase-in schedule provides additional flexibility for manufacturers with too few product offerings to allow for a sufficiently gradual transition into the Final Tier 3 requirements, as described below. In either case, Interim Tier 3 vehicles not yet meeting the Tier 3 PM standards must at a minimum meet the Tier 2 PM full useful life FTP PM standard of 10 mg/mi and the SFTP PM weighted composite standard of 70 mg/mi.

i. Primary PM Percent Phase-In Schedule

It is important to note that the percent phase-in of the new Tier 3 PM standards and the declining fleet average NMOG+NO X standards that we are finalizing are separate and independent elements of the Tier 3 program. “Phase-in” in the context of Tier 3 PM standards means the fraction of a manufacturer's fleet that is required to meet the new Tier 3 PM standards in a given model year. We expect that manufacturer fleets may consist of a mix of vehicle models certified to Tier 2, LEV II, LEV III and Tier 3 standards throughout the percent phase-in period.

As discussed above, vehicles originally certified to Tier 2, LEV II, and LEV III may be carried over into the Tier 3 program as Interim Tier 3 vehicles. A vehicle will be considered a Final Tier 3 vehicle when it is certified to one of the Tier 3 bins, meets the new Tier 3 PM standards for FTP (3mg/mi) and US06 (10 or 6 mg/mi), certifies to the 150,000 useful life value (as applicable), and certifies on the new Tier 3 test fuel. Table IV-11 below presents the PM phase-in schedule for Final Tier 3 vehicles.

Table IV-11—PM Phase-In Schedule for Final Tier 3 Vehicles
Model year201720182019202020212022 and later
Manufacturer's Fleet (%)20 a 204070100100
Vehicle Types≤ 6,000 lbs GVWRAll vehicles ≤ 8,500 lbs GVWR and MDPVs    

The PM percent-of-sales phase-in schedule described above will allow manufacturers with multiple vehicle models to plan the phase-in of those models based on anticipated volumes of each vehicle model. However, manufacturers certifying only a few vehicle models might not benefit from this schedule. This is because, in order to satisfy the phase-in schedule percentages, they may have to over-comply with the required percentages earlier than will a manufacturer with many vehicle models available for the phase-in.

For instance, a manufacturer with only two models that each equally account for 50 percent of their sales will be required to introduce (at least) one of the models in MY 2017 to meet the PM phase-in requirement of 20 percent in the first year. Because it represents 50 percent of the manufacturer's sales, this model will then also meet the requirements for MY 2018 (20 percent) and MY 2019 (40 percent). To meet the MY 2020 requirement of 70 percent of sales, however, the manufacturer will need to introduce the second Tier 3 vehicle that year. Thus the manufacturer will have introduced 100 percent of its Tier 3 models one year earlier than required of a manufacturer that is able to delay the final 30 percent of its fleet until MY 2021 (by distributing its models over the entire phase-in period).

To provide for more equivalent phasing in of the PM requirements among all manufacturers in the early years of the program, we are finalizing, as proposed, an optional “indexed” PM phase-in schedule that can be used by a manufacturer to meet its PM percent phase-in requirements. A manufacturer that exceeds the phase-in requirements in any given year will be allowed to, in effect, offset some of the phase-in requirements in a later model year. The optional phase-in schedule will be acceptable if it passes a mathematical test. The mathematical test is designed to provide manufacturers a benefit from certifying to the standards at higher volumes than they are obligated to under the normal phase-in schedule, while ensuring that significant numbers of vehicles are meeting the new Tier 3 requirements during each year of the optional phase-in schedule. In this approach, manufacturers weight the earlier years by multiplying their percent phase-in by the number of years prior to MY 2022 (i.e., the second year of the 100 percent phase-in requirement).

The mathematical equation for applying the optional PM phase-in is as follows: (5 × APP2017) + (4 × APP2018) + (3 × APP2019) + (2 × APP2020) + (1 × APP2021) = 540, where APP is the actual phase-in percentage for the referenced model year.

The sum of the calculation must be greater than or equal to 540, which is the result when the optional phase-in equation is applied to the primary percent phase-in schedule (i.e., 5 × 20% + 4 × 20% + 3 × 40% + 2 × 70% + 1 × 100% = 540).

Applying the optional PM phase-in equation to the hypothetical manufacturer in the example above, the manufacturer can postpone its model introductions by one year each, to MY 2018 and MY 2021. Its calculation is (5 × 0% + 4 × 50% + 3 × 50% + 2 × 50% + 1 × 100% = 550, and thus the phase-in is acceptable.

i. In-Use Standards
i. NMOG+NO

The Tier 3 emission standards will require a substantial migration of emission control technology historically used only on a small percent of the fleet and typically limited to smaller vehicles and engines. While we believe that these technologies can generally be used on any vehicle and are applicable to the entire fleet, manufacturers have less experience with the in-use performance of these technologies across the fleet. For example, technologies that accelerate catalyst warm-up such as catalyst location close to the engine exhaust ports and other advanced thermal management approaches will be new to certain vehicle types, particularly larger vehicles (i.e., LDT3/4s), which have historically not relied on these technologies to meet emission standards.

As proposed, to help manufacturers address the lack of in-use experience and associated challenges with the expanded introduction of these technologies, particularly in the larger vehicles, we are finalizing temporarily-relaxed in-use NMOG+NO X standards that will apply to all vehicles certified to Bins 70 and cleaner as Interim or Final Tier 3 vehicles. The in-use standards will apply during the entire percent phase-in period (i.e., through MY 2021). The in-use standards are 40 percent less stringent than the certification standards, providing a significant but reasonable temporary cushion for the uncertainties associated with new technologies (or new applications of existing technologies) over the life of the vehicles.

The in-use NMOG+NO X standards are shown in Table IV-12.

Table IV-12—FTP In-Use Standards for Light Duty Vehicles and MDPVs
BinNMOG+NO X (mg/mi)
Bin 160160
Bin 125125
Bin 7098
Bin 5070
Bin 3042
Bin 2028
ii. PM

As with the NMOG+NO X standards, the introduction of new emission control technologies or new applications of existing technologies (e.g., GDI, turbocharging, downsized engines) will create significant uncertainties for manufacturers about in-use performance over the vehicle's useful life. We are finalizing as proposed a temporary in-use FTP standard for PM of 6 mg/mi for all light duty vehicles certified to the Tier 3 full useful life 3 mg/mi standard. Since the Tier 3 FTP PM standard has a percent phase-in schedule spread over several years, starting in 2017 with full phase-in completed in 2022, we are finalizing the requirement that the in-use standard apply to all vehicles certified to the new PM standards during the entire percent phase-in period (i.e., through MY 2021).

We also proposed temporarily-relaxed in-use US06 PM standards. As described in Section IV.A.4.b above, we are finalizing an in-use US06 PM standard of 10 mg/mi for the intermediate years of the program (MYs 2019 through 2023) in response to industry concerns about emissions variability as the new standards become effective.

j. FFVs

Because of the physical and chemical differences in how emissions are generated and controlled between vehicles operating on different blends of gasoline and ethanol, manufacturers of vehicles designed for high-percentage blends of ethanol (usually called Flexible Fuel Vehicles, or FFVs) may face unique compliance challenges under the Tier 3 program. Historically, under the Tier 2 program, FFVs have only been required to meet all Tier 2 emission standards, FTP and SFTP, while operating on gasoline (E0); when operating on the alternative fuel (generally this means a blend that is nominally 85 percent ethanol, or E85), they have only been required to meet the FTP emission standards.

However, E85 use may rise considerably in the future as ethanol use increases in response to the Renewable Fuels Standards (RFS). Thus, as the Tier 3 program is implemented, it is increasingly important that FFVs maintain their emission performance when operating on E85 across different operating conditions.

We believe that at standard test conditions, requiring manufacturers to meet the Tier 3 standards on any blend of gasoline and ethanol will not be significantly more challenging technologically than compliance on lower ethanol blends, including the E10 Tier 3 test fuel we are adopting. We are thus finalizing, as proposed, the requirement that in addition to complying with the Tier 3 requirements when operating on Tier 3 test fuel, FFVs also comply with both the FTP and the SFTP emission standards when operating on E85. This includes the requirement to meet emission standards for both Tier 3 test fuel and E85 for the FTP, highway test, and SFTP emission standards at standard test temperatures (i.e., 68 °F to 86 °F). Since FFVs can operate on any blend of gasoline and ethanol (up to a nominal 85 percent ethanol), the emission requirements apply to operation at all levels of the alternative fuel that can be achieved with commercially available fuels. However, for exhaust emission compliance demonstration purposes, we will test on Tier 3 test fuel and on fuel with the highest available ethanol content.

k. Credit for Direct Ozone Reduction (DOR) Technology

Since the late 1990s, technologies have been commercialized with which vehicles can remove ozone from the air that flows over the vehicle's coolant radiator. In such direct ozone reduction (DOR) technology, a catalytic coating on the radiator is designed to convert ambient ozone into gaseous oxygen, as a way of addressing the air quality concerns about ozone. Detailed technical analyses for the California LEV II and the federal Tier 2 programs showed that when properly designed these systems can remove sufficient ozone from the air to be equivalent to a quantifiable reduction in tailpipe NMOG emissions. In the earlier programs, both California and EPA provided methodologies through which a manufacturer could demonstrate the capability and effectiveness of the ozone-reducing technology and be granted an NMOG credit. A small number of vehicle models with DOR applications received credit under the LEV II program; no manufacturer formally applied for credits under the federal Tier 2 program.

Some manufacturers have expressed an interest in the continued availability of a DOR credit as a part of their potential LEV III and Tier 3 compliance strategies. EPA believes that when a DOR system is shown to be effective in reducing ozone, a credit toward Tier 3 compliance is warranted. We are finalizing a provision, as proposed, that manufacturers following the California methodology for demonstrating effectiveness and calculating a appropriate credit for a DOR system be granted a specific credit toward the NMOG portion of the NMOG+NO X standard. (334) As with the California program, such a credit may not exceed 5 mg/mi NMOG.

l. Credit for Adopting a 150,000-Mile Emissions Warranty

Under the Tier 3 standards, manufacturers are expected to design their emission control systems to continue to operate effectively for a useful life of 150,000 miles (120,000 miles for some smaller vehicles). However, manufacturers are only required to replace failed emission control components or systems on customers' vehicles for a limited time period, specified in the Clean Air Act (80,000 miles/8 years for key emission control components). EPA believes that voluntary extension of this warranty obligation by manufacturers would provide additional emission reductions by helping ensure that controls continue to operate effectively in actual operation through the full life of the vehicle.

We are finalizing as proposed that a manufacturer providing its customers with a robust emission control system warranty of 15 years or 150,000 miles be eligible for a modest credit of 5 mg/mi NMOG+NO X. (335) Because of the significant liability that manufacturers would be accepting, we do not expect that the use of this credit opportunity will be widespread. However, based on our modeling of the expected deterioration of the emissions of future Tier 3 vehicles absent repair/replacement of failed emission controls, we anticipate that the value to the environment of long emissions warranties in terms of reduced real-world emissions would significantly exceed the 5 mg/mi NMOG+NO X credit. (336)

We will use the same criteria for approving such a credit as does the parallel California program. (337) Thus, in addition to committing to customers that failing emission controls will be repaired or replaced for 15 years/150,000 miles, manufacturers will also need to accept the liability that in the event that a specific emissions control device fails on greater than 4 percent of a vehicle model's production, they will recall the entire production of that model for repair.

m. Averaging, Banking, and Trading of Credits

We proposed and are finalizing an averaging, banking, and trading (ABT) program similar to those that have historically been a part of most EPA emission control programs. For the Tier 3 final rule, the ABT program is consistent with the other Tier 3 program elements, the heavy duty exhaust emission standards and the evaporative emission standards programs, with the only exception being credit life during the longer phase in for the light duty program as described below. The ABT program is intended to provide an opportunity for manufacturers to deploy their Tier 3 vehicle models more efficiently, especially during the transition years, and to avoid excessive delays in the necessary technological improvements across the fleet. We have designed the Tier 3 ABT program to provide for credits to be generated by certifying vehicles that perform better than the fleet-average NMOG+NO X standards. These credits may be used within a company to offset vehicles that perform worse than the standards, they may be banked for later use, or they may be traded to other manufacturers.

We are also finalizing limitations on the use of credits for the light-duty fleet. We proposed that Tier 3 credits expire after 5 model years following the model year they are generated and solicited comment on the Tier 3 credit life. In communications regarding the proposed rule, representatives of the auto industry expressed to EPA that the value of the ABT program during the MY 2017-2025 phase-in of the primary program would be improved if credits had a longer credit life. (338) We determined that, with certain restrictions, Tier 3 credit life can be temporarily extended with no adverse impacts on the overall emission reductions of the program. Specifically, we are finalizing a credit life of 8 years for credits generated in MYs 2017-2022 for the FTP and SFTP NMOG+NO X fleet average standards for the primary program only. For the heavier light-duty vehicles, the 8-year credit life begins for credits generated in MY 2018. Note that, as proposed, credits generated under the Early Tier 3 Credit provision (Section IV.A.7.a) are limited to 5-year life, and are not affected by the longer credit life.

For credits generated in MYs 2023-2025, the credit life declines by one year of credit life annually, with credit life stabilizing at 5 years for credits generated in MYs 2025 and later. That is, credits generated in MY 2023 have a 7-year life, in MY 2024 a 6-year life, and in MY 2025 and later a 5-year life. However, while credits can be generated, banked, and used internally for the extended time periods, credits cannot be traded to other manufacturers after 5 years.

After considering the views expressed by manufacturers as well as the implementation schedules of this Tier 3 rule and the 2017 light-duty GHG rule, we believe that the temporary up-to-8-year credit life available to manufacturers during the phase-in period provides substantial flexibility to address manufacturer uncertainties about future technology development and product planning during implementation of the Tier 3 program. We also believe this longer credit life provision will alleviate most if not all concerns expressed by manufacturers with respect to the challenges they may encounter by simultaneous implementation of the two programs.

As proposed, we are finalizing a provision for a manufacturer to create a credit deficit, at certification or at the end of the production year, if its fleet average emissions exceed the standard. A manufacturer would be required to use all of its banked credits, if any, before creating a credit deficit. A credit deficit would need to be resolved before the fourth model year after the deficit was created; that is, a manufacturer may not maintain a credit deficit more than 3 consecutive model years.

n. Tier 3 Transitional Emissions Bins

During the development of the proposed rule and in their comments, manufacturers pointed out that they may continue to produce some vehicles as late as MY 2019 that could be certified to Tier 2 Bin 3 or Bin 4 standards. In order to provide manufacturers flexibility in meeting the fleet average standards and to further facilitate the transition, we will allow manufacturers to certify to the combined NMOG+NO X levels of these Tier 2 bins through MY 2019. We are finalizing two transitional Tier 3 bins, Bin 110 and Bin 85, that have FTP NMOG+NO X standards of 110 mg/mi and 85 mg/mi, respectively (i.e., the sum of the NMOG and NO X values from the Tier 2 bins). The associated FTP standards for CO, PM, and HCHO corresponding to these bins are identical to those for vehicles certified to the Tier 3 Bin 125. Tier 3 SFTP standards will apply to these vehicles, and these vehicles will be included in the Tier 3 PM percent phase-in calculations.

o. Compliance Demonstration

In general, we are finalizing requirements that manufacturers demonstrate compliance with the Tier 3 light-duty vehicle emission standards in a very similar manner to existing Tier 2 vehicle compliance (see § 86.1860 of the regulatory language). However, for Tier 3, manufacturers must calculate their compliance with the fleet average standards and percent phase-in standards based on annual nationwide sales, including sales in California and Clean Air Act Section 177 states. We believe that this approach represents another step toward achieving the goal of an effectively nationwide program as early as possible, which has been a basic principle in EPA's development of this program and broadly supported by vehicle manufacturers. We also believe that basing compliance on nationwide sales may reduce the need for manufacturers to project future sales and track past years' sales in a disaggregated way. Because the Tier 3 provisions become increasingly consistent with LEV III provisions as the Tier 3 program phases in, we believe that any disproportionate impacts of different mixes of vehicles in different states are unlikely to occur.

This nationwide compliance calculation approach applies to vehicles as they become subject to the Tier 3 provisions, either the declining fleet-average NMOG+NO X curves or the percent phase-in PM standards. Were any manufacturer to choose to use the alternative FTP and SFTP phase-ins, which are not a part of the LEV III program, the manufacturer would not include sales in California or in the Section 177 states in its compliance calculations.

B. Tailpipe Emissions Standards for Heavy-Duty Vehicles

1. Overview and Scope of Vehicles Regulated

After considering the comments we received, we are adopting the Tier 3 exhaust emissions standards that we proposed for chassis-certified heavy-duty vehicles (HDVs) between 8501 and 14,000 lbs gross vehicle weight rating (GVWR). Vehicles in this GVWR range are often referred to as Class 2b (8501-10,000 lbs) and Class 3 (10,001-14,000 lbs) vehicles, and are typically full-size pickup trucks and work vans certified as complete vehicles. (339) Medium-duty passenger vehicles (MDPVs), although in the Class 2b GVWR range, are subject to Tier 3 standards discussed in Section IV.A. To a large extent, we are also adopting the Tier 3 certification testing and compliance provisions that we proposed for HDVs. There are, however, a number of improvements we are making in response to comments, as discussed in detail below.

The Tier 3 program for HDVs will bring substantial reductions in harmful emissions from this large fleet of work trucks and vans, a fleet that is used extensively on every part of the nation's highway, rural, and urban roadway system. The fully-phased in Tier 3 standards levels for non-methane organic gas (NMOG) plus oxides of nitrogen (NO X), and for particulate matter (PM), are on the order of 60 percent lower than the current standards levels.

We proposed to require that diesel-fueled Class 2b and 3 complete vehicles, like their gasoline-fueled counterparts, be certified to the Tier 3 standards on the chassis test; we also proposed to include these vehicles in the Tier 3 HDV averaging, banking, and trading (ABT) program. Currently only gasoline-fueled Class 2b/3 complete HDVs are required to chassis certify.

The International Council for Clean Transportation (ICCT) provided comments in support of this requirement, arguing that it is needed to stop manufacturers from making trucks marginally above 8500 lbs GVWR to avoid light-duty emission standards. The Truck and Engine Manufacturers Association (EMA) opposed mandatory chassis certification for any class of engines or vehicles over 8500 lbs GVWR, arguing that the existing flexibility is needed to minimize unnecessary costs and certification burdens. EMA commented that, at a minimum, EPA should maintain optional certification of diesel engines used in complete Class 3 vehicles. In their joint comments, the Alliance of Automobile Manufacturers and the Association of Global Automakers also requested that EPA retain the option for complete Class 3 diesel vehicles and engines, arguing that otherwise manufacturers may be required to dual certify vehicle models that include variants both under and over 14,000 lbs.

We are sensitive to this issue but remain concerned that the fleet average standard program we are finalizing would not work well if a major fleet component, such as complete Class 3 diesel trucks, can be left in or taken out of the fleet calculation based on what each manufacturer considers to be most advantageous. We believe the resulting competitive issues and uncertainties would be problematic, given the wide variance in gasoline/diesel HDV sales among the manufacturers, our provision for averaging across each manufacturers' entire Class 2b/3 fleet, and the overwhelming preponderance of diesels in the Class 3 market. It would also create uncertainties in the Tier 3 environmental benefits, given the pronounced difference between these Tier 3 standards and the heavy-duty diesel engine standards we set 13 years ago, which we expect to remain in effect for the foreseeable future.

As a result, we are finalizing these provisions as proposed, except that we are providing that manufacturers, instead of certifying complete diesel Class 3 HDVs, may install diesel engines that have been engine-certified for any model year that the engine family has less than half of its sales being installed in such non-chassis-certified complete Class 3 vehicles. For example, if a company has a certified diesel engine family with 10,001 sales in MY 2020, up to 5,000 of those engines may be installed in complete Class 3 HDVs that are not chassis-certified for exhaust emissions. This provision is intended to help address manufacturers' concern about dual certification, while at the same time ensuring a coherent fleetwide standards regimen in this vehicle class. It also better harmonizes with California's low-emission vehicle (LEV) III program which does not mandate chassis certification for diesel Class 3 vehicles. By only allowing engine-certified vehicles in the case of engines that are primarily produced for other purposes, we believe this approach adequately guards against potential abuse. In the case of complete diesel Class 3 HDVs produced by a company other than the engine certifier, the responsibility for ensuring the sales limit is not exceeded remains with the vehicle manufacturer, who will need to coordinate with the engine supplier to ensure compliance.

Manufacturers of incomplete HDVs that are sold to secondary manufacturers for subsequent completion (less than 10 percent of the Class 2b and 3 U.S. market) are also allowed under existing EPA regulations to certify via either the chassis or engine test, and those who choose to chassis-certify in the future will be subject to Tier 3 requirements. We asked for comment on mandating chassis certification of incomplete Class 2b and 3 vehicles, noting that California's LEV III program includes such a requirement for Class 2b. Commenters expressed opposition to this extension of mandatory chassis certification, despite their general support for harmonization with LEV III; as a result, we are not mandating chassis certification for any incomplete HDVs.

The key elements of the Tier 3 program for HDVs parallel those for passenger cars and light-duty trucks (LDTs), with adjustments in standards levels, emissions test requirements, and implementation schedules, appropriate to this sector. These key elements include:

  • A combined NMOG+NO X declining fleet average standard beginning in 2018 and reaching the final, fully phased-in level in 2022,
  • creation of a bin structure for standards, including standards for carbon monoxide (CO) and formaldehyde,
  • PM standards phasing in separately on a percent-of-sales basis,
  • changes to the test fuel for gasoline- and ethanol-fueled vehicles,
  • extension of the regulatory useful life to 150,000 miles,
  • a new requirement to meet standards over the supplemental federal test procedure (SFTP) that addresses real-world driving modes not well-represented by the federal test procedure (FTP) cycle alone, and
  • special flexibility provisions for small businesses and small volume manufacturers described in Section IV.G.

As in the light-duty Tier 3 program, we have put a strong emphasis on coordinating HDV Tier 3 program elements with California's LEV III program for Class 2b and 3 vehicles, referred to in LEV III as medium-duty vehicles (MDVs). The goal is to create a coordinated “national program” in which California would accept compliance with Tier 3 standards as sufficient to also satisfy LEV III requirements, thus allowing manufacturers to comply nationwide by marketing a single vehicle fleet. As part of this effort, we proposed that manufacturers of Tier 3 HDVs calculate compliance with the fleet average standards and percent phase-in standards based on annual nationwide sales, including sales in California and in states implementing California standards under Clean Air Act section 177. Commenters expressed emphatic support for this approach and we are finalizing it as a key element of the Tier 3 program.

2. HDV Exhaust Emissions Standards
a. Bin Standards

Manufacturers will certify HDVs to Tier 3 requirements by having them meet the standards for NMOG+NO X, PM, CO and formaldehyde for one of the bins listed in Table IV-13. Manufacturers choose bins for their vehicles based on their product plans and corporate strategy for compliance with the fleet average standards discussed in Section IV.B.2.b, and once a vehicle's bin is designated, those bin standards apply throughout its useful life. Because the fleet average standards become more stringent over time, the bin mix will gradually shift from higher to lower bins.

As in the past, there are numerically higher standards levels for Class 3 vehicles than for Class 2b vehicles, reflective of the added challenge in reducing per-mile emissions from large work trucks designed to carry and tow heavier loads. Also, the standards levels for both Class 2b and Class 3 HDVs are significantly higher than those being adopted for light-duty trucks due to marked differences in vehicle size and capability, and to our requirement to test HDVs in a loaded condition (at the adjusted loaded vehicle weight (ALVW)). By conducting emissions testing with loaded vehicles, the heavy-duty program ensures that emissions controls are effective when these vehicles are performing one of their core functions: hauling heavy loads. This is a key difference between the heavy-duty and light-duty truck programs. The bin structure and standards levels are consistent with those in California's LEV III program. We requested comment on the usefulness of creating additional bins between Bin 0 and the next lowest bin in each vehicle class, as a means of encouraging clean technologies and adding flexibility, but commenters saw no need for these.

Table IV-13 FTP Standards for HDVs
NMOG+NO X (mg/mi)PM(mg/mi)CO(g/mi)Formaldehyde(mg/mi)
Class 2b (8501-10,000 lbs GVWR)    
Bin 395 (interim)39586.46
Bin 340 (interim)34086.46
Bin 25025086.46
Bin 20020084.26
Bin 17017084.26
Bin 15015083.26
Bin 00000
Class 3 (10,001-14,000 lbs GVWR)    
Bin 630 (interim)630107.36
Bin 570 (interim)570107.36
Bin 400400107.36
Bin 270270104.26
Bin 230230104.26
Bin 200200103.76
Bin 00000

The NMOG+NO X standards levels for the highest bins in each class (Class 2b Bin 395 and Class 3 Bin 630) are equal to the sum of the current non-methane hydrocarbon (NMHC) and NO X standards levels that took full effect in 2009, as well as to equivalent LEV standards in California's LEV II program. These bins are intended as carryover bins. That is, we expect them to be populated with vehicles that are designed to meet the current standards, and that are being phased out as new lower-emitting vehicle designs phase in to satisfy the Tier 3 fleet average NMOG+NO X standard. We also consider the next highest bins (Class 2b Bin 340 and Class 3 Bin 570) to be carryover bins, because they likewise can be readily achieved by vehicles designed for today's EPA and California LEV II emissions programs. As the 2018-2022 phase-in progresses, it will become increasingly difficult to produce vehicles in these bins and still meet the fleet average standard. Therefore vehicles in these bins (as well as some others not yet designed to meet Tier 3 PM standards described in Section IV.B.2.d) will be considered “interim Tier 3” vehicles, and the bins themselves will be considered “interim bins.”

To facilitate their use in this carryover function, the interim bins do not require manufacturers to meet Tier 3 exhaust emissions standards on the SFTP, over the longer useful life, or with the new gasoline test fuel discussed in Section IV.F, although testing on this fuel will be allowed. These requirements do apply in all other bins.

In the context of these relaxed requirements for the interim bins, we proposed two additional measures to help ensure these bins are focused on their function of helping manufacturers transition to the long-term Tier 3 emissions levels. First, we proposed that the interim bins would be available only in the phase-in years of the program; that is, through model year (MY) 2021, as is appropriate to their interim status. Second, vehicles in the interim bins would meet separate NMOG and NO X standards rather than combined NMOG+NO X standards. The goal was to ensure that a manufacturer does not redesign or recalibrate a vehicle model under combined NMOG+NO X Tier 3 standards for such purposes as reducing fuel consumption, through means that result in higher NO X or NMOG emissions than exhibited by today's vehicles, contrary to the intended carryover function of the interim bins. Industry commenters objected to both the proposed sunsetting of the interim bins and the proposed separate NO X and NMOG standards, arguing that they overly restrict manufacturer flexibility and work against harmonization with LEV III. However, commenters did not address EPA's concern regarding increased NO X emissions at the interim bin levels.

After considering the comments, we believe a modified approach to the interim bins can at least partly address the industry concerns regarding harmonization while still precluding backsliding on NO X levels. We are finalizing the interim bins with combined NMOG+NO X standards as requested by the commenters, but are adopting a restriction on deterioration-adjusted NO X levels in certification testing, to the levels allowed under the current standards in 40 CFR 86.1816-08. These are 0.2 and 0.4 g/mi for Class 2b and Class 3, respectively. This restriction will not apply to vehicles in use, and does not impose a parallel NMOG restriction. Given our continuing concerns about NO X increases that would be allowed by the combined standards at the interim bin levels, we believe that this approach and the associated certification burden are reasonable, noting that manufacturers already must obtain NO X test results in certifying to an NMOG+NO X standard, and the differing NO X and NMOG deterioration mechanisms will likely dictate that they be considered separately in obtaining deteriorated NMOG+NO X levels for certification.

We believe that making the interim bins available indefinitely would run counter to their limited purpose as an aid to making the transition to Tier 3 emissions levels. Making these bins permanent would, we believe, necessitate that they take on other key elements of the Tier 3 program such as longer useful life, SFTP compliance, and the use of Tier 3 test fuel. These requirements in turn would negate the usefulness of these bins in helping to carry over some pre-Tier 3 vehicle designs during the transition years in which the declining fleet average standard levels are high enough to accommodate their continued sale. By MY 2022, the fleetwide standard will be stringent enough to effectively eliminate the ability of manufacturers to use interim bins while meeting the declining fleet average standard levels. We are therefore adopting the sunsetting of the interim bins as proposed, making them available only through MY 2021.

b. Fleet Average NMOG+NO

As in the light-duty Tier 3 program, a key element of the program we are finalizing for HDVs is a fleet average NMOG+NO X standard that becomes more stringent in successive model years: in the case of HDVs, from 2018 to 2022. Each HDV sold by a manufacturer in each model year contributes to this fleet average based on the mg/mi NMOG+NO X level of the bin declared for it by the manufacturer. Manufacturers may also earn or use credits for fleet average NMOG+NO X levels below or above the standard in any model year, as described in Section IV.B.4. As proposed, we are adopting the separate Class 2b and Class 3 fleet average standards shown in Table IV-14, though a manufacturer can effectively average the two fleet classes using credits (see Section IV.B.4). We believe this split-curve approach is superior to a single phase-in covering all HDVs because it recognizes the different Class 2b/Class 3 fleet mixes among manufacturers and the differing challenge in meeting mg/mi standards for Class 3 vehicles compared to Class 2b vehicles, while still allowing for a corporate compliance strategy based on a combined HDV fleet through the use of credits.

We are adopting the proposed fleet average NMOG+NO X standards. These are consistent with those set for the LEV III MDV program in model years 2018 and later. As proposed, we are also adopting provisions allowing manufacturers to voluntarily meet bin and fleet average standards in model years 2016 and 2017 that are consistent with the MDV LEV III standards in those years, for the purpose of generating credits that can be used later or traded to others. These voluntary standards are shown in Table IV-14. This voluntary opt-in program serves the important purpose of furthering consistency between the federal and California programs, such that manufacturers who wish to can produce a single vehicle fleet for sale nationwide, with the opportunity for reciprocal certification in affected model years. It further incentivizes pulling ahead of Tier 3 technologies, with resulting environmental benefits, by providing for early compliance credits in this nationwide fleet. Commenters expressed support for this harmonized array of HDV emissions standards.

Manufacturers choosing to opt into this early compliance program could start in either model year 2016 or 2017. They would have to meet the full complement of applicable bin standards and requirements for the bins they choose for their vehicles in meeting the 2016/2017 MY fleet average FTP NMOG+NO X standards, including SFTP standards in the bins that have SFTP standards. However, they do not need to meet the Tier 3 PM FTP and SFTP standards discussed in Sections IV.B.2.d and IV.B.3.a, or the evaporative emissions standards discussed in Section IV.C, because these requirements phase in on a later schedule. We are not extending the voluntary compliance opportunity to the 2015 model year, based on manufacturer comments indicating it would be of little value.

Table IV-14—HDV Fleet Average NMOG+NO X Standards
VoluntaryRequired program     
Model Year2016201720182019202020212022 and later.
Class 2b333310278253228203178.
Class 3548508451400349298247.

We believe that the voluntary program provisions will benefit the environment, the regulated industry, and vehicle purchasers, because it has potential to accomplish early emissions reductions while maintaining the goal of a cost-effective, nationwide vehicle program in every model year going forward.

Although manufacturers will be allowed to meet the fleet average NMOG+NO X standard through whatever combination of bin-specific vehicles they choose, it is instructive to note that the fully phased in fleet average standard for model years 2022 and later will be the equivalent of a Class 2b fleet mix of 90 percent Bin 170 and 10 percent Bin 250 vehicles, and a Class 3 fleet mix of 90 percent Bin 230 and 10 percent Bin 400 vehicles. Therefore, it is appropriate to consider Bin 170 Class 2b vehicles and Bin 230 Class 3 vehicles to be representative of Tier 3-compliant HDVs in the long term.

c. Alternative NMOG+NO

We believe the fleet average phase-in described above will be flexible, effective, and highly compatible with manufacturers' desire to market vehicles nationwide, because of its close alignment with California's LEV III program for medium-duty vehicles. However, for any HDV manufacturers seeking four years of lead time and three years of stability as specified in Clean Air Act section 202(a)(3)(C), we proposed an alternative compliance path. (340) This alternative approach was crafted to be equivalent to the NMOG+NO X declining fleet average in the above-described LEV III-harmonized alternative in every model year, except that the period for the voluntary program in the alternative approach would extend an extra model year—through 2018. To ensure that this approach meets the Act's stability requirement, instead of being structured around an annually declining fleet average standard, the alternative approach requires a manufacturer to demonstrate compliance (including through use of credits) with a schedule of annually increasing percent-of-sales of HDVs certified to the fully phased in 178 mg/mi (Class 2b) and 247 mg/mi (Class 3) standards, as shown in Table IV-15. We are adopting the alternative percent-of-sales phase-in largely as proposed, with limited changes described below.

Table IV-15—Percent-of-Sales Alternative NMOG+NO X Phase-In
VoluntaryRequired program     
Model Year201620172018 a 2019202020212022 and later.
Class 2b29%39%54%65%77%88%100%.
Class 321%32%47%60%73%87%100%.

The availability of emissions averaging under our alternative phase-in, discussed below, makes the two alternatives functionally equivalent, not just in the annual emissions reductions they achieve, but also in how manufacturers may design their mix of products to meet the phase-in standards. Commenters who disagreed with this assessment for HDVs did not provide their reasoning, beyond referring to similar comments they had on the parallel light-duty (above 6000 lbs GVWR) alternative phase-in. However, that proposed alternative differs from the one we proposed for HDVs, and the elements in it that were found objectionable by the manufacturers are not in the HDV alternative. (See Section IV.A.3 for discussion of comments on the light-duty alternative.)

Commenters objected that the proposed percent-of-sales alternative has not been shown by EPA to be feasible, or in fact is infeasible because it mandates the early phase-in of low-emitting vehicles certified to the final standards. Such comments miss the fact that, with ABT, every manufacturer can produce the same mix of vehicles in any model year to comply with either HDV phase-in alternative, with the exception that MY 2018 is a voluntary phase-in year under the alternative phase-in and a required year under the LEV III-harmonized phase-in. The ABT provisions enable a manufacturer to adopt a fleet average compliance strategy while utilizing the percent-of-sales phase-in that is identical to what would be required under the LEV III-harmonized phase-in's fleet average standards. By no means are manufacturers forced to make only vehicles certified to the final standards. The percent-of-sales phase-in is thereby no more stringent than the LEV III-harmonized phase-in, and the feasibility analysis provided in Section IV.B.5, which expressly addresses the LEV III-harmonized phase-in, serves to demonstrate the feasibility of both alternatives.

Some comments seem to assert that the percent-of-sales framework for the alternative was chosen by EPA to make this alternative so stringent (by requiring some vehicles to meet final standards four years early) that no reasonable company would use it. This is incorrect, both in regard to its actual effect (which as explained above is not more stringent), and in regard to our intent. The percent-of-sales framework for the alternative was proposed and is being adopted for the purpose of providing manufacturers with a phase-in alternative that explicitly meets the applicable Clean Air Act stability requirement.

We are making one change to the percent-of-sales alternative, necessitated by the fact that this final rule is being signed in 2014, not 2013 as envisioned in the proposal. HDV models for which the 2019 model year begins before the fourth anniversary of the signature date of this final rule may be excluded from the Tier 3 fleet average compliance calculations and all other Tier 3 requirements. These excluded vehicles would instead need to comply with the applicable pre-Tier 3 standards and requirements for the entire production of these models throughout the 2019 MY. This limited allowance ensures that the alternative meets EPA's obligation for four years of lead time under the Clean Air Act. It is similar to a phase-in alternative we provided in the light-duty vehicle Tier 2 rule (see 65 FR 6747, February 10, 2000). Note that 40 CFR 86.1803-01 defines “model year” as “the manufacturer's annual production period (as determined by the Administrator) which includes January 1 of such calendar year: Provided that if the manufacturer has no annual production period, the term `model year' shall mean the calendar year.” Additional regulations pertaining to the definition of a model year are in 40 CFR 85, subpart X.

This allowance remains optional within the percent-of-sales alternative—a manufacturer may voluntarily include these early-starting 2019 MY vehicles in the Tier 3 program, and in this case these vehicles would be treated no differently under the alternative than vehicles with a later-starting 2019 MY, including with regard to whether manufacturers choose to make them part of the “phase-in” fleet (vehicles counting toward the phase-in percentages) or the “phase-out” fleet (vehicles not counting toward the phase-in percentages).

Although it is conceivable that manufacturers would commence an early start of the 2019 model year specifically for the purpose of delaying Tier 3 obligations, we do not think this is likely, given the many important constraints and decisions that typically factor into setting this date, and the fact that signature of this final rule is occurring relatively early in the calendar year, well before typical model year start dates. We believe this is a reasonable way to provide a viable percent-of-sales phase-in alternative that has four years of lead time without making the 2019 model year voluntary for all vehicles or putting new constraints on the timing of a manufacturer's model year.

To help ensure that the percent-of-sales alternative is fully equivalent to the LEV III-harmonized alternative in terms of fleet-wide emissions control and technology mix choices, we are including some additional provisions, as proposed. First, the Tier 3 vehicles being phased in under the percent-of-sales alternative, in addition to meeting the fully phased-in FTP NMOG+NO X standards, must also meet all other FTP and (as described below) SFTP standards required by the LEV III-harmonized alternative. These include the CO and formaldehyde FTP standards, the 150,000 mile (15 year) useful life requirement, exhaust emissions testing with the new test fuel for gasoline- and ethanol-fueled vehicles discussed in Section IV.F, and the NMOG+NO X and CO SFTP standards in Table IV-16. The specific standards are those for the bins in these tables closest to the fully phased-in NMOG+NO X standards: Bin 170 for Class 2b and Bin 230 for Class 3. (The PM and evaporative emissions standards phase in on separate schedules under both alternatives, as discussed in Sections IV.B.2.d and IV.C.)

Second, we are making an ABT program available for the percent-of-sales alternative, structured like the one created for the LEV III-harmonized alternative. This involves certifying the vehicles in a manufacturer's HDV fleet to the bin standards, and demonstrating compliance with the fleet average standards for the LEV III-harmonized alternative in each model year, including through the use of ABT credits as in the LEV III-harmonized alternative. We are using the fleet average calculation method for purposes of ABT because, as explained above, we have determined that making this demonstration is equivalent to demonstrating compliance with the percent-of-sales requirement, and we see no value in complicating the program with another set of calculations.

However, we are establishing one difference between the LEV III-harmonized and percent-of-sales alternatives with respect to ABT provisions. Unlike in the LEV III-harmonized alternative, manufacturers will not have to certify all vehicles into bins in order to take advantage of the ABT provisions under the percent-of-sales alternative. Rather they could choose to certify any “phase-out” vehicles (that is, those not counting toward the percent-of-sales phase-in) to the pre-Tier 3 NMHC and NO X standards, provided these vehicles do not have family emission limits (FELs) above those standards. These non-Tier 3 vehicles will not be subject to the Tier 3 standards or other vehicle-specific elements of the Tier 3 compliance program. There were no comments on these specific compliance and ABT provisions associated with the percent-of-sales alternative.

d. Phase-In of PM Standards

Consistent with the light-duty Tier 3 program discussed in Section IV.A, we are phasing in the PM standards for HDVs as an increasing percentage of a manufacturer's production of chassis-certified HDVs (combined Class 2b and 3) per year. In addition to concerns regarding the availability and required upgrades of test facilities used for both light-duty and heavy-duty vehicle testing, manufacturers have expressed uncertainty about PM emissions with new engine and emissions control technologies entering the market as a result of new greenhouse gas (GHG) standards. Therefore we are adopting the same phase-in schedule as for the light-duty sector in model years 2018-2019-2020-2021: 20-40-70-100 percent, respectively. This will apply to HDVs certified under either NMOG+NO X phase-in alternative. The California Air Resources Board (CARB) is phasing in the LEV III PM standards for HDVs on the same schedule, except that LEV III will also involve a 10 percent PM phase-in in the 2017 model year. We asked for comment on our adding this to our voluntary program for 2017, but received no comments on it and are not including it in the Tier 3 program.

For manufacturers choosing the declining fleet average NMOG+NO X compliance path, the PM phase-in requirement for HDVs will be completely independent of the NMOG+NO X phase-in, with no requirement that both phase-ins be met on the same vehicles. As a result, vehicles certified to any of the bin standards for NMOG+NO X need not necessarily meet Tier 3 PM standards before the 2021 model year. Instead, the current 0.02 g/mi PM standard will apply for those vehicles not yet phased into the Tier 3 PM standards. We are requiring that manufacturers choosing the percent-of-sales phase-in alternative for NMOG+NO X meet the PM phase-in requirements with only those vehicles certified to the Tier 3 NMOG+NO X standard, except in the 2019 and earlier model years when the standards, including the PM standards, are voluntary, and in the 2021 model year when the 100 percent PM phase-in requirement exceeds the 87-88 percent NMOG+NO X phase-in requirement. This is appropriate given the ability of manufacturers to build “phase-out” vehicles (those not counting toward the phase-in percentages) under the percent-of-sales NMOG+NO X alternative that are certified entirely to pre-Tier 3 standards while still participating in the Tier 3 ABT program, discussed above.

We will consider any vehicle under either compliance path that is not certified to Tier 3 standards for PM and NMOG+NO X (as well as the other, concomitant Tier 3 standards and requirements such as the extended useful life), an “interim Tier 3” vehicle. This term also applies to vehicles certified in one of the interim bins, as discussed above.

Note that compliance with Tier 3 evaporative emissions requirements follows a separate phase-in schedule as described in Section IV.C. As a result, a vehicle in an exhaust emissions family that the manufacturer has phased in to the new useful life and test fuel requirements may be in an evaporative emissions family that has not yet phased in the Tier 3 useful life and test fuel for evaporative emissions compliance and testing.

i. Optional PM Phase-In

The percent-of-sales phase-in schedule for the PM standard, described above, will allow manufacturers with multiple vehicle models to determine and plan the phase-in of those models based on anticipated sales volumes of each model. However, manufacturers certifying only a few vehicle models may not be able to take meaningful advantage of this schedule. This is because their limited number of models may force them to over-comply to reach the required minimum percentages, compared to a manufacturer with many vehicle models available from which to choose a phase-in pathway.

For instance, a manufacturer with only two models that each equally account for 50 percent of its sales would be required to introduce (at least) one of the models in MY 2018 to meet the phase-in requirement of 20 percent in the first year. At the 50 percent level, this model would then also meet the requirements for MY 2019 (40 percent). To meet the MY 2020 requirement of 70 percent of sales, however, the manufacturer would need to introduce the second Tier 3 vehicle that year. Thus the manufacturer would have introduced 100 percent of its Tier 3 models one year earlier compared to a manufacturer that was able to delay the final 30 percent of its fleet until MY 2021 by distributing its redesign of models over the entire phase-in period.

To provide for more equal application of this benefit among all manufacturers in the early years of the program, we are adopting the proposed optional “indexed” phase-in schedule that could be used by a manufacturer to meet the phase-in requirements. A manufacturer that exceeds the phase-in requirements in any given year will be allowed to, in effect, offset some of the phase-in requirements in a later model year. The optional phase-in schedule will be acceptable if it passes a mathematical test. The mathematical test is designed to provide manufacturers a benefit from certifying to the standards at higher volumes than they are obligated to under the normal phase-in schedule, while ensuring that the overall population of complying vehicles at the end of the phase-in is roughly the same as under the fixed percentage approach. In this alternative approach, manufacturers will weight Tier 3 PM-compliant vehicles in the earlier years by multiplying their percent phase-in by the number of years prior to MY 2022 (that is, the second year of the 100 percent phase-in requirement).

The mathematical equation for applying the optional phase-in is as follows:

(4 × APP2018) + (3 × APP2019) + (2 × APP2020) + (1 × APP2021) ≥ 440,

where APP is the actual phase-in percentage for the referenced model year. The sum of the calculation will need to be greater than or equal to 440, which is the result when the optional phase-in equation is applied to the primary percent phase-in schedule (4 × 20% + 3 × 40% + 2 × 70% + 1 × 100% = 440). Commenters supported this optional PM phase-in approach.

3. Supplemental FTP Standards for HDVs

Unlike passenger cars and light trucks, HDVs are not currently subject to SFTP standards. SFTP standards are intended to ensure vehicles have robust emissions control over a wide range of real-world driving patterns not well-covered by the FTP drive cycle. Even though HDVs are not typically driven in the same way as passenger cars and LDTs, especially as they frequently carry or tow heavy loads, we believe some substantial portion of real world heavy-duty pickup and van driving is not well-represented on the FTP cycle.

The goal in setting the SFTP standards levels is not to force manufacturers to add expensive new control hardware for off-FTP cycle conditions, but rather to ensure a robust overall control program that precludes high off-FTP cycle emissions by having vehicle designers consider them in their choice of compliance strategies. High off-FTP cycle emissions, even if encountered relatively infrequently in real-world driving, could create a substantial inadequacy in the Tier 3 program, which aims to achieve very low overall emissions in use. The SFTP provisions will also help make the HDV program more consistent with the heavy-duty engine program, which for several years has included “not-to-exceed” provisions to control off-cycle emissions. Therefore, in addition to the SFTP provisions, we are further limiting enrichment on spark ignition engines in all areas of operation unless absolutely necessary.

a. SFTP NMOG+NO

The SFTP standards levels are provided in Table IV-16. These are consistent with those adopted in the LEV III program.

Table IV-16—SFTP Standards for HDVs
Vehicles in FTP binsNMOG+NO X (mg/mi)PM(mg/mi)CO(g/mi)
Class 2b with hp/GVWR ≤ 0.024 hp/lb a    
FTP Bins 200, 250550722.0
FTP Bins 150, 170350712.0
Class 2b   
FTP Bins 200, 2508001022.0
FTP Bins 150, 1704501012.0
Class 3   
FTP Bins 270, 40055076.0
FTP Bins 200, 23035074.0

We are linking Tier 3 SFTP implementation for HDVs directly to the Tier 3 FTP phase-in and bins for these vehicles. That is, an HDV certified to any of the Tier 3 FTP bin standards must meet the SFTP standards for that bin as well. However, because the FTP PM standard phases in on a separate schedule, we will require that SFTP PM compliance be linked to the same schedule. That is, an HDV certified to the Tier 3 FTP PM standard must meet the applicable SFTP PM standard as well. This approach recognizes the complementary nature of FTP and SFTP provisions and helps to ensure that Tier 3 emissions controls are robust in real world driving. CARB expressed support in its written comments for this approach to linking FTP and SFTP requirements and an intent to propose aligning LEV III with it once the Tier 3 program is finalized.

There are no SFTP requirements for the interim Tier 3 bins in each class (Class 2b Bins 340 and 395 and Class 3 Bins 570 and 630), because these are essentially carry-over bins from the previous standards to help facilitate the transition to Tier 3, and therefore are not intended to take on new requirements that might prompt a redesign. These implementation provisions are consistent with the approach taken in the LEV III program, except that California applies more of the Tier 3 requirements for SFTP and extended useful life to vehicles in the interim bins.

To help ensure a robust SFTP program that achieves good control over a wide range of real world conditions, we proposed to use a weighted-average composite SFTP cycle, with NMOG+NO X emissions calculated from results of testing over three cycles: the US06, the FTP, and the SC03, weighting these results by 0.28, 0.35, and 0.37, respectively. However, at proposal, we determined that the full US06 component of the composite cycle, along with the ALVW loaded test condition, would not be sufficiently representative of real-world driving for two groups of HDVs: Those with low power-to-weight ratios and Class 3 vehicles.

Therefore, as discussed in the proposal, SFTP testing of Class 2b vehicles with power-to-weight ratios at or below 0.024 hp/lb, may, at the manufacturer's option replace the full US06 component of the composite SFTP emissions with the test results from only the second of the three emissions sampling bags in the US06 test, generally referred to as the “highway” portion of the US06. HDVs so tested will be subject to the correspondingly lower SFTP standards levels shown in the table above. These vehicles will be driven during the test in the same way as the higher power-to-weight Class 2b vehicles (over the full US06 cycle), using best effort (maximum power) if the vehicle cannot maintain the driving schedule. The large majority of Class 2b vehicles—those with power-to-weight above 0.024 hp/lb—will be required to include emissions over the full US06 cycle in the composite SFTP. We believe that this approach provides a robust but repeatable and reliable test for the full range of Class 2b vehicles, as the highway portion of the US06 retains broad coverage of vehicle speed/acceleration combinations measured in real-world driving. Any testing conducted by EPA would follow the manufacturer's test path for the vehicle.

For Class 3 vehicles, which range up to 14,000 lbs GVWR, we are also concerned that the full US06 cycle would not provide a representative drive cycle for SFTP testing. These vehicles are much larger than the light-duty vehicles that formed the basis for development of the US06 cycle, and loading them to ALVW for the SFTP test yields a very heavy test vehicle, not likely to be safely driven in the real world in a manner that is typified by this aggressive cycle. We believe that the LA-92 (or “Unified”) driving cycle developed by CARB is more representative of Class 3 truck driving patterns and will produce more robust results for use in SFTP evaluations. Therefore we are adopting the proposed LA-92 cycle for use in place of the US06 component of the composite SFTP for Class 3 HDVs.

HDVs do not have SC03 emissions requirements under the current HDV standards. Manufacturers of HDVs have indicated that they expect the SC03 emissions to be consistently lower than either the US06 or the FTP emissions levels, and therefore the added SC03 testing burden may be unnecessary. We are therefore providing HDV manufacturers with the option to substitute the FTP emissions levels for the SC03 emissions results for purposes of compliance. However, we will retain the ability to determine the composite emissions using SC03 test results in confirmatory or in-use testing. We received no adverse comments on this proposed approach.

The set of composite SFTP cycles and standards we proposed and are adopting for HDVs is consistent with the MDV LEV III program. We received no adverse comments on them, except with regard to in-use testing as discussed in Section IV.B.6.a.

b. Enrichment Limitation for Spark-Ignition Engines

To prevent emissions from excessive enrichment in areas not fully encountered in the SFTP cycles, we proposed and are adopting limitations in the frequency and magnitude of enrichment episodes for spark-ignition HDVs. These limitations are identical to those for light-duty vehicles. See Section IV.A.4.c for discussion of the requirements and relevant comments received.

4. HDV Emissions Averaging, Banking, and Trading

This section describes how exhaust emissions credits may be earned and used. See Section V.C for similar provisions that apply for evaporative emissions. We are continuing the practice of allowing manufacturers to satisfy standards through the averaging of emissions, as well as through the banking of emissions credits for later use and the trading of credits with others.

There are a number of facets of the Tier 3 ABT program for HDVs that are different from the existing program. First, instead of separate NMHC and NO X credits, manufacturers earn combined credits, consistent with the form of the standards.

Second, manufacturers may accrue a deficit in their credit balance. Deficits incurred in a model year may be carried forward but a manufacturer will not be permitted to have a negative overall HDV credit balance in more than 3 consecutive model years. Manufacturers will have to use any new credits to offset any shortfall before those credits can be traded or banked for additional model years. Credits not used within 5 years after they are earned will be forfeited. These 5/3-year credit/deficit life provisions are consistent with our light-duty Tier 3 approach, the California LEV III program for MDVs, and EPA programs for controlling GHG emissions from light- and heavy-duty vehicles.

Third, as part of our new requirement for chassis certification of complete diesel HDVs, we are allowing the chassis-certified diesel HDVs to participate in the Tier 3 ABT program without restriction. Prior to Tier 3 they have not been allowed to earn or use ABT credits. We are not restricting or adjusting credit exchange between diesel and gasoline-fueled HDVs, consistent with our shift to combined NMOG+NO X standards that helps to ensure comparable stringency for these two engine types, and consistent also with the LEV III MDV program.

Credits earned by a chassis-certified Tier 3 HDV may be used to demonstrate compliance with NMOG+NO X standards for any other chassis-certified Tier 3 HDV, regardless of size and without adjustment. This effectively allows manufacturers to plan a comprehensive HDV compliance strategy for their entire Class 2b and Class 3 product offering, by balancing credits so as to demonstrate compliance with the standards for both classes.

Industry commenters argued that EPA should align the HDV credit provisions with the light-duty program by allowing early Tier 3 credits to be generated in MYs 2016 and 2017, calculated relative to the highest Class 2b and Class 3 bin NMOG+NO X levels (395 and 630 mg/mi, respectively), and capped at a level proportional to the California level in MY 2018. However, these highest bin levels correspond to those of the existing HDV standards for NMHC and NO X, and are significantly higher than the MY 2016 and 2017 LEV III levels. Thus vehicles designed to just meet the LEV III standards in these years could generate a large preliminary number of credits under the industry's Tier 3 early credits proposal, credits they would not earn in LEV III, thereby potentially thwarting the harmonization of the two programs. Truncating that credit bank for each manufacturer in 2018 such that it is proportional to their LEV III balance could perhaps, with additional restrictions on trading and banking, restore a harmonized credit status in that year. However, it constitutes an unnecessarily complex and uncertain pathway to the same result as that achieved under EPA's early opt-in provisions.

Commenters requested that we provide for the conversion of pre-Tier 3 HDV credits for use in Tier 3. However, as discussed in the proposal, we are not including provisions for doing so. We believe that by providing an early Tier 3 opt-in program for HDVs, capable of generating credits for two model years before the mandatory standards take effect (even longer under the alternative percent-of-sales phase-in approach), we are giving ample opportunity for the manufacturers to accumulate early credits.

Manufacturers commented that the proposed fleet average compliance approach is incongruous with California's LEV III method based on vehicle equivalent credits (VECs). Although expressing that they have no preference for the method since the stringency is equivalent, they recommended that EPA foster harmonization by providing a compliance option based on VECs. We believe that such an option would add unnecessary complexity to the Tier 3 program, and is made even more unnecessary by the intent expressed in CARB's written comments to propose a fleet average option for LEV III that is identical to EPA's approach.

In the past we have set upper bounds, called family emission limit (FEL) caps, on how high emissions can be for credit-using vehicles, regardless of how many credits might be available. Under our Tier 3 bin structure, we believe that exhaust emission FEL caps are no longer relevant for Tier 3 HDVs, as every vehicle must meet whatever standards apply in the bin chosen for the vehicle by the manufacturer. (The bin standard becomes the effective FEL.) Indeed, because credits and deficits are calculated based on the difference between a manufacturer's fleet average emissions and the fleet average standards for a given model year, credits are not calculated for individual vehicle families at all. Thus the standard for NMOG+NO X in the highest allowable bin serves the purpose of the FEL caps in previous programs.

Consistent with our proposal, we are not creating an averaging program for the HDV SFTP program, because we believe that the bin structure and FTP-centered NMOG+NO X ABT program provide adequate flexibility for smooth program implementation, especially in light of our aim to have the FTP standards be the primary technology forcers. A separate ABT program for SFTP compliance would add substantial complexity with little benefit, and, by making it possible to demonstrate robust SFTP emissions control on a vehicle that lacks commensurate FTP control, could prove at odds with the primary goal of the supplemental test for HDVs.

5. Feasibility of HDV Standards

The feasibility assessment, discussed in more detail in Chapter 1 of the RIA, recognizes that the Tier 3 program is composed of several new requirements for Class 2b and 3 heavy-duty vehicles, which include primarily large gasoline and diesel pick-up trucks and vans with diverse application-specific designs. These new exhaust emissions requirements include stringent NMOG+NO X and PM standards for the FTP and the SFTP, that will as a whole require new emissions control strategies and hardware in order to achieve the standards. The type of new hardware that will be required will vary depending on the specific application and emissions challenges. Additionally, gasoline and diesel vehicles will require different emissions control strategies and hardware. The level of stringency for the SFTP NMOG+NO X standards will generally only require additional precise control of the engine parameters not necessitated in the past because of the lack of SFTP requirements. Similarly, the new PM standards on both the FTP and SFTP cycles will require more precise control of engine operation on gasoline vehicles while diesels already equipped with diesel particulate filters will require minimal changes. The new PM standards may also require that manufacturers consider the durability of their engines to the 150,000 miles useful life requirement with respect to engine wear resulting in increased oil consumption and potentially higher PM emissions.

In order to assess the technical feasibility of NMOG+NO X national fleet average FTP standards of 178 mg/mi for Class 2b vehicles and 247 mg/mi for Class 3 vehicles, we conducted an analysis of certification data for the HDVs certified in the 2010 and 2011 MYs. For this final rule, we also reviewed certification records for 2012 and 2013 MY vehicles, and determined that these primarily involve carryover engines and emission control hardware. Therefore we did not update the NPRM analysis however any new or updated certification results in the 2012 or 2013 MYs are included in the RIA chapter 1 discussion. This analysis provided a baseline for the current HDV fleet emissions performance, as well as the emissions performance specific to the Class 2b and 3 vehicles. The emissions performance of each heavy-duty vehicle class specific to gasoline and diesel is shown in Table IV-17 below. It is important to note that the emissions results are only the 4000 mile test point results and do not incorporate any deterioration which manufacturers must account for when certifying to a full useful life standard. Designs limiting the deterioration of emission control hardware are critical to meeting the emission standards at the useful life of the Tier 3 program. Deterioration factors to adjust the values to the Tier 3 useful life standard of 150,000 miles were not available. However, deterioration factors to adjust to 120,000 miles useful life, and their implications for performance at higher miles, are discussed in the RIA Chapter 1.

The analysis also reflects the importance of the combined NMOG+NO X standard approach, where diesels and gasoline HDVs can balance their combined NMOG and NO X levels. Diesel vehicles in the analysis produce very low NMHC emissions (NMOG is not reported for diesels) but higher NO X emissions, while gasoline vehicles have opposite performance. The combined standard allows manufacturers to determine the proper balance of the unique emissions challenges of a diesel or gasoline vehicle.

Table IV—17 2010/11 Certification Test Results at 4,000 Miles
NMHCNMOGNO X CONMOG+NO X
GasolineClass 2b0.0500.0520.0411.6480.092
Class 30.0800.0830.0732.3730.156
NMHC+NO X      
DieselClass 2b0.0370.1380.1950.174
Class 30.0190.2490.1580.268
Combined Class 2b0.0430.0260.0890.9220.133 
Combined Class 30.0500.0410.1611.2650.212 

Manufacturers typically certify their vehicles at emissions levels well below the numerical standards. This difference is referred to as “compliance margin” and is a result of manufacturers' efforts to address all the sources of variability that could occur during the certification or in-use testing processes and during in-use operation. These sources of variability include: Test-to-test variability, test location, build variation and manufacturing tolerances, vehicle operation (for example: Driving habits, ambient temperature, etc.), and the deleterious effects of sulfur and other oil and fuel contaminants. To meet the NMOG+NO X standard of 178 mg/mi for Class 2b and 247 mg/mi for Class 3 vehicles and establish a compliance margin for these sources of variability, manufacturers will need to reduce their emission levels considerably from the levels indicated in this data set, particularly for diesel vehicles.

However, as discussed above, these emission results do not include the expected emissions deterioration which will be determined by manufacturers during development and certification testing. Therefore, manufacturers will need to further reduce emissions levels in anticipation of the unavoidable emissions deterioration that will occur during the useful life of the vehicle. Further, deterioration is a function of several factors, but it is predominantly due to emissions control hardware thermal exposure (high temperatures), which is typically a significant issue on vehicles used for performing work like Class 2b and 3 vehicles.

We also expect that the 2011 heavy-duty GHG rule will present new challenges to manufacturers' emissions performance goals as vehicles begin to use new engines designed to meet the new GHG requirements. (341) Some of these new technologies may result in emissions challenges that are specific to certain operating conditions. For example, downsized gasoline engines will likely have improved FTP exhaust emissions but have increased challenge with the high-load SFTP requirements. Diesel-fueled vehicles may need to carefully balance engine controls which reduce GHG emissions but can increase criteria emissions (NO X).

With regard to the ability of the heavy-duty fleet to meet the PM standards for the FTP and the SFTP, we based our conclusions on some testing of current heavy-duty gasoline vehicles (HDGVs) and the PM performance of the existing light-duty fleet with similar engines. Testing of two HDGVs with the highest sales volume (Ford F250 and Chevrolet Silverado 2500), albeit not aged to full useful life, confirmed that they have similar PM emissions levels as the light-duty counterparts and therefore also meet the standards for both the Class 2b and Class 3 configurations. Data from light-duty gasoline vehicles with similar or common engines with their heavy-duty “sister” vehicle models demonstrates that these vehicles are currently meeting the Tier 3 FTP PM standards at the Tier 2 useful life mileage of 120,000 miles. Heavy-duty diesel vehicles all are equipped with DPFs and have no challenges meeting the FTP or SFTP PM standards being set for Tier 3.

The SFTP test data from the same two heavy-duty vehicles described above indicates that gasoline vehicles can achieve the standards for SFTP NMOG+NO X and PM. Since heavy-duty vehicles are not currently required to comply with any of the SFTP requirements, manufacturers have not focused on improving the emissions performance specifically over the SFTP cycles (US06, LA-92, and SC03). Therefore, although the limited testing results had a high degree of variability, several tests met the PM standards for the high power-to-weight Class 2b vehicles. Consistent with light-duty, vehicles that are demonstrating high PM on the US06 will need to control enrichment and oil consumption from engine wear. Recently manufacturers have already been implementing product changes to reduce oil consumption to address both customer satisfaction issues and to reduce cost of vehicle ownership.

Given the technologies likely to be applied to meet the HDV exhaust emissions standards, discussed below, we consider the lead time available before the standards take effect under all of the alternatives to be sufficient. HDV manufacturers are already adopting some of the complying technologies, especially for their light-duty vehicles, and these can readily be adapted for heavy-duty applications. In addition, manufacturers have already begun developing these technologies for HDVs, including diesels, in response to California's recently adopted LEV III MDV standards which begin to take effect in the 2015 model year. Finally, as described above in Sections IV.B.2, IV.B.3, and IV.B.4, our program incorporates a number of phase-in and alternative compliance provisions that will ease the transition to final standards without disrupting heavy-duty pickup and van product redesign cycles. Among these is an alternative phase-in that does not begin mandatory standards until model year 2019.

Comments we received on the proposed HDV standards did not specifically address our analysis of their technical feasibility. The Manufacturers of Emission Controls Association (MECA) outlined diesel and gasoline-engine technologies that they expect will be used to achieve the Tier 3 standards cost-effectively, generally consistent with our draft RIA. Vehicle and engine industry commenters argued that the case we made for feasibility relied too heavily on extending light-duty truck test data, supplemented by testing of only two HDVs, neither of which were fully aged or representative of future vehicles designed to meet our new GHG standards. However, commenters did not question the feasibility, durability, implementability, or effectiveness of the technologies we identified, or their ability to achieve the proposed standards. Instead, the focus of these comments was on statutory provisions for lead time and stability, and on how relaxed standards for in-use testing and testing at high altitudes would help to implement the standards within the allotted lead time. These issues, including changes we are making in response to the comments, are addressed in Sections IV.B.2.c, IV.B.6.a, and IV.B.6.f.

i. Technologies Likely To Be Applied

The technologies expected to be applied to vehicles to meet the lower standards levels will address the emissions control system's ability to control emissions during cold start. Current vehicle emissions control systems depend on the time it takes for the catalyst to light-off, which is typically defined as the catalyst reaching a temperature of 250 °C. While the specific emissions challenge is somewhat different for gasoline engines than for diesel engines, achieving the necessary temperatures in the catalysts is a common challenge. In order to improve catalyst light-off, the manufacturers will likely add technologies that provide heat from combustion more readily to the catalyst or improve the catalyst efficiency at lower temperatures. These technologies could include calibration changes, thermal management, close-coupled catalysts, catalyst Platinum Group Metal (PGM) loading, and possibly secondary air injection. In some cases, where the catalyst light-off response and efficiency are not enough to address the cold start emissions, hydrocarbon adsorbers may be applied to trap hydrocarbons until such time that the catalyst is lit-off. Note that with the exception of hydrocarbon adsorbers each of these technologies addresses both NMOG and NO X performance. Key potential technologies are described in greater detail below.

  • Engine Control Calibration Changes—These include changes to retard spark and/or adjust air/fuel mixtures such that more combustion heat is created during the cold start on gasoline engines. Diesel engines may use unique injection timing strategies or other available engine control parameters. Engine calibration changes can affect NMOG, NO X and PM emissions.
  • Thermal Management—This technology includes all design attributes meant to conduct the combustion heat into the catalyst with minimal cooling on both gasoline and diesel engines. This includes insulating the exhaust piping between the engine and the catalyst, reducing the wetted area of the exhaust path and/or reducing the thermal mass of the exhaust system. Close-coupling of catalysts (packaging the catalysts as close to the head of the engine as possible to mitigate the cooling effects of longer exhaust piping) can also be effective, but is more difficult to employ than in light-duty applications because of durability concerns with highly loaded operation and the potential increase in fuel consumption to protect the catalyst from high temperatures.
  • Catalyst PGM Loading—Additional PGM loading in the catalyst provides a greater number of sites to catalyze emissions and addresses NMOG, NO X and PM emissions.
  • Selective Catalytic Reduction Optimization—Diesel applications will continue to refine this NO X emissions control strategy through improved hardware design and implementation in vehicle applications. Additional engineering enhancements in the control of the SCR system and related processes will also help reduce emissions levels.
6. Other HDV Provisions
a. In-Use Emissions

The proposal requested comment on the need for relaxation of NMOG+NO X and PM standards for in-use vehicle testing. The LEV III program includes these on an interim basis in the more stringent bins in both FTP and SFTP testing. However, in its written comments, CARB expressed the view that the technologies required for SFTP compliance are well-established, and that sufficient lead time is provided such that interim in-use standards for SFTP are not needed. As a result, CARB expressed an intent to propose aligning the LEV III program with the approach EPA proposed on this matter after the Tier 3 program is finalized. The manufacturers commented that relaxed interim in-use standards are needed in the HDV sector, both for FTP and SFTP standards. The reasons cited were a need to harmonize with LEV III, the scarcity of data on which to establish standards that apply over the full useful life, the extension of that useful life to 150,000 miles, the need for manufacturers to address customer concerns with new products and technologies, uncertainties that accompany the new SFTP cycles and part 1066 testing requirements (especially for PM), and the introduction of innovative technologies required to meet GHG standards in the same timeframe.

After considering the comments we have concluded that relaxed interim in-use standards are appropriate for HDVs, both for FTP and SFTP testing. We are adopting HDV in-use standards levels that are identical to those adopted for LEV III, as shown in Table IV-18. We consider these levels reasonable, in line with relaxed in-use standards adopted in past programs, and helpful toward harmonization. We are not applying interim in-use NMOG+NO X standards to the interim (two highest) bins for the FTP standards, because these bins are intended for carry-over of existing designs, and there should be little uncertainty over their in-use emissions performance. Interim bin vehicles certified to the Tier 3 PM standards shall, however, be subject to the relaxed in-use PM standards in the same way as for HDVs in other bins. Bin 0 standards are driven by specific zero-emissions technologies for which in-use margins would not be appropriate, and so we are not setting in-use standards for Bin 0.

We are also adopting the general approach taken in LEV III of making these interim standards available during the phase-in period (model years 2016-2022) for the first two model years that a test group is newly certified to a Tier 3 NMOG+NO X or PM standard. Test groups subsequently recertified to a more stringent NMOG+NO X bin standard may begin the two year cycle over again. A test group that is first certified into a Tier 3 bin in model year 2022 or later may not take advantage of the relaxed interim in-use standards. LEV III adopted somewhat different applicability years, for the most part ending earlier, in model year 2020. However, we believe that the modest extension is appropriate to facilitate the Tier 3 phase-in. If a vehicle test group is certified into a Tier 3 bin, but not yet to the Tier 3 PM standard, the in-use standard for PM shall apply for the first two model years it is first certified to the PM standard. In order to better harmonize with LEV III, the availability of these in-use standards includes the voluntary model years.

Table IV-18—Interim In-Use Standards for HDVs
FTP (mg/mi)NMOG+NO X PMSFTP (mg/mi)NMOG+NO X PM
Class 2b    
Bin 395 (interim)(a)16(a)(a)
Bin 340 (interim)(a)16(a)(a)
Bin 25037016 b 770/1120 b 12/15
Bin 20030016 b 770/1120 b 12/15
Bin 17025016 b 490/630 b 12/15
Bin 15022016 b 490/630 b 12/15
Bin 0(a)(a)(a)(a)
Class 3    
Bin 630 (interim)(a)20(a)(a)
Bin 570 (interim)(a)20(a)(a)
Bin 4006002077012
Bin 2704002077012
Bin 2303402049012
Bin 2003002049012
Bin 0(a)(a)(a)(a)
b. HDV Useful Life

Currently the HDV regulatory useful life, the period of use or time during which emissions standards apply, is 120,000 miles or 11 years, whichever occurs first (40 CFR 86.1805-4). For Tier 3 vehicle criteria emissions we are extending the useful life to 150,000 miles or 15 years, whichever occurs first. This change better reflects the improvements in vehicle durability and longevity that have occurred in the several years since the 120,000 mile useful life was established, and maintains consistency with the LEV III MDV program and with our Tier 3 program for large LDTs, for which the same useful life period is being adopted.

The new useful life requirement applies to Tier 3 HDVs in all bins except those designated as interim bins, consistent with the purpose of the interim bins to provide for limited carry-over of pre-Tier 3 vehicle designs during the phase-in period. Although the percentage application in each year will therefore depend on each manufacturer's fleet binning strategy, the declining NMOG+NO X fleet average standard will ensure a robust phase-in of the new useful life requirement over the 2018-2022 model years, such that it is expected to be about 50 percent in 2018, and necessarily reaches 100 percent by 2022 when the interim bins are no longer available. For those manufacturers choosing to certify to the voluntary standards, the new useful life will apply even earlier, in model year 2016 or 2017. For manufacturers choosing the alternative percent-of-sales NMOG+NO X alternative, the new useful life requirement applies to all HDVs counted toward the phase-in requirement, resulting in a generally equivalent useful life phase-in rate to that of the LEV III-harmonized alternative.

See Section IV.F.5 for further discussion of useful life requirements with regard to GHG standards. Manufacturers may optionally retain the 120,000 mile/11 year useful life for PM on interim Tier 3 vehicles that are not phased in to the Tier 3 PM standards. We received no adverse comments on these useful life provisions.

c. Heavy-Duty Alternative Fuel Vehicles

As in the light-duty program, manufacturers must demonstrate heavy-duty flexible fuel vehicle (FFV) and dual-fuel vehicle compliance with both the FTP and the SFTP emissions standards when operating on both the conventional petroleum-derived fuel and the alternative fuel. Dedicated alternative fuel vehicles must demonstrate compliance with both the FTP and SFTP emission standards while operating on the alternative fuel. For all of these vehicles, this includes the requirement to meet FTP emissions standards when conducting fuel consumption and GHG emissions testing, and also to meet the FTP and highway test requirements at high altitudes (see Sections IV.B.6.e and f). Because FFVs can operate on various combinations of their conventional and alternative fuel, the emissions requirements apply to operation at any mix of the fuels achievable in the fuel tank with commercially available fuels, including for compliance at high altitudes, even though the required demonstration of compliance is limited to the conventional and alternative fuels designated for certification testing. We received no adverse comments on these provisions.

d. Existing Provision To Waive HDV PM Testing

EPA's existing program includes a provision for manufacturers to waive measurement of PM emissions in non-diesel heavy-duty vehicle emissions testing. As proposed, we are eliminating this provision. We believe that the Tier 3 PM standards for these vehicles are of sufficient stringency that routine waiver of testing is not appropriate. The CARB LEV III program also reflects this view. We do not expect this change to be onerous for manufacturers, as the number of heavy-duty vehicle families is not large. We received no adverse comments on this change.

e. Meeting HDV Standards in Fuel Consumption and GHG Emissions Testing

As with the light-duty Tier 3 program, HDVs must meet the FTP bin standards when tested over both the city and highway test cycles. We do not believe this adds a very significant test burden as vehicle emissions are already required to be measured when these tests are run for GHG and fuel consumption determinations. Nor do we believe that this requirement is design forcing. Rather, we are creating this requirement to ensure that test vehicle calibrations are not set by manufacturers to minimize fuel consumption and GHG emissions, at the expense of causing high criteria pollutant emissions. Considering the additional work involved in measuring PM emissions and the reduced likelihood of high PM emissions on the highway test, we are not mandating that PM emissions testing be included in this requirement. We received no adverse comments on these proposed provisions.

f. HDV Altitude Requirements

As in the past, we intend that HDV Tier 3 standards result in emissions controls that are effective over a full range of operating altitudes. We proposed that HDVs be required to meet the FTP bin standards (but not the SFTP standards) at high altitudes, and expressed our expectation that compliance with the FTP standards would require neither the use of special hardware nor adjustment to the level of the standards.

The manufacturers argued in their comments that the reasons EPA cited in proposing relief at high altitudes for light-duty vehicles apply for HDVs as well, and requested that relaxed NMOG+NO X standards be adopted in the more stringent bins for testing of HDVs at high altitudes. Ford argued that the challenges could be even greater for HDVs because they are designed to operate at high altitudes with heavy payloads and towed trailers, and this may necessitate the locating of emissions systems farther from exhaust manifolds, thereby increasing catalyst lightoff delays.

Although we agree to a certain extent about the performance of gasoline-fueled HDVs at high altitudes and their similarity to LDVs, the comments did not alter our view that the compliance margins provided in the HDV FTP bin standards compared to what the control technologies can achieve, and the freedom manufacturers have to shift to the more stringent bins gradually as the program phases in, are adequate to account for these effects at altitude. The manufacturers provided no data to counter this view.

We note that our adoption of relaxed interim in-use standards for vehicles in these bins will be directionally helpful to address any remaining concerns by manufacturers regarding emissions at altitude (Section IV.B.6.a). This is because testing at high altitudes is often not required for certification (typically manufacturers use an engineering analysis instead), and thus the relaxed in-use standards will help to facilitate Tier 3 implementation for any HDV designs in which in-use problems at high altitudes surface in the initial model years.

C. Evaporative Emissions Standards

Gasoline vapor emissions from vehicle fuel systems, which are a mixture of hydrocarbon compounds, occur when a vehicle is in operation, when it is parked, and when it is being refueled. Evaporative emissions which occur daily from gasoline-powered vehicles are primarily functions of air and fuel temperature, fuel vapor pressure, and vehicle driving. EPA first instituted evaporative emissions standards in the early 1970s to address hydrocarbon emissions when vehicles are parked after being driven. These are commonly referred to as hot soak and diurnal emissions. Over the subsequent years the test procedures have been modified and improved, the standards have been revised to be more stringent, and we have addressed emissions which arose from new fuel system designs by establishing new requirements such as running loss emission standards and test procedure provisions which address resting losses (e.g., permeation). Onboard refueling vapor recovery (ORVR) requirements for control of refueling emissions first began to phase-in for light-duty vehicles (LDVs) and light-duty trucks (LDTs) in the 1998 MY. These were later expanded to cover medium-duty passenger vehicles (MDPVs) and some heavy-duty gasoline vehicles (HDGVs).

Even though evaporative and refueling emission control systems have been in place for most of these vehicles for many years, evaporative emissions still contribute 30-40 percent of the on-road mobile source hydrocarbon inventory. The rate of these emissions in grams/day (hot soak and diurnal), grams/mile (running loss) or grams per gallon (refueling) depends on (1) the stringency of the applicable emission standards, (2) ambient and fuel temperature, (3) fuel vapor pressure, and (4) the presence/state of repair of the fuel/evaporative control system.

These fuel vapor emissions are ozone and PM precursors, and also contain air toxics such as benzene. Even though there are mature evaporative emission control programs in place, further hydrocarbon emission reductions are needed and can be achieved from further evaporative emission controls on gasoline-powered highway motor vehicles.

This section discusses the vehicle evaporative emission standards and related provisions for LDVs, LDTs, MDPVs, and HDGVs. The evaporative emissions program has six basic elements: (1) The early allowance program (MY 2015-2016), (2) the transitional program (MY 2017), (3) the Tier 3 evaporative emission phase-in program (MY 2018-2021), (4) the fully phased-in standards (MY2022+), (5) requirements for HDGVs including ORVR for the 2018MY, and (6) a leak standard and test procedure which become mandatory for Tier 3 vehicles in the 2018MY. As discussed below, we are finalizing more stringent standards that will apply for the 2- and 3-day evaporative emissions tests, a canister bleed test procedure and emission standard, and a new certification test fuel specification. (342) As discussed in section IV.D, we are also adding a fuel/vapor system leak standard and test procedure for LDVs, LDTs, and MDPVs. EPA is not changing any existing light-duty running loss or refueling emission standards with the Tier 3 FRM, with the exception of the certification test fuel specification and the addition of a refueling emission controls for complete HDGVs over 10,000 lbs gross vehicle weight rating (GVWR). This section also describes phase-in flexibilities, credit and allowance programs, and other issues related to evaporative emissions control.

In this rule, the vehicle classifications, LDVs, LDTs, MDPVs, and HDGVs, remain unchanged from Tier 2 (see 40 CFR 86.1803-010). For purposes of this discussion of the Tier 3 evaporative emissions program, the vehicle standards can be further placed in four categories: (1) “zero evaporative emission” PZEV vehicles certified by CARB as part of the ZEV program, (2) vehicles certified by CARB to meet LEV III evaporative emission program requirements on CARB certification fuel (7 RVP E10) as early as 2014 MY, (3) vehicles meeting the Tier 3 evaporative emissions program requirements using the Tier 3 certification test fuel (9 RVP E10), and (4) transitional vehicles meeting existing EPA evaporative requirements on Tier 2 certification test fuel (9 RVP E0). (343 344) For ease of reference these four categories may be referred to as PZEV evap, LEV III evap, Tier 3 evap, and Tier 2/MSAT evap in this section. (345)

1. Tier 3 Evaporative Emission Standards
a. Final Standards

The Tier 3 program for evaporative emissions builds on previous EPA requirements as well as the evaporative emissions portion of CARB's recent LEV III rule which starts mandatory phase-in with the 2018 MY. The level of the standards, the timing of their implementation, and related provisions are designed in great measure to allow manufacturers to design, certify, and build one control system for each evaporative/refueling family to meet CARB and EPA requirements so that these vehicles can be sold in all 50 states. Commenters supported this approach and no commenter opposed the stringency or timing of the evaporative emission standards and related test procedures. We believe the program is appropriate since it will require new more stringent evaporative emissions control technology in new vehicles and also achieve improved in-use system performance.

Section IV.C.1.a.i, which follows, describes the basic emission standard levels for LDVs, LDTs, MDPVs, and HDGVs. Section IV.C.1.a.ii, describes a new canister bleed standard and testing requirement for measuring emissions from the evaporative canister. Section IV.C.1.a.iii discusses the optional use of the CARB LEV III Option 1 evaporative emission standards during a transition period. Next, Section IV.C.1.a.iv discusses interim use of CARB PZEV zero evap data based on CARB Phase II fuel. Finally section IV.C.1.a.iv, discusses the ongoing requirement to meet running loss emission standards.

i. Hot Soak Plus Diurnal Standards

The Tier 3 hot soak plus diurnal emission standards are designed to bring into the broader motor vehicle fleet the “zero evap” technology used by the manufacturers in their partial zero emission vehicles (PZEVs). Manufacturers developed this “zero evap” technology as part of their response to meeting the requirements of the CARB Zero Emission Vehicle (ZEV) program. This program, which is in effect in 11 other states, allows manufacturers to meet their ZEV mandate percentages (totally or in-part) by the use of vehicles which among other characteristics have very low fuel vapor emissions.

The hot soak plus diurnal emission standards we are adopting (presented in Table IV-19) are designed to be met with technology that limits Tier 3 vehicles to essentially zero fuel vapor emissions. For the Tier 3 evaporative emissions program, we are not changing the basic 2-and 3-day evaporative emission test procedures other than the certification fuel requirements. The level of the standards primarily accommodates what is often referred to as new vehicle background hydrocarbon emissions. These emissions arise from the off-gassing of volatile hydrocarbons from plastics, rubbers, and other polymers found in new vehicles (e.g., new tires, interiors, seats, fuel system components, paints, and adhesives). In the field these emissions decrease over time as the vehicle ages, but this cannot necessarily be replicated in the time that manufacturers typically allocate for vehicle certification or with the techniques normally used for vehicle pre-conditioning. Provisions related to vehicle pre-conditioning before evaporative emissions certification testing are discussed further below.

In the past EPA has set relatively uniform (but not identical) evaporative emission standards for LDVs and LDTs and somewhat higher values for MDPVs and HDGVs. The Tier 3 hot soak plus diurnal emission standards follow this approach, because in general the vehicles have higher levels of non-fuel background emissions as they get larger.

As described in more detail in Section IV.C.2.d below, EPA is finalizing a program that will allow manufacturers to demonstrate compliance with the hot soak plus diurnal evaporative emission standards using averaging concepts. A manufacturer may comply by averaging within each of the four vehicle categories but for the reasons discussed below, may not rely on averaging across categories. The technical approaches to meeting the standards are discussed in Section IV.C.2.

Table IV-19 Final Evaporative Emission Standards
Vehicle category/averaging setsHighest hot soak +diurnal level(over both 2-day and3-day diurnal tests)
LDV, LDT10.300
LDT20.400
LDT3, LDT4, MDPV0.500
HDGVs0.600
ii. Canister Bleed Emission Standard

In addition to more stringent hot soak plus diurnal standards, EPA is finalizing a new canister bleed emission test procedure and standard as part of the Tier 3 program. The canister bleed test procedure is described in Section IV.C.6 below. EPA is adopting the canister bleed standard because it is an important tool in moving Tier 3 evaporative emissions control toward zero fuel vapor emissions. No commenter opposed the canister bleed standard or commented on the test procedure. The new test and standard align with the California LEV III requirements and help to ensure that near-zero fuel vapor emissions are being emitted by vehicles from the fuel tank through the evaporative emission canister. Manufacturers will be required to measure diurnal emissions over the 2-day diurnal test procedure from just the fuel tank and the evaporative emission canister using Tier 3 certification fuel and comply with a 0.020 g/test standard for all LDVs, LDTs, and MDPVs and 0.030 g/test for HDGVs. The feasibility of this standard is discussed in Section IV.C.3 below. The canister bleed test and standard drives canister design elements such as total gasoline working capacity, internal architecture, and the type of carbon used. These are also key elements of canister design for the hot soak plus diurnal emission standards.

The canister bleed standard will be implemented differently than the hot soak plus diurnal standard. EPA is not applying the averaging program to this new bleed test standard as compliance is relatively straightforward and low in cost. Therefore, each evaporative/refueling emission family certified by manufacturers will need to demonstrate compliance with their respective standard. As discussed below, the canister bleed standard will not apply at high altitude, but proportional control is expected. Since the performance of the canister is also evaluated in the hot soak plus diurnal evaporative emissions sealed housing for evaporative determination (SHED) test the canister bleed emission standard will not be included in the In-Use Verification Program of under 40 CFR 86.1845 through 1853, but it must be met in use. We will not have canister bleed specific family criteria for certification but the test will have to be completed and the standard met for each evaporative/refueling family including potentially twice if there are two canisters used. A deterioration factor will not be required, but the manufacturer must certify that the standard will be met for the full useful life. As mentioned above, the standard will have to be met in-use and could be evaluated in EPA confirmatory testing.

The canister bleed standard will have to be met using the same fuels and test procedures used for the hot soak plus diurnal standards. We will accept results on either CARB or EPA test fuels/test temperatures for the canister bleed test provided the same are used for the hot soak plus diurnal test.

iii. Hot Soak Plus Diurnal Standard With the Fuel System Rig Test

As part of its LEV III program, CARB has included an alternative set of evaporative emission standards, referred to as Option 1 standards. These are shown in Table IV-20.

Table IV-20 CARB—Option 1 Evaporative Emission Standards
Vehicle categoryHighest hot soak + diurnal level (over both 2- and 3-day diurnal tests) (g/test)Vehicle SHEDRig SHEDRunning loss(g/mile)
Passenger Car0.3500.00.05
LDT 6,000 lbs GVWR0.5000.00.05
All other vehicles > 6,000 lbs GVWR0.7500.00.05

The Option 1 standards include evaporative emission standards (hot soak plus diurnal) that are slightly higher numerically than our final standards. Vehicles certified under this option may not use averaging in the CARB LEV III program because they basically represent the same evaporative emission standards as exist for PZEVs under CARBs ZEV program wherein averaging is not permitted. Option 1 also includes an additional SHED test of the vehicle fuel system (rig test) that pre-dates development of the canister bleed emission standard. The rig SHED test is discussed in Section IV.C.6. From a practical perspective, this test is more difficult to conduct than the bleed test discussed above and is intended to force manufacturers to demonstrate at certification that their stand alone (not in chassis) fuel/vapor control system designs have 54 mg fuel vapor emissions. (346) While one commenter was in favor of permanently including Option 1 in the EPA final rule based on what it viewed to be favorable pre-production engineering design features of the rig SHED test, EPA is including Option 1 only as interim compliance alternative for a limited period of time but not as a permanent option in the Tier 3 evaporative emission program. While we see the value to vehicle manufacturers of the rig SHED test as an engineering design and development tool, by its very nature, the rig SHED test and standard is not implementable as an enforceable standard because a fuel system cannot be removed from a vehicle and reconstructed in a SHED for testing without compromising its fundamental structural and mechanical integrity as it existed on the vehicle. We believe that the hot soak plus diurnal SHED test and standard and the canister bleed test and standard will accomplish the objective of keeping fuel vapor emissions to a minimum while doing so in an enforceable manner.

EPA believes most manufacturers will prefer to certify to the averaging based standards in Table IV-1 (similar in stringency and program construct to CARB Option 2). However, because some manufacturers may have vehicle models meeting the CARB Option 1 standards and emission requirements now or in the near future, EPA will allow compliance with the CARB Option 1 standards as an acceptable interim alternative to compliance with the Tier 3 evaporative emission standards if the model is certified by CARB to LEV III requirements before the 2017 MY. These vehicles could then be certified using carryover provisions through the 2021 MY as part of the evaporative emissions phase-in described below. This is two model years longer than in the proposal, but this extension is reasonable given the life cycle of most fuel/vapor control systems and the goal of aligning with the LEV III program for a national program where possible. (347) As noted in the following sections, vehicles certified under this provision will count toward the phase-in percentage requirements and could earn allowances as discussed below, but the vehicles will not be eligible to earn or use credits for the evaporative emissions averaging program. Carryover vehicles will have to meet the EPA leak standard and the high altitude emission standard to be counted toward the sales percentage requirements for 2018 and later model years.

iv. Interim Carryover of PZEV Evap Data for Tier 3 Certification

To earn credits toward compliance with the CARB Zero Evaporative Emissions (ZEV) program requirements, many manufacturers have certified LDVs and LDTs to 150,000 mile useful life emission standards similar to those found in Table IV-20. These vehicles have used CARB Phase II fuel (E0) and met the rig SHED test requirement in lieu of the canister bleed standard, but otherwise have employed the same basic technology EPA expects for the LEV III and Tier 3 programs. EPA is permitting data generated from certification of these vehicles in the 2015 and 2016 MYs to be used for Tier 3 evaporative emissions purposes through the 2019 MY.

v. Running Loss Emission Standards

EPA has required vehicles to meet running loss emission standards since the 1996 model year. These requirements, which are specified in 40 CFR 86.134-96, apply to all gasoline-powered highway motor vehicles. EPA is not changing either the test procedures or emission standard for the running loss test. However, the change in certification test fuel will apply to testing for such standards. This is appropriate based on the rationale for implementing a certification fuel change and is necessary since the running loss test is part of the overall test sequence for the 3-day hot soak plus diurnal test. EPA does not anticipate that the change in certification test fuel will impact the stringency of the running loss test and standards or the manufacturers' ability to comply as part of Tier 3.

b. High-Altitude Requirements

Prior to this rule, the most recent vehicle evaporative emission standards were adopted in 2007. (348) The new standards adopted in 2007 apply only to testing under low-altitude conditions. (349) In the 2007 rule, we decided to continue to apply the previous “Tier 2” standards for testing under high-altitude conditions. This was necessary to achieve an equivalent level of overall stringency for high-altitude testing, accounting for the various effects of altitude and lower atmospheric pressure on vapor generation rates, canister loading and purging dynamics, and other aspects of controlling evaporative emissions due primarily to lower air density and vapor concentrations at altitude. While it is important for vehicles to have effective emission controls at high altitudes, we do not want the high-altitude standards and test procedures to dictate the fundamental design of the Tier 3 evaporative emission control systems since the high altitude vehicle population is only about five percent of the national total. Therefore, we believe it is appropriate to address this goal by applying the current 2-day low altitude evaporative emission standards and requirements for high-altitude testing. (350) The vehicle categories for the high altitude standards in this rule are the same as for the low altitude standards. The standards are presented below in Table IV-21. This will both reduce evaporative emissions at high altitude and again create a requirement to confirm that emission controls function effectively at high altitude without forcing manufacturers to apply altitude-specific technologies. The leak standard presented in Section IV.D below will apply equally at low and high altitude testing as compliance is not dependent on air density and vapor concentrations.

Table IV-21—Final High-Altitude Evaporative Emission Standards
Vehicle categoryHighest hot soak +diurnal level(over both 2-day and3-day tests) (g/test)
LDV, LDT10.65
LDT20.85
LDT3, LDT41.15
MDPV1.25
HDGVs ≤ 14,000 lbs GVWR1.75
HDGVs > 14,000 lbs GVWR2.3

A few additional points should be noted about our Tier 3 high altitude evaporative emissions control program. First, EPA does not expect manufacturers to produce vehicles with high-altitude only evaporative control systems. Given the nature of evaporative emission control technology, there should be emission reductions at high altitude proportional to those achieved at lower altitudes. We are not applying the canister bleed test and emission standard at high altitude, but we expect similar emission reductions to those which will occur at low altitude. These vehicles will have to meet the canister bleed emission standard at low altitude and canister bleed emission reductions at high altitude should be proportional as is the case with the low altitude hot soak plus diurnal standards. Any adjustment to meet the standard at high altitude to account for canister adsorption and desorption effects of higher altitudes would result in fundamentally the same technology with an increase in the testing burden but not necessarily more emissions control. Therefore, we believe the low-altitude canister bleed test is sufficient for achieving the level of emission control for operation in both low-altitude and high-altitude conditions. Second, for vehicles certified with FELs above or below the applicable standard for testing at low altitude, the same differential will apply to the FELs for high-altitude. For example, if an LDV was certified with an FEL of 0.400 g instead of the 0.300 g standard, the high-altitude FEL will be 0.75 g (0.65g+0.10g). This high-altitude FEL will not be used for any emission-credit calculations, but it will be used as the emission standard for compliance purposes. Third, gasoline RVP for certification test fuel will be set at 7.8 RVP with 10 percent ethanol, as specified in Section IV.F. Finally, we are finalizing a minor adjustment to the high altitude test procedures. The existing 2- and 3-day test procedures apply equally at low and high altitude. We are keeping the same basic requirement but will allow for a downward adjustment of 5 °F in the temperatures related to the running loss test within the 3-day test cycle. Thus, the applicable ambient temperatures at § 86.134-96 (f) and (g) will be 90±5 °F instead of 95±5 °F for high altitude testing, and the entire fuel temperature profile from § 86.129-94(d) shifts down by 5 °F. EPA believes this is appropriate given the differences in atmospheric conditions at low versus high altitude and will still result in equivalent control of running loss emissions at higher altitudes. EPA requested comment on the alternative approach of keeping test temperatures the same, but omitting the 3-day test cycle for testing at high altitude. This was supported by one set of commenters, but at this time EPA does not have the data needed to drop such a fundamental test requirement.

As mentioned above, emission data from vehicles meeting the current CARB PZEV zero evap and CARB LEV III Option 1 requirements could be used to qualify that vehicle to meet the Tier 3 evaporative emission regulations for the 2017-2021 MYs. To qualify for a federal certificate, the vehicle will also have to meet the Tier 3 high altitude evaporative emission requirements. CARB does not require vehicles to meet EPA high altitude requirements, so for these vehicles we are giving the manufacturers the option to certify either by providing SHED test data or based on an engineering demonstration using data and analysis and the application of good engineering judgment. For the 2015-2017 MYs, manufacturers can use data based on either Tier 2 or Tier 3 test fuel. Beginning in the 2018 model year, for Tier 3 vehicle certification to the high altitude standard, the data must be based on Tier 3 fuel.

c. Useful Life

Trends indicate that vehicle lifetimes are increasing. It is important that emission control systems be designed to meet requirements while vehicles are in use. As discussed in Section IV.A.7 and IV.B.6 of this preamble, along with the new emission standards, we are finalizing a longer useful life of 150,000 miles/15 years, whichever comes first, for LDTs up to 6,000 lbs GVWR but over 3,750 lbs loaded vehicle weight (LVW) (LDT2s), all LDTs over 6,000 lbs GVWR (LDT3/4), MDPVs, and HDGVs. The longer useful life will apply to all certifications to the Tier 3 evaporative emission standards (see Table IV-19 and Table IV-20 above). For an evaporative/refueling family certified to 150,000 miles/15 year useful life for evaporative emissions this useful life will also apply to the hot soak plus diurnal, running loss, canister bleed, fuel system rig, refueling, leak, and high altitude standards. All of these standards impact the fuel and vapor control systems and it is technologically consistent to require the same useful life for these standards because they all rely on the mechanical integrity, durability, and operational performance of the same components in the evaporative emissions control system.

Due to limitations in the CAA, for LDVs and for LDTs up to 6,000 lbs GVWR and at or below 3,750 lbs LVW (LDT1s), we are keeping the current useful life of 120,000 miles/10 years unless, as described in Section IV.A.7, a manufacturer elects alternative exhaust emission requirements that are associated with 150,000 mile/15 year useful life for these vehicles. For manufacturers that select those optional standards, the useful life of 150,000 miles/15 years will apply for all Tier 3 evaporative emission requirements as listed in the previous paragraph.

During the early, transition, and phase-in program periods and until the final year of the allowed phase-in period for the Tier 3 evaporative emission program (MY 2015-2021) the differences between the exhaust and evaporative emission phase-in programs presents the possibility that in some cases a manufacturer could certify a model to the Tier 3 exhaust requirements (or CARB equivalents) but not necessarily to the Tier 3 evaporative emission requirements. (351) In those situations, the final rule provides that a family could have a 150,000 miles/15 years useful life for exhaust emissions but maintain the current useful life for all of the evaporative and refueling emission standards since the vehicle does not yet meet Tier 3 evaporative emission requirements. During the phase-in period, if a family is certified to the Tier 3 evaporative emission requirements but not yet certified for Tier 3 exhaust emission requirements, then the useful life could be 150,000 miles/15 years for evaporative and refueling emissions standards but the existing useful life for exhaust emissions. However, by the 2022 MY the useful life for all of these requirements will be 150,000 miles/15 years for LDT2/3/4s, MDPVs, and HDGVs since by that model year all vehicles must be certified using Tier 3 certification fuel and test procedures and meet Tier 3 evaporative emission standards or CARB equivalents. (352)

OBD regulations call for the systems to operate effectively over the useful life of the vehicle. We are not changing that requirement, but rather want to clarify that during the early, transition, and phase-in years of the program (MY 2015-2021), all of the OBD monitoring requirements have the same useful life as that for the exhaust emission standard except for the evaporative system leak monitoring requirement which has the same as that required for the evaporative and refueling emission standards control systems.

d. What requirements must a vehicle meet to qualify as a Tier 3 vehicle for evaporative emissions?

As mentioned above, there are three different revised or new evaporative emision requirements applicable to Tier 3 vehicles. These are the hot soak plus diurnal standards, the canister bleed standard, and the leak standard. In addition the refueling, running loss, and spit back standards are unchanged but will have to be met on Tier 3 certification fuel. Compliance with these requirements is potentially complicated by the fact that the CARB ZEV and LEVIII programs will bring zero evap technology into the market place before or at the same time that Tier 3 implementation begins but with test fuel and test procedure differences. In order to qualify as a Tier 3 vehicle for evaporative emission purposes the vehicle must meet all applicable requirements on the specified fuel. Unless otherwise specified (e.g., HDGV refueling spit back), if a vehicle does not meet all evaporative emission program requirements, including both the applicable standards and test fuel then it does not qualify as a Tier 3 vehicle for evaporative emission purposes. Table IV-22, below summarizes the requirements that vehicles in various categories must meet to qualify as a Tier 3 vehicle for evaporative emission purposes as a function of model year. The entries in the cells of the table specify the required test fuel. The table is for reference of the reader in reviewing subsequent sections of this preamble. Refer to the regulatory text for specific requirements for the various programs.

Table IV-22—Requirements for Vehicle To Qualify for Tier 3 Evaporative Emissions Program and Test Fuel Requirements
Model yearProgram/zero evap stdsHS+DI/running lossRigCanister bleedLeak (except HHDGV) *High altitude & refueling/spit back **
MY 2017 TRANSITION PROGRAM      
2017Percentage—PZEV zero evap (carryover)CA Ph. 2CA Ph. 2N/AN/AEPA Tier 2 or Tier 3.
2017Percentage—LEV III Opt. 1CA Ph. 3CA Ph. 3N/AN/AEPA Tier 3.
2017Percentage—LEV III Opt. 2CA Ph. 3N/ACA Ph. 3N/AEPA Tier 3.
2017Percentage—Tier 3Tier 3N/AEPA Tier 3N/AEPA Tier 3.
2017PZEV zero evap only (carryover)CA Ph. 2CA Ph. 2N/AN/AEPA Tier 2 or Tier 3.
201720/20—PZEV zero evap (carryover)CA Ph. 2CA Ph. 2N/AEPA Tier 3EPA Tier 2 or Tier 3.
201720/20—LEV III Opt. 1CA Ph. 3CA Ph. 3N/AEPA Tier 3EPA Tier 3.
201720/20—LEV III Opt. 2CA Ph. 3N/ACA Ph. 3EPA Tier 3EPA Tier 3.
201720/20—Tier 3Tier 3N/AEPA Tier 3EPA Tier 3EPA Tier 3.
MY 2018-2021 PHASE-IN PROGRAM      
2018-2019PZEV zero evap (carryover)CA Ph. 2CA Ph. 2N/AEPA Tier 2 or Tier 3EPA Tier 2 or Tier 3.
2018-2021LEV III Opt. 1CA Ph. 3CA Ph. 3N/AEPA Tier 3EPA Tier 3.
2018-2021LEV III Opt. 2CA Ph. 3N/ACA Ph. 3EPA Tier 3EPA Tier 3.
2018-2021Tier 3Tier 3N/AEPA Tier 3EPA Tier 3EPA Tier 3.
MY 2022+ FULLY PHASED-IN PROGRAM      
2022+LEV III Opt. 2CA Ph. 3N/ACA Ph. 3EPA Tier 3EPA Tier 3.
2022+Tier 3Tier 3N/AEPA Tier 3EPA Tier 3EPA Tier 3.
2. Program Structure and Implementation Flexibilities
a. Percentage Phase-In Requirements

As proposed, the final Tier 3 evaporative emission standards will be phased in over a period of six MYs 2017-2022. Manufacturers supported the proposed phase-in schedule and there were no issues raised with regard to lead time for any vehicle class. As discussed below, there will be three options for the 2017 MY. For the 2018-2019 MYs, the requirement will apply to 60 percent of a manufacturer's nationwide sales of all LDVs, LDTs, MDPVs, and HDGVs (including vehicles sold in California and the section 177 states). This will increase to 80 percent for MYs 2020 and 2021 and by MY 2022 it will apply to 100 percent of sales in these four categories. Beginning in MY 2018 any vehicle included in the percentage phase-in, except vehicles that had earned allowances, will have to meet the leak standard discussed in section IV.D.

Evaporative emission requirements for the MY 2017 apply only to LDVs, LDT1s, and LDT2s as defined in 40 CFR 86.1803-01. To be consistent with the start date for Tier 3 exhaust standards, phase-in requirements will not include vehicles over 6,000 lbs GVWR until the 2018 MY. The manufacturers will have three options. The first, which we are calling the “primary” or “percentage” option, requires that a value equal to 40 percent of a manufacturer' s LDVs, LDT1s, and LDT2s sold outside of California and the states that have adopted the CARB ZEV or LEV III programs must meet the Tier 3 evaporative emission requirements on average. The 40 percent is calculated based on vehicles at or below 6,000 lbs GVWR but compliance can be based on vehicles regardless of their GVWR. The second which we are calling the “PZEV zero evap only” option, requires a manufacturer to sell all of the LDVs, LDT1s, and LDT2s certified with CARB as meeting the PZEV evaporative emission requirements (zero evap) in MY 2017 throughout all of the U.S. and not to offer for sale any non-PZEV zero evap version of those specific vehicle models/configurations in any state whose vehicles are covered by the Tier 3 evaporative emission standards. Thus, this will apply to sales in any state except for California and states that have adopted the CARB ZEV or LEV III programs under section 177 of the Clean Air Act. Under this second option, no tracking of sales or end of year compliance calculation will be required. Some manufacturers may find this option attractive, as they have more limited product offerings and find tracking of production and sales more difficult. The third option, which we are terming the 20/20 option, requires that 20 percent of a manufacturer's LDVs, LDT1s, and LDT2s (e.g., equal to or less than 6,000 lbs GVWR) sold outside of California and the states that have adopted the CARB ZEV or LEV III programs meet the Tier 3 evaporative emission requirements on average and that this 20 percent or another 20 percent of vehicles in the three groups listed above meet the leak standard discussed in section IV.D. Each percentage requirement must be met, (i.e., there is no flexibility to permit meeting shortfalls of the hot soak plus diurnal or leak standard percentages with higher values from the leak standard category). However, as was the case with the 40 percent option above, compliance can be based on vehicles regardless of their GVWR. The third option was supported by several commenters as a means to address 2017 MY transition issues related to phase-out of current products and phase-in of future products. EPA believes that for these vehicles the leak standard will provide emission reduction benefits comparable in magnitude to the Tier 3 evaporative emission standards. Thus, under this approach, the manufacturers' product transition concerns can be addressed while achieving the overall evaporative emission reductions from 2017 MY vehicles. It should be noted that these vehicles must also meet the 0.020 inch evaporative system leak monitoring requirement which also takes effect in the 2017 model year.

As discussed below, beginning in the 2018 MY, to be counted toward the percentages needed to meet the Tier 3 phase-in percentages (e.g., 60% in 2018 and 2019 MYs) a Tier 3 compliant vehicle must also meet the leak standard.

At the time of certification, manufacturers will identify which families will be included in their Tier 3 evaporative emission percentage calculations (this could be families above or below the individual Tier 3 evaporative emission standards for the given class of vehicles (Table IV-19) as well as vehicles meeting CARB's PZEV zero evap or LEV III Option 1 standards (Table IV-20) and could also include earned allowances as discussed below. The manufacturers will use projected sales information for these families plus allowances as desired and available, to show how they expect to meet the phase-in percentage requirements for the model year of interest. At the end of the model year reconciliation the manufacturers will be expected to show that the percentages were met. If the percentages are not met, the manufacturers will either use additional allowances and/or bring more vehicle families/vehicles into the calculation until the sales percentage is met. This step is being required because the initial demonstration of compliance with the fixed percentage at certification is based on projected sales. If the manufacturers did not have to demonstrate that the fixed percentages were met, the percentage would then be a goal and not a requirement and there would be no means to capture the emission reduction shortfalls. This step is unique to the evaporative emission program relative to the NMOG+NO X and PM programs because the evaporative program involves both fixed percentages and ABT. The NMOG+NO X program involves ABT but does not involve fixed percentages and the PM program involves fixed percentages but does not involve ABT.

The additional vehicles added to meet the percentage could only be meeting the Tier 2 hot soak plus diurnal requirements. In this case, use the larger of the 2- or 3-day hot soak plus diurnal certification emission levels. Adding these vehicle families/vehicles into the calculations (discussed below) may result in a credit deficit for that model year for a given averaging set. A manufacturer could not have an unresolved deficit for more than three consecutive model years as discussed below. The deficit would have to be eliminated with positive credits not later than the ABT calculation and credit reconciliation which occurs after the fourth model year.

As discussed above for exhaust emissions, while unlikely, it is possible that a manufacturer could in its annual certification preview meeting with EPA, indicate that its technology mix is such that it will have a credit deficit when the sales percentages requirement is met. This could occur if the fleet average evaporative emission value for Tier 3 vehicles did not meet the Tier 3 hot soak plus diurnal standard for the Tier 3 vehicles in any given averaging set. Also, a manufacturer could have a deficit from a previous model. In these situations, certifying with a projected or actual deficit would require EPA approval after submission of a plan from the manufacturer which explains how it will eliminate the deficit within the model years permitted. Even if a manufacturer had projected or actual deficits for two or three consecutive model years, all accrued deficits would have to be eliminated by the reconciliation which occurs after the fourth model year. Within this plan, which would have to be submitted and approved at each annual certification preview meeting, EPA would expect to see progress toward compliance as indicated by such factors as improved emissions performance for future test groups, a substantiated trend toward a more favorable fleet technology sales mix, no backsliding in projected fleet average values, and perhaps other situation specific criteria.

Requiring a showing at the time of certification based on projected sales requires due diligence by the manufacturers and EPA, but the Tier 3 evaporative emissions program allows for fleet averaging, so a validation or “truing up” of these sales projections after the end of the model year is necessary for determining compliance with the requirements of the standard. This is discussed in Section IV.C.2.d.iii. As discussed further below, validated sales information will also be used for earning early allowances and to show compliance with the alternative phase-in schedule approach.

For these purposes, vehicles included in the phase-in percentage could be: (1) Families which certified to PZEV zero evap or CARB LEV III Option 1 requirements in MYs 2015 and 2016, (2) families certified to meet Tier 3 evaporative emission requirements, (3) any vehicle family certified to the CARB LEV III Option 2 hot soak plus diurnal evaporative emission standards, and (4) vehicles from the early allowance program. To qualify as a Tier 3 certification for evaporative emission purposes, any new evaporative/refueling emission family certifications will have to meet the EPA Tier 3 certification requirements for both test procedure and certification test fuel for the evaporative (hot soak plus diurnal and canister bleed, running loss), refueling, and spit back emission standards. The leak standard will apply in the 2018 and later MYs to all Tier 3 vehicles except HHDGVs and those from the early allowance program. Furthermore, assuming the EPA provisions related to carryover of emissions data are met, 2015-2016 MY CARB PZEV zero evap evaporative emissions certifications could be carried over until the end of the 2019 MY and included as compliant vehicles within the Tier 3 program if they meet the other applicable Tier 3 requirements. The same is true for CARB LEVIII Option 1 certification, except carryover would be permitted through the 2021 MY if they meet the other Tier 3 requirements. See Table IV-4 for more detail on the program options and fuel requirements by model year.

The phase-in percentages for MYs 2017 through 2022 reflect a percentage phase-in concept applied successfully by EPA in previous rules involving evaporative and refueling emissions control. The phase-in provides an appropriate balance between the needed emission reductions and time for the manufacturers to make an orderly transition to the new technology on such a broad scale. The higher initial percentage here is appropriate because the expected evaporative emission control technology is already being used to varying degrees by 12 manufacturers on over 50 vehicle models today and is projected to gain even deeper penetration by 2017 due to the partial zero emission vehicles (PZEV) option within the CARB ZEV program. (353)

b. Alternative Phase-In Percentage Scheme

As part of program flexibility, we are allowing manufacturers to demonstrate compliance with the phase-in percentage requirements of the evaporative emissions program by using a manufacturer-determined alternative phase-in percentage scheme. The alternative phase-in percentage provisions allow manufacturers to use a phase-in more consistent with product plans such as beginning with a lower percentage(s) than required under the primary phase-in during the early years or to benefit from producing and selling more than the minimum percentage of compliant vehicles early. This flexibility could also be helpful in the event that a manufacturer elects to put some vehicles on different phase-in schedules for meeting Tier 3 exhaust and evaporative emission standards. As explained further below, with some limitations, allowances could be used toward compliance with the alternative phase-in scheme values for any given model year.

This approach, which was widely supported in comments by the manufacturers, would be available beginning in the 2017 MY for all manufacturers, except for any manufacturer which used the “PZEV zero evap only” nationwide option for the 2017 MY for whom the approach would be available beginning in 2018 MY. Vehicle and fuel eligibility requirements for the program are summarized in Table IV-23. Refer to the regulatory text for specific requirements.

Table IV-23—Vehicle Qualifications for 2017-2022MY Alternative Phase-in Percentage Schemes & Test Fuel Requirements
Model yearProgram zero evap stds.HS+DI/running lossRigCanister bleedLeak (except HHDGV) *High altitude & refueling/Spit back **
2017PZEV evap (carryover)CA Ph. 2CA Ph. 2N/AN/AEPA Tier 2 or Tier 3.
2017LEV III Opt. 1CA Ph. 3CA Ph. 3N/AN/AEPA Tier 3.
2017LEV III Opt. 2CA Ph. 3N/ACA Ph. 3N/AEPA Tier 3.
2017Tier 3Tier 3N/AEPA Tier 3N/AEPA Tier 3.
2018-2019PZEV evap(carryover)CA Ph. 2CA Ph. 2N/AEPA Tier 2 or Tier 3EPA Tier 2 or Tier 3.
2018-2021LEV III Opt. 1CA Ph. 3CA Ph. 3N/AEPA Tier 3EPA Tier 3.
2018-2022LEV III Opt. 2CA Ph. 3N/ACA Ph. 3EPA Tier 3EPA Tier 3.
2018-2022Tier 3Tier 3N/AEPA Tier 3EPA Tier 3EPA Tier 3.

Under this approach, before the 2017 MY (2018 MY for a manufacturer which used the “PZEV zero evap only” nationwide option for the 2017 MY), a manufacturer will present a plan to EPA which demonstrates that the sum of the products of a weighting factor and the percentages of their U.S. vehicle sales for each model year from 2017 (2018) through 2022 is greater than or equal to 1280 if the program started in the 2017 MY (or 1040 if the program started in the 2018 MY). The 1280 and 1040 numerical values are equal to the sum of the product of the weighting factors and the percentage requirements for MY 2017 or 2018 start dates, respectively, as applicable through MY 2022. These are calculated in the following manner: [(6)(2017MY%)+(5)(2018MY%)+4(2019MY%)+3(2020MY%)+2(2021MY%)+(1)(2022MY%)]. The 2017 MY portion of the calculation would not be included if the manufacturer used the “PZEV zero evap only” nationwide option and thus started the alternative phase-in scheme in the 2018 MY. Under the regulations, EPA has the authority to question elements of the plan and to seek clarifications and potential changes as needed. EPA could disapprove the plan and potentially not allow the use of an alternative phase-in scheme for the model year of interest if the manufacturer does not present a viable explanation and rationale as to how the required numerical sum for the phase-in would be achieved.

EPA also sought comment on including the 20 percent value hot soak plus diurnal value from the 20/20 option described above for 2017 MY in this calculation. Manufacturers generally supported including the 2017 MY in the calculation but did not clearly state whether the 40 percent or 20/20 option approach or both were supported. EPA has decided to include both options for the 2017 MY in the alternative phase-in percentage scheme; 40 percent as described above or 20 percent, with the stipulation that any vehicle used to meet the 20 percent requirement in the 2017 MY would also have to meet the OBD evaporative leak monitoring requirements and the leak standard. In other words, the flexibility of using different vehicles as allowed for the 20/20 option in the primary phase-in scheme is not included in the alternative phase-in. Including this restriction avoids the complexity that would be added if two different sets of vehicles were allowed to meet the two elements of the 20/20 option for the 2017 MY, as in the primary phase-in (e.g., expanding the calculation and tracking requirements and incorporating leak standard compliance and OBD evaporative system monitoring as part of the alternative phase-in scheme). If a manufacturer's hot soak plus diurnal value exceeded 20 percent then that larger value could be used in the alternative phase-in calculation. However, the leak standard value cannot be less than 20 percent and for the first 20 percent the hot soak plus diurnal and the leak must be on the same vehicle and that vehicle must meet the 0.020 inch OBD evaporative system leak monitoring requirement. Compliance would be calculated in the following manner: [(6)(2017MY%)+(5)(2018MY%)+(2019MY%)+3(2020MY%)+2(2021MY%)+(1)(2022MY%)]. If choosing the 20/20 option approach for MY 2017, the value to be met or exceeded in the alternative phase-in would be 1160 which is based on substituting the required phase-in percentages for MYs 2017-2022 in the equation. Under this option as above, before the 2017 MY, the manufacturer would have to submit a plan to EPA which demonstrates that the sum of the products of a weighting factor and the percentages of their U.S. vehicle sales for each model year from 2017 through 2022 is greater than or equal to 1160. A manufacturer that over complies with the targets (i.e., 1040, 1160, 1280) may not trade the excess to another manufacturer. Also, a manufacturer must include all of its affected products in program, not just specific vehicle categories or subcategories.

A manufacturer's alternative phase-in plan must be approved by EPA prior to the start of production for a given model year and will have to be reviewed with EPA each subsequent model year to confirm that the manufacturer's target percentages are being met. This would be expected to occur at the annual certification preview meeting. Manufacturers not meeting their target goals must present revised plans for EPA approval to show how the target percentages and equivalent emission standards will be met. Manufacturers using the alternative phase-in percentage scheme must still show compliance with the hot soak plus diurnal standards in each year as discussed in Section IV.C.2.d. iii even if they fall short of their individual target goal percentages for a given year. EPA is not requiring that manufacturers include Tier 2 vehicles in the calculation for a given model year if they fall short of projections (e.g., if a manufacturer projects 25% in a given model year but only achieves 22%) because it will have to be made up in a subsequent year using a lower multiplier.

c. Allowance Program

We are finalizing incentives for early introduction of vehicles compliant with the Tier 3 evaporative emission regulations. Manufacturers can take advantage of these incentives prior to MY 2018 by selling vehicles that meet the Tier 3 evaporative emission regulations earlier than required or in greater numbers than required. Vehicle eligibility requirements for the allowance program are summarized in Table IV-24. Refer to the regulatory text for specific provisions.

Table IV-24—Vehicle Eligibility To Earn Allowances & Test Fuel Requirements
Model year & programVehicle categoryHS+DI/running lossRigCanister bleedHigh altitude & refueling/spitback *
EVAPORATIVE EMISSIONS     
2015-2016 PZEV zero evap carryoverLDV, LDTCA Ph. 2CA Ph. 2N/AEPA Tier 2/Tier 3.
2015-2016 LEV III Option 1LDV, LDTCA Ph. 3CA Ph. 3N/AEPA Tier 2/Tier 3.
2015-2016 LEV III Option 2AllCA Ph. 3N/ACA Ph. 3EPA Tier 2/Tier 3.
2015-2016 Tier 3AllTier 3N/ATier 3EPA Tier 3.
2017 “PZEV evap only” carryoverLDT 3&4CA Ph. 2CA Ph. 2N/AEPA Tier 2/Tier 3.
2017 “Percentage” option—LEV III Option 1LDT3 &4 MDPV, HDGVCA Ph. 3CA Ph. 3N/AEPA Tier 3.
2017 “Percentage” option LEV III—Option 2LDT3/4 MDPV, HDGVCA Ph. 3CA Ph. 3CA Ph. 3EPA Tier 3.
2017 “Percentage” option Tier 3LDT3/4 MDPV, HDGVTier 3EPA Tier 3EPA Tier 3EPA Tier 3.
2017 “20/20” and all MY alt phase-in schemesNot available.
2018+ LDV, LDT, MDPV & HDGVNot available.
ORVR **     
2015-2017 Early ORVRComplete HDGV >10,000 but ≤14,000 lbs. GVWREPA Tier 2/Tier 3.
2015-2021 Early ORVRComplete HDGV >14,000 lbs. GVWREPA Tier 2/Tier 3.
2015-2021 ORVRIncomplete HDGV >8,500 lbs. GVWREPA Tier 2/Tier 3.

As described below, manufacturers can earn “allowances” for selling any vehicle meeting the Tier 3 evaporative emission program requirements as specified in Table IV-22 earlier than required. The vehicles may be LDVs, LDTs, MDPVs, or HDGVs. Specifically, the allowance program includes the following: (1) For MYs 2015 and 2016, any LDVs and any LDTs meeting the Tier 3 evaporative emission program requirements as specified in Table IV-22 which are sold outside of California and the states that have adopted CARB's ZEV or LEV III programs, (2) for MYs 2015-2017, any MDPV or HDGV meeting the Tier 3 evaporative emission program requirements as specified in Table IV-22 early and sold in any state, (3) for MY 2017, any LDT3/4 meeting the Tier 3 evaporative emission program requirements as specified in Table IV-22 and sold outside of California and the states that have adopted CARB's LEV III or ZEV programs, and (4) for MYs 2015-2017, any complete or incomplete HDGV with a GVWR greater than 10,000 lbs meeting the EPA refueling emissions regulations and sold outside of California and the states that have adopted CARB's LEV III program. EPA asked for comment on extending the ORVR requirement to all HDGVs, complete and incomplete. As discussed in section IV.C.4.b, we are extending ORVR to all complete vehicles over 14,000 lbs GVWR, but are not including incomplete vehicles over 8,500 lbs GVWR in the ORVR requirement at this time. However, we are permitting complete vehicles over 14,000 lbs GVWR and incomplete HDGVs meeting the refueling emission standard to earn allowances through the 2021 MY. Any complete or incomplete HDGV eligible to earn allowances for the model years and areas discussed above will earn them at a 1:1 rate for refueling emissions compliance purposes and at a 2:1 rate for Tier 3 evaporative emissions purposes because the refueling emission reductions are much larger.

Furthermore, for the 2017 MY, manufacturers choosing EPA's “percentage” option (see Section IV.C.2.a) could earn allowances for sales of LDT3s, LDT4s, MDPVs, and HDGVs that meet the CARB LEV III or Tier 3 evaporative emission standards and related requirements assuming their LDV, LDT1/2 sales meet the 40 percent requirement. Similarly, manufacturers choosing EPA's “PZEV zero evap only” option could earn allowances in MY 2017 for LDT3/4s, MDPVs, and HDGVs that meet the “PZEV zero evap” evaporative emission standards, CARB LEV III, or EPA Tier 3 evaporative emission standards and related requirements. EPA has decided not to include allowances for the 2017MY for any manufacturer using the 20/20 option since it would involve identifying not only the vehicles exceeding the 20 percent for the Tier 3 evaporative emission requirements but also the vehicles exceeding the 20 percent for the leak standard and these may be different vehicles. For both the “percentage” and “PZEV zero evap only” options for the 2017 model year, to avoid double counting, the allowances will be earned only for those vehicles sold outside of California and the states that have adopted CARB's LEV III/ZEV program requirements.

To qualify as a Tier 3 vehicle for evaporative emission allowance purposes the vehicle must meet the requirements summarized in Table IV-22. Manufacturers will earn one allowance for each qualifying vehicle sold. Manufacturers can use these allowances in MY 2017 through 2022 to help demonstrate compliance with the phase-in percentage requirements and fleet average evaporative emission standards for those years. Since credits and allowances serve primarily the same purpose and allowing for splits of allowances/credits greatly complicates program implementation, the final rule provides that manufacturers can only earn allowances in MYs 2015-2016 for any LDVs and LDT1/2s meeting the Tier 3 evaporative emission regulations which are sold outside of California and the states that have adopted CARB's ZEV or LEV III programs and for MYs 2015-2017 for any qualifying LDT3/4, MDPV, and HDGV. (354)

Allowances will be used in the compliance determination in the following manner. Vehicles qualifying for allowances can be used in the fleet average evaporative emission standard calculation for any year during the phase-in. This applies to the primary phase-in and alternative phase-in programs. Allowance vehicles will be entered into the compliance calculation with an emission value equivalent to the evaporative emission standard for their vehicle category from Table IV-19 even if it was certified to CARB PZEV zero evap or LEV III Option 1 standards (Table IV-20). For the percent phase-in requirement in either the primary or alternative phase-in schemes, allowance vehicles will count for one vehicle for each allowance used within their vehicle category. For the primary scheme this will be counted as one vehicle, but for the alternative phase-in option the value will be multiplied by the weighting factor (6 for 2017, 5 for 2018, 4 for 2019, 3 for 2020, etc). Within the alternative phase-in scheme the manufacturer will be limited to using these early allowances for no more than 10 percentage points of the phase-in requirements in any given model year (e.g., MYs 2017-2022). EPA believes this limitation is appropriate since early use in the alternative phase-in scheme is multiplied and early introduction of “zero evap” technology should be encouraged, but not necessarily at the expense of its widespread use across the various vehicle categories as the phase-in progresses. The allowances are designed primarily to facilitate manufacturer transition during the program phase-in. As such, they may not be traded between manufacturers and unused allowances will expire after the 2022 MY.

An example here may be helpful in demonstrating how allowances will work. Take a hypothetical manufacturer who earned a total of 10,000 allowances in MYs 2015 and 2016 and sells 100,000 units per year. In MY 2018, the manufacturer will have a phase-in requirement of 60 percent or 60,000 vehicles. For the primary phase-in option the manufacturer could use part or all of its allowances in 2018 without restriction. For the alternative phase-in scheme assume the manufacturer set its alternative phase-in value at 60 percent for the 2018 MY. The final regulations limit the use of allowances to 10 percentage points of the 60 percent or in this case 10,000 vehicles out of 60,000. Without a multiplier this will require the use of all 10,000 allowances in 2018, but with the multiplier of 5 for MY 2018 only 2,000 allowances are needed to reach the 10 percentage point maximum. Using a similar calculus, the manufacturer could use another 10 percentage points in MY 2019, but it will require 2,500 allowances to reach this level since the multiplier is 4 assuming sales remain at 100,000 units per year. The number of allowances to reach the 10 percentage point level will increase each year as the multiplier decreases.

d. Evaporative Emissions Averaging, Banking, and Trading
i. Introduction

Throughout EPA's programs for mobile source emission controls, we have often included emission averaging programs for exhaust emissions. An emission averaging program is an important factor we take into consideration in setting emission standards under the Clean Air Act. An emission averaging program can reduce the cost and improve the technological feasibility of achieving standards, helping to ensure the standards achieve the greatest achievable reductions, considering cost and other relevant factors, in a time frame that is earlier than might otherwise be possible. Manufacturers gain flexibility in product planning and the opportunity for a more cost-effective introduction of product lines meeting a new standard. Emission averaging programs also create an incentive for the early introduction of new technology, which allows certain emission families to act as leaders for new technology. This can help provide valuable information to manufacturers on the technology before they apply the technology throughout their product line.

These programs generally involve averaging and banking, and sometimes trading (ABT). Averaging allows a manufacturer to certify one or more families at emission levels above the applicable emission standards as long as the increased emissions are offset by one or more families certified below the applicable standards. These are referred to as individual family emission limits (FELs). The over-complying families generate credits that are used by the under-complying families. Compliance is determined on a total mass emissions basis to account for differences in production volume, and on other factors as necessary such as useful life. The average of all emissions for a particular manufacturer's production within a vehicle category must be at or below the level of the applicable emission standards. Banking allows a manufacturer to generate emission credits and bank them for future use in its own averaging program in later years. Trading allows a manufacturer to sell credits or obtain credits from another manufacturer.

EPA proposed and is finalizing an emissions ABT program for the Tier 3 hot soak plus diurnal evaporative emissions standards. The evaporative emissions ABT program is generally structured and operates the same as that for exhaust emissions as discussed in Section IV.A.7.m. The major difference is the added requirement to reconcile compliance with the fixed percentage requirement as discussed in detail in Section IV.C.2.a. Also, there is a five year credit life for evaporative emissions as opposed to the longer interim values for NMOG+NO X FTP and SFTP credits.

This is the EPA's first averaging type program for evaporative emissions from light-duty or heavy-duty vehicles. It does not apply to the canister bleed standard or the leak standard because it is the low altitude “zero evap” hot soak plus diurnal standard which will drive the fundamental approach used to comply with all of these requirements. We sought comment on the value of including trading in the program. The comments from the Alliance of Automobile Manufacturers and the Association of Global Automakers very generally supported the inclusion of trading but provided no detail. Upon follow-up from EPA no manufacturer provided any further explanation on the need for the program or how they might use it. (355) In past similar programs for exhaust emissions there have been only a few trades, but incorporating trading within the program adds a degree of flexibility if a manufacturer finds itself in a credit deficit situation. Thus, we have decided to include trading, but credit trades are limited based on the same averaging set restrictions as discussed below for averaging and banking.

The evaporative emissions ABT program will start with the 2017 MY for the percentage and 20/20 options. Prior to the 2017 MY and for other options as discussed in Section IV.C.2.b, manufacturers may earn allowances. The programs will continue for the 2018 MY and beyond for all manufacturers regardless of their 2017 MY option and will not sunset, as does the allowance program. Vehicles generating ABT credits in the 2017 MY or later will not be permitted to also generate allowances as this would be double counting.

A key element of an averaging program is the identification of the averaging sets. This establishes the basis within which evaporative emission families can be averaged for purposes of compliance as well as credit and deficit determinations. As proposed, we are finalizing four averaging sets and the applicable emission standard for each of the averaging sets as shown in Table IV-19. Except as noted in Section IV.C.2.d.2 below, credit exchanges between averaging sets will not be permitted. Participation in ABT is voluntary since a manufacturer could elect to certify each family within the averaging set to its individual standard as if there were no averaging program.

An evaporative emission ABT calculation and assessment involves two distinct steps. The first is the determination of the credit/deficit status of each family relative to its applicable standard from Table IV-19. The second is the role of ABT calculations in the overall compliance demonstration which is discussed in Section IV.C.2.d.

ii. Family Emission Limits

A manufacturer choosing to participate in the evaporative emissions ABT program will certify each emission family to an FEL that applies for the hot soak plus diurnal standard for low altitude testing. The FEL selected by the manufacturer becomes the emission standard for that emission family. Emission credits (or deficits) are based on the difference between the emission standard that applies (by vehicle category) and the FEL. The vehicles will have to meet the FEL for all emission testing. As mentioned in Section IV.C.1.b., above, for vehicles certified with FELs above or below the applicable standard for testing at low altitude, the same differential will apply to the FELs for high-altitude. This high-altitude FEL will not be used for any emission-credit calculations, but it will be used as the emission standard for compliance purposes.

The final rule provides that the FELs selected by the manufacturer must be selected at 0.025 g/test increments above or below the applicable Tier 3 evaporative emission standards for each vehicle category. For example, for LDVs the increments for the FELs would be +/− 0.025 from 0.300 g/test (e.g., 0.225, 0.250, 0.275, 0.300, 0.325, 0.350, 0.375 . . . 0.500). The FEL is used in the compliance demonstration not the certified level. The certified level must be below the FEL, but the FEL could be a higher value than the closest increment value. For example, a certified value of 0.235 g/test could support an FEL of 0.250 g/test or any other higher increment value. One commenter asked that the gradation be finer than 0.025 g/test, but EPA believes this is the appropriate increment, since the standard itself is the sum of two values and rounding of the measured values is involved.

FELs are capped such that they cannot be set any higher than 0.500 g/test for LDVs, 0.650 g/test for LDT1s and LDT2s, 0.900 g/test for LDT3s and LDT4s, 1.000 g/test for MDPVs, 1.4 g/test for HDGVs at or below 14,000 lbs GVWR, and 1.9 g/test for those above 14,000 lbs GVWR, respectively. These FEL caps are the 3-day hot soak plus diurnal emission standards applicable under EPA's existing regulations. While we asked for input on these FEL caps and vehicle groupings, no party provided comment.

Total evaporative emission credits (or deficits) under the Tier 3 hot soak plus diurnal ABT program will be calculated differently in the 2017 model year and the 2018 and later model years. For 2017 calculations will be based on sales in the U.S. excluding California and the section 177 states which have adopted the LEV III/ZEV programs. For 2018 and later model years it will be based on all 50 states. Calculations will use the following equation: Credits = (fleet average standard−fleet average FEL) × “U.S. sales”. The “fleet average standard” term here is the applicable Tier 3 hot soak plus diurnal standard for the vehicle category from Table IV-19. The sales number used in the 2018 and later MY calculation will be the number of vehicles of the evaporative emission families in that category sold in the U.S. which are subject to the Tier 3 evaporative emission standards. Emission credits banked under the evaporative emission ABT program will have a five year credit life and will not be discounted. This means the credits will maintain their full value through the fifth model year after the model year in which they are generated. At the beginning of the sixth model year after they are generated, the credits will expire and cannot be used by the manufacturer. We are limiting credit life so there is a reasonable overlap between credit generating and credit using vehicles. As mentioned above, for purposes of the compliance calculation, allowance vehicles will have an FEL equivalent to the EPA emission standard (Table IV-19) for their respective vehicle category.

iii. Compliance Demonstration

Demonstration of compliance with the evaporative emissions standards is done after the end of each model year. There are two steps. In the first step, as discussed above, manufacturers must show compliance with the applicable phase-in percentages from the primary phase-in scheme (i.e., 40, 60, 80, and 100), the 20/20 option for MY 2017, or an alternative phase-in percentage scheme. It is sales from these families together with their respective FELs which will be used to make the demonstration of compliance with the emission standard on average within each vehicle averaging set. Compliant vehicle types for these purposes are the same as described in Section IV.C.1.c above for projected sales. If the required sales percentages are not met by direct sales or allowances, non-Tier 3 vehicles would have to be identified to make up the shortfall in this calculation but would not be subject to the canister bleed or leak standard requirements.

In the second step, using the FELs, manufacturers calculate the sales-weighted average emission levels within each of the four vehicle categories using sales for each family. (356) Manufacturers are allowed to use credits only within a defined averaging set. The averaging sets are: (1) LDVs and LDT1s, (2) LDT2s, (3) LDT3s, LDT4s, and MDPVs, and (4) HDGVs. These sales-weighted calculated values must be at or below the emission standard for that vehicle category as shown in Table IV-19, (unless credits from ABT are used). If the difference between the standard and the sales-weighted average FEL is a positive value this could generate banked credit available for future use. If the difference between the standard and the sales-weighted average FEL is a negative value this would be a credit deficit which could be covered by previously banked credits. Credit deficits will be allowed to be carried forward through negative banking. However, manufacturers are required to make up any deficits within the three subsequent model years with credits from vehicles in the same averaging set, except as described below. That is, after calculations for the fourth model year are complete, all previous deficits from the preceding model years will have to be resolved by credits generated by the manufacturer or acquired through trading from vehicles within the same averaging set. As an illustration, a credit deficit accumulated in MY 2017 would have to be eliminated not later than the time that the 2020 MY ABT calculation is submitted to EPA. In no case will a manufacturer be permitted to carry a deficit (negative credit balance) for more than three consecutive model years. Using a similar illustration, all credit deficits accumulated in MYs 2017, 2018, and 2019 would have to be eliminated not later than the time that the 2020 MY ABT calculation is submitted to EPA.

As discussed above, manufacturers are required to identify and include in the calculations for each of the four averaging sets, vehicle families from each of the vehicle categories (see Table IV-19) until the total annual nationwide sales in the given model year equals or exceeds the prescribed percentages. This could include non-Tier 3 vehicles. If the inclusion of non-Tier 3 vehicles results in an exceedance of the hot soak plus diurnal emission standard for that category of vehicles, the credit deficit would have to be made up in a subsequent model year. Credits from banking and trading can be used to cover deficits at any time within the appropriate averaging set.

Allowances can also be used to demonstrate compliance with the percentage phase-in requirements and the vehicle category average emission standard. For purposes of the percentage phase-in requirements, vehicles which have earned allowances are counted as compliant in the percentage calculation. For purposes of the calculations for compliance with the emission standard, allowance vehicles enter into the evaporative emissions compliance calculation as having an emission rate equivalent to the standard for that category of vehicle. Thus, allowance vehicles can help in demonstrating compliance with the percentage phase-in requirement (up to ten percentage points per model year in the alternative phase-in scheme) and can help in reducing deficits since their calculation value is equivalent to the level of the standard.

As presented in detail above, during the 2017-2021 MYs EPA is allowing manufacturers limited flexibility to meet the percentage phase-in requirements using carryover certification data from vehicles certified to CARB PZEV zero evap and CARB LEV III Option 1 standards in the 2015 or 2016 model years. These vehicles may have certification values slightly higher than those of EPA's Tier 3 program for the given vehicle and vehicle category. Since the emission standard values in Table IV-19 and Table IV-20 are very similar for any given vehicle category, for purposes of simplification during the phase in, EPA in the final rule provides that any CARB PZEV zero evap or CARB LEV III Option 1 vehicles used in the 2017-2021MYs emission standard compliance determination be entered into the calculation with the emission level equivalent to the Tier 3 vehicle category in which the vehicle model would otherwise fit. However, we are not allowing manufacturers to generate emission credits for families certified with EPA based on carryover CARB PZEV zero evap or CARB LEV III Option 1 evaporative emissions data as provided for in Table IV-20. We are not including these vehicles in the ABT program since the programs are not directly comparable, and the structure of the current CARB ZEV program, which is the genesis of most PZEV zero evap offerings, allows for a different number of PZEV sales as a function of manufacturer size and CARB LEV III Option 1 does not permit averaging.

As mentioned above, we are limiting use of credits to only within a defined averaging set. Cost effective technology is available to meet the hot soak plus diurnal emission standards on average within each of the vehicle categories in the averaging sets, especially since the standards are designed to accommodate nonfuel hydrocarbon background emissions. Thus, further flexibility is not needed. Moreover, we are constraining averaging to within these sets because of equity issues for the manufacturers. We are concerned that in the absence of such constraints the four or five manufacturers with a wide variety of product offerings in most or all of these categories would have a competitive advantage over the majority of manufacturers which have more limited product lines. This effect could be even more pronounced if the number of evaporative families is considered, since larger more diverse manufacturers have more models and thus more evaporative families.

Nonetheless, manufacturer use of credits from different averaging sets to demonstrate compliance is permitted in limited cases. As noted above, if a manufacturer has a credit deficit at the end of a model year in a given averaging set, they will have to use credits from the same averaging set during the next three model years to make up the deficit. However, if a deficit still exists at the end of the third year (i.e., the deficit has existed for three consecutive model years), we are incorporating provisions to permit a manufacturer to use banked or traded credits from a different averaging set to cover the remaining deficit in the fourth model year's ABT calculation, with the following limitations. Manufacturers are able to use credits from the LDV and LDT1 averaging set to address remaining deficits in the LDT2 averaging set, and vice versa. Furthermore, manufacturers are permitted to use credits from the LDT3, LDT4, and MDPV averaging set to address remaining deficits in the HDGV averaging set, and vice versa. No other use of credit exchanges across different averaging sets is allowed. These restrictions are being finalized because of equity concerns caused by the different nature and size of various manufacturer product lines.

For both the percentage phase-in and sales-weighted average calculation steps above, we are basing the calculation on nationwide sales (excluding California and the section 177 states which have adopted the LEVIII/ZEV programs) in the 2017 MY since the anti-backsliding provisions of the LEV III evaporative emissions program are in place through the 2017 MY. The program uses annual nationwide sales beginning in the 2018 MY. We believe this approach is consistent with the manufacturers' plans for 50-state vehicles. A program design which enables a nationwide program has been an important premise of this rulemaking. Furthermore, this is simpler for the manufacturers and for EPA since it relieves the need to project future model year sales or track past model year sales at a disaggregated level. We recognize that decisions by the manufacturers on a national fleet versus a bifurcated approach such as exists today (California and the section 177 states which have adopted the LEVIII/ZEV programs separate from the rest of U.S. sales) may not yet have been made. The CARB LEV III and EPA phase-in requirements are identical beginning in 2018, so EPA sees little need for concern that a nationwide-based accounting approach could lead to disproportionate state by state impacts or the encouragement of practices which would lead to any particular state or area not receiving the anticipated emission reductions with this nationwide approach to the calculation.

As discussed above, manufacturers not meeting the percentage phase-in requirements will need to include non-Tier 3 vehicles in the count and include their emissions in the overall calculation of compliance with the hot soak plus diurnal standard and resolve shortfalls in compliance with the emission standard with future reductions, earned allowances, or credits. These non-Tier 3 vehicles would not be subject to leak standard or canister bleed standard requirements. The additional vehicles could only be meeting the Tier 2 hot soak plus diurnal requirements and adding these vehicle families/vehicles into the calculation may result in a credit deficit. A manufacturer could not have an unresolved deficit for more than three model years as discussed below. The deficit would have to be eliminated with positive credits not later than the ABT calculation and credit reconciliation which occurs after the fourth model year.

Resolving this sales percentage shortfall problem becomes a bit more complicated for the 2017 MY 20/20 option because it requires that 20 percent of vehicles meet the Tier 3 evaporative emission requirements and that 20 percent meet the leak standard. These may or may not be the same vehicles. As a means to resolve this potential problem, EPA is requiring that any shortfall of either of the 20 percent values (Tier 3 evaporative or leak standard) for the 2017 MY be covered by allowances or by future sales of vehicles meeting the Tier 3 evaporative emission requirements in excess of the evaporative emission percentage sales requirement for that MY or some combination of MYs. For example, if a manufacturer was five percentage points short of either the 20 percentage points for the hot soak plus diurnal or the 20 percentage points for the leak standard in the 2017 MY, then it will have to accelerate sales of vehicles meeting Tier 3 evaporative emission requirements in the 2018-2021 MYs to cover the 5 percentage points (e.g., 65 percent in 2018 instead of 60 percent or 63 percent in 2018 MY and 62 percent in the 2019 MY, etc.). These vehicles as Tier 3 vehicles in MY 2018 or later would also have to meet the leak standard.

e. Small Volume Manufacturers

As flexibility, we are establishing provisions for small volume manufacturers and for those small business manufacturers and operationally independent small volume manufacturers with average annual nationwide sales of 5,000 units or less. (357) These manufacturers would be permitted to delay meeting the Tier 3 evaporative emission standards, including the requirement to use EPA certification test fuel, until the 2022 MY. See pages 29892 and 29998-29999 of the preamble to the NPRM and Section IV.G.5 below for a discussion of the 5,000 vehicle threshold. This includes the hot soak plus diurnal standards, the canister bleed emission standard, and the leak standard. In the interim, these vehicles must meet the existing evaporative and refueling emission standards. The initial determination of whether a manufacturer is under the 5,000 unit threshold will be based on the three year average of actual nationwide sales for MYs 2012-2014. This allowance would not be affected if a qualifying manufacturer's nationwide sales later exceed that value before 2022. Similarly, new market entrants (not in the market in the 2012 MY) with projected sales of less than 5,000 units could be covered by the small volume manufacturer provisions. However, in this case if actual running average nationwide sales exceed 5,000 units per year in any three consecutive model years they will have to meet the Tier 3 evaporative requirements in the third model year thereafter. For example, if a new market entrant in 2015 projects nationwide production of 4,000 units per year and the average of actual values in 2015-2017 exceeds 5,000 units per year they will have to meet Tier 3 evaporative requirements by the 2020 MY.

3. Technological Feasibility

Evaporative/refueling emission control systems are an integral part of the overall vehicle engine and fuel system. EPA is establishing two revised and three new standards in this rule (2-/3-day hot soak plus diurnal standards, high altitude standards canister bleed standards, fuel rig SHED standard, leak standard) and a new test fuel which applies to these standards as well as the current running loss, refueling, and spit back emission standards.

Hot soak plus diurnal emissions are fuel vapors which arise from the fuel system when it is parked immediately after operation (hot soak) and during daily ambient heating and cooling or by means of permeation when the vehicle is at rest. Control of hot soak plus diurnal emissions is primarily achieved by routing fuel vapors to a canister filled with activated carbon. These vapors are stored on the carbon and purged in the engine during vehicle operation. Hot soak plus diurnal emission rates vary with fuel vapor pressure, temperature, and fuel system design. Permeation emissions have been reduced by improving fuel tank and fuel line materials. Permeation emissions are sensitive to the gasoline ethanol content. While EPA has required ethanol in the fuel used for assessing evaporative system durability since 2004, Tier 3 is the first rule to require the certification test fuel for gasoline-fueled vehicles to include ethanol (E10).

Canister bleed emissions are fuel vapors which diffuse from the canister vent as a result of the normal redistribution of vapors within the activated carbon while the vehicle is at rest. The emission rate depends on the tank volume, its fill quantity, the size and architecture of the canister and the characteristics of the carbon itself. While the biggest effect of this vapor redistribution is a uniform vapor concentration within the canister, it can also cause vapors to escape through the canister vent even without continued canister loading resulting from fuel tank heating.

Vapor leaks in the vehicle fuel/evaporative system can arise from micro-cracks or other flaws in various fuel/evaporative system component structures or welds, problems with component installations, and more generally from connections between components and fuel lines and vapor lines. Control of leaks is especially important to achieving full useful life emission control system performance.

In Tier 3, the emissions test fuel is changing from 9 RVP E0 to 9 RVP E10. EPA does not expect the change in emissions test fuel to affect refueling, spit back, or running loss compliance technology or strategies.

While these elements of the evaporative/refueling program are separate requirements for compliance purposes, the integrated nature of the design and operation of the evaporative/refueling control systems and the vehicle engine/fuel systems often leads to co-benefits when technology is added or upgraded. In some cases technology to meet one of the new or revised evaporative emission requirements will either help in efforts to meet other evaporative type requirements or enhance durability. For example, technology used to address the canister bleed standard will also reduce hot soak plus diurnal emissions and technology to meet the leak standard will reduce hot soak plus diurnal emissions and enhance durability.

Based on review of current certification data and the documentation in current professional literature, there is no doubt that the technology is available to meet the final evaporative emission standards described in this rule. (358) There are at least 50 vehicle models which met the requirements in 2013. (359) There are many technologies manufacturers can consider which will reduce emissions and enhance durability. Manufacturer compliance options and cost considerations are also addressed by the phase-in flexibilities and as the ABT program.

In the NPRM we described a variety of technology approaches and calibrations which manufacturers could use to meet the Tier 3 evaporative emission requirements. No comments were provided on the stringency of the standards, the technologies, the feasibility of the standards, or the costs of compliance. Nonetheless, we updated our technology analysis in light of new certification data and vehicle technology projections. As in the analysis supporting the NPRM, we identified technologies on the basis of their control effectiveness and cost to implement. Not every model will use every technology described below. Rather we expect manufacturers to apply the technologies needed on any given model to meet the compliance target level. The technologies could be broadly grouped into two segments. The first are those expected to see widespread use based on their effectiveness and cost to implement. The second are those which are in relatively widespread use today, but could be optimized if necessary to achieve further reductions. In many cases the reductions available from this second group are relatively small and the costs are slightly higher than for the other strategies. The anticipated control technologies to comply with the hot soak plus diurnal, canister bleed/rig, and leak standards are described briefly below and are grouped in these two basic segments. A more detailed analysis for each vehicle category is found in Chapter 1 of the Regulatory Impact Analysis (RIA).

a. Technologies expected to see widespread use: Engine/fuel system conversion: As projected in our RIA for the 2017-2025 light-duty GHG emissions final rule, EPA projects a significant movement from port fuel injection (PFI) engines to gasoline direct injection (GDI) engines. This ranges from 60-100 percent of products for all categories except gasoline-powered trucks over 14,000 lbs GVWR. This reduces air induction systems emissions by 90 percent.

Air Induction System (AIS) Scrubber: For vehicles/engine models not converted to GDI, EPA projects the use of an AIS scrubber as is now used on some PZEV models. These would reduce air induction system emissions by 85 percent.

Canister honeycomb: This is a lower gasoline working capacity activated carbon device designed to load and purge very easily and quickly. This device reduces canister bleed emissions by 90 percent but also provides control for the hot soak plus diurnal test.

Reduce leaks from connections and improve seals and o-rings: Vapor leaks from connections and the emission rates from these leaks is exacerbated if poor sealing techniques or low grade seal materials are use in connectors such as o-rings. Reducing connections in the fuel and evaporative systems and improving techniques and materials would reduce these emissions by 90 percent. This would reduce hot soak plus diurnal emissions, improve durability, and help to assure compliance with the leak standard.

Move parts into the fuel tank: Another means to reduce leak-related vapor emissions is to move fuel evaporative system parts which are external to the fuel tank to the inside. Emissions from these parts would be completely eliminated. This would reduce hot soak plus diurnal emissions, improve durability, and help to assure compliance with the leak standard.

OBD evaporative system leak monitoring: Beginning in the 2017 model year, the OBD system will need to be able to find, confirm, and signal a leak in the evaporative system of 0.020 inches cumulative diameter or greater. This is currently done on most vehicles less than 14,000 lbs GVWR as a result of the manufacturers' response to meeting CARB requirements, but will be mandatory under EPA regulations.

b. Technologies expected to be optimized if necessary to achieve further reductions:

In the NPRM, EPA discussed a number of other technologies with the demonstrated potential to further reduce evaporative emissions. These included: (1) Upgrading the activated carbon canister and optimizing purge calibrations (especially for larger displacement engines), (2) upgrading fuel line materials to reduce permeation, (3) improving the fuel tank barrier layer to reduce permeation, (4) improving fuel tank manufacturing processes to reduce tank seam permeation emissions, (5) upgrading the fuel tank fill tube material to reduce permeation, and (6) improving the security of the fill tube connection to the fuel tank. While each of these approaches reduces evaporative emissions, they are to large degrees in use today. Thus their further application may be limited to specific situations. It is worth noting, that the use of these technologies has contributed to the relatively large compliance margins under the existing hot soak plus diurnal standards.

The reductions required and cost of compliance for any given vehicle model will depend on its current certification level and the type of evaporative emission control technology applied. The baseline emission values for 2-day hot soak plus diurnal evaporative emission certification for current models range from 0.42-0.96 grams per test (g/test). Achieving the desired compliance targets (at least 25 percent below the Tier 3 standard) would require reductions ranging from 0.12 g/test for LDT2s to 0.51 g/test for HDGVs. (360) EPA estimates 2025MY costs in the range of $9-15 per vehicle with a fuel cost savings of about $2 over the vehicle life. The application of the technologies expected to see widespread use under Tier 3 will create the margins need for compliance and in some cases create excess reductions which could be used to generate credits for ABT.

4. Heavy-Duty Gasoline Vehicle (HDGV) Requirements
a. Background on HDGV

HDGVs are gasoline-powered vehicles with either a GVWR of greater than 8,500 lbs, or a vehicle curb weight of more than 6,000 lbs, or a basic vehicle frontal area in excess of 45 square feet. (361) HDGVs are predominantly but not exclusively commercial vehicles, mostly trucks and other work type vehicles built on a truck chassis. EPA often discusses HDGVs in three basic categories for regulatory purposes according to their GVWR class. These are Class 2b (8,501-10,000 lbs GVWR), Class 3 (10,001-14,000 lbs GVWR), and Class 4 and above (over 14,000 lbs GVWR). These are further sub-categorized into complete and incomplete vehicles. (362) Class 2b HDGVs are mostly produced by the manufacturers as complete vehicles and are very similar to lower GVWR LDTs of the same basic model sold by the manufacturers. Class 3 HDGVs are also built from LDT chassis with fuel system designs that are similar to their Class 2b and LDT counterparts, but these are on some occasions sent to secondary manufacturers as incomplete vehicles to attach a load carrying device or container. EPA estimates that more than 95 percent of Class 2b/3 vehicles are complete when they leave the original equipment manufacturer (OEM). Class 4 and above HDGVs are built on a more traditional heavy-truck chassis and in most cases leave the OEM as an incomplete vehicle. For Class 2b/3 vehicles, it is common to certify the vehicle for emissions purposes (exhaust, evaporative, etc) as a full chassis, while for Class 4 and above the vehicle is certified as a chassis for evaporative emissions while the engine is dynamometer certified for exhaust emissions.

HDGVs have been subject to evaporative emission standards since the mid 1980s. Recently, the timing of the standards has lagged requirements for LDVs and LDTs by several years, but the standards are of comparable stringency when vehicle size and fuel tank volume are considered. The most recent 2/3 day hot soak plus diurnal standards for HDGVs took effect in 2008. Refueling control requirements apply to complete Class 2b vehicles only. These requirements phased-in over the period from 2004-2006.

b. HDGV Evaporative Emission Control Requirements

As discussed above, EPA is including HDGVs within the Tier 3 evaporative emissions program. The hot soak plus diurnal and canister bleed test emission standards that will apply to these HDGVs are presented in Table IV-19 and-Table IV-20 and the high altitude standard is presented in Table IV-21. These vehicles will be included in the averaging calculation beginning in the 2018 MY and will be eligible for creating and using allowances and credits.

Furthermore, for the reasons discussed below, EPA is requiring that all complete HDGVs regardless of their GVWR be required to meet the refueling emission standards and use the test procedures currently required for LDVs and LDTs and complete Class 2b vehicles. (See § 86.1813-17). In their comments, manufacturers expressed concern about the amount of gasoline used in the development and certification of refueling emission control systems for HDGVs (due to the larger fuel tanks). To address this concern, EPA will permit manufacturers to certify using two separate processes for vehicles with tanks of 40 gallons or larger. The first will be the engineering evaluation of canister and purge data from lighter weight HDGVs certified in the SHED to show that similar or scaled-up systems on heavier HDGVs have the purge volume and canister working capacity to pass the refueling standard. This could include a comparison of control system design elements such as canister shape, canister internal architecture, total canister volume, and total gasoline working capacity as well as purge air volume over the Federal Test Procedure. This would be subject to the application of good engineering judgment. The second is application of the provisions of 40 CFR 86.153-98 (a) through (b)(1) on a bench set up for a tank of the appropriate volume in lieu of a vehicle test to show the efficacy of the fill neck seal. Such a test could be conducted in a conventional SHED.

The ORVR requirement applies to complete Class 3 vehicles by the 2018MY and all other complete HDGVs by the 2022MY. EPA proposed these requirements for Class 3 HDGVs and asked for comment on extending the requirements to all HDGVs. The manufacturers expressly commented that HDGV ORVR requirements should be limited to complete HDGVs. (363) There are only four manufacturers of HDGVs. Of these, three offer complete products in the Class 3 weight range and none offer complete products in the Class 4 and above weight range. As mentioned above, Class 3 vehicles have largely the same vehicle chassis and fuel system configurations as Class 2b vehicles. The manufacturers of complete Class 2b vehicles indicated to the CARB and EPA that they carry across their Class 2b fuel evaporative control system designs onto Class 3 and this includes the onboard refueling vapor recovery (ORVR) system used for control of refueling emissions. Thus, applying refueling emission controls to complete Class 3 vehicles adds no cost and has little additional emission reduction benefit. However, it does set a requirement to continue these controls in future model years. There are no complete Class 4 and above HDGVs and neither manufacturer who certifies incomplete HDGVs above 14,000 lbs GVWR objected to establishing an ORVR requirement for complete HDGVs. (364) This sector is made of incomplete HDGV chassis and diesel-powered products. However, setting a requirement for potential future Class 4 and above designs establishes certainty for manufacturers but brings no near term cost burden or emission reductions.

Incomplete HDGVs make up 15-20 percent of all HDGV sales. Of this, approximately 80 percent are Class 2b/3 and 20 percent are Class 4 and above. EPA is not extending the refueling emission control requirement to incomplete HDGVs at this time. The control system designs would be essentially the same as on complete HDGVs, but manufacturers have indicated to EPA that they would have to establish additional measures to ensure that the steps taken to complete the vehicle by the secondary manufacturer do not compromise the integrity and safety of the fuel/evaporative control system (including ORVR) and that the ORVR system continues to perform properly with regard to emissions control. While there are relatively few of these vehicles, their contributions to the inventory are larger than might be expected due to their lower fuel economy. Given these contributions, EPA may consider proposing to apply ORVR to incomplete HDGVs in a future action.

EPA is also including a provision that manufacturers be permitted to comply with the refueling emission standard as early as the 2015 MY to earn on a one-to-one basis allowances which could be used to phase-in the Class 3 refueling emission control requirement or as an allowance on a 2:1 basis under the Tier 3 evaporative emission program. EPA believes this is appropriate since the expected daily average reduction in vehicle refueling emissions for this class of vehicles is large relative to the reduction in evaporative emissions expected under Tier 3. This would also apply to any incomplete HDGV a manufacturer voluntarily certified to the refueling emission standards. Any certifications, including those done early, must use EPA Tier 3 test procedures and certification test fuels or CARB LEV III equivalents.

c. Other Program Elements for HDGVs

In the NPRM, EPA sought comment on several provisions related to Tier 3 certification test fuel and evaporative emission control requirements.

First, EPA sought comment on whether heavy-duty gasoline engines (HDGEs) not subject to new Tier 3 exhaust emission standards (those certified for exhaust emissions using an engine dynamometer) which are used in HDGVs subject to Tier 3 evaporative emission standards should certify for exhaust emissions on Tier 3 emissions test fuel. (365) Manufacturers responded by asking that the use of Tier 3 fuel for HDGE exhaust emissions certification be voluntary, but agreed that the use of Tier 3 certification fuel would not change the stringency of the current dynamometer-based emission standards or the costs of compliance. Based on consultations with manufacturers, EPA has decided to require that all HDGEs be certified on Tier 3 fuel by the 2022MY. (366) To provide flexibility for very unique applications or circumstances, EPA will allow up to five percent of a manufacturer's dynamometer-certified HDGE sales in any given model year to be certified using Tier 2 certification fuel. This flexibility is limited to certification based on carryover data beginning in the 2022MY.

Second, as discussed in Section IV.F.5 for light-duty vehicles, we are committed to the principle of ensuring that any change in test fuel for heavy-duty gasoline vehicles/engines will not affect the stringency of either the fuel consumption or GHG emissions standards. As part of the separate rulemaking discussed in Section IV.F.5, we expect to establish the appropriate test procedure adjustment for HD engine fuel consumption standards and to determine the need for any test procedure adjustment for GHG emissions standards based on the change in certification test fuels.

Third, to simplify the evaporative emission regulations for HDGVs and to bring them more in line with the current structure of the product offerings in this sector, we are finalizing provisions to permit evaporative emissions certification by engineering analysis for vehicles above 14,000 lbs GVWR (instead of above 26,000 lbs GVWR as permitted in the existing regulations). We are also finalizing regulatory language to clarify how these provisions are to be implemented. This applies to the hot soak plus diurnal, running loss, and canister bleed standards. These HDGVs will remain subject to the emission standards when tested using the specified procedures. This is the same cut point allowed by CARB and will allow for one certification method. Even though it was supported by one commenter, we are not including specific provisions for design-based certification for HDGVs over 14,000 lbs GVWR. EPA believes that the option to certify using engineering analysis and data serves the same purpose.

Fourth, we are finalizing a revised description of evaporative emission families that does not reference sealing methods for carburetors or air cleaners as this technology is now obsolete for HDGEs.

Fifth, EPA is finalizing regulatory language permitting HDGVs over 14,000 lbs GVWR to be grouped with those between 10,001 and 14,000 lbs GVWR for purposes of complying with evaporative and refueling emission control standards and related provisions. In these cases, we require these HDGVs to meet all the requirements applicable to the group in which they are being included (e.g., useful life, OBD, etc.).

Finally, the regulations at 40 CFR part 86, subpart M, describe how to test heavy-duty vehicles above 14,000 lbs GVWR to demonstrate compliance with evaporative emission standards. Most of these provisions are identical to those that apply under 40 CFR part 86, subpart B. We are eliminating subpart M and replacing it with a simple instruction to test these heavy-duty vehicles using the procedures of subpart B, with a small number of appropriate modifications noted as exceptions to the light-duty test procedures. Relying on references to subpart B instead of largely copying them into subpart M eliminates many pages of unnecessary regulatory text and makes it easier to maintain a consistent set of requirements. Changing a provision in subpart B in the future will automatically apply for evaporative testing of both light-duty and heavy-duty vehicles unless otherwise provided in the particular rulemaking.

In response to comments received, we are specifying that heavy-duty vehicles above 14,000 lbs GVWR must use the same drive schedules and test fuels that apply for light-duty vehicles. Subpart M already allows light-duty drive schedules and certification test fuels as an alternative to using those for heavy-duty vehicles, and most if not all manufacturers of these vehicles already use the light-duty drive schedules, which facilitates testing simplicity and coordination of design parameters with light-duty vehicles. The heavy-duty drive schedule generally involves less driving, which makes this the more stringent test option for designing purge. Omitting this more stringent option therefore does not change the effective stringency of the applicable standards.

With these changes from the proposed rule, there are only two aspects of testing that are different for heavy-duty vehicles above 14,000 lbs GVWR. First, the regulations specify that the exhaust emission measurements are not required for the driving portion of the test between canister pre-conditioning and diurnal testing. Exhaust emission standards in this vehicle size range apply based on engine testing only. Second, wider engine speed tolerances apply. This is captured in part 1066 by specifying wider engine speed tolerances for any testing that does not require exhaust emission measurements since the greater allowance has no effect on emissions measurements. This applies, for example, for pre-conditioning drives for light-duty vehicles, and it also applies for pre-conditioning related to evaporative emissions of heavy-duty vehicles above 14,000 lbs GVWR.

There are some differences in the existing test provisions in subparts B and M that we are not preserving. Some of these differences arose from changes to subpart B that were inadvertently not carried over to subpart M. In other cases, there may have been an intentional distinction that no longer applies (such as provisions related to slippage on twin-roll dynamometers). Also, we are not retaining distinctions in subpart M related to procedures for determining road load settings and for operating manual or automatic transmissions. Additional differences we are not preserving include gas divider specifications, SHED and dynamometer calibration procedures, and some provisions for alternative canister loading and vehicle pre-conditioning. We are also restoring the content of § 86.1235(b) through (i) related to dynamometer operating procedures, which were inadvertently removed in an earlier rulemaking.

5. Evaporative Emission Requirements for FFVs

A flexible fuel vehicle (FFV) as defined in 40 CFR 86.1301-01 means any motor vehicle engineered and designed to be operated on a petroleum fuel and on a methanol or ethanol fuel or any mixture of the petroleum fuel and methanol or ethanol. Many manufacturers have one or more FFVs in their product offerings. These include many different LDV and LDT vehicle chassis styles including passenger cars, mini-vans, pick-ups, sport utility vehicles and even a few HDGVs.

The EPA regulations implementing the FFV provisions for ethanol FFVs, including those in 40 CFR 86.1811-04 and 86.1811-09, have been applied primarily for FFVs capable of operating on gasoline/ethanol mixtures up to E85. As a matter of policy, EPA has not required certification testing for evaporative and refueling emissions on the full range of E0-E85 fuel blends, but instead has allowed the option to use a blend created when Tier 2 fuel (9 RVP E0) is splash blended with ethanol to a 10 percent gasoline/ethanol blend. This simulates what often occurs in the vehicle fuel tank when Tier 2 fuel (9 RVP E0) is dispensed into a tank containing mostly E85. This yields a blend which has a Reid vapor pressure of about 10 psi. Nearly all manufacturers have certified using this option. The California ARB LEV III program has no special evaporative or refueling emission test fuel requirements for FFVs.

In the Tier 3 NPRM, EPA proposed to revise the certification test fuel for evaporative emissions, to revise the hot soak plus diurnal emission standard, and to add a canister bleed emission standard and a leak standard. These standards apply to FFVs and non-FFVs. EPA proposed to revise the ethanol content of the certification test fuel for refueling emissions but did not otherwise propose to change the fuel vapor pressure, the level of the refueling emission standard or the test procedure. Furthermore, in the NPRM, EPA sought comment on leaving unchanged the basic approach to FFV certification test fuel for Tier 3 evaporative and refueling emissions, except that the certification test fuel would be 9 RVP E0 splash blended with E15 such that the blend would have a 10 psi vapor pressure, i.e., the RVP of the evaporative emissions test fuel used by nearly all manufacturers. Manufacturers commented that the Tier 3 certification test fuel should be the same for FFVs and non-FFVs and that carryover should be permitted from Tier 2 to Tier 3. EPA met with several manufacturers to clarify their comments and to discuss issues affecting the evaporative and refueling emissions certification fuel for FFVs. (367)

For FFVs, EPA has several factors to consider for evaporative and refueling emission certification test fuel. First, EPA is finalizing a 9 RVP E10 certification test fuel for non-FFVs for evaporative and refueling emissions. This is consistent with our broader policy objective to allow the manufacturers to sell the same vehicles in all 50 states. Second, 10 psi RVP certification test fuel for the Tier 3 evaporative emission standards for FFVs could result in more evaporative emission reductions than a 9 psi RVP test fuel, but this would be counter to the broader policy objective regarding a national program since CARB has no separate FFV evaporative emission standards and likely would affect the stringency of the final evaporative emission standards. Specifically, finalizing 10 psi RVP certification test fuel for the Tier 3 evaporative emission standards as applied to FFVs would increase the stringency of the evaporative emission standards for FFVs both compared to the Tier 3 evaporative emission standards with 9 psi RVP test fuel for non-FFVs and compared to the Tier 2/MSAT evaporative emission standards with 10 psi RVP test fuel for FFVs. Third, we are not changing the level of the refueling emission standard (though we are adding ethanol to the test fuel and extending ORVR to complete Class 3 HDGVs) and we did not examine how a potential change from the existing 10 psi RVP test fuel for FFV refueling would affect in-use emission reductions or the stringency of the refueling standard for FFVs. A change in the test fuel vapor pressure likely would likely lead to a change in the stringency of the refueling emission standard as they are now applied to FFVs. Retaining the current requirements for refueling emissions for FFVs does not affect the national program since CARB currently follows Federal Test Procedures and test fuels for ORVR.

Balancing all of these factors, EPA is adopting a bifurcated scheme for evaporative and refueling emission certification for Tier 3. Evaporative emission requirements for the hot soak plus diurnal, canister bleed, running loss, spit back, and leak standards will be based on Tier 3 certification fuel (9 RVP E10) for FFVs. This will permit reciprocity between the LEVIII and Tier 3 evaporative emission standards programs and subject the manufacturers to only one set of evaporative emission tests for FFVs and non-FFVs. However, for the refueling emission standard, EPA is retaining the 10 psi certification test fuel requirement for FFVs because the worst case in-use RVP conditions when E0 and E85 are commingled will still be possible. In current systems, the fuel vapor pressure in the refueling emission test drives the total gasoline working capacity of the activated carbon canister that is necessary in the integrated evaporative/refueling control system. Although a 10 psi RVP certification fuel for evaporative emissions control could be viewed as more stringent, we believe that keeping the fuel vapor pressure at 10 psi in the refueling test, which is what was proposed for comment, will help to assure that the in-use emission reduction benefits of current evaporative systems on FFVs are retained. We expect that total canister gasoline working capacities will still be driven by the 10 psi RVP fuel used in the refueling test and therefore the higher in-use RVP conditions which impact evaporative emissions will still be addressed.

EPA is specifying a 10 RVP E10 test fuel specification for FFV refueling emissions certification. However, as a compliance alternative EPA will continue to permit certification based on in vehicle fuel tank blending of two different fuels (i.e., vehicle fuel tank filled to 10 percent of capacity with E85 and then refueled to at least 95 percent of capacity with (9 RVP E0). Either of these approaches will also meet CARB certification test fuel requirements as the test fuel vapor pressure would be higher than with EPA's 9 RVP E10 or CARB's 7 RVP E10 test fuel. In addition, we are not changing existing requirements that all IUVP testing for evaporative and refueling tests are done on the non-FFV fuel (i.e., Tier 2 IUVP vehicles are tested on 9 RVP E0 and Tier 3 IUVP vehicles are tested on 9 RVP E10.

In their comments on the Tier 3 NPRM, manufacturers asked that EPA allow carryover of certification emission data from Tier 2 to Tier 3. Since the regulatory approach for refueling emissions is basically the same as what is currently being used by the manufacturers, we believe there should be opportunity for carryover of refueling emission data under the current regulatory program. Manufacturers also expressed concern that the refueling emission standard would require them to keep a 10 RVP E10 or 9 RVP E0 test fuel solely for refueling emission standard certification purposes. To help address this concern, in certification testing, EPA would consider approving other refueling test fuel blends with 10 percent ethanol and 10 psi such as a refueling event where a tank is filled initially with 10 percent E85 and during refueling test is filled with 90 percent 9 RVP E0. EPA would also permit manufacturers the option to seek EPA approval to certify by attestation using alternative procedures or through engineering analysis based on similar evaporative/refueling emission system configurations and emission test results and data on similar vehicles showing that the vehicle could pass the refueling emission standard and meet the requirements in use on 10 psi RVP E10 fuel. They would remain subject to confirmatory testing on 10 RVP E10. Both of these options could only be implemented with approval of the Administrator.

6. Test Procedures and Certification Test Fuel
a. Review and Update of Testing Requirements

EPA adopted the current test requirements for controlling evaporative emissions in 1993. (368) Those changes included: (1) Diurnal testing based on heating and cooling the ambient air in the SHED  (369) instead of forcing fuel temperatures through a specified temperature excursion; (2) repeated 24-hour diurnal measurements to capture both permeation and diurnal emissions; (3) high-temperature hot soak testing; (4) high-temperature running-loss measurements with a separate standard, including controlled fuel temperatures according to a fuel-temperature profile developed for the vehicle; and (5) canister preconditioning to ensure that vehicles could effectively create canister capacity to prepare for several days of non-driving.

These test procedures are generally referred to as “enhanced evap” testing. EPA adopted these “enhanced evap” test procedures in coordination with CARB. The test requirements include two separate test sequences to demonstrate the effectiveness of evaporative emission controls. The “2-day sequence” involves canister loading to two-gram breakthrough, followed by driving for the exhaust test (about 31 minutes), a hot soak test, and two days of cycled ambient temperatures. The “3-day sequence” involves canister loading with 50 percent more vapor than needed to reach breakthrough, followed by driving for the exhaust test, driving for the running loss test (about 97 minutes total), a high-temperature hot-soak test, and three days of cycled ambient temperature.

The 2-day sequence was intended primarily to insure a purge strategy which would create enough canister capacity to capture two days of diurnal emissions after limited driving. The two-day measurement period is also effective for requiring control of permeation and other fugitive emissions. The 3-day sequence was intended to establish a design benchmark for achieving adequate canister storage capacity to allow for several days of parking on hot summer days, in addition to requiring vehicle designs that prevent emissions during high-temperature driving and shutdown conditions.

After adopting these evaporative test procedures, we set new standards for refueling emissions control which called for onboard refueling vapor recovery (ORVR). (370) Manufacturers have typically designed their ORVR systems to be integrated with their evaporative controls, using a single canister and purge strategy to manage all fuel vapors vented from the fuel tank. Due to the magnitude of the refueling emission load and the manner in which the load rates affect activated carbon capture efficiency, it has become clear that ORVR testing with these integrated systems serves as the benchmark for achieving adequate canister storage capacity.

In the nearly 20 years since adopting these test procedures, manufacturers have made great strides in developing designs and technologies to manage canister loading and purging and to reduce permeation emissions. Except as discussed below, we are not changing the test procedures for demonstrating compliance with the Tier 3 emission standards. As described above, we are adopting a new standard based on measured values over a canister bleed test, and a fuel system rig test. These are intended to measure only fuel vapors which diffuse from the evaporative canister or permeate/leak from a fuel system. CARB developed these procedures as a means for setting standards that are not affected by nonfuel background emissions. The canister bleed test procedure is a variation of the established two-day test sequence. The canister is preconditioned by purging and loading to breakthrough, then attached to an appropriate test vehicle for driving over the duty cycle for the exhaust test. The canister is then attached to a fuel tank for measurement. After a stabilization period, the tank and canister undergo two days of temperature cycling. Canister emissions are measured using a flame ionization detector (FID), with a conventional SHED approach or by collecting emissions in a bag and measuring the mass. Rather than repeating CARB's regulations, we are incorporating those regulations by reference into the CFR. (371) This will avoid the possibility of complications related to minor differences that may occur with separate test procedures. The fuel system rig test is a bench test where a complete vehicle fuel system (without the vehicle chassis) is constructed in the SHED and evaluated over the 3-day cycle in both a “wet” and “dry” state. (372)

CARB adopted the fuel system “rig test” as an optional approach to demonstrate control of evaporative emissions without the effects of the nonfuel hydrocarbon emissions that are seen in testing the whole vehicle in the SHED. We generally expect manufacturers to comply with the EPA requirements which include the canister bleed test and emission standard instead of CARB LEV III Option 1 which includes the rig test and emission standard. However, since we are accepting PZEV zero evap and CARB LEV III Option 1 certifications for the 2017-2018 MYs and 2017-2021 MYs, respectively, we are also incorporating by reference CARB's rig test into the CFR to accommodate those manufacturers that do in fact rely on this approach.

Also, as discussed further below, we are adopting a new leak test procedure which will be used to measure leak rates for the leak standard. The leak test standard test procedure is contained in the regulatory text.

Manufacturers have raised a pair of related concerns regarding the current test procedures. First, hybrid vehicles and new engine designs for meeting fuel economy standards and CO 2 emission standards increase the challenge of maintaining an adequate purge volume to prepare vehicles for the diurnal test. For hybrid vehicles this is related to the amount of time the engine is running. For other technologies this is related to the trend toward decreasing available vacuum in the intake manifold, which is the principal means of drawing purge air through the canister. Second, preconditioning the canister by loading to breakthrough serves as a disincentive for some control strategies that might otherwise be effective at reducing emissions, such as designs involving greater canister capacity or better containment of fuel vapors inside the fuel tank. In addition, we have learned from studying in-use emissions and in-use driving behaviors and usage patterns that it is not uncommon for vehicles to go for an extended period with little or no opportunity to purge the canister.

In the NPRM, we requested comment on an optional adjustment to the test procedure intended to address these three concerns. In this alternative, for designs involving pressurized tanks, manufacturers would determine an alternative vapor load to precondition the canister before the exhaust test. If, for example, a fuel system is designed to stay sealed up to 1 psi and to vent vapors to the canister if rising temperatures trigger a pressure-relief valve, the manufacturer could quantify the actual vapor load to the canister during three consecutive days of cycling through diurnal test temperatures. This three-day vapor load would be the amount of fuel vapor used to precondition the canister (loaded at the established rate of 15 grams per hour). This canister loading may also involve butane instead of fuel vapor, but we would likely require a greater mass of butane to account for the fact that it is easier to remove the butane from the activated carbon in the canister. This approach would be flexible to accommodate any design target for pressurizing fuel tanks. Canister preconditioning for the ORVR test (for integrated and nonintegrated systems) would remain unchanged. EPA sees merit in further consideration of such test procedure flexibilities, but auto manufacturers did not provide support these concepts in their comments and we are not adopting the proposed optional adjustment.

b. Test Fuel for Certification

EPA is changing the certification test fuel specifications as described in Section IV.F. Here we discuss some implications for evaporative and refueling emissions testing beyond those discussed above for FFVs. We are revising the certification test fuel specification in conjunction with the Tier 3 standards, principally to include ethanol and reduce sulfur such that the test fuel better aligns with the current and projected in-use fuel. Although we received unsolicited comment asking that we set durability test fuel specifications for evaporative and refueling emission control systems to be the same as those for the certification test fuel (9 RVP E10 in this final rule), we are not changing durability fuel specifications in this rule other than to remove minimum sulfur content requirements. In particular, we are not changing the existing requirement that “any mileage accumulation method for evaporative emissions must employ gasoline fuel for the entire mileage accumulation period which contains ethanol in, at least, the highest concentration permissible in gasoline under federal law and that is commercially available in any state in the United States”. See §§ 86.1824-08(f)(1) and 86.113-04(a)(3)(i). EPA believes this is prudent policy to ensure that emission control systems are designed for the fuels with the potential to adversely affect durability and there is no reason to change the existing approach especially since E15 fuel is now legally permissible and commercially available for appropriate vehicles and there is potential for its market penetration to increase in the future. Any bench aging using E15 fuel must simulate the effects of alcohol in-use fuels on evaporative emission system components.

Since there are already vehicles in the market which employ the technology needed to meet the new hot soak plus diurnal requirements, EPA is taking a flexible approach to the phase-in of the certification test fuel. This is summarized in Table IV-22.

To accommodate vehicles already designed to meet CARB PZEV zero evap evaporative emission requirements, EPA's phase-in provides that PZEV zero evap vehicles which qualify for carryover can use CARB Phase 2 fuel for evaporative emissions (hot soak plus diurnal and running loss standards) and rig test certification for MYs 2015-2019. For CARB PZEV zero evap vehicles, high altitude, refueling, and spit back standard certification may use either EPA Tier 2 or Tier 3 fuel in MYs 2015-2019. For the leak standard in the 2018 and later MYs, they must use Tier 3 test fuel. Beginning in the 2017 MY, the use of PZEV zero evap data is limited to carryover of data from 2015 or 2016 MY certifications.

Those using CARB LEV III Option 1 can use CARB Phase 3 fuel for evaporative emissions (hot soak plus diurnal and running loss standards) and rig test certification for MYs 2015-2021. For CARB Option 1, high altitude, refueling, and spit back standard certification must may use Tier 2 or Tier 3 fuel in MYs 2015-2016 but in the 2017 and later MYs all LEV III option 1 certifications for the high altitude, refueling, spit back, and leak standards must use EPA Tier 3 fuel.

CARB LEV III Option 2 evaporative emission vehicles may use CARB Phase 3 fuel to meet evaporative (hot soak plus diurnal and running loss standards) and canister bleed standards beginning in 2015 MY and following. High altitude, refueling, and spit back may use Tier 2 or Tier 3 fuel in model years 2015 and 2016. For 2017 and later model years CARB LEV III option 2 evaporative families must use Tier 3 test fuel for high altitude, refueling, spit back, and leak standard certifications.

Tier 3 evaporative emission vehicles must use Tier 3 fuel to meet evaporative emission (hot soak plus diurnal and running loss standards), high altitude, canister bleed, and refueling/spit back emission standards beginning in the 2015 MY and following. Beginning in the 2018 MY, Tier 3 vehicles must use Tier 3 emission test fuel to demonstrate compliance with the leak standard requirements. (373)

When the program is fully phased-in, any Tier 3 evaporative emission certification will have to use Tier 3 certification test fuel and test procedures or CARB equivalent test procedures and fuels. This could be done as early as the 2015 MY and will be required for all vehicle models by the 2022 MY. (374) As indicated above and in Table IV-22, we are further applying the new test fuel at the same time to ORVR testing. Therefore, beginning in the 2017 MY if manufacturers do any new testing to demonstrate compliance with the Tier 3 evaporative emission standards (using Tier 3 or LEV III fuel), they will need to submit test data to demonstrate compliance with the refueling emission standards using the new certification test fuel as well as the leak (when applicable), spit back, canister bleed, running loss, and high altitude emission standards. Any family that is not yet captured within the Tier 3 phase-in percentage may remain on Tier 2 certification fuel through the 2021 MY. By the 2022 MY all evaporative and refueling emission certifications will have to be on EPA test procedures and certification fuels or CARB equivalents as identified in the regulations. Policies regarding test procedures and test fuels for EPA confirmatory and other post certification testing are discussed in Section IV.C.6.e below.

Finally, we are including provisions to allow any vehicle certified to the refueling spit back standard separately (mostly incomplete HDGVs)to continue to do so using Tier 2 current certification fuel until the 2022 MY even if its evaporative emissions are certified on Tier 3 certification fuel. This is reasonable since the fill quality of the vehicle and eliminating spit back are not necessarily related to the ethanol or sulfur content of the gasoline. The manufacturers must meet this requirement through testing, as the engineering evaluation flexibility available for HDGVs over 14,000 lbs GVWR does not apply to this standard.

c. Correction for Ethanol Portion of the SHED Measurement

Another issue related to adding ethanol to the certification test fuel relates to the emission measurement in the SHED. Emissions are detected by flame ionization detectors (FID), which are less responsive to ethanol than gasoline. This effect causes under-reporting from the ethanol portion of the fuel vapor. Fuel-related emissions from the vehicle may be slightly more weighted toward ethanol than gasoline, depending on how the different fuel constituents permeate through various fuel-system materials, how they evaporate from the bulk fuel in the tank at varying temperatures, and how they adsorb onto and desorb from the activated carbon in the canister. We proposed to address this issue by the use of a prescribed correction factor. Under this approach manufacturers would simply multiply their SHED measurement results by a fixed value to adjust upward for the difference in the FID response to ethanol. Data available to EPA at the time of the NPRM suggested that a value of approximately 1.1 would be appropriate for E15. (375) For an E10 certification fuel, California ARB finalized a value of 1.08.

In their comments, the manufacturers supported the use of a correction factor, but stipulated that the value put forth by EPA was too large and they should be given the option to measure the ethanol fraction of the vapor in the SHED through procedures and instrumental approaches prescribed in the regulations (see 40 CFR 1065.269, 1065.369, and 1065.805) instead of using a fixed correction value. Two manufacturers provided data based on testing with E10 test fuel which generally showed lower ethanol fractions than represented by the 1.1 value proposed by EPA for hot soak plus diurnal emissions, and uniformly showed very low ethanol fractions for refueling measurements. (376)

EPA has reviewed the data provided by the manufacturers and has considered their comment that they should be given the option to measure the ethanol fraction and adjust the SHED results rather than be required to use a fixed correction factor. Based on these considerations, EPA is establishing the following approach with regard to ethanol corrections. First, EPA will permit measurement or the use of a fixed correction factor on an evaporative family by evaporative family basis. However, once the manufacturer selects an approach for any given evaporative family, that approach must be used in all subsequent testing of all vehicles certified using that data including carry over. For example, if a manufacturer chooses to measure the ethanol fraction for purposes of certification of a test group in a given model year, that same method must be used in any manufacturer confirmatory testing as well as IUVP or IUCP testing of all vehicles in that test group. Alternatively, if a manufacturer uses the fixed correction factor in certification it must also use it for all evaporative emission tests covered by the requirement for a given test group and for all follow on testing. Second, the decision on measurement or correction factor must be uniform on a test group basis for all evaporative emission standards covered by the correction requirement. In this case this includes hot soak, diurnal, high altitude, running loss, and rig test measurements. Third, in terms of a fixed correction factor, EPA believes that the 1.08 value adopted by California is consistent with the data and is specifying that value for hot soak plus diurnal (low and high altitude), running loss, and rig test measurement corrections for any testing conducted with 10 percent ethanol. Based on the data provided by the manufacturers, EPA is not requiring a fixed correction value or measurement for refueling, spit back, or canister bleed measurements for testing conducted with 10 percent ethanol. This aligns with the expectation that ethanol concentrations will be very low with FID-based measurements and that mass-based measurements will capture any ethanol adequately without a need for correction. Finally, EPA will use the method selected by the manufacturer in any confirmatory or surveillance testing. However, since corrections will always be zero or greater, no correction is needed to make a failure determination if the FID value exceeds the emission standard or FEL. With regard to the 1.08, EPA remains open to future revisions to this value, in coordination with CARB, if a fuller data set representative of various vehicle models, SHED FID ethanol response values, FID designs (analog vs. digital), ethanol calculation approaches (photo acoustic and impinger), and test sites demonstrates that a different value would be technically appropriate and adequately conservative relative to the direct measurement methods permitted in 40 CFR 1065.

For higher ethanol blends (such as E85), the regulation already specifies measurement and calculation procedures to adjust for this effect. We are not making any changes to these procedures.

d. Vehicle Preconditioning for Nonfuel Hydrocarbon Emissions for the Tier 3 Evaporative Emission Standards

The Tier 3 hot soak plus diurnal, leak, and canister bleed emission standards taken together are expected to bring about the widespread use of technology which effectively eliminates fuel vapor emissions. The fuel rig, canister bleed, and leak standards are not influenced by nonfuel hydrocarbon emissions from the vehicle. Nonfuel hydrocarbon emissions from the vehicle are measured as part of SHED emission testing, and are indistinguishable from fuel hydrocarbons when a FID is used to measure the concentration. The level of these nonfuel hydrocarbon emissions vary by vehicle and component design and material. These emissions arise from paint, adhesives, plastics, fuel/vapor lines, tires, and other rubber or polymer components and are generally greater with larger size vehicles. These nonfuel hydrocarbon emissions are usually highest with newly manufactured vehicles and decrease relatively quickly over time.

Currently, manufacturers normally conduct some preconditioning to reduce or eliminate the effects of these nonfuel hydrocarbon emissions on evaporative emissions measurements in the SHED. In the past, this practice has not been addressed through regulatory provisions. However, given the stringent level of the Tier 3 hot soak plus diurnal evaporative emission standards, and that nonfuel hydrocarbon emissions are expected to be a significant portion of the hydrocarbon emissions measured in the SHED, EPA believes that some sort of preconditioning before certification testing is appropriate and that a regulatory provision addressing this practice is warranted. Providing some recognition of and allowance for this practice will help to create the proper balance between necessary and proper preconditioning to address high nonfuel hydrocarbon emissions and excessive preconditioning which could undermine the intent of the hot soak plus diurnal emission standard (~ 50 mg or less of fuel evaporative emissions). EPA believes the goal of evaporative emissions preconditioning should be to get nonfuel hydrocarbon emissions to what we call vehicle background levels. A working definition of vehicle background level might be the level which will occur naturally twelve months after production. A provision in the regulations which addresses preconditioning reduces ambiguity for the manufacturers and could reduce or eliminate any uncertainty in the true meaning of certification test results.

Manufacturer activity with regard to preconditioning often involves two practices. First, manufacturers in some cases “bake” their test vehicles at temperatures of 50 °C or higher for periods of up to ten or more days to accelerate the off-gassing of these nonfuel hydrocarbon emissions before testing is conducted. While this practice is common, there is no standardized method or protocol for this preconditioning prior to new vehicle certification testing. For example, some manufacturers bake for a set period of time in a climate chamber while others bake in the climate chamber and periodically measure nonfuel background in a SHED until an acceptable or stable level of nonfuel hydrocarbon emissions is achieved. Second, manufacturers often remove, modify, or clean certain components which are the largest source of nonfuel hydrocarbon emissions. Preconditioning could also include measures to eliminate minor fuel drips, spills, or other fuel remnants which occur as a result of vehicle preparation for testing.

We are not specifying standardized pre-conditioning practices or protocols with regard to addressing nonfuel hydrocarbon emissions before evaporative emission certification testing. However, we are finalizing general provisions in four areas. First, we specify in the regulations that preconditioning for the purpose of addressing nonfuel hydrocarbon emissions is permitted. Second, we specify that any preconditioning is voluntary. Third, we specify that if preconditioning is conducted, the details must be specified to EPA before certification testing, (i.e., at the time of the pre-certification planning meeting). The goal of this preconditioning should be to get nonfuel hydrocarbon emissions to vehicle background levels as discussed above. The specifics to be discussed with EPA could include details on vehicle baking practices such the temperature and time duration in the climate chamber and practices conducted as an alternative or complement to vehicle baking such as installing used tires (drive and spare) on certification vehicles, and allowing the windshield washer tank to be filled only with water. EPA's goal in these discussions is to gain certainty that manufacturers are not preconditioning vehicles so severely that they create a level of nonfuel hydrocarbons that is artificially low and would not occur in use and thereby creating a false additional compliance margin for fuel hydrocarbons in the certification test. Fourth, except as discussed below we are providing in the regulations that no pre-conditioning is permitted for testing of any vehicle aged more than twelve months from its date of manufacture. This restriction for vehicles older than 12 months includes certification, confirmatory and in-use testing for any vehicle certified to the Tier 3 evaporative emission standards. For these vehicles, nonfuel hydrocarbon emissions will presumably be reduced to a stable level due to natural off gassing which begins after the vehicle is manufactured. Emissions from any replacement parts or other vehicle maintenance will presumably be encompassed within the margin below the standard created by this natural off-gassing.

EPA received several comments concerning the proposed restriction on pre-conditioning of vehicles older than 12 months from the date of manufacture. The Alliance of Automobile Manufacturers and the Association of Global Automakers asked that baking be permitted if such a vehicle is found to have identifiable contamination due to causes such as a fuel spill, refrigerant leak, or washer fluid leak and that the manufacturer be given the option to age the tires (tires only) from any vehicle where the tires are less than twelve months from manufacture as indicated on the sidewall. CARB asked that EPA only allow the use of an aged spare tire in any testing and not spare tire removal. EPA generally agrees with these commenters and is finalizing provisions for limited flexibility subject to EPA approval. Under these provisions manufacturers may be permitted to clean any spills or leaks but not to bake the entire vehicle. Baking of tires less than 12 months old may also be permitted with EPA prior approval. Vehicles must be tested with a spare tire in place since emissions from the spare tire were considered as the standard was developed. Manufacturers may exchange a new spare tire for one that is baked or aged. Finally, one manufacturer indicated that there may be circumstances where the base chassis for a certification vehicle was used in previous certification but that this base chassis was modified for a new model year and cleaned, reconfigured, and recertified with new components which affect background emissions. (377) While EPA believes this would be a rare occurrence, regulatory provisions in this rule allow EPA to approve additional pre-conditioning for vehicles in this situation upon manufacturer request and justification.

e. Reciprocity With CARB

Over the past 15 years EPA's “enhanced evap” test procedures have been based on testing with 9 pound per square inch (psi) RVP gasoline with test temperatures representing a summer day with peak temperatures of about 96 °F. CARB adopted the same basic procedures, but specified that testing should occur with 7 psi RVP gasoline at temperatures of up to 105 °F. EPA and CARB agreed that certification could be based on testing with either EPA or CARB conditions and that these provided equivalent stringency for purposes of evaporative control system design. However, the provision allowing for this equivalence of test data preserved EPA's ability to also test with either EPA or CARB temperature conditions and related test fuels. CARB always specified EPA test conditions for refueling as they were deemed worst case. CARB recently changed their certification test fuel to a 7 RVP gasoline with 10 percent ethanol and as discussed in Section IV.F, we are changing the Federal certification test fuel specification to a 9 RVP gasoline with 10 percent ethanol.

During the development of this FRM we carefully considered the practice of CARB/EPA reciprocity with regard to certification test fuels, hot soak plus diurnal test procedures, running loss test procedures, and emission test results when it comes to evaporative emissions certification. Based on these considerations and the alignment of the ethanol content for the EPA and CARB certification fuels, we have decided to retain our current approach with regard to CARB/EPA reciprocity for evaporative and refueling emissions. EPA and CARB have agreed to continue accepting emission test data on each other's test fuels and temperature conditions for certification such that a uniform national program for certification test fuel will be able to exist. For model years during the evaporative emissions standard phase-in discussed above (ending after the 2021 MY), EPA will conduct any post certification testing on any vehicle in the Tier 3 program manufactured in the 2015-2021 MYs using the fuel and temperatures used by the manufacturer for certification. This approach covers families certified using carry over PZEV evaporative emissions data (through the 2019 MY) and LEV III Option 1 certifications (through the 2021 MY). Our program flexibility in the area of test fuels for hot soak plus diurnal, running loss and SHED rig/canister bleed emission standards is summarized in Table IV-25. After the 2021 model year, EPA will retain the option to test on either set of temperatures/fuels. This applies to all evaporative emission standards (hot soak plus diurnal, running loss, and canister bleed). For the other emission standards (refueling, leak, spit back, and high altitude hot soak plus diurnal) EPA will use the test fuel used by the manufacturer through the 2019 model year. For the 2020 model year and later we may use Tier 3 fuel or California Phase 3 if its use is permitted for certification. Please refer to the regulatory text for specific provisions.

EPA will review all Tier 3 program evaporative emissions data. If the data shows that the EPA and CARB based test requirements give fully equivalent results, in the future we may revise our regulations so that a vehicle is always tested on the fuel used for its initial certification.

Table IV-25—Tier 3 Evaporative Emissions Program Options and Test Fuels
Vehicle programStart MYProgram standardsCert fuelEPA test fuel for confirmatory,surveillance & IUVPEnd MY for use in Tier 3
PZEV zero evap2015Hot soak + diurnal, running loss & SHED rigCA Ph. 2Fuel used by the manufacturerAfter 2019 MY.
LEV III Opt. 12015Hot soak + diurnal, running loss & SHED rigCA Ph. 3CA Phase 3 through 2019 MY, after EPA may use Tier 3 or CA Phase 3After 2021 MY.
LEV III Opt. 22015Hot soak + diurnal, running loss & canister bleedCA Ph. 3CA Phase 3 through 2019 MY, after EPA may use Tier 3 or CA Phase 3N/A.
Tier 32015Hot soak + diurnal, running loss & canister bleedTier 3Tier 3N/A.
As shown in Table IV-22, to qualify as a Tier 3 vehicle for evaporative emission purposes vehicles must meet the hot soak + diurnal, high altitude, rig/canister bleed, running loss, refueling, and spit back standards. The leak standard applies beginning in the 2018 MY and the SHED rig/canister bleed tests are program specific.     

Generally, a vehicle test group using Tier 3 certification fuel and test procedures for meeting the various evaporative and refueling emission standards will qualify for inclusion in the Tier 3 evaporative emission standards phase-in. However, EPA recognizes that the California and federal evaporative emission standard programs are starting from different bases and that the transition provisions are different in some ways. For example, the EPA program starts in the 2017 MY but after that has the same basic program construct as CARB in 2018. However, prior to the 2017 MY, CARB has a ZEV program provision which will continue to bring zero evap technology into the fleet before the 2017 MY and CARB also allows early LEV III Option 1 and Option 2 evaporative emission certifications. To capitalize on this technology and to facilitate transition, we are finalizing provisions that any CARB evaporative emission test data from MYs 2015 and 2016 PZEV zero evap certifications (hot soak plus diurnal and running loss) can be used in federal certification for those evaporative families through the 2019 MY. Similarly, we are finalizing provisions that CARB LEV III Option 1 certifications (hot soak plus diurnal and running loss) can be used in federal certification for those evaporative families through the 2021 MY. Assuming the vehicle test groups also meet the Tier 3 high altitude evaporative emission standards, the refueling emission standard, the spit back standard, and the leak standard when applicable, they could be included in the percentage phase-in calculations as Tier 3 vehicles. If the vehicles do not meet the Tier 3 evaporative emission requirements manufacturers could potentially sell them nationwide, but they could not be included as Tier 3 compliant vehicles in the percentage phase-in calculation. Table IV-22 provides a concise summary of the requirements a vehicle must meet to qualify as a Tier 3 vehicle during the program's early, transition, and phase-in periods.

EPA proposed a similar provision for a manufacturer who elects to use the CARB test procedures and test fuels to meet the refueling emission standard. However, no manufacturer indicated interest in their comments and we have decided not to include reciprocity for this provision in the Tier 3 program. While experimental data based on field bench testing suggests that the CARB test fuel RVP and dispensed temperature together would give the same results as the EPA test fuel RVP and dispensed temperature there are no vehicle test data in the record at this time. CARB has always accepted refueling and spit back certification on EPA test fuel and will continue to do so in the future. This provision would have added another layer of complexity to the program and was not necessary since the refueling and evaporative tests are done separately.

f. Evaporative and Refueling Emission Standards for Various Fuels

The evaporative and refueling emission standards apply in different ways to different fuels. First, with regard to the evaporative emission standards, Clean Air Act section 202(k) specifies that gasoline-fueled vehicles must be certified to evaporative emission standards. Section 202(a) authorizes EPA to establish evaporative emission standards for other fuels. Today evaporative emission standards apply to LDVs, LDTs, MDPVs, and HDVs fueled by gasoline methanol, ethanol, natural gas, and liquified petroleum gas (LPG). For the refueling emission standard the situation is quite different. Section 202(a)(6) of the Clean Air Act specifies that the refueling emission standards apply to all LDVs regardless of the fuel used. Section 202(a) of the Clean Air Act authorizes EPA to establish emission standards for other fuels and classes of vehicles. Prior to the Tier 3 final rule, the refueling emission standards applied to all vehicles less than 10,000 lbs GVWR regardless of the fuel used.

In the NPRM, EPA requested comment on applying the refueling standards to all vehicles regardless of fuel used. This would include all volatile fuels. (378) The evaporative standards apply today to all volatile fuels  (379) (except for diesel) and we asked for comment on explicitly including dedicated ethanol as well as fuel-cell vehicles, and electric vehicles. EPA also requested comment on applying the refueling and evaporative standards only to vehicles using volatile liquid fuels instead of all volatile fuels.

EPA received four comments on this issue. One commenter expressed the view that evaporative requirements should be expanded to apply to volatile liquid fuels plus liquified petroleum gas (LPG) and liquified natural gas (LNG) while the three other commenters did not see the need to apply the requirements to any gaseous fueled vehicle or other vehicle using a non-volatile liquid fuel because these vehicle fuel systems are sealed and rarely vent during normal operation or never vent at all.

As is discussed further in the Summary and Analysis of comments, based on the comments, the fuel properties, and current industry fuel system design practices, EPA has decided to retain the requirement that the evaporative and refueling emission standards apply to vehicles using any volatile fuel. For gaseous fueled vehicles (LPG and LNG/CNG vehicles), only the Tier 3 3-day hot soak plus diurnal and running loss standards apply. For the other volatile fuels all of the Tier 3 evaporative emission standards apply. For the refueling emission standard the requirements apply to all complete vehicles less than 10,000 lbs GVWR regardless of the fuel used. This is not being changed, except that the requirement will not apply to diesel-powered LDTs and HDVs vehicles. For vehicles over 10,000 lbs GVWR, the refueling emission standards will apply only to complete vehicles. This includes LPG, CNG, LNG, and dedicated ethanol or methanol vehicles. While the test procedures for these standards would apply, EPA is including regulatory provisions to permit manufacturers to certify based on related data, engineering analysis, and compliance with published consensus standards. We are not applying these requirements to electric or fuel cell vehicles.

For vehicles equal to or less than 8,500 lbs GVWR, the Tier 3 evaporative and refueling emission standards for alternative fuel vehicles apply to each vehicle of a vehicle evaporative/refueling family as the family is included in the manufacturer's phase-in for the Tier 3 evaporative emission standards. For vehicles over 8,500 lbs GVWR, the application of the Tier 3 evaporative emission standards depends on the Job 1 (first build) date for the vehicle evaporative family. If the Job 1 date for a vehicle model is before the fourth anniversary date of the signature of the rule then the Tier 3 evaporative emission standards do not apply until the next model year. If the Job 1 date is after the fourth anniversary date, the Tier 3 evaporative emission standards apply in that model year. This determines when the vehicle is to be included in the denominator of the percentage phase-in calculation. The refueling emission standard applies only to complete vehicles and we are applying the same phase-in requirements as for complete HDGVs. For complete vehicles between 10,000 and 14,000 lbs GVWR the refueling emission standard applies in the 2018 model year. For complete vehicles with a GVWR in excess of 14,000 lbs GVWR, compliance is required in the 2022 model year. Finally, for all small businesses, the Tier 3 evaporative and refueling emission standards do not apply until the 2022 model year.

g. Other Changes and Future Considerations

This rulemaking included consideration of several amendments or clarifications to existing requirements related to evaporative emissions. As part of this process, EPA has concluded that the following provisions warrant adjustment, clarification, or correction:

  • Even though the evaporative emission standards in 40 CFR part 86 apply to the same engines and vehicles that must meet exhaust emission standards, we require a separate certificate for complying with evaporative and refueling emission standards. An important related point to note is that the evaporative and refueling emission standards always apply to the vehicle, while the exhaust emission standards may apply to either the engine or the vehicle. Since we plan to apply evaporative/refueling/leak standard and the recently adopted greenhouse gas standards to vehicle manufacturers, we believe it will be advantageous to have the regulations related to their certification requirements written together as much as possible to reduce burden and increase efficiency. Therefore, for 2015 and later model years, we are moving the emission standards and certification requirements for HDGVs from 40 CFR part 86 to the new 40 CFR part 1037, which was originally used for greenhouse gas standards for heavy-duty highway vehicles. This is not intended to change the requirements that apply to these vehicles, except as noted in this section.
  • Section 86.1810-01 contains specifications addressing whether diesel fuel vehicles can be waived from demonstrating compliance with the refueling emission standard through testing. In the existing regulation the potential for a waiver from testing depended on the diesel fuel having an RVP equal to or less than 1 psi and the fuel tank having a temperature which does not exceed 130 °F. We have examined this provision and are withdrawing the fuel temperature limit specification. Short of fuel spillage in the SHED, EPA sees no likelihood that a diesel fueled vehicle with RVP less than 1 psi could fail the refueling emission standard even at fuel tank temperatures above 130 °F. This is due to the inherently low vapor pressure of diesel at these temperatures and the likelihood that vapor shrinkage conditions will occur in the fuel tank during refueling since the dispensed fuel will be much cooler than the tank fuel.
  • When adopting the most recent prior set of evaporative emission regulatory changes we did not carry through the changes applying evaporative emission standards to vehicles using methanol-fueled compression-ignition engines. This final rule corrects this oversight.
  • We are finalizing provisions to address which standards apply when an auxiliary (nonroad) engine is installed in a motor vehicle, which is currently not directly addressed in the highway regulation. The approach requires testing complete vehicles with any auxiliary engines (and the corresponding fuel-system components). Incomplete vehicles are to be tested without the auxiliary engines, but any such engines and the corresponding fuel-system components will need to meet the standards that apply under our nonroad program as specified in 40 CFR part 1060.
  • We are removing the option for secondary vehicle manufacturers to use a larger fuel tank capacity than is specified by the certifying manufacturer without re-certifying the vehicle. Secondary vehicle manufacturers needing a greater fuel tank capacity must either work with the certifying manufacturer to include the larger tank, or go through the effort to re-certify the vehicle. This provision has not been used and is better handled as part of certification rather than managing a separate process. We are including corresponding changes to the emission control information label.
  • We are revising the provisions for setting the vehicle air conditioning controls during the running loss portion of the evaporative emissions test cycle to simply reference the specifications for exhaust emission testing described in 40 CFR part 1066. This allows test labs to use a uniform set of test procedures for setting up test vehicles. This change is expected to have no effect on the stringency of the running loss test.
  • EPA regulations at § 86.1824-01 permit manufacturers to develop their full-useful life deterioration factors for evaporative and refueling emission standards based on the use of good engineering judgment. These factors are additive in nature, and when added to the “undeteriorated low mileage” test value the sum must be less than the applicable emission standard or FEL. Manufacturers usually certify such that this summed value falls below the emission standard or FEL enough to provide a margin for in-use compliance and to address variability and other uncertainty. Regulations (at § 86.1824-08) require that evaporative emissions durability assessments must employ gasoline fuel for the entire mileage accumulation period which contains ethanol in, at least, the highest concentration permissible in gasoline under federal law and that is commercially available in any state in the United States (currently E15). In their comments the Alliance of Automobile Manufacturers and the Association of Global Automakers asked to be able to use evaporative emissions deterioration factors from Tier 2/LEV II assessments even if the assessed or measured full life emission value used to determine the deterioration factor from the Tier 2/LEV II 2 testing is above the Tier 3/LEV III emission standard for the vehicle category of interest. (This situation, which is often referred to as line crossing, is not prohibited in the EPA regulation.)  (380) Thus, EPA is permitting the use of this data but requires that: (1) The manufacturers use good engineering judgment in the testing used to develop their deterioration factors and the assessment and application of this data in developing deterioration factors, (2) the manufacturers use the evaporative/refueling emissions test fuel as stipulated in the regulations for Tier 3, and (3) the addition of the deterioration factor to the low mileage test result does not result in an exceedance of the emission standard or the FEL cap for that category of vehicles.

D. Improvements to In-Use Performance of Fuel Vapor Control Systems

1. Reasons for Adding a Leak Test Standard

As emission standards approach zero, as in the “zero evap” standards discussed above, in-use performance becomes critical for vehicles to meet the standards over their useful life periods and provide the expected emission reductions. Fuel vapor control system leaks are not a new problem, in fact it was one of the main reasons for replacing the canister method for assessing evaporative emissions with the enclosure (SHED test) method used today. (381) However, as emission standards have become more stringent, test procedures have improved, and vehicle lifetimes have increased, any malfunction or deterioration in the system causes significant emissions increases. Even a small leak can cause large amounts of HC vapor. Therefore, the prevalence of leaks in the fleet can have a significant effect on the average evaporative emissions overall.

As discussed in detail in the NPRM, recent laboratory and field data  (382) show very high emissions from vehicles with liquid/vapor leaks. Field studies have indicated approximately 10 percent of overall fleet have significantly elevated evaporative emissions. The studies show that this frequency increases as vehicles age. The Coordinating Research Council (CRC) E-77 programs randomly recruited sixteen vehicles and almost half had some type of leak. Emissions related to these leaks grew in magnitude over the course of the program which lasted a few years. In addition, the EPA recently completed a test program to gather information on running loss emissions with implanted leaks of varying sizes, locations and fuel volatility. (383) Data from this study is not included in the modeling analysis for this final rule, but the results show that there are significant emissions from leaks while driving as the fuel tank temperature rises. Therefore the reductions from the future prevention of leaks will be larger than our current estimates. These data led EPA to examine the OBD-based evaporative system leak data available from I/M programs from several states to more accurately gauge the rate of leaks above the 0.020 inch monitoring threshold met by most manufacturers as a result of CARB's 2004 model year OBD II requirements. (384) These are important data because even a vehicle with a fuel/evaporative system leak as small as 0.020 inches would be expected to fail the Tier 3 evaporative emission standard in a SHED test and in fact emit 4-5 times above the Tier 3 emission standard on a daily basis due to the number of vehicle trips per day.

We examined data for vehicles meeting CARB's OBDII evaporative emission leak monitoring requirements as well as either the CARB/EPA enhanced evaporative emission or Tier2/LEV II evaporative emission standards. Since the data were gathered by the states under different protocols and time periods, the content of the data sets is not identical. To provide some degree of uniformity in our analysis, we examined the data for model years 2000 and later, but within each state we only looked at calendar years of data beginning after the initial state I/M exemption period had passed (2-6 calendar years depending on the state). Thus the analysis focused on I/M OBD information for calendar years 2004-2012.

Examined together, the data generally indicate the following.

  • For all our States analyzed, the trend lines show that between 2-4 percent of the vehicles entering the I/M program (at about 2 years old) have a “not ready” evaporative monitor. The percentage increased to between 8-11 percent as the vehicle aged to 8 years old with a rate increase of approximately 1 percent per year as the vehicle ages.
  • The model years and time periods analyzed for the four States shows approximately 0.7-2.5 percent of vehicles overall with a “ready” evap monitor had one or more stored evap DTCs, indicating a potential evaporative emissions-related problem as defined in the OBD regulations.
  • A further review of the data shows that, overall, in the three States with an enforced OBD program approximately 0.7-1.6 percent of vehicles with a “ready” evap monitor had one or more stored evaporative emissions related DTCs. The fourth State, which does not enforce the OBD test, had a higher percentage (2.5 percent) of evap monitor “ready” vehicles that had stored evap related DTCs.
  • For the same model years and time periods analyzed for the three States with enforced OBD programs, EPA estimates about 0.5 percent of vehicles with a “ready” evap monitor evaluated at four years old in an I/M program had a stored DTC. This rate increased at a rate of about 0.15 percent per year and was about 1.1 percent for vehicles at 8 years old. For the fourth state, which does not enforce OBD evaporative results, EPA estimates about 1.4 percent of vehicles evaluated at four years old had a stored DTC. This rate increased at a rate of about 0.5 percent per year and was about 3.5 percent for vehicles at 8 years old.
  • Analyzing each state's data for specific evaporative DTCs, over 50 percent of all evaporative codes were for evaporative system leaks. The second most common category (15-20 percent) involved some sort of error in the operation of the purge flow control which could also contribute to evaporative leaks.
  • The monitor “ready” rates are relatively uniform for all States analyzed, but the percentage of evaporative emissions related MILs illuminated and the percentage of evaporative system leak related DTCs were larger in the fourth State. EPA believes this is the case because OBD is advisory only in this State's I/M program, meaning that a vehicle could pass its I/M requirement with a MIL illuminated and not have to repair it.

In considering this information for the fleet as a whole, a few other factors must be considered. First, a vehicle can pass its I/M requirements (based on provisions of individual State I/M programs) with the evaporative emissions monitor “not ready”. Second, the vehicle can pass with a pending DTC. Third, it is not uncommon for vehicle repair related to an OBD MIL to occur just before I/M visits. Based on factors such as these, the values presented above are likely to be conservative on a fleet average basis. Beyond this, as discussed in the NPRM, earlier research conducted by EPA and the state of Colorado indicated that OBD is not designed to catch every evaporative system leak and sometimes misses leaks it should have found but did not for various reasons (some determined and some unknown). (385) This suggests that overall leak prevalence is higher than indicated by the OBD data alone.

Estimating a nationwide fleet average leak rate is possible with the limited data available if some informed assumptions are made. Only about 24.5 percent of vehicles in the U.S. are in I/M areas and of these only 20.8 percentage points (~4/5) are in areas which rely on OBD as part of the pass/fail protocol. There is at present no data on the prevalence of evaporative system leaks for vehicles in areas without I/M. However, based on these data it reasonable to assume that the rates in these areas are no less than for areas with I/M (where I/M mandates repair) and are likely similar to or larger than those for the one state analyzed where OBD is advisory only. Under those assumptions, the average leak rate across the country is much higher than for I/M areas alone. For example, if one considers data from the eight year age point in the I/M data for states which require repair, the leak prevalence rate is about 1.4 percent and in the state where OBD is advisory it is 3.5 percent. Weighted by the fleet percentages given above, this indicates a leak rate of about 3.0 percent in the fleet for the eight year age point. This is a conservative estimate based on historic evaporative I/M data. (386)

The propensity for leaks in the vehicle fleet has the potential to reduce the benefits of the Tier 3 evaporative emission standards substantially. If on any given day, as few as 3 percent of Tier 3 vehicles have a leak(s) of 0.020 inches or greater this will cause in-use emissions equivalent to essentially all of the projected emission reductions from the Tier 3 evaporative emission standards on that day. (387)

The leak standard we are adopting will help technology to meet the Tier 3 evaporative emission standards and to improve in use durability. These technology measures (see Section IV.C.3) coupled with the upgrade to the OBD evaporative emissions certification and monitoring requirements to signal problems at smaller threshold diameters (discussed in Section IV.E below) and additions to the IUVP program focused on testing a larger sample of vehicles for fuel/evaporative system leaks in IUVP than for evaporative emission standards alone will help to ensure improved in-use performance of evaporative emission control systems.

Based on the above discussion, there needs to be an increased focus on evaporative emissions durability. Nevertheless, there is no question of the value of OBD leak monitoring for evaporative systems, especially when owners complete needed repairs in response to the DTCs set. The I/M OBD statistics and associated in-use leak values discussed above would be higher without OBD evaporative system leak monitoring. However, these data suggest that EPA OBD regulations in place for 2004 and later model year vehicles will not alone be sufficient to address concerns regarding the emission effects of vapor leaks from the fuel and evaporative control systems. (388)

In the NPRM, EPA included a substantial discussion of the work we conducted on high evaporative emission rates and our rationale for the need for a leak standard to help address these concerns. No commenter challenged the data or the premises for our conclusion that a leak standard was needed. Manufacturers asked that the leak standard be phased-in with the Tier 3 evaporative emission standards and that use of upgraded OBDII evaporative system monitoring capability be included as part of the in-use verification program (IUVP) provisions. Both elements are contained in this final rule. CARB fully supported the proposed leak standard and test procedure and indicated its intent to adopt such provisions after the Tier 3 FRM is adopted.

2. Nature, Scope and Timing of Leak Standard

The evaporative emission standards in this FRM will help to promote widespread use of improved technology and materials which will reduce evaporative emissions in-use. The new requirement for a leak standard and test procedure will help to ensure the durability of Tier 3 evaporative emission control systems nationwide. As discussed in the technological feasibility discussion in Section IV.C above, the actions of manufacturers to meet the Tier 3 evaporative emission standards are expected to address fuel/evaporative system design features which currently have a greater propensity for developing leaks and thus improve in-use durability for evaporative control systems compared to vehicles meeting previous evaporative emission standards. The leak standard will provide added assurance that as the manufacturers design for “zero evap” standards they also design the systems to avoid leaks over the full useful life.

Based on the information described above concerning evaporative emissions in-use, we believe a leak standard is necessary to ensure that vehicles meeting Tier 3 evaporative emission requirements not have evaporative emissions in excess of the Tier 3 standards for their full useful life. Toward that end, we are finalizing a leak standard to be met both at new vehicle certification and in use for IUVP testing. The leak standard will apply beginning in the 2017 MY to vehicles in the 20/20 option for that year and in the 2018 MY and later model years to any vehicle certified to the Tier 3 evaporative emission standards or a CARB carryover vehicle counted toward the sales percentage phase-in requirements discussed in Section IV.C, including LDVs, LDTs, MDPVs, and complete HDGVs up to 14,000 lbs GVWR. The standard will be applicable for the same useful life period as for the evaporative emission standards that apply to the vehicle. The standard will apply to vehicles using volatile fuel (e.g., gasoline, FFV, and methanol fuel vehicles, but not diesel or CNG vehicles).

To be compatible with CARB OBD requirements being met by most manufacturers and the OBD requirements included in this rule, we are specifying that the leak standard be expressed in the form of a cumulative equivalent orifice diameter. We are finalizing a value of 0.02 inches. (389) The standard basically requires that the cumulative equivalent diameter of any orifices or “leaks” in the system not exceed 0.02 inches. This is consistent with California OBD requirements (and those being finalized in this rule as well) that the OBD system be capable of identifying leaks in the fuel/evaporative system of a cumulative equivalent diameter of 0.020 inches. EPA believes a standard at this level is feasible since earlier testing programs identified vehicles with essentially no leaks and it is essentially equivalent to that required for CARB OBD evaporative system leak monitoring. We are finalizing a leak standard of 0.02 inches which with rounding is a bit less stringent than the 0.020 inch OBD evaporative system leak monitoring requirement. EPA believes this level of precision is sufficient to accomplish the air quality objective and yet provides some compliance margin between the standard and the monitor requirement such as is reflected through multipliers for the exhaust emission standards established for other OBD monitors. The leak standard will be specified to one significant digit (e.g., 0.02 inches) but will have to be measured and reported to at least two significant digits.

The leak standard will apply at the time of certification as well as during confirmatory and in-use verification program testing. We do not expect that new vehicles being certified will have a leak problem, and since a vehicle with a leak would likely fail the evaporative emissions SHED test, there is little value in mandating a leak test at certification. Thus, EPA will permit a manufacturer to attest to compliance with the leak standard at certification.

To implement the leak standard within the existing regulatory structure a few minor rule changes are being made. First, existing EPA regulations such as those at § 86.098-24, specify criteria for evaporative/refueling emission families. EPA believes this basic structure is appropriate for the leak standard, with the additional criteria that vehicles in the same evaporative/refueling family must use the same basic approach to OBD leak detection. Significantly different volume fuel tanks would likely also be a family determinant, but we believe this is already covered by the evaporative/refueling family criteria. Second, since the leak standard is a pass/fail requirement and not an emission rate, there is no requirement for the application of a deterioration factor. Third, EPA requires that the manufacturers recommend two or more leak test points for each test group. One of these points should be near the canister/purge valve (ideally in the vapor line between the canister/purge valve and the fuel tank) and the other in the gas cap/fill pipe area. Three points are required for vehicles with two separate evaporative and refueling canisters such as non-integrated ORVR systems which employ two activated carbon canisters and four points are required for vehicles with dual fuel tanks and two separate evaporative/refueling control systems.

EPA believes that linking the timing of the leak standard to the beginning of the phase-in of the Tier 3 evaporative emission standards in the 2018 model year provides adequate lead time and is consistent with the technical rationale supporting the feasibility of the Tier 3 evaporative emission standard.

3. Leak Standard Test Procedure

The fundamental concepts underlying fuel/evaporative system leak test are not new to the manufacturers. There is already a simple leak check in 40 CFR 86.608-98(a)(1)(xii)(A) and in the past at least three states included a fuel/evaporative system pressure leak test in I/M programs. More importantly, all LDVs, LDTs, MDPVs and HDGVs manufactured today have the onboard capability to run a pressure or vacuum leak based check on the vehicle's evaporative emission system as part of OBD evaporative system leak monitoring. These systems employ either positive or negative pressure leak detection pumps or operate based on natural vacuum for negative pressure leak detection. EPA is finalizing a test based on a similar concept of placing the system under a slight positive pressure (but from an external source), measuring the flow needed to maintain that pressure in the fuel/evaporative control system, and converting that flow rate to an equivalent orifice diameter. With regard to the test procedure we will first discuss where the leak test can occur in the FTP test sequence. We will then discuss how the test is to be conducted. EPA proposed this test procedure as part of the NPRM and discussed it extensively in the preamble to the proposed rule, and provided a full draft of the Recommended Practice for comment as an Appendix to the RIA. No comments were received. We are finalizing this test procedure as proposed. (390)

First, when conducted, the leak test should be completed immediately following the first two preconditioning steps within the FTP sequence (see Figure B96-10 in 40 CFR 86.130-96). Thus, the vehicle preconditioning steps for the leak test are: (1) Fill the vehicle fuel tank to 40 percent of capacity using the appropriate certification test fuel and then (2) let the vehicle soak for a minimum of a six hour period at a temperature in the range of 68-86 °F. EPA requires that the test be conducted with 9 RVP E10 test fuel for both certification and IUVP. (391) After preconditioning is complete, the leak test is conducted and the test sequence proceeds as prescribed in subpart B or testing is terminated if the purpose is only to conduct leak testing. EPA believes this modest level of preconditioning is sufficient to create standard conditions which enable repeatable and reliable measurement results. Preconditioning cannot include any prescreening for leaks nor will any tightening of fittings or connections be permitted.

After preconditioning is complete, manufacturers then run the leak test. To fully complete testing on a vehicle, two or more test points are required depending on the fuel evaporative system configuration. All points must pass for the vehicle test to be a pass. As discussed above, one of these points should be near the canister/purge valve (ideally in the vapor line between the canister/purge valve and the fuel tank) and the other in the gas cap/fill pipe area. Three points are required for vehicles with two separate evaporative and refueling canisters such as non-integrated ORVR systems which employ two activated carbon canisters and four points are required for vehicles with dual fuel tanks and two separate evaporative/refueling control systems such as dual tank LDTs. If the fuel/evaporative system has an embedded evaporative system test port then that point can be used. Also, a manufacturer can develop a test rig such as a “fill pipe extension” which screws into the fill pipe opening using cap threads at one end and on the other end has threads to screw the fill pipe vehicle cap in place. Within this extension there must be an access port for the leak test equipment to be attached. Thus, the full system could be tested without any direct intrusion or the need for a separate gas cap assessment. The manufacturer must specify the test points at the time of the pre-certification meeting. If the manufacturer selects an entry point which requires the fuel cap to be removed, then the cap will have to undergo a separate test as is now done in many I/M stations. (392) In this case, tests from both points combined must pass the standard. Manufacturers commented that only one test point was needed, but when asked by EPA they offered no data to counter that provided by EPA in the NPRM which showed the potential for different results at different test point locations for the same vehicle.

The procedure is conducted as follows:

  • Calibrate the testing apparatus and otherwise verify testing apparatus is ready and able to complete the procedure.
  • Seal fuel system so as to pressure test entire system (purge valve, cap, etc.).
  • Attach test apparatus to vehicle's fuel system at selected test point.
  • Pressurize fuel system with nitrogen or another inert gas to at least 2.4 kilopascals (kPa).
  • Allow flow and pressure to stabilize in accordance with specification provided in the regulatory text.
  • Calculate effective leak orifice diameter from measured output flow rate and temperature and pressure data or use apparatus with built in computer providing an equivalent digital readout. Calculate to the nearest 0.01 inch.
  • Calculated effective orifice diameter must be less than or equal to the standard.
  • If leak test is conducted at the fuel cap opening then the manufacturer must also show evidence that the vehicle's fuel cap is performing properly. (393)
  • Use two or more separate test points, near the evaporative canister/purge valve and the other near the fuel cap are required. This is especially important if the fuel cap/fill neck area is isolated from the rest of the fuel/evaporative system as a result of the 40 percent fill or if dual tanks are not otherwise connected through vapor lines.
  • Tests can be void if the test apparatus fails, becomes disconnected, fails to maintain a stable flow rate or pressure, or the test was stopped before completion due to safety considerations or some other relevant vehicle issue.
  • Leak tests at all points (2 or more depending on the fuel tank/evaporative system configuration) must pass for a vehicle to pass. This includes performance within specification for the fuel cap if it is removed for testing.

The test procedure presented above is based on current fuel system designs. In the future, it is reasonable to expect changes in designs of the fuel systems such that the procedure above may need adjustment. EPA will monitor these fuel system changes and modify the test procedure provisions as needed. Furthermore, existing EPA regulations (see § 1065.10(c)) contain provisions which provide the opportunity for manufacturers to seek approval for special or alternate test procedures if from a practical perspective their systems cannot be evaluated under EPA requirements or they have an approach deemed equivalent or better. Any such special or alternative procedures must be reported under § 86.004-21(b)(9).

4. Certification and Compliance

As part of the Compliance Assistance Program (CAP 2000) in-use verification program (IUVP)  (394) the manufacturers began testing the evaporative emissions performance of small samples of in-use vehicles owned and used by the public. These regulations can be found at 40 CFR 86 1845-01, and 1845-04. In 2000, EPA extended this requirement to cover chassis-certified HDVs, which for these purposes are basically all HDGVs up to 14,000 lbs GVWR. (395) The in-use testing for evaporative emissions started in 2004 for 2001 MY LDVs, LDTs, and MDPVs and in 2008 for 2007 MY chassis certified HDGVs. Current IUVP data for evaporative emissions (including LDVs, LDTs, MDPVs, and HDGVs up to 14,000 lbs GVWR) covers about 1800 vehicle tests. These data show that when evaluated in the laboratory using certification test procedures, the vast majority (over 95 percent) of the vehicles pass the evaporative emission standards to which they were certified. While this information is indicative of good in-use performance, it has limitations. First, the test results are for small sample sizes. For the approximately 150 million LDVs, LDTs, MDPVs, and chassis-certified HDGVs produced between 2001 (the start of the IUVP program) and 2010 (latest available data), only about 0.001 percent of vehicles were tested. Second, the IUVP regulations place limits on the age/mileage for vehicle testing. Each model year is tested in two “batches,” nominally at the one and four year age points. One year old vehicles must have at least 10,000 miles and four year old vehicles must have at least 50,000 miles with at least one within the higher mileage group having an odometer reading of at least 75 percent of useful life (90,000 miles for most Tier 2 vehicles). With the even longer useful life periods under Tier 3, attention to in-use durability for evaporative systems becomes even more important. Including the leak standard within the IUVP protocol, as structured in the discussion below, will provide better information to EPA and manufacturers concerning evaporative system performance and help to focus manufacturer efforts on using designs and hardware with full useful life durability in mind.

a. In-Use Verification Program (IUVP) Requirements for the Leak Standard
i. Introduction

We believe it is important to identify leaks since vehicles with leaks are expected to have daily emission rates above the Tier 3 evaporative emissions standards, and the recent laboratory and field data 382 suggest a propensity for the diameter of vehicle leak orifice to get larger over time and thus to have even higher emissions. This is also important because evaporative leak emissions occur virtually every day whether the vehicle is driven or not. Thus identifying potential leak problems is important to capturing the emission benefits of the Tier 3 evaporative emission requirements.

Toward that end, EPA is including assessment of compliance with the leak standard within the IUVP program. In developing the proposed rule, we considered expanding the evaporative emission testing portion of the IUVP program as a means to assess leaks, but we decided to focus on the leak standard because it is less burdensome than a full evaporative emissions SHED test and is a cost effective step toward assessing many aspects of evaporative emissions performance in-use.

EPA believes adding a leak test requirement does not create an unreasonable burden. The test procedure described above is simple to run, inexpensive to conduct in terms of equipment and labor, and can be completed relatively quickly compared to an evaporative emissions test. However, we are retaining the evaporative emissions testing requirements currently in IUVP to monitor broader evaporative control system effectiveness (e.g., purge, canister control efficiency, permeation).

ii. IUVP Test Requirements

We are requiring that the leak test be conducted for each and every vehicle assessed in IUVP for exhaust emissions under 40 CFR 86.1845-04. This will begin for 2017 MY vehicles meeting the leak standard under the 20/20 option and more fully in the 2018 MY certifications for all test groups meeting the new leak standard. The leak test IUVP requirement includes the low and high mileage tests for any exhaust vehicle evaluated for exhaust emissions plus a requirement that there be at least one representative of each evaporative/refueling/leak family evaluated at each mileage/year point. We are finalizing this approach to implementing IUVP for the leak standard in lieu of creating a new set of requirements which would require another set of vehicles to be procured for testing. We are not including the leak test with any evaporative emissions test in IUVP, since a leak will be evident in the results of the evaporative emissions test.

The existing IUVP regulations at § 86.1845-04, Table S04-07, call for test sample sizes on a sliding scale based on annual vehicle sales by test group. This can vary from zero for very small sales test groups to six vehicles for test groups with sales exceeding 250,000. There are more exhaust emission test groups than there are evaporative/refueling test families and exhaust emission test groups may cover one or more of the same evaporative/refueling/leak families, so we expect to receive multiple leak test results for most evaporative/refueling/leak families. This will expand the amount of IUVP data we receive in this important area and improve our ability to assess the overall leak performance for a given evaporative/refueling/leak family and the fleet as a whole.

As discussed above, EPA believes that the fuel and evaporative control system leaks are heavily influenced by age as well as design and other factors. EPA asked comment on extending the age point for leak testing for IUVP beyond the four year point to better assess this effect. However, in the past, manufacturers have expressed concern about the implications of testing older vehicles and about finding vehicles still within their warranty and recall liability periods. EPA believes further consideration of longer year test points is merited for exhaust, evaporative, refueling and leak tests but because such a change could potentially affect all four tests we have decided to defer that action to a broader IUVP program review. Extending the time point for the leak test alone would create a different programmatic test burden in terms of more vehicle procurements than the program laid out above.

iii. Assessment of IUVP Leak Emission Standard Test Results

The existing regulations contain provisions addressing follow-on testing requirements for exhaust emissions for vehicles which fail to meet various performance thresholds within IUVP (see 40 CFR 86.1846-01). As mentioned above, we expect that it will be common to get more than one leak test result over the course of each model year's mileage testing point for each evaporative/refueling/leak family as a result of the requirement to assess leaks with each exhaust IUVP test. However, the leak standard is basically pass/fail at 0.02 inches and it is difficult to establish a threshold criteria for a pass/fail standard such as has been done for exhaust emissions where there is a multiplier applied to the level of the individual exhaust emission standard.

Given the importance of the leak standard in assuring in-use evaporative emissions control, we are finalizing a set of criteria for assessing leak standard results from IUVP. These criteria can be summarized as follows for each low and high mileage test point for each model year tested:

  • lf 50 percent or more of all vehicles evaluated in an evaporative/refueling/leak emission family for any given model year pass the leak standard, testing is complete. This applies to cumulative testing for that family throughout the model year for that mileage group. This is consistent with the exhaust emission requirements for IUVP and EPA believes it is reasonable since vehicles are tested in the “as received” condition from consumers.
  • If only one representative of the evaporative/refueling/leak family is tested in a mileage group for that model year's vehicles and it passes the leak standard testing is complete. If that vehicle does not pass the leak standard a manufacturer may test an additional vehicle to achieve the 50 percent rate.
  • If an evaporative/refueling/leak emission family fails to achieve the 50 percent rate, it is presumed that the family will enter into In-Use Confirmatory Testing Program (IUCP).

Before IUCP begins, the manufacturer may ask for engineering analysis discussions with EPA to evaluate and understand the technical reasons for the testing outcomes and the implications for the broader fleet. Technical information for these discussions could include but will not be limited to detailed system design, calibration, and operating information, technical explanations as to why the individual vehicles tested failed the leak standard, and comparisons to other similar families from the same manufacturer. Relevant information from the manufacturer such as data or other information on owner complaints, technical service bulletins, service campaigns, special policy warranty programs, warranty repair data, state I/M data, and data available from other manufacturer specific programs or initiatives could help inform understanding of implications for the broader fleet. As part of this process a manufacturer could elect to provide evaporative emissions SHED test data on the individual vehicle(s) that did not pass the leak standard during IUVP. With an adequate technical basis, the outcome of this engineering analysis discussion could result in an EPA decision not to require IUCP testing.

We will operate within the basic structure of the IUCP program in the existing regulations. Prior to commencing IUCP testing the manufacturer, after consultation with EPA submits a written plan describing the details of the vehicle procurement, maintenance, and testing procedures. This plan could include inclusion of a hot soak plus diurnal SHED test to supplement leak test results. EPA must approve this plan before testing begins. As prescribed in the IUCP regulations for exhaust, if five vehicles are tested and all pass the leak standard then testing will be complete. If all five vehicles do not pass, then five more are tested. More vehicles can be tested at the manufacturer's discretion but all testing must be completed within the time period specified in the regulations. EPA and the manufacturer then enter into discussions regarding interpretation, technical understanding, and compliance/enforcement implications of the test results, if any.

iv. Optional Test Procedure Approach for IUVP/IUCP

With the implementation of the OBD regulation changes in Section IV.E below regarding evaporative system leak rate monitoring, EPA is finalizing an optional approach to a portion of the leak test procedure. This optional testing approach is included in the IUVP/IUCP testing program for the leak standard, but will not be used for certification testing for the leak standard. EPA can also use this procedure for conducting compliance assessments. Under this optional approach manufacturers will be able to rely upon the operation of their OBD evaporative system leak detection hardware and operating protocols in lieu of running the stand alone in-use leak test to check for the presence of a 0.02 inch leak in the fuel/evaporative system.

Quite simply, if a vehicle is brought in for IUVP or IUCP testing and a scan tool query of the onboard computer indicates that the vehicle has successfully completed a full OBD-based evaporative system leak monitoring check within the last 750 miles and no evaporative system leak problems for any diameter above 0.020 inches are indicated (no pending or confirmed diagnostic trouble code(s) P0440, P0442, P0446, P0455, P0456, or P0457), the vehicle would be deemed to have met and passed the leak standard test requirement. However, if the system has not successfully completed a full OBD-based evaporative system leak check within 750 miles with no problem indicated then the manufacturer will have the option to run its OBD-based evaporative system leak check in the laboratory after prescribed preconditioning. This OBD-based approach is sometimes used in auto manufacturer dealerships and repair facilities to diagnose and fix evaporative system leaks found by the OBD system. If the vehicle completes the full OBD-based leak test in the laboratory then the vehicle's pass/fail results for the 0.02 inch cumulative equivalent diameter orifice will be based on the OBD test result. This optional protocol can apply to every leak standard test after certification unless not approved by EPA for IUCP under 40 CFR 1846.01(i). Replicate tests will not be required or allowed but void tests could be repeated.

Furthermore, EPA will permit the manufacturer to run the stand alone EPA leak test in several situations. First, manufacturers can conduct the stand alone test to confirm that a problem identified by the OBD-based evaporative system monitoring leak check is a leak and not a problem with the OBD leak monitor itself. Second, a manufacturer can run the stand alone EPA leak test to confirm that the leak value identified by the OBD system is truly above the level of the leak standard. Third, it can be used for vehicles which have not successfully completed a full OBD-based evaporative system leak monitoring check within the last 750 miles. Fourth, it can be used to confirm that a DTC set within the last 750 miles actually indicates the presence of a leak(s) greater than the standard. However, if a manufacturer elects to use only OBD-based evaporative system leak based monitoring in its IUVP testing; these results will be the basis for decisions regarding IUCP. As required in the existing IUVP regulations, all test data whether OBD based or based on EPA's stand alone test procedure must be reported to EPA.

There may be some advantages to this option since it employs a pressure/vacuum approach manufacturers understand and creates positive/negative pressures manufacturers have accommodated within their fuel/evaporative system. One potential downside is that under current designs vehicle engines will have to be operating to create the pressure or vacuum and because the engine is operating this will require the OBD-based leak test to be stand alone after the preconditioning sequence is complete. This will be more challenging for natural vacuum leak detection systems unless extended driving is involved to create the fuel system heat needed for a natural vacuum event or this is done through a climate chamber or SHED based diurnal heat build.

Allowing for this approach raises at least two implementation questions. The first is related to the value of conducting the OBD-based test for a vehicle with a confirmed or pending leak DTC already set in the computer and/or an MIL indicated. In this case, EPA will permit the manufacturer to run the OBD-based leak test and/or the stand alone EPA leak test or concede that the vehicle will not pass the leak standard and count the result. Second is the question of gas caps. This is among the most common codes found in OBD records and is often related to operator error such as not tightening the gas cap properly. Codes of this nature have no value in this leak assessment, so a manufacturer will be permitted to correct the problem before testing and clear this OBD code before testing or run the stand alone EPA leak test.

E. Onboard Diagnostic System Requirements

1. Onboard Diagnostic (OBD) System Regulation Changes—Timing

EPA first adopted OBD requirements for 1994 and later model year LDVs and LDTs. While EPA has extended its requirements from LDVs and LDTs to larger and heavier vehicles, (396) EPA's last broad upgrade to its basic OBD regulation was in the 2005 timeframe. Since that time, CARB has adopted and the manufacturers have implemented a number of additional provisions to enhance the effectiveness of their OBD programs. These provisions include new requirements for OBD evaporative system leak detection as well as provisions to help insure that systems are built and operate as designed over their full useful life, give reliable results (find and signal only true deficiencies), and operate frequently during in use operation. It is permitted in existing EPA regulations and is common practice for the industry to certify their OBD systems with CARB and for EPA to accept CARB OBD certifications as satisfying EPA requirements. EPA is continuing that practice and we are updating our regulations to be consistent with the latest CARB regulations.

EPA proposed to adopt, with a few adjustments, the CARB regulatory requirements related to OBD II (see California Code of Regulations (CCR) 1968.2 dated May 18, 2010). We received comment from CARB that since our NPRM was issued, they were s completing an update of their OBD II regulations and that EPA should adopt these provisions in lieu of the May 18, 2010 provisions. (397) We have reviewed these updates and concur with the commenters, so we are adopting the provisions officially approved by CARBs Office of Administrative Law on July 31, 2013. We are also adding provisions and continuing the exceptions as discussed below. The changes we are adopting do not include any changes to requirements for engines used in vehicles over 14,000 lbs GVWR or to vehicles over 14,000 lbs GVWR, except for HDGVs optionally certified using chassis procedures. To be consistent with the manner in which the Tier 3 exhaust emission standards are being implemented for the heavy-duty vehicles between 8,501 and 14,000 lbs GVWR, the OBD requirements will be based on the Job 1 (first production) date for the vehicle/engine model. If the vehicle/engine model Job 1 date is before the fourth anniversary date of the signature of the Tier 3 rule the requirements will not be mandatory in that model year. If the Job 1 date is on or after the fourth anniversary of the signature date of the Tier 3 rule the OBD requirements will apply in that model year. The Tier 3 OBD requirements will apply to all 8,501-14,000 lb HDVs in the 2020 model year. To be consistent with the manner in which the Tier 3 exhaust emission standards.

We are taking this approach to OBD for three basic reasons. First, this is consistent with the goal of a national program and one vehicle technology for all 50 states. Second, compliance with the current CARB OBDII requirements is now demonstrated technology, compliance with these requirements is common within the industry today, and we expect that to continue in the future with the 2013 CARB changes. Thus, the added burden is minimal since essentially all manufacturers certify their CARB OBD systems nationwide with EPA. Third, the latest OBD systems run frequently on in-use vehicles to identify potential exhaust and evaporative system performance problems, so adopting these provisions will create the opportunity for OBD to serve a more prominent role in ensuring the Tier 3 emission standards are met in-use.

Alignment with the existing CARB OBD II requirements will be required by the 2017 MY, except as discussed below. Manufacturers requested a phase-in compliance approach in lieu of a fixed compliance date, but no specific justification was provided by the commenters and EPA could not establish a need for this accommodation since the most recent changes to CARB OBDII regulations (2013) did not meaningfully affect provisions regarding vehicles/engines under 14,000 lbs GVWR which have been in place since 2006. LDVs, LDTs, MDPVs and vehicles under 14,000 lbs GVWR already comply with CARB OBDII requirements and use the CARB certification as the basis for EPA certification.

There is an important link between OBD provisions related to evaporative emission control system leak monitoring and the leak standard. They each provide an important incentive to design fuel/evaporative systems with fewer propensities to develop leaks in use but each addresses the issue from a different perspective. The distinction is that the leak standard prohibits leaks of greater than 0.02 inches cumulative equivalent diameter, while the OBD evaporative system leak monitoring provision requires that the OBD system find leaks larger than 0.020 inches cumulative equivalent orifice diameter and notify the owner, but with no explicit requirement to repair the problem. Thus adopting a 0.020 inch cumulative equivalent orifice diameter aligns these two programs and, as discussed above, facilitates the use of OBD evaporative system leak monitoring hardware/strategy as an optional leak detection test procedure for in-use testing.

With regard to OBD evaporative system leak detection, EPA received comment that we should permit a phase-in for compliance with the 0.020 inch evaporative system leak monitoring requirement. Even though the 0.020 inch leak monitoring requirement has been in place since the 2004MY for CARB OBDII, and essentially manufacturers have met it for years, the existing EPA regulation actually only requires monitoring at the 0.040 inch threshold level. After considering the comments received, EPA is permitting a limited and minimal phase-in for the 0.020 inch leak detection criterion for the OBD evaporative system monitoring requirement. We are permitting this phase-in, because a few vehicle models still only meet the 0.040 inch monitoring threshold in their Federal configuration and complying with the 0.020 inch CARB OBD II requirement entails validating performance in high altitude and cold weather regimes not seen in California. Thus, the 0.020 inch requirement would be new for those few models currently certified only to the EPA evaporative leak monitoring requirement. We are, therefore, implementing the following phase-in provision for the 0.020 inch leak detection criterion for the OBD evaporative system monitoring requirement. First, if a vehicle model meets the 0.020 inch requirement in the 2016 model year it is not eligible for the phase-in provision. No backsliding is permitted. Second, for manufacturers with models not meeting the CARB OBDII evaporative system leak monitoring requirement in the 2016 MY (see 13 CCR 1968.2(e)(4)), they will be permitted to delay product-wide compliance with the 0.020 inch leak provision of the evaporative system monitoring requirements until the 2018 model year by engaging in a voluntary early phase-in. This phase-in would begin in the 2016 model year and conclude in the 2018 model year at a 100 percent implementation rate. For example, a manufacturer could delay attaining 100 percent compliance with the OBD evaporative system leak monitoring requirement until the 2018 model year by complying in the 2016 model year using a percentage which is at least as large as the delay for the 2017 model year (e.g., 40% in 2016 MY, 60% in 2017MY, and 100% in 2018MY).

2. Revisions to EPA OBD Regulatory Requirements

As discussed above, we are updating our OBD regulations to be consistent with current California OBD II requirements. We are incorporating by reference section 1968.2 of the California Code of Regulations as adopted July 31, 2013 (13 CCR 1968.2). This includes paragraphs (c) through (j) in their entirety. These paragraphs are entitled: (c) Definitions, (d) General Requirements, (e) Monitoring Requirements for Gasoline/Spark Ignited Engines, (f) Monitoring Requirements for Diesel/Compression Ignition Engines, (g) Standardization Requirements, (h) Monitoring System Demonstration Requirements for Certification, (i) Certification Documentation, (j) Production Vehicle Evaluation Testing. The substance of many of these provisions is already contained in existing EPA OBD requirements for LDVs, LDTs, MDPVs, and complete HDGVs less than 14,000 lbs GVWR. (398 399) EPA will continue to accept certifications with CARB OBD requirements as satisfying EPA OBD requirements.

The most noteworthy changes we are finalizing are summarized below. The CCR below is the California Code of Regulations cite for each pertinent provision.

  • EPA is adding a 0.020 inch leak detection monitoring threshold upstream of the purge valve for all 4 vehicle categories LDV, LDT, MDPV, and complete HDGVs up to 14,000 lbs GVWR except for those with fuel tanks larger than 25 gallons capacity (see 13 CCR 1968.2(e)). OBD leak monitoring systems will have to identify, store, and if required signal any leak(s) equal to or greater than 0.020 inches cumulative equivalent diameter. This will thus include diagnostic trouble codes (DTC) P0440, P0442, P0446, P0455, P0456, and P0457.
  • EPA is incorporating by reference the full array of rate based monitoring requirements (see 13 CCR1968.2 (d)(3)-(6)). Meeting the rate based monitoring requirements will help to insure that, even with enable criteria, the exhaust and evaporative system monitors run frequently enough that on average a problem would be identified and signaled to the owner in operation within two weeks. This will help to improve the fraction of time monitors are ready to find a potential problem.
  • EPA is incorporating by reference provisions regarding monitoring system demonstration requirements for certification. We are incorporating by reference CARB provisions in this area and accepting submissions to CARB for purposes of compliance demonstration (see 13 CCR 1968.2(h)). Adopting current CARB monitoring system demonstration requirements assures that monitoring systems operate as designed when installed on certification vehicles.
  • EPA is incorporating by reference the CARB production vehicle evaluation data program. This program requires manufacturers to demonstrate that the OBD system functions as designed and certified when installed on production vehicles. (See 13 CCR 1968.2(j)).

In addition, we are adding two new requirements, and retaining three minor exceptions. Each of these actions is described separately below.

  • We are adding the requirement that before certification a manufacturer must demonstrate the ability of its OBD leak monitoring system to detect and report a 0.020 inch leak in the fuel/evaporative system. Current CARB protocols within 13 CCR 1968.2(h)(3) do not require this demonstration as part of certification. This requirement helps to ensure the OBD system's capability to function as designed and the OBD-based evaporative system leak monitoring hardware to be used as an optional test procedure for IUVP testing for the leak standard. This requirement being added for the same vehicles that are subject to monitoring system demonstration requirements for certification under CARB OBD regulations under 1968.2(h)(3). (400) EPA test procedures are contained in 40CFR 86.1806-17(b). In the spirit of aligning CARB and EPA OBD provisions, if CARB ultimately adopts this demonstration requirement and CARB's test procedure provisions fulfill the purpose of the EPA requirement, EPA will strongly consider proposing to adopt the CARB test procedures in lieu of those in 40 CFR 86.1806-17(b).

This requirement applies to any vehicle test group certified to the OBD 0.020 inch evaporative system leak monitoring requirement. Since the regulation requires only a relative few test groups each model year per manufacturer, we will permit the manufacturers either to meet the requirement for the remainder of its test groups on production vehicles of a previous model year which used the identical monitoring hardware and strategies or to certify by attestation that each of their remaining test groups meets the requirement based on development, calibration, and other information. If a manufacturer chooses to certify by attestation for some test groups for a given model year, the regulations are structured such that over several model years a manufacturer would evaluate through testing all test groups as new groups are selected in subsequent model years.

  • For the OBD evaporative system leak monitoring requirement, EPA is establishing a requirement for a scan tool readable function (a new InfoType $14 in Service $09 of SAE J1979DA) which can be used to obtain the distance traveled since the OBD leak monitoring diagnostic was last completed successfully, i.e., the system passed or failed (identified any leak above 0.020 inches) during that monitoring event (unless it is otherwise already required in other OBD system modes). The purpose of this requirement is to facilitate implementation of the leak standard within IUVP, by permitting the use of OBD evaporative system monitoring results as a tool to make pass/fail determinations during IUVP. As discussed in section IV.D above, if a vehicle successfully completed an evaporative system leak monitoring within the most recent 750 miles then the manufacturer could use this result for its IUVP requirement for the leak standard. EPA asked for comment on how best to implement this requirement within the OBD system, in what model year(s) it should be required and to which vehicle classes it should apply.

Manufacturers supported this requirement, and suggested a lower cost approach which we are adopting in the final rule. Rather than requiring that the distance and monitoring results be stored in NVRAM to avoid false results based on a user induced code clear or battery disconnect, the manufacturers suggested that the “distance since evap monitoring decision” InfoType be reset to the maximum value ($FFFF/65,535km) when codes are cleared or after a reprogramming event (e.g., battery disconnect). The InfoType would be reset to zero km when an evaporative monitoring pass/fail decision is later made, allowing the mileage to be read directly at IUVP. In the usual situation where no user induced code clear or reprogramming event (e.g., battery disconnect) occurred, the mileage since the last decision could be read directly. In either circumstance, the presence of an evaporative system leak related DTC (P0440, P0442, P0446, P0455, P0456, and P0457 or manufacturer specific equivalent DTC) will indicate a failure and the lack of such a DTC will indicate a pass. The mileage and the pass/fail results will then be taken together for purposes of the 750 mile option in the IUVP assessment for the leak standard. (401)

This requirement applies to all vehicle categories subject to the leak test including LDVs, LDTs, MDPVs, and complete HDGVs less than 14,000 lbs GVWR. Manufacturers commented that this requirement should apply only to vehicles/test groups meeting the leak standard. Since the leak standard phases-in between 2018 and 2022 model years (2017 for manufacturers using the 20/20 evaporative emission option), a manufacturer may phase-in compliance with this requirement as well.

  • The minor exceptions which are contained in EPA's existing OBD regulations are to be continued. Compliance with 13 CCR 1968.2(d)(1.4), pertaining to tampering protection is not required. Also, the deficiency provisions of 13 CCR 1968.2(k) are not being adopted. In addition, demonstration of compliance with 13 CCR 1968.2(e)(15.2.1)(C), to the extent it applies to the verification of proper alignment between the camshaft and crankshaft, will apply only to vehicles equipped with variable valve timing. For all model years, the deficiency provisions of paragraph (i) of the existing EPA regulations apply only to alternative fuel vehicle/engine manufacturers selecting this paragraph for demonstrating compliance.

These changes, taken together will improve the performance, reliability, general utility, and effectiveness of OBD systems for Tier 3 exhaust and evaporative emission controls. Furthermore, these changes create the opportunity for OBD evaporative system leak monitoring systems to serve a more prominent role in ensuring compliance with the leak standard. EPA believes that they can be implemented for minimal cost since most manufacturers are meeting them today and will have to for LEV III vehicles. The provisions we are incorporating by reference give manufacturers the flexibility to seek a revision to the emission threshold for a malfunction on any diagnostic required if the most reliable monitoring method developed requires a higher threshold to prevent significant errors of commission in detecting a malfunction. (402) Any decision on a potential exception would be preceded by a consultation between EPA and CARB.

As discussed below, the OBD requirements will apply to small entities in the 2022 model year, if they choose to take advantage of one of the revised implementation schedules for small volume manufacturers and small businesses. However, as is the case for larger manufacturers, no backsliding is permitted meaning that if they voluntarily meet the OBD requirements on their Federal configurations in the 2016 model year as a result of compliance with CARB regulations they must continue to meet the requirements on the Federal configurations in the 2017 and later model years. Small alternative fuel converters will still be able to meet the OBD requirements using the provisions of 40 CFR 85, subpart F. Finally, it should be noted that as CARB updates its OBD regulations in the future EPA will consider these changes and propose to adopt them or incorporate them by reference, if appropriate.

3. Provisions for Emergency Vehicles

It is common for emergency vehicles such as law enforcement, medical response, and fire protection vehicles operated by government entities to be derived from similar publicly available vehicle configurations. However, these vehicles often have chassis configurations, auxiliary equipment packages, and performance requirements different from the standard publicly available configurations. These emergency response vehicles typically meet the various EPA emission standards based on the engineering calibrations and emission control hardware used in the publicly available configuration. OBD requirements also apply to these vehicles and occasionally their unique design and/or operating characteristics may prevent them from meeting one or more of the various OBD requirements.

In comments on the NPRM, one manufacturer raised a concern that EPA's proposed adoption of the current CARB OBDII requirements for the 2017 model year would create a compliance problem for two of their law enforcement vehicle configurations. These two vehicle configurations cannot meet one element of the current CARB OBDII requirements (CCR 1968.2 (e)(6.2.1)(C)) without compromising the performance expected by law enforcement personnel. (403) To address this issue CARB provided these vehicles an exemption from this provision, by permitting it to meet Federal requirements as permitted by the California Vehicle Code. This solved the problem because the CARB OBD II provision of interest did not exist within the Federal OBD requirements at that time.

This raises both a near term and a broader policy issue related to emergency vehicles. First, we are incorporating a definition for emergency vehicle that is specific to the Tier OBD requirements. (404) Second, with regard to the two law enforcement vehicle configurations identified by the manufacturer, EPA has reviewed the manufacturer's technical information and agrees with CARB's previous assessment. (405) Thus, EPA will grant the manufacturer a three model year exemption from the requirement as requested by the manufacturer (MY2017-2019 inclusive). Specifically, we are delaying the need to comply with the requirements of CCR 1968.2 (e)(6.2.1)(C)—incorporated by reference by EPA—until the 2020MY for any emergency vehicle which does not meet the requirement in the 2016 model year. This specifically applies to the two test groups identified by the commenter. Second, in a broader context, there is a need to address the potential future need for a deficiency or an exemption for emergency response vehicles. If CARB grants a deficiency for emergency response vehicles under CCR 1968.2(k) we would expect this to be done in consultation with EPA. Furthermore, we are incorporating provisions to address a potential situation where an emergency vehicle needs a deficiency (a temporary or permanent allowance for manufacturers to be non-compliant with a specific requirement of the OBD regulations as long as certain requirements are met) or exemption which is not addressed by CARB under CCR 1968.2(k). EPA is adopting a provision which authorizes us to address these circumstances based on an application from the manufacturer. Under this provision, EPA may approve a request for a deficiency or in extreme circumstances a temporary or potentially permanent exemption from a given OBD requirement. In considering decisions to approve/disapprove this request, EPA will consider the provisions of CCR 1968.2 (k)(1) plus engineering information and vehicle emission and performance data provided by the manufacturer which demonstrates significant vehicle engineering or system performance issues (e.g., vehicle speed, acceleration, handling, safety, fuel economy, cost) related to complying with the OBD requirements.

4. Future Considerations

EPA and CARB coordinate closely on OBD II requirements. When changes to the requirements occur, CARB provisions often precede those from EPA. Since LEV III begins before Tier 3, EPA expects that CARB will revise any OBD II requirements related to the LEV III before EPA would do so for Tier 3. EPA expects to work with CARB on any potential changes to OBD II requirements related to LEV III and to consider proposing such changes in a future action since we expect great commonality between Tier 3 and LEV III exhaust and evaporative emission control systems. Two presentations related to CARB's initial thinking for LEV III related OBDII revisions are available in the docket. (406) In the interim, for any Tier 3 exhaust emission bin which does not have a corresponding bin value in the Tier 2 program, the threshold for the exhaust emission malfunction criteria is that of the next higher bin in the Tier 2 regulation as prescribed for the latest model year in CCR 1968.2(e)(1)-(3).

In the NPRM, EPA discussed the basics of evaporative emission control technology and laid out concerns regarding the loss of evaporative and refueling emission control which occurs if a canister is not purged. This can potentially occur if the purge hardware fails or if the flow of purge air through the canister is impeded by foreign matter collecting at the inlet port or on the carbon itself, canister poisoning due to fuel or water intrusion, or activated carbon breakdown from phenomena such as road vibration. Failure of purge hardware is already covered by OBD and a recent study indicates that this is a relatively rare evaporative system problem. (407) Failure of the activated carbon to purge due to problems such as those mentioned above are not covered by OBD. EPA is undertaking a study to better characterize the causes and frequency of such potential problems, and may propose in a future rulemaking an OBD-based monitoring requirement related to activated carbon/canister capture should the study indicate a significant frequency of loss of canister efficiency in-use and loss in emissions control relative to other evaporative system failure modes.

In the NPRM we also asked for comment on several other issues related to the role of OBD in future technology fuel/evaporative control systems. This included pursuing a monitoring threshold less than the 0.020 inches cumulative diameter that we are finalizing in this rule for non- pressurized and pressurized fuel systems. We asked about the feasibility and cost of requiring the OBD leak detection monitoring system to detect and signal the presence of a smaller diameter orifice, such as 0.010 inch upstream of the purge valve for a pressurized system with a designed in-use operating pressure threshold in excess of 0.36 psi (10 inches water). Also, for the pressurized system, we asked for comment on a potential provision to require that the fuel tank vent to the canister at key off if the OBD system identifies a leak. In their comments manufacturers indicated concerns about the need for such provisions or their value in reducing emissions relative to current requirements. EPA believes both of these provisions merit further investigation, but at the present time we lack the data to assess the feasibility and emission reduction benefits associated with each approach and so are not taking action on them.

Finally, in the NPRM we sought input on whether the operation of a vacuum pump or similar device used to assist or supplement vehicle engine vacuum purge or any device otherwise used to enhance or control purge flows, rates, or schedules should be required to be monitored as part of OBD. In their comments the manufacturers indicated their view that this would be covered by current OBD provisions, and we are not taking further actions.

F. Emissions Test Fuel

In-use gasoline has changed considerably since EPA last revised specifications for the gasoline used in emissions testing of light- and heavy-duty vehicles. Sulfur and benzene levels have been reduced and, perhaps most importantly, gasoline containing 10 percent ethanol by volume (E10) has replaced non-oxygenated gasoline (E0) across the country. This trend has had second-order effects on other gasoline properties. In-use fuel is projected to continue to change as refiners adjust their gasoline production to reflect the renewable fuel volumes required under the RFS2 program, as well as further sulfur reduction under the Tier 3 rule. (408) As a result, we are updating federal emission test fuel specifications to better match in-use fuel. The revised test fuel specifications apply for exhaust emissions testing, fuel economy/greenhouse gas testing, and emissions testing for non-exhaust emissions (with some exceptions discussed elsewhere in this preamble, e.g., for refueling tests in flex-fuel vehicles). The revised gasoline specifications, found at § 1065.710 and discussed below, apply to emissions testing of light-duty cars and trucks as well as heavy-duty gasoline vehicles certified on the chassis test, where the vehicles are certified to the Tier 3 standards. (409)

1. Gasoline Emissions Test Fuel: Ethanol Content and Volatility
a. Emission Test Fuel Ethanol Content

In the NPRM, EPA proposed that the emissions test gasoline be changed from E0 to E15 as a forward-looking position based on indications following the 2011 E15 waiver decision that the market would move in that direction. (410) Since the time when we developed the proposal, several relevant factors have led EPA to reconsider that position, including limited proliferation on a national scale of stations offering E15 and the complexities E15 test fuel would introduce for long-term harmonization of the Tier 3 vehicle emission regulations with California's LEVIII program (which uses E10 for emissions testing).

We received comments supporting use of E10 as emissions test fuel from the automotive and oil industries, as well as states and NGOs citing the fact that this was most representative of current market conditions. Other stakeholders involved in fuel marketing and distribution cited significant infrastructure cost and liability concerns in making E15 widely available at existing stations. Ethanol industry commenters generally supported E15 certification fuel as proposed, but provided no specific timeline on which this blend level would become representative of in-use fuel. The most recent surveys of the market show that E10 now comprises nearly 100% of in-use gasoline, with very small amounts of E0 and E15 being sold in limited areas where there is specific interest. (411) Based on this information and considering comments, EPA is finalizing E10 as the ethanol blend level in emissions test gasoline for Tier 3 light-duty and heavy-duty gasoline vehicles. We will continue to monitor the in-use gasoline supply and based on such review may initiate rulemaking action to revise the specifications for emissions test fuel to include a higher ethanol blend level.

As discussed above in Sections IV.A.7.d (tailpipe emission testing) and IV.C.5.b (evaporative emission testing), we are requiring all light-duty and chassis-certified heavy-duty gasoline vehicles to be certified to Tier 3 standards on federal E10 test fuel. As described in those sections, EPA will accept emission certification test results performed according to CARB's LEVIII procedures including CARB's E10 test fuel. Confirmatory and in-use exhaust or evaporative testing of vehicles certified on CARB's E10 test fuel will be performed using that same test fuel through MY 2019. After MY 2019, EPA will continue the practice of accepting emission data at certification on the LEVIII test fuel; however confirmatory and in-use testing may be performed using Tier 3 E10 test fuel at the discretion of the Agency.

b. Certification Fuel Volatility (RVP) Specification

In deciding to finalize E10 as the emissions test fuel it is appropriate to consider whether a change in the volatility of the test fuel is warranted, typically expressed as in pounds per square inch (psi) Reid Vapor Pressure (RVP) or dry vapor pressure equivalent (DVPE). The Clean Air Act (Section 211(h)(1)) sets a national limit on summertime RVP in northern conventional gasoline areas of 9.0 psi to control ozone pollution. However, Congress included a waiver allowance (Section 211(h)(4)) granting an additional 1 psi RVP to 10% ethanol blends, meaning that E10 could have an RVP up to 10 psi in these conventional gasoline areas unless specifically prohibited by state or local rules. Under Section 211(h)(4), E15 is not covered by the waiver and thus is restricted to 9 psi nationwide.

The automakers submitted comments that recommended leaving the RVP of emissions test fuel at 9 psi on the basis that raising the specification to 10 psi would increase the stringency of the proposed evaporative emission standards significantly. We agree that the resulting increased vapor generation rates during the refueling test would increase emissions (by about 10 percent and during the hot soak, diurnal, canister bleed, and running loss tests by as much as 25 percent in total). While the likely increase in canister volume in response to higher certification fuel RVP would not be difficult for automakers to accommodate in most cases, there are additional uncertainties regarding cost and feasibility of strategies for removing the larger vapor loads from the canister during vehicle operation (vapor “purging”). Some vehicles have adequate engine vacuum available to accomplish the increased vapor purge, while others may require new or innovative approaches to increase purge volume or efficiency (as discussed in the evaporative emissions technology discussion in Section IV.C.3).

Several other commenters, such as NGOs and environmental groups, supported setting certification gasoline RVP to 10 psi to be representative of the worst-case volatility vehicles may see in the market, making the test procedure more stringent than in the proposed program and further reducing evaporative emissions.

Raising the certification test fuel RVP to 10 psi would also impact the equivalency of CARB and EPA hot soak plus diurnal evaporative emission test procedures. (California requires the use of 7 psi RVP test fuel, which, in conjunction with higher test temperatures, produces equivalent results to the federal test procedures using 9 psi fuel.) If we were to adopt 10 psi test fuel, we would likely need to develop and adopt new test procedure adjustments in order to maintain the equivalency of CARB and EPA evaporative procedures (and allow reciprocal acceptance of test data generated under either agency's program).

In addition, the 1 psi RVP waiver for E10 does not apply to gasoline with higher ethanol levels; for example, under current regulations E15 is subject to an RVP limit of 9 psi. If EPA had adopted 10 psi test fuel in this rule and if gasoline with higher ethanol levels than E10 were to become commonly used nationwide, maintaining alignment with in-use fuel could necessitate a change in emissions test fuel back to 9 psi.

A review of 2011 gasoline batch data submitted to EPA shows that just under half of summertime gasoline was conventional gasoline at 10 psi RVP. An additional third was RFG at approximately 7 psi RVP, with the remainder having intermediate RVPs under local volatility control programs. A volume-weighted average of these data is approximately 8.7 psi RVP. Thus, an emissions test gasoline volatility at 9 psi aligns well with the average nationwide in-use RVP today. In addition, virtually all of the areas that have elevated summertime ozone levels where excess evaporative VOC emissions would be of greatest concern already control in-use gasoline RVP to levels less than 9 psi. Furthermore, under section 211(a)(5), governors can request that the 1 psi waiver for E10 not apply in their state if it causes an emissions increase that contributes to air pollution. Any state exercising this authority would have in-use E10 RVP levels limited to 9 psi.

After considering these technical and policy issues in the context of the information available and comments received, we conclude that the most appropriate approach is to set an RVP of 9 psi for Tier 3 emissions test fuel.

c. Durability Test Fuel

EPA's motor vehicle emissions standards typically require a level of performance over a specified test procedure, with emissions measured while the engine or the vehicle is operated using the specified test fuel and operated in a specified manner. The test fuel specifications typically apply for all emissions testing used to determine compliance with the standard, including emissions testing to obtain a certificate of conformity, as well as compliance testing for newly produced or in-use engines or vehicles. While this test fuel is sometimes referred to as “certification fuel,” the test fuel specifications are not limited to certification related emissions testing, but also apply to compliance related emissions testing after the certificate of conformity has been issued. The certification process also typically involves a process to ensure that the emissions controls system is durable over the regulatory useful life of the vehicle or engine. This can involve long-term or accelerated aging of a vehicle or engine prior to emissions testing. The fuel used for such aging is commonly referred to as service accumulation or durability fuel, and in many cases is specified as commercial gasoline that will be generally available through retail outlets (§ 86.113-04(a)(3)), or in some cases may be specified as gasoline which contains ethanol in, at least, the highest concentration permissible in gasoline under federal law and that is commercially available in any state in the United States, such as for durability aging of evaporative emissions system (§ 86.1824-08(f)). EPA is not changing the specifications for fuel used during durability related aging that is part of the certification process. The regulatory changes in this final rule only apply to the test fuel used during emissions testing, both for purposes of certification and for later compliance related testing.

We are not changing the exhaust or evaporative durability fuel requirements outlined in the provisions of § 86.113-04(a)(3), except to remove the minimum sulfur content (15 ppm) specified at § 86.113-04(a)(3)(i). Those provisions require that “[u]nless otherwise approved by the Administrator, unleaded gasoline representative of commercial gasoline that will be generally available through retail outlets must be used in service accumulation.” We expect that manufacturers will use service accumulation fuels that are generally representative of the national average in-use fuels (or worst case for durability) during the model year which is being certified, including, for example, the ethanol content (for exhaust emissions), sulfur level, and fuel additive package. For exhaust emission bench aging durability programs as allowed under the provisions of § 86.1823-08(d) and (e), the bench aging program should be designed using good engineering judgment to account for the effects of in-use fuels on exhaust emissions, including the effects of future in-use fuels on catalytic converters, oxygen sensors, fuel injectors, and other emission-related components.

For evaporative emissions, durability fuel requirements are the same as for exhaust emissions (as outlined above), plus an additional requirement in the provisions of § 86.1824-08(f), that the service accumulation fuel “contains ethanol in, at least, the highest concentration permissible in gasoline under federal law and that is commercially available in any state in the United States. Unless otherwise approved by the Administrator, the manufacturer must determine the appropriate ethanol concentration by selecting the highest legal concentration commercially available during the calendar year before the one in which the manufacturer begins its mileage accumulation.” Thus, we expect that E15 service accumulation fuel will be used for whole vehicle evaporative durability programs. Similarly, evaporative bench aging durability programs allowed under the provisions of § 86.1824-08(d) and (e), should be designed using good engineering judgment to account for the durability effects of in-use fuels on evaporative emissions, bleed emissions, and leakage emissions.

2. Other Gasoline Emissions Test Fuel Specifications

Where possible, we are changing test fuel specifications to be consistent with CARB's LEV III gasoline test fuel specifications. (412) In addition to the ethanol and volatility specifications discussed above, below is an overview of some of the key changes. Table IV-26 provides a summary of the new test fuel properties. For more information on how we arrived at the test fuel property ranges and ASTM test methods, refer to Chapter 3 of the RIA.

  • Octane—lowering gasoline octane to around 87 (R+M)/2 to be representative of in-use fuel, i.e., regular-grade E10 gasoline. Manufacturers can continue to use high-octane gasoline for testing of premium-required  (413) vehicles and engines as well as for testing unrelated to exhaust emissions. Historically, the high octane rating of test fuel has not had any real emissions implications. However, as manufacturers begin introducing new advanced vehicle technologies (e.g., turbocharged/downsized), this may no longer be the case. For those vehicles where operation on high-octane gasoline is required by the manufacturer, we are allowing the manufacturer to test on a fuel with a minimum octane rating of 91 (R+M)/2 (in lieu of the 87 (R+M)/2 specified for general test fuel). According to the regulations found at § 1065.710(d), vehicles or engines are considered to require premium fuel if they are designed specifically for operation on high-octane fuel and the manufacturer requires the use of premium gasoline as part of their warranty as indicated in the owner's manual. Cases where premium gasoline is not required but is recommended to improve performance would not qualify as a vehicle or engine that requires the use of premium fuel. For qualifying vehicles and engines, all emission tests must use the specified high-octane fuel. For vehicles and engines certified on high-octane gasoline, all EPA confirmatory and in-use testing would also be conducted on high-octane gasoline. All other test fuel specifications are the same as those described in Table IV-26.
  • Distillation Temperatures—adjusting gasoline distillation temperatures to better reflect in-use E10 gasoline. This includes minor T10, T90 and FBP adjustments based on AAM fuel surveys and refinery batch data. These data show that T50 varies widely in in-use fuel, from around 150 °F to 220 °F. Adopting a wide specification range for test fuel may have undesirable effects on consistency of results between facilities and over time. Therefore, we have chosen a range of 190-210 °F to maintain some overlap with CARB's specification of 205-215 °F but extending somewhat lower to better capture federal in-use fuel. For more information on how we arrived at the distillation temperatures in Table IV-26, refer to Chapter 3 of the RIA.
  • Sulfur—lowering the sulfur content of test fuel to 8-11 ppm to be consistent with our new Tier 3 gasoline sulfur standards. The 10 ppm annual average sulfur standard for in-use gasoline standard is expected to result in two-thirds less sulfur nationwide so it is appropriate to lower the gasoline test fuel specification in concert.
  • Benzene—setting a benzene test fuel specification of 0.5-0.7 volume percent to represent in-use fuel under the MSAT2 regulations. (414) The MSAT2 standards, which took effect January 1, 2011, limit the gasoline pool to 0.62 volume percent benzene on average.
  • Total Aromatics—lowering the range of aromatics content in the test fuel to better match today's in-use E10 gasoline, and narrowing the range to limit variability of results. Data from recent gasoline batch data as well as AAM surveys support a specification of 22-26 volume percent. (415)
  • Distribution of Aromatics—in addition to total aromatics and benzene, the updated test fuel requirements place boundaries on the distribution of aromatics by carbon number (i.e., prescribed volume percent ranges for each of C7, C8, C9, and C10+ hydrocarbons). There is evidence that the heaviest aromatics in gasoline contribute disproportionately to PM emissions, so compliance with emission standards should be demonstrated on fuel with a composition representative of in-use gasoline. For more information on the aromatics specifications, refer to Chapter 3 of the RIA.
  • Olefins—adjusting the olefins specification to a range of 4-10 volume percent to better match in-use E10 gasoline.
  • Other Specifications—adding distillation residue, total content of oxygenates other than ethanol, copper corrosion, solvent-washed gum, and oxidation stability specifications to better control other performance properties of test fuel. These specifications are consistent with ASTM's D4814 gasoline specifications and CARB's LEV III test fuel requirements.
  • Updates to Gasoline Test Methods—updating some of the gasoline test methods previously specified in § 86.113 with more appropriate, easier to use, or more precise test methods for ethanol-blended gasoline. Key changes include replacement of ASTM D323 with ASTM D5191 for measuring vapor pressure; replacement of ASTM D1319 with ASTM D5769 for measuring aromatics and benzene; and replacement of ASTM D1266 with three alternative ASTM test methods (D2622, D5453 or D7039) for measuring sulfur.
Table IV-26—Gasoline Emissions Test Fuel Properties
PropertyUnitSpecificationGeneral testingLow-temperaturetestingHigh altitude testingReference procedure a
Antiknock Index (R+M)/287.0—88.4 b 87.0 MinimumASTM D2699 and D2700. 
Sensitivity (R-M)7.5 MinimumASTM D2699 and D2700.  
Dry Vapor Pressure Equivalent (DVPE) c, d kPa (psi)60.0-63.4(8.7-9.2)77.2-81.4(11.2-11.8)52.4-55.2(7.6-8.0)ASTM D5191.
Distillation e      
10% evaporated°C (°F)49-60(120-140)43-54(110-130)49-60(120-140)ASTM D86.
50% evaporated°C (°F)88-99 (190-210)   
90% evaporated°C (°F)157-168 (315-335)   
Evaporated final boiling point°C (°F)193-216 (380-420)   
Residuemilliliter2.0 Maximum   
Total Aromatic Hydrocarbonsvolume %21.0-25.0ASTM D5769.  
C6 Aromatics (benzene)volume %0.5-0.7   
C7 Aromatics (toluene)volume %5.2-6.4   
C8 Aromaticsvolume %5.2-6.4   
C9 Aromaticsvolume %5.2-6.4   
C10+ Aromaticsvolume %4.4-5.6   
Olefins 5 mass %4.0-10.0ASTM D6550.  
Ethanol blendedvolume %9.6-10.0See § 1065.710(b)(3).  
Ethanol confirmatory f volume %9.4-10.2ASTM D4815 or D5599.  
Total Content of Oxygenates Other than Ethanol f volume %0.1 MaximumASTM D4815 or D5599.  
Sulfurmg/kg8.0-11.0ASTM D2622, D5453 or D7039.  
Leadg/liter0.0026 MaximumASTM D3237.  
Phosphorusg/liter0.0013 MaximumASTM D3231.  
Copper CorrosionNo. 1 MaximumASTM D130.  
Solvent-Washed Gum Contentmg/100 milliliter3.0 MaximumASTM D381.  
Oxidation Stabilityminute1000 MinimumASTM D525.  

As mentioned earlier, we will continue to allow manufacturers to test vehicles on premium-grade gasoline should the vehicles require it. In addition, since we cannot predict all future changes in gasoline vehicle technologies and in-use fuels, we will allow vehicle manufacturers to specify an alternative test fuel under certain situations. Under this provision, if manufacturers were to design vehicles that required operation on a higher octane, higher ethanol content gasoline (e.g., dedicated E30 vehicles or FFVs optimized to run on E30 or higher ethanol blends), under 40 CFR 1065.701(c), they can petition the Administrator for approval of a higher octane, higher ethanol content test fuel if they can demonstrate that such a fuel would be used by the operator and would be readily available nationwide, vehicles would not operate appropriately on other available fuels, and such a fuel would result in equivalent emissions performance. For vehicles certified on high-octane, high-ethanol gasoline, all EPA confirmatory and in-use testing would also be conducted on high-octane, high-ethanol gasoline. This could help manufacturers who wish to raise compression ratios to improve vehicle efficiency as a step toward complying with the 2017 and later light-duty greenhouse gas and CAFE standards. This in turn could help provide a market incentive to increase ethanol use beyond E10 and enhance the environmental performance of ethanol as a transportation fuel by using it to enable more fuel efficient engines.

We received comments in general support of allowing certification on higher octane fuels if the vehicles require it, although some commenters believe that the criteria EPA is specifying for such an allowance are too severe. We have considered these comments, and as discussed in the Summary and Analysis of Comments document, we continue to believe that our approach is appropriate, and we are finalizing these provisions as proposed.

3. Flexible Fuel Vehicle Exhaust Emissions Test Fuel

We are also finalizing specifications for the fuel used in flexible fuel vehicles (FFV) exhaust emissions testing including certification testing. EPA is establishing specifications for FFV test fuel to resolve confusion and inconsistency among FFV manufacturers in carrying out their certification and other testing requirements and to ensure that FFV emissions are appropriately controlled over the range of in-use fuels. The FFV exhaust emissions test fuel specifications will phase in on the same schedule as the E10 standard gasoline test fuel specifications for light- and heavy-duty gasoline vehicles (described in Section IV.F.4). These FFV exhaust emissions test fuel specifications may be used voluntarily prior to when they are required to be used. The base fuel stock used to formulate FFV exhaust emissions test fuel must comply with the specifications finalized today for the standard E10 emissions test fuel as described in preamble Sections IV.F.1 and 2. This practice avoids the need to specify the ranges for a number of fuel parameters as we have done for gasoline test fuel in Table IV-26 and helps to minimize the number of test fuels that a vehicle manufacturer must store. Denatured fuel ethanol (DFE) that meets the specifications discussed in preamble Section V.G. must be blended into this base fuel stock to attain an ethanol content of 80 to 83 volume percent in the finished test fuel. Commercial grade normal butane can be added as a volatility trimmer to meet a 6.0 to 6.5 psi RVP specification for the finished test fuel. (416)

As an alternative to the use of DFE to manufacture FFV test fuel, neat (undenatured) fuel grade ethanol can be used. As an alternative to using a finished E10 standard gasoline test fuel in the manufacture of FFV test fuel, the gasoline blendstock used by the fuel provider to produce a compliant E10 test fuel can also be used to manufacture the FFV test fuel. This would allow ethanol to be blended only once to produce FFV test fuel. In such cases, a sample of the subject gasoline blendstock must be tested after the addition of ethanol to produce a finished standard E10 gasoline test fuel to demonstrate that the blend meets all of the requirements for standard gasoline test fuel described in Section IV.F.

The public comments were supportive of EPA establishing specifications for FFV exhaust emissions test fuel. However, some commenters stated that the ethanol content and RVP specifications for FFV exhaust emissions test fuel should be based on typical values for in-use E85. (417) Automobile manufacturers commented that EPA should wait to finalize FFV test fuel specifications until a review of in-use E51-83 fuel quality can be completed in later 2013. They stated that this would allow the FFV test fuel specifications to be representative of the change to in-use “E85” composition since ASTM reduced the minimum ethanol concentration from 68 to 51 volume percent.

Substantial publicly available literature exists to demonstrate that the ethanol content of fuel used in FFVs has a significant effect on vehicle emissions. The effect of ethanol content on FFV emissions becomes more pronounced with increasing ethanol concentration. The current ASTM specification for E85 provides that the ethanol content of E85 may vary from 51 to 83 volume percent depending on climactic conditions. (418) Consistent with our long standing policy regarding the exhaust emissions testing of FFVs, we continue to believe that FFVs must comply with all emissions control requirements while using any fuel that they have the potential to operate on in-use. This ensures vehicles are designed and calibrated for emissions performance across the full range of potential in-use fuel formulations. FFVs are required to have exhaust emissions certification testing conducted using both E10 and FFV exhaust emissions test fuel to account for the effect on emissions of the full range of potential ethanol blend formulations. To ensure that FFV certification testing adequately accounts for in-use emissions performance, we are finalizing the ethanol content of FFV exhaust emissions test fuel at 81-83 volume percent as proposed. Exhaust emissions testing conducted using a fuel containing 81-83 volume percent ethanol will provide results that represent the effect of ethanol on FFV emissions performance when this effect is most pronounced. The complimentary emissions certification testing required for FFVs on E10 will ensure that the effect on FFV emissions from the full range of potential in-use ethanol concentrations is represented. Given the need to ensure that FFV emissions certification testing is representative of the full range of potential in-use ethanol blends, it would be inappropriate to set the required ethanol concentration for FFV emissions test fuel based on typical in-use levels as suggested by some of the commenters.

Similarly, the RVP of FFV exhaust emissions test fuel must assure emissions performance over the range of in-use fuels. When ethanol and gasoline are blended to produce high level ethanol blends, the RVP can be and often is very low. As a result, ASTM instituted a minimum RVP for E51-83 of 5.5 psi. Given that low volatility fuels can make the control of cold start emissions more challenging, we are finalizing the RVP of FFV exhaust emissions test to be near the minimum RVP that will be encountered in-use. The 6.0 to 6.5 RVP specification finalized today will help to ensure that FFVs are designed and calibrated to maintain their exhaust emissions performance across the range of in-use fuels.

The levels of other fuel parameters for in-use E51-83 are determined by the levels of these parameters present in the gasoline blendstock used as diluted by the addition of ethanol. Therefore, we believe that requiring that the levels of these other fuel parameters present in FFV exhaust emissions test fuel be determined by the dilution of the levels present in standard gasoline emissions test fuel appropriately reflects their potential effect on emissions performance. Given the considerations discussed above in determining the FFV exhaust emissions test fuel specifications finalized today, we do not believe that there would be a substantial benefit in waiting for the completion of the E51-83 fuel quality survey currently underway to finalize FFV test fuel specifications.

As discussed in preamble Section V.H., the Agency is also considering finalizing the in-use fuel quality standards for higher level ethanol blends on which we sought comment in the NPRM. These standards included an in-use RVP standard of 9.0 psi matching that of conventional gasoline. They also contained provisions to allow the production of high-level ethanol blends for use in FFVs from natural gasoline and other higher volatility components. Were we to finalize these in-use standards, we would also consider raising the RVP for the FFV exhaust emissions test fuel.

We are revising the definition of “alcohol” in 40 CFR part 600 to align with the change in the ASTM specification for in-use fuels. Under the revised regulation, we consider an alcohol-fueled vehicle to be one that is designed to operate exclusively on a fuel containing 51 percent or more ethanol or other alcohol by volume. This is not intended to change the applicability, procedures, or requirements for the fuel economy provisions in 40 CFR part 600.

4. Implementation Schedule

As described earlier in this Section IV, we are establishing Tier 3 exhaust and evaporative emission standards. The changes in the specifications for test fuel apply to vehicles certified to these new standards. The program is designed to transition to the new test fuel during the first few years as the Tier 3 standards are phasing in. Testing requirement with the new Tier 3 test fuel starts with light-duty vehicles certified to Tier 3 bin standards at or below Bin 70, and heavy-duty vehicles certified to Tier 3 bin standards at or below Bin 250 (for Class 2b) and Bin 400 (for Class 3). For light-duty vehicles, Table IV-27 below describes the implementation schedule of the new Tier 3 gasoline test fuels for each of the program elements in addition to the all the gasoline test fuel options available during the transition period. Table IV-3 below similarly describes the heavy-duty gasoline vehicle test fuel implementation schedule and gasoline test fuel options. The new Tier 3 PM requirements for both light-duty vehicles and heavy duty vehicles which phase-in independent of other vehicle exhaust emission requirements must be met using the certification test fuel for meeting the NMOG+NO X standards.

Starting with model years 2020 for light-duty and 2022 for heavy-duty, all manufacturers will use the new test fuel for all exhaust emission testing (with the exception of small volume manufacturers and small businesses, which can delay using the new test fuel for all vehicles until model year 2022). Manufacturers also need to comply with cold temperature CO and NMHC standards using the new test fuel for any models that use the new test fuel for meeting the light-duty Tier 3 exhaust emission standards as indicated in the tables below. These same tests will also provide the basis for meeting GHG requirements under 40 CFR part 86 and fuel economy requirements under 40 CFR part 600, as described in the following section.

Table IV-27—Exhaust Emissions Gasoline Test Fuels for LDVs, LDTs, and MDPVs
Emission compliance programTest purpose: demonstration of compliance to the emissions standards:Test cyclesFTP City/HWFE/SFTPCold CO and NMHCHigh altitude
Tier 2Certification(1)(2)(3)(4)(1)(1)
Confirmatory and In-useCertification fuel and/or (1)*(1)(1)
Tier 3 Early 2015 to 2017Certification(1)**(2)***(3)(4)(1)**(3)(1)**(3)
Confirmatory and In-useCertification fuel(1)**(3)(1)**(3)
Tier 3 phase-in 2017 to 2019Certification(1)**(2)***(3)(4)(1)**(3)(1)**(3)
Confirmatory and In-useCertification fuel(1)**(3)(1)**(3)
Tier 3 complete 2020+Certification(3)(4)(3)(3)
Confirmatory and In-useCertification fuel and/or (3)*(3)(3)
Table IV-28—Exhaust Emissions Gasoline Test Fuels for Heavy Duty Vehicles
Emission complianceprogramTest purpose: demonstration of compliance to the emissions standardsTest cyclesFTP City/HWFE/SFTPHigh altitude
Pre-Tier 3Certification(1)(2)(3)(4)(1)
Confirmatory and In-useCertification fuel and/or (1)*(1)
Tier 3 Early 2016 to 2017Certification(1)** (3)(4)(1)**(3)
Confirmatory and In-useCertification fuel(1)**(3)
Tier 3 phase-in 2018 to 2021Certification(1)** (3)(4)(1)**(3)
Confirmatory and In-useCertification fuel(1)**(3)
Tier 3 complete 2022+Certification(3)(4)(3)
Confirmatory and In-useCertification fuel and/or (3)*(3)

Additionally, heavy-duty gasoline engines (HDGEs) not subject to new Tier 3 exhaust emission standards (those certified for exhaust emissions using an engine dynamometer) are required to be certified on Tier 3 fuel by MY 2022. Further discussion can be found in Section IV.C.4.c.

For evaporative emission testing, manufacturers will need to use the new test fuel for any models that are to be certified to the Tier 3 evaporative emission standards. To the extent that these models are different than those used for exhaust emission testing with the new test fuel, manufacturers will need to do additional testing to demonstrate compliance with all applicable standards. They may alternatively use the new test fuel earlier than the regulations specify to avoid additional testing. We further require that manufacturers submit certification data based on the new test fuel to demonstrate compliance with refueling emission standards for any vehicles that are certified to the Tier 3 evaporative emission standards.

5. Implications of Emission Test Fuel Changes on CAFE Standards, GHG Standards, and Fuel Economy Labels
a. Test Fuel

Under regulations in 40 CFR part 600, vehicles use the same test fuel in emission testing conducted for CAFE standards, greenhouse gas (GHG) emissions, and the fuel economy label as that used for emission testing for criteria pollutants. This includes the test fuel used for testing on all five cycles (FTP, highway fuel economy test (HFET), US06, SC03, and Cold FTP). In the Tier 3 NPRM, EPA proposed a change in emissions test fuel used to determine compliance with criteria pollutant standards and this test fuel change would also apply to CAFE and GHG standards and the fuel economy label such that a common test fuel under 40 CFR part 600 was retained. At the same time, EPA indicated its commitment to the principle that the change in test fuel would not affect the stringency of the CAFE or GHG standards and that the labeling calculations would be updated in a future action to reflect the change in test fuel properties.

The NPRM indicated that more data and time were needed to assess the effects on stringency and implementation of these programs. While EPA's initial review of available data suggested that the change in test fuel would not impact the GHG standards, more time and data were needed to confirm this initial view regarding the GHG standards and to determine what adjustments if any would need to be made to the CAFE program and fuel economy label calculation procedures to account for the change in test fuel. EPA indicated we would defer action on appropriate adjustments, if any, for the GHG and CAFE programs until data were available to assess how the difference in the fuel properties (Tier 3 fuel compared to Tier 2 fuel) would impact the stringency of the CAFE and GHG standards for Tier 3 technology vehicles and the calculations for the fuel economy label. EPA indicated that any adjustments or changes in the regulatory text would be done through a future action.

Manufacturers commented that EPA should take action on the necessary adjustments to compliance calculations as part of the Tier 3 final rule. The methodologies for addressing some elements of the changes in fuel properties such as the difference in energy density are already addressed in the regulation. One key element, the “R” factor found in the equation of 40 CFR 600.113-12(h)(1) is intended to capture inefficiencies and differences in how vehicles respond to changes in the energy content of the fuel. This factor is empirically based, developed using vehicle test data. This value is presently set at 0.6 and is shown in the denominator of the aforementioned equation. While there has been some data evaluated to assess the impact of changing the emission test fuel on the “R” factor, EPA did not propose a value in the NPRM and specifically stated that we would continue to investigate this issue and if necessary address it as part of a future action, as opposed to changing it in the Tier 3 final rule. Furthermore, as discussed above, there is a need for more data to fully understand how other changes in certification fuel for Tier 3, such as the octane specification, may affect the stringency of the CAFE and GHG standards which were based on Tier 2 emission fuel, as well as any implications for the fuel economy label. These potential effects are best understood using emission data generated on Tier 3/LEV III vehicles tested on both Tier 3 and Tier 2 test fuel.

In addition, the manufacturers commented that even with the use of “analytically derived data” as permitted under current EPA regulations and guidance,  (419) EPA should finalize an appropriate test procedure adjustment in the Tier 3 rulemaking, including adoption of an “R” factor of 1.0, and should allow manufacturers the option of using Tier 2 fuel for CAFE, GHG, and fuel economy labeling at least through MY2019 to provide time for adjusting to the new test fuel.

In the NPRM, EPA indicated that we would not be changing the “R” factor or implementing other adjustments or changes in the regulatory text in the FRM. In follow-up meetings with the manufacturers, we expressed a willingness to consider permitting GHG and CAFE to continue on Tier 2 fuel until the future rulemaking action to address the “R” factor and other potential changes was complete and in effect. The manufacturers responded that under this approach the existing regulations would require a significant amount of additional emission testing for any model certified to the Tier 3/LEVIII exhaust emission standards before the future rulemaking is completed and in effect. (420) This is because Tier 3 test fuels would be used in emission data vehicles (EDVs) evaluated for compliance with Tier 3 criteria pollutant standards, but these same EDVs would also have to be tested on Tier 2 fuel for GHG, fuel economy label, and CAFE program data purposes. Also, while Tier 2 fuel would apply to EDVs and fuel economy data vehicles (FEDVs) evaluated for fuel economy label, CAFE program data values, and compliance with GHG standards, these same FEDVs would have to be retested on Tier 3 fuel to show compliance with the Tier 3 criteria pollutant standards. This additional testing would also extend to in-use verification program (IUVP) testing under 40 CFR 86.1845 through 86.1853.

In response to the concerns expressed by the manufacturers, EPA has identified five interim changes to existing regulations to both clarify testing requirements and to provide the manufacturers a reasonable opportunity to continue to test for CAFE, GHG, and labeling purposes on Tier 2 test fuels for each EDV and FEDV until such time as EPA determines appropriate adjustments, if any, related to a change to Tier 3 test fuels. EPA believes these changes can be implemented without impacting the integrity of the testing conducted for the criteria pollutant, CAFE, and GHG standards or values generated for determination of fuel economy labels. It is very important to note that the emission test data generated by these early Tier 3/LEVIII vehicles covering both Tier 2 and Tier 3 test fuel will provide data needed to assess the “R” value and the impact of the fuel change on the stringency of the CAFE and GHG standards, and the calculations for the fuel economy labeling program. These data will be instrumental in developing any appropriate adjustments to maintain equivalent stringency for the CAFE and GHG standards and to update the fuel economy labeling calculations, as needed. At the present time, EPA expects to have the needed data in early to mid 2015 and will then be in a position to conduct a thorough assessment of the impacts of different emission test fuels on Tier 3/LEV III vehicles and develop any appropriate adjustments and changes, in consultation and coordination with NHTSA.

These interim changes which are presented below and shown in Table IV-29, apply only to vehicles certified to the Tier 3 and/or LEV III exhaust emission standards in the model years before the future action mentioned above takes effect. These are reflected in 40 CFR 80.600.117.

1. For any given EDV or FEDV, our regulations will require that testing related to CAFE and GHG standards and the fuel economy label must still be done on Tier 2 fuel even if criteria pollutant testing is done on Tier 3 or LEV III fuel. The “R” value used in the fuel economy equation would remain at 0.6 until any change is made in a future rulemaking.

2. The requirement continues that FEDVs are expected to meet the criteria pollutant emission standards. As a flexibility, rather than requiring FEDVs to retest on Tier 3 fuel to show that they pass the criteria pollutant emission standards, we are providing in the regulations that FEDVs may meet these standards using Tier 2 fuel on each of the five cycles (as applicable) or be subject to retesting and passing on Tier 3 fuel if they do not meet requirements on Tier 2 fuel or otherwise do not comply with 40 CFR 86.1835-01(b) and 40 CFR 600.008(b). In these circumstances, assuming a retested vehicle meets criteria pollutant standards on Tier 3 fuel, the emissions results on the Tier 2 fuel will still be used for CAFE, GHG, and fuel economy labeling purposes. Retesting on Tier 3 fuel is only required for those cycles where the FEDV did not meet the criteria pollutant standards on Tier 2 fuel.

3. As a flexibility, if EDV testing is conducted on Tier 3/LEV III fuel for criteria pollutants (all 5 cycles), then we are requiring the EDV testing to be conducted on Tier 2 fuel for only 2 cycles (FTP and HFET) for GHG and CAFE purposes. These emission results on Tier 2 fuel are expected to meet the Tier 3 criteria pollutant standards. Our regulations then require manufacturers to use these EDV Tier 2 fuel test results (FTP and HFET) for the CAFE, and GHG standards. The EDV Tier 2 fuel test results (FTP and HFET) would also be used for fuel economy label calculations except in rare cases where the EDV does not pass the litmus test or if the manufacturer voluntarily elects to use the vehicle specific 5-cycle method to determine fuel economy label values. In those two cases, the EDV would need to be tested on Tier 2 test fuel on each of the five cycles.

4. As a flexibility, during the interim model years, manufacturers may use either Tier 2 or Tier 3/LEVIII test fuel emission results to conduct the litmus evaluations for fuel economy labeling under 40 CFR 600.115-11. All emission results for the five tests involved used must be from the same test fuel. EPA believes this is appropriate since the litmus evaluation is based on a comparison of the percent differences of 2 and 5 cycle values rather than absolute differences in the values. If a manufacturer chooses to conduct the litmus evaluation using LEVIII fuel, the cold FTP test must still use Tier 3 fuel. In the situation where the manufacturer uses Tier 3/LEV III test fuel for the litmus test the R-factor will be 0.6. EPA will provide guidance on determining the values for the other fuel quality parameters needed for the fuel economy calculations when Tier 3/LEVIII fuel is used.

5. Exhaust emission testing for IUVP for GHGs shall be conducted using the same test fuel as used for criteria pollutant certification, unless the manufacturer uniformly elects to conduct its IUVP GHG testing on Tier 2 fuel. This relieves the need to conduct IUVP testing for criteria pollutants on Tier 3 fuel and GHG testing on Tier 2 fuel. EPA believes this is an acceptable interim regulatory flexibility, since the IUVP testing for GHGs does not involve the IUCP provisions of 40 CFR 86.1846-01.

Table IV-29—Interim Testing Requirements for EDVs and FEDVs on Tier 3 and Tier 2 Test Fuel
Tests/ cyclesTest fuel requirementsCriteria pollutant (EDVs) 5-cycle Tier 3 fuel or other transition option**GHG/Label/CAFE (EDVs and FEDVs) (tier 2 fuel)5-cycle (label)Tier 2 fuel2-cycle (CAFE/GHG/label)Tier 2 fuelTier 3 fuelLitmus calculation5-cycleTier 2 or Tier 3 fuelIUVPCriteria pollutantTier 3 fuelGHG*Tier 2 fuel
FTPXXXShow criteria pollutant standards are met using Tier 2 fuel or must retest on Tier 3 fuelUse 5-cycle Tier 3 fuel or Tier 2 fuel test resultsXX
HFETXXXXX
US06XXX 
SC03XX     
Cold FTPXX     

Manufacturers may use LEVIII fuel (California Phase 3) in lieu of Tier 3 fuel, but any cold FTP testing must be done using Tier 3 Cold FTP fuel. LEV III fuel is 7 RVP E10, Tier 3 fuel is 9 RVP E10, and Tier 2 fuel is 9 RVP E0.

Manufacturers have presented two points of view with regard to when the potential new requirements (including a revised “R” factor and other possible test procedure changes/adjustments related to CAFE, GHG, and fuel economy labeling) should take effect once the future rulemaking action mentioned above is complete. Some have stated that use of the new provisions should be available for use as soon as possible after the rule is completed. This would minimize the need for any future duplicate testing and put manufacturers on course for fully aligning with the new requirements quickly. Others have asked that there be lead time provided before the application of the new requirements becomes mandatory. The manufacturers have expressed concern that the use of the new requirements more quickly by one manufacturer versus another could create a competitive imbalance. At the same time, manufacturers do not necessarily want to be forced to certify all products to the new requirements by a cut-off date (e.g., 2020 model year) without EPA consideration of phase-in or phase-out provisions and data carryover.

EPA understands the manufacturers' various issues and concerns in this area. Based on the information available at this time, EPA is expecting to allow the optional use of any future adjustments for compliance calculation and labeling purposes as soon as the future rule mentioned above becomes effective. Furthermore, we expect that the mandatory use of any such new adjustments with all Tier 3 certifications would be required for the 2020 MY. These initial timing projections are subject to revision based on timing of the completion of the future action and the data and record developed in that future rulemaking.

b. Useful Life for GHG Standards

As stated above, EPA is committed to retaining equivalent stringency for GHG emissions compliance beginning in MY 2017. We need more emissions test data to better understand the GHG emission impacts of Tier 3 fuel in Tier 3 technology vehicles. However, we believe that certifying a vehicle to a longer useful life for any emission constituent would have only a beneficial effect on emissions. To address potential concerns about changes in the stringency of the GHG standards resulting from a longer useful life, we are not requiring a longer useful life for GHG emission standards, although manufacturers can optionally certify GHG emissions to a 150,000 mile, 15 year useful life.

6. Consideration of Test Fuel for Nonroad Engines and Highway Motorcycles

As described earlier in Section IV.F., we are adopting new specifications for the gasoline emissions test fuel used for testing highway vehicles subject to the Tier 3 standards. Earlier in the development of this rulemaking, EPA also considered changing the test fuel specifications for other categories of engines, vehicles, equipment, and fuel system components that use gasoline. These include a wide range of applications, including small nonroad engines used in lawn and garden applications, recreational vehicles such as ATVs and snowmobiles, recreational marine applications, and highway motorcycles. While engines in some of these categories employ advanced technologies similar to light-duty vehicles and trucks, the vast majority of these engines employ much simpler designs, with many of the engines being carbureted with no electronic controls. Because of the lower level of technology, emissions from these engines are potentially much more sensitive to changes in fuel quality.

EPA is not applying the new emissions test fuel specifications to these other categories of engines, vehicles, equipment, and fuel system components. In discussing the potential change in test fuel specifications with the large number of businesses potentially impacted by such a change, many companies supported such a change. However, a number of manufacturers raised concerns about the level of ethanol in the new fuel, the cost of recertifying emission families on the new fuel, the impact on nationwide product offerings, and the cost impact of complying with the existing standards on the new test fuel. EPA believes it is important that the test fuel for these other categories reflect real-world fuel qualities but has elected to defer moving forward now pending additional analysis of the impacts of changing the test fuel specifications for the wide range of engines, vehicles, equipment and fuel system components that could be impacted. These impacts include the impact on the emissions standards, as well as the other issues raised by the manufacturers. EPA plans to explore such a change in a separate future action.

While we are not changing the test fuel specifications for these other types of vehicles and engines, we are updating the reference standards associated with specific parameters and making minor adjustments to calculation methods. For certified engines and vehicles that have already been using the test fuel specified in § 1065.710, we are clarifying that the RVP is calculated using the same equation described above for the new fuel specified for Tier 3 vehicles. We are also taking the opportunity to align and update test methods for the various gasoline test fuels in 40 CFR part 86. Specifically, we are revising §§ 86.113 and 86.213 to (1) use both ASTM D2699 and ASTM D2700 for octane measurements involving both research and motor octane specifications (including octane sensitivity), (2) use ASTM D2622 for all sulfur measurements, which is widely used and provides superior results compared with the methods that have been referenced in the regulations, (3) use ASTM D5191 for measuring fuel volatility, including the calculation described above. We are also updating the regulations to reference a newer version of the following currently referenced procedures: ASTM D86, ASTM D1319 ASTM D2699, ASTM D3231, and ASTM D3237. All these changes and updates align with fuel specifications in 40 CFR part 1065.

7. CNG and LPG Emissions Test Fuel Specifications

There are currently no sulfur specifications for the test fuel used for certifying natural gas (CNG) vehicles. There is also no sulfur specification in 86.113 for the test fuel used for certifying liquefied petroleum gas (LPG) light duty vehicles. The LPG certification test fuel for heavy-duty highway engines and for nonroad engines in 1065.720 includes an 80 ppm maximum sulfur specification. We requested comment on the appropriateness of changing 86.113 to reference 40 CFR part 1065 for all natural gas and LPG test fuels. We further requested comment on amending these specifications to better reflect in-use fuel characteristics, and in particular on the appropriateness of aligning the sulfur specifications with those that apply for gasoline test fuel. We noted that changing the sulfur specifications would depend on establishing that the new specification is consistent with the range of properties expected from in-use fuels.

The Alliance of Automobile Manufacturers (the Alliance) stated that EPA should adopt a 10 ppm maximum sulfur specification for CNG and LPG vehicle certification test fuels. They also stated that § 86.113 should reference part 1065 for CNG and LPG test fuels (for light and heavy duty vehicles). The American Petroleum Institute (API), the Association of Fuel and Petrochemical Manufacturers (AFPM), and several individual refiners stated that EPA should not establish new sulfur standards for CNG and LPG vehicle certification test fuels until additional data are available on the sulfur content of in-use CNG/LPG fuels. The National Propane Gas Association (NPGA) stated that they are opposed to a change in the sulfur specifications for LPG vehicle certification test fuels given that they are unaware of any issue that would warrant such a change.

As discussed in Section V.J. of today's preamble, additional time is needed for EPA to work with industry to collect data on current CNG/LPG sulfur content, to determine whether additional control of in-use CNG/LPG sulfur content is needed, and to evaluate the feasibility and costs associated with potential additional sulfur controls. Therefore, we are deferring finalizing in-use quality and certification test fuel specifications for CNG and LPG at this time.

G. Small Business Provisions

We are adopting special flexibility provisions for small businesses that are subject to the Tier 3 emissions standards. Such businesses are typically vehicle manufacturers, independent commercial importers (ICIs), or alternative fuel vehicle converters. We are also providing Tier 3 flexibility to companies that, though they may not meet the eligibility requirements for small businesses, sell less than 5,000 vehicles per year in the United States, and thus qualify as small volume manufacturers (SVMs). These companies and small businesses typically face similar challenges in implementing new EPA vehicle standards.

As in previous vehicle emissions rulemakings in which we have provided such flexibilities, our reason for doing so is that these entities generally have more implementation difficulty than larger companies. Small companies generally have more limited resources to carry out necessary research and development; they can be a lower priority for emission control technology suppliers than larger companies; they have lower vehicle production volumes over which to spread compliance costs; and they have a limited diversity of product lines, which limits their ability to take advantage of the phase-in and averaging provisions that are major elements of the Tier 3 program.

We proposed small business provisions largely based on the recommendations of the Small Business Advocacy Review (SBAR) Panel, described in Section XIII.C of the Notice of Proposed Rulemaking (NPRM). We proposed provisions for additional lead time, reduced testing requirements, and opportunities for hardship relief to help small entities to leverage technological developments by others and to spread the availability of needed engineering, supplier, and capital resources. Based on the comments we received, we have improved on the proposed provisions in the final rule as described in detail below.

1. Lead Time and Relaxed Interim Standards

We proposed that small businesses and SVMs be allowed to postpone compliance with the standards and other Tier 3 requirements, including use of the new certification test fuel, until model year (MY) 2022. For MY 2022 and later, they would be subject to the same Tier 3 requirements as other manufacturers, including the declining fleet average NMOG+NO X standards and the fully phased in 30 mg/mi FTP standard for MYs 2025 and later. We requested comment on adopting relaxed FTP NMOG+NO X standards for small companies in the light-duty market segment, noting that LEV III provides light-duty SVMs with relaxed FTP NMOG+NO X standards of 125 and 70 mg/mi in MYs 2022 and 2025, respectively.

We did not receive comments from non-SVM small businesses subject to the Tier 3 vehicle standards about our proposed small entity phase-in provisions. However, we received comments from SVMs, as well as the Alliance of Automobile Manufacturers and the Association of Global Automakers, arguing that the proposed phase-in did not provide adequate lead time relief for SVMs, and that the long-term Tier 3 standards for light-duty vehicles are not technologically feasible for SVMs. They highlighted the ability of large manufacturers to offset high emissions from high-performance, luxury models by averaging with their low-emitting models, while competing SVM products must be designed to actually achieve low emissions while still meeting customers' performance expectations. Their limited production can also result in emission control technology suppliers placing a lower priority on SVM orders than on those of larger, high-volume manufacturers.

Because of these factors, SVMs suggested that their companies meet a slightly more stringent NMOG+NO X standard (125 mg/mi) than what we proposed for SVMs in the early years of the program, and a permanently relaxed standard of 51 mg/mi beginning in MY 2022. Ferrari suggested a compliance schedule for SVMs similar to the California LEV III program, with either a permanently relaxed standard (matching the California LEV III 70 mg/mi long-term standard) or a delay until MY 2030 to meet the primary 30 mg/mi Tier 3 standard (when they suggest that SVMs could potentially comply). CARB comments supported Tier 3 adoption of its LEV III provisions for SVMs, including the long-term 70 mg/mi standard beginning in MY 2025. VNG, a natural gas fuel network provider, commented that gaseous-fuel small-volume test groups should be given extended phase-in opportunities identical to those proposed for SVMs, regardless of company size. As justification, VNG pointed to challenges unique to converting vehicles to operate on natural gas: thermal management of direct injection fueling and engine oil systems, adaptation of gasoline direct injection (GDI) controls to natural gas port fuel injection, and improvement of turbocharger response times.

After considering the comments, we agree with SVMs that their unique logistical and technological challenges, especially in the later years of the primary FTP NMOG+NO X standards phase-in schedule, warrant a significant period of relaxed standards for these manufacturers. However, we have found no fundamental reason why, given sufficient lead time, all manufacturers, regardless of company size and vehicle characteristics, will not be able to meet the Tier 3 standards. Thus, we are finalizing an optional program for SVMs, available to non-SVM small businesses as well, under which they can choose an alternative 3-stage FTP NMOG+NO X fleet average standard phase-in schedule: an initial standard of 125 mg/mi for MYs 2017 through 2021, a more stringent standard of 51 mg/mi for MYs 2022 through 2027, and the final Tier 3 standard of 30 mg/mi thereafter.

Because companies choosing this 3-stage compliance option are certifying to Tier 3 bin standards in MY 2017, we are requiring that other exhaust emissions standards, including SFTP and PM standards, apply for their vehicles as well, to the same degree and on the same schedule as for other manufacturers. Application of evaporative emissions and onboard diagnostics (OBD) standards, on the other hand, is not affected by choice of the 3-stage compliance option for the FTP NMOG+NO X standards, and small companies may separately choose to delay compliance with evaporative emissions and OBD standards (except as noted in Section IV.G.3) until MY 2022, as proposed. In addition, small companies choosing the 3-stage compliance option may delay the longer useful life and new test fuel requirements for exhaust emissions standards until MY 2022 to align these changes with the 3-stage schedule. This option would not preclude use of other applicable small entity flexibility provisions discussed in this subsection.

Although we are adopting this revised implementation schedule for SVMs and small businesses, we believe the proposed approach of allowing postponement of Tier 3 compliance until MY 2022 may be useful for small companies needing more lead time to begin certifying Tier 3 vehicles. Therefore we are finalizing the proposed approach as an additional but separate option for such companies, including SVMs, ICIs, and alternative fuel vehicle converters. Furthermore, because the optional 3-stage SVM implementation schedule, and the record of comments that prompted it, are specific to the light-duty sector, we are not extending it to heavy-duty vehicles and instead are finalizing only the proposed approach of allowing postponement of Tier 3 compliance until MY 2022 for any SVMs and small businesses in the heavy-duty sector.

Companies that take advantage of one of the SVM and small business implementation schedule provisions in either the light-duty or heavy-duty sector are not allowed to generate or use Tier 3 exhaust emissions credits in that sector while or before they are subject to significantly less stringent standards than other manufacturers. That is, they cannot earn or use Tier 3 exhaust emissions credits before MY 2022 under the 3-stage light-duty SVM revised implementation schedule, and they also cannot do so before MY 2022 if they are using the postponed compliance schedule that we proposed, unless they choose to end their use of these SVM implementation options earlier than MY 2022.

We disagree with VNG's assessment that small-volume test groups of large manufacturers should have until 2022 to comply with Tier 3. The technical challenges outlined by VNG have to do with converting gasoline vehicles to run reliably and durably on natural gas. Although these conversion challenges may be exacerbated for the new generation of turbocharged GDI vehicles, we have no evidence or comments from a vehicle manufacturer indicating that meeting Tier 3 standards is significantly more difficult for natural gas vehicles than for gasoline vehicles. Note that we are providing some relief for small volume test groups in the form of assigned deterioration factors (discussed below), but not because of feasibility concerns. Rather, we believe that assigned deterioration factors provide a sufficient alternative to the extensive process of developing a unique factor for each low-volume vehicle model. We find no justification to delay compliance with Tier 3 standards for larger manufacturers' low-volume models as requested by VNG.

2. Assigned Deterioration Factors

In Tier 3 as in past programs, manufacturers must demonstrate compliance with emissions standards throughout the vehicle's useful life. This is generally done by testing vehicles at low mileage and then applying a deterioration factor to the measured emissions levels. The deterioration factors are determined by testing emissions control systems before and after an aging process. In the past we have allowed small entities to use deterioration factors assigned by EPA instead of performing the extended testing, and we proposed to do so again for demonstrating compliance with Tier 3 exhaust and evaporative emissions standards. We did not propose specific assigned deterioration factors, but noted that the proposed delay in the small entity compliance schedule to MY 2022 would allow sufficient deterioration data from large manufacturers to accumulate for timely development of these factors.

We are adopting the assigned deterioration factor provisions for small businesses and SVMs (as well as for small volume test groups), as proposed. Commenters expressed support, and asked that the Agency commit itself to keeping these factors up to date as durability data accumulates. In response, we can state that we are committed to periodically updating and publishing these assigned deterioration factors. Given that SVMs will be allowed to use the revised implementation schedule described above, starting in MY 2017, it becomes necessary to consider assigned deterioration factors in stages. Because there may not be a sufficient base of accumulated durability data on Tier 3 vehicles by MY 2017, we expect that the current set of assigned factors based on Tier 2 vehicles may continue in place for some time, noting that the MY 2017-2021 SVM fleet average of 125 mg/mi is not too much different from the average of today's Tier 2 vehicle emissions. By MY 2022, when the SVM NMOG+NO X fleet average standard drops to 51 mg/mile, we expect to have new assigned factors available. We note that small businesses and SVMs may also, with advance EPA approval, use deterioration factors developed by another manufacturer (40 CFR 86.1826-01(b)).

3. Reduced Testing Burden and OBD Requirements

Under our existing regulations, manufacturers must perform in-use testing on their vehicles and demonstrate that their in-use vehicles comply with the emissions standards. These regulations provide for reduced levels of testing for small companies with annual sales under 15,000, and for no in-use testing for those with annual sales up to 5,000. We received no adverse comments on our proposal to continue this approach in Tier 3, and are retaining it.

As described in Sections IV.A and IV.B, we are requiring manufacturers to test for PM emissions from vehicles of all fuel types, a change from previous practice in which non-diesel vehicles could be waived from PM testing. However, we proposed and have decided to continue the PM testing waiver in Tier 3 for small businesses and SVMs. In lieu of testing, these companies are required to make a statement of compliance with the Tier 3 PM standards, and their vehicles are still subject to the standards. We may however measure PM emissions to determine compliance in EPA confirmatory or in-use testing.

We proposed to apply CARB's OBD requirements to Tier 3 vehicles, except that small alternative fuel vehicle converters would be allowed to instead meet our existing OBD requirements (40 CFR 86.1806-05). The natural gas fuel network provider VNG objected that the proposed exception disadvantages larger vehicle manufacturers and should be made equally available to all vehicle manufacturers' small volume test groups. We expect that larger manufacturers wishing to produce alternative fuel vehicles will be familiar with CARB's OBD requirements and well-positioned to implement these requirements in Tier 3. We note that larger OEMs themselves did not request to be covered by an extension of this provision.

We are finalizing the exception to the Tier 3 OBD requirements as proposed. Note that the optional delay in Tier 3 implementation until MY 2022 that is available to small businesses, discussed above, includes a delay in the Tier 3 OBD requirement to MY 2022, as proposed, except that vehicles already meeting this requirement in MY 2017 must continue to do so in subsequent years. We are also adopting this Tier 3 OBD delay to MY 2022 for small companies taking advantage of the revised light-duty 3-stage implementation schedule discussed above, even though other Tier 3 requirements start for them in MY 2017, in order to avoid overburdening these manufacturers with multiple sets of new OBD design constraints.

4. Hardship Relief

We proposed and are adopting provisions for small businesses and SVMs in hardship situations to apply for additional time to meet the Tier 3 standards. Such appeals will need to include evidence that the noncompliance would occur despite the manufacturer's best efforts to comply, and that severe economic hardship would occur if the relief is not granted, though the company need not show that its solvency will be in jeopardy without the relief. (This showing is required in other EPA programs granting hardship relief under 40 CFR 1068.250.) The duration of relief will be established on a case-by-case basis for Tier 3 and is not being limited by regulation. Commenters supported these proposed provisions, within the context of a revised approach to SVM lead time, discussed above.

5. Eligibility for the Flexibilities

As proposed, we are using the federal Small Business Administration (SBA) criteria to define small businesses eligible for the special provisions. SBA defines small business vehicle manufacturers as those with less than 1,000 employees, and small business ICIs and alternative fuel vehicle converters are evaluated using SBA criteria based on annual revenues. See Section IV.H.3 for a discussion of additional provisions that apply specifically to ICIs. Also, as proposed, we are defining SVMs in 40 CFR 86.1838-01 for purposes of Tier 3 as companies with nationwide annual U.S. sales volumes at or below 5,000 vehicles, though the 15,000 vehicle threshold used in Tier 2 continues to apply in a few regulatory provisions that Tier 3 changes are not impacting. Eligibility will be evaluated using an average of 2012-2014 MY sales. For companies with no 2012 MY sales, projected sales may be used, but their eligibility will be re-evaluated thereafter using a three-year running average.

VNG commented that the proposed 5,000 vehicle threshold could potentially limit the ability (or willingness) of natural gas SVMs to scale up production by forcing a tradeoff between sales and regulatory burden, pointing also to the fact that 15,000 vehicles is only 0.1% of annual light-duty vehicle sales. We do not believe that the SVM relief provisions are so advantageous as to cause self-limiting of sales, except possibly in the unlikely case of a company very near the threshold. Even if this were to happen, moving the threshold to 15,000 would not prevent the same dynamic from happening at that sales level. Furthermore, our use of a three-year average of sales for determining SVM eligibility protects the SVMs from being penalized for having an especially good year not reflective of its long-term growth trend. See the MY 2017 and later light-duty GHG final rule for a discussion of our basis for adopting the 5,000 vehicle threshold (77 FR 62793, October 15, 2012).

We requested comment on extending eligibility for the Tier 3 SVM provisions to small manufacturers that are owned by large manufacturers but are able to demonstrate that they are operationally independent. We established such a provision in the light-duty greenhouse gas (GHG) program, and CARB did so in LEV III. Comments from CARB and Ferrari supported this extension. No commenters opposed it; however, Advanced Biofuels USA recommended caution to avoid advantaging SVMs capable of leveraging parent company resources to drastically increase U.S. market share within 2-3 years. Given the establishment of this provision in our GHG program, and the value of this extension for harmonization with LEV III, we are adopting this change into Tier 3 using the same eligibility criteria as in our GHG program, set forth in 40 CFR 86.1838-01(d). We believe these criteria are sufficiently strict and objective to address the concerns expressed by Advanced Biofuels USA.

To qualify as SVMs in either the light-duty or heavy-duty Tier 3 programs, the company's total sales of vehicles subject to standards under 40 CFR part 86, subpart S count toward the vehicle sales limit, including both light- and heavy-duty vehicles. Companies so qualified may take advantage of SVM provisions in both sectors.

H. Compliance Provisions

1. Exhaust Emission Test Procedures

We are finalizing most of the amendments we proposed to 40 CFR part 1066 as part of the effort to migrate test requirements from 40 CFR part 86. We began this process a couple of years ago when we established part 1066, but we applied these test procedures only to certain vehicles above 14,000 lbs gross vehicle weight rating (GVWR) for the purpose of measuring greenhouse gas emissions (76 FR 57470, September 15, 2011). This final rule extends these procedures, with some amendments, to vehicles at or below 14,000 lbs GVWR for measurement of both criteria pollutants and greenhouse gas emissions. The procedures in part 1066 cover the same requirements that have been included in 40 CFR part 86, but include more detailed specifications for how to measure exhaust emissions using a chassis dynamometer. They also reference large portions of 40 CFR part 1065 to align test specifications that apply equally to engine-based and vehicle-based testing, such as CVS and analyzer specifications, calibrations, test fuels, calculations, and definitions of many terms. Overall, the part 1066 procedures represent a modernization of the part 86 procedures rather than fundamentally different procedures.

Until this rule, testing requirements related to chassis dynamometers have relied on a combination of regulatory provisions, EPA guidance documents, and extensive learning from industry experience that has led to a good understanding of best practices for operating a vehicle in the laboratory to measure emissions. The revisions we are finalizing capture this range of material, integrating and organizing these specifications and procedures to include a complete set of provisions to ensure that emission measurements are accurate and repeatable.

This final rule includes the following revisions to part 1066:

  • Clarification of regulatory requirements.
  • Migration of mass-based emission calculations from part 86 to part 1066.
  • Introduction of a new NMOG calculation.
  • Revision of 40 CFR part 1066, subpart B, to increase the specificity with which part 1065 references are made as they pertain to testing equipment, test fluids, test gases, and calibration standards.
  • Addition of coastdown procedures for light-duty vehicles.
  • Reordering of the test sequence with respect to vehicle preparation and running a test.
  • Specifying part 1065 procedures for PM measurement, including certain deviations from part 1065 for chassis testing.
  • Insertion of detailed test specifications for vehicles certified under 40 CFR part 86, subpart S.
  • Addition of provisions related to testing with four-wheel drive dynamometers, as described below.
  • Correction of typographical errors.

We are finalizing the use of part 1065 for PM measurement with slight adjustments to the dilution air temperature, minimum dilution ratio, and background measurement requirements. By controlling the parameters that affect PM formation (dilution air temperature, dilution factor, sample residence time, filter face temperature, and filter face velocity), the procedures will reduce lab-to-lab and test-to-test variability.

The regulations being finalized will provide alternative approaches to sample PM onto different combinations of filters. One option is to collect a sample for phases 1 and 2 of the FTP on a single filter, and collect a sample for phases 2 and 3 of the FTP onto a second filter. Another option is to collect a sample for phases 1, 2, and 3 on a single filter. A final option is to sample PM emissions from two full UDDS cycles; however manufacturers choosing this option must still run a separate three-bag test for evaporative emission testing. We will continue to allow sampling under the traditional FTP methodology of a bag or filter per test phase (3 phases in total) instead of these new methods. We are also finalizing new PM sampling and calculation methods as proposed.

We are revising the chassis dynamometer specifications in part 1066 by removing the maximum roll diameter and by requiring speed and force measurements at a minimum frequency of 10 hertz (Hz). Some manufacturers may be interested in testing with nonstandard dynamometer configurations, such as new flat-track dynamometers or old twin-roll dynamometers. We may approve the use of these and other nonstandard dynamometer configurations as alternative procedures under 40 CFR 1065.10(c)(7).

We proposed that EPA may test vehicles with the capability of all-wheel drive operation with dynamometers operating in either two-wheel drive or four-wheel drive mode, regardless of the type of dynamometer that the manufacturer used for certifying the vehicle. However, the final regulations specify that we will conduct our testing using the same drive mode as the manufacturer. Vehicle manufacturers commented that differences in test results between a vehicle tested on a two-wheel drive and a four-wheel drive dynamometer might be due to differences in dynamometer characteristics more than in vehicle operation. Results of a government-industry study that tested vehicles on both two-wheel and four-wheel drive dynamometers indicated fuel economy differences in the range of ±4%, although the study was inconclusive with respect to the cause of the differences. (421) Based on the results of this study, we will continue to test vehicles during confirmatory tests using the manufacturer's dynamometer configuration for that vehicle, and that test will be the official certification result. We are, however, finalizing revisions to 40 CFR 1066.410(g) to clarify that we may also test the manufacturer's vehicle in a different dynamometer configuration than what was used for certification testing for information-gathering purposes. If we decide to perform this testing, we will depend on the manufacturer to cooperate in reconfiguring the test vehicle for our testing. We will continue to investigate the effects of four-wheel drive dynamometers on emission results and will not rule out possible future test procedure changes that might require certification of, or allow EPA to perform confirmatory testing on, any vehicle on a four-wheel drive dynamometer.

In their comments to this rulemaking, vehicle manufacturers stressed the importance to them that EPA use the same test procedures that they used for their certification testing when we perform confirmatory testing on their vehicles. Although the manufacturers did not explain the reasons for their comment, we presume that the manufacturers' concern relates to situations where EPA test procedures would lead to higher emission levels than those resulting from a slightly different test procedure used by a manufacturer. If so, the concern is misplaced. The purpose of EPA's test procedure flexibility provisions is not to allow manufacturers to use test procedures as a tool to enable compliance with the standards—in other words, to demonstrate compliance for engines that in the absence of the regulatory test procedure flexibility would not meet the standard. Rather, the purpose is to reduce the burden of testing. We go through the rulemaking process to establish the specified default test procedures as a means of creating an objective measure of compliance with emission standards. Where we also include alternative procedures, they generally are not intended to change the conclusions from the rulemaking related to the stringency of the emission standards, or to lead to a different conclusion regarding compliance relative to the specified test procedures. EPA has addressed this issue previously for engine testing in § 1065.10(a), where we note that we condition the allowance to use alternate procedures on the provision that they would “not affect your ability to show that your engines comply with the applicable emission standards.” We note further that this provision “generally requires emission levels to be far enough below the applicable emission standards so that any errors caused by greater imprecision or inaccuracy do not affect your ability to state unconditionally that the engines meet all applicable emission standards.”

In a related context, § 1065.10(c)(1) explains that the intent of the test procedures is “to produce emission measurements equivalent to those that would result from measuring emissions during in-use operation”. This provision, which also applies for vehicle testing, envisions a process in which both the manufacturer and EPA can apply their engineering judgment to improve the representativeness of the testing. It would be appropriate for a manufacturer to ask EPA to modify our test procedures if the manufacturer believed EPA's test procedures would lead to results that were unrepresentative of in-use operation. However, it would not be appropriate to ask us to modify our test procedures to make them less representative of in-use operation.

The proposed rule included discussion of SI units as part of emission measurement procedures. At this time we are not converting emission standards to SI units. Note however that like part 86, part 1066 relies extensively on calculations involving physical parameters to calculate emission rates and perform various calibrations and verifications. As already reflected in part 1066, manufacturers have used a variety of units to perform these calculations. We would expect that dynamometers and other laboratory equipment are all capable of operating in SI units even if current practice in some laboratories is to use other units. Moving toward standardized units for calculations will allow us to more carefully and appropriately specify precision values for various measured and calculated parameters. This will also simplify calculations, facilitate review of results from different laboratories, and help with communications regarding any round robin testing that might occur.

As proposed, we will phase in the part 1066 test procedures for certifying all sizes of chassis-tested vehicles. All aspects of part 1066 related to PM testing must be met at the start of MY 2017 for vehicles certified to the PM standards. All other aspects of part 1066 must be met starting with the certification of MY 2022 vehicles. Manufacturers may begin using the part 1066 procedures before these deadlines, including step-wise changes to migrate gradually to part 1066 procedures. The regulations will require that good engineering judgment be used during this transition to ensure that the effective stringency of the standards is not changed. We recognize that individual differences between part 86 and part 1066 test procedures may have a slight upward or downward impact on measured emissions, even though the combined overall impact will be negligible. Thus, during the migration, care must be taken to avoid applying an unbalanced mix of changes that could bias emissions.

As described in Section IV.D, we are finalizing new test fuel specifications for E10 gasoline test fuel in 40 CFR part 1065. The test fuels specified for natural gas and liquefied petroleum gas, while not used for very many engine families, are currently following different specifications under 40 CFR part 86 and part 1065. We intend to revisit these fuel specifications in the future in the hope of adopting single, comprehensive fuel specifications for natural gas and liquefied petroleum gas that properly represent in-use fuels for highway and nonroad applications.

The proposal also included various technical amendments to 40 CFR part 1065, which we are finalizing largely as proposed. See the Summary and Analysis of Comments for a discussion of changes we have made in response to comments. Of particular note is the revision to subpart F specific to preconditioning engines with exhaust aftertreatment devices. We are also adopting test procedures for unregulated pollutants such as semi-volatile compounds (PAHs, etc.). These technical amendments, which have no effect on the stringency of any emission standards, include several minor changes to clarify regulatory requirements, align with chassis-testing procedures where appropriate, and correct typographical errors.

2. Reduced Test Burden

We are updating the regulatory provisions that allow manufacturers to omit testing for certification, in-use testing, and selective enforcement audits in certain circumstances. Sections IV.A.3, IV.B.6, and IV.G.3 describe how this applies for demonstrating that vehicles meet the Tier 3 PM standards. We are also allowing manufacturers to omit PM measurements for fuel economy and GHG emissions testing that goes beyond the testing needed for certifying vehicles to the Tier 3 standards. Requiring such measurement would add a significant burden with very limited additional assurance that vehicles adequately control PM. We are also allowing manufacturers to ask us to omit PM and formaldehyde measurement for selective enforcement audits. If there is a concern that any type of vehicle would not meet the Tier 3 PM or formaldehyde standards, we will not approve a manufacturer's request to omit measurement of these emissions during a selective enforcement audit.

The existing regulations have allowed for waived formaldehyde testing for gasoline- and diesel-fueled vehicles. The Tier 3 NMOG+NO X emission standards are stringent enough that it is unlikely that vehicles will comply with the NMOG+NO X standards while exceeding the formaldehyde standards. We are therefore continuing this waiver practice, such that manufacturers of Tier 3 vehicles do not need to submit formaldehyde data for certification.

3. Miscellaneous Provisions

The following additional certification and compliance provisions are included in the final rule:

  • The certification practice for assigned deterioration factors that are available for both small volume manufacturers and small volume test groups has matured significantly since it was first adopted. We are revising § 86.1826 to more carefully reflect the current practice. For example, the existing regulations specified that manufacturers with sales volumes between 300 and 15,000 units per year should propose their own deterioration factors based on engineering analysis of emission data from other families. We believe it is best for EPA to develop a set of assigned deterioration factors that can apply to all small volume manufacturers and small volume test groups. The revised regulation accordingly spells out a process for EPA to use available information to establish assigned deterioration factors that can be used for any number of manufacturers and test groups.
  • The regulations in 40 CFR part 86 rely on rounding procedures specified in ASTM E29. This standard is revised periodically. The newer versions are not likely to change in a way that affects the regulation, but the updates make it difficult to maintain a coordinated reference to the current protocol. We are addressing this by specifying that the rounding protocol described in 40 CFR 1065.20(e) applies, unless specified otherwise. We are not changing all the references in part 86; rather, we are defining “round” in subparts A and S to have the meaning given in 40 CFR part 1065 so that all new regulatory text would rely on this new description. The rounding specifications in 40 CFR part 1065 are intended to be identical to those in the latest versions of ASTM E29 and NIST SP811. For example, this now includes procedures for nonstandard rounding, such as rounding to the nearest 25 units, or the nearest 0.05, where that is appropriate.
  • Independent Commercial Importers (ICIs) are companies that import specialized vehicles into the U.S. and are subject to EPA requirements specified in 40 CFR part 85, subpart P. The standards that apply to the imported vehicles depend in part on the vehicle's model year. Therefore, vehicles imported by ICIs in the future will eventually be subject to the Tier 3 standards. Because all existing ICIs are small businesses, the Tier 3 standards generally do not apply until 2022 at the earliest. In addition, the certification practices for ICIs have matured significantly since they were first adopted. EPA is adopting two changes to update how the regulations affect ICIs. First, we are adopting a requirement for ICIs to use electric dynamometers when running exhaust emission tests. Electric dynamometers have been required for many years for vehicle manufacturers, and EPA believes it is time to require that ICIs use such test equipment. In cases where an ICI can demonstrate that they will incur a substantial increase in compliance costs, the regulations include a provision allowing EPA to approve requests on a case-by-case basis to allow testing on other types of dynamometers until the ICI is able to use an electric dynamometer meeting current specifications. Second, we are adopting an allowance for ICIs to use a specific set of reduced testing procedures for up to 300 vehicles each year that have been modified to a U.S.-certified configuration. This has been allowed for ICIs since 1999 and was approved under EPA's authority to establish equivalent alternate test procedures. (422) Instead of running a full set of emission tests, the reduced-testing requirements allow ICIs to run an FTP for exhaust emissions, a highway fuel economy test, and the hot soak test and the one-hour diurnal emission test that applied prior to the evaporative emission test procedures that involve 24-hour cycling of ambient temperatures. We do not believe these changes will have any significant cost impacts on ICIs. Most ICIs have electric dynamometers or can upgrade for a relatively small cost. The reduced testing burden provisions keep the cost of testing low, compared to the cost of running a full set of emission tests that would otherwise be required.
  • We are adopting CARB's onboard diagnostic requirements for light-duty vehicles, light-duty trucks, and heavy-duty vehicles at or below 14,000 lbs GVWR, as described in Section IV.C.5.d. We currently allow for this as an option, and almost all manufacturers do this already to avoid certifying multiple systems. Now that we are adopting evaporative provisions that are largely based on California's regulatory specifications and we are making efforts to adopt a single, national regulatory program, we believe this is an appropriate step. These changes apply starting in MY 2017 for vehicles subject to Tier 3 standards. In the case of alternative fuel conversions, we continue to apply the requirements of 40 CFR 86.1806-05.
4. Manufacturer In-Use Verification Program (IUVP) Requirements

The fuel on which an in-use vehicle will be operated and tested is considered an integral part of the vehicle's emission control system design. The Tier 2 program recognized that to achieve the desired emission reductions, vehicles must operate on the same fuel that the emission control system was originally designed to encounter in-use and during testing. In the Tier 2 program, we acknowledged that during the transition of the in-use fuel from sulfur levels of 300 ppm to 30 ppm average level, vehicles designed for 30 ppm could encounter in-use sulfur levels well above the level for which their emission control systems were designed. To address this issue, we allowed manufacturers, with agency approval, to perform specific preconditioning test procedures during the IUVP testing to ensure that potential exposure to high sulfur fuel would not impact the emission test results. These procedures included specific drive cycles or maneuvers not regularly encountered during normal in-use operation that would result in removal of sulfur contamination from the emission control system.

Consistent with the Tier 2 program, EPA continues to recognize the importance of the fuel to the emission control system design, particularly on Tier 3 vehicles designed to meet the most stringent emission levels of the program (i.e., Bin 70 and cleaner). Under the requirements of this final rule, in-use fuel will transition from an average sulfur level of 30 ppm to a new average level of 10 ppm. These sulfur requirements are average standards. Thus, even after the transition to the 10 pm average sulfur level, vehicles may still encounter sulfur levels during in-use operation that are above 10 ppm, and as high as the 95 ppm cap, which could adversely impact the emission control system. Tier 3 vehicles tested by manufacturers in IUVP that have been exposed to such sulfur levels could experience sulfur-related impacts, which in turn could cause the vehicle to temporarily exceed emission standards.

To address the potential emission impact on Tier 3 vehicles from exposure to higher sulfur levels, we are modifying the IUVP testing process based in part on what was allowed under the Tier 2 program. Tier 3 vehicles tested in the IUVP are to be tested initially without allowing any sulfur cleanout procedure, such as a US06 test run prior to the FTP or Highway Fuel Economy (HFET) tests. If a vehicle fails the NMOG+NO X standard for the FTP or HFET cycle during the initial round of testing, manufacturers may perform a sulfur cleanout procedure before repeating the FTP or HFET, consisting of up to two US06 cycles. The measured US06 cycle and a preconditioning US06 cycle, if performed as part of the initial measured tests would serve as the cleanout procedure and therefore no additional US06 cycles would be allowed. Alternative sulfur cleanout procedures would require EPA approval. Following the sulfur cleanout procedure, the manufacturer will prep and soak the vehicles and then repeat the FTP and HFET tests. Manufacturers choosing to perform the sulfur cleanout procedure would need to submit evidence that the vehicle encountered high sulfur levels in the fuel just prior to emission testing. This would need to include an analysis of a fuel sample from the vehicle fuel system as received from in-use operation just prior to testing. If the fuel sample indicated that the vehicle had been operating on fuel containing 15 ppm or higher sulfur levels, only the emission results of the tests following the cleanout procedure would be used to determine emission compliance and whether to enter the in-use compliance program (IUCP). We intend to monitor the emission results of in-use testing and sulfur-related test failures to determine if further reductions in the sulfur cap are required to ensure that Tier 3 vehicles are meeting the standards under in-use driving conditions.

The changes to the IUVP testing described above apply for light-duty vehicles, light-duty trucks, and MDPVs. These changes are not applicable to heavy-duty vehicles tested in the IUVP program. Also, as described in Section IV.D, we are incorporating leak testing into the IUVP test protocol.

V. Fuel Program

Under today's Tier 3 program, we are finalizing reductions in gasoline sulfur levels nationwide. These standards will help reduce current levels of sulfur that contribute to ambient levels of air pollution that endanger public health and welfare. It will also help prevent the significant impairment of the emission control systems expected to be used in Tier 3 technology, significantly improve the efficiency of emissions control systems currently in use, and continue prevention of the substantial adverse effects of sulfur levels on the performance of vehicle emissions control systems.

A. Overview

1. Background
a. History of Gasoline Sulfur Control

Sulfur is naturally occurring in crude oil. Crude oil containing higher concentrations of sulfur (i.e., greater than 0.5 percent) is called “sour” and crude containing lower sulfur concentrations (e.g., West Texas Intermediate) is referred to as “sweet.” Regardless of the concentration, because sulfur is naturally occurring in crude oil, it is also naturally occurring in gasoline. As discussed in Section IV.A, sulfur impairs the performance of today's vehicle emission control technologies (i.e., precious metal catalytic converters), reducing the emission benefits of current and advanced vehicles. As explained below, in 2000 EPA took action to reduce gasoline sulfur levels under what is known as the Tier 2 Program  (423) and we are taking further action with today's Tier 3 Program.

Tier 2 was a major, comprehensive program designed to reduce emissions from passenger cars, light trucks, and large passenger vehicles (including sport utility vehicles, minivans, vans, and pick-up trucks) and the sulfur content of gasoline. Under this program, automakers were required to manufacture low-emission vehicles when operated on low-sulfur gasoline, and refiners were required to produce low-sulfur gasoline nationwide.

Required reductions in gasoline sulfur under the Tier 2 program began in 2004 with refinery and importer caps of 300 ppm and a corporate average cap of 120 ppm. For most refiners and importers, compliance with the final sulfur standards (30 annual average and 80 per-gallon cap) was required beginning in 2006. The Tier 2 program was fully implemented on January 1, 2011 (the ultra-low sulfur diesel program allowed for some extensions of the Tier 2 gasoline program flexibility provisions). The Tier 2 gasoline sulfur program also included an averaging, banking, and trading (ABT) program that allowed companies to generate credits for implementing the required changes earlier than their required start date, and allowed ongoing flexibility to meet the 30 average sulfur standard.

At full implementation, the Tier 2 program (treating vehicles and fuels as a system) required passenger vehicles to be over 77 percent cleaner and gasoline sulfur to be reduced by up to 90 percent from pre-program levels.

b. Need for Additional Gasoline Sulfur Control

The authority under which we are lowering the existing gasoline sulfur standards comes from Clean Air Act section 211(c)(1). This is because emission products of gasoline with current levels of sulfur cause or contribute to air pollution which may reasonably be anticipated to endanger public health or welfare, and because emission products of gasoline with current levels of sulfur will impair to a significant degree the emissions control device or systems on the vehicles subject to today's final Tier 3 standards. For more on our legal authority to set gasoline sulfur standards, refer to Section V.M.

As explained in Section IV.A, robust data from many sources show that gasoline sulfur at current levels (i.e., around 30 ppm on average) continues to degrade vehicle catalytic converter performance during normal operation. NO X emissions are the most significantly affected by this degradation. The NMOG+NO X vehicle emission standards, representing an 80 percent reduction from current Tier 2 standards, will not be possible without the gasoline sulfur controls we are finalizing today. Today's 10 ppm sulfur standard should enable vehicle manufacturers to certify their entire product line of vehicles to the final Tier 3 fleet average standards. Tier 3 vehicles must achieve essentially zero warmed-up NO X emissions to comply and must maintain this performance for up to 150,000 miles. An increase in emissions of only a few milligrams per mile due to sulfur could make compliance impossible for some vehicles. The standards are projected to be especially challenging for larger SUVs and pick-up trucks. Based on testing of these vehicles, as shown in Section IV.A, reducing gasoline sulfur to 10 ppm should enable these vehicles to maintain their emission performance in-use over their useful life. Lowering gasoline sulfur will also help reduce emissions of pollutants that endanger public health and welfare from vehicles already on the road today. As also discussed above in Section IV.A, we have tested a wide range of vehicles to better understand the impact that even lower gasoline sulfur could have on emissions. Our test data showed significant NO X and VOC reductions when vehicles were tested on low sulfur gasoline. As also explained in more detail in Section III.B, lowering average gasoline sulfur from 30 to 10 ppm will result in approximately 260,000 less tons of NO X and 50,000 less tons of VOC almost immediately as the Tier 3 gasoline sulfur standards take effect.

2. Summary of Final Tier 3 Fuel Program Standards

The major elements of the fuel program being finalized today are summarized below. Please refer to sections V.B through V.J for more discussion on each of the elements summarized here.

a. Annual Average Sulfur Standard

Under today's final Tier 3 fuel program, gasoline and any ethanol-gasoline blend will be required to have a sulfur level of 10 ppm or less on an annual average basis beginning January 1, 2017. The 10 ppm average will apply to a refiner or importer's annual gasoline production. Similar to the Tier 2 gasoline program, the Tier 3 program applies to gasoline in the United States and the U.S. territories of Puerto Rico and the Virgin Islands, excluding California. Please see Section V.B for a more detailed discussion of the annual average sulfur standard.

b. Per-Gallon Sulfur Caps

Refiners and importers will continue to be subject to refinery gate per-gallon sulfur caps of 80 ppm. Similarly, gasoline downstream of the refinery gate (e.g., at terminals, retail stations, etc.) will continue to be subject to a 95 ppm per-gallon sulfur cap. We are also committing to continue to evaluate if reductions in the per-gallon sulfur caps are warranted.

A more detailed discussion on our decision to continue the current 80 and 95 ppm per-gallon sulfur caps, and elements of an in-use study, can be found below in Section V.C.

c. Small Refiner and Small Volume Refinery Provisions

As described in further detail in Section V.E.1, approved gasoline small refiners and small volume refineries must produce gasoline meeting the 10 ppm annual average sulfur standard beginning January 1, 2020. Small refiners and small volume refineries who meet the 10 ppm sulfur standard prior to this date may generate credits for early Tier 3 program compliance.

d. Averaging, Banking, and Trading (ABT) Program

Section V.D discusses our averaging, banking, and trading (ABT) program. Refiners and importers may continue to generate credits for reductions in their gasoline sulfur levels below the current (Tier 2) 30 ppm average gasoline sulfur standard through December 31, 2016; and for reductions below the new 10 ppm average standard beginning January 1, 2017. These credits can be used for compliance with either the Tier 2 standard through 2016 or the Tier 3 standard beginning in 2017. The Tier 3 ABT program will have similar credit use provisions as the Tier 2 ABT program. These provisions include: Five-year credit life from the year of generation; two-trade limit for inter-company trading; and the ability to use credits internally, bank for future use, or trade to other refiners/importers. Although credits generated prior to January 1, 2017 will be valid for five years or until December 31, 2019, whichever is earlier.

e. Gasoline Additive Cap

As discussed further in Section V.C., manufacturers of gasoline additives that are used downstream of the refinery at less than 1.0 volume percent will be required to limit the sulfur contribution to the finished gasoline from the use of the additive to less than 3 ppm when the additive is used at the maximum recommended treatment rate. For each batch of additive produced, the manufacturer must retain sulfur test records for 5 years, and must make these records available to EPA upon request.

Parties that introduce additives to gasoline at over 1.0 volume percent will be required to satisfy all of the obligations of a refiner and fuel manufacturer, including demonstration that the finished blend meets the applicable sulfur specification.

f. Requirements for Denatured Fuel Ethanol and Other Gasoline Oxygenates

Today's rule finalizes a 10 ppm sulfur cap for denatured fuel ethanol (DFE). While DFE is the predominant gasoline oxygenate currently in use, these standards also apply to other gasoline oxygenates. Today's rule finalizes a 3.0 volume percent limit on ethanol denaturant concentration. We are adopting the current ASTM International specifications that only natural gasoline, gasoline blendstocks, or gasoline may be used as denaturants for DFE. (424) As discussed in the Summary and Analysis of Comments, we believe it is not necessary to finalize the proposed additional limits on the potential denaturants that may be used at this time. We are also finalizing regulatory text to state that DFE must be composed solely of carbon, hydrogen, oxygen, and sulfur. Testing, recordkeeping, and reporting obligations are also being finalized to implement these new standards as discussed in Section V.G. Sulfur testing using approved analytical methods or volumetric blending records and denaturant product transfer documents (PTDs) can be used by manufacturers/importers of DFE in demonstrating compliance with the 10 ppm sulfur cap for DFE finalized today.

g. Fuel Used in Flexible Fuel Vehicles

As discussed in Section V.H., we are deferring finalizing additional fuel quality requirements for E16-50 and E51-83 at this time. We continue to believe in the importance of implementing additional fuel quality standards for higher-level ethanol blends and will continue to work with stakeholders in their development following the publication of this final rule.

h. Standards for Butane and Pentane

As discussed further in Section V.I, we are finalizing a 10 ppm sulfur cap for butane blended into gasoline effective January 1, 2017. This is consistent with the Tier 3 10 ppm refinery average sulfur specification finalized today. In addition, as discussed below in Section VI.A.4, we are also finalizing provisions to allow pentane to be blended into gasoline downstream of the refinery. These provisions are similar to the existing provisions for butane blending. This allowance will become effective June 27, 2014; a 30 ppm sulfur cap will apply to pentane blended into gasoline (consistent with the existing sulfur cap for butane under the Tier 2 program) until December 31, 2016, after which a 10 ppm sulfur cap will apply.

i. CNG/LPG

As discussed below in Section V.J., we are deferring establishing in-use sulfur requirements for compressed natural gas (CNG) and liquid propane gas (LPG) to provide additional time to work with stakeholders to collect data on current CNG/LPG sulfur content; to determine whether additional control of in-use CNG/LPG sulfur content is needed; and to evaluate the feasibility and costs associated with potential additional sulfur controls.

B. Annual Average Sulfur Standard

Under today's final Tier 3 fuel program, gasoline and any ethanol-gasoline blend will be required to have a sulfur level of 10 ppm or less on an annual average basis beginning January 1, 2017. The 10 ppm average will apply to a refiner or importer's annual gasoline production. Similar to the Tier 2 gasoline program, the Tier 3 program applies to gasoline in the United States and the U.S. territories of Puerto Rico and the Virgin Islands, excluding California. We are finalizing the 10 ppm average sulfur standard both to enable the new vehicle fleet to meet the Tier 3 vehicle standards being finalized today pursuant to CAA section 211(c)(1)(B), and to reduce emissions from the existing in-use vehicle fleet that endanger public health and welfare pursuant to section 211(c)(1)(A) of the Clean Air Act (CAA).

We received numerous comments both in support of and against the proposed 10 ppm annual average sulfur standard. Commenters opposing the standard believe that 10 ppm is too low and/or is not needed to enable Tier 3 vehicle technologies. Some commenters suggested that EPA should consider setting a less stringent sulfur standard than the proposed 10 ppm annual average (detailed information regarding the comments can be found in the Summary and Analysis of Comments document, which is located in the docket for this rulemaking). We believe that a 10 ppm annual average standard will help reduce current levels of sulfur that contribute to ambient levels of air pollution that endanger public health and welfare. It will also help prevent the significant impairment of the emission control systems expected to be used in Tier 3 technology, significantly improve the efficiency of emissions control systems currently in use, and continue prevention of the substantial adverse effects of sulfur levels on the performance of vehicle emissions control systems. This level is also feasible, and is the level that appropriately balances costs with the emission reductions that it provides and enables.

As discussed in Section IV.A.6., and further in Chapter 5 of the RIA, we believe that a standard of 10 ppm is appropriate, and when combined with the advances in emissions control technologies will be sufficient to meet the Tier 3 emissions standards. The feasibility of the 30 mg/mi NMOG+NO X fleet average depends on exhaust catalyst systems that require gasoline with average sulfur levels of 10 ppm or less. Further, annual average sulfur levels greater than 10 ppm would significantly impair the emission control technology that we expect will be used to meet the Tier 3 standards and to ensure in-use compliance over a vehicle's useful life. This is particularly a concern for some larger vehicles that will need to reduce NO X to near-zero levels, due to greater difficulty in reducing cold-start NMOG, in order to meet a combined NMOG+NO X standard. As discussed in Section IV.A.6, increasing gasoline sulfur from 10 ppm to 20 ppm or 30 ppm would make it impossible for vehicle manufacturers to meet the Tier 3 standards. Achieving Tier 3 standards would require offsetting the resultant higher emissions but EPA is not aware of existing technology or developing technology that could address these higher emissions when taking into consideration the entire vehicle fleet. Increasing gasoline sulfur from 10 ppm to 20 ppm or 30 ppm would also forego the very large immediate reductions from the existing fleet.

We also do not believe a sulfur standard lower than 10 ppm is necessary to enable vehicles to meet the Tier 3 standards. As also discussed in Section IV.A, reducing sulfur below 10 ppm would further reduce vehicle emissions and allow the Tier 3 vehicle standards to be achieved more easily. However, we believe that a 10 ppm average standard is sufficient to allow vehicles to meet the Tier 3 standards. Furthermore, as discussed below, there are significant challenges associated with reducing sulfur below 10 ppm.

As explained in Section IV.A, sulfur in fuel oxidizes in the exhaust and coats the sites where chemical reactions can take place on the precious metal catalysts used in vehicles to reduce emissions of VOC, NO X, PM, CO, and toxics. Accordingly, any sulfur in gasoline causes vehicle emissions to increase. Sulfur can be burned off the catalyst during high-temperature, rich operation of the vehicle (i.e., aggressive driving conditions), but as long as there is any sulfur in the fuel, exhaust emissions will increase. Because any amount of sulfur in the fuel can have this effect, the lower the sulfur the better. Refiners experience the same phenomenon with precious metal catalysts used in the reformer and isomerization units at their refineries. (425) To protect the precious metal catalysts in these units, refiners reduce the sulfur in the feed to these units to 1 ppm or below. Thus, it is technically possible for refiners to reduce their gasoline sulfur levels to virtually zero. While refiners did not have reason to reduce the sulfur in FCC gasoline until Tier 2 required such reductions, some refiners have achieved reductions in this stream at some of their refineries for other reasons such as: (1) Protecting the FCC catalyst from the contaminants in the gas oil feed, (2) reducing stack emissions from the regenerator of the FCC unit, and most importantly (3) increasing gasoline yields from the FCC unit. For most refineries, FCC gasoline accounts for about one-third of gasoline and before Tier 2 was the source of over 95 percent of the sulfur in gasoline. Under Tier 2, most refiners significantly desulfurized FCC gasoline to around 70 to 80 ppm, yet FCC gasoline continues to contribute the majority of sulfur in gasoline today.

An annual average sulfur level of 10 ppm will achieve very large immediate reductions from the existing fleet, as discussed in Sections III and IV. Because any sulfur in gasoline will continue to impair vehicle catalyst performance, reducing sulfur levels to zero would maximize vehicle emission reductions. However, there are two reasons why we believe a 10 ppm average sulfur standard is sufficient and further reductions (e.g., 10 ppm cap or 5 ppm average) are not necessary at this time. First, our analysis shows that a 10 ppm annual average is sufficient to enable vehicles to reach the Tier 3 standards. Consequently, while reducing sulfur levels further would continue to yield reductions from the in-use fleet, they would not be necessary to enable the new Tier 3 vehicle standards to be met. Second, while sulfur levels would continue to reduce emissions from the existing fleet, reducing sulfur further below 10 ppm becomes increasingly difficult and costly. FCC naphtha is very rich in high-octane olefins. As the severity of desulfurization increases, more olefins are saturated, further sacrificing the octane value of this stream and further increasing hydrogen consumption. Making up for this lost octane represents a significant portion of the sulfur control costs. Furthermore, as desulfurization severity increases, there is an increase in the amount of sulfur removed (in the form of hydrogen sulfide) which recombines with the olefins in the FCC naphtha, thus offsetting the principal desulfurization reactions. There are means to deal with the recombination reactions, but they result in even greater capital investments. In addition, while FCC gasoline contributes the majority of sulfur to the finished gasoline, as the sulfur level drops below 10 ppm, the sulfur level of the various other gasoline streams within the refinery also become important. Any necessary treatment of these additional streams increases both capital and operating costs.

U.S. refineries are currently in different positions, both technically and financially. In general, they are configured to handle the different crude oils they process and turn them into a widely varying product slate to match available markets. Those processing heavier, sour crudes may have a more challenging time reducing gasoline sulfur under the Tier 3 program. Also, those with higher sulfur levels in other refinery streams may have a more difficult time desulfurizing gasoline. Perhaps most important, U.S. refineries vary greatly in size (atmospheric crude capacities range from less than 5,000 to more than 500,000 barrels per day) and thus have different economies of scale for adding capital to their refineries. Therefore, it can be less costly per gallon for some larger refineries to get down to 10 ppm than for smaller refineries, as discussed in Chapter 5 of the RIA. As a result, with a 10 ppm average standard, the flexibility afforded by the ABT program helps those refineries with very high costs. They have the option of staying above 10 ppm if they can acquire credits from other refineries that were able to lower their sulfur level below 10 ppm. However, if the gasoline sulfur standard were lower, this would essentially end the ability of refiners to average sulfur reductions across their refineries. There simply would not be enough opportunity to generate credits at levels much below 10 ppm.

As discussed further in Chapter 5 of the RIA, we assessed the potential costs of an annual average standard lower than 10 ppm (e.g., 5 ppm). Our analysis shows that sulfur control costs for refineries to meet a standard below 10 ppm could be on the order of two times more costly per ppm-gallon of gasoline sulfur reduced. In addition, a standard below 10 ppm could be cost-prohibitive for more challenged refineries. Further, such a standard would also introduce additional costs to address the contribution to gasoline sulfur from gasoline additives, transmix, ethanol denaturants, and contamination in the distribution system.

Therefore, we believe that the 10 ppm annual average standard will help reduce current levels of sulfur that contribute to ambient levels of air pollution that endanger public health and welfare. It will also help prevent significant impairment of the emission control systems expected to be used in Tier 3 technology, significantly improve the efficiency of emissions control systems currently in use, and continue prevention of the substantial adverse effects of sulfur on the performance of vehicle emissions control systems. The level is also feasible (especially considering its associated ABT provisions, described in Section V.D), and is the point which appropriately balances costs with the emission reductions that it provides and enables.

C. Per-Gallon Sulfur Caps

1. Standards

The final Tier 3 program is composed of a 10 ppm refinery annual average sulfur standard (discussed above in Section V.B) with an 80 ppm per-gallon cap at the refinery gate and a 95 ppm per-gallon cap downstream; these per-gallon caps currently exist under the Tier 2 program. We believe this is the most prudent approach for lowering in-use sulfur while maintaining flexibility considering cost and other factors. These per-gallon caps are important in the context of an average sulfur standard to provide an upper limit on the sulfur concentration that vehicles must be designed to tolerate. The caps also limit downstream sulfur contamination and enable the enforcement of the gasoline sulfur standard in-use. Our 10 ppm average standard with higher per-gallon caps compares to a 10 ppm cap standard in much of Europe, Japan, and Korea. In addition to the gasoline standards we are finalizing today, we are also finalizing caps on the sulfur content of gasoline additives, to limit their contribution to the overall in-use gasoline sulfur level.

a. What We Proposed

We proposed two options for the per-gallon sulfur caps—maintaining the Tier 2 80 ppm refinery gate sulfur and 95 ppm downstream sulfur caps and, beginning January 1, 2020, lowering to 50 ppm refinery gate and 65 ppm downstream caps. The 50 ppm refinery gate cap was proposed to take effect on January 1, 2020, as this is the date when the small refiner, small volume refinery, and early credit use provisions would expire; and also to avoid forcing additional refinery investments during the early credit usage period. We also requested comment on lowering the caps to 20 ppm at the refinery gate and 25 ppm downstream.

We received comments on both of the proposed per-gallon cap options of 80/95 ppm and 50/65 ppm, as well as comments on finalizing lower caps of 20/25 ppm and a 20 ppm overall cap. Comments supporting lower caps noted potential environmental benefits, greater certainty that vehicles would see lower and more uniform gasoline sulfur levels, and enabling new vehicle technologies that require very low sulfur levels. Comments in support of maintaining the current Tier 2 caps cited concerns on cost, flexibility for turnarounds/unplanned shut downs (due to refinery fire, natural disaster, etc.), and potential impacts on gasoline supply and pricing. Detailed information regarding the comments we received on the per-gallon sulfur caps is provided in the Summary and Analysis of Comments document, which is available in the rulemaking record.

b. Final Refinery Gate Sulfur Cap

In today's action, we are retaining the 80 ppm refinery gate cap. The refinery gate cap provides flexibility for batch-to-batch variability that naturally occurs at a refinery due to the varying types of crude that refineries process, variations in unit operations, and variations in product mix. It further provides for flexibility during unit turnarounds, and unplanned upsets (e.g., refinery fires, natural disasters, etc.), to avoid a complete refinery shutdown. A lower cap could create situations where refiners would need to store more off-spec gasoline for future processing. However, if a refinery does not have adequate tankage for storing this product, and/or if its processing units are not large enough to “catch up” in refining off-spec product, it could result in significant impacts to fuel pricing or supply. For a refiner that produces multiple products, any potential supply impacts could also impact other fuel markets (e.g., diesel, jet fuel, etc.). Additionally, the refinery gate cap is a “hard” limit—a refinery's actual production has to be well below this limit to account for in-use testing tolerances, safety margins, and any additives that a refiner may need to add prior to the fuel leaving the refinery. An 80 ppm refinery gate cap will provide refiners needed flexibility, and more certainty that they will be able to continue producing and distributing at least some gasoline during turnarounds/upsets to avoid a total shutdown. It will further provide more certainty for transmix processors, additive manufacturers and other downstream parties.

As described below in Section VII, we believe that most refineries would not have significant costs as a result of the Tier 3 program because they will be able to meet the 10 ppm average sulfur standard largely through revamps and operational changes at their facilities, rather than installing grassroots units. Lowering to a cap of 50 ppm would directionally increase the costs of the Tier 3 program. The American Petroleum Institute (API) provided a detailed study with their comments  (426) quantifying the additional costs associated with successively more stringent per-gallon caps. While we do not agree with the study's overall cost analysis, we do agree that with a refinery gate cap of 50 ppm, a number of refiners would incur higher capital costs due to the decreased ability to handle off-spec product with a lower refinery gate cap. As refiners must ensure that they can continue to produce saleable product and meet demand in the event of an upset or an off-spec batch of fuel, the need for installation of additional tankage and/or increased refinery processing capability would be greater with a 50 ppm refinery gate cap. While at the time of the proposal we believed that a cap of 50 ppm would have little cost impact, our more recent analysis shows that a 50 ppm cap would increase the cost of the Tier 3 gasoline sulfur standards by approximately 10 percent (see RIA Chapter 5.2.2.4). At the same time, the more stringent cap with its associated increase in cost would be unlikely to provide significant additional emission benefits nationwide. As discussed previously in Sections III and IV above, the emissions benefits associated with the Tier 3 program are mainly driven by the reduction in the average sulfur content of gasoline from 30 to 10 ppm, since vehicle emissions are proportional to the sulfur content of the fuel. Changes in the cap would not affect this. In the context of the final ABT provisions, a higher cap does allow for increases in emissions on a temporal basis as one batch of fuel is allowed to have higher sulfur levels. However, this is then offset by reductions in emissions from batches of fuel that are then required to be below the 10 ppm average standard. Similarly, the final ABT provisions allow for the possibility that the fuel from different refineries will cause varying emission reductions as one refinery's higher average sulfur levels would lead to less emission reductions in-use. However, this is then offset by greater reductions in emissions due to the fuel produced by refineries with sulfur levels below the average standard.

Based on our cost analysis, which is discussed below in Section VII.B., we project nearly 40% of the gasoline pool would be at 5 ppm, about 45% at 10 ppm and the remaining approximately 15% at levels higher than 10 ppm. The sulfur level for this 15% in our analysis ranges from 11 ppm all the way up to 70 ppm. However, as discussed in Section VII.B., these high sulfur levels are more a function of the limitations of our analysis where we could only model these refineries as remaining at their current Tier 2 sulfur levels. We anticipate that in most (if not all) cases refineries will make operational changes and/or investments in order to reduce their credit burden and reduce their compliance costs. This anticipation, along with the fact that a 10 ppm average standard by definition limits the amount of gasoline that can remain at higher sulfur levels (regardless of the cap), means that we anticipate most refineries, including those using credits, will still average less than 20 or 30 ppm in their physical gasoline production. Nevertheless, the final ABT program does allow for the possibility (regardless of the cap) that were this higher sulfur fuel to be concentrated in any certain geographical area, it would not receive the full emission reductions from the Tier 3 program. We have considered the potential for areas to consistently receive fuel that might be predominantly higher than the 10 ppm average. Because refineries generating credits and using credits are interspersed across the country, and because most areas receive a considerable portion of their fuel by pipeline, barge, rail, or truck from refineries in other areas, we expect the variation in average sulfur levels across the country to be too limited to warrant lowering the per-gallon cap to 50 ppm. Given the stringency of the 10 ppm average standard, we predict that in-use sulfur levels will generally be well below 50 ppm.

Further reductions in the refinery gate cap are also not needed to enable the vehicle emissions standards, as the vehicle standards are a function of the 10 ppm annual average sulfur standard. While vehicle manufacturers have expressed concerns about the potential impacts on emissions performance if individual vehicles are exposed to gasoline above 10 ppm due to higher per-gallon caps and/or credit usage, (427) we believe that vehicles will see sulfur levels closer to the 10 ppm average rather than the 80 ppm cap due to the fact that the 10 ppm average will drive reductions in gasoline sulfur levels.

Thus, we believe it is prudent at this time to retain an 80 ppm refinery gate cap. However, we are committing to monitor and further evaluate in-use sulfur levels and their impact on vehicle emissions. If it is warranted, we will reassess the sulfur cap level and the need for potential future regulatory action. Such ongoing evaluation will include analyses of: In-use fuel surveys; batch data that refineries are required to submit; and the sulfur credit market. It will also include the evaluation of any issues or concerns that might arise during implementation of the program. Finally, we will also carry out an ongoing evaluation of data submitted by the vehicle manufacturers on the performance of their Tier 3 vehicles in-use.

c. Downstream Sulfur Cap

With regard to the downstream sulfur cap, we believe that maintaining a 15 ppm differential between the refinery gate sulfur cap and the downstream sulfur cap will provide pipeline operators, transmix processors, and gasoline additive users the same flexibility as was provided under the Tier 2 program. As was the case under the Tier 2 program, allowing a 15 ppm differential is needed to ensure adequate flexibility in accommodating gasoline produced from transmix, instances of contamination during distribution, and for the use of necessary (sulfur-containing) additives. In rare circumstances when the sulfur contribution from all these sources are coincidently at their maximum levels, a very limited number of batches of gasoline at the 95 ppm downstream sulfur cap may be present in the distribution system. However, we expect that this will not have a substantial impact on the average sulfur content of in-use gasoline. Comments received on this issue were generally in support of maintaining the 15 ppm delta.

Pipeline operators are currently allowed to blend limited volumes of transmix into gasoline in their systems provided that the resulting gasoline meets all fuel quality specifications and the endpoint of the blended gasoline does not exceed 437 °F. (428) This enables pipeline operators to avoid the installation of additional transmix storage and loading equipment at a number of remote locations to facilitate shipping small volumes of transmix to processing facilities by truck.

Currently transmix processors must produce gasoline sufficiently below the 95 ppm downstream sulfur cap to accommodate any downstream sulfur increases from the use of gasoline additives and contamination from further distribution. The sulfur content of the gasoline produced by transmix processors is determined by the sulfur content of the transmix they receive, which in turn is primarily a function of the sulfur content of gasoline and jet fuel components in the transmix. (429) Transmix processors do not handle sufficient volumes to support the installation of currently available desulfurization units. (430)

d. Accounting for Ethanol Blending in the Determination of Compliance With Gasoline Sulfur Requirements

In demonstrating compliance with the gasoline sulfur standards finalized today, gasoline refiners and importers may adjust the sulfur levels in the gasoline and blendstocks for oxygenate blending (BOBs) that they produce/import to account for the downstream addition of ethanol. We proposed that the sulfur content of denatured fuel ethanol (DFE) used for downstream blending would be assumed to be 10 ppm in making such demonstrations of compliance. Refiners commented that refiners and importers should be allowed to either use the actual sulfur value of the DFE or conduct laboratory hand blends of a representative sample of DFE to determine the effect on the sulfur content of the blended fuel from the addition of DFE. We agree that refiners and importers should be allowed to use the actual sulfur content of DFE when a sulfur test result is available and when the refiner can demonstrate that the test result was derived from a representative sample of the DFE that was blended with the gasoline or BOB. The sulfur content of in-use DFE will typically be lower than the 10 ppm sulfur cap finalized today for DFE. We assumed that DFE would have an average sulfur content of 5 ppm in conducting the refinery analysis to support this rule. Therefore, today's final rule requires that in determining their compliance with today's sulfur standards, refiners and importers must either use the actual sulfur content of the DFE established through testing of the DFE actually blended or assume a 5 ppm sulfur content for the DFE added downstream. To prevent potential bias, a refiner or importer must choose to use only one method during each annual compliance period.

2. Requirements for Gasoline Additives

Today's action finalizes the requirement that manufacturers of gasoline additives used downstream of the refinery at less than 1 volume percent must limit the sulfur contribution to the finished gasoline from the use of their additive to no more than 3 ppm when the additive is used at the maximum recommended treatment rate. The additive manufacturer will be required to maintain records of its additive production quality control activities which demonstrate that the sulfur content of additive production batches is such that when the additive is used at its maximum recommended treatment rate it will add no more than 3 ppm to sulfur content of the finished gasoline. We received comments in support of our proposed requirements (these comments can be found in the docket to this rulemaking, and are summarized in the Summary and Analysis of Comments document, which is also located in the docket). An environmental organization commented that the sulfur contribution from additives can have a material effect on emissions performance given the level of vehicle emissions control that is being finalized today. We also received comments from gasoline additive manufactures were in favor of the proposed controls.

The requirements finalized today are designed to prevent the potential dumping of high sulfur materials into gasoline under the guise of the addition of gasoline additives. We continue to believe that all current gasoline additives contribute less than 3 ppm to the sulfur content of the finished fuel when used at the maximum recommended treatment rate (with 3 ppm being the extreme). Normal additive production quality control practices already have had to consider the sulfur contribution of the additive to finished gasoline as a result of the Tier 2 gasoline sulfur requirements. The maximum recommended treatment rate is already stated on product transfer document or packaging for the additive. Additive manufacturers are to retain production quality control records for 5 years and make these available to EPA upon request. Therefore, the requirements finalized today will not constrain the use of genuine gasoline additives or result in significant additional costs to gasoline additive manufacturers. Parties that introduce additives to gasoline at over 1.0 volume percent are required to satisfy all of the obligations of a refiner and fuel manufacturer including demonstration that the finished blend meets the applicable sulfur specification.

We also received comments from an environmental organization requesting that EPA promulgate limits on the combined sulfur contribution for all additives blended into a batch of gasoline in addition to controlling the sulfur contribution from individual additives. We believe that such additional controls are not necessary, would add an unwarranted additional compliance burden, and could interfere with the use of necessary downstream additives. Certain additives that provide critical fuel performance characteristics (e.g., corrosion control, demulsifiers) contain sulfur-containing compounds as an essential functional component. Such additives are used to remedy specific instances of gasoline quality problems, and their treatment rate is governed by the desire to limit the added cost from their use.

D. Averaging, Banking, and Trading Program

In today's rule, we are finalizing an ABT program that will reduce the compliance costs and promote the feasibility of the Tier 3 gasoline sulfur program, because it will allow refiners and importers to choose the most economical compliance strategy (i.e., investment in technology, credits, or both) to meet the 10 ppm average sulfur standard. In response to comments received on our proposal, we have simplified and added flexibility to the ABT program. The ABT program allows refiners and importers to continue to generate credits for overcompliance with the current Tier 2 30 ppm average gasoline sulfur standard through 2016, and the new 10 ppm average standard beginning in 2017. (Small refiners and small volume refineries have a January 1, 2020 compliance date, as described below.) These credits can be used for compliance with either the Tier 2 standard through 2016 or the Tier 3 standard beginning in 2017. Credits can also be banked for future use or transferred to other refineries for compliance with the average sulfur standard. In addition, we are allowing refiners and importers to also use any valid credits banked from 2012 and 2013 under the Tier 2 program toward compliance with either the Tier 2 or Tier 3 sulfur programs. We believe these provisions will provide a seamless transition from Tier 2 to the Tier 3 program.

1. How will the ABT program assist with compliance?

The Tier 3 ABT program allows refiners and importers the flexibility to: (1) Have varying gasoline sulfur levels for their batches of fuel as long as they meet the 10 ppm average over the course of the year; (2) use credits generated at one of its refineries to offset higher sulfur gasoline produced at another of its refineries; (3) bank generated credits for future use; and (4) participate in trading (via buying and/or selling) of credits from another refiner to help lower costs. The ABT program allows for the generation of credits by refiners and importers for over-compliance with the 10 ppm sulfur standard (on a refinery/facility basis), and for the transfer of these credits to other refiners (for use a refinery) to reduce or eliminate their need to make capital investments to meet the 10 ppm standard. The ABT program will provide refiners and importers with multiple approaches to compliance, and each can choose the approach that best minimizes their costs.

2. ABT Modeling

For the proposed rule, we modeled the effects of an ABT program on refinery compliance. Our modeling determined the lowest cost approach on a refinery-by-refinery basis under two scenarios: The first, in which every refinery has the opportunity to make credit transfers with every other refinery in the nation, and a more limited scenario in which credit transfers would only occur within companies that own more than one refinery.

In developing today's final program, we also analyzed the Tier 2 credit market and found that, currently, Tier 2 sulfur credits are both transferred within companies (intra-company) and traded between companies (inter-company). As discussed in Chapter 4.3 of the RIA, in 2012 approximately 56% of the Tier 2 credit transactions were inter-company trades. The remaining 44% were intra-company trades. This analysis shows that there is a functioning, well-established gasoline sulfur credit trading market. There does not appear to be any hindrance to credit trading currently or in the future. We anticipate that a significant number of refineries will take advantage of the opportunity to generate or use credits, thus lowering their compliance burden for the Tier 3 program. For a more complete discussion of our analysis of credit trading and the associated cost impacts of the ABT program, refer to Chapter 4.3 of the RIA.

3. Eligibility

Consistent with our proposal, under today's final program, sulfur credits may be generated by both U.S. refiners and importers of gasoline into the U.S. only for gasoline that is subject to the sulfur requirements as described in the regulations at § 80.1603. This excludes gasoline produced or imported for use in California (“California gasoline”) and gasoline designated for export, but includes gasoline produced by California refineries for use outside the state. We sought comment in the proposed rule on whether or not to include California. All comments received on this issue were against the inclusion of such gasoline in the ABT program, largely because it would cause additional burden but provide no appreciable credits to the market to justify the additional compliance burden (such as batch reporting). We agree with these comments, and are finalizing the provision that California gasoline and gasoline for export will not be included in the ABT program. In order to exclude exported gasoline and California gasoline, refiners must keep records to demonstrate that the excluded gasoline was designated for export or as California gasoline, and was actually exported or used in California.

While under existing fuel programs (e.g., MSAT2), we precluded importers from generating “early” credits (credits generated before a program start date), we are allowing importers to generate early credits in the Tier 3 ABT program, consistent with our proposal. Importers were previously precluded from generating early credits because they generally did not need additional lead time to comply with our fuel standards (as they most likely would not be investing in new refining technologies) and we also thought it would be difficult for them to establish representative baselines from which early credits could be generated. However, early credit baselines are not a part of the Tier 3 program, as discussed above. In addition, commenters noted that while importers may not necessarily have to take actions to desulfurize fuel like domestic refiners do, these parties may have to pay premiums on obtaining lower sulfur fuel, thus their efforts to provide lower sulfur fuel should also be able to generate early credits. Thus, we are finalizing provisions allowing importers to generate early credits for sending over-compliant gasoline to the United States prior to January 1, 2017.

We proposed to limit credit generation to refiners and importers. Under the Tier 2 gasoline program, we allow refiners who produce gasoline by combining blendstocks together or adding blendstocks to previously certified gasoline (refiner-blenders) to participate in the in the sulfur ABT program, but do not allow butane blenders who comply with the reduced sampling, testing, and reporting requirements to participate in the sulfur ABT program. (431) We are extending these provisions to butane and pentane blenders under the Tier 3 program. (432) Under today's rule, any refiner who uses the reduced sampling, testing, and reporting provisions for blending butane or pentane into gasoline will not be allowed to generate credits. Refiner-blenders who comply with the full suite of sampling, testing, and reporting requirements will, however, be allowed to generate credits.

We received several comments from ethanol producers and Growth Energy (representing ethanol producers), who commented that DFE should be subject to an annual average sulfur standard and that ethanol producers should be able to participate in the ABT program that is available to refiners and importers of gasoline. These commenters stated that preventing them from participating in the ABT program would provide refiners with pathways to allow delayed adoption of cleaner fuel standards at their expense. Some ethanol producers commented that they should be allowed to participate in the ABT program as a means of offsetting the additional cost of the proposed per-batch sulfur testing and reporting requirements. We also received comments from refiners stating that credit generation should be permitted only for refiners and importers.

As in existing EPA fuel programs, we continue to believe that it is not appropriate expand the ABT provisions to cover ethanol producers and oxygenate blenders for several reasons. First, expanding the ABT program beyond refiners and importers could greatly increase the number of parties participating thereby potentially complicating EPA compliance assurance activities while having little overall impact on the sulfur credit pool. Second, the current ABT program under the Tier 2 gasoline sulfur program, which is limited to gasoline refiners and importers, has functioned effectively with few compliance irregularities. Third, experience with the unleaded gasoline program suggests widespread abuse and fraud when credits have been allowed to be generated or sold by parties other than refiners or importers subject to the regulations. Fourth, it would require a considerably more complicated compliance structure, including the application of all refiner responsibilities to ethanol producers and blenders. Fifth, there is no need for DFE producers to generate credits in order to recoup the value for any lower sulfur content of their product. The value of any lower sulfur content will be reflected in the market price of DFE, similar to the octane value to refiners. Sixth, the sulfur ABT provisions were included to ease the burden of compliance for refiners who have to make capital changes to their facilities to meet today's more stringent sulfur standards. In addition to reducing the cost of today's gasoline sulfur program, the ABT provisions allow for an earlier effective date of the sulfur standards than would otherwise be possible. Such considerations are not applicable to ethanol producers since capital expenditures for desulfurization equipment or other equipment will not be needed at their facilities to comply with the sulfur standards finalized today. (433) Finally, overcompliance with the per-gallon sulfur cap for DFE is not a valid basis for credit generation. We expect that in all cases, the DFE sulfur level will be below the cap. To allow credit generation for these parties, we would need to set an additional annual average sulfur standard for DFE at some level below 10 ppm and allow credits to be generated for overcompliance with that standard. Accordingly, we do not believe it is appropriate to allow ethanol producers or blenders to generate sulfur credits under the Tier 3 gasoline sulfur program, and as such, we are not finalizing such a provision. The Tier 3 rule prohibits any person downstream of the refinery or importer that produced or imported gasoline, CBOB, or RBOB who adds oxygenate to such product from including the volume and sulfur content of the oxygenate in any compliance calculations for credit generation.

4. Credit Generation and Use

Under the Tier 3 ABT program, the credit generation provisions are nearly the same as those under the Tier 2 program. In essence, the Tier 2 program simply continues with a lower standard below which credits are generated. Refiners and importers are allowed to average within and across companies to meet the standard in the most cost-effective manner possible, including generating and using credits. For Tier 3, refiners and importers can generate credits for overcomplying with the 10 ppm standard on a volume-weighted annual average basis beginning January 1, 2017. Credit generation periods remain 12 months long and continue to be synchronized with annual compliance demonstration periods. The final Tier 3 ABT program provisions for approved small refiners and small volume refineries are discussed below in Section V.D.6.

Consistent with our proposal, and to encourage early gasoline desulfurization and give the refining industry flexibility to stagger their investments over time, we are also finalizing provisions to allow refiners and importers to generate credits prior to January 1, 2017 (i.e., early credits).

We proposed an early credit program for overcompliance with the current Tier 2 30 ppm gasoline sulfur standard from January 1, 2014 through December 31, 2016. These early credits were proposed to have a credit life of three years, and could be used through December 31, 2019. We also proposed that refiners and importers who generated early credits would then have to designate them at the end of the compliance year as either Tier 2 credits or Tier 3 early credits, and the credits would have to be used in accordance with the designation.

We received several comments on our early credit provisions. Some commenters stated that the requirement to designate credits as either Tier 2 or Tier 3 would not provide refiners and importers the intended amount of flexibility, as they may not necessarily be able to predict if they will need to use those early credits for either Tier 2 or Tier 3 compliance. Further, commenters were concerned that if more refiners than anticipated opt to bank the credits for possible future use at their own refineries, it could result in less credits being made available to the market. We also received comments both in support of and against the proposed three-year credit life. Some commenters stated that they believed that all credits should have a five-year life to provide maximum flexibility and more certainty on the availability of credits; while other commenters supported the three-year credit life limit, because it would coincide with the end of the small refiner/small refinery provisions thus providing more certainty of the benefits of the Tier 3 program. Finally, while not proposed, we received many comments requesting that EPA clarify what would become of “banked” Tier 2 credits (Tier 2 credits generated in 2012 and/or 2013 that were not used before the end of their five-year credit life and would have expired when the Tier 3 program began, under the terms of the proposed ABT program). Commenters further stated that these credits should not expire, but rather should be allowed to receive their full five-year credit life, as they represent actual reductions in refinery sulfur levels.

As discussed above, we are finalizing a more flexible approach to the ABT program than what we proposed. We believe this will provide for a more seamless transition from the Tier 2 to the Tier 3 credit programs and more certainty on credit availability.

Consistent with our proposal, refiners and importers may begin generating early credits on January 1, 2014, and continue through December 31, 2016. These credits will be generated on an annual average basis, from a demonstration that the refiner or importer's annual average gasoline sulfur level is below the current Tier 2 30 ppm annual average sulfur standard. We believe this simple early credit approach is possible because U.S. gasoline was averaging around 30 ppm as we started developing the Tier 3 program, based on compliance data. Since refiners and importers would need to continue to comply with the existing Tier 2 sulfur standards during the early credit period, absent Tier 3, they would need to maintain this level of performance on an industry average basis. Accordingly, any additional gasoline sulfur reductions beyond 30 ppm will be attributed to the proposed Tier 3 program.

In response to comments received, we are not requiring that credits be designated as either Tier 2 or Tier 3 and only used for the program for which they were designated. Refiners and importers may use early credits either for ongoing compliance with the Tier 2 program, or bank them for future compliance with the Tier 3 program (within the limits of the credit life restrictions). Essentially, the Tier 2 credit generation provisions simply continue, and any banked credits generated in 2014 through 2016 that were not used for compliance with the Tier 2 standards would be carried over for use in complying with Tier 3. We believe that this will allow for more certainty of credit availability before refiners must make their Tier 3 investment decisions, thus reducing the cost of the program. It will also avoid any incentive for refiners to use up banked Tier 2 credits prior to 2017 causing increased in-use sulfur levels and emissions.

Based on our analysis of the Tier 2 credit market for 2012, we believe that there will be a balance of 2012 banked credits equivalent to approximately two months of compliance, and we anticipate a similar amount (perhaps more) for 2013. Although we did not propose to allow for banked 2012 and 2013 Tier 2 credits to be used for compliance with the Tier 3 program, we are finalizing this provision for a number of reasons. First, the Tier 2 banked credits represent real reductions—refiners and importers are currently generating these credits for overcompliance with the Tier 2 gasoline sulfur standard. Second, allowing these banked credits to receive their full five-year credit life will provide more assurance of the credit availability for trading for those who need them to comply with Tier 3 without a large capital investment. As previously explained, this will allow for more certainty of credits available far before making Tier 3 investment decisions, thus reducing the cost of the program. A lack of certainty in the credit trading market could lead to refiners banking more credits than usual for their own use rather than allowing these credits to be available in the market for trading. As shown in Chapter 4.3 of the RIA, refiners tend to hold credits as an insurance policy until they approach the end of their credit life. If credit-generating refiners continue with this approach, credits generated in 2012 and 2013 will likely be available for purchase in 2017 and 2018 for those refiners that may want to rely on them for compliance (along with additional early credits generated in 2014 through 2016). Finally, as we anticipate that these credits will be equal to about four months of compliance, the allowance of 2012 and 2013 banked Tier 2 credits makes for a more flexible program by effectively allowing for a small amount of additional lead time without adversely affecting the overall benefits of the Tier 3 program. As discussed previously, this will avoid any incentive for refiners to use up banked Tier 2 credits prior to 2017 causing increased in-use sulfur levels and emissions. We believe these provisions will allow the Tier 3 program to begin on January 1, 2017, with more certainty regarding the availability of credits for those refiners needing (or choosing) to defer investment to better align with their existing turnaround/shutdown schedules.

All credits generated before January 1, 2017 will be valid for five years or until December 31, 2019, whichever is earlier—no early credits may be used for compliance beginning January 1, 2020. Thus, banked Tier 2 credits generated in 2012 and 2013 will receive their full five-year life and will not expire at the start of the Tier 3 program. However, credits generated in 2015 and 2016 that are unused as of December 31, 2019 will expire and become invalid. We believe that structuring the early credit program this way will offer considerable flexibility to refiners phasing in Tier 3 gasoline sulfur controls, while still placing a date at which point the intended sulfur program will be fully implemented and enforceable—January 1, 2020 (the same date small refiners and small refineries must begin complying with the 10 ppm sulfur standard). This will also provide a date certain to give auto manufacturers greater confidence for the design of their vehicles that all vehicles in-use are running on 10 ppm average fuel. Otherwise, it is possible that the greater ease of generating early credits relative to 30 ppm sulfur (as opposed to 10 ppm in 2017 and beyond) would allow higher sulfur levels to continue well beyond 2019.

Consistent with our proposal, all credits generated beginning January 1, 2017 will be for overcompliance with the Tier 3 10 ppm annual average sulfur standard, and will have a five-year credit life. We believe five years will give refiners and importers sufficient time to use credits generated in previous years while still placing limitations on credit life to help with enforcement. Five years is consistent with the Tier 2 ABT program, as well as the current credit life and recordkeeping provisions for other 40 CFR part 80 fuels programs, and coincides with the applicable statute of limitations for violations by parties who generate invalid credits. Credits must be used within five years from the year they were generated (regardless of when/if they are traded), otherwise they will expire and become invalid. For example, credits generated in 2017 can be applied towards 2018-2022 compliance, as well as 2017 compliance. After March 31, 2023 (when reports for the 2022 annual compliance period will be due), credits generated in 2017 will expire and become invalid. Similarly, credits generated in 2018 can be applied towards 2019-2023 compliance, as well as 2018 compliance. After March 31, 2024, credits generated in 2018 expire, and so on and so forth.

5. Credit Trading Provisions

We are also finalizing provisions for credit trading in Tier 3 that are identical to those under the current Tier 2 program. As in that program, it is possible that sulfur credits could be generated by one party, subsequently transferred or used in good faith by another, and later found to have been calculated or created improperly or otherwise determined to be invalid. As in the current Tier 2 program as well as other 40 CFR part 80 fuel programs, if this occurs, we are requiring that both the seller and purchaser will have to adjust their sulfur calculations to reflect the proper credits and either party (or both) could be determined to be in violation of the standards and other requirements if the adjusted calculations demonstrate noncompliance with the 10 ppm standard.

Sulfur credits must be transferred directly from the refiner or importer generating them to the party using them for compliance purposes. This ensures that the parties purchasing them are better able to assess the likelihood that the credits are valid. As proposed, we are also finalizing an exception for the case where a credit generator transfers credits to a refiner or importer who inadvertently cannot use all the credits. In this case, the credits can be transferred a second time to another refiner or importer. After the second trade, the credits must be used or they will expire. Allowing a maximum of two trades is consistent with other recent fuel programs and we believe it is sufficiently flexible while still preserving adequate means for enforcement. While some commenters stated that they believe the two-trade maximum is not necessary given the fact that credits are only being traded within a small part of industry, we believe that unlimited trading could result in an unenforceable program and potentially lead to problems with invalid credit trading. Given the widespread use of credit trading between different companies under the Tier 2 program despite this provision, there appears to already be sufficient flexibility in the program. We received comments requesting that we clarify that intra-company trading will continue to be unlimited as it is in the existing Tier 2 ABT program. Intra-company trading will in fact remain unlimited, and we have added language to the final Tier 3 regulations to clarify this.

There are currently no prohibitions against brokers facilitating the transfer of credits from one party to another. Any person can act as a credit broker, regardless of whether such person is a refiner or importer, as long as the title to the credits is transferred directly from the generating refiner or importer to the using refiner or importer. This prohibition on outside parties taking ownership of credits was promulgated in response to problems encountered during implementation of the unleaded gasoline program, and has since been extended to subsequent fuels rulemakings. We continue to believe that maintaining this prohibition will allow for maximum program enforceability and consistency with all of our other ABT programs for mobile sources and their fuels.

6. ABT Provisions for Small Refiners and Small Volume Refineries

Consistent with our proposal, approved small refiners and small volume refineries must comply with the 10 ppm annual average standard by January 1, 2020, which allows for an additional three years for compliance. This is the primary form of relief offered to small refiners and small volume refineries under the Tier 3 gasoline sulfur program (discussed further in Section V.E, below). Approved small refiners and small volume refineries may continue to generate credits for overcomplying with the 30 ppm Tier 2 standard before January 1, 2020. Prior to January 1, 2017, credits generated by small refiners and small volume refineries can be traded/sold to non-small refiners for use by December 31, 2019, and the credit revenues could be used to help offset their Tier 3 investments.

When the Tier 3 program begins on January 1, 2017, small refiners and small volume refineries may continue to generate credits for overcompliance with the 30 ppm sulfur standard (as they will still be subject to the Tier 2 standards through December 1, 2019), or they may generate credits for overcompliance with the Tier 3 10 ppm sulfur standard. We are finalizing that small refiners and small volume refineries must designate their credits as being generated for either the Tier 2 or Tier 3 ABT program, as proposed. Credits designated and generated as Tier 2 credits may only be traded with other small refiners and small volume refineries (and these credits may only be used for compliance through December 31, 2019). However, credits designated and generated as Tier 3 credits may be traded with non-small refiners as well. Additionally, from January 1, 2017 through December 31, 2019, if a small refiner's annual average sulfur level is below 10 ppm, they may elect to split the designation and generation of credits between both the 10 ppm and 30 ppm standards (without double-counting). For example, in 2017, a small refiner with an annual gasoline sulfur average of 8 ppm could generate 20 ppm-volume Tier 2 credits (30 ppm-10 ppm) that could be used by other small refiners and small volume refineries, or banked by the refinery for future Tier 2 compliance. This small refiner would also generate 2 ppm-volume Tier 3 credits (10 ppm-8 ppm) that could be sold to refiners and importers subject to Tier 3, or banked by the refiner for future Tier 3 compliance.

7. Deficit Carryforward

Under the final Tier 3 sulfur program, we are finalizing deficit carryforward provisions similar to the existing Tier 2 program, whereby an individual refinery that does not meet the 10 ppm standard in a given year may carry a credit deficit forward for 1 year. Under this deficit carryforward allowance, the refinery will have to make up the credit deficit and come into compliance with the Tier 3 sulfur standard the next calendar year. We received comments expressing concern that it will be more challenging for refineries to make up their credit deficit in one year with a 10 ppm sulfur standard, and requesting that the deficit carryforward allowance be extended to two or three years. We disagree with these comments primarily because of concerns with the enforceability of allowing for a deficit beyond one year. In addition, we believe that an extended deficit carryforward will further delay Tier 3 sulfur reductions. While we acknowledge that there might be an increased hurdle for some refiners to make up their own credit shortfall, we believe the ABT program provides ample opportunity to purchase credits from others. However, in recognition of unanticipated circumstances, such as where credits are unavailable or are prohibitively expensive such that the refiner could not make up the deficit in one year, the Tier 3 hardship provisions provide EPA with the authority to allow for extended deficit carryforward, if a refiner's hardship petition demonstrates that it meets the criteria. Thus, we are finalizing that a refiner could carry a deficit forward for up to 3 years only in cases of hardship situations, as described below in Section V.E.2.

E. Additional Program Flexibilities

1. Regulatory Flexibility Provisions
a. Small Business Regulatory Flexibility Provisions

We are finalizing several regulatory flexibility provisions for small entities in the fuels industry to reduce the burden that the Tier 3 program could have on them. As in previous fuel rulemakings, our justification for including provisions specific to small businesses is that these entities generally have a greater degree of difficulty in complying with the standards compared to other entities.

In developing the Tier 3 gasoline sulfur program, we evaluated the environmental need as well as the technical and financial ability of refiners and others in the fuel industry to meet the sulfur standards as expeditiously as possible. We believe it is necessary and feasible for the vast majority of the program to be implemented in the established time frame to achieve the air quality benefits as soon as possible. Based on information available from small refiners and others (as discussed further in the Regulatory Flexibility Analysis description in Section XII.C), we believe that the category of entities classified as small generally face unique circumstances with regard to compliance with environmental programs, compared to larger entities. Thus, as discussed below, we are finalizing several regulatory flexibility provisions for small refiners to reduce the burden that the Tier 3 program could have on them.

Small entities as a category generally lack the resources that are available to larger companies to raise capital for investing in a new regulatory program, such as shifting of internal funds, securing of financing, or selling of assets. Small entities are also likely to have more difficulty in competing for any needed engineering and construction resources. This is because the magnitude of their projects tends to be both smaller and less profitable for the contracted firms. As such, we are including provisions in today's rule that would provide assistance for small entities in meeting the 10 ppm sulfur standards. This proposed approach would allow the overall program to begin as early as possible; achieving the air quality benefits of the program as soon as possible, while helping to ensure that small entities have adequate time to raise capital for new fuel desulfurization equipment or to make any other needed changes. We also believe that small business regulatory flexibilities can provide these entities with additional help and/or time to accumulate capital internally or to secure capital financing from lenders, and could spread out the availability of any needed engineering and construction resources in a manner that they are available by the time they are needed.

i. Delayed Standards for Small Refiners

We are finalizing a compliance date of January 1, 2020 for small refiners, allowing small refiners to postpone compliance with the Tier 3 program for up to three years. Small refiners will have from January 1, 2017 through December 31, 2019 to continue production of gasoline with an average sulfur level of 30 ppm (per the Tier 2 gasoline sulfur program). This delayed compliance schedule for small refiners is not intended as an opportunity for those refiners to increase their production of gasoline with sulfur levels greater than 10 ppm, but rather will help small refiners with compliance with the program. Since the compliance costs for their competitors may rise during these three years and since their gasoline will be sold into the same fungible market, this delay will not only provide them more lead time, but also financial support towards later compliance. Compliance with the 10 ppm annual average sulfur standard will begin on January 1, 2020 for small refiners. Further, as discussed in greater detail in Section V.D.5, a small refiner would be allowed to continue using Tier 2 gasoline sulfur credits through December 31, 2019 to meet their refinery average 30 ppm sulfur standard.

ii. Refinery Gate and Downstream Caps

During the Small Business Regulatory Enforcement Fairness Act (SBREFA) Panel process, small refiners raised the concern that a refinery gate cap of 20 ppm could cause problems during a refinery turnaround or an upset because a cap of this level could result in a refiner not being able