Skip to Content

Rule

Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles-Phase 2

This document was corrected by an document published on 06/30/2017. View Correction

Document Details

Information about this document as published in the Federal Register.

Published Document

This document has been published in the Federal Register. Use the PDF linked in the document sidebar for the official electronic format.

Start Preamble Start Printed Page 73478

AGENCY:

Environmental Protection Agency (EPA) and National Highway Traffic Safety Administration (NHTSA), Department of Transportation (DOT).

ACTION:

Final rule.

SUMMARY:

EPA and NHTSA, on behalf of the Department of Transportation, are establishing rules for a comprehensive Phase 2 Heavy-Duty (HD) National Program that will reduce greenhouse gas (GHG) emissions and fuel consumption from new on-road medium- and heavy-duty vehicles and engines. NHTSA's fuel consumption standards and EPA's carbon dioxide (CO2) emission standards are tailored to each of four regulatory categories of heavy-duty vehicles: Combination tractors; trailers used in combination with those tractors; heavy-duty pickup trucks and vans; and vocational vehicles. The rule also includes separate standards for the engines that power combination tractors and vocational vehicles. Certain requirements for control of GHG emissions are exclusive to the EPA program. These include EPA's hydrofluorocarbon standards to control leakage from air conditioning systems in vocational vehicles and EPA's nitrous oxide (N2 O) and methane (CH4) standards for heavy-duty engines. Additionally, NHTSA is addressing misalignment between the Phase 1 EPA GHG standards and the NHTSA fuel efficiency standards to virtually eliminate the differences. This action also includes certain EPA-specific provisions relating to control of emissions of pollutants other than GHGs. EPA is finalizing non-GHG emission standards relating to the use of diesel auxiliary power units installed in new tractors. In addition, EPA is clarifying the classification of natural gas engines and other gaseous-fueled heavy-duty engines. EPA is also finalizing technical amendments to EPA rules that apply to emissions of non-GHG pollutants from light-duty motor vehicles, marine diesel engines, and other nonroad engines and equipment. Finally, EPA is requiring that engines from donor vehicles installed in new glider vehicles meet the emission standards applicable in the year of assembly of the new glider vehicle, including all applicable standards for criteria pollutants, with limited exceptions for small businesses and for other special circumstances.

DATES:

This final rule is effective on December 27, 2016. The incorporation by reference of certain publications listed in this regulation is approved by the Director of the Federal Register as of December 27, 2016.

ADDRESSES:

EPA and NHTSA have established dockets for this action under Docket ID No. EPA-HQ-OAR-2014-0827 (for EPA's docket) and NHTSA-2014-0132 (for NHTSA's docket). All documents in the docket are listed on the https://www.regulations.gov 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 https://www.regulations.gov or in hard copy at the following locations:

EPA: Air and Radiation Docket and Information Center, EPA Docket Center, EPA/DC, EPA WJC West Building, 1301 Constitution Ave. NW., Room 3334, 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.

NHTSA: Docket Management Facility, M-30, U.S. Department of Transportation, West Building, Ground Floor, Rm. W12-140, 1200 New Jersey Avenue SE., Washington, DC 20590. The telephone number for the docket management facility is (202) 366-9324. The docket management facility is open between 9 a.m. and 5 p.m. Eastern Time, Monday through Friday, except Federal Holidays.

Start Further Info

FOR FURTHER INFORMATION CONTACT:

EPA: Tad Wysor, Office of Transportation and Air Quality, Assessment and Standards Division (ASD), Environmental Protection Agency, 2000 Traverwood Drive, Ann Arbor, MI 48105; telephone number: (734) 214-4332; email address: wysor.tad@epa.gov.

NHTSA: Ryan Hagen, Office of Chief Counsel, National Highway Traffic Safety Administration, 1200 New Jersey Avenue SE., Washington, DC 20590. Telephone: (202) 366-2992; ryan.hagen@dot.gov.

End Further Info End Preamble Start Supplemental Information

SUPPLEMENTARY INFORMATION:

A. Does this action apply to me?

This action will affect companies that manufacture, sell, or import into the United States new heavy-duty engines and new Class 2b through 8 trucks, including combination tractors, all types of buses, vocational vehicles including municipal, commercial, recreational vehicles, and commercial trailers as well as 3/4-ton and 1-ton pickup trucks and vans. The heavy-duty category incorporates all motor vehicles with a gross vehicle weight rating of 8,500 lbs. or greater, and the engines that power them, except for medium-duty passenger vehicles already covered by the greenhouse gas standards and corporate average fuel economy standards issued for light-duty model year 2017-2025 vehicles.[1] Regulated categories and entities include the following:

CategoryNAICS code aExamples of potentially affected entities
Industry336111Motor Vehicle Manufacturers, Engine Manufacturers, Truck Manufacturers, Truck Trailer Manufacturers.
336112
333618
336120
336212
Industry541514Commercial Importers of Vehicles and Vehicle Components.
811112
Start Printed Page 73479
811198
Industry336111Alternative Fuel Vehicle Converters.
336112
422720
454312
541514
541690
811198
Note:
a North American Industry Classification System (NAICS).

This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely covered by these rules. This table lists the types of entities that the agencies are aware may 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 the referenced regulations. You may direct questions regarding the applicability of this action to the persons listed in the preceding FOR FURTHER INFORMATION CONTACT section.

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

This regulatory action is supported by influential scientific information. Therefore, EPA conducted a peer review consistent with OMB's Final Information Quality Bulletin for Peer Review. As described in Section II.C, a peer review of updates to the vehicle simulation model (GEM) for the Phase 2 standards has been completed. This version of GEM is based on the model used for the Phase 1 rule, which was peer reviewed by a panel of four independent subject matter experts. The peer review report and EPA's response to the peer review comments are available in Docket ID No. EPA-HQ-OAR-2014-0827. We note that this rulemaking is based on a vast body of existing peer-reviewed work, i.e., work that was peer-reviewed outside of this action, as noted in the references throughout this Preamble, the Regulatory Impacts Analysis, and the rulemaking docket. EPA also notified the SAB of its plans for this rulemaking and on June 11, 2014, the chartered SAB discussed the recommendations of its work group on the planned action and agreed that no further SAB consideration of the supporting science was merited.

C. Executive Summary

(1) Commitment to Greenhouse Gas Emission Reductions and Vehicle Fuel Efficiency

In June 2013, the President announced a comprehensive Climate Action Plan for the United States to reduce carbon pollution, prepare for the impacts of climate change, and lead international efforts to address global climate change.[2] In this plan, President Obama reaffirmed his commitment to reduce U.S. greenhouse gas emissions in the range of 17 percent below 2005 levels by 2020. More recently, in December 2015, the U.S. was one of over 190 signatories to the Paris Climate Agreement, widely regarded as the most ambitious climate change agreement in history. The Paris agreement reaffirms the goal of limiting global temperature increase to well below 2 degrees Celsius, and for the first time urged efforts to limit the temperature increase to 1.5 degrees Celsius. The U.S. submitted a non-binding intended nationally determined contribution (NDC) target of reducing economy-wide GHG emissions by 26-28 percent below its 2005 level in 2025 and to make best efforts to reduce emissions by 28 percent.[3] This pace would keep the U.S. on a trajectory to achieve deep economy-wide reductions on the order of 80 percent by 2050.

As part of his Climate Action plan, the President specifically directed the Environmental Protection Agency (EPA) and the Department of Transportation's (DOT) National Highway Traffic Safety Administration (NHTSA) to set the next round of standards to reduce greenhouse gas (GHG) emissions and improve fuel efficiency for heavy-duty vehicles pursuant to and consistent with the agencies' existing statutory authorities.[4] More than 70 percent of the oil used in the United States and 26 percent of GHG emissions come from the transportation sector, and since 2009 EPA and NHTSA have worked with industry, states, and other stakeholders to develop ambitious, flexible standards for both the fuel economy and GHG emissions of light-duty vehicles and the fuel efficiency and GHG emissions of heavy-duty vehicles.[5 6] The standards here (referred to as Phase 2) will build on the light-duty vehicle standards spanning model years 2012 to 2025 and on the initial phase of standards (referred to as Phase 1) for new medium and heavy-duty vehicles (MDVs and HDVs) and engines in model years 2014 to 2018. Throughout every stage of development for these programs, EPA and NHTSA (collectively, the agencies, or “we”) have worked in close partnership not only with one another, but also with the vehicle manufacturing industry, environmental community leaders, and the State of California among other entities to create a single, effective set of national standards.

Through two previous rulemakings, EPA and NHTSA have worked with the auto industry to develop new fuel economy and GHG emission standards for light-duty vehicles. Taken together with NHTSA's 2011 CAFE standards, the light-duty vehicle standards span model years 2011 to 2025 and are the first significant improvement in fuel economy in approximately two decades. Under the final program, average new car and light truck fuel economy is expected to nearly double by 2025 Start Printed Page 73480compared to 2010 vehicles.[7] In the 2012 rule, the agencies projected the standards would save consumers $1.7 trillion at the pump—roughly $8,200 per vehicle for a MY 2025 vehicle—reducing oil consumption by 2.2 million barrels a day in 2025 and slashing GHG emissions by 6 billion metric tons over the lifetime of the vehicles sold during this period.[8] These fuel economy standards are already delivering savings for American drivers. Between model years 2008 and 2013, the unadjusted average test fuel economy of new passenger cars and light trucks sold in the United States has increased by about four miles per gallon. Altogether, light-duty vehicle fuel economy standards finalized after 2008 have already saved nearly one billion gallons of fuel and avoided more than 10 million tons of carbon dioxide emissions.[9]

Similarly, EPA and NHTSA have previously developed joint GHG emission and fuel efficiency standards for MDVs and HDVs. Prior to these Phase 1 standards, heavy-duty trucks and buses—from delivery vans to the largest tractor-trailers—were required to meet pollution standards for soot and smog-causing air pollutants, but no requirements existed for the fuel efficiency or carbon pollution from these vehicles.[10] By 2010, total fuel consumption and GHG emissions from MDVs and HDVs had been growing, and these vehicles accounted for 23 percent of total U.S. transportation-related GHG emissions [11] and about 20 percent of U.S. transportation-related energy use. In August 2011, the agencies finalized the groundbreaking Phase 1 standards for new MDVs and HDVs in model years 2014 through 2018. This program, developed with support from the trucking and engine industries, the State of California, Environment and Climate Change Canada, and leaders from the environmental community, set standards based on the use of off-the-shelf technologies. These standards are expected to save a projected 530 million barrels of oil and reduce carbon emissions by about 270 million metric tons, representing one of the most significant programs available to reduce domestic fuel consumption and emissions of GHGs.[12] The Phase 1 program, as well as the many additional actions called for in the President's 2013 Climate Action Plan [13] including this Phase 2 rulemaking, not only result in meaningful decreases in GHG emissions and fuel consumption, but also support—indeed are critical for—United States leadership to encourage other countries to also achieve meaningful GHG reductions and fuel conservation.

This rule builds on our commitment to robust collaboration with stakeholders and the public. It follows an expansive and thorough outreach effort in which the agencies gathered input, data and views from many interested stakeholders, involving over 400 meetings with heavy-duty vehicle and engine manufacturers, technology suppliers, trucking fleets, truck drivers, dealerships, environmental organizations, and state agencies.[14] As with the previous light-duty rules and the heavy-duty Phase 1 rule, the agencies have consulted frequently with the California Air Resources Board (CARB) staff during the development of this rule, given California's unique ability among the states to adopt their own GHG standards for on-highway engines and vehicles. Through this close coordination, the agencies are finalizing a Phase 2 program that will be fully aligned between EPA and NHTSA, while providing CARB with the opportunity to adopt a Phase 2 program that will allow manufacturers to continue to build a single fleet of vehicles and engines.

(2) Overview of Phase 1 Medium- and Heavy-Duty Vehicle Standards

The Phase 1 program covers new trucks and heavy vehicles in model years 2014 and later. That program includes specific standards for combination tractors, heavy-duty pickup trucks and vans, and vocational vehicles and includes separate standards for both vehicles and engines. The program offers extensive flexibility, allowing manufacturers to reach standards through average fleet calculations, a mix of technologies, and the use of various credit and banking programs.

The Phase 1 program was developed by the agencies through close consultation with industry and other stakeholders, resulting in standards tailored to the specifics of each different class of vehicles and engines.

  • Heavy-duty combination tractors. Combination tractors—semi trucks that typically pull trailers—are regulated under nine subcategories based on weight class, cab type, and roof height. These vehicles represent approximately 60 percent of the fuel consumption and GHG emissions from MDVs and HDVs.
  • Heavy-duty pickup trucks and vans. Heavy-duty pickup and van standards are based on a “work factor” attribute that combines a vehicle's payload, towing capabilities, and the presence of 4-wheel drive. These vehicles represent about 23 percent of the fuel consumption and GHG emissions from MDVs and HDVs.
  • Vocational vehicles. Specialized vocational vehicles, which consist of a very wide variety of truck and bus types (e.g., delivery, refuse, utility, dump, cement, transit bus, shuttle bus, school bus, emergency vehicles, and recreational vehicles) are regulated in three subcategories based on engine classification. These vehicles represent approximately 17 percent of the fuel consumption and GHG emissions from MDVs and HDVs. The Phase 1 program includes EPA GHG standards for recreational vehicles, but not NHTSA fuel efficiency standards.[15]
  • Heavy-duty engines. The Phase 1 rule has independent standards for heavy-duty engines to assure they contribute to reducing GHG emissions and fuel consumption because the Phase 1 tractor and vocational vehicle standards do not account for the contributions of engine improvements to reducing fuel consumption and GHG emissions.

The Phase 1 standards were premised on utilization of technologies that were already in production on some vehicles at the time of the Phase 1 FRM and are adaptable to the broader fleet. The Phase 1 program provides flexibilities that facilitate compliance. These flexibilities help provide sufficient lead time for manufacturers to make necessary technological improvements and reduce the overall cost of the program, without compromising overall environmental and fuel consumption objectives. The primary flexibility provisions are an engine averaging, banking, and trading (ABT) program and a vehicle ABT program. These ABT programs allow for emission and/or fuel consumption credits to be averaged, banked, or traded within each of the averaging sets.

The Phase 1 program was projected to save 530 million barrels of oil and avoid 270 million metric tons of GHG emissions.[16] At the same time, the Start Printed Page 73481program was projected to produce $50 billion in fuel savings and $49 billion of net societal benefits. Today, the Phase 1 fuel efficiency and GHG reduction standards are already reducing GHG emissions and U.S. oil consumption, and producing fuel savings for America's trucking industry. The market appears to be very accepting of the Phase 1 technologies.

(3) Overview of Phase 2 Medium- and Heavy-Duty Vehicle Standards

The Phase 2 GHG and fuel efficiency standards for MDVs and HDVs are a critical next step in improving fuel efficiency and reducing GHG emissions. The Phase 2 national program carries forward our commitment to meaningful collaboration with stakeholders and the public, as they build on more than 400 meetings with manufacturers, suppliers, trucking fleets, dealerships, state air quality agencies, non-governmental organizations (NGOs), and other stakeholders; over 200,000 public comments; and two public hearings to identify and understand the opportunities and challenges involved with this next level of fuel-saving technology. These meetings and public feedback, in addition to close coordination with CARB, have been invaluable to the agencies, enabling the development of a program that appropriately balances all potential impacts, effectively minimizes the possibility of unintended consequences, and allows manufacturers to continue to build a single fleet of vehicles and engines.

Phase 2 will include technology-advancing standards that will phase in over the long-term (through model year 2027) to result in an ambitious, yet achievable program that will allow manufacturers to meet standards through a mix of different technologies at reasonable cost. The terminal requirements go into effect in 2027, and would apply to MY 2027 and subsequent model year vehicles, unless modified by future rulemaking. The Phase 2 standards will maintain the underlying regulatory structure developed in the Phase 1 program, such as the general categorization of MDVs and HDVs and the separate standards for vehicles and engines. However, the Phase 2 program will build on and advance Phase 1 in a number of important ways including the following: basing standards not only on currently available technologies but also on utilization of technologies now under development or not yet widely deployed while providing significant lead time to assure adequate time to develop, test, and phase in these controls; developing first-time GHG and fuel efficiency standards for trailers; further encouraging innovation and providing flexibility; including vehicles produced by small business manufacturers with appropriate flexibilities for these companies; incorporating enhanced test procedures that (among other things) allow individual drivetrain and powertrain performance to be reflected in the vehicle certification process; and using an expanded and improved compliance simulation model.

The Phase 2 program will provide significant GHG reductions and save fuel by:

  • Strengthening standards to account for ongoing technological advancements. Relative to the baseline as of the end of Phase 1, these final standards are projected to achieve vehicle fuel savings as high as 25 percent, depending on the vehicle category. While costs are higher than for Phase 1, benefits greatly exceed costs, and payback periods are short, meaning that consumers will see substantial net savings over the vehicle lifetime. Payback is estimated at about two years for tractors and trailers, about four years for vocational vehicles, and about three years for heavy-duty pickups and vans. The agencies are finalizing a program that phases in the MY 2027 standards with interim standards for model years 2021 and 2024 (and for certain types of trailers, EPA is finalizing model year 2018 phase-in standards as well). The final program includes both significant strengthening of certain standards from the NPRM as well as adjustments to better align other standards with new data, analysis, and stakeholder and public feedback received since the time of the proposal.
  • Setting standards for trailers for the first time. In addition to retaining the vehicle and engine categories covered in the Phase 1 program, the Phase 2 standards include fuel efficiency and GHG emission standards for trailers used in combination with tractors. Although the agencies are not finalizing standards for all trailer types, the majority of new trailers will be covered.
  • Encouraging technological innovation while providing flexibility and options for manufacturers. For each category of HDVs, the standards will set performance targets that allow manufacturers to achieve reductions through a mix of different technologies and generally leave manufacturers free to choose any means of compliance. For tractor standards, for example, different combinations of improvements like advanced aerodynamics, engine improvements and waste-heat recovery, automated transmission, lower rolling resistance tires, and automatic tire inflation can be used to meet standards. For tractors and vocational vehicles, enhanced test procedures and an expanded and improved compliance simulation model enable the vehicle standards to encompass more of the complete vehicle than the Phase 1 program and to account for engine, transmission and driveline improvements. With the addition of the powertrain and driveline to the compliance model, representative drive cycles and vehicle baseline configurations become critically important to assure the standards promote technologies that improve real world fuel efficiency and GHG emissions. This rule updates drive cycles and vehicle configurations to better reflect real world operation. The final program includes adjustments to technical elements of the proposed compliance program, e.g., test procedures, reflecting the significant amount of stakeholder and public comment the agencies received on the program. Additionally, the agencies' analyses indicate that this rule should have no adverse impact on vehicle or engine safety.
  • Providing flexibilities to help minimize effect on small businesses. All small businesses are exempt from the Phase 1 standards. The agencies are regulating small business entities under Phase 2 (notably certain trailer manufacturers), but we have conducted extensive proceedings pursuant to section 609 of the Regulatory Flexibility Act, and engaged in extensive consultation with stakeholders, and developed an approach to provide targeted flexibilities geared toward helping small businesses comply with the Phase 2 standards. Specifically, the agencies are delaying the initial implementation of the Phase 2 standards by one year and simplifying certification requirements for small businesses. We are also adopting additional flexibilities and exemptions adapted to particular vehicle categories.

The following tables summarize the impacts of the Heavy-Duty Phase 2 rule.Start Printed Page 73482

Summary of the Phase 2 Medium- and Heavy-Duty Vehicle Rule Impacts to Fuel Consumption, GHG Emissions, Benefits and Costs Over the Lifetime of Model Years 2018-2029 ab

3%7%
Fuel Reductions (billion gallons)71-82
GHG Reductions (MMT, CO2eq)959-1098
Pre-Tax Fuel Savings ($billion)149-16980-87
Discounted Technology Costs ($billion)24-2716-18
Value of reduced emissions ($billion)60-6948-52
Total Costs ($billion)29-3119-20
Total Benefits ($billion)225-260136-151
Net Benefits ($billion)197-229117-131
Notes:
a Ranges reflect two analysis methods: Method A with the 1b baseline and Method B with the la baseline. For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the “flat” baseline, 1a, and the “dynamic” baseline, 1b, please see Section X.A.1.
b Benefits and net benefits (including those in the 7% discount rate column) use the 3 percent average Social Cost of CO2, the Social Cost of CH4, and the Social Cost of N2O.

Summary of the Phase 2 Medium- and Heavy-Duty Vehicle Annual Fuel and GHG Reductions, Program Costs, Benefits and Net Benefits in Calendar Years 2040 and 2050 a

20402050
Fuel Reductions (Billion Gallons)10.813.0
GHG Reduction (MMT, CO2eq)166.8199.3
Vehicle Program Costs (including Maintenance; Billions of 2013$)−$6.5−$7.5
Fuel Savings (Pre-Tax; Billions of 2013$)$53.1$63.4
Benefits (Billions of 2013$)$24.8$31.7
Net Benefits (Billions of 2013$)$71.4$87.6
Note:
a Benefits and net benefits (including those in the 7% discount rate column) use the 3 percent average Social Cost of CO2, the Social Cost of CH4, and the Social Cost of N2O. Values reflect the final program using Method B relative to the flat baseline (a reference case that projects very little improvement in new vehicle fuel economy absent new standards).

Summary of the Phase 2 Medium- and Heavy-Duty Vehicle Program Expected Per-Vehicle Fuel Savings, GHG Emission Reductions, and Cost for Key Vehicle Categories

MY 2021MY 2024MY 2027
Maximum Vehicle Fuel Savings and Tailpipe GHG Reduction (%):
Tractors b132025
Trailers a579
Vocational Vehicles b122024
Pickups/Vans2.51016
Per Vehicle Cost ($)cd (% Increase in Typical Vehicle Price):
Tractors$6,400-$6,480 (6%)$9,920-$10,100 (10%)$12,160-$12,440 (12%)
Trailers$850-$870 (3%)$1,000-$1,030 (4%)$1,070-$1,110 (4%)
Vocational Vehicles$1,110-$1,160 (1%)$1,980-$2,020 (2%)$2,660-$2,700 (3%)
Pickups/Vans$520-$750 (1%)$760-$960 (2%)$1,340-$1,360 (3%)
Notes:
a Note that the EPA standards for trailers begin in model year 2018
b All engine costs are included
c Please refer to Preamble Chapters 6 and 10 for additional information on the reference fleet used to analyze costs and benefits of the rule. Please also refer to these chapters for impacts of the rule under more dynamic baseline assumptions for pickups and vans.
d Ranges reflect two analysis methods: Method A with the 1b baseline and Method B with the la baseline. For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the “flat” baseline, 1a, and the “dynamic” baseline, 1b, please see Section X.A.1.
e For this table, we use an approximate minimum vehicle price today of $100,000 for tractors, $25,000 for trailers, $100,000 for vocational vehicles and $40,000 for HD pickups/vans.
Start Printed Page 73483

Payback Periods for MY 2027 Vehicles Under the Final Standards, Based on both Analysis Methods A and B

[Payback occurs in the year shown; using 7% discounting]

Final standards
Tractors/Trailers2nd.
Vocational Vehicles4th.
Pickups/Vans a3rd.
Note:
a Please refer to Preamble Chapters 6 and 10 for additional information on the reference fleet used to analyze costs and benefits of the rule. Please also refer to these chapters for impacts of the rule under more dynamic baseline assumptions for pickups and vans.

(4) Issues Addressed in This Final Rule

This Preamble contains extensive discussion of the background, elements, and implications of the Phase 2 program, as well as updates made to the final program from the proposal based on new data, analysis, stakeholder feedback and public comments. Section I includes information on the MDV and HDV industry, related regulatory and non-regulatory programs, summaries of Phase 1 and Phase 2 programs, costs and benefits of the final standards, and relevant statutory authority for EPA and NHTSA. Section II discusses vehicle simulation, engine standards, and test procedures. Sections III, IV, V, and VI detail the final standards for combination tractors, trailers, vocational vehicles, and heavy-duty pickup trucks and vans. Sections VII and VIII discuss aggregate GHG impacts, fuel consumption impacts, climate impacts, and impacts on non-GHG emissions. Section IX evaluates the economic impacts of the final program. Sections X and XI present the alternatives analyses and consideration of natural gas vehicles. Finally, Sections XII and XIII discuss the changes that the Phase 2 rules will have on Phase 1 standards and other regulatory provisions. In addition to this Preamble, the Regulatory Impact Analysis (RIA),[17] provides additional data, analysis and discussion of the standards, and the Response to Comments Document for Joint Rulemaking (RTC) provides responses to comments received on the Phase 2 rulemaking through the public comment process.[18]

Table of Contents

A. Does this action apply to me?

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

C. Executive Summary

I. Overview

A. Background

B. Summary of Phase 1 Program

C. Summary of the Phase 2 Standards and Requirements

D. Summary of the Costs and Benefits of the Final Rules

E. EPA and NHTSA Statutory Authorities

F. Other Issues

II. Vehicle Simulation and Separate Engine Standards for Tractors and Vocational Chassis

A. Introduction

B. Phase 2 Regulatory Structure

C. Phase 2 GEM and Vehicle Component Test Procedures

D. Engine Test Procedures and Engine Standards

III. Class 7 and 8 Combination Tractors

A. Summary of the Phase 1 Tractor Program

B. Overview of the Phase 2 Tractor Program and Key Changes From the Proposal

C. Phase 2 Tractor Standards

D. Feasibility of the Final Phase 2 Tractor Standards

E. Phase 2 Compliance Provisions for Tractors

F. Flexibility Provisions

IV. Trailers

A. The Trailer Industry

B. Overview of the Phase 2 Trailer Program and Key Changes From the Proposal

C. Phase 2 Trailer Standards

D. Feasibility of the Trailer Standards

E. Trailer Standards: Compliance and Flexibilities

V. Class 2b-8 Vocational Vehicles

A. Summary of Phase 1 Vocational Vehicle Standards

B. Phase 2 Standards for Vocational Vehicles

C. Feasibility of the Vocational Vehicle Standards

D. Compliance Provisions for Vocational Vehicles

VI. Heavy-Duty Pickups and Vans

A. Summary of Phase 1 HD Pickup and Van Standards

B. HD Pickup and Van Final Phase 2 Standards

C. Use of the CAFE Model in Heavy-Duty Rulemaking

D. NHTSA CAFE Model Analysis of the Regulatory Alternatives for HD Pickups and Vans: Method A

E. Analysis of the Regulatory Alternatives for HD Pickups and Vans: Method B

F. Compliance and Flexibility for HD Pickup and Van Standards

VII. Aggregate GHG, Fuel Consumption, and Climate Impacts

A. What methodologies did the agencies use to project GHG emissions and fuel consumption impacts?

B. Analysis of Fuel Consumption and GHG Emissions Impacts Resulting From Final Standards

C. What are the projected reductions in fuel consumption and GHG emissions?

D. Climate Impacts and Indicators

VIII. How will these rules impact non-GHG emissions and their associated effects?

A. Health Effects of Non-GHG Pollutants

B. Environmental Effects of Non-GHG Pollutants

C. Emissions Inventory Impacts

D. Air Quality Impacts of Non-GHG Pollutants

IX. Economic and Other Impacts

A. Conceptual Framework

B. Vehicle-Related Costs Associated With the Program

C. Changes in Fuel Consumption and Expenditures

D. Maintenance Expenditures

E. Analysis of the Rebound Effect

F. Impact on Class Shifting, Fleet Turnover, and Sales

G. Monetized GHG Impacts

H. Monetized Non-GHG Health Impacts

I. Energy Security Impacts

J. Other Impacts

K. Summary of Benefits and Costs

L. Employment Impacts

M. Cost of Ownership and Payback Analysis

N. Safety Impacts

X. Analysis of the Alternatives

A. What are the alternatives that the agencies considered?

B. How do these alternatives compare in overall fuel consumption and GHG emissions reductions?

XI. Natural Gas Vehicles and Engines

A. Natural Gas Engine and Vehicle Technology

B. GHG Lifecycle Analysis for Natural Gas Vehicles

C. Projected Use of LNG and CNG

D. Natural Gas Emission Control Measures

E. Dimethyl Ether

XII. Amendments to Phase 1 Standards

A. EPA Amendments

B. Other Compliance Provisions for NHTSA

XIII. Other Regulatory Provisions

A. Amendments Related to Heavy-Duty Highway Engines and Vehicles

B. Amendments Affecting Glider Vehicles and Glider Kits

C. Applying the General Compliance Provisions of 40 CFR Part 1068 to Light-Duty Vehicles, Light-Duty Trucks, Chassis-Certified Class 2b and 3 Heavy-Duty Vehicles and Highway Motorcycles

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

E. Amendments to Light-Duty Greenhouse Gas Program Requirements

F. Amendments to Highway and Nonroad Test Procedures and Certification Requirements

G. Amendments Related to Locomotives in 40 CFR Part 1033

H. Amendments Related to Nonroad Diesel Engines in 40 CFR Part 1039

I. Amendments Related to Marine Diesel Engines in 40 CFR Parts 1042 and 1043

J. Miscellaneous EPA Amendments

K. Competition Vehicles

L. Amending 49 CFR Parts 512 and 537 To Allow Electronic Submissions and Defining Data Formats for Light-Duty Vehicle Corporate Average Fuel Economy (CAFE) Reports

XIV. Statutory and Executive Order Reviews

A. Executive Order 12866: Regulatory Planning and Review and Executive Start Printed Page 73484Order 13563: Improving Regulation and Regulatory Review

B. National Environmental Policy Act

C. Paperwork Reduction Act

D. Regulatory Flexibility Act

E. Unfunded Mandates Reform Act

F. Executive Order 13132: Federalism

G. Executive Order 13175: Consultation and Coordination With Indian Tribal Governments

H. Executive Order 13045: Protection of Children From Environmental Health Risks and Safety Risks

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

J. National Technology Transfer and Advancement Act and 1 CFR Part 51

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

L. Endangered Species Act (ESA)

M. Congressional Review Act (CRA)

XV. EPA and NHTSA Statutory Authorities

A. EPA

B. NHTSA

List of Subjects

I. Overview

The agencies issued a Notice of Proposed Rulemaking (NPRM) on July 13, 2015, that proposed Phase 2 GHG and fuel efficiency standards for heavy-duty engines and vehicles.[19] The agencies also issued a Notice of Data Availability (NODA) on March 2, 2016, to solicit comment on new material not available at the time of the NPRM.[20] The agencies have revised the proposed standards and related requirements to address issues raised in public comments. Nevertheless, the final rules being adopted today remain fundamentally similar to the proposed rules.

Although the agencies describe the final requirements in this document, readers are encouraged to also read supporting materials that have been place into the public dockets for these rules. In particular, the agencies note:

  • The Final Regulatory Impact Analysis (RIA), provides additional technical information and analysis
  • The Response to Comments Document for Joint Rulemaking (RTC), provides a detailed summary and analysis of public comments, including comments received in response to the NODA
  • The NHTSA Final Environmental Impact Statement (FEIS)

This overview of the final Phase 2 GHG emissions and fuel efficiency standards includes a description of the heavy-duty truck industry and related regulatory and non-regulatory programs, a summary of the Phase 1 GHG emissions and fuel efficiency program, a summary of the Phase 2 standards and requirements being finalized, a summary of the costs and benefits of the Phase 2 standards, discussion of EPA and NHTSA statutory authorities, and other issues.

A. Background

For purposes of this Preamble (and consistent with all terminology used at proposal), the terms “heavy-duty” or “HD” are used to apply to all highway vehicles and engines that are not within the range of light-duty passenger cars, light-duty trucks, and medium-duty passenger vehicles (MDPV) covered by separate GHG and Corporate Average Fuel Economy (CAFE) standards.[21] (The terms also do not include motorcycles). Thus, in this rulemaking, unless specified otherwise, the heavy-duty category incorporates all vehicles with a gross vehicle weight rating above 8,500 lbs, and the engines that power them, except for MDPVs.[22 23 24] Note also that the terms heavy-duty truck and heavy-duty vehicle are sometimes used interchangeably, even though commercially the term heavy-duty truck can have a narrower meaning.

Consistent with the President's direction, over the past three years as we have developed this rulemaking, the agencies have met on an on-going basis with a very large number of diverse stakeholders. This includes meetings, and in many cases site visits, with truck, trailer, and engine manufacturers; technology supplier companies and their trade associations (e.g., transmissions, drivelines, fuel systems, turbochargers, tires, catalysts, and many others); line haul and vocational trucking firms and trucking associations; the trucking industries owner-operator association; truck dealerships and dealers associations; trailer manufacturers and their trade association; non-governmental organizations (NGOs, including environmental NGOs, national security NGOs, and consumer advocacy NGOs); state air quality agencies; manufacturing labor unions; and many other stakeholders. In addition, EPA and NHTSA have consulted on an on-going basis with the California Air Resources Board (CARB) over the past three years as we developed the Phase 2 rule. CARB staff and managers have also participated with EPA and NHTSA in meetings with many external stakeholders, including those with vehicle OEMs and technology suppliers.[25]

EPA and NHTSA staff also participated in a large number of technical and policy conferences over the past three years related to the technological, economic, and environmental aspects of the heavy-duty trucking industry. The agencies also met with regulatory counterparts from several other nations who either have already or are considering establishing fuel consumption or GHG requirements, including outreach with representatives from the governments of Canada, the European Commission, Japan, and China.

These comprehensive outreach actions by the agencies provided us with information to assist in our identification of potential technologies that can be used to reduce heavy-duty GHG emissions and improve fuel efficiency. The outreach has also helped the agencies to identify and understand the opportunities and challenges involved with these standards for the heavy-duty trucks, trailers, and engines detailed in this Preamble, including time needed for implementation of various technologies and potential costs and fuel savings. The scope of this outreach effort to gather input for the proposal and final rulemaking included well over 400 meetings with stakeholders. These meetings and conferences have been invaluable to the agencies. We believe they enabled us to refine the proposal in such a way as to appropriately consider all of the potential impacts and to minimize the possibility of unintended consequences in the final rules.Start Printed Page 73485

(1) Brief Overview of the Heavy-Duty Truck Industry

The heavy-duty sector is diverse in several respects, including the types of manufacturing companies involved, the range of sizes of trucks and engines they produce, the types of work for which the trucks are designed, and the regulatory history of different subcategories of vehicles and engines. The current heavy-duty fleet encompasses vehicles from the “18-wheeler” combination tractor-trailers one sees on the highway to the largest pickup trucks and vans, as well as vocational vehicles covering the range between these extremes. Together, the HD sector spans a wide range of vehicles with often specialized form and function. A primary indicator of the diversity among heavy-duty trucks is the range of load-carrying capability across the industry. The heavy-duty truck sector is often subdivided by vehicle weight classifications, as defined by the vehicle's gross vehicle weight rating (GVWR), which is a measure of the combined curb (empty) weight and cargo carrying capacity of the truck.[26] Table I-1 below outlines the vehicle weight classifications commonly used for many years for a variety of purposes by businesses and by several Federal agencies, including the Department of Transportation, the Environmental Protection Agency, the Department of Commerce, and the Internal Revenue Service.

Table I-1—Vehicle Weight Classification

Class2b345678
GVWR (lb.)8,501-10,00010,001-14,00014,001-16,00016,001-19,50019,501-26,00026,001-33,000>33,000

In the framework of these vehicle weight classifications, the heavy-duty truck sector refers to “Class 2b” through “Class 8” vehicles and the engines that power those vehicles.[27]

Unlike light-duty vehicles, which are primarily used for transporting passengers for personal travel, heavy-duty vehicles fill much more diverse operator needs. Heavy-duty pickup trucks and vans (Classes 2b and 3) are used chiefly as work trucks and vans, and as shuttle vans, as well as for personal transportation, with an average annual mileage in the range of 15,000 miles. The rest of the heavy-duty sector is used for carrying cargo and/or performing specialized tasks. “Vocational” vehicles, which span Classes 2b through 8, vary widely in size, including smaller and larger van trucks, utility “bucket” trucks, tank trucks, refuse trucks, urban and over-the-road buses, fire trucks, flat-bed trucks, and dump trucks, among others. The annual mileage of these vehicles is as varied as their uses, but for the most part tends to fall in between heavy-duty pickups/vans and the large combination tractors, typically from 15,000 to 150,000 miles per year.

Class 7 and 8 combination tractor-trailers—some equipped with sleeper cabs and some not—are primarily used for freight transportation. They are sold as tractors and operate with one or more trailers that can carry up to 50,000 lbs or more of payload, consuming significant quantities of fuel and producing significant amounts of GHG emissions. Together, Class 7 and 8 tractors and trailers account for approximately 60 percent of the heavy-duty sector's total CO2 emissions and fuel consumption. Trailer designs vary significantly, reflecting the wide variety of cargo types. However, the most common types of trailers are box vans (dry and refrigerated), which are a focus of this Phase 2 rulemaking. The tractor-trailers used in combination applications can and frequently do travel more than 150,000 miles per year and can operate for 20-30 years.

Heavy-duty vehicles differ significantly from light-duty vehicles in other ways. In particular, we note that heavy-duty engines are much more likely to be rebuilt. In fact, it is common for Class 8 engines to be rebuilt multiple times. Commercial heavy-duty vehicles are often resold after a few years and may be repurposed by the second or third owner. Thus issues of resale value and adaptability have historically been key concerns for purchasers.

EPA and NHTSA have designed our respective standards in careful consideration of the diversity and complexity of the heavy-duty truck industry, as discussed in Section I.C.

(2) Related Regulatory and Non-Regulatory Programs

(a) History of EPA's Heavy-Duty Regulatory Program and Assessments of the Impacts of Greenhouse Gases on Climate Change

To provide a context for EPA's program to reduce greenhouse gas emissions from motor vehicles, this subsection provides an overview of two important related areas. First, we summarize the history of EPA's heavy-duty regulatory program, which provides a basis for the compliance structure of this rulemaking. Next we summarize EPA prior assessments of the impacts of greenhouse gases on climate change, which provides a basis for much of the analysis of the environmental benefits of this rulemaking.

(i) History of EPA's Heavy-Duty Regulatory Program

Since the 1980s, EPA has acted several times to address tailpipe emissions of criteria pollutants and air toxics from heavy-duty vehicles and engines. During the last two decades these programs have primarily addressed emissions of particulate matter (PM) and the primary ozone precursors, hydrocarbons (HC) and oxides of nitrogen (NOX). These programs, which have successfully achieved significant and cost-effective reductions in emissions and associated health and welfare benefits to the nation, were an important basis of the Phase 1 program. See e.g. 66 FR 5002, 5008, and 5011-5012 (January 18, 2001) (detailing substantial public health benefits of controls of criteria pollutants from heavy-duty diesel engines, including bringing areas into attainment with primary (public health) PM NAAQS, or contributing substantially to such attainment); National Petrochemical Refiners Association v. EPA, 287 F. 3d 1130, 1134 (D.C. Cir. 2002) (referring to the “dramatic reductions” in criteria pollutant emissions resulting from the EPA on-Start Printed Page 73486highway heavy-duty engine standards, and upholding all of the standards).

As required by the Clean Air Act (CAA), the emission standards implemented by these programs include standards that apply at the time that the vehicle or engine is sold and continue to apply in actual use. EPA's overall program goal has always been to achieve emissions reductions from the complete vehicles that operate on our roads. The agency has often accomplished this goal for many heavy-duty truck categories by regulating heavy-duty engine emissions. A key part of this success has been the development over many years of a well-established, representative, and robust set of engine test procedures that industry and EPA now use routinely to measure emissions and determine compliance with emission standards. These test procedures in turn serve the overall compliance program that EPA implements to help ensure that emissions reductions are being achieved. By isolating the engine from the many variables involved when the engine is installed and operated in a HD vehicle, EPA has been able to accurately address the contribution of the engine alone to overall emissions.

(ii) EPA Assessment of the Impacts of Greenhouse Gases on Climate Change

In 2009, the EPA Administrator issued the document known as the Endangerment Finding under CAA section 202(a)(1).[28] In the Endangerment Finding, which focused on public health and public welfare impacts within the United States, the Administrator found that elevated concentrations of GHG emissions in the atmosphere may reasonably be anticipated to endanger public health and welfare of current and future generations. See also Coalition for Responsible Regulation v. EPA, 684 F. 3d 102, 117-123 (D.C. Cir. 2012) (upholding the endangerment finding in all respects). The following sections summarize the key information included in the Endangerment Finding.

Climate change caused by human emissions of GHGs threatens public health in multiple ways. By raising average temperatures, climate change increases the likelihood of heat waves, which are associated with increased deaths and illnesses. While climate change also decreases the likelihood of cold-related mortality, evidence indicates that the increases in heat mortality will be larger than the decreases in cold mortality in the United States. Compared to a future without climate change, climate change is expected to increase ozone pollution over broad areas of the U.S., including in the largest metropolitan areas with the worst ozone problems, and thereby increase the risk of morbidity and mortality. Other public health threats also stem from projected increases in intensity or frequency of extreme weather associated with climate change, such as increased hurricane intensity, increased frequency of intense storms and heavy precipitation. Increased coastal storms and storm surges due to rising sea levels are expected to cause increased drownings and other adverse health impacts. Children, the elderly, and the poor are among the most vulnerable to these climate-related health effects. See also 79 FR 75242 (December 17, 2014) (climate change, and temperature increases in particular, likely to increase O3 (ozone) pollution “over broad areas of the U.S., including the largest metropolitan areas with the worst O3 problems, increas[ing] the risk of morbidity and mortality”).

Climate change caused by human emissions of GHGs also threatens public welfare in multiple ways. Climate changes are expected to place large areas of the country at serious risk of reduced water supplies, increased water pollution, and increased occurrence of extreme events such as floods and droughts. Coastal areas are expected to face increased risks from storm and flooding damage to property, as well as adverse impacts from rising sea level, such as land loss due to inundation, erosion, wetland submergence and habitat loss. Climate change is expected to result in an increase in peak electricity demand, and extreme weather from climate change threatens energy, transportation, and water resource infrastructure. Climate change may exacerbate ongoing environmental pressures in certain settlements, particularly in Alaskan indigenous communities. Climate change also is very likely to fundamentally rearrange U.S. ecosystems over the 21st century. Though some benefits may balance adverse effects on agriculture and forestry in the next few decades, the body of evidence points towards increasing risks of net adverse impacts on U.S. food production, agriculture and forest productivity as temperature continues to rise. These impacts are global and may exacerbate problems outside the U.S. that raise humanitarian, trade, and national security issues for the U.S. See also 79 FR 75382 (December 17, 2014) (welfare effects of O3 increases due to climate change, with emphasis on increased wildfires).

As outlined in Section VIII.A of the 2009 Endangerment Finding, EPA's approach to providing the technical and scientific information to inform the Administrator's judgment regarding the question of whether GHGs endanger public health and welfare was to rely primarily upon the recent, major assessments by the U.S. Global Change Research Program (USGCRP), the Intergovernmental Panel on Climate Change (IPCC), and the National Research Council (NRC) of the National Academies. These assessments addressed the scientific issues that EPA was required to examine, were comprehensive in their coverage of the GHG and climate change issues, and underwent rigorous and exacting peer review by the expert community, as well as rigorous levels of U.S. government review. Since the administrative record concerning the Endangerment Finding closed following EPA's 2010 Reconsideration Denial, a number of new major, peer-reviewed scientific assessments have been released. These include the IPCC's 2012 “Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation” (SREX) and the 2013-2014 Fifth Assessment Report (AR5), the USGCRP's 2014 “Climate Change Impacts in the United States” (Climate Change Impacts), and the NRC's 2010 “Ocean Acidification: A National Strategy to Meet the Challenges of a Changing Ocean” (Ocean Acidification), 2011 “Report on Climate Stabilization Targets: Emissions, Concentrations, and Impacts over Decades to Millennia” (Climate Stabilization Targets), 2011 “National Security Implications for U.S. Naval Forces” (National Security Implications), 2011 “Understanding Earth's Deep Past: Lessons for Our Climate Future” (Understanding Earth's Deep Past), 2012 “Sea Level Rise for the Coasts of California, Oregon, and Washington: Past, Present, and Future,” 2012 “Climate and Social Stress: Implications for Security Analysis” (Climate and Social Stress), and 2013 “Abrupt Impacts of Climate Change” (Abrupt Impacts) assessments.

EPA has reviewed these new assessments and finds that the improved understanding of the climate system they present further strengthens the case that GHG emissions endanger public health and welfare.

In addition, these assessments highlight the urgency of the situation as the concentration of CO2 in the atmosphere continues to rise. Absent a reduction in emissions, a recent Start Printed Page 73487National Research Council assessment projected that concentrations by the end of the century would increase to levels that the Earth has not experienced for millions of years.[29] In fact, that assessment stated that “the magnitude and rate of the present greenhouse gas increase place the climate system in what could be one of the most severe increases in radiative forcing of the global climate system in Earth history.” [30] What this means, as stated in another NRC assessment, is that:

Emissions of carbon dioxide from the burning of fossil fuels have ushered in a new epoch where human activities will largely determine the evolution of Earth's climate. Because carbon dioxide in the atmosphere is long lived, it can effectively lock Earth and future generations into a range of impacts, some of which could become very severe. Therefore, emission reductions choices made today matter in determining impacts experienced not just over the next few decades, but in the coming centuries and millennia.[31]

Moreover, due to the time-lags inherent in the Earth's climate, the Climate Stabilization Targets assessment notes that the full warming from any given concentration of CO2 reached will not be realized for several centuries.

The most recent USGCRP “National Climate Assessment” [32] emphasizes that climate change is already happening now and is happening in the United States. The assessment documents the increases in some extreme weather and climate events in recent decades, as well as the resulting damage and disruption to infrastructure and agriculture, and projects continued increases in impacts across a wide range of peoples, sectors, and ecosystems.

These assessments underscore the urgency of reducing emissions now. Today's emissions will otherwise lead to raised atmospheric concentrations for thousands of years, and raised Earth system temperatures for even longer. Emission reductions today will benefit the public health and public welfare of current and future generations.

Finally, it should be noted that the concentration of carbon dioxide in the atmosphere continues to rise dramatically. In 2009, the year of the Endangerment Finding, the average concentration of carbon dioxide as measured on top of Mauna Loa was 387 parts per million.[33] The average concentration in 2015 was 401 parts per million, the first time an annual average has exceeded 400 parts per million since record keeping began at Mauna Loa in 1958, and for at least the past 800,000 years according to ice core records.[34] Moreover, 2015 was the warmest year globally in the modern global surface temperature record, going back to 1880, breaking the record previously held by 2014; this now means that the last 15 years have been 15 of the 16 warmest years on record.[35]

(b) The EPA and NHTSA Light-Duty National GHG and Fuel Economy Program

On May 7, 2010, EPA and NHTSA finalized the first-ever National Program for light-duty cars and trucks, which set GHG emissions and fuel economy standards for model years 2012-2016 (see 75 FR 25324). More recently, the agencies adopted even stricter standards for model years 2017 and later (77 FR 62624, October 15, 2012). The agencies have used the light-duty National Program as a model for the HD National Program in several respects. This is most apparent in the case of heavy-duty pickups and vans, which are similar to the light-duty trucks addressed in the light-duty National Program both technologically as well as in terms of how they are manufactured (i.e., the same company often makes both the vehicle and the engine, and several light-duty manufacturers also manufacture HD pickups and vans).[36] For HD pickups and vans, there are close parallels to the light-duty program in how the agencies have developed our respective heavy-duty standards and compliance structures. However, HD pickups and vans are true work vehicles that are designed for much higher towing and payload capabilities than are light-duty pickups and vans. The technologies applied to light-duty trucks are not all applicable to heavy-duty pickups and vans at the same adoption rates, and the technologies often produce a lower percent reduction in CO2 emissions and fuel consumption when used in heavy-duty vehicles. Another difference between the light-duty and the heavy-duty standards is that each agency adopts heavy-duty standards based on attributes other than vehicle footprint, as discussed below.

Due to the diversity of the remaining HD vehicles, there are fewer parallels with the structure of the light-duty program. However, the agencies have maintained the same collaboration and coordination that characterized the development of the light-duty program throughout the Phase 1 rulemaking and the continued efforts for Phase 2. Most notably, as with the light-duty program, manufacturers will continue to be able to design and build vehicles to meet a closely coordinated, harmonized national program, and to avoid unnecessarily duplicative testing and compliance burdens. In addition, the averaging, banking, and trading provisions in the HD program, although structurally different from those of the light-duty program, serve the same purpose, which is to allow manufacturers to achieve large reductions in fuel consumption and emissions while providing a broad mix of products to their customers. The agencies have also worked closely with CARB to provide harmonized national standards.

(c) EPA's SmartWay Program

EPA's voluntary SmartWay Transport Partnership program encourages businesses to take actions that reduce fuel consumption and CO2 emissions while cutting costs by working with the shipping, logistics, and carrier communities to identify low carbon strategies and technologies across their transportation supply chains. SmartWay provides technical information, benchmarking and tracking tools, market incentives, and partner recognition to facilitate and accelerate the adoption of these strategies. Through the SmartWay program and its related technology assessment center, EPA has worked closely with truck and trailer manufacturers and truck fleets over the past 12 years to develop test procedures to evaluate vehicle and component performance in reducing fuel consumption and has conducted testing and has established test programs to verify technologies that can achieve these reductions. SmartWay partners have demonstrated these new and emerging technologies in their business operations, adding to the body of technical data and information that EPA can disseminate to industry, researchers and other stakeholders. Over the last several years, EPA has developed hands-on experience testing the largest heavy-duty trucks and trailers and evaluating improvements in tire and vehicle aerodynamic performance. In developing the Phase 1 Start Printed Page 73488program, the agencies drew from this testing and from the SmartWay experience. In the same way, the agencies benefitted from SmartWay in developing the Phase 2 trailer program.

(d) DOE's SuperTruck Initiative

The U.S. Department of Energy launched its SuperTruck I initiative in 2009. SuperTruck I was a DOE partnership with four industry teams, who at this point have either met the SuperTruck I 50 percent fuel efficiency improvement goal (relative to a 2009 best-in-class truck) or have laid the groundwork to succeed. Teams from Cummins/Peterbilt, Daimler, and Volvo exceeded the 50 percent efficiency improvement goal, with Navistar on track to exceed this target later this year. Research vehicles developed under SuperTruck I are Class 8 combination tractor-trailers that have dramatically increased fuel and freight efficiency through the use of advanced technologies. These technologies include tractor and trailer aerodynamic devices, engine waste heat recovery systems, hybrids, automated transmissions and lightweight materials. In March 2016 DOE announced SuperTruck II, which is an $80M follow-on to SuperTruck I, where DOE will continue to partner with industry teams to collaboratively fund new projects to research, develop, and demonstrate technologies to further improve heavy-truck freight efficiency—by more than 100 percent, relative to a manufacturer's best-in-class 2009 truck. Achieving these kinds of Class 8 truck efficiency increases will require an integrated systems approach to ensure that the various components of the vehicle work well together. SuperTruck II projects will utilize a wide variety of truck and trailer technology approaches to achieve performance targets, such as further improvements in engine efficiency, drivetrain efficiency, aerodynamic drag, tire rolling resistance, and vehicle weight.

The agencies leveraged the outcomes of SuperTruck I by projecting how these tractor and trailer technologies could continue to advance from this early developmental stage toward the prototype and production stages. For a number of the SuperTruck technologies, the agencies are projecting advancement into production, given appropriate lead time. For example, a number of the aerodynamic and transmission technologies are projected to be in widespread production by 2021, and the agencies are finalizing 2021 standards based in part on performance of these SuperTruck technologies. For other more advanced SuperTruck technologies, such as organic Rankine cycle waste heat recovery systems, the agencies are projecting that additional lead time is needed to ensure that these technologies will be effective and reliable in production. For these technologies, the agencies are finalizing 2027 standards whose stringency reflects a significant market adoption rate of advanced technologies, including waste heat recovery systems. Furthermore, the agencies are encouraged by DOE's announcement of SuperTruck II. We believe that the combination of HD Phase 2 and SuperTruck II will provide both a strong motivation and a proven means for manufacturers to fully develop these technologies within the lead times we have projected.

(e) The State of California

California has established ambitious goals for reducing GHG emissions from heavy-duty vehicles and engines as part of an overall plan to reduce GHG emissions from the transportation sector in California.[37] Heavy-duty vehicles are responsible for one-fifth of the total GHG emissions from transportation sources in California. In the past several years, the California Air Resources Board (CARB) has taken a number of actions to reduce GHG emissions from heavy-duty vehicles and engines. For example, in 2008, CARB adopted regulations to reduce GHG emissions from heavy-duty tractors that pull box-type trailers through improvements in tractor and trailer aerodynamics and the use of low rolling resistance tires.[38] The tractor-trailer operators subject to the CARB regulation are required to use SmartWay-certified tractors and trailers, or retrofit their existing fleet with SmartWay-verified technologies, consistent with California's state authority to regulate both new and in-use vehicles. In December 2013, CARB adopted regulations that establish its own parallel Phase 1 program with standards consistent with EPA Phase 1 standards. On December 5, 2014, California's Office of Administrative Law approved CARB's adoption of the Phase 1 standards, with an effective date of December 5, 2014.[39] Complementary to its regulatory efforts, CARB and other California agencies are investing significant public capital through various incentive programs to accelerate fleet turnover and stimulate technology innovation within the heavy-duty vehicle market (e.g., Air Quality Improvement, Carl Moyer, Loan Incentives, Lower-Emission School Bus and Goods Movement Emission Reduction Programs).[40] Recently, California Governor Jerry Brown established a target of up to 50 percent petroleum reduction by 2030.

California has long had the unique ability among states to adopt its own separate new motor vehicle standards per section 209 of the Clean Air Act (CAA). Although section 209(a) of the CAA expressly preempts states from adopting and enforcing standards relating to the control of emissions from new motor vehicles or new motor vehicle engines (such as state controls for new heavy-duty engines and vehicles), CAA section 209(b) directs EPA to waive this preemption under certain conditions. Under the waiver process set out in CAA section 209(b), EPA has granted CARB a waiver for its initial heavy-duty vehicle GHG regulation.[41] Even with California's ability under the CAA to establish its own emission standards, EPA and CARB have worked closely together over the past several decades to largely harmonize new vehicle criteria pollutant standard programs for heavy-duty engines and heavy-duty vehicles. In the past several years EPA and NHTSA also consulted with CARB in the development of the Federal light-duty vehicle GHG and CAFE rulemakings for the 2012-2016 and 2017-2025 model years.

As discussed above, California operates under state authority to establish its own new heavy-duty vehicle and engine emission standards, including standards for CO2, methane, N2 O, and hydrofluorocarbons. EPA recognizes this independent authority, and we also recognize the potential benefits for the regulated industry if the Federal Phase 2 standards could result Start Printed Page 73489in a single, National Program that would meet the EPA and NHTSA's statutory requirements to set appropriate and maximum feasible standards, and also be equivalent to potential future new heavy-duty vehicle and engine GHG standards established by CARB (addressing the same model years as addressed by the final Federal Phase 2 program and requiring the same technologies). In order to further the opportunity for maintaining coordinated Federal and California standards in the Phase 2 timeframe (as well as to benefit from different technical expertise and perspective), EPA and NHTSA consulted frequently with CARB while developing the Phase 2 rule. Prior to the proposal, the agencies' technical staff shared information on technology cost, technology effectiveness, and feasibility with the CARB staff. We also received information from CARB on these same topics. In addition, CARB staff and managers participated with EPA and NHTSA in meetings with many external stakeholders, in particular with vehicle OEMs and technology suppliers. The agencies continued significant consultation during the development of the final rules.

EPA and NHTSA believe that through this information sharing and dialog we have enhanced the potential for the Phase 2 program to result in a National Program that can be adopted not only by the Federal agencies, but also by the State of California, given the strong interest from the regulated industry for a harmonized State and Federal program. In its public comments, California reiterated its support for a harmonized State and Federal program, although it identified several areas in which it believed the proposed program needed to be strengthened.

(f) Environment and Climate Change Canada

On March 13, 2013, Environment and Climate Change Canada (ECCC), which is EPA's Canadian counterpart, published its own regulations to control GHG emissions from heavy-duty vehicles and engines, beginning with MY 2014. These regulations are closely aligned with EPA's Phase 1 program to achieve a common set of North American standards. ECCC has expressed its intention to amend these regulations to further limit emissions of greenhouse gases from new on-road heavy-duty vehicles and their engines for post-2018 MYs. As with the development of the current regulations, ECCC is committed to continuing to work closely with EPA to maintain a common Canada-United States approach to regulating GHG emissions for post-2018 MY vehicles and engines. This approach will build on the long history of regulatory alignment between the two countries on vehicle emissions pursuant to the Canada-United States Air Quality Agreement.[42] In furtherance of this coordination, EPA participated in a workshop hosted by ECCC on March 3, 2016 to discuss Canada's Phase 2 program.[43]

The Government of Canada, including ECCC and Transport Canada, has also been of great assistance during the development of this Phase 2 rule. In particular, the Government of Canada supported aerodynamic testing, and conducted chassis dynamometer emissions testing.

(g) Recommendations of the National Academy of Sciences

In April 2010, as mandated by Congress in the EISA, the National Research Council (NRC) under the National Academy of Sciences (NAS) issued a report to NHTSA and to Congress evaluating medium- and heavy-duty truck fuel efficiency improvement opportunities, titled “Technologies and Approaches to Reducing the Fuel Consumption of Medium- and Heavy-duty Vehicles.” That NAS report was far reaching in its review of the technologies that were available and that might become available in the future to reduce fuel consumption from medium- and heavy-duty vehicles. In presenting the full range of technical opportunities, the report included technologies that may not be available until 2020 or even further into the future. The report provided not only a valuable list of off-the-shelf technologies from which the agencies drew in developing the Phase 1 program, but also provided useful information the agencies have considered when developing this second phase of regulations.

In April 2014, the NAS issued another report: “Reducing the Fuel Consumption and Greenhouse Gas Emissions of Medium- and Heavy-Duty Vehicles, Phase Two, First Report.” [44] This study outlines a number of recommendations to the U.S. Department of Transportation and NHTSA on technical and policy matters to consider when addressing the fuel efficiency of our nation's medium- and heavy-duty vehicles. In particular, this report provided recommendations with respect to:

  • The Greenhouse Gas Emission Model (GEM) simulation tool used by the agencies to assess compliance with vehicle standards
  • Regulation of trailers
  • Natural gas-fueled engines and vehicles
  • Data collection on in-use operation

The agencies are adopting many of these recommendations into the Phase 2 program, including recommendations relating to the GEM simulation tool and to trailers.

B. Summary of Phase 1 Program

(1) EPA Phase 1 GHG Emission Standards and NHTSA Phase 1 Fuel Consumption Standards

The EPA Phase 1 mandatory GHG emission standards commenced in MY 2014 and include increased stringency for standards applicable to MY 2017 and later MY vehicles and engines. NHTSA's fuel consumption standards were voluntary for MYs 2014 and 2015, due to lead time requirements in EISA, and apply on a mandatory basis thereafter. They also increase in stringency for MY 2017. Both agencies allowed voluntary early compliance starting in MY 2013 and encouraged manufacturers' participation through credit incentives.

Given the complexity of the heavy-duty industry, the agencies divided the industry into three discrete categories for purposes of setting our respective Phase 1 standards—combination tractors, heavy-duty pickups and vans, and vocational vehicles—based on the relative degree of homogeneity among trucks within each category. The Phase 1 rules also include separate standards for the engines that power combination tractors and vocational vehicles. For each regulatory category, the agencies adopted related but distinct program approaches reflecting the specific challenges in these segments. In the following paragraphs, we briefly summarize EPA's Phase 1 GHG emission standards and NHTSA's Phase 1 fuel consumption standards for the three regulatory categories of heavy-duty vehicles and for the engines powering vocational vehicles and Start Printed Page 73490tractors. See Sections II, III, V, and VI for additional details on the Phase 1 standards. To respect differences in design and typical uses that drive different technology solutions, the agencies segmented each regulatory class into subcategories. The category-specific structure enabled the agencies to set standards that appropriately reflect the technology available for each regulatory subcategory of vehicles and the engines for use in each type of vehicle. The Phase 1 program also provided several flexibilities, as summarized in Section I.B.(3).

The agencies proposed and are adopting Phase 2 standards based on test procedures that differ from those used for Phase 1, including the revised GEM simulation tool. Significant revisions to GEM are discussed in Section II and in the RIA Chapter 4, and other test procedures are discussed further in the RIA Chapter 3. The pre-proposal revisions from Phase 1 GEM reflected input from both the NAS and from industry.[45] Changes since the proposal generally reflect comments received from industry and other key stakeholders. It is important to note that due to these test procedure changes, the Phase 1 and Phase 2 standards are not directly comparable in an absolute sense. In particular, the revisions being made to the 55 mph and 65 mph highway cruise cycles for tractors and vocational vehicles have the effect of making the cycles more challenging (albeit more representative of actual driving conditions). We are not applying these revisions to the Phase 1 program because doing so would significantly change the stringency of the Phase 1 standards, for which manufacturers have already developed engineering plans and are now producing products to meet. Moreover, the changes to GEM address a broader range of technologies not part of the projected compliance path for use in Phase 1.

Because the numeric values of the Phase 2 tractor and vocational standards are not directly comparable to their respective Phase 1 standards, the Phase 1 numeric standards were not appropriate baseline values to use to determine Phase 2's improvements. To address this situation, the agencies applied all of the new Phase 2 test procedures and GEM software to tractors and vocational vehicles equipped with Phase 1 compliant levels of technology. The agencies used the results of this approach to establish appropriate Phase 1 baseline values, which are directly comparable to the Phase 2 standards. For example, in this rulemaking we present Phase 2 per vehicle percent reductions versus Phase 1, and for tractors and vocational vehicles these percent reductions were all calculated versus Phase 1 compliant vehicles, where we applied the Phase 2 test procedures and GEM software to determine these Phase 1 vehicles' results.

(a) Class 7 and 8 Combination Tractors

Class 7 and 8 combination tractors and their engines contribute the largest portion of the total GHG emissions and fuel consumption of the heavy-duty sector, approximately 60 percent, due to their large payloads, their high annual miles traveled, and their major role in national freight transport. These vehicles consist of a cab and engine (tractor or combination tractor) and a detachable trailer. The primary manufacturers of combination tractors in the United States are Daimler Trucks North America, Navistar, Volvo/Mack, and PACCAR. Each of the tractor manufacturers and Cummins (an independent engine manufacturer) also produce heavy-duty engines used in tractors. The Phase 1 standards require manufacturers to reduce GHG emissions and fuel consumption for these tractors and engines, which we expect them to do through improvements in aerodynamics and tires, reductions in tractor weight, reduction in idle operation, as well as engine-based efficiency improvements.[46]

The Phase 1 tractor standards differ depending on gross vehicle weight rating (GVWR) (i.e., whether the truck is Class 7 or Class 8), the height of the roof of the cab, and whether it is a “day cab” or a “sleeper cab.” The agencies created nine subcategories within the Class 7 and 8 combination tractor category reflecting combinations of these attributes. The agencies set Phase 1 standards for each of these subcategories beginning in MY 2014, with more stringent standards following in MY 2017. The standards represent an overall fuel consumption and CO2 emissions reduction up to 23 percent from the tractors and the engines installed in them when compared to a baseline MY 2010 tractor and engine.

For Phase 1, tractor manufacturers demonstrate compliance with the tractor CO2 and fuel consumption standards using a vehicle simulation tool described in Section II. The tractor inputs to the simulation tool in Phase 1 are the aerodynamic performance, tire rolling resistance, vehicle speed limiter, automatic engine shutdown, and weight reduction.

In addition to the Phase 1 tractor-based standards for CO2, EPA adopted a separate standard to reduce leakage of hydrofluorocarbon (HFC) refrigerant from cabin air conditioning (A/C) systems from combination tractors, to apply to the tractor manufacturer. This HFC leakage standard is independent of the CO2 tractor standard. Manufacturers can choose technologies from a menu of leak-reducing technologies sufficient to comply with the standard, as opposed to using a test to measure performance. Given that HFC leakage does not relate to fuel efficiency, NHTSA did not adopt corresponding HFC standards.

(b) Heavy-Duty Pickup Trucks and Vans (Class 2b and 3)

Heavy-duty vehicles with a GVWR between 8,501 and 10,000 lb. are classified as Class 2b motor vehicles. Heavy-duty vehicles with a GVWR between 10,001 and 14,000 lb. are classified as Class 3 motor vehicles. Class 2b and Class 3 heavy-duty vehicles (referred to in these rules as “HD pickups and vans”) together emit about 23 percent of today's GHG emissions from the heavy-duty vehicle sector.[47]

The majority of HD pickups and vans are 3/4-ton and 1-ton pickup trucks, 12- and 15-passenger vans,[48] and large work vans that are sold by vehicle manufacturers as complete vehicles, with no secondary manufacturer making substantial modifications prior to registration and use. These vehicles can also be sold as cab-complete vehicles (i.e., incomplete vehicles that include complete or nearly complete cabs that are sold to secondary manufacturers). The majority of heavy-duty pickups and vans are produced by companies with major light-duty markets in the United States. Furthermore, the technologies available to reduce fuel consumption and GHG emissions from this segment are similar to the technologies used on light-duty pickup trucks, including both engine efficiency improvements (for gasoline and diesel engines) and vehicle efficiency improvements. For these reasons, EPA and NHTSA concluded Start Printed Page 73491that it was appropriate to adopt GHG standards, expressed as grams per mile, and fuel consumption standards, expressed as gallons per 100 miles, for HD pickups and vans based on the whole vehicle (including the engine), consistent with the way these vehicles have been regulated by EPA for criteria pollutants and also consistent with the way their light-duty counterpart vehicles are regulated by EPA and NHTSA. This complete vehicle approach adopted by both agencies for HD pickups and vans was consistent with the recommendations of the NAS Committee in its 2010 Report.

For the light-duty GHG and fuel economy standards, the agencies based the emissions and fuel economy targets on vehicle footprint (the wheelbase times the average track width). For those standards, passenger cars and light trucks with larger footprints are assigned higher GHG and lower fuel economy target levels reflecting their inherent tendency to consume more fuel and emit more GHGs per mile. For HD pickups and vans, the agencies believe that setting standards based on vehicle attributes is appropriate, but have found that a work-based metric is a more appropriate attribute than the footprint attribute utilized in the light-duty vehicle rulemaking, given that work-based measures such as towing and payload capacities are critical elements of these vehicles' functionality. EPA and NHTSA therefore adopted standards for HD pickups and vans based on a “work factor” attribute that combines their payload and towing capabilities, with an added adjustment for 4-wheel drive vehicles.

Each manufacturer's fleet average Phase 1 standard is based on production volume-weighting of target standards for all vehicles, which in turn are based on each vehicle's work factor. These target standards are taken from a set of curves (mathematical functions), with separate curves for gasoline and diesel vehicles.[49] However, both gasoline and diesel vehicles in this category are included in a single averaging set. EPA phased in the CO2 standards gradually starting in the 2014 MY, at 15-20-40-60-100 percent of the MY 2018 standards stringency level in MYs 2014-2015-2016-2017-2018, respectively (i.e., the 2014 standards requires only 15 percent of the reduction required in 2018, etc.). The phase-in takes the form of a set of target curves, with increasing stringency in each MY.

NHTSA allowed manufacturers to select one of two fuel consumption standard alternatives for MYs 2016 and later. The first alternative defined individual gasoline vehicle and diesel vehicle fuel consumption target curves that will not change for MYs 2016-2018, and are equivalent to EPA's 67-67-67-100 percent target curves in MYs 2016-2017-2018-2019, respectively. The second alternative defined target curves that are equivalent to EPA's 40-60-100 percent target curves in MYs 2016-2017-2018, respectively. NHTSA allowed manufacturers to opt voluntarily into the NHTSA HD pickup and van program in MYs 2014 or 2015 at target curves equivalent to EPA's target curves. If a manufacturer chose to opt in for one category, they would be required to opt in for all categories. In other words, a manufacturer would be unable to opt in for Class 2b vehicles, but opt out for Class 3 vehicles.

EPA also adopted an alternative phase-in schedule for manufacturers wanting to have stable standards for model years 2016-2018. The standards for heavy-duty pickups and vans, like those for light-duty vehicles, are expressed as set of target standard curves, with increasing stringency in each model year. The Phase 1 EPA standards for 2018 (including a separate standard to control air conditioning system leakage) are estimated to represent an average per-vehicle reduction in GHG emissions of 17 percent for diesel vehicles and 12 percent for gasoline vehicles (relative to pre-control baseline vehicles). The NHTSA standard will require these vehicles to achieve up to about 15 percent reduction in fuel consumption by MY 2018 (relative to pre-control baseline vehicles). Manufacturers demonstrate compliance based on entire vehicle chassis certification using the same duty cycles used to demonstrate compliance with criteria pollutant standards.

(c) Class 2b-8 Vocational Vehicles

Class 2b-8 vocational vehicles include a wide variety of vehicle types, and serve a vast range of functions. Some examples include service for parcel delivery, refuse hauling, utility service, dump, concrete mixing, transit service, shuttle service, school bus, emergency, motor homes, and tow trucks. In Phase 1, we defined Class 2b-8 vocational vehicles as all heavy-duty vehicles that are not included in either the heavy-duty pickup and van category or the Class 7 and 8 tractor category. EPA's and NHTSA's Phase 1 standards for this vocational vehicle category generally apply at the chassis manufacturer level. Class 2b-8 vocational vehicles and their engines emit approximately 17 percent of the GHG emissions and burn approximately 17 percent of the fuel consumed by today's heavy-duty truck sector.[50]

The Phase 1 program for vocational vehicles has vehicle standards and separate engine standards, both of which differ based on the weight class of the vehicle into which the engine will be installed. The vehicle weight class groups mirror those used for the engine standards—Classes 2b-5 (light heavy-duty or LHD in EPA regulations), Classes 6 and 7 (medium heavy-duty or MHD in EPA regulations) and Class 8 (heavy heavy-duty or HHD in EPA regulations). Manufacturers demonstrate compliance with the Phase 1 vocational vehicle CO2 and fuel consumption standards using a vehicle simulation tool described in Section II. The Phase 1 program for vocational vehicles limited the simulation tool inputs to tire rolling resistance. The model assumes the use of a typical representative, compliant engine in the simulation, resulting in one overall value for CO2 emissions and one for fuel consumption.

(d) Engine Standards

The agencies established separate Phase 1 performance standards for the engines manufactured for use in vocational vehicles and Class 7 and 8 tractors.[51] These engine standards vary depending on engine size linked to intended vehicle service class. EPA's engine-based CO2 standards and NHTSA's engine-based fuel consumption standards are being implemented using EPA's existing test procedures and regulatory structure for criteria pollutant emissions from heavy-duty engines. EPA also established engine-based N2 O and CH4 emission standards in Phase 1.

(e) Manufacturers Excluded From the Phase 1 Standards

Phase 1 deferred greenhouse gas emissions and fuel consumption standards for any manufacturers of heavy-duty engines, manufacturers of combination tractors, and chassis manufacturers for vocational vehicles that meet the “small business” size criteria set by the Small Business Administration (SBA). 13 CFR 121.201 Start Printed Page 73492defines a small business by the maximum number of employees; for example, this is currently 1,500 for heavy-duty truck manufacturing and 1,000 for engine manufacturing.[52] In order to utilize this exemption, qualifying small businesses must submit a declaration to the agencies. See Section I.F.(1)(b) for a summary of how Phase 2 applies for small businesses.

The agencies stated that they would consider appropriate GHG and fuel consumption standards for these entities as part of a future regulatory action. This includes both U.S.-based and foreign small-volume heavy-duty manufacturers that introduce new products into the U.S.

(2) Costs and Benefits of the Phase 1 Program

Overall, EPA and NHTSA estimated that the Phase 1 HD National Program will cost the affected industry about $8 billion, while saving vehicle owners fuel costs of nearly $50 billion over the lifetimes of MY 2014-2018 vehicles. The agencies also estimated that the combined standards will reduce CO2 emissions by about 270 million metric tons and save about 530 million barrels of oil over the life of MY 2014 to 2018 vehicles. The agencies estimated additional monetized benefits from CO2 reductions, improved energy security, reduced time spent refueling, as well as possible dis-benefits from increased driving crashes, traffic congestion, and noise. When considering all these factors, we estimated that Phase 1 of the HD National Program will yield $49 billion in net benefits to society over the lifetimes of MY 2014-2018 vehicles.

EPA estimated the benefits of reduced ambient concentrations of particulate matter and ozone resulting from the Phase 1 program to range from $1.3 to $4.2 billion in 2030.[53]

In total, we estimated the combined Phase 1 standards will reduce GHG emissions from the U.S. heavy-duty fleet by approximately 76 million metric tons of CO2-equivalent annually by 2030. In its Environmental Impact Statement for the Phase 1 rule, NHTSA also quantified and/or discussed other potential impacts of the program, such as the health and environmental impacts associated with changes in ambient exposures to toxic air pollutants and the benefits associated with avoided non-CO2 GHGs (methane, nitrous oxide, and HFCs).

(3) Phase 1 Program Flexibilities

As noted above, the agencies adopted numerous provisions designed to give manufacturers a degree of flexibility in complying with the Phase 1 standards. These provisions, which are essentially identical in structure and function in EPA's and NHTSA's regulations, enabled the agencies to consider overall standards that are more stringent and that will become effective sooner than we could consider with a more rigid program, one in which all of a manufacturer's similar vehicles or engines would be required to achieve the same emissions or fuel consumption levels, and at the same time.[54]

Phase 1 included four primary types of flexibility: Averaging, banking, and trading (ABT) provisions; early credits; advanced technology credits (including hybrid powertrains); and innovative technology credit provisions. The ABT provisions were patterned on existing EPA and NHTSA ABT programs (including the light-duty GHG and fuel economy standards) and will allow a vehicle manufacturer to reduce CO2 emission and fuel consumption levels further than the level of the standard for one or more vehicles to generate ABT credits. The manufacturer can use those credits to offset higher emission or fuel consumption levels in the same averaging set, “bank” the credits for later use, or “trade” the credits to another manufacturer. As also noted above, for HD pickups and vans, we adopted a fleet averaging system very similar to the light-duty GHG and CAFE fleet averaging system. In both programs, manufacturers are allowed to carry-forward deficits for up to three years without penalty. The agencies provided in the ABT programs flexibility for situations in which a manufacturer is unable to avoid a negative credit balance at the end of the year. In such cases, manufacturers are not considered to be out of compliance unless they are unable to make up the difference in credits by the end of the third subsequent model year.

In total, the Phase 1 program divides the heavy-duty sector into 14 subcategories of vehicles and 4 subcategories of engines. These subcategories are grouped into 4 vehicle averaging sets and 4 engine averaging sets in the ABT program. For tractors and vocational vehicles, the fleet averaging sets are: Light heavy-duty (Classes 2b-5); medium heavy-duty (Class 6-7); and heavy heavy-duty (Class 8). Complete HD pickups and vans (both spark-ignition and compression-ignition) are the final vehicle averaging set. For engines, the fleet averaging sets are spark-ignition engines, compression-ignition light heavy-duty engines, compression-ignition medium heavy-duty engines, and compression-ignition heavy heavy-duty engines. ABT allows the exchange of credits within an averaging set. This means that a Class 8 day cab tractor can exchange credits with a Class 8 sleeper tractor but not with a smaller Class 7 tractor. Also, a Class 8 vocational vehicle can exchange credits with a Class 8 tractor. However, we did not allow trading between engines and chassis (i.e. vehicles).

In addition to ABT, the other primary flexibility provisions in the Phase 1 program involve opportunities to generate early credits, advanced technology credits (including for use of hybrid powertrains), and innovative technology credits.[55] For the early credits and advanced technology credits, the agencies adopted a 1.5x multiplier, meaning that manufacturers would get 1.5 credits for each early credit and each advanced technology credit. In addition, advanced technology credits for Phase 1 can be used anywhere within the heavy-duty sector (including both vehicles and engines). Put another way, as a means of promoting these promising technologies, the Phase 1 rule does not restrict averaging or trading by averaging set in this instance.

For other vehicle or engine technologies that can reduce CO2 and fuel consumption, but whose benefits are not reflected if measured using the Phase 1 test procedures, the agencies wanted to encourage the development of such innovative technologies, and therefore adopted special “innovative technology” credits. These innovative technology credits apply to technologies that are shown to produce emission and fuel consumption reductions that are not adequately recognized on the Phase 1 test procedures and that were not yet in widespread use in the heavy-duty sector before MY 2010. Manufacturers Start Printed Page 73493need to quantify the reductions in fuel consumption and CO2 emissions that the technology is expected to achieve, above and beyond those achieved on the Phase 1 test procedures. As with ABT, the use of innovative technology credits is allowed only among vehicles and engines of the same defined averaging set generating the credit, as described above. The credit multiplier likewise does not apply for innovative technology credits.

(4) Implementation of Phase 1

Manufacturers have already begun complying with the Phase 1 standards. In some cases manufacturers voluntarily chose to comply early, before compliance was mandatory. The Phase 1 rule allowed manufacturers to generate credits for such early compliance. The market appears to be very accepting of the new technologies, and the agencies have seen no evidence of “pre-buy” effects in response to the standards. In fact sales have been higher in recent years than they were before Phase 1. Moreover, manufacturers' compliance plans indicate intention to utilize the Phase 1 flexibilities, and we have yet to see significant non-compliance with the standards.

(5) Litigation on Phase 1 Rule

The D.C. Circuit rejected all challenges to the agencies' Phase 1 regulations. The court did not reach the merits of the challenges, holding that none of the petitioners had standing to bring their actions, and that a challenge to NHTSA's denial of a rulemaking petition could only be brought in District Court. See Delta Construction v. EPA, 783 F. 3d 1291 (D.C. Cir. 2015).

C. Summary of the Phase 2 Standards and Requirements

The agencies are adopting new standards that build on and enhance existing Phase 1 standards, and are adopting as well the first-ever standards for certain trailers used in combination with heavy-duty tractors. Taken together, the Phase 2 program comprises a set of largely technology-advancing standards that will achieve greater GHG and fuel consumption savings than the Phase 1 program. As described in more detail in the following sections, the agencies are adopting these standards because, based on the information available at this time and careful consideration of all comments, we believe they best fulfill our respective statutory authorities when considered in the context of available technology, feasible reductions of emissions and fuel consumption, costs, lead time, safety, and other relevant factors.

The Phase 2 standards represent a more technology-forcing [56] approach than the Phase 1 approach, predicated on use of both off-the-shelf technologies and emerging technologies that are not yet in widespread use. The agencies are adopting standards for MY 2027 that we project will require manufacturers to make extensive use of these technologies. The standards increase in stringency incrementally beginning in MY 2018 for trailers and in MY 2021 for other segments, ensuring steady improvement to the MY 2027 stringency levels. For existing technologies and technologies in the final stages of development, we project that manufacturers will likely apply them to nearly all vehicles, excluding those specific vehicles with applications or uses that prevent the technology from functioning properly. We also project as one possible compliance pathway that manufacturers could apply other more advanced technologies such as hybrids and waste engine heat recovery systems, although at lower application rates than the more conventional technologies. Comments on the overall stringency of the proposed Phase 2 program were mixed. Many commenters, including most non-governmental organizations, supported more stringent standards with less lead time. Many technology and component suppliers supported more stringent standards but with the proposed lead time. Vehicle manufacturers did not support more stringent standards and emphasized the importance of lead time. To the extent these commenters provided technical information to support their comments on stringency and lead time, it is discussed in Sections II through VI.

The standards being adopted provide approximately ten years of lead time for manufacturers to meet these 2027 standards, which the agencies believe is appropriate to implement the technologies industry could use to meet these standards. For some of the more advanced technologies production prototype parts are not yet available, though they are in the research stage with some demonstrations in actual vehicles.[57] In the respective sections of Chapter 2 of the RIA, the agencies explain what further steps are needed to successfully and reliably commercialize these prototypes in the lead time afforded by the Phase 2 standards. Additionally, even for the more developed technologies, phasing in more stringent standards over a longer timeframe will help manufacturers to ensure better reliability of the technology and to develop packages to work in a wide range of applications.

As discussed later, the agencies are also adopting new standards in MYs 2018 (trailers only), 2021, and 2024 to ensure that manufacturers make steady progress toward the 2027 standards, thereby achieving steady and feasible reductions in GHG emissions and fuel consumption in the years leading up to the MY 2027 standards.

Providing additional lead time can often enable manufacturers to resolve technological challenges or to find lower cost means of meeting new regulatory standards, effectively making them more feasible in either case. See generally NRDC v. EPA, 655 F. 2d 318, 329 (D.C. Cir. 1981). On the other hand, manufacturers and/or operators may incur additional costs if regulations require them to make changes to their products with less lead time than manufacturers would normally have when bringing a new technology to the market or expanding the application of existing technologies. After developing a new technology, manufacturers typically conduct extensive field tests to ensure its durability and reliability in actual use. Standards that accelerate technology deployment can lead to manufacturers incurring additional costs to accelerate this development work, or can lead to manufacturers beginning production before such testing can be completed. Some industry stakeholders have informed EPA that when manufacturers introduced new emission control technologies (primarily diesel particulate filters) in response to the 2007 heavy-duty engine standards they did not perform sufficient product development validation, which led to additional costs for operators when the technologies required repairs or resulted in other operational issues in use. Thus, the issues of costs, lead time, and reliability are intertwined for the Start Printed Page 73494agencies' determination of whether standards are reasonable and maximum feasible, respectively.

Another important consideration was the possibility of disrupting the market, which would be a risk if compliance required application of new technologies too suddenly. Several of the heavy-duty vehicle manufacturers, fleets, and commercial truck dealerships informed the agencies that for fleet purchases that are planned more than a year in advance, expectations of reduced reliability, increased operating costs, reduced residual value, or of large increases in purchase prices can lead the fleets to pull-ahead by several months planned future vehicle purchases by pre-buying vehicles without the newer technology. In the context of the Class 8 tractor market, where a relatively small number of large fleets typically purchase very large volumes of tractors, such actions by a small number of firms can result in large swings in sales volumes. Such market impacts would be followed by some period of reduced purchases that can lead to temporary layoffs at the factories producing the engines and vehicles, as well as at supplier factories, and disruptions at dealerships. Such market impacts also can reduce the overall environmental and fuel consumption benefits of the standards by delaying the rate at which the fleet turns over. See International Harvester v. EPA, 478 F. 2d 615, 634 (D.C. Cir. 1973). A number of commenters stated that the 2007 EPA heavy-duty engine criteria pollutant standard precipitated pre-buy for the Class 8 tractor market.[58] The agencies understand the potential impact that fleets pulling ahead purchases can have on American manufacturing and labor, dealerships, truck purchasers, and on the program's environmental and fuel savings goals, and have taken steps in the design of the program to avoid such disruption (see also our discussion in RTC Section 11.7). These steps include the following:

  • Providing considerable lead time
  • Adopting standards that will result in significantly lower operating costs for vehicle owners (unlike the 2007 standard, which increased operating costs)
  • Phasing in the standards
  • Structuring the program so the industry will have a significant range of technology choices to be considered for compliance, rather than the one or two new technologies the OEMs pursued to comply with EPA's 2007 criteria pollutant standard
  • Allowing manufacturers to use emissions averaging, banking and trading to phase in the technology even further

As discussed in the Phase 1 final rule, NHTSA has certain statutory considerations to take into account when determining feasibility of the preferred alternative.[59] EISA states that NHTSA (in consultation with EPA and the Secretary of Energy) will develop a commercial medium- and heavy-duty fuel efficiency program designed “to achieve the maximum feasible improvement.” [60] Although there is no definition of maximum feasible standards in EISA, NHTSA is directed to consider three factors when determining what the maximum feasible standards are. Those factors are, appropriateness, cost-effectiveness, and technological feasibility,[61] which modify “feasible” beyond its plain meaning.

NHTSA has the broad discretion to weigh and balance the aforementioned factors in order to accomplish EISA's mandate of determining maximum feasible standards. The fact that the factors may often be at odds gives NHTSA significant discretion to decide what weight to give each of the competing factors, policies and concerns and then determine how to balance them—as long as NHTSA's balancing does not undermine the fundamental purpose of the EISA: Energy conservation, and as long as that balancing reasonably accommodates “conflicting policies that were committed to the agency's care by the statute.” [62]

EPA also has significant discretion in assessing, weighing, and balancing the relevant statutory criteria. Section 202(a)(2) of the Clean Air Act (42 U.S.C. 7521(a)(2)) requires that the standards “take effect after such period as the Administrator finds necessary to permit the development and application of the requisite technology, giving appropriate consideration to the cost of compliance within such period.” This language affords EPA considerable discretion in how to weight the critical statutory factors of emission reductions, cost, and lead time (76 FR 57129-57130). Section 202(a)(2) also allows (although it does not compel) EPA to adopt technology-forcing standards. Id. at 57130.

Sections II through VI of this Preamble explain the consideration that the agencies took into account based on careful assessment and balancing of the statutory factors under Clean Air Act section 202(a)(1) and (2), and under 49 U.S.C. 32902(k).

(1) Carryover From Phase 1 Program and Compliance Changes

Phase 2 is carrying over many of the compliance approaches developed for Phase 1, with certain changes as described below. Readers are referred to the regulatory text for much more detail. Note that the agencies have adapted some of these Phase 1 provisions in order to address new features of the Phase 2 program, notably provisions related to trailer compliance. The agencies have also reevaluated all of the compliance provisions to ensure that they will be effective in achieving the projected reductions without placing an undue burden on manufacturers.

The agencies received significant comments from vehicle manufacturers emphasizing the potential for the structure of the compliance program to impact stringency. Although the agencies do not agree with all of these comments (which are discussed in more detail in later sections), we do agree that it is important to structure the compliance program so that the effective stringency of standards is consistent with levels established by regulation. The agencies have made appropriate improvements to the compliance structure in response to these comments.

(a) Certification

EPA and NHTSA are applying the same general certification procedures for Phase 2 as are currently being used for certifying to the Phase 1 standards. Tractors and vocational vehicles will continue to be certified using the vehicle simulation tool (GEM). The agencies, however, revised the Phase 1 GEM simulation tool to develop a new version, Phase 2 GEM, that more specifically reflects improvements to engines, transmissions, and drivetrains.[63] Rather than the GEM simulation tool using default values for engines, transmissions and drivetrains, most manufacturers will enter measured or tested values as inputs reflecting performance of the actual engine, transmission and drivetrain technologies.

Start Printed Page 73495

The Phase 1 certification process for engines used in tractors and vocational vehicles was based on EPA's process for showing compliance with the heavy-duty engine criteria pollutant standards using engine dynamometer testing, and the agencies are continuing it for Phase 2. We also will continue certifying HD pickups and vans using the Phase 1 chassis dynamometer testing results and vehicle certification process, which is very similar to the light-duty vehicle certification process. The Phase 2 trailer certification process will resemble the Phase 2 tractor certification approach, but with a simplified version of Phase 2 GEM. The trailer certification process allows trailer manufacturers to use a simple equation to determine GEM-equivalent g/ton-mile emission rates without actually running GEM.

EPA and NHTSA are also clarifying provisions related to confirming a manufacturer's test data during certification (i.e., confirmatory testing) and verifying a manufacturer's vehicles are being produced to perform as described in the application for certification (i.e., selective enforcement audits or SEAs). The EPA confirmatory testing provisions for engines, vehicles, and components are in 40 CFR 1036.235 and 1037.235. The SEA provisions are in 40 CFR 1036.301 and 1037.301-1037.320. The NHTSA provisions are in 49 CFR 535.9(a). As we proposed, these clarifications will also apply for Phase 1 engines and vehicles.

In response to comments, we are making several changes to the proposed EPA confirmatory testing provisions. First, the regulations being adopted specify that EPA will conduct triplicate tests for engine fuel maps to minimize the impact of test-to-test variability. The final regulations also state that we will consider entire fuel maps rather than individual points. Engine manufacturers objected to EPA's proposal that individual points could be replaced based on a single test, arguing that it effectively made the vehicle standards more stringent due to point-to-point and test-to-test variability. We believe that the changes being adopted largely address these concerns. We are also applying this approach for axle and transmission maps for similar reasons.

As described in Sections III and IV, EPA has also modified the SEA regulations for verifying aerodynamic performance. These revised regulations differ somewhat from the standard SEA regulations to address the unique challenges of measuring aerodynamic drag. In particular EPA recognizes that for coastdown testing, test-to-test variability is expected to be large relative to production variability. This differs fundamentally from traditional compliance testing, in which test-to-test variability is expected to be small relative to production variability. To address this difference, the modified regulations call for more repeat testing of the same vehicle, but fewer test samples. These revisions were generally supported by commenters. See Section III and IV for additional discussion.

Some commenters suggested that the agencies should apply a compliance margin to confirmatory and SEA test results to account for test variability. However, other commenters supported following EPA's past practice, which has been to base the standards on technology projections that assume manufacturers will apply compliance margins to their test results for certification. In other words, they design their products to have emissions below the standards by some small margin so that test-to-test or lab-to-lab variability would not cause them to exceed any applicable standards. Consequently, EPA has typically not set standards precisely at the lowest levels achievable, but rather at slightly higher levels—expecting manufacturers to target the lower levels to provide compliance margins for themselves. As discussed in Sections II through VI, the agencies have applied this approach to the Phase 2 standards.

(b) Averaging, Banking and Trading (ABT)

The Phase 1 ABT provisions were patterned on established EPA ABT programs that have proven to work well. In Phase 1, the agencies determined this flexibility would provide an opportunity for manufacturers to make necessary technological improvements and reduce the overall cost of the program without compromising overall environmental and fuel economy objectives. Commenters generally supported this approach for engines, pickups/vans, tractors, and vocational vehicles. Thus, we are generally continuing this Phase 1 approach with few revisions to the engine and vehicle segments. However, as described in Section IV, in response to comments, we are finalizing a much more limited averaging program for trailers that will not go into effect until 2027. We are adopting some other provisions for certain vocational vehicles, which are discussed in Section V.

The agencies see the overall ABT program as playing an important role in making the technology-advancing standards feasible, by helping to address many issues of technological challenges in the context of lead time and costs. It provides manufacturers flexibilities that assist the efficient development and implementation of new technologies and therefore enable new technologies to be implemented at a more aggressive pace than without ABT.

ABT programs are more than just add-on provisions included to help reduce costs. They can be, as in EPA's Title II programs generally, an integral part of the standard setting itself. A well-designed ABT program can also provide important environmental and energy security benefits by increasing the speed at which new technologies can be implemented (which means that more benefits accrue over time than with later-commencing standards) and at the same time increase flexibility for, and reduce costs to, the regulated industry and ultimately consumers. Without ABT provisions (and other related flexibilities), standards would typically have to be numerically less stringent since the numerical standard would have to be adjusted to accommodate issues of feasibility and available lead time. See 75 FR 25412-25413. By offering ABT credits and additional flexibilities the agencies can offer progressively more stringent standards that help meet our fuel consumption reduction and GHG emission goals at a faster and more cost-effective pace.[64]

(i) Carryover of Phase 1 Credits and Credit Life

The agencies proposed to continue the five-year credit life provisions from Phase 1, and not to adopt any general restriction on the use of banked Phase 1 credits in Phase 2. In other words, Phase 1 credits in MY 2019 could be used in Phase 1 or in Phase 2 in MYs 2021-2024. CARB commented in support of a more restrictive approach for Phase 1 credits, based on the potential for manufacturers to delay implementation of technology in Phase 2 by using credits generated under Phase 1. We also received comments asking the agencies to provide a path for manufacturers to generate credits for applying technologies not explicitly included in the Phase 1 program. In response to these comments, the agencies have analyzed the potential impacts of Phase 1 credits on the Phase 2 program for each sector and made appropriate adjustments in the program. For example, as described in Section II.D.(5), the agencies are adopting some restrictions on the carryover of windfall Phase 1 engine credits that result from the Phase 1 vocational engine standards. Start Printed Page 73496Also, as described in Section III, the agencies are projecting that Phase 1 credit balances for tractor manufacturers will enable them to meet more stringent standards for MY 2021-2023, so the agencies have increased the stringency of these standards accordingly.

In contrast to the Phase 1 tractor program, the Phase 1 vocational chassis program currently offers fewer opportunities to generate credits for potential carryover into Phase 2. To address comments related to this particular situation and also to provide a new Phase 1 incentive to voluntarily apply certain Phase 2 technologies, which are available today but currently not being adopted, the agencies are finalizing a streamlined Phase 1 off-cycle credit approval process for these Phase 2 technologies. For vocational chassis, these technologies include workday idle reduction technologies such as engine stop-start systems, automatic engine shutdown systems, shift-to-neutral at idle automatic transmissions, automated manual transmissions, and dual-clutch transmissions. The agencies are also finalizing a streamlined Phase 1 off-cycle credit approval process for Phase 2 automatic tire inflation systems (ATIS), for both tractors and vocational chassis. The purpose for offering these streamlined off-cycle approval processes for Phase 1 is to encourage more early adoption of these Phase 2 technologies during the remaining portion of the Phase 1 program (e.g., model years 2018, 2019, 2020). Earlier adoption of these technologies would help demonstrate that these newer, but not advanced, technologies are effective, reliable and well-accepted into the marketplace by the time the agencies project that they would be needed for compliance with the Phase 2 standards.

The agencies are also including a provision allowing exempt small business manufacturers of vocational chassis to opt into the Phase 1 program for the purpose of generating credits which can be used throughout the Phase 2 program, as just described.

In conjunction with this provision allowing manufacturers to receive credit in Phase 1 for pulling ahead certain Phase 2 technologies, the agencies are providing an extended credit life for the Light and Medium heavy-duty vocational vehicle averaging sets (see next subsection) to provide additional Phase 2 transition flexibility for these vehicles. Unlike the HD Phase 1 pickup/van and tractor programs, where the averaging sets are broad; where manufacturers have many technology choices from which to earn credits (e.g., tractor aerodynamic and idle reduction technologies, pickup/van engine and transmission technologies); and where we project manufacturers to have sufficient pickup/van and tractor credits to manage the transition to the Phase 2 standards, transitioning to the new Light and Medium vocational vehicle standards may be more challenging. Manufacturers selling lower volumes of these lighter vehicles may find themselves with fewer overall credits to manage the transition to the new standards, especially the 2027 standards. To facilitate this transition and better assure adequate lead time, the agencies are extending the credit life for the Light and Medium heavy-duty vehicle averaging sets (typically vehicles in Classes 2b through 7) so that all credits generated in 2018 and later will last at least until 2027. We are not doing this for the Heavy heavy-duty vocational vehicle category (typically Class 8) because tractor credits may be used within this averaging set. Because we project that manufacturers will have sufficient tractor credits, we believe that they will be able to manage the Heavy vocational transition to each set of new standards, without the extended credit life that we are finalizing for Light and Medium vocational averaging sets. Nevertheless, we will continue to monitor the manufacturers' progress in transitioning to the Phase 2 standards for each category, and we may reconsider the need for additional transitional flexibilities, such as extending other categories' credit lives.

Although, as we have already noted, the numerical values of Phase 2 standards are not directly comparable in an absolute sense to the existing Phase 1 standards (in other words, a given vehicle would have a different g/ton-mile emission rate when evaluated using Phase 1 GEM than it would when evaluated using Phase 2 GEM), we believe that the Phase 1 and Phase 2 credits are largely equivalent. Because the standards and emission levels are included in a relative sense (as a difference), it is not necessary for the Phase 1 and Phase 2 standards to be directly equivalent in an absolute sense in order for the credits to be equivalent.

This is best understood by examining the way in which credits are calculated. For example, the credit equations in 40 CFR 1037.705 and 49 CFR 535.7 calculate credits as the product of the difference between the standard and the vehicle's emission level (g/ton-mile or gallon/1,000 ton-mile), the regulatory payload (tons), production volume, and regulatory useful life (miles). The Phase 2 payloads, production volumes, and useful lives for tractors, medium and heavy heavy-duty engines, or medium and heavy heavy-duty vocational vehicles are equivalent to those of Phase 1. However, EPA is changing the regulatory useful lives of HD pickups and vans, light heavy-duty vocational vehicles, spark-ignited engines, and light heavy-duty compression-ignition engines. Because useful life is a factor in determining the value of a credit, the agencies proposed to apply interim adjustment factors to ensure banked credits maintain their value in the transition from Phase 1 to Phase 2.

For Phase 1, EPA aligned the useful life for GHG emissions with the useful life already in place for criteria pollutants. After the Phase 1 rules were finalized, EPA updated the useful life for criteria pollutants as part of the Tier 3 rulemaking.[65] The new useful life implemented for Tier 3 is 150,000 miles or 15 years, whichever occurs first. This same useful life is being adopted in Phase 2 for HD pickups and vans, light heavy-duty vocational vehicles, spark-ignited engines, and light heavy-duty compression-ignition engines.[66] The numeric value of the adjustment factor for each of these regulatory categories depends on the Phase 1 useful life. These are described in detail below in this Preamble in Sections II, V, and VI. Without these adjustment factors the changes in useful life would effectively result in a discount of banked credits that are carried forward from Phase 1 to Phase 2, which is not the intent of the changes in the useful life. With the relatively flat deterioration generally associated with CO2, EPA does not believe the changes in useful life will significantly affect the feasibility of the Phase 2 standards.

We note that the primary purpose of allowing manufacturers to bank credits is to provide flexibility in managing transitions to new standards. The five-year credit life is substantial, and allows credits generated in either Phase 1 or early in Phase 2 to be used for the intended purpose. The agencies believe a credit life longer than five years is unnecessary to accomplish this transition. Restrictions on credit life serve to reduce the likelihood that any manufacturer will be able to use banked credits to disrupt the heavy-duty vehicle market in any given year by effectively limiting the amount of credits that can be held. Without this limit, one manufacturer that saved enough credits over many years could achieve a significant cost advantage by using all the credits in a single year. The agencies Start Printed Page 73497believe that allowing a five-year credit life for all credits, and as a consequence allowing use of Phase 1 credits in Phase 2, creates appropriate flexibility and appropriately facilitates a smooth transition to each new level of standards.

(ii) Averaging Sets

EPA has historically restricted averaging to some extent for its HD emission standards to avoid creating unfair competitive advantages or environmental risks due to credits being inconsistent. It also helps to ensure a robust and manageable compliance program. Under Phase 1, averaging, banking and trading can only occur within and between specified “averaging sets” (with the exception of credits generated through use of specified advanced technologies). As proposed, we will continue this regime in Phase 2, retaining the existing vehicle and engine averaging sets, and creating new trailer averaging sets. We are also continuing the averaging set restrictions from Phase 1 in Phase 2. (See Section V for certain other provisions applicable to vehicles certified to special standards.) These general averaging sets for vehicles are:

  • Complete pickups and vans
  • Other light heavy-duty vehicles (Classes 2b-5)
  • Medium heavy-duty vehicles (Class 6-7)
  • Heavy heavy-duty vehicles (Class 8)
  • Long dry and refrigerated van trailers [67]
  • Short dry and refrigerated van trailers

We are not allowing trading between engines and chassis, even within the same vehicle class. Such trading would essentially result in double counting of emission credits, because the same engine technology would likely generate credits relative to both standards (and indeed, certain engine improvements are reflected exclusively in the vehicle standards the agencies are adopting). We similarly limit trading among engine categories to trades within the designated averaging sets:

  • Spark-ignition engines
  • Compression-ignition light heavy-duty engines
  • Compression-ignition medium heavy-duty engines
  • Compression-ignition heavy heavy-duty engines

The agencies continue to believe that maintaining trading to be only within the classes listed above will provide adequate opportunities for manufacturers to make necessary technological improvements and to reduce the overall cost of the program without compromising overall environmental and fuel efficiency objectives, and it is therefore appropriate and reasonable under EPA's authority and maximum feasible under NHTSA's authority, respectively. We do not expect emissions from engines and vehicles—when restricted by weight class—to be dissimilar. We therefore expect that the lifetime vehicle performance and emissions levels will be very similar across these defined categories, and the credit calculations will fairly ensure the expected fuel consumption and GHG emission reductions.

These restrictions have generally worked well for Phase 1, and we continue to believe that these averaging sets create flexibility without creating an unfair advantage for manufacturers with integrated portfolios, including engines and vehicles. See 76 FR 57240.

(iii) Credit Deficits

The Phase 1 regulations allow manufacturers to carry-forward deficits for up to three years. This is an important flexibility because the program is designed to address the diversity of the heavy-duty industry by allowing manufacturers to sell a mix of engines or vehicles that have very different emission levels and fuel efficiencies. Under this construct, manufacturers can offset sales of engines or vehicles not meeting the standards by selling others (within the same averaging set) that perform better than the standards require. However, in any given year it is possible that the actual sales mix will not balance out, and the manufacturer may be short of credits for that model year. The three-year provision allows for this possibility and creates additional compliance flexibility to accommodate it.

(iv) Advanced Technology Credits

At the time of the proposal, the agencies believed it was no longer appropriate to provide extra credit for any of the technologies identified as advanced technologies for Phase 1, although we requested comment on this issue. The Phase 1 advanced technology credits were adopted to promote the implementation of advanced technologies that were not included in our basis of the feasibility of the Phase 1 standards. Such technologies included hybrid powertrains, Rankine cycle waste heat recovery systems on engines, all-electric vehicles, and fuel cell vehicles (see 40 CFR 86.1819-14(k)(7), 1036.150(h), and 1037.150(p)). The Phase 2 heavy-duty engine and vehicle standards are premised on the use of some of these technologies, making them equivalent to other fuel-saving technologies in this context. We believe the Phase 2 standards themselves will provide sufficient incentive to develop those specific technologies.

Although the agencies proposed to eliminate all advanced technology incentives, we remained open to targeted incentives that would address truly advanced technology. We specifically requested comment on this issue with respect to electric vehicle, plug-in hybrid, and fuel cell technologies. Although the Phase 2 standards are premised on some use of Rankine cycle waste heat recovery systems on engines and hybrid powertrains, none of these standards are based on projected utilization of these other even more-advanced technologies (e.g., all-electric vehicles, fuel cell vehicles). 80 FR 40158. Commenters generally supported providing credit multipliers for these advanced technologies. However, Allison supported ending the incentives for hybrids, fuel cells, and electric vehicles in Phase 2. ATA, on the other hand, commented that the agencies should preserve the advanced technology credits which provide a credit multiplier of 1.5 in order to promote the use of hybrid and electric vehicles in larger vocational vehicles and tractors. ARB supported the use of credit multipliers even more strongly and provided suggestions for values larger than 1.5 that could be used to incentivize plug-in hybrids, electric vehicles, and fuel cell vehicles. Eaton recommended the continuation of advanced technology credits for hybrid powertrains until a sufficient number are in the market. Overall, the comments indicated that there is support for such incentives among operators, suppliers, and states. Upon further consideration, the agencies are adopting advanced technology credits for these three types of advanced technologies, as shown in Table I-2 below.

Table I-2—Advanced Technology Multipliers

TechnologyMultiplier
Plug-in hybrid electric vehicles3.5
All-electric vehicles4.5
Fuel cell vehicles5.5

Our intention in adopting these multipliers is to create a meaningful incentive to those considering adopting these qualifying advanced technologies into their vehicles. The values being Start Printed Page 73498adopted are consistent with values recommended by CARB in their supplemental comments.[68] CARB's values were based on a cost analysis that compared the costs of these technologies to costs of other conventional technologies. Their costs analysis showed that adopting multipliers in this range would make these technologies much more competitive with the conventional technologies and could allow manufacturers to more easily generate a viable business case to develop these technologies for heavy-duty and bring them to market at a competitive price.

Another important consideration in the adoption of these larger multipliers is the tendency of the heavy-duty sector to significantly lag the light-duty sector in the adoption of advanced technologies. There are many possible reasons for this, such as:

  • Heavy-duty vehicles are more expensive than light-duty vehicles, which makes it a greater monetary risk for purchasers to invest in unproven technologies.
  • These vehicles are work vehicles, which makes predictable reliability even more important than for light-duty vehicles.
  • Sales volumes are much lower for heavy-duty vehicles, especially for specialized vehicles.

As a result of factors such as these, adoption rates for these advanced technologies in heavy-duty vehicles are essentially non-existent today and seem unlikely to grow significantly within the next decade without additional incentives.

The agencies believe it is appropriate to provide such large multipliers for these very advanced technologies at least in the short term, because they have the potential to provide very large reductions in GHG emissions and fuel consumption and advance technology development substantially in the long term. However, because they are so large, we also believe that we should not necessarily allow them to continue indefinitely. Therefore, the agencies are adopting them as an interim program that will continue through MY 2027. If the agencies determine that these credit multipliers should be continued beyond MY 2027, we could do so in a future rulemaking.

As discussed in Section I.C.(1)(d), the agencies are not specifically accounting for upstream emissions that might occur from production of electricity to power these advanced vehicles. This approach is largely consistent with the incentives offered for electric vehicles in the light-duty National Program. 77 FR 62810. For light-duty vehicles, the agencies also did not require manufacturers to account for upstream emissions during the initial years, as the technologies are being developed. While we proactively sunset this allowance for light-duty due to concerns about potential impacts from very high sales volumes, we do not have similar concerns for heavy-duty. Nevertheless, in this program we are only adopting these credit multipliers through MY 2027, and should we not promulgate a future rulemaking to extend them beyond MY 2027, these multipliers would essentially sunset in MY 2027.

One feature of the Phase 1 advanced technology program that is not being continued in Phase 2 is the allowance to use advanced technology credits across averaging sets. We believe that combined with the very large multipliers being adopted, there could be too large a risk of market distortions if we allowed the use of these credits across averaging sets.

(v) Transition Flexibility for Meeting the Engine Standards

Some manufacturers commented that the proposed engine regulations did not offer sufficient flexibility. Although these commenters acknowledge that the tractor and vocational vehicle standards will separately drive engine improvements, they nonetheless maintain that the MY 2024 engine standards may constrain potential compliance paths too much. Some commented that advanced technologies (such as waste heat recovery) may need to be deployed before the technologies are fully reliable for every engine manufacturer, and may lead to the development and implementation of additional engine technologies outside of scheduled engine redesign cycles, which could cause manufacturers to incur costs which were not accounted for in the agencies' analyses. These costs could include both product development and equipment costs for the engine manufacturer, and potential increased costs for vehicle owners associated with potential reliability issues in-the-field.

The agencies have considered these comments carefully. See, e.g., RIA Section 2.3.9 and RTC Section 3.4. The agencies recognize the importance of ensuring that there is adequate lead time to develop, test, and otherwise assure reliability of the technologies projected to be needed to meet the standards and for the advanced engine technologies in particular. See Section I.C above; see also responses regarding waste heat recovery technology in RTC Section 3.4, and Response 3.4.1. The agencies are therefore adopting an alternative, optional ABT flexibility for heavy-heavy and medium-heavy engines in partial response to these comments. This optional provision would affect only the MYs 2021 and 2024 standards for these engines, not the final MY 2027 engine standards, and to the extent manufacturers elect the provision would increase fuel consumption and GHG reduction benefits, as explained below.

This optional provision has three aspects:

  • A pull ahead of the engine standards to MY 2020
  • Extended credit life for engine credits generated against MYs 2018-2019 Phase 1 standards, the MY 2020 pull-ahead Phase 2 engine standards, and the MYs 2021-2024 Phase 2 engine standards
  • Slightly relaxed engine standards for MYs 2024-2026 tractor engine standards [69]

Thus, the final rule provides the option of an extended credit life for the medium heavy-duty and heavy heavy-duty engines so that all credits generated in MY 2018 and later will last at least until MY 2030.[70] To be eligible for this allowance, manufacturers would need to voluntarily certify all of their HHD and/or MHD MY 2020 engines (tractor and vocational) to MY 2021 standards.[71] Manufacturers could elect to apply this provision separately to medium heavy-duty and heavy heavy-duty engines, since these remain separate averaging sets. Credits banked by the manufacturer in Phase 1 for model year 2018 and 2019 engines would be eligible for the extended credit life for manufacturers satisfying the pull ahead requirement. Such credits could be used in any model year 2021 through Start Printed Page 734992030. Manufacturers that voluntarily certify their engines to MY 2021 standards early would then also be eligible for slightly less stringent engine tractor standards in MYs 2024-2026, as shown in the following table.

Table I-3—Optional ABT Flexibility Standards for Heavy-Heavy and Medium-Heavy Engines

Model yearsMedium heavy-duty—tractorHeavy heavy-duty—tractor
EPA CO2 standard (g/bhp-hr)NHTSA fuel consumption standard (gal/100bhp-hr)EPA CO2 standard (g/bhp-hr)NHTSA fuel consumption standard (gal/100bhp-hr)
2020-20234734.64644474.3910
2024-20264674.58744424.3418

Once having opted into this alternative compliance path, engine manufacturers would have to adhere to that path for the remainder of the Phase 2 program. The choice would be made when certifying MY 2020 engines. Instead of certifying engines to the final year of the Phase 1 engine standards, manufacturers electing the alternative would indicate that they are instead certifying to the MY 2021 Phase 2 engine standard.

Because these engine manufacturers would be reducing emissions of engines otherwise subject to the MY 2020 Phase 1 engine standards (and because engine reductions were not reflected in the Phase 1 vehicle program), there would be a net benefit to the environment. These engines would not generate credits relative to the Phase 1 standards (that is, MY 2020 engines would only use or generate credits relative to the pulled ahead MY 2021 Phase 2 engines standards) which would result in net reductions of CO2 and fuel consumption of about 2 percent for each engine. Thus, if every engine manufacturer chooses to use this flexibility, there could be resulting reductions of an additional 12MMT of CO2 and saving of nearly one billion gallons of diesel fuel.

This alternative also does not have adverse implications for the vehicle standards. As just noted, the vehicle standards themselves are unaffected. Thus, these voluntary standards would not reduce the GHG reductions or fuel savings of the program. Vehicle manufacturers using the alternative MYs 2024-2026 engines would need to adopt additional vehicle technology (i.e. technology beyond that projected to be needed to meet the standard) to meet the vehicle standards. This means the vehicles would still achieve the same fuel efficiency in use.[72]

In sum, the agencies view this alternative as being positive from the environmental and energy conservation perspectives, and believe it will provide significant flexibility for manufacturers that may reduce their compliance costs. It also provides a hedge against potential premature introduction of advanced engine technologies, providing more lead time to assure in-use reliability.

(c) Innovative Technology and Off-Cycle Credits

The agencies are continuing the Phase 1 innovative technology program (reflecting certain streamlining features as just discussed), but re-designating it as an off-cycle program for Phase 2. In other words, beginning in MY 2021 technologies that are not accounted for in the GEM simulation tool, or by compliance dynamometer testing (for engines or chassis certified vehicles) will be considered “off-cycle,” including those technologies that may no longer be considered innovative technologies.

The final rules provide that in order for a manufacturer to receive these credits for Phase 2, the off-cycle technology will still need to meet the requirement that it was not in common use prior to MY 2010. Although we have not identified specific off-cycle technologies at this time that should be excluded, we believe it is prudent to continue this requirement to avoid the potential for manufacturers to receive windfall credits for technologies that they were already using before MY 2010, and that are therefore reflected in the Phase 2 (and possibly Phase 1) baselines. However, because the Phase 2 program will be implemented in MY 2021 and extend at least through MY 2027, the agencies and manufacturers may have difficulty in the future determining whether an off-cycle technology was in common use prior to MY 2010. In order to avoid this approach becoming an unnecessary hindrance to the off-cycle program, the agencies will presume that off-cycle technologies were not in common use in 2010 unless we have clear evidence to the contrary. Neither the agencies nor manufacturers will be required to demonstrate that the technology meets this 2010 criteria. Rather, the agencies will simply retain the authority to deny a request for off-cycle credits if it is clear that the technology was in common use in 2010 and thus part of the baseline.

Manufacturers will be able to carry over innovative technology credits from Phase 1 into Phase 2, subject to the same restrictions as other credits. Manufacturers will also be able to carry over the improvement factor (not the credit value) of a technology, if certain criteria are met. The agencies will require documentation for all off-cycle requests similar to those required by EPA for its light-duty GHG program.

Additionally, the agencies will not grant any off-cycle credits for crash avoidance technologies. The agencies will also require manufacturers to consider the safety of off-cycle technologies and will request a safety assessment from the manufacturer for all off-cycle technologies.

Similar principles apply to off-cycle credits in this heavy-duty Phase 2 program as under the light-duty vehicle rules. Thus, technologies which are part of the basis of a Phase 2 standard would not be eligible for off-cycle credits. Their benefits have been accounted for in developing the stringency of the Phase 2 standard, as have their costs. See 77 FR 62835 (October 15, 2012). In addition, technologies which are integral or inherent to the basic vehicle design and are recognized in GEM or under the FTP (for pickups and vans), including engine, transmission, mass reduction, passive aerodynamic design, and base tires, will not be eligible for off-cycle credits. 77 FR 62836. Start Printed Page 73500Technologies integral or inherent to basic vehicle design are fully functioning and are thus recognized in GEM, or operate over the entirety of the FTP/HFET and therefore are adequately captured by the test procedure.

Just as some technologies that were considered off-cycle for Phase 1 are being adopted as primary technologies in Phase 2 on whose performance standard stringency is calculated, the agencies may revise the regulation in a future rulemaking to create a more direct path to recognize technologies currently considered off-cycle. For example, although we are including specific provisions to recognize certain electrified accessories, recognizing others would require the manufacturer to go through the off-cycle process. However, it is quite possible that the agencies could gather sufficient data to allow us to adopt specific provisions in a future rulemaking to recognize other accessories in a simpler manner. Because such a change would merely represent a simpler way to receive the same credit as could be obtained under the regulations being adopted today (rather than a change in stringency), it would not require us to reconsider the standards.

(d) Alternative Fuels and Electric Vehicles

The agencies will largely continue the Phase 1 approach for engines and vehicles fueled by fuels other than gasoline and diesel.[73] Phase 1 engine emission standards applied uniquely for gasoline-fueled and diesel-fueled engines. The regulations in 40 CFR part 86 implement these distinctions for alternative fuels by dividing engines into Otto-cycle and Diesel-cycle technologies based on the combustion cycle of the engine. However, as proposed, the agencies are making a small change that is described in Section II. Under this change, we will require manufacturers to divide their natural gas engines into primary intended service classes, like the current requirement for compression-ignition engines. Any alternative fuel-engine qualifying as a heavy heavy-duty engine will be subject to all the emission standards and other requirements that apply to compression-ignition engines. Note that this small change in approach will also apply with respect to EPA's criteria pollutant program.

We are also applying the Phase 2 standards at the vehicle tailpipe. That is, compliance is based on vehicle fuel consumption and GHG emission reductions, and does not reflect any so-called lifecycle emission properties. The agencies have explained why it is reasonable that the heavy-duty standards be fuel neutral in this manner and adhere to this reasoning here. See 76 FR 57123; see also 77 FR 51705 (August 24, 2012) and 77 FR 51500 (August 27, 2012). In particular, EPA notes that there is a separate, statutorily-mandated program under the Clean Air Act which encourages use of renewable fuels in transportation fuels, including renewable fuel used in heavy-duty diesel engines. This program considers lifecycle greenhouse gas emissions compared to petroleum fuel. NHTSA notes that the fuel efficiency standards are necessarily tailpipe-based, and that a lifecycle approach would likely render it impossible to harmonize the fuel efficiency and GHG emission standards, to the great detriment of our goal of achieving a coordinated program. 77 FR 51500-51501; see also 77 FR 51705 (similar finding by EPA); see also Section I.F.(1)(a) below, Section 1.8 of the RTC, and Section XI.B.

The agencies received mixed comments on this issue. Many commenters supported the proposed approach, generally agreeing with the agencies' arguments. However, some other commenters opposed this approach. Opposing commenters generally fell into two categories:

  • Commenters concerned that upstream emissions of methane occurring during the production and distribution of natural gas would offset some or all of the GHG emission reductions observed at the tailpipe.
  • Commenters concerned that tailpipe-only standards ignore the GHG benefits of using renewable fuels.

The agencies are not issuing rules that effectively would turn these rules into a fuel program, rather than an emissions reduction and fuel efficiency program. Nor will the agencies disharmonize the program by having GHG standards reflect upstream emissions having no relation to fuel efficiency. See e.g. 77 FR 51500-51501; see also 77 FR 51705. We thus will continue to measure compliance at the tailpipe. Issues relating to whether to consider in the emission standards upstream emissions related to natural gas exploration and production are addressed in detail in Section XI below. It is sufficient to state here that the agencies carefully investigated the potential use of natural gas in the heavy-duty sector and the impacts of such use. We do not believe that the use of natural gas is likely to become a major fuel source for heavy-duty vehicles during the Phase 2 time frame. Thus, since we project natural gas vehicles to have little impact on both overall GHG emissions and fuel consumption during the Phase 2 time frame, the agencies see no need to make fundamental changes to the Phase 1 approach for natural gas engines and vehicles.

The agencies note further that a consequence of the tailpipe-based approach is that the agencies will treat vehicles powered by electricity the same as in Phase 1. In Phase 1, EPA treated all electric vehicles as having zero tailpipe emissions of CO2, CH4, and N2 O (see 40 CFR 1037.150(f)). Similarly, NHTSA adopted regulations in Phase 1 that set the fuel consumption standards based on the fuel consumed by the vehicle. The agencies also did not require emission testing for electric vehicles in Phase 1. The agencies considered the potential unintended consequence of not accounting for upstream emissions from the charging of heavy-duty electric vehicles. In our reassessment for Phase 2, we have found only one all-electric heavy-duty vehicle manufacturer that has certified through 2016. As we look to the future, we project limited adoption of all-electric vehicles into the market. Therefore, we believe that this provision is still appropriate. Unlike the 2017-2025 light-duty rule, which included a cap whereby upstream emissions would be counted after a certain volume of sales (see 77 FR 62816-62822), we believe there is no need to establish a cap for heavy-duty vehicles because of the small likelihood of significant production of EV technologies in the Phase 2 timeframe. Commenters specifically addressing electric vehicles generally supported the agencies' proposal. However, some commenters did support accounting for emissions from the generation of electricity in the broader context of supporting full life-cycle analysis. As noted above, and in more detail in Section I.F.(2)(f) as well as Section 1.8 of the RTC, the agencies are not predicating the standards on a full life-cycle approach.

(e) Phase 1 Interim Provisions

EPA adopted several flexibilities for the Phase 1 program (40 CFR 86.1819-14(k), 1036.150 and 1037.150) as interim provisions. Because the existing regulations do not have an end date for Phase 1, most of these provisions did not have an explicit end date. NHTSA adopted similar provisions. With few exceptions, the agencies are not continuing these provisions for Phase 2. These will generally remain in effect for the Phase 1 program. In particular, the agencies note that we are not continuing the blanket exemption for small Start Printed Page 73501manufacturers. Instead, in Phase 2 the agencies are providing more targeted relief for these entities.

(f) In-Use Standards and Recall

Section 202(a)(1) of the CAA specifies that EPA is to adopt emissions standards that are applicable for the useful life of the vehicle and for the engine. EPA finalized in-use standards for the Phase 1 program, whereas NHTSA's rules do not include these standards. For the Phase 2 program, EPA will carry-over its in-use provisions, and NHTSA is adopting EPA's useful life requirements for its vehicle and engine fuel consumption standards to ensure manufacturers consider in the design process the need for fuel efficiency standards to apply for the same duration and mileage as EPA standards. If EPA determines a manufacturer fails to meet its in-use standards, civil penalties may be assessed.

CAA section 207(c)(1) requires “the manufacturer” to remedy certain in-use problems. The remedy process is to recall the nonconforming vehicles and bring them into conformity with the standards and the certificate. The regulations for this process are in 40 CFR part 1068, subpart F. EPA is also adopting regulatory text addressing recall obligations for component manufacturers and other non-certifying manufacturers. We note that the CAA does not limit this responsibility to certificate holders, consistent with the definition of a “manufacturer” as “any person engaged in the manufacturing or assembling of new motor vehicles, new motor vehicle engines, new nonroad vehicles or new nonroad engines, or importing such vehicles or engines for resale, or who acts for and is under the control of any such person in connection with the distribution of new motor vehicles, new motor vehicle engines, new nonroad vehicles or new nonroad engines, but shall not include any dealer with respect to new motor vehicles, new motor vehicle engines, new nonroad vehicles or new nonroad engines received by him in commerce.”

As discussed in Section I.E.(1) below, this definition was not intended to restrict the definition of “manufacturer” to a single person per vehicle. Under EPA regulations, we can require any person meeting the definition of manufacturer for a nonconforming vehicle to participate in a recall. However, we would normally presume the certificate holder to have the primary responsibility.

EPA requested comment on adding regulatory text that would explicitly apply these provisions to tire manufacturers. Comments from the tire industry generally opposed this noting that they are not the manufacturer of the vehicle. These comments are correct that tires are not incomplete vehicles and hence that the recall authority does not apply for companies that only manufacture the tires. However, EPA remains of the view that in the event that vehicles (e.g. trailers) do not conform to the standards in-use due to nonconforming tires, tire manufacturers would have a role to play in remedying the problem. In this (hypothetical) situation, a tire manufacturer would not only have produced the part in question, but in the case of a trailer manufacturer or other small vehicle manufacturer, would have significantly more resources and knowledge regarding how to address (and redress) the problem. Accordingly, EPA would likely require that a component manufacturer responsible for the nonconformity assist in the recall to an extent and in a manner consistent with the provisions of CAA section 208(a). This section specifies that component and part manufacturers “shall establish and maintain records, perform tests where such testing is not otherwise reasonably available under this part and part C of this subchapter (including fees for testing), make reports and provide information the Administrator may reasonably require to determine whether the manufacturer or other person has acted or is acting in compliance with this part and part C of this subchapter and regulations thereunder, or to otherwise carry out the provision of this part and part C of this subchapter. . .”. Any such action would be considered on a case-by-case basis, adapted to the particular circumstances at the time.

(g) Vehicle Labeling

EPA proposed to largely continue the Phase 1 engine and vehicle labeling requirements, but to eliminate the requirement for tractor and vocational vehicle manufacturers to list emission control on the label. The agencies consider it crucial that authorized compliance inspectors are able to identify whether a vehicle is certified, and if so whether it is in its certified condition. To facilitate this identification in Phase 1, EPA adopted labeling provisions for tractors that included several items. The Phase 1 tractor label must include the manufacturer, vehicle identifier such as the Vehicle Identification Number (VIN), vehicle family, regulatory subcategory, date of manufacture, compliance statements, and emission control system identifiers (see 40 CFR 1037.135). EPA proposed to apply parallel requirements for trailers.

In Phase 1, the emission control system identifiers are limited to vehicle speed limiters, idle reduction technology, tire rolling resistance, some aerodynamic components, and other innovative and advanced technologies. However, the number of emission control systems for greenhouse gas emissions in Phase 2 has increased significantly for tractors and vocational vehicles. For example, all aspects of the engine transmission and drive axle; accessories; tire radius and rolling resistance; wind averaged drag; predictive cruise control; idle reduction technologies; and automatic tire inflation systems are controls which can be evaluated on-cycle in Phase 2 (i.e. these technologies' performance can now be input to GEM), but could not be in Phase 1. Due to the complexity in determining greenhouse gas emissions in Phase 2, the agencies do not believe that we can unambiguously determine whether or not a vehicle is in a certified condition through simply comparing information that could be made available on an emission control label with the components installed on a vehicle. Therefore, EPA proposed to remove the requirement to include the emission control system identifiers required in 40 CFR 1037.135(c)(6) and in Appendix III to 40 CFR part 1037 from the emission control labels for vehicles certified to the Phase 2 standards. The agencies received comments on the emission control labels from Navistar, which supported the elimination of the emission control information from the vehicle GHG label.

Although we are largely finalizing the proposed labeling requirements, we remain interested in finding a better approach for labeling. Under the agencies' existing authorities, manufacturers must provide detailed build information for a specific vehicle upon our request. Our expectation is that this information should be available to us via email or other similar electronic communication on a same-day basis, or within 24 hours of a request at the latest. The agencies have started to explore ideas that would provide inspectors with an electronic method to identify vehicles and access on-line databases that would list all of the engine-specific and vehicle-specific emissions control system information. We believe that electronic and Internet technology exists today for using scan tools to read a bar code or radio frequency identification tag affixed to a vehicle that could then lead to secure on-line access to a database of manufacturers' detailed vehicle and Start Printed Page 73502engine build information. Our exploratory work on these ideas has raised questions about the level of effort that would be required to develop, implement and maintain an information technology system to provide inspectors real-time access to this information. We have also considered questions about privacy and data security. We requested comment on the concept of electronic labels and database access, including any available information on similar systems that exist today and on burden estimates and approaches that could address concerns about privacy and data security.

Although we are not finalizing such a program in this rulemaking, we remain very interested in the use of electronic labels that could be used by the agencies to access vehicle information and may pursue these in a future rulemaking. Such a rulemaking would likely consider the feasibility of accessing dynamic link libraries in real-time to view each manufacturer's build records (and perhaps pending orders). The agencies envision that this could be very useful for our inspectors by providing them access to the build information by VIN to confirm that each vehicle has the proper emission control features.

(h) Model Year Definition

The agencies proposed to continue the Phase definitions of “model year” for compliance with GHG emissions and fuel efficiency standards. However, in response to comments, the agencies are revising the definition slightly for Phase 2 tractors and vocational vehicles to match the model years of the engines installed in them. The revised definition generally sets the vehicle model year to be the calendar year of manufacture, but allows the vehicle manufacturer the option to select the prior year if the vehicle uses an engine manufactured in the prior model year.[74] Because Phase 2 vehicle standards are based in part on engine performance, some commenters stated that the engine model year should dictate the vehicle's GHG and fuel efficiency compliance model year, and that the emissions and fuel efficiency compliance model year should be presented on the vehicle emissions label. This would allow manufacturers to market a vehicle and certify it to NHTSA's safety standards based on the standards applicable on the date of manufacture, but certify the vehicle for GHG emissions and fuel efficiency purposes based on the engine model compliance year. For example, a 2023 model year tractor might have a 2022 model year engine in it. The tractor would be marketed as a model year 2023 tractor, certified as complying with NHTSA's safety standards applicable at the time when certifying the vehicle, but would have an “emissions and fuel efficiency compliance model year” of 2022 for purposes of emissions and fuel efficiency standards. In today's action, NHTSA and EPA are finalizing standards that allow for the use of an “emissions and fuel efficiency compliance model year.” This is consistent with past program practice, in which certain manufacturers have been able to reclassify tractors to the previous model year for emissions purposes when the tractors use engines from the previous model year.

(2) Phase 2 Standards

This section briefly summarizes the Phase 2 standards for each category and identifies the technologies that the agencies project will be needed to meet the standards. Given the large number of different regulatory categories and model years for these standards, the actual numerical standards are not listed. Readers are referred to Sections II through IV for the tables of standards.

(a) Summary of the Engine Standards

The agencies are continuing the basic Phase 1 structure for the Phase 2 engine standards. There will be separate standards and test cycles for tractor engines, vocational diesel engines, and vocational gasoline engines. However, as described in Section II, we are adopting a revised test cycle for tractor engines to better reflect actual in-use operation. After consideration of comments, including those specifically addressing whether the agencies should adopt an alternative with accelerated stringency targets, the agencies are adopting engine standards that can generally be characterized as more stringent than the proposed alternative.

Specifically, for diesel tractor engines, the agencies are adopting standards for MY 2027 that are more stringent than the preferred alternative from the proposal, and require reductions in CO2 emissions and fuel consumption that are 5.1 percent better than the 2017 baseline for tractor engines.[75] We are also adopting standards for MY 2021 and MY 2024, requiring reductions in CO2 emissions and fuel consumption of 1.8 to 4.2 percent better than the 2017 baseline tractor engines. For vocational diesel engines, the new standards will require reductions of 2.3, 3.6, and 4.2 percent in MYs 2021, 2024, and 2027, respectively. These levels are more stringent than the proposed standards for these same MYs, and approximately as stringent in MY 2021 and MY 2024 as the Alternative 4 standards discussed at proposal.[76]

The agencies project that these reductions will be maximum feasible and reasonable for diesel engines based on technological changes that will improve combustion and reduce energy losses. For most of these improvements, the agencies project (i.e., the agencies have set out a potential, but by no means mandatory, compliance path) that manufacturers will begin applying improvements to about 45 percent of their heavy-duty engines by 2021, and ultimately apply them to about 95 percent of their heavy-duty engines by 2024. However, for some of these improvements we project more limited application rates. In particular, we project a more limited use of waste exhaust heat recovery systems in 2027, projecting that about 10 percent of tractor engines will have turbo-compounding systems, and an additional 25 percent of tractor engines will employ Rankine-cycle waste heat recovery. We do not project that turbo-compounding or Rankine-cycle waste heat recovery technology will be utilized in vocational engines due to vocational vehicle drive cycles under which these technologies would not show significant benefit, and also due to low sales volumes, limiting the ability to invest in newer technologies for these vehicles.

As described in Section III.D.(1)(b)(i), the agencies project that some engine manufacturers will be able to achieve larger reductions for at least some of their tractor engines. So in developing the tractor vehicle standards, we projected slightly better fuel efficiency for the average tractor engine than is required by the engine standards. We are projecting that similar over-compliance will occur for heavy heavy-duty vocational engines.

For gasoline vocational engines, we are not adopting more stringent engine standards. Gasoline engines used in Start Printed Page 73503vocational vehicles are generally the same engines as are used in the complete HD pickups and vans in the Class 2b and 3 weight categories, although the operational demands of vocational vehicles often require use of the largest, most powerful SI engines, so that some engines fitted in complete pickups and vans are not appropriate for use in vocational vehicles. Given the relatively small sales volumes for gasoline-fueled vocational vehicles, manufacturers typically cannot afford to invest significantly in developing separate technology for these vocational vehicle engines. Thus, we project that in general, vocational gasoline engines will incorporate much of the technology that will be used to meet the pickup and van chassis standards, and this will result in some real world reductions in CO2 emissions and fuel consumption. The agencies received many comments suggesting that technologies be applied to increase the stringency of the SI engine standard, which technologies in fact are already presumed to be adopted at 100 percent to meet the MY 2016 engine standard. The commenters did not identify any additional engine technologies that are not already fully considered by the agencies in setting the MY 2016 engine standard, that could be recognized over the HD SI Engine FTP test cycle. We did, however, consider some additional technologies recommended by commenters, which can be recognized over the GEM vehicle cycles. As a result, the Phase 2 vehicle standards for gasoline-fueled vocational vehicles are predicated on adoption of engine technologies beyond what is required to meet the separate engine standard, those additional technologies being advanced engine friction reduction and cylinder deactivation. As described in Section V, we are projecting these technologies to improve fuel consumption over the GEM cycles by nearly one percent in MY 2021, MY 2024, and MY 2027. In other words, this improvement is reflected in the vehicle standards rather than in the engine standards. To the extent any SI engines do not incorporate the projected engine technologies, manufacturers of gasoline-fueled vocational vehicles would need to achieve equivalent reductions from some other technology to meet the GEM-based vehicle standards. The engine standards are summarized in Table I-4.

Table I-4—Summary of Phase 1 and Phase 2 Requirements for Engines in Combination Tractors and Vocational Vehicles

Phase 1 programFinal 2027 standards
Covered in this categoryEngines installed in tractors and vocational chassis.
Share of HDV fuel consumption and GHG emissionsCombination tractors and vocational vehicles account for approximately 85 percent of fuel use and GHG emissions in the heavy duty truck sector.
Per vehicle fuel consumption and CO2 improvement5%-9% improvement over MY 2010 baseline, depending vehicle application. Improvements are in addition to improvements from tractor and vocational vehicle standards4%-5% improvement over MY 2017 for diesel engines. Note that improvements are captured in complete vehicle tractor and vocational vehicle standards, so that engine improvements and the vehicle improvement shown below are not additive.
Form of the standardEPA: CO2 grams/horsepower-hour and NHTSA: Gallons of fuel/horsepower-hour.
Example technology options available to help manufacturers meet standardsCombustion, air handling, friction and emissions after-treatment technology improvementsFurther technology improvements and increased use of all Phase 1 technologies, plus waste heat recovery systems for tractor engines (e.g., turbo-compound and Rankine-cycle).
FlexibilitiesABT program which allows emissions and fuel consumption credits to be averaged, banked, or traded (five year credit life). Manufacturers allowed to carry-forward credit deficits for up to three model years. Interim incentives for advanced technologies, recognition of innovative (off-cycle) technologies not accounted for by the HD Phase 1 test procedures, and credits for certifying earlySame ABT and off-cycle program as Phase 1. Adjustment factor of 1.36 for credits carried forward from Phase 1 to Phase 2 for SI and LHD CI engines due to change in useful life. Revised multipliers for Phase 2 advanced technologies. No Phase 2 early credit multipliers.

(b) Summary of the Tractor Standards

As explained in Section III, the agencies will largely continue the structure of the Phase 1 tractor program, but adopt new standards and update test procedures, as summarized in Table I-6. The tractor standards for MY 2027 will achieve up to 25 percent lower CO2 emissions and fuel consumption than a 2017 model year Phase 1 tractor. The agencies project that the 2027 tractor standards could be met through improvements in the:

  • Engine [77] (including some use of waste heat recovery systems)
  • Transmission
  • Driveline
  • Aerodynamic design
  • Tire rolling resistance
  • Idle performance
  • Other accessories of the tractor.

The agencies have enhanced the Phase 2 GEM vehicle simulation tool to recognize these technologies, as described in Section II.C. The agencies' evaluation shows that some of these technologies are available today, but have very low adoption rates on current vehicles, while others will require some lead time for development and deployment. In addition to the proposed alternative for tractors, the agencies solicited comment on an alternative that reached similar ultimate stringencies, but at an accelerated pace.

We have also determined that there is sufficient lead time to introduce many of these tractor and engine technologies into the fleet at a reasonable cost starting in the 2021 model year. The Start Printed Page 735042021 model year standards for combination tractors and engines will achieve up to 14 percent lower CO2 emissions and fuel consumption than a 2017 model year Phase 1 tractor, the 2024 model year standards will achieve up to 20 percent lower CO2 emissions and fuel consumption, and as already noted, the 2027 model year standards will achieve up to 25 percent lower CO2 emissions and fuel consumption.

In addition to the CO2 emission standards for tractors, EPA is adopting new particulate matter (PM) standards which effectively limit which diesel fueled auxiliary power units (APUs) can be used as emission control devices to reduce main engine idling in tractors, as shown in Table I-5. Additional details are discussed in Section III.C.3.

Table I-5—PM Standards Related to Diesel APUs

Tractor MYPM emission standard (g/kW-hr)Expected control technology
2018-20230.15In-cylinder PM control.
20240.02DPF.

Table I-6—Summary of Phase 1 and Phase 2 Requirements for Class 7 and Class 8 Combination Tractors

Phase 1 programFinal 2027 standards
Covered in this categoryTractors that are designed to pull trailers and move freight.
Share of HDV fuel consumption and GHG emissionsCombination tractors and their engines account for approximately sixty percent of fuel use and GHG emissions in the heavy duty vehicle sector.
Per vehicle fuel consumption and CO2 improvement10%-23% improvement over MY 2010 baseline, depending on tractor category. Improvements are in addition to improvements from engine standards19%-25% improvement over tractors meeting the MY 2017 standards.
Form of the standardEPA: CO2 grams/ton payload mile and NHTSA: Gallons of fuel/1,000 ton payload mile.
Example technology options available to help manufacturers meet standardsAerodynamic drag improvements; low rolling resistance tires; high strength steel and aluminum weight reduction; extended idle reduction; and speed limitersFurther technology improvements and increased use of all Phase 1 technologies, plus engine improvements, improved transmissions and axles, tire pressure systems, and predictive cruise control (depending on tractor type).
FlexibilitiesABT program which allows emissions and fuel consumption credits to be averaged, banked, or traded (five year credit life). Manufacturers allowed to carry-forward credit deficits for up to three model years. Interim incentives for advanced technologies, recognition of innovative (off-cycle) technologies not accounted for by the HD Phase 1 test procedures, and credits for certifying earlySame ABT and off-cycle program as Phase 1. Revised multipliers for Phase 2 advanced technologies.

(c) Summary of the Trailer Standards

The final rules contain a set of GHG emission and fuel consumption standards for manufacturers of new trailers that are used in combination with tractors. These standards will significantly reduce CO2 and fuel consumption from combination tractor-trailers nationwide over a period of several years. As described in Section IV, there are numerous aerodynamic and tire technologies available to manufacturers to achieve these standards. Many of these technologies have already been introduced into the market through EPA's voluntary SmartWay program and California's tractor-trailer greenhouse gas requirements.

The agencies are adopting Phase 2 standards that will phase-in beginning in MY 2018 and be fully phased-in by 2027. These standards are predicated on use of aerodynamic and tire improvements, with trailer OEMs making incrementally greater improvements in MYs 2021 and 2024 as standard stringency increases in each of those model years. EPA's GHG emission standards will be mandatory beginning in MY 2018, while NHTSA's fuel consumption standards will be voluntary beginning in MY 2018, and be mandatory beginning in MY 2021. In general, the trailer standards being finalized apply only for box vans, flatbeds, tankers, and container chassis.

As described in Section XIV.D and Chapter 12 of the RIA, the agencies are adopting special provisions to minimize the impacts on small business trailer manufacturers. These provisions have been informed by and are largely consistent with recommendations from the SBAR Panel that EPA conducted pursuant to section 609(b) of the Regulatory Flexibility Act (RFA). Broadly, these provisions provide additional lead time for small business manufacturers, as well as simplified testing and compliance requirements. The agencies also are not finalizing standards for various trailer types, including most specialty types of non-box trailers. Excluding these specialty trailers also reduces the impacts on small businesses.Start Printed Page 73505

Table I-7—Summary of Phase 2 Requirements for Trailers

Phase 1 programFinal 2027 standards
Covered in this categoryAll lengths of dry vans, refrigerated vans, tanks, flatbeds, and container chassis hauled by low, mid, and high roof day and sleeper cab tractors.
Share of HDV fuel consumption and GHG emissionsTrailers are modeled together with combination tractors and their engines. Together, they account for approximately sixty percent of fuel use and GHG emissions in the heavy duty truck sector.
Per vehicle fuel consumption and CO2 improvementN/ABetween 3% and 9% improvement over MY 2018 baseline, depending on the trailer type.
Form of the standardN/AEPA: CO2 grams/ton payload mile and NHTSA: Gallons/1,000 ton payload mile.
Example technology options available to help manufacturers meet standardsN/ALow rolling resistance tires and tire pressure systems for most trailers, plus weight reduction and aerodynamic improvements such as side and rear fairings, gap closing devices, and undercarriage treatment for box vans (e.g., dry and refrigerated).
FlexibilitiesN/AOne year delay in implementation for small businesses, trailer manufacturers may use pre-approved aerodynamic data in lieu of additional testing, averaging program available in MY 2027 for manufacturers of dry and refrigerated box vans.

(d) Summary of the Vocational Vehicle Standards

As explained in Section V, the agencies are adopting new vocational vehicle standards that expand upon the Phase 1 Program. These new standards reflect further subcategorization from Phase 1, with separate standards based on mode of operation: Urban, regional, and multi-purpose. The agencies are also adopting optional separate standards for emergency vehicles and other custom chassis vehicles.

The agencies project that the vocational vehicle standards could be met through improvements in the engine, transmission, driveline, lower rolling resistance tires, workday idle reduction technologies, weight reduction, and some application of hybrid technology. These are described in Section V of this Preamble and in Chapter 2.9 of the RIA. These MY 2027 standards will achieve up to 24 percent lower CO2 emissions and fuel consumption than MY 2017 Phase 1 standards. The agencies are also making revisions to the compliance program for vocational vehicles. These include: The addition of two idle cycles that will be weighted along with the other drive cycles for each vocational vehicle; and revisions to Phase 2 GEM to recognize improvements to the engine, transmission, and driveline.

Similar to the tractor program, we have determined that there is sufficient lead time to introduce many of these new technologies into the fleet starting in MY 2021. Therefore, we are adopting new standards for MY 2021 and 2024. Based on our analysis, the MY 2021 standards for vocational vehicles will achieve up to 12 percent lower CO2 emissions and fuel consumption than a MY 2017 Phase 1 vehicle, on average, and the MY 2024 standards will achieve up to 20 percent lower CO2 emissions and fuel consumption.

In Phase 1, EPA adopted air conditioning (A/C) refrigerant leakage standards for tractors, as well as for heavy-duty pickups and vans, but not for vocational vehicles. For Phase 2, EPA believes that it will be feasible to apply similar A/C refrigerant leakage standards for vocational vehicles, beginning with the 2021 model year. The certification process for vocational vehicles to certify low-leakage A/C components is identical to that already required for tractors.

Table I-8—Summary of Phase 1 and Phase 2 Requirements for Vocational Vehicle Chassis

Phase 1 programFinal 2027 standard
Covered in this categoryClass 2b—8 chassis that are intended for vocational services such as delivery vehicles, emergency vehicles, dump truck, tow trucks, cement mixer, refuse trucks, etc., except those qualified as off-highway vehicles.
Because of sector diversity, vocational vehicle chassis are segmented into Light, Medium and Heavy Heavy-Duty vehicle categories and for Phase 2 each of these segments are further subdivided using three duty cycles: Regional, Multi-purpose, and Urban.
Share of HDV fuel consumption and GHG emissionsVocational vehicles account for approximately 17 percent of fuel use and GHG emissions in the heavy duty truck sector categories.
Per vehicle fuel consumption and CO2 improvement2% improvement over MY 2010 baseline. Improvements are in addition to improvements from engine standardsUp to 24% improvement over MY 2017 standards.
Form of the standardEPA: CO2 grams/ton payload mile and NHTSA: Gallons of fuel/1,000 ton payload mile.
Example technology options available to help manufacturers meet standardsLow rolling resistance tiresFurther technology improvements and increased use of Phase 1 technologies, plus improved engines, transmissions and axles, weight reduction, hybrids, and workday idle reduction systems.
Start Printed Page 73506
FlexibilitiesABT program which allows emissions and fuel consumption credits to be averaged, banked, or traded (five year credit life). Manufacturers allowed to carry-forward credit deficits for up to three model years. Interim incentives for advanced technologies, recognition of innovative (off-cycle) technologies not accounted for by the HD Phase 1 test procedures, and credits for certifying earlySame ABT and off-cycle program as Phase 1. Adjustment factor of 1.36 for credits carried forward from Phase 1 to Phase 2 due to change in useful life. Revised multipliers for Phase 2 advanced technologies. No Phase 2 early credit multipliers. Chassis intended for emergency vehicles, cement mixers, coach buses, school buses, transit buses, refuse trucks, and motor homes may optionally use application-specific Phase 2 standards using a simplified version of GEM.

(e) Summary of the Heavy-Duty Pickup and Van Standards

The agencies are adopting new Phase 2 GHG emission and fuel consumption standards for heavy-duty pickups and vans that will be applied in largely the same manner as the Phase 1 standards. These standards are based on the extensive use of most known and proven technologies, and could result in some use of mild or strong hybrid powertrain technology. These standards will commence in MY 2021. By 2027, these standards are projected to be 16 percent more stringent than the 2018-2019 standards.

Table I-9—Summary of Phase 1 and Phase 2 Requirements for HD Pickups and Vans

Phase 1 programFinal 2027 standard
Covered in this categoryClass 2b and 3 complete pickup trucks and vans, including all work vans and 15-passenger vans but excluding 12-passenger vans which are subject to light-duty standards.
Share of HDV fuel consumption and GHG emissionsHD pickups and vans account for approximately 23% of fuel use and GHG emissions in the heavy duty truck sector.
Per vehicle fuel consumption and CO2 improvement15% improvement over MY 2010 baseline for diesel vehicles, and 10% improvement for gasoline vehicles16% improvement over MY 2018-2019 standards.
Form of the standardPhase 1 standards are based upon a “work factor” attribute that combines truck payload and towing capabilities, with an added adjustment for 4-wheel drive vehicles. There are separate target curves for diesel-powered and gasoline-powered vehicles. The Phase 2 standards are based on the same approach.
Example technology options available to help manufacturers meet standardsEngine improvements, transmission improvements, aerodynamic drag improvements, low rolling resistance tires, weight reduction, and improved accessoriesFurther technology improvements and increased use of all Phase 1 technologies, plus engine stop-start, and powertrain hybridization (mild and strong).
FlexibilitiesTwo optional phase-in schedules; ABT program which allows emissions and fuel consumption credits to be averaged, banked, or traded (five year credit life). Manufacturers allowed to carry-forward credit deficits for up to three model years. Interim incentives for advanced technologies, recognition of innovative (off-cycle) technologies not accounted for by the HD Phase 1 test procedures, and credits for certifying earlySame as Phase 1, with phase-in schedule based on year-over-year increase in stringency. Same ABT and off-cycle program as Phase 1. Adjustment factor of 1.25 for credits carried forward from Phase 1 to Phase 2 due to change in useful life. Revised multipliers for Phase 2 advanced technologies. No Phase 2 early credit multipliers.

Similar to Phase 1, the agencies are adopting for Phase 2 a set of continuous equation-based standards for HD pickups and vans. Please refer to Section VI for a description of these standards, including associated tables and figures.

D. Summary of the Costs and Benefits of the Final Rules

This section summarizes the projected costs and benefits of the NHTSA fuel consumption and EPA GHG emission standards. See Sections VII through IX and the RIA for additional details about these projections.

For these rules, the agencies used two analytical methods for the heavy-duty pickup and van segment by employing both DOT's CAFE model and EPA's MOVES model. The agencies used EPA's MOVES model to estimate fuel consumption and emissions impacts for tractor-trailers (including the engine that powers the tractor), and vocational vehicles (including the engine that powers the vehicle). Additional calculations were performed to determine corresponding monetized program costs and benefits. For heavy-duty pickups and vans, the agencies performed separate analyses, which we refer to as “Method A” and “Method B.” In Method A, a new version of the CAFE model was used to project a pathway the industry could use to comply with each regulatory alternative and the estimated effects on fuel consumption, emissions, benefits and costs. In Method B, the CAFE model from the NPRM was used to project a pathway the industry could use to comply with each regulatory alternative, along with resultant impacts on per-vehicle costs. However, the MOVES model was used to calculate corresponding changes in total fuel consumption and annual emissions for pickups and vans in Method B. Additional calculations were performed to determine corresponding Start Printed Page 73507monetized program costs and benefits. NHTSA considered Method A as its central analysis and Method B as a supplemental analysis. EPA considered the results of Method B. The agencies concluded that these methods led the agencies to the same conclusions and the same selection of these standards. See Section VII for additional discussion of these two methods.

(1) Reference Case Against Which Costs and Benefits Are Calculated

The No Action Alternatives for today's analysis, alternatively referred to as the “baselines” or “reference cases,” assume that the agencies did not issue new rules regarding MD/HD fuel efficiency and GHG emissions. These are the baselines against which costs and benefits for these standards are calculated. The reference cases assume that model year 2018 engine, tractor, vocational vehicle, and HD pickup and van standards will be extended indefinitely and without change. They also assume that no new standards would be adopted for trailers.

The agencies recognize that if these Phase 2 standards had not been adopted, manufacturers would nevertheless continue to introduce new heavy-duty vehicles in a competitive market that responds to a range of factors, and manufacturers might have continued to improve technologies to reduce heavy-duty vehicle fuel consumption. Thus, as described in Section VII, both agencies fully analyzed these standards and the regulatory alternatives against two reference cases. The first case uses a baseline that projects no improvement in new vehicles in the absence of new Phase 2 standards, and the second uses a more dynamic baseline that projects some significant improvements in vehicle fuel efficiency. NHTSA considered its primary analysis to be based on the dynamic baseline, where certain cost-effective technologies are assumed to be applied by manufacturers to improve fuel efficiency beyond the Phase 1 requirements in the absence of new Phase 2 standards. EPA considered both reference cases. The results for all of the regulatory alternatives relative to both reference cases, derived via the same methodologies discussed in this section, are presented in Section X of the Preamble.

The agencies received limited comments on these reference cases. Some commenters expressed support for a flat baseline in the context of the need for the regulations, arguing that little improvement would occur without the regulations. Others supported the less dynamic baseline because they believe it more fully captures the costs. A number of commenters expressed that purchasers are willing to and do pay for fuel efficiency improving technologies, provided the cost for the technology is paid back through fuel savings within a certain period of time; this is the premise for a dynamic baseline. Some commenters thought it reasonable that the agencies consider both baselines given the uncertainty in this area. No commenters opposed the consideration of both baselines.

The agencies have continued to analyze two different baselines for the final rules because we recognize that there are a number of factors that create uncertainty in projecting a baseline against which to compare the future effects of this action and the remaining alternatives. The composition of the future fleet—such as the relative position of individual manufacturers and the mix of products they each offer—cannot be predicted with certainty at this time. Additionally, the heavy-duty vehicle market is diverse, as is the range of vehicle purchasers. Heavy-duty vehicle manufacturers have reported that their customers' purchasing decisions are influenced by their customers' own determinations of minimum total cost of ownership, which can be unique to a particular customer's circumstances. For example, some customers (e.g., less-than-truckload or package delivery operators) operate their vehicles within a limited geographic region and typically own their own vehicle maintenance and repair centers within that region. These operators tend to own their vehicles for long time periods, sometimes for the entire service life of the vehicle. Their total cost of ownership is influenced by their ability to better control their own maintenance costs, and thus they can afford to consider fuel efficiency technologies that have longer payback periods, outside of the vehicle manufacturer's warranty period. Other customers (e.g., truckload or long-haul operators) tend to operate cross-country, and thus must depend upon truck dealer service centers for repair and maintenance. Some of these customers tend to own their vehicles for about four to seven years, so that they typically do not have to pay for repair and maintenance costs outside of either the manufacturer's warranty period or some other extended warranty period. Many of these customers tend to require seeing evidence of fuel efficiency technology payback periods on the order of 18 to 24 months before seriously considering evaluating a new technology for potential adoption within their fleet (NAS 2010, Roeth et al. 2013, and Klemick et al. 2014). Purchasers of HD pickups and vans wanting better fuel efficiency tend to demand that fuel consumption improvements pay back within approximately one to three years, but some HD pickup and van owners accrue relatively few vehicle miles traveled per year, such that they may be less likely to adopt new fuel efficiency technologies, while other owners who use their vehicle(s) with greater intensity may be even more willing to pay for fuel efficiency improvements. Regardless of the type of customer, their determination of minimum total cost of ownership involves the customer balancing their own unique circumstances with a heavy-duty vehicle's initial purchase price, availability of credit and lease options, expectations of vehicle reliability, resale value and fuel efficiency technology payback periods. The degree of the incentive to adopt additional fuel efficiency technologies also depends on customer expectations of future fuel prices, which directly impacts customer payback periods. Purchasing decisions are not based exclusively on payback period, but also include the considerations discussed above and in Section X.A.1. For the baseline analysis, the agencies use payback period as a proxy for all of these considerations, and therefore the payback period for the baseline analysis is shorter than the payback period industry uses as a threshold for the further consideration of a technology. See Section X.A.1 of this Preamble and Chapter 11 of the RIA for a more detailed discussion of baselines. As part of a sensitivity analysis, additional baseline scenarios were also evaluated for HD pickups and vans, including baseline payback periods of 12, 18 and 24 months. See Section VI of this Preamble and Chapter 10 of the RIA for a detailed discussion of these additional scenarios.

(2) Costs and Benefits Projected for the Phase 2 Standards

The tables below summarize the benefits and costs for the program in two ways: First, from the perspective of a program designed to improve the Nation's energy security and to conserve energy by improving fuel efficiency and then from the perspective of a program designed to reduce GHG emissions. The individual categories of benefits and costs presented in the tables below are defined more fully and presented in more detail in Chapter 8 of the RIA.

Lifetime fuel savings, GHG reductions, benefits, costs and net benefits for model years 2018 through Start Printed Page 735082029 vehicles as presented below. This is consistent with the NPRM analysis and allows readers to compare the costs and benefits of the final program with those projected for the NPRM. It also includes for modeling purposes at least three model years for each standard.

Table I-10 shows benefits and costs for these standards from the perspective of a program designed to improve the Nation's energy security and conserve energy by improving fuel efficiency. From this viewpoint, technology costs occur when the vehicle is purchased. Fuel savings are counted as benefits that occur over the lifetimes of the vehicles produced during the model years subject to the Phase 2 standards as they consume less fuel.

Table I-10—Lifetime Fuel Savings, GHG Reductions, Benefits, Costs, and Net Benefits for Model Years 2018-2029 Vehicles Using Analysis Method A

[Billions of 2013$] ab

Category3% discount rate7% discount rate
Fuel Reductions (Billion Gallons)71.1-77.7
GHG reductions (MMT CO2 eq)959-1049
Vehicle Program: Technology and Indirect Costs, Normal Profit on Additional Investments23.7 to 24.416.1 to 16.6
Additional Routine Maintenance1.7 to 1.70.9 to 0.9
Congestion, Crashes, Fatalities and Noise from Increased Vehicle Use d3.1 to 3.21.8 to 1.9
Total Costs28.5 to 29.318.8 to 19.4
Fuel Savings (valued at pre-tax prices)149.1 to 163.079.7 to 87.0
Savings from Less Frequent Refueling3.0 to 3.21.6 to 1.7
Economic Benefits from Additional Vehicle Use5.4 to 5.53.4 to 3.5
Reduced Climate Damages from GHG Emissions c33.0 to 36.0
Reduced Health Damages from Non-GHG Emissions27.1 to 30.014.6 to 16.1
Increased U.S. Energy Security7.3 to 7.93.9 to 4.2
Total Benefits225 to 246136 to 149
Net Benefits197 to 216117 to 129
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline, 1a, and dynamic baseline, 1b, please see Section X.A.1.
b Range reflects two reference case assumptions 1a and 1b.
c Benefits and net benefits use the 3 percent global average SCC value applied only to CO2 emissions; GHG reductions include CO2, CH4, N2O and HFC reductions, and include benefits to other nations as well as the U.S. See Draft RIA Chapter 8.5 and Preamble Section IX.G for further discussion.
d “Congestion, Crashes, Fatalities and Noise from Increased Vehicle Use” includes NHTSA's monetized value of estimated reductions in the incidence of highway fatalities associated with mass reduction in HD pickup and vans, but this does not include these reductions from tractor-trailers or vocational vehicles. This likely results in a conservative overestimate of these costs.

Table I-11 shows benefits and cost from the perspective of reducing GHG. As shown below in terms of MY lifetime GHG reductions, and in RIA Chapter 5 in terms of year-by-year GHG reductions, the final program is expected to reduce more GHGs over the long run than the proposed program. In general, the greater reductions can be attributed to increased market penetration and effectiveness of key technologies, based on new data and comments, leading to increases in stringency such as with the diesel engine standards (Section I.C.(2)(a) above).

Table I-11—Lifetime Fuel Savings, GHG Reductions, Benefits, Costs and Net Benefits for Model Years 2018-2029 Vehicles Using Analysis Method B

[Billions of 2012$] ab

Category3% discount rate7% discount rate
Fuel Reductions (Billion Gallons)73-82
GHG reductions (MMT CO2eq)976-1,098
Vehicle Program (e.g., technology and indirect costs, normal profit on additional investments)−$26.5 to −$26.2−$17.6 to −$17.4
Additional Routine Maintenance−$1.9 to −$1.9−$1.0 to −$1.0
Fuel Savings (valued at pre-tax prices)$149.3 to $169.1$76.8 to $87.2
Energy Security$6.9 to $7.8$3.5 to $4.0
Congestion, Crashes, and Noise from Increased Vehicle Use−$3.2 to −$3.2−$1.8 to −$1.8
Savings from Less Frequent Refueling$3.4 to $4.0$1.8 to $2.1
Economic Benefits from Additional Vehicle Use$10.4 to $10.5$5.7 to $5.7
Benefits from Reduced Non-GHG Emissions c$28.3 to $31.9$13.4 to $15.0
Start Printed Page 73509
Reduced Climate Damages from GHG Emissions d$33.0 to $37.2
Net Benefits$200 to $229$114 to $131
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline, 1a, and dynamic baseline, 1b, please see Section X.A.1.
b Range reflects two baseline assumptions 1a and 1b.
c Range reflects both the two baseline assumptions 1a and 1b using the mid-point of the low and high $/ton estimates for calculating benefits.
d Benefits and net benefits use the 3 percent average directly modeled SC-GHG values applied to direct reductions of CO2, CH4 and N2O emissions; GHG reductions include CO2, CH4 and N2O reductions.

Table I-12 breaks down by vehicle category the benefits and costs for these standards using the Method A analytical approach. For additional detail on per-vehicle break-downs of costs and benefits, please see RIA Chapter 10.

Table I-12—Per Vehicle Category Lifetime Fuel Savings, GHG Reductions, Benefits, Costs and Net Benefits for Model Years 2018-2029 Vehicles Using Analysis Method A (Billions of 2013$), Relative to Baseline 1b a

Key costs and benefits by vehicle category3% discount rate7% discount rate
Tractors, Including Engines, and Trailers
Fuel Reductions (Billion Gallons)50
GHG Reductions (MMT CO2 eq)685
Total Costs13.89.0
Total Benefits161.096.8
Net Benefits147.285.5
Vocational Vehicles, Including Engines
Fuel Reductions (Billion Gallons)12
GHG Reductions (MMT CO2 eq)162
Total Costs7.34.8
Total Benefits37.822.7
Net Benefits30.515.3
HD Pickups and Vans
Fuel Reductions (Billion Gallons)10
GHG Reductions (MMT CO2 eq)111
Total Costs7.45.1
Total Benefits26.016.7
Net Benefits18.611.6
Notes:
a For an explanation of analytical Methods A and B, please see Section I.D; for an explanation of the flat baseline, 1a, and dynamic baseline, 1b, please see Section X.A.1.

Table I-13—Per Vehicle Costs, Using Method A (2013$), Relative to Baseline 1b

MY 2021MY 2024MY 2027
Per Vehicle Cost ($): a
Tractors$6,400$9,920$12,160
Trailers8501,0001,070
Vocational Vehicles1,1102,0202,660
Pickups/Vans7507601,340
Note:
a Per vehicle costs include new engine and vehicle technology only; costs associated with increased insurance, taxes and maintenance are included in the payback period values.
Start Printed Page 73510

Table I-14—Per Vehicle Costs Using Method B Relative to Baseline 1a

MY 2021MY 2024MY 2027
Per Vehicle Cost ($): a
Tractors$6,484$10,101$12,442
Trailers8681,0331,108
Vocational Vehicles1,1102,0222,662
Pickups/Vans5249631,364
Note:
a Per vehicle costs include new engine and vehicle technology only; costs associated with increased insurance, taxes and maintenance are included in the payback period values.

An important metric to vehicle purchasers is the payback period that can be expected on any new purchase. In other words, there is greater willingness to pay for new technology if that new technology “pays back” within an acceptable period of time. The agencies make no effort to define the acceptable period of time, but seek to estimate the payback period for others to make the decision themselves. The payback period is the point at which reduced fuel expenditures outpace increased vehicle costs, including increased maintenance, insurance premiums and taxes. The payback periods for vehicles meeting the standards considered for the final year of implementation are shown in Table I-15, and are similar for both Method A and Method B.

Table I-15—Payback Periods for MY 2027 Vehicles Relative to Baseline 1a

[Payback cccurs in the year shown; using 7% discounting]

Tractors/Trailers2nd.
Vocational Vehicles4th.
Pickups/Vans3rd.

Table I-16—Payback Periods for MY 2027 Vehicles Relative to Baseline 1b

[Payback occurs in the year shown; using 7% discounting]

Tractors/Trailers2nd.
Vocational Vehicles4th.
Pickups/Vans3rd.

(3) Cost Effectiveness

These regulations implement section 32902(k) of EISA and section 202(a)(1) and (2) of the Clean Air Act. Through the 2007 EISA, Congress directed NHTSA to create a medium- and heavy-duty vehicle fuel efficiency program designed to achieve the maximum feasible improvement by considering appropriateness, cost effectiveness, and technological feasibility to determine maximum feasible standards.[78] The Clean Air Act requires that any air pollutant emission standards for heavy-duty vehicles and engines take into account the costs of any requisite technology and the lead time necessary to implement such technology. Both agencies considered overall costs, overall benefits and cost effectiveness in developing the Phase 2 standards. Although there are different ways to evaluate cost effectiveness, the essence is to consider some measure of costs relative to some measure of impacts.

Considering that Congress enacted EPCA and EISA to, among other things, address the need to conserve energy, the agencies have evaluated these standards in terms of costs per gallon of fuel conserved. We also considered the similar metric of cost of technology per ton of CO2 e removed, consistent with the objective of CAA section 202(a)(1) and (2) to reduce emissions of air pollutants which contribute to air pollution which endangers public health and welfare. As described in the RIA, the agencies also evaluated these standards using the same approaches employed in HD Phase 1. Together, the agencies have considered the following three ratios of cost effectiveness:

1. Total social costs per gallon of fuel conserved

2. Technology costs per ton of GHG emissions reduced (CO2 eq)

3. Technology costs minus fuel savings per ton of GHG emissions reduced

By all three of these measures, the total heavy-duty program will be highly cost effective.

As discussed below, the agencies estimate that over the lifetime of heavy-duty vehicles produced for sale in the U.S. during model years 2018-2029, these standards will cost about $30 billion and conserve about 75 billion gallons of fuel, such that the first measure of cost effectiveness will be about 40 cents per gallon. Relative to fuel prices underlying the agencies' analysis, the agencies have concluded that today's standards will be cost effective.

With respect to the second measure, which is useful for comparisons to other GHG rules, these standards will have overall $/ton costs similar to the HD Phase 1 rule. As Chapter 7 of the RIA shows, social costs will amount to about $30 per metric ton of GHG (CO2 eq) for the entire HD Phase 2 program. This compares well to both the HD Phase 1 rule, which was also estimated to cost about $30 per metric ton of GHG (without fuel savings), and to the agencies' estimates of the social cost of carbon.[79] Thus, even without accounting for fuel savings, these standards will be cost-effective.

The following table include the overall per-unit costs of both gallons of fuel conserved and metric tons of GHG emissions abated using both a 3 percent and a 7 percent discount rate. Table I-16 gives these values under the Method A analysis.

Start Printed Page 73511

Table I-17—Method A Cost Per-Unit of Fuel Savings and GHG Emission Reductions by Vehicle Class

[Relative to the dynamic baseline]

Per-unit costs (2013$/Unit) by vehicle category3% Discount rate7% Discount rate
Tractors, Including Engines, and Trailers
Cost per Gallon of Fuel Saved$0.28$0.18
Cost per Ton of GHG Emissions Saved2013
Vocational Vehicles, Including Engines
Cost per Gallon of Fuel Saved0.610.40
Cost per Ton of GHG Emissions Saved4530
HD Pickups and Vans
Cost per Gallon of Fuel Saved0.760.52
Cost per Ton of GHG Emissions Saved6746
Total Program
Cost per Gallon of Fuel Saved0.400.26
Cost per Ton of GHG Emissions Saved3020

When considering these values, it is important to emphasize two points:

1. As is shown throughout this rulemaking, the Phase 2 standards represent the most stringent standards that are technologically feasible and reliably implementable within the lead time provided.

2. These are not the marginal cost-effectiveness values.

Without understanding these two points, some readers might assume that because the tractor-trailer standards are more cost-effective overall than the other standards that manufacturers would choose to over-comply with the more cost-effective tractor or trailer standards and do less for other vehicles. However, the agencies believe this is not a technologically feasible option. Because the tractor and trailer standards represent maximum feasible standards, they will effectively require manufacturers to deploy all available technology to meet the standards. The agencies do not project that manufacturers would be able to over-comply with the 2027 standards by a significant margin.

The third measure deducts fuel savings from costs, which also is useful for comparisons to other GHG rules. As shown in Table I-18, the agencies have also calculated the cost per metric ton of CO2 e emission reductions including the savings associated with reduced fuel consumption. The calculations presented here include all engine-related costs but do not include benefits associated with the final program such as those associated with criteria pollutant reductions or energy security benefits (discussed in Chapter 8 of this RIA). On this basis, net costs per ton of GHG emissions reduced will be negative under these standards. This means that the value of the fuel savings will be greater than the technology costs, and there will be a net cost saving for vehicle owners. In other words, the technologies will pay for themselves (indeed, more than pay for themselves) in fuel savings.

Table I-18—Annual Net Cost per Metric Ton of CO2eq Emissions Reduced in the Final Program Vs. the Flat Baseline and Using Method B for Calendar Year 2030

[Dollar values are 2013$] a

Calendar yearVehicle & maintenance costs ($billions)Fuel savings ($billions)GHG reduced (MMT)Net cost ($/metric ton) b
HDE Pickups and Vans1.63.9150
Vocational Vehicles1.53.5140
Tractor-Trailers2.316640
All Vehicles5.523940
Notes:
a For an explanation of analytical Methods A and B, please see the beginning of this Section I.D; for an explanation of the flat baseline, 1a, and dynamic baseline, 1b, please see Section X.A.1. GHG reductions include CO2 and CO2 equivalents of CH4, and N2O.
b For each category, fuel savings exceed cost so there is no net cost per ton of GHG reduced.

In addition, while the net economic benefits (i.e., total benefits minus total costs) of these standards is not a traditional measure of their cost effectiveness, the agencies have concluded that the total costs of these standards are justified in part by their significant economic benefits. As discussed in the previous subsection and in Section IX, this rule will provide benefits beyond the fuel conserved and GHG emissions avoided. The rule's net benefits is a measure that quantifies each of its various benefits in economic terms, including the economic value of the fuel it saves and the climate-related damages it avoids, and compares their sum to the rule's estimated costs. The agencies estimate that these standards will result in net economic benefits exceeding $100 billion, making this a highly beneficial program.

EPA and NHTSA received many comments suggesting that more Start Printed Page 73512stringent standards were feasible because many cost effective technologies exist for future vehicle designs. While the agencies agree that many cost effective technologies exist, and indeed, we reflect the potential for many of those technologies to be applied in our analysis for today's final rule, commenters who focused on the cost-effectiveness of technologies did not consistently recognize certain real-world constraints on technology implementation. Manufacturers and suppliers have limited research and development capacities, and although they have some ability to expand (by adding staff or building new facilities), the process of developing and applying new technologies is inherently constrained by time. Adequate lead time is also necessary to complete durability, reliability, and safety testing and ramp up production to levels that might be necessary to meet future standards. If the agencies fail to account for lead time needs in determining the stringency of the standards, we could create unintended consequences, such as technologies that are applied before they are ready and lead to maintenance and repair problems. In addition to cost-effectiveness, then, lead time constraints can also be highly relevant to feasibility of more stringent standards.

E. EPA and NHTSA Statutory Authorities

This section briefly summarizes the respective statutory authority for EPA and NHTSA to promulgate the Phase 1 and Phase 2 programs. For additional details of the agencies' authority, see Section XV of this document as well as the Phase 1 rule.[80]

(1) EPA Authority

Statutory authority for the emission standards in this rule is found in CAA section 202(a)(1) and (2) (which requires EPA to establish standards for emissions of pollutants from new motor vehicles and engines which emissions cause or contribute to air pollution which may reasonably be anticipated to endanger public health or welfare), and in CAA sections 202(a)(3), 202(d), 203-209, 216, and 301 (42 U.S.C. 7521 (a)(1) and (2), 7521(d), 7522-7543, 7550, and 7601).

Title II of the CAA provides for comprehensive regulation of mobile sources, authorizing EPA to regulate emissions of air pollutants from all mobile source categories. When acting under Title II of the CAA, EPA considers such issues as technology effectiveness, its cost (both per vehicle, per manufacturer, and per consumer), the lead time necessary to implement the technology, and based on this the feasibility and practicability of potential standards; the impacts of potential standards on emissions reductions of both GHGs and non-GHG emissions; the impacts of standards on oil conservation and energy security; the impacts of standards on fuel savings by customers; the impacts of standards on the truck industry; other energy impacts; as well as other relevant factors such as impacts on safety.

This action implements a specific provision from Title II, section 202(a). Section 202(a)(1) of the CAA states that “the Administrator shall by regulation prescribe (and from time to time revise) . . . standards applicable to the emission of any air pollutant from any class or classes of new motor vehicles . . ., which in his judgment cause, or contribute to, air pollution which may reasonably be anticipated to endanger public health or welfare.” With EPA's December 2009 final findings that certain greenhouse gases may reasonably be anticipated to endanger public health and welfare and that emissions of GHGs from section 202(a) sources cause or contribute to that endangerment, section 202(a) requires EPA to issue standards applicable to emissions of those pollutants from new motor vehicles. See Coalition for Responsible Regulation v. EPA, 684 F. 3d at 116-125, 126-27 cert. granted by, in part Util. Air Regulatory Group v. EPA, 134 S. Ct. 418 (2013), affirmed in part and reversed in part on unrelated grounds by Util. Air Regulatory Group v. EPA, 134 S. Ct. 2427 (2014) (upholding EPA's endangerment and cause and contribute findings, and further affirming EPA's conclusion that it is legally compelled to issue standards under section 202(a) to address emission of the pollutant which endangers after making the endangerment and cause or contribute findings); see also id. at 127-29 (upholding EPA's light-duty GHG emission standards for MYs 2012-2016 in their entirety).

Other aspects of EPA's legal authority, including its authority under section 202(a), its testing authority under section 203 of the Act, and its enforcement authorities under sections 205 and 207 of the Act are discussed fully in the Phase 1 rule, and need not be repeated here. See 76 FR 57129-57130.

In this final rule, EPA is establishing first-time CO2 emission standards for trailers hauled by tractors. 80 FR 40170. Certain commenters, notably the Truck Trailer Manufacturers Association (TTMA), maintained that EPA lacks authority to adopt requirements for trailer manufacturers, and that emission standards for trailers could be implemented, if at all, by requirements applicable to the entity assembling a tractor-trailer combination. The argument is that trailers by themselves are not “motor vehicles” as defined in section 216(2) of the Act, that trailer manufacturers therefore do not manufacture motor vehicles, and that standards for trailers can be imposed, if at all, only on “the party that joined the trailer to the tractor.” Comments of TTMA, p. 4; Comments of TTMA (March 31, 2016) p. 2.

EPA also proposed a number of changes and clarifications for rules respecting glider kits and glider vehicles. 80 FR 40527-40530. As shown in Figure I.1, a glider kit is a tractor chassis with frame, front axle, interior and exterior cab, and brakes.

Start Printed Page 73513

It is intended for self-propelled highway use, and becomes a glider vehicle when an engine, transmission, and rear axle are added. Engines are often salvaged from earlier model year vehicles, remanufactured, and installed in the glider kit. The final manufacturer of the glider vehicle, i.e. the entity that installs an engine, is typically a different manufacturer than the original manufacturer of the glider kit. The final rule contains emission standards for glider vehicles, but does not contain separate standards for glider kits.[81]

Many commenters to both the proposed rule and the NODA supported EPA's interpretation. However, a number of commenters, including Daimler, argued that glider kits are not motor vehicles and so EPA lacks the authority to impose any rules respecting their sale or configuration. Comments of Daimler, pp. 122-23; Comments of Daimler Trucks (April 1, 2016) pp. 2-3. We respond to these comments below, with a more detailed response appearing in RTC Section 1.3.1 and 14.2.

Under the Act, “motor vehicle” is defined as “any self-propelled vehicle designed for transporting persons or property on a street or highway.” CAA section 216(2). At proposal, EPA maintained that tractor-trailers are motor vehicles and that EPA therefore has the authority to promulgate emission standards for complete and incomplete vehicles—both the tractor and the trailer. 80 FR 40170. The same proposition holds for glider kits and glider vehicles. Id. at 80 FR 40528. The argument that a trailer, or a glider kit, standing alone, is not self-propelled, and therefore is not a motor vehicle, misses the key issues of authority under the Clean Air Act to promulgate emission standards for motor vehicles produced in discrete segments, and the further issue of the entities—namely “manufacturers”—to which standards and certification requirements apply. Simply put, EPA is authorized to set emission standards for complete and incomplete motor vehicles, manufacturers of complete and incomplete motor vehicles can be required to certify to those emission standards, and there can be multiple manufacturers of a motor vehicle, each of which can be required to certify.

(a) Standards for Complete Vehicles—Tractor-Trailers and Glider Vehicles

Section 202(a)(1) authorizes EPA to set standards “applicable to the emission of any air pollutant from any . . . new motor vehicles.” There is no question that EPA is authorized to establish emission standards under this provision for complete new motor vehicles, and thus can promulgate emission standards for air pollutants emitted by tractor-trailers and by glider vehicles.

Daimler maintained in its comments that although a glider vehicle is a motor vehicle, it is not a “new” motor vehicle because “glider vehicles, when constructed retain the identity of the donor vehicle, such that the title has already been exchanged, making the vehicles not `new' under the CAA.” Daimler Comments p. 121; see also the similar argument in Daimler Truck Comments (April 1, 2016), p. 4. Daimler maintains that because title to the powertrain from the donor vehicle has already been transferred, the glider vehicle to which the powertrain is added cannot be “new.” Comments of April 1, 2016 p. 4. Daimler also notes that NHTSA considers a truck to be “newly manufactured” and subject to Federal Motor Vehicle Safety Standards when a new cab is used in its assembly, “unless the engine, transmission, and drive axle(s) (as a minimum) of the assembled vehicle are not new, and at least two of these components were taken from the same vehicle.” 49 CFR 571.7(e). Daimler urges EPA to adopt a parallel provision here.

First, this argument appears to be untimely. In Phase 1, EPA already indicated that glider vehicles are new motor vehicles, at least implicitly, by Start Printed Page 73514adopting an interim exemption for them. See 76 FR 57407 (adopting 40 CFR 1037.150(j) indicating that the general prohibition against introducing a vehicle not subject to current model year standards does not apply to MY 2013 or earlier engines). Assuming the argument that glider vehicles are not new can be raised in this rulemaking, EPA notes that the Clean Air Act defines “new motor vehicle” as “a motor vehicle the equitable or legal title to which has never been transferred to an ultimate purchaser” (section 216(3)). Glider vehicles are typically marketed and sold as “brand new” trucks. Indeed, one prominent assembler of glider kits and glider vehicles advertises that “Fitzgerald Glider Kits offers customers the option to purchase a brand new 2016 tractor, in any configuration offered by the manufacturer . . . Fitzgerald Glider Kits has mastered the process of taking the `Glider Kit' and installing the components to work seamlessly with the new truck.” [82] The purchaser of a “new truck” necessarily takes initial title to that truck.[83] Daimler would have it that this `new truck' terminology is a mere marketing ploy, but it obviously reflects reality. As shown in Figure I.1 above, the glider kit constitutes the major parts of the vehicle, lacking only the engine, transmission, and rear axle. The EPA sees nothing in the Act that compels the result that adding a used component to an otherwise new motor vehicle necessarily vitiates classification of the motor vehicle as “new.” See 80 FR 40528. Rather, reasonable judgments must be made, and in this case, the agency believes it reasonable that the tail need not wag the dog: Adding the engine and transmission to the otherwise-complete vehicle does not prevent the glider vehicle from being “new”—as marketed. The fact that this approach is reasonable, if not mandated, is confirmed by the language of the Act's definition of “new motor vehicle engine,” which includes any “engine in a new motor vehicle” without regard to whether or not the engine was previously used. EPA has also previously addressed the issue of used components in new engines and vehicles explicitly in regulations in the context of locomotives and locomotive engines in 40 CFR part 1033. There we defined remanufactured locomotives and locomotive engines to be “new” locomotives and locomotive engines. See 63 FR 18980; see also Summary and Analysis of Comments on Notice of Proposed Rulemaking for Emission Standards for Locomotives and Locomotive Engines (EPA-420-R-97-101 (December 1997)) at pp. 10-14. This is a further reason that the model year of the engine is not determinative of whether a glider vehicle is “new.” As to the suggestion to adopt a provision parallel to the NHTSA definition, EPA notes that the NHTSA definition was developed for different purposes using statutory authority which differs from the Clean Air Act in language and intent. There consequently is no basis for requiring EPA to adopt such a definition, and doing so would impede meaningful control of both GHG emissions and criteria pollutant emissions from glider vehicles.

(b) Standards for Incomplete Vehicles

Section 202(a)(1) not only authorizes EPA to set standards “applicable to the emission of any air pollutant from any . . . new motor vehicles,” but states further that these standards are applicable “whether such vehicles . . . are designed as complete systems or incorporate devices to prevent or control such pollution.” The Act in fact thus not only contemplates, but in some instances, directly commands that EPA establish standards for incomplete vehicles and vehicle components. See CAA section 202(a)(6) (standards for onboard vapor recovery systems on “new light-duty vehicles,” and requiring installation of such systems); section 202(a)(5)(A) (standards to control emissions from refueling motor vehicles, and requiring consideration of, and possible design standards for, fueling system components); 202(k) (standards to control evaporative emissions from gasoline-fueled motor vehicles). Both TTMA and Daimler argued, in effect, that these provisions are the exceptions that prove the rule and that without this type of enumerated exception, only entire, complete vehicles can be considered to be “motor vehicles.” This argument is not persuasive. Congress did not indicate that these incomplete vehicle provisions were exceptions to the definition of motor vehicle. Just the opposite. Without amending the new motor vehicle definition, or otherwise indicating that these provisions were not already encompassed within Title II authority over “new motor vehicles”, Congress required EPA to set standards for evaporative emissions from a portion of a motor vehicle. Congress thus indicated in these provisions: (1) That standards should apply to “vehicles” whether or not the “vehicles” were designed as complete systems; (2) that some standards should explicitly apply only to certain components of a vehicle that are plainly not self-propelled. Congress thus necessarily was of the view that incomplete vehicles can be motor vehicles.

Emission standards EPA sets pursuant to this authority thus can be, and often are focused on emissions from the new motor vehicle, and from portions, systems, parts, or components of the vehicle. Standards thus apply not just to exhaust emissions, but to emissions from non-exhaust portions of a vehicle, or from specific vehicle components or parts. See the various evaporative emission standards for light duty vehicles in 40 CFR part 86, subpart B (e.g., 40 CFR 86.146-96 and 86.150-98 (refueling spitback and refueling test procedures); 40 CFR 1060.101-103 and 73 FR 59114-59115 (various evaporative emission standards for small spark ignition equipment); 40 CFR 86.1813-17(a)(2)(iii) (canister bleed evaporative emission test procedure, where testing is solely of fuel tank and evaporative canister); see also 79 FR 23507 (April 28, 2014) (incomplete heavy duty gasoline vehicles could be subject to, and required to certify compliance with, evaporative emission standards)). These standards are implemented by testing the particular vehicle component, not by whole vehicle testing, notwithstanding that the component may not be self-propelled until it is installed in the vehicle or (in the case of non-road equipment), propelled by an engine.[84]

EPA thus can set standards for all or just a portion of the motor vehicle notwithstanding that an incomplete motor vehicle may not yet be self-propelled. This is not to say that the Act authorizes emission standards for any part of a motor vehicle, however insignificant. Under the Act it is reasonable to consider both the significance of the components in comparison to the entire vehicle and the significance of the components for achieving emissions reductions. A vehicle that is complete except for an ignition switch can be subject to standards even though it is not self-Start Printed Page 73515propelled. Likewise, as just noted, vehicle components that are significant for controlling evaporative emissions can be subject to standards even though in isolation the components are not self-propelled. However, not every individual component of a complete vehicle can be subjected to standards as an incomplete vehicle. To reflect these considerations, EPA is adopting provisions stating that a trailer is a vehicle “when it has a frame with one or more axles attached,” and a glider kit becomes a vehicle when “it includes a passenger compartment attached to a frame with one or more axles.” Section 1037.801 definition of “vehicle,” paragraphs (1)(ii) and (iii); see also Section XIII.B below.

TTMA and Daimler each maintained that this claim of authority is open-ended, and can be extended to the least significant vehicle part. As noted above, EPA acknowledges that lines need to be drawn, but whether looking at the relation between the incomplete vehicle and the complete vehicle, or looking at the relation between the incomplete vehicle and the emissions control requirements, it is evident that trailers and glider kits should properly be treated as vehicles, albeit incomplete ones.[85] They properly fall on the vehicle side of the line. When one finishes assembling a whole aggregation of parts to make a finished section of the vehicle (e.g. the trailer), that is sufficient. You have an entire, complete section made up of assembled parts. Everything needed to be a trailer is complete. This is not an engine block, a wheel, or a headlight. Similarly, glider kits comprise the largely assembled tractor chassis with front axles, frame, interior and exterior cab, and brakes. This is not a few assembled components; rather, it is an assembled truck with a few components missing. See CAA section 216(9) of the Act, which defines “motor vehicle or engine part manufacturer” as “any person engaged in the manufacturing, assembling or rebuilding of any device, system, part, component or element of design which is installed in or on motor vehicles or motor vehicle engines.” Trailers and glider kits are not “installed in or on” a motor vehicle. A trailer is half of the tractor-trailer, not some component installed on the tractor. And one would more naturally refer to the donor drivetrain being installed on the glider kit than vice versa. See Figure I.1 above. Furthermore, as discussed below, the trailer and the glider kit are significant for purposes of controlling emissions from the completed vehicle.

Incomplete vehicle standards must, of course, be reasonably designed to control emissions caused by that particular vehicle segment. The standards for trailers would do so and account for the tractor-trailer combination by using a reference tractor in the trailer test procedure (and, conversely, by use of a reference trailer in the tractor test procedure). The Phase 2 rule contains no emission standards for glider kits in isolation, but the standards for glider vehicles necessarily reflect the contribution of the glider kit.

(c) Application of Emission Standards to Manufacturers

In some ways, the critical issue is to whom these emission standards apply. As explained in this section, the emission standards apply to manufacturers of motor vehicles, and manufacturers thus are required to test and to certify compliance to those standards. Moreover, the Act contemplates that a motor vehicle can have multiple manufacturers. With respect to the further question of which manufacturer certifies and tests in multiple manufacturer situations, EPA rules have long contained provisions establishing responsibilities where a vehicle has multiple manufacturers. We are applying those principles in the Phase 2 rules. The overarching principle is that the entity with most control over the particular vehicle segment due to producing it is usually the most appropriate entity to test and certify.[86] EPA is implementing the trailer and glider vehicle emission standards in accord with this principle, so that the entities required to test and certify are the trailer manufacturer and, for glider kits and glider vehicles, either the manufacturer of the glider kit or glider vehicle, depending on which is more appropriate in individual circumstances.

(i) Definition of Manufacturer

Emission standards are implemented through regulation of the manufacturer of the new motor vehicle. See, e.g. section 206(a)(1) (certification testing of motor vehicle submitted by “a manufacturer”); 203(a)(1) (manufacturer of new motor vehicle prohibited from introducing uncertified motor vehicles into commerce); 207(a)(1) (manufacturer of motor vehicle to provide warranty to ultimate purchaser of compliance with applicable emission standards); 207(c) (recall authority); 208(a) (recordkeeping and testing can be required of every manufacturer of new motor vehicle).

The Act further distinguishes between manufacturers of motor vehicles and manufacturers of motor vehicle parts. See, e.g. section 206(a)(2) (voluntary emission control system verification testing); 203(a)(3)(B) (prohibition on parts manufacturers and other persons relating to defeat devices); 207(a)(2) (parts manufacturer may provide warranty certification regarding use of parts); 208(a) (recordkeeping and testing requirements for manufacturers of vehicle and engine “parts or components”).

Thus, the question here is whether a trailer manufacturer or glider kit manufacturer can be a manufacturer of a new motor vehicle and thereby become subject to the certification and related requirements for manufacturers, or must necessarily be classified as a manufacturer of a motor vehicle part or component. EPA may reasonably classify trailer manufacturers and glider kit manufacturers as motor vehicle manufacturers.

Section 216(1) defines a “manufacturer” as “any person engaged in the manufacturing or assembling of new motor vehicles, new motor vehicle engines, new nonroad vehicles or new nonroad engines, or importing such vehicles or engines for resale, or who acts for and is under the control of any such person in connection with the distribution of new motor vehicles, new motor vehicle engines, new nonroad vehicles or new nonroad engines, but shall not include any dealer with respect to new motor vehicles, new motor vehicle engines, new nonroad vehicles or new nonroad engines received by him in commerce.”

It appears plain that this definition was not intended to restrict the definition of “manufacturer” to a single person per vehicle. The use of the conjunctive, specifying that a manufacturer is “any person engaged in the manufacturing or assembling of new motor vehicles . . . or who acts for and is under the control of any such person Start Printed Page 73516. . .” (emphasis added) indicates that Congress anticipated that motor vehicles could have more than one manufacturer, since in at least some cases those will plainly be different people. The capacious reference to “any person engaged in the manufacturing of motor vehicles” likewise allows the natural inference that it could apply to multiple entities engaged in manufacturing.[87]

The provision also applies both to entities that manufacture and entities that assemble, and does so in such a way as to encompass multiple parties: Manufacturers “or” (rather than ‘and’) assemblers are included. Nor is there any obvious reason that only one person can be engaged in vehicle manufacture or vehicle assembling.

Reading the Act to provide for multiple motor vehicle manufacturers reasonably reflects industry realities, and achieves important goals of the CAA. Since title II requirements are generally imposed on “manufacturers” it is important that the appropriate parties be included within the definition of manufacturer—“any person engaged in the manufacturing or assembling of new motor vehicles.” Indeed, as set out in Chapter 1 of the RIA, most heavy duty vehicles are manufactured or assembled by multiple entities; see also Comments of Daimler (October 1, 2015) p. 103.[88] One entity produces a chassis; a different entity manufactures the engine; specialized components (e.g. garbage compactors, cement mixers) are produced by still different entities. For tractor-trailers, one person manufactures the tractor, another the trailer, a third the engine, and another typically assembles the trailer to the tractor. Installation of various vehicle components occurs at different and varied points and by different entities, depending on ultimate desired configurations. See, e.g. Comments of Navistar (October 1, 2015), pp. 12-13. The heavy duty sector thus differs markedly from the light duty sector (and from manufacturing of light duty pickups and vans), where a single company designs the vehicle and engine (and many of the parts), and does all assembling of components into the finished motor vehicle.

(ii) Controls on Manufacturers of Trailers

It is reasonable to view the trailer manufacturer as “engaged in” (section 216(1)) the manufacturing or assembling of the tractor-trailer. The trailer manufacturer designs, builds, and assembles a complete and finished portion of the tractor-trailer. All components of the trailer—the tires, axles, flat bed, outsider cover, aerodynamics—are within its control and are part of its assembling process. The trailer manufacturer sets the design specifications that affect the GHG emissions attributable to pulling the trailer. It commences all work on the trailer, and when that work is complete, nothing more is to be done. The trailer is a finished product. With respect to the trailer, the trailer manufacturer is analogous to the manufacturer of the light duty vehicle, specifying, controlling, and assembling all aspects of the product from inception to completion. GHG emissions attributable to the trailer are a substantial portion of the total GHG emissions from the tractor-trailer.[89] Moreover, the trailer manufacturer is not analogous to the manufacturer of a vehicle part or component, like a tire manufacturer, or to the manufacturer of a side skirt. The trailer is a significant, integral part of the finished motor vehicle, and is essential for the tractor-trailer to carry out its commercial purpose. See 80 FR 40170. Although it is true that another person may ultimately hitch the trailer to a tractor (which might be viewed as completing assembly of the tractor-trailer), as noted above, EPA does not believe that the fact that one person might qualify as a manufacturer, due to “assembling” the motor vehicle, precludes another person from qualifying as a manufacturer, due to “manufacturing” the motor vehicle. Given that section 216(1) does not restrict motor vehicle manufacturers to a single entity, it appears to be consistent with the facts and the Act to consider trailer manufacturers as persons engaged in the manufacture of a motor vehicle.

This interpretation of section 216(1) is also reasonable in light of the various provisions noted above relating to implementation of the emissions standards—certification under section 206, prohibitions on entry into commerce under section 203, warranty and recall under section 207, and recordkeeping/reporting under section 208. All of these provisions are naturally applied to the entity responsible for manufacturing the trailer, which manufacturer is likewise responsible for its GHG emissions.

TTMA maintains that if a tractor-trailer is a motor vehicle, then only the entity connecting the trailer to the tractor could be subject to regulation.[90] This is not a necessary interpretation of section 216(1), as explained above. TTMA does not discuss that provision, but notes that other provisions refer to “a” manufacturer (or, in one instance, “the” manufacturer), and maintains that this shows that only a single entity can be a manufacturer. See TTMA Comment pp. 4-5, citing to sections 206(a)(1), 206(b), 207, and 203(a). This reading is not compelled by the statutory text. First, the term “manufacturer” in all of these provisions necessarily reflects the underlying definition in section 216(1), and therefore is not limited to a single entity, as just discussed. Second, the interpretation makes no practical sense. An end assembler of a tractor-trailer is not in a position to certify and warrant performance of the trailer, given that the end-assembler has no control over how trailers are designed, constructed, or even which trailers are attached to the tractor. It makes little sense for the entity least able to control the outcome to be responsible for that outcome. The EPA doubts that Congress compelled such an ungainly implementation mechanism, especially given that it is well known that vehicle manufacture responsibility in the heavy duty vehicle sector is divided, and given further that title II includes requirements for EPA to promulgate emission standards for portions of vehicles.

(iii) Controls on Manufacturers of Glider Kits

Application of these same principles indicate that a glider kit manufacturer is a manufacturer of a motor vehicle and, as an entity responsible for assuring that glider vehicles meet the Phase 2 vehicle emission standards, can be a party in the certification process as either the certificate holder or the entity which provides essential test information to the glider vehicle manufacturer. As noted above, glider kits include the entire tractor chassis, cab, tires, body, and brakes. Glider kit manufacturers thus control critical elements of the Start Printed Page 73517ultimate vehicle's greenhouse gas emissions, in particular, all aerodynamic features and all emissions related to steer tire type. Glider kit manufacturers would therefore be the entity generating critical GEM inputs—at the least, those for aerodynamics and tires. Glider kit manufacturers also often know the final configuration of the glider vehicle, i.e. the type of engine and transmission which the final assembler will add to the glider kit.[91] This is because the typical glider kit contains all necessary wiring, and it is necessary, in turn, for the glider kit manufacturer to know the end configuration in order to wire the kit properly. Thus, a manufacturer of a glider kit can reasonably be viewed as a manufacturer of a motor vehicle under the same logic as above: There can be multiple manufacturers of a motor vehicle; the glider kit manufacturer designs, builds, and assembles a substantial, complete and finished portion of the motor vehicle; and that portion contributes substantially to the GHG emissions from the ultimate glider vehicle. A glider kit is not a vehicle part; rather, it is an assembled truck with a few components missing.

EPA rules have long provided provisions establishing responsibilities where there are multiple manufacturers of motor vehicles. See 40 CFR 1037.620 (responsibilities for multiple manufacturers), 40 CFR 1037.621 (delegated assembly), and 40 CFR 1037.622 (shipment of incomplete vehicles to secondary vehicle manufacturers). These provisions, in essence, allow manufacturers to determine among themselves as to which should be the certificate holder, and then assign respective responsibilities depending on that decision. The end result is that incomplete vehicles cannot be introduced into commerce without one of the manufacturers being the certificate holder.

Under the Phase 1 rules, glider kits are considered to be incomplete vehicles which may be introduced into commerce to a secondary manufacturer for final assembly. See 40 CFR 1037.622(b)(1)(i) and 1037.801 (definition of “vehicle” and “incomplete vehicle”) of the Phase 1 regulations (76 FR 57421). Note that 40 CFR 1037.622(b)(1)(i) was originally codified as 40 CFR 1037.620(b)(1)(i). EPA is expanding somewhat on these provisions, but in essence, as under Phase 1, glider kit and glider vehicle manufacturers could operate under delegated assembly provisions whereby the glider kit manufacturer would be the certificate holder. See 40 CFR 1037.621 of the final regulations. Glider kit manufacturers would also continue to be able to ship uncertified kits to secondary manufacturers, and the secondary manufacturer must assemble the vehicle into certifiable condition. 40 CFR 1037.622.[92]

(d) Additional Authorities Supporting EPA's Actions

Even if, against our view, trailers and glider kits are not considered to be “motor vehicles,” and the entities engaged in assembling trailers and glider kits are not considered to be manufacturers of motor vehicles, the Clean Air Act still provides authority for the testing requirements adopted here. Section 208 (a) of the Act authorizes EPA to require “every manufacturer of new motor vehicle or engine parts or components” to “perform tests where such testing is not otherwise reasonably available.” This testing can be required to “provide information the Administrator may reasonably require to determine whether the manufacturer . . . has acted or is acting in compliance with this part,” which includes showing whether or not the parts manufacturer is engaged in conduct which can cause a prohibited act. Testing would be required to show that the trailer will conform to the vehicle emission standards. In addition, testing for trailer manufacturers would be necessary here to show that the trailer manufacturer is not causing a violation of the combined tractor-trailer GHG emission standard either by manufacturing a trailer which fails to comply with the trailer emission standards, or by furnishing a trailer to the entity assembling tractor-trailers inconsistent with tractor-trailer certified condition. Testing for glider kit manufacturers is necessary to prevent a glider kit manufacturer furnishing a glider kit inconsistent with the tractor's certified condition. In this regard, we note that section 203 (a)(1) of the Act not only prohibits certain acts, but also prohibits “the causing” of those acts. Furnishing a trailer not meeting the trailer standard would cause a violation of that standard, and the trailer manufacturer would be liable under section 203 (a)(1) for causing the prohibited act to occur. Similarly, a glider kit supplied in a condition inconsistent with the tractor standard would cause the manufacturer of the glider vehicle to violate the GHG emission standard, so the glider kit manufacturer would be similarly liable under section 203 (a)(1) for causing that prohibited act to occur.

In addition, section 203 (a)(3)(B) prohibits use of `defeat devices'—which include “any part or component intended for use with, or as part of, any motor vehicle . . . where a principal effect of the part or component is to . . . defeat . . . any . . . element of design installed . . . in a motor vehicle” otherwise in compliance with emission standards. Manufacturing or installing a trailer not meeting the trailer emission standard could thus be a defeat device causing a violation of the emission standard. Similarly, a glider kit manufacturer furnishing a glider kit in a configuration that would not meet the tractor standard when the specified engine, transmission, and axle are installed would likewise cause a violation of the tractor emission standard. For example, providing a tractor with a coefficient of drag or tire rolling resistance level inconsistent with tractor certified condition would be a violation of the Act because it would cause the glider vehicle assembler to introduce into commerce a new tractor that is not covered by a valid certificate of conformity. Daimler argued in its comments that a glider kit would not be a defeat device because glider vehicles use older engines which are more fuel efficient since they are not meeting the more rigorous standards for criteria pollutant emissions. (Daimler Truck Comment, April 1, 2016, p. 5). However, the glider kit would be a defeat device with respect to the tractor vehicle standard, not the separate engine standard. A non-conforming glider kit would adversely affect compliance with the vehicle standard, as just explained. Furthermore, as explained in RTC Section 14.2, Daimler is incorrect that glider vehicles are more fuel efficient than Phase 1 2017 and later vehicles, much less Phase 2 vehicles.

In the memorandum accompanying the Notice of Data Availability, EPA solicited comment on adopting additional regulations based on these principles. EPA has decided not to adopt those provisions, but again notes Start Printed Page 73518that the authorities in CAA sections 208 and 203 support the actions EPA is taking here with respect to trailer and glider kit testing.

(e) Standards for Glider Vehicles and Lead Time for Those Standards

At proposal, EPA indicated that engines used in glider vehicles are to be certified to standards for the model year in which these vehicles are assembled. 80 FR 40528. This action is well within the agency's legal authority. As noted above, the Act's definition of “new motor vehicle engine,” includes any “engine in a new motor vehicle” without regard to whether or not the engine was previously used. Given the Act's purpose of controlling emissions of air pollutants from motor vehicle engines, with special concern for pollutant emissions from heavy-duty engines (see, e.g., section 202(a)(3)(A) and (B)), it is reasonable to require engines placed in newly-assembled vehicles to meet the same standards as all other engines in new motor vehicles. Put another way, it is both consistent with the plain language of the Act and reasonable and equitable for the engines in “new trucks” (see Section I.E.(1)(a) above) to meet the emission standards for all other engines installed in new trucks.

Daimler challenged this aspect of EPA's proposal, maintaining that it amounted to regulation of vehicle rebuilding, which (according to the commenter) is beyond EPA's authority. Comments of Daimler, p. 123; Comments of Daimler Trucks (April 1, 2016) p. 3. This comment is misplaced. The EPA has authority to regulate emissions of pollutants from engines installed in new motor vehicles. As explained in subsection (a) above, glider vehicles are new motor vehicles. As also explained above, the Act's definition of “new motor vehicle engine” includes any “engine in a new motor vehicle” without regard to whether or not the engine was previously used. CAA section 216(3). Consequently, a previously used engine installed in a glider vehicle is within EPA's multiple authorities. See CAA sections 202(a)(1) (GHGs), 202(a)(3)(A) and (B)(ii) (hydrocarbon, CO, PM and NOX from heavy-duty vehicles or engines), and 202(a)(3)(D) (pollutants from rebuilt heavy duty engines).[93]

As explained in more detail in Section XIII.B, the final rule requires that as of January 1, 2017, glider kit and glider vehicle production involving engines not meeting criteria pollutant standards corresponding to the year of glider vehicle assembly be allowed at the highest annual production for any year from 2010 to 2014. See section 1037.150(t)(3). (Certain exceptions to this are explained in Section XIII.B.) The rule further requires that as of January 1, 2018, engines in glider vehicles meet criteria pollutant standards and GHG standards corresponding to the year of the glider vehicle assembly, but allowing certain small businesses to introduce into commerce vehicles with engines meeting criteria pollutant standards corresponding to the year of the engine for up to 300 vehicles per year, or up to the highest annual production volume for calendar years 2010 to 2014, whichever is less. Section 1037.150(t)(1)(ii) (again subject to various exceptions explained in Section XIII.B). Glider vehicles using these exempted engines will not be subject to the Phase 1 GHG vehicle standards, but will be subject to the Phase 2 vehicle standards beginning with MY 2021. As explained in Section XIII.B, there are compelling environmental reasons for taking these actions in this time frame.

With regard to the issue of lead time, EPA indicated at proposal that the agency has long since justified the criteria pollutant standards for engines installed in glider kits. 80 FR 40528. EPA further proposed that engines installed in glider vehicles meet the emission standard for the year of glider vehicle assembly, as of January 1, 2018 and solicited comment on an earlier effective date. Id. at 40529. The agency noted that CAA section 202(a)(3)(D) [94] requires that standards for rebuilt heavy-duty engines take effect “after a period . . . necessary to permit the development and application of the requisite control measures.” Here, no time is needed to develop and apply requisite control measures for criteria pollutants because compliant engines are immediately available. In fact, manufacturers of compliant engines, and dealers of trucks containing those compliant engines, commented that they are disadvantaged by manufacturing more costly compliant engines while glider vehicles avoid using those engines. Not only are compliant engines immediately available, but (as commenters warned) there can be risk of massive pre-buys. Moreover, EPA does not envision that glider manufacturers will actually modify the older engines to meet the applicable standards. Rather, they will either choose from the many compliant engines available today, or they will seek to qualify under other flexibilities provided in the final rule. See Section XIII.B. Given that compliant engines are immediately available, the flexibilities provided in the final rule for continued use of donor engines for traditional glider vehicle functions and by small businesses, and the need to expeditiously prevent further perpetuation of use of heavily polluting engines, EPA sees a need to begin constraining this practice on January 1, 2017. However, the final rule is merely capping glider production using higher-polluting engines in 2017 at 2010-2014 production levels, which would allow for the production of thousands of glider vehicles using these higher polluting engines, and unlimited production of glider vehicles using less polluting engines.

Various commenters, however, argued that the EPA must provide four years lead-time and three-year stability pursuant to section 202(a)(3)(C) of the Act, which applies to regulations for criteria pollutant emissions from heavy duty vehicles or engines. For criteria pollutant standards, CAA section 202(a)(3)(C) establishes lead time and stability requirements for “[a]ny standard promulgated or revised under this paragraph and applicable to classes or categories of heavy duty vehicles or engines.” In this rule, EPA is generally requiring large manufacturers of glider vehicles to use engines that meet the standards for the model year in which a vehicle is manufactured. EPA is not promulgating new criteria pollutant standards. The NOX and PM standards that apply to heavy duty engines were promulgated in 2001.

We are not amending these provisions or promulgating new criteria pollutant standards for heavy duty engines here. EPA interprets the phrase “classes or categories of heavy duty vehicles or engines” in CAA section 202(a)(3)(C) to refer to categories of vehicles established according to features such as their weight, functional type, (e.g. tractor, vocational vehicle, or pickup truck) or engine cycle (spark-ignition or compression-ignition), or weight class of the vehicle into which an engine is installed (LHD, MHD, or HHD). EPA has established several different categories Start Printed Page 73519of heavy duty vehicles (distinguished by gross vehicle weight, engine-cycle, and other criteria related to the vehicles' intended purpose) and is establishing in this rule GHG standards applicable to each category.[95] By contrast, a “glider vehicle” is defined not by its weight or function but by its method of manufacture. A Class 8 tractor glider vehicle serves exactly the same function and market as a Class 8 tractor manufactured by another manufacturer. Similarly, rebuilt engines installed in glider vehicles (i.e. donor engines) are not distinguished by engine cycle, but rather serve the same function and market as any other HHD or MHD engine. Thus, EPA considers “glider vehicles” to be a description of a method of manufacturing new motor vehicles, not a description of a separate “class or category” of heavy duty vehicles or engines. Consequently, EPA is not adopting new standards for a class or category of heavy duty engines within the meaning of section 202(a)(3)(C) of the Act.

EPA believes this approach is most consistent with the statutory language and the goals of the Clean Air Act. The date of promulgation of the criteria pollutant standards was 2001. There has been plenty of lead time for the criteria pollutant standards and as a result, manufacturers of glider vehicles have many options for compliant engines that are available on the market today—just as manufacturers of other new heavy-duty vehicles do. We are even providing additional compliance flexibilities to glider manufacturers in recognition of the historic practice of salvaging a small number of engines from vehicles involved in crashes. See Section XIII.B. We do not believe that Congress intended to allow changes in how motor vehicles are manufactured to be a means of avoiding existing, applicable engine standards. Obviously, any industry attempts to avoid or circumvent standards will not become apparent until the standards begin to apply. The commenters' interpretation would effectively preclude EPA from curbing many types of avoidance, however dangerous, until at least four years from detection.

As to Daimler's further argument that the lead time provisions in section 202(3)(C) not only apply but also must trump those specifically applicable to heavy duty engine rebuilding, the usual rule of construction is that the more specific provision controls. See, e.g. HCSC-Laundry v. U.S., 450 U.S.1, 6 (1981). Daimler's further argument that section 202(a)(3)(C) lead time provisions also apply to engine rebuilding because those provisions fall within the same paragraph would render the separate lead time provisions for engine rebuilding a virtual nullity. The sense of the provision is that Congress intended there to be independent lead time consideration for the distinct practice of engine rebuilding. In any case, as just explained, it is EPA's view that section 202(a)(3)(C) does not apply here.

(2) NHTSA Authority

The Energy Policy and Conservation Act (EPCA) of 1975 mandates a regulatory program for motor vehicle fuel economy to meet the various facets of the need to conserve energy. In December 2007, Congress enacted the Energy Independence and Security Act (EISA), amending EPCA to require, among other things, the creation of a medium- and heavy-duty fuel efficiency program for the first time.

Statutory authority for the fuel consumption standards in this final rule is found in EISA section 103, 49 U.S.C. 32902(k). This section authorizes a fuel efficiency improvement program, designed to achieve the maximum feasible improvement to be created for commercial medium- and heavy-duty on-highway vehicles and work trucks, to include appropriate test methods, measurement metrics, standards, and compliance and enforcement protocols that are appropriate, cost-effective and technologically feasible.

NHTSA has responsibility for fuel economy and consumption standards, and assures compliance with EISA through rulemaking, including standard-setting; technical reviews, audits and studies; investigations; and enforcement of implementing regulations including penalty actions. This rule continues to fulfill the requirements of section 103 of EISA, which instructs NHTSA to create a fuel efficiency improvement program for “commercial medium- and heavy-duty on-highway vehicles and work trucks” by rulemaking, which is to include standards, test methods, measurement metrics, and enforcement protocols. See 49 U.S.C. 32902(k)(2).

Congress directed that the standards, test methods, measurement metrics, and compliance and enforcement protocols be “appropriate, cost-effective, and technologically feasible” for the vehicles to be regulated, while achieving the “maximum feasible improvement” in fuel efficiency. NHTSA has broad discretion to balance the statutory factors in section 103 in developing fuel consumption standards to achieve the maximum feasible improvement.

As discussed in the Phase 1 final rule, NHTSA has determined that the five year statutory limit on average fuel economy standards that applies to passengers and light trucks is not applicable to the HD vehicle and engine standards. As a result, the Phase 1 HD engine and vehicle standards remain in effect indefinitely at their 2018 or 2019 MY levels until amended by a future rulemaking action. As was contemplated in that rule, NHTSA is finalizing a Phase 2 rulemaking action. Therefore, the Phase 1 standards will not remain in effect at their 2018 or 2019 MY levels indefinitely; they will remain in effect until the MY Phase 2 standards begin. In accordance with section 103 of EISA, NHTSA will ensure that not less than four full MYs of regulatory lead-time and three full MYs of regulatory stability are provided for in the Phase 2 standards.

With respect to the proposal, many stakeholders opined in their comments as to NHTSA's legal authority to issue the Phase 2 medium- and heavy-duty standards (Phase 2 standards), in whole or in part. NHTSA addresses these comments in the following discussion.

Allison Transmission, Inc. (Allison) questioned NHTSA's authority to issue the Phase 2 Standards. Allison stated that the Energy Independence and Security Act of 2007 (EISA) [96] directs NHTSA to undertake “a rulemaking proceeding,” (emphasis added) predicated on a study by the National Academy of Sciences (NAS). Allison and the Truck Trailer Manufacturers Association (TTMA) asserted that because NAS has published a study on medium- and heavy duty vehicles and NHTSA promulgated the Phase 1 medium- and heavy-duty vehicle standards (Phase 1 standards), NAS and NHTSA have fulfilled their statutory duties under EISA. Thus, Allison stated, NHTSA has no authority to issue standards beyond the Phase 1 standards.

NHTSA maintains that EISA allows the agency to promulgate medium- and heavy duty fuel efficiency standards beyond the Phase 1 standards. EISA states that NHTSA: [97]

by regulation, shall determine in a rulemaking proceeding how to implement a commercial medium- and heavy-duty on-highway vehicle and work truck fuel Start Printed Page 73520efficiency program designed to achieve the maximum feasible improvement, and shall adopt and implement appropriate test methods, measurement metrics, fuel economy standards, and compliance and enforcement protocols . . . for commercial medium- and heavy-duty on-highway vehicles and work trucks.[98]

Allison equates the process by which Congress specified NHTSA promulgate standards—a rulemaking proceeding—to mean a limitation or constraint on NHTSA's ability to create, amend, or update the medium- and heavy duty fuel efficiency program. NHTSA believes the charge in 49 U.S.C. 32902(k)(2) discusses “a rulemaking proceeding” only insofar as the statute specifies the process by which NHTSA would create a medium- and heavy-duty on-highway vehicle and work truck fuel efficiency improvement program and its associated standards.

Allison and TTMA commented that EISA only refers to an initial NAS study, meaning EISA only specified that NHTSA issue one set of standards based on that study. As NHTSA stated in the NPRM, EISA requires NAS to issue updates to the initial report every five years through 2025.[99] With that in mind, NAS issued an interim version of its first update to inform the Phase 2 NPRM. EISA's requirement that NAS update its initial report, which examines existing and potential fuel efficiency technologies that can practically be integrated into medium- and heavy-duty vehicles, is consistent with the conclusion that EISA intended the medium- and heavy-duty standards to function as part of an ongoing program [100] and not a single rulemaking.

Allison also noted that the language in EISA discussing lead time and stability refers to a single medium- and heavy-duty on-highway vehicle and work truck fuel economy standard.[101] NHTSA believes the language highlighted by Allison serves the purpose of noting that each medium- and heavy-duty segment standard included in its program shall have the requisite amount of lead-time and stability. As discussed in 49 U.S.C. 32902(k)(2), “[t]he Secretary may prescribe separate standards for different classes of vehicles . . .” Since NHTSA has elected to set standards for particular classes of vehicles, this language ensures each particular standard shall have the appropriate lead-time and stability required by EISA.

TTMA asserted that NHTSA has no more than 24 months from the completion of the NAS study to issue regulations related to the medium- and heavy-duty program and therefore regulations issued after 2013 “lack congressional authorization.” This argument significantly misinterprets the Congressional purpose of this provision. Section 32902(k)(2) requires that, 24 months after the completion of the NAS study, NHTSA begin implementing through a rulemaking proceeding a commercial medium- and heavy-duty on-highway vehicle and work truck fuel efficiency improvement program. Congress therefore authorized NHTSA to implement through rulemaking a “program,” which the dictionary defines as “a plan of things that are done in order to achieve a specific result.” [102] Contrary to TTMA's assertion, Congress did not limit NHTSA to the establishment of one set of regulations, nor did it in any way limit NHTSA's ability to update and revise this program. The purpose of the 24 month period was simply to ensure that NHTSA exercised this authority expeditiously after the NAS study, which NHTSA accomplished by implementing the first phase of its fuel efficiency program in 2011.[103] Today's rulemaking merely continues this program and clearly comports with the statutory language in 49 U.S.C. 32902(k). Further, the specific result sought by Congress in establishing the medium- and heavy-duty fuel efficiency program was a program focused on continuing fuel efficiency improvements. Specifically, Congress emphasized that the fuel efficiency program created by NHTSA be “designed to achieve the maximum feasible improvement,” allowing NHTSA to ensure the regulations implemented throughout the program encourage regulated entities to achieve the maximum feasible improvements. Congress did not limit, restrict, or otherwise suggest that the phrase “designed to achieve the maximum feasible improvement” be confined to the issuance of one set of standards. NHTSA actions are, therefore, clearly consistent with the authority conferred upon it in 49 U.S.C. 32902(k).

POP Diesel stated that the word “fuel” has not been defined by Congress, and therefore NHTSA should use its authority to define the term “fuel” as “fossil fuel,” allowing the agencies to assess fuel efficiency based on the carbon content of the fuels used in an engine or vehicle. Congress has already defined the term “fuel” in 49 U.S.C. 32901(a)(10) as gasoline, diesel oil, or other liquid or gaseous fuel that the Secretary decides to include. As Congress has already spoken to the definition of fuel, it would be inappropriate for the agency to redefine “fuel” as “fossil fuel.”

Additionally, POP Diesel asserted that NHTSA's metric for measuring fuel efficiency is contrary to the mandate in EISA. Specifically, POP Diesel stated that many dictionaries define “efficiency” as a ratio of work performed to the amount of energy used, and NHTSA's load specific fuel consumption metric runs afoul of the plain meaning of statute the Phase 2 program implements. POP Diesel noted that Congressional debate surrounding what is now codified at 49 U.S.C. 32902(k)(2) included a discussion that envisioned NHTSA and EPA having separate regulations, despite having overlapping jurisdiction.

NHTSA continues to believe its use of load specific fuel consumption is an appropriate metric for assessing fuel efficiency as mandated by Congress. 49 U.S.C. 32902(k)(2) states, as POP Diesel noted, that NHTSA shall develop a medium- and heavy-duty fuel efficiency program. The section further states that NHTSA “. . . shall adopt and implement appropriate test methods [and] measurement metrics . . . for commercial medium- and heavy-duty on-highway vehicles and work trucks.” In the Phase 1 rulemaking, NHTSA, aided by the National Academies of Sciences (NAS) report, assessed potential metrics for evaluating fuel efficiency. NHTSA found that fuel economy would not be an appropriate metric for medium- and heavy-duty vehicles. Instead, NHTSA chose a metric that considers the amount of fuel consumed when moving a ton of freight (i.e., performing work).[104] This metric, delegated by Congress to NHTSA to formulate, is not precluded by the text of the statute. It is a reasonable way by which to measure fuel efficiency for a program designed to reduce fuel consumption.

Start Printed Page 73521

(a) NHTSA's Authority To Regulate Trailers

As contemplated in the Phase 1 proposed and final rules, the agencies proposed standards for trailers in the Phase 2 rulemaking. Because Phase 1 did not include standards for trailers, NHTSA did not discuss its authority for regulating them in the proposed or final rules; that authority is described here.

NHTSA is finalizing fuel efficiency standards applicable to heavy-duty trailers as part of the Phase 2 program. NHTSA received several comments on the proposal relating to the agency's statutory authority to issue standards for trailers as part of the Phase 2 program. In particular, TTMA commented that NHTSA does not have the authority to regulate trailers as part of the medium- and heavy-duty standards. TTMA took issue with NHTSA's use of the National Traffic and Motor Vehicle Safety Act as an aid in defining an undefined term in EISA. Additionally, TTMA stated that EISA's use of GVWR instead of gross combination weight rating (GCWR) to define the vehicles subject to these regulations was intended to exclude trailers from the regulation.

As stated in the proposal, EISA directs NHTSA to “determine in a rulemaking proceeding how to implement a commercial medium- and heavy-duty on-highway vehicle and work truck fuel efficiency improvement program designed to achieve the maximum feasible improvement . . . .” [105] EISA defines a commercial medium- and heavy-duty on-highway vehicle to mean “an on-highway vehicle with a GVWR of 10,000 lbs or more.” A “work truck” is defined as a vehicle between 8,500 and 10,000 lbs GVWR that is not an MDPV. These definitions do not explicitly exclude trailers, in contrast to MDPVs. Because Congress did not act to exclude trailers when defining these terms by GVWRs, despite demonstrating the ability to exclude MDPVs, it is reasonable to interpret the provision to include them.

Both the tractor and the trailer are vehicles subject to regulation by NHTSA in the Phase 2 program. Although EISA does not define the term “vehicle,” NHTSA's authority to regulate motor vehicles under its organic statute, the Motor Vehicle Safety Act (“Safety Act”), does. The Safety Act defines a motor vehicle as “a vehicle driven or drawn by mechanical power and manufactured primarily for use on public streets, roads, and highways. . . .” [106] NHTSA clearly has authority to regulate trailers under this Act as they are vehicles that are drawn by mechanical power—in this instance, a tractor engine—and NHTSA has exercised that authority numerous times.[107] Given the absence of any apparent contrary intent on the part of Congress in EISA, NHTSA believes it is reasonable to interpret the term “vehicle” as used in the EISA definitions to have a similar meaning that includes trailers.

Additionally, it is worth noting that the dictionary definition of “vehicle” is “a machine used to transport goods or persons from one location to another.” [108] A trailer is a machine designed for the purpose of transporting goods. With these foregoing considerations in mind, NHTSA interprets its authority to regulate commercial medium- and heavy-duty on-highway vehicles, including trailers.

TTMA pointed to language in the Phase 1 NPRM where the agencies stated that GCWR included the weight of a loaded trailer and the vehicle itself. TTMA interprets this language to mean that standards applicable to vehicles defined by GVWR must inherently exclude trailers. The language TTMA cited is a clarification from a footnote in an introductory section describing the heavy-duty trucking industry. This statement was not a statement of NHTSA's legal authority over medium- and heavy-duty vehicles. NHTSA continues to believe a trailer is a vehicle under EISA if its GVWR fits within the definitions in 49 U.S.C. 32901(a), and is therefore subject to NHTSA's applicable fuel efficiency regulations.

Finally, in a comment on the Notice of Data Availability, TTMA stated that because NHTSA's statutory authority instructs the agency to develop a fuel efficiency program for medium- and heavy-duty on-highway vehicles, and trailers themselves do not consume fuel, trailers cannot be regulated for fuel efficiency. The agency disagrees with this assertion. A tractor-trailer is designed for the purpose of holding and transporting goods. While heavy-duty trailers themselves do not consume fuel, they are immobile and inoperative without a tractor providing motive power. Inherently, trailers are designed to be pulled by a tractor, which in turn affects the fuel efficiency of the tractor-trailer as a whole. As previously discussed, both a tractor and trailer are motor vehicles under NHTSA's authority. Therefore it is reasonable to consider all of a tractor-trailer's parts—the engine, the cab-chassis, and the trailer—as parts of a whole. As such they are all parts of a vehicle, and are captured within the scope of NHTSA's statutory authority. As EPA describes above, the tractor and trailer are both incomplete without the other. Neither can fulfill the function of the vehicle without the other. For this reason, and the other reasons stated above, NHTSA interprets its authority to regulate commercial medium- and heavy-duty on-highway vehicles, including tractor-trailers, as encompassing both tractors and trailers.

(b) NHTSA's Authority To Regulate Recreational Vehicles

NHTSA did not regulate recreational vehicles as part of the Phase 1 medium- and heavy-duty fuel efficiency standards, although EPA did regulate them as vocational vehicles for GHG emissions. In the Phase 1 NPRM, NHTSA interpreted “commercial medium- and heavy duty on-road vehicle” to mean that recreational vehicles, such as motor homes, were not to be included within the program because recreational vehicles are not commercial. Following comments to the Phase 1 proposal, NHTSA reevaluated its statutory authority and proposed that recreational vehicles be included in the Phase 2 standards, and that early compliance be allowed for manufacturers who want to certify during the Phase 1 period.

The Recreational Vehicle Industry Association (RVIA) and Newell Coach Corporation (Newell) asserted that NHTSA does not have the authority to regulate recreational vehicles (RVs). RVIA and Newell stated that NHTSA's authority under EISA is limited to commercial medium- and heavy-duty vehicles and that RVs are not commercial. RVIA pointed to the fact that EISA gives NHTSA fuel efficiency authority over “commercial medium- and heavy-duty vehicles” and “work trucks,” the latter of which is not prefaced with the word “commercial.” Because of this difference, RVIA argued that NHTSA is ignoring a limitation on its authority—that is, that NHTSA only has authority over medium- and heavy-duty vehicles that are commercial in nature. RVIA stated that RVs are not used for commercial purposes, and are therefore not subject to Phase 2.

NHTSA's authority to regulate medium- and heavy-duty vehicles under EISA extends to “commercial medium- and heavy-duty on-highway vehicles” Start Printed Page 73522and “work truck[s].” [109] If terms in the statute are defined, NHTSA must apply those definitions. Both terms highlighted by RVIA have been defined in EISA, therefore, NHTSA will use their defined meanings. “Work truck” means a vehicle that is rated between 8,500 and 10,000 pounds GVWR and is not an MDPV.[110] “Commercial medium- and heavy-duty on-road highway vehicle” means an on-highway vehicle with a gross vehicle weight rating (GVWR) of 10,000 pounds or more.[111] Based on the definitions in EISA, recreational vehicles would be regulated as class 2b-8 vocational vehicles. Neither statutory definition requires that those vehicles encompassed be commercial in nature, instead dividing the medium- and heavy-duty segments based on weight. The definitions of “work truck” and “commercial medium- and heavy-duty on-highway vehicles” collectively encompass the on-highway motor vehicles not covered in the light duty CAFE standards.

RVIA further stated that NHTSA's current fuel efficiency regulations are not consistent with EISA and do not purport to grant NHTSA authority to regulate vehicles simply based on weight. NHTSA's regulations at 49 CFR 523.6 define, by cross-reference the language in 49 U.S.C. 32901(a)(7) and (19), and consistent with the discussion above, include recreational vehicles.

Finally, NHTSA notes that excluding recreational vehicles in Phase 2 could create illogical results, including treating similar vehicles differently, as determinations over whether a given vehicle would be covered by the program would be based upon either its intended or actual use, rather than the actual characteristics of the vehicle. Moreover, including recreational vehicles under NHTSA regulations furthers the agencies' goal of one national program, as EPA regulations will continue to regulate recreational vehicles. NHTSA will allow early compliance for manufacturers that want to certify during the Phase 1 period.

F. Other Issues

In addition to establishing new Phase 2 standards, this document addresses several other issues related to those standards. The agencies are adopting some regulatory provisions related to the Phase 1 program, as well as amendments related to other EPA and NHTSA regulations. These other issues are summarized briefly here and discussed in greater detail in later sections.

(1) Opportunities for Further Oxides of Nitrogen (NOX) Reductions From Heavy-Duty On-Highway Engines and Vehicles

The EPA has the authority under section 202 of the Clean Air Act to establish, and from time to time revise, emission standards for certain air pollutants emitted from heavy-duty on-highway engines and vehicles. The emission standards that EPA has developed for heavy-duty on-highway engines have become progressively more stringent over the past 40 years, with the most recent NOX standards for new heavy-duty on-highway engines fully phased in with the 2010 model year. NOX emissions standards for heavy-duty on-highway engines have contributed significantly to the overall reduction in the national NOX emissions inventory. Nevertheless, a need for additional NOX reductions remains, particularly in areas of the country with elevated levels of air pollution. As discussed further below, in response to EPA's responsibilities under the Clean Air Act, the significant comments we received on this topic during the public comment period, the recent publication by the California Air Resources Board (CARB) of its May 2016 Mobile Source Strategy report and Proposed 2016 Strategy for the State implementation Plan [112] and a recent Petition for Rulemaking,[113] EPA plans to further engage with stakeholders after the publication of this Final Rule to discuss the opportunities for developing more stringent federal standards to further reduce the level of NOX emissions from heavy-duty on-highway engines through a coordinated effort with CARB.

NOX is one of the major precursors of tropospheric ozone (ozone), exposure to which is associated with a number of adverse respiratory and cardiovascular effects, as described in Section VIII.A.2 below. These effects are particularly pronounced among children, the elderly, and among people with lung disease such as asthma. NOX is also a major contributor to secondary PM2.5 formation, and exposure to PM2.5 itself has been linked to a number of adverse health effects (see Section VIII.A.1), such as heart attacks and premature mortality. In addition, NO2 exposure is linked to asthma exacerbation and possibly to asthma development in children (see Section VIII.A.3). EPA has already adopted many emission control programs that are expected to reduce ambient ozone levels. However, the U.S. Energy Information Administration's AEO 2015 predicts that vehicles miles travelled (VMT) for heavy-duty trucks will increase in the coming years,[114] and even with the implementation of all current state and federal regulations, some of the most populous counties in the United States are expected to have ozone air quality that exceeds the National Ambient Air Quality Standards (NAAQS) into the future. As of April 22, 2016, there were 44 ozone nonattainment areas for the 2008 ozone NAAQS composed of 216 full or partial counties, with a population of more than 120 million. These nonattainment areas are dispersed across the country, with counties in the west, northeastern United States, Texas, and several Great Lakes states. The geographic diversity of this problem necessitates action at the national level. In California, the San Joaquin Valley and the South Coast Air Basin are highly-populated areas classified as “extreme nonattainment” for the 2008 8-hour ozone standard, with an attainment demonstration deadline of 2031 (one year in advance of the actual 2032 attainment date). In addition, EPA lowered the level of the primary and secondary NAAQS for the 8-hour standards from 75 ppb to 70 ppb in 2015 (2015 ozone NAAQS),[115] with plans to finalize nonattainment designations for the 2015 ozone NAAQS in October 2017. Further NOX reductions would provide reductions in ambient ozone levels, helping to prevent adverse health impacts associated with ozone exposure and assisting states and local areas in attaining and maintaining the applicable ozone NAAQS. Reductions in NOX emissions would also improve air quality and provide Start Printed Page 73523public health and welfare benefits throughout the country by (1) reducing PM formed by reactions of NOX in the atmosphere; (2) reducing concentrations of the criteria pollutant NO2; (3) reducing nitrogen deposition to sensitive environments; and (4) improving visibility.

In the past year, EPA has received requests from several state and local air quality districts and other organizations asking that EPA establish more stringent NOX standards for heavy-duty on-highway engines to help reduce the public's exposure to air pollution. In its comments, CARB estimated that heavy-duty on-highway vehicles currently contribute about one-third of all NOX emissions in California. In order to achieve the 2008 ozone NAAQS, California has estimated that the state's South Coast Air Basin will need an 80 percent reduction in NOX emissions by 2031. California has the unique ability among states to adopt its own separate new motor engine and vehicle emission standards under section 209 of the CAA; however, CARB commented that EPA action to establish a new federal low-NOX standard for heavy-duty trucks is critical, since California standards alone are not sufficient to demonstrate compliance with either the 2008 ozone NAAQS or the 2015, even more stringent ozone NAAQS. CARB has developed a comprehensive mobile source strategy which for heavy-duty on-highway vehicles includes: Lowering the emissions from the in-use fleet; establishing more stringent NOX standards for new engines; and accelerating the deployment of zero and near-zero emissions technology.[116] In September of 2015, CARB published a draft of this strategy, Mobile Source Strategy Discussion Draft, after which CARB held a public workshop and provided opportunity for public comment. On May 16, 2016, CARB issued a final Mobile Source Strategy report.[117] In this report, CARB provides a comprehensive strategy plan for the future of mobile sources and goods movement in the State of California for how mobile sources in California can meet air quality and climate goals over the next fifteen years. Among the many programs discussed are plans for a future on-highway heavy-duty engine and vehicle NOX control regulatory program for new products with implementation beginning in 2024. CARB states “The need for timely action by U.S. EPA to establish more stringent engine performance standards in collaboration with California efforts is essential. About 60 percent of total heavy-duty truck VMT in the South Coast on any given day is accrued by trucks purchased outside of California, and are exempt from California standards. U.S. EPA action to establish a federal low-NOXstandard for trucks is critical.” CARB lays out a time line for a California specific action for new highway heavy-duty NOX standards with CARB action in 2017-2019 that would lead to new standards that could begin with the model year 2023. CARB also requests that the U.S. EPA work on a Federal rulemaking action in the 2017-2019 time frame which could result in standards that could begin with the model year 2024. The CARB Mobile Source Strategy document also states “Due to the preponderance of interstate trucking's contribution to in-state VMT, federal action would be far more effective at reducing in-state emissions than a California-only standard. However, California is prepared to develop a California-only standard, if needed, to meet federal attainment targets.” CARB goes on to state “[C]ARB will begin development of new heavy-duty low NOX emission standard in 2017 with Board action expected in 2019. ARB may also petition U.S. EPA in 2016 to establish new federal heavy-duty engine emission standards . . . . If U.S. EPA begins the regulatory development process for a new federal heavy-duty emission standard by 2017, ARB will coordinate its regulatory development efforts with the federal regulation.” On May 17, 2016, CARB published its “Proposed 2016 State Strategy for the State Implementation Plan.” [118] This document contains CARB staff's proposed strategy to attain the health-based federal air quality standards over the next fifteen years. With respect to future on-highway heavy-duty NOX standards, the proposed State Implementation Plan is fully consistent with the information published by CARB in the Mobile Source Strategy report. EPA intends to work with CARB to consider the development of a new harmonized Federal and California program that would apply lower NOX emissions standards at the national level to heavy-duty on-highway engines and vehicles.

In addition to CARB, EPA received compelling letters and comments from the National Association of Clean Air Agencies, the Northeast States for Coordinated Air Use Management, the Ozone Transport Commission, and the South Coast Air Quality Management District explaining the critical and urgent need to reduce NOX emissions that significantly contribute to ozone and fine particulate air quality problems in their represented areas. The comments describe the challenges many areas face in meeting both the 2008 and recently strengthened 2015 ozone NAAQS. These organizations point to the significant contribution of heavy-duty vehicles to NOX emissions in their areas, and call upon EPA to begin a rulemaking to require further NOX controls for the heavy-duty sector as soon as possible. Commenters such as the American Lung Association, Environmental Defense Fund, Union of Concerned Scientists, the California Interfaith Power and Light, Coalition for Clean Air/California Cleaner Freight Coalition, and the Moving Forward Network similarly describe the air quality and public health need for NOX reductions and request EPA to lower NOX emissions standards for heavy-duty vehicles. Taken as a whole, the numerous comments, the expected increase in heavy-duty truck VMT, and the fact that ozone challenges will remain across the country demonstrate the critical need for more stringent nationwide NOX emissions standards. Such standards are vital to improving air quality nationwide and reducing public health effects associated with exposure to ozone and secondary PM2.5, especially for vulnerable populations and in highly impacted regions.

On June 3, 2016, the EPA received a Petition for Rulemaking from the South Coast Air Quality Management District (California), the Pima County Department of Environmental Quality (Arizona), the Bay Area Air Quality Management District (California), the Connecticut Department of Energy and Environmental Protection Agency, the Delaware Department of Energy and Environmental Protection, the Washoe County Health District (Nevada), the New Hampshire Department of Environmental Services, the New York City Department of Environmental Protection, the Akron Regional Air Quality Management District (Ohio), the Washington State Department of Ecology, and the Puget Sound Clean Air Start Printed Page 73524Agency (Washington).[119 120] In a June 15, 2016 letter to EPA, the Commonwealth of Massachusetts also joined this petition. On June 22, 2016, the San Joaquin Valley Air Pollution Control District (California) also submitted a petition for rulemaking to EPA.[121] In these Petitions, the Petitioners request that EPA establish a new, lower NOX emission standard for on-road heavy-duty engines. The Petitioners request that EPA implement a new standard by January 1, 2022, and that EPA establish this new standard through a Final Rulemaking issued by December 31, 2017. EPA is not formally responding to this Petition in this Final Rule, but we will do so in a future action. In the petitions, the Petitioners include a detailed discussion of their views and underlying data regarding the need for large scale reduction in NOX emissions from heavy-duty engines, why they believe new standards can be achieved, and their legal views on EPA's responsibilities under the Clean Air Act.

Since the establishment of the current heavy-duty on-highway standards in January of 2001,[122] there has been continued progress in emissions control technology. EPA and CARB are currently investing in research to evaluate opportunities for further NOX reductions from heavy-duty on-highway vehicles and engines. Programs and research underway at CARB, as well as a significant body of work in the technical literature, indicate that reducing NOX emissions significantly below the current on-highway standard of 0.20 grams per brake horsepower-hour (g/bhp-hr) is potentially feasible.[123 124] Opportunities for additional NOX reductions include reducing emissions over cold start operation as well as low-speed, low-load off-cycle operation. Reductions are being accomplished through the use of improved engine management, advanced aftertreatment technologies (improvements in SCR catalyst design/formulation), catalyst positioning, aftertreatment thermal management, and heated diesel exhaust fluid dosing. At the same time, the effect of these new technologies on cost and GHG emissions is being carefully evaluated,124 since it is important that any future NOX control technologies be considered in the context of the final Phase 2 GHG standards. During the Phase 2 program public comment period, EPA received some comments stressing the need for careful evaluation of emerging NOX control technologies and urging EPA to consider the relationship between CO2 and NOX before setting lower NOX standards (commenters include American Trucking Association, Caterpillar, Daimler Trucks North America, Navistar Inc., PACCAR Inc., Volvo Group, Truck and Engine Manufacturers Association, Diesel Technology Forum, National Association of Manufacturers, and National Automobile Dealers Association). EPA also received comments pointing to advances in NOX emission control technologies that would lower NOX without reducing engine efficiency (commenters include Advanced Engine Systems Institute, Clean Energy, Manufacturers of Emission Controls Association, and Union of Concerned Scientists). EPA will continue to evaluate both opportunities and challenges associated with lowering NOX emissions from the current standards, and over the coming months we intend to engage with many stakeholders as we develop our response to the June 2016 Petitions for Rulemaking discussed above.

EPA believes the opportunity exists to develop, in close coordination with CARB and other stakeholders, a new, harmonized national NOX reduction strategy for heavy-duty on-highway engines which could include the following:

  • Substantially lower NOX emission standards;
  • Improvements to emissions warranties;
  • Consideration of longer useful life, reflecting actual in-use activity;
  • Consideration of rebuilding/remanufacturing practices;
  • Updated certification and in-use testing protocols;
  • Incentives to encourage the transition to next-generation cleaner technologies as soon as possible;
  • Improvements to test procedures and test cycles to ensure emission reductions occur in the real-world, not only over the applicable certification test cycles.

Based on the air quality need, the requests described above, the continued progress in emissions control technology, and the June 2016 petitions for rulemaking, EPA plans to engage with a range of stakeholders to discuss the opportunities for developing more stringent federal standards to further reduce the level of NOX emissions from heavy-duty on-highway engines, after the publication of this Final Rule. Recognizing the benefits of a nationally harmonized program and given California's unique ability under CAA section 209 to be allowed to regulate new motor vehicle and engine emission standards if certain criteria are met, EPA intends to work closely with CARB on this effort. EPA also intends to engage with truck and engine manufacturers, suppliers, state air quality agencies, NGOs, labor, the trucking industry, and the Petitioners over the next several months as we develop our formal response to the June 2016 Petitions for Rulemaking.

(2) Issues Related to Phase 2

(a) Natural Gas Engines and Vehicles

This combined rulemaking by EPA and NHTSA is designed to regulate two separate characteristics of heavy duty vehicles and engines: GHGs and fuel consumption. In the case of diesel or gasoline powered vehicles, there is a one-to-one relationship between these two characteristics. For alternatively fueled vehicles, which use no petroleum, the situation is different. For example, a natural gas vehicle that achieves approximately the same fuel efficiency as a diesel powered vehicle will emit 20 percent less CO2; and a natural gas vehicle with the same fuel efficiency as a gasoline vehicle will emit 30 percent less CO2. Yet natural gas vehicles consume no petroleum. The agencies are continuing Phase 1 approach, which the agencies have previously concluded balances these facts by applying the gasoline and diesel CO2 standards to natural gas engines based on the engine type of the natural gas engine. Fuel consumption for these vehicles is then calculated according to their tailpipe CO2 emissions. In essence, this applies a one-to-one relationship between fuel efficiency and tailpipe CO2 emissions for all vehicles, including natural gas vehicles. The agencies determined that this approach will likely create a small balanced incentive for natural gas use. In other words, it created a small incentive for the use of natural gas engines that appropriately balanced concerns about the climate impact methane emissions against other factors such as the energy security Start Printed Page 73525benefits of using domestic natural gas. See 76 FR 57123.

(b) Alternative Refrigerants

In addition to use of low-leak components in air conditioning system design, manufacturers can also decrease the global warming impact of any refrigerant leakage emissions by adopting systems that use alternative, lower global warming potential (GWP) refrigerants, to replace the refrigerant most commonly used today, HFC-134a (R-134a). HFC-134a is a potent greenhouse gas with a GWP 1,430 times greater than that of CO2.

Under EPA's Significant New Alternatives Policy (SNAP) Program,[125] EPA has found acceptable, subject to use conditions, three alternative refrigerants that have significantly lower GWPs than HFC-134a for use in A/C systems in newly manufactured light-duty vehicles: HFC-152a, CO2 (R-744), and HFO-1234yf.[126] HFC-152a has a GWP of 124, HFO-1234yf has a GWP of 4, and CO2 (by definition) has a GWP of 1, as compared to HFC-134a which has a GWP of 1,430.[127] CO2 is nonflammable, while HFO-1234yf and HFC-152a are flammable. All three are subject to use conditions requiring labeling and the use of unique fittings, and where appropriate, mitigating flammability and toxicity. Currently, the SNAP listing for HFO-1234yf is limited to newly manufactured A/C systems in light-duty vehicles, whereas HFC-152a and CO2 have been found acceptable for all motor vehicle air conditioning applications, including heavy-duty vehicles.

None of these alternative refrigerants can simply be “dropped” into existing HFC-134a air conditioning systems. In order to account for the unique properties of each refrigerant and address use conditions required under SNAP, changes to the systems will be necessary. Typically these changes will need to occur during a vehicle redesign cycle but can also occur during a refresh. For example, because CO2, when used as a refrigerant, is physically and thermodynamically very different from HFC-134a and operates at much higher pressures, a transition to this refrigerant would require significant hardware changes. A transition to A/C systems designed for HFO-1234yf, which is more thermodynamically similar to HFC-134a than is CO2, requires less significant hardware changes that typically include installation of a thermal expansion valve and can potentially require resized condensers and evaporators, as well as changes in other components. In addition, vehicle assembly plants require re-tooling in order to handle new refrigerants safely. Thus a change in A/C refrigerants requires significant engineering, planning, and manufacturing investments.

EPA is not aware of any significant development of A/C systems designed to use alternative refrigerants in heavy-duty vehicles.[128] However, all three lower GWP alternatives are in use or under various stages of development for use in LD vehicles. Of these three refrigerants, most manufacturers of LD vehicles have identified HFO-1234yf as the most likely refrigerant to be used in that application. For that reason, EPA anticipates that HFO-1234yf will be a primary candidate for refrigerant substitution in the HD market in the future if it is listed as an acceptable substitute under SNAP for HD A/C applications.

As mentioned above, EPA has listed as acceptable, subject to use conditions, two lower-GWP refrigerants, R-744 (CO2) and HFC-152a, for use in HD vehicles. On April 18, 2016, EPA also proposed to list HFO-1234yf as acceptable, subject to use conditions, in A/C systems for newly manufactured MDPVs, HD pickup trucks, and complete HD vans (81 FR 22810). In that action, EPA proposed to list HFO-1234yf as acceptable, subject to use conditions, for those vehicle types for which human health and environmental risk could be assessed using the currently available risk assessments and analysis on LD vehicles. Also in that action, EPA requested “information on development of HFO-1234yf MVAC systems for other HD vehicle types or off-road vehicles, or plans to develop these systems in the future.” EPA also stated “This information may be used to inform a future listing” (81 FR 22868).

In another rulemaking action under the SNAP program, on July 20, 2015, EPA published a final rule (80 FR 42870) that will change the listing status of HFC-134a to unacceptable for use in newly manufactured LD motor vehicles beginning in MY 2021 (except as allowed under a narrowed use limit for use in newly manufactured LD vehicles destined for use in countries that do not have infrastructure in place for servicing with other acceptable refrigerants through MY 2025). In that same rule, EPA listed the refrigerant blends SP34E, R-426A, R-416A, R-406A, R-414A, R-414B, HCFC Blend Delta, Freeze 12, GHG-X5, and HCFC Blend Lambda as unacceptable for use in newly manufactured light-duty vehicles beginning in MY 2017. EPA's decisions were based on the availability of other substitutes that pose less overall risk to human health and the environment, when used in accordance with required use conditions. Neither the April 2016 proposed rule nor the July 2015 final rule consider a change of listing status for HFC-134a in HD vehicles.

LD vehicle manufacturers are currently making investments in systems designed for lower-GWP refrigerants, both domestically and on a global basis. In support of the LD GHG rule, EPA projected a full transition of LD vehicles to lower-GWP alternatives in the United States by MY 2021. We expect the costs of transitioning to decrease over time as alternative refrigerants are adopted across all LD vehicles and trucks, in part due to increased availability of components and the continuing increases in refrigerant production capacity, as well as knowledge gained through experience. As lower-GWP alternatives become widely used in LD vehicles, some HD vehicle manufacturers may wish to also transition their vehicles. Transitioning could be advantageous for a variety of reasons, including platform standardization and company environmental stewardship policies.

In the proposal for this Phase 2 HD rule, EPA proposed another action related to alternative refrigerants. EPA proposed to allow a manufacturer to be “deemed to comply” with the leakage standard if its A/C system used a refrigerant other than HFC-134a that was both listed as an acceptable substitute refrigerant for heavy-duty A/C systems under SNAP, and was identified in the LD GHG regulations at 40 CFR 86.1867-12(e). 80 FR 40172. By slightly reducing the regulatory burden of compliance with the leakage standard for a manufacturer that used an alternative refrigerant, the “deemed to comply” provision was intended to provide a modest incentive for the use of such refrigerants. There were comments in support of this approach, Start Printed Page 73526including from Honeywell and Chemours, both of which manufacture HFO-1234yf.

For several reasons, EPA has reconsidered the proposed “deemed to comply” provision for this rule, and instead, the Phase 2 program retains the Phase 1 requirement that manufacturers attest that they are using low-leak components, regardless of the refrigerant they use. CARB and several NGO commenters expressed concerns about the proposed “deemed to comply” provision, primarily citing the potential for manufacturers to revert to less leak-tight components if they were no longer required to attest to the use of low-leak A/C system components because they used a lower-GWP refrigerant. In general, we expect that the progress LD vehicle manufacturers are making toward more leak-tight A/C systems will continue and that this progress will transfer to HD A/C systems. Still, we agree that continued improvements in low-leak performance HD vehicles is an important goal, and that continuing the Phase 1 leakage requirements in the Phase 2 program should discourage manufacturers from reverting to higher-leak and potentially less expensive components. It is also important to note that there is no “deemed to comply” option in the parallel LD-GHG program—manufacturers must attest to meeting the leakage standard. There is no compelling reason to have a different regime for heavy duty applications.

Although leakage of lower-GWP refrigerants is of less concern from a climate perspective than leakage of higher GWP refrigerants, we also agree with several commenters that expressed a concern related to the servicing of lower-GWP systems with higher-GWP refrigerants in the aftermarket. We agree that this could result due to factors such as price differentials between aftermarket refrigerants. However, as is the case for Phase 1, as a part of certification, HD manufacturers will attest both to the use of low-leak components as well as to the specific refrigerant used. Thus, in the future, a manufacturer wishing to certify a vehicle with an A/C system designed for an alternative refrigerant will attest to the use of that specific refrigerant. In that situation, any end-user servicing and recharging that A/C system with any other refrigerant would be considered tampering with an emission-related component under Title II of the CAA. For example, recharging an A/C system certified to use a lower-GWP refrigerant, such as HFO-1234yf, with any other refrigerant, including but not limited to HFC-134a, would be considered a violation of Title II tampering provisions.

At the same time, EPA does not believe that finalizing the “deemed to comply” provision would have had an impact on any future transition of the HD industry to alternative refrigerants. As discussed above, two lower-GWP refrigerants are already acceptable for use in HD vehicles, and EPA has proposed to list HFO-1234yf as acceptable, subject to use conditions, for limited HD vehicle types. As also discussed above, and especially in light of the rapid expansion of alternative refrigerants that has been occurring in the LD vehicle market, similar trends may develop in the HD vehicle market, regardless of EPA's action regarding leakage of alternative refrigerants in this final rule.

(c) Small Business Issues

The Regulatory Flexibility Act (RFA) generally requires an agency to prepare a regulatory flexibility analysis of any rule subject to notice and comment rulemaking requirements under the Administrative Procedure Act or any other statute unless the agency certifies that the rule will not have a significant economic impact on a substantial number of small entities. See generally 5 U.S.C. 601-612. The RFA analysis is discussed in Section XIV.

Pursuant to section 609(b) of the RFA, as amended by the Small Business Regulatory Enforcement Fairness Act (SBREFA), EPA also conducted outreach to small entities and convened a Small Business Advocacy Review Panel to obtain advice and recommendations of representatives of the small entities that potentially will be subject to the rule's requirements. Consistent with the RFA/SBREFA requirements, the Panel evaluated the assembled materials and small-entity comments on issues related to elements of the Initial Regulatory Flexibility Analysis (IRFA). A copy of the Panel Report was included in the docket for this rule.

The agencies previously determined that the Phase 2 regulations could potentially have a significant economic impact on small entities. Specifically, the agencies identified four categories of directly regulated small businesses that could be impacted:

  • Trailer Manufacturers
  • Alternative Fuel Converters
  • Vocational Chassis Manufacturers
  • Glider Vehicle [129] Assemblers

To minimize these impacts the agencies are adopting certain regulatory flexibilities—both general and category-specific. In general, we are delaying new requirements for EPA GHG emission standards by one initial year and simplifying certification requirements for small businesses. Even with this one year delay, small businesses will be required to comply with EPA's standards before NHTSA's fuel efficiency standards are mandatory. Because of this timing, compliance with NHTSA's regulations will not be delayed, as small business manufacturers will be accommodated through EPA's initial one year delay. The agencies are also providing the following specific relief:

  • Trailers: Adopting simpler requirements for non-box trailers, which are more likely to be manufactured by small businesses; reduced reliance on emission averaging; and making third-party testing easier for certification.
  • Alternative Fuel Converters: Omitting recertification of a converted vehicle when the engine is converted and certified; reduced N2 O testing; and simplified onboard diagnostics and delaying required compliance with each new standard by one model year.
  • Vocational Chassis: Less stringent standards for certain vehicle categories; opportunity to generate credits under the Phase 1 program.
  • Glider Vehicle Assemblers:[130] Exempting existing small businesses, but limiting the small business exemption to a capped level of annual production (production in excess of the capped amount will be allowed, but subject to all otherwise applicable requirements including the Phase 2 standards). Providing additional flexibility for newer engines.

These flexibilities are described in more detail in Section XIV, in RIA Section 12 and in the Panel Report. Flexibilities specific to glider vehicle assemblers are described in Section XIII.

(d) Confidentiality of Test Results and GEM Inputs

The agencies received mixed comments regarding the question of whether GEM inputs should be made available to public. Some commenters supported making this information available, while others thought it should Start Printed Page 73527be protected as confidential business information (CBI). In accordance with Federal statutes, EPA does not release information from certification applications (or other compliance reports) that we determine to be CBI under 40 CFR part 2. Consistent with section 114(c) of the CAA, EPA does not consider emission test results to be CBI after introduction into commerce of the certified engine or vehicle. (However, we have generally treated test results as protected before the introduction into commerce date). EPA has not yet made a final determination for Phase 1 or Phase 2 certification test results. Nevertheless, at this time we expect to continue this policy and consider it likely that we would not treat any test results or other GEM inputs as CBI after the introduction into commerce date as identified by the manufacturer.

With regard to NHTSA's treatment of confidential business information, manufacturers must submit a request for confidentiality with each electronic submission specifying any part of the information or data in a report that it believes should be withheld from public disclosure as trade secret or other confidential business information. A form is available through the NHTSA Web site to request confidentiality. NHTSA does not consider manufacturers to continue to have a business case for protecting pre-model report data after the vehicles contained within that report have been introduced into commerce.

(e) Delegated Assembly and Secondary Manufacturers

In EPA's existing regulations (40 CFR 1068.261), we allow engine manufacturers to sell or ship engines that are missing certain emission-related components if those components will be installed by the vehicle manufacturer. These provisions already apply to Phase 1 vehicles as well, providing a similar allowance for vehicle manufacturers to sell or ship vehicles that are missing certain emission-related components if those components will be installed by a secondary vehicle manufacturer. See section 1037.620. EPA has found this provision to work well and is finalizing certain amendments in this rule. See 40 CFR 1037.621. Under the amended rule, as conditions of this allowance, manufacturers will be required to:

  • Have a contractual obligation with the secondary manufacturer to complete the assembly properly and provide instructions about how to do so
  • Keep records to demonstrate compliance
  • Apply a temporary label to the incomplete vehicles
  • Take other reasonable steps to ensure the assembly is completed properly
  • Describe in its application for certification how it will use this allowance

Under delegated assembly, it is the upstream manufacturer that holds the certificate and assumes primary responsibility for all compliance requirements. Our experience applying this approach has shown that holding the upstream manufacturer responsible ensures that they will exercise due diligence throughout the process.

EPA proposed to apply this new section broadly. However, commenters raised valid questions about whether it is necessary to apply this formal process as broadly as proposed. In response, we have reconsidered the proposed approach and have determined that it would be appropriate to allow a less formal process with components for which market forces will make it unlikely that a secondary manufacturer would not complete assembly properly. In those cases, the certifying manufacturers will be required to provide sufficiently detailed installation instructions to the secondary manufacturers, who would then be obligated to complete assembly properly before the vehicles are delivered to the ultimate purchasers.

One example of a case for which market forces could ensure that assembly is completed properly would be air conditioning leakage requirements. Purchasers will have the expectation that the systems will not leak, and a secondary manufacturer should have no incentive to not follow the certifying manufacturer's instructions.

As revised, § 1037.621 will require the formal delegated assembly process for the following technologies if they are part of the OEM's certified configuration but not shipped with the vehicle:

  • Auxiliary power units
  • Aerodynamic devices
  • Hybrid components
  • Natural gas fuel tanks

Certificate holders will remain responsible for other certified components, but will not automatically be required to comply with the formal delegated assembly requirements. That determination will be made case-by-case as part of the certification process. We are also explicitly making the flexibility in 40 CFR 1037.621 available for HD pickups and vans certified to the standards in 40 CFR part 86. As is currently specified in 40 CFR 1068.261, EPA will retain the authority to apply additional necessary conditions (at the time of certification) to the allowance to delegate assembly of emission to secondary manufacturers (when emission control equipment is not shipped with the vehicle to the secondary manufacturer, as just noted). In particular, we would likely apply such additional conditions for manufacturers that we determine to have previously not completed assembly properly. Issues of delegated assembly are addressed in more detail in Section 1.4.4 of the RTC.

(f) Engine/Vehicle Useful Life

We received comment on what policies we should adopt to address the situation where the engine and the vehicle are subject to emission standards over different useful-life periods. For example, a medium heavy-duty engine may power vehicles in weight classes ranging from 2b to 8, with correspondingly different regulatory useful lives for those vehicles. As provided in 40 CFR 1037.140 of the final regulations, we have structured the vehicle regulations to generally apply the same useful life for the vehicle that applies for the engines. However, these regulations also allow vehicle manufacturers to certify their vehicles to longer useful lives. The agencies see no problem with allowing vehicles to have longer useful lives than the engines.

(g) Compliance Reports

The agencies received comment on the NPRM from two environmental organizations requesting that the agencies make available to the public data and information that would enable the public to track trends in technology sales over time, as well as track company-specific compliance data. The commenters suggested that this should include an agency publication of an annual compliance report for the Heavy-duty Phase 2 program. The commenters requested this information to allow all stakeholders to see how individual companies, as well as the industry overall, were performing relative to their compliance obligations (see comments from ACEEE and NRDC).

The agencies agree with this comment. In the context of the light-duty vehicle GHG standards, EPA has already published four annual compliance reports which has made available to the public detailed information regarding both how individual light-duty vehicle companies have been meeting their compliance obligations, as well as summary information at the light-duty fleet level. NHTSA makes the up-to-date information on the light-duty fuel economy program available through its Start Printed Page 73528CAFE Public Information Center (http://www.nhtsa.gov/​CAFE_​PIC/​CAFE_​PIC_​Home.htm). Information includes manufacturer and overall fleet standards and CAFE performance, credit status, and civil penalty status. This information has been helpful to increase transparency to all stakeholders and to allow the public to see how companies are progressing from one year to the next with respect to their compliance requirements. It is EPA's intention to publish a similar annual compliance report for the heavy duty GHG program, covering both the existing Phase 1 program, as well as the Phase 2 standards contained in this final rule. It is NHTSA's intention to expand the Public Information Center to include the medium- and heavy-duty fuel efficiency program and to make up-to-date information collected in the heavy-duty fuel efficiency compliance process available publicly. Both the EPA and NHTSA compliance reports will provide available information at the vehicle subclass level for each of the four vehicle categories (i.e. Tractors, Trailers, Vocational, and Heavy-Duty Pickups and Vans), and EPA will provide available information for the other GHG standards, such as N2 O and refrigerant leak detection standards. Prior to issuing the compliance reports, EPA and NHTSA will work with regulated manufacturers to reconcile concerns over the release of claimed confidential business information, consistent with 40 CFR part 2 and 49 CFR 512.

(3) Life Cycle Emissions

The agencies received many comments expressing concerns about establishing the GHG and fuel consumption standards as tailpipe standards that do not account for upstream emissions or other life cycle impacts. However, many other commenters supported this approach. Comments specifically related to alternative fuels or electric vehicles are addressed in Section I.C.(1)(d) and in Section XI.B. This section addresses the issue more broadly.

As discussed below, the agencies do not see how we could accurately account for life cycle emissions in our vehicle standards, nor have commenters shown that such an accounting is needed. In addition, NHTSA has already noted that the fuel efficiency standards are necessarily tailpipe-based, and that a lifecycle approach would likely render it impossible to harmonize the fuel efficiency and GHG emission standards, to the great detriment of our goal of achieving a national, harmonized program. See 76 FR 57125.

It is also worth noting that EPA's engine and vehicle emission standards and NHTSA's vehicle fuel consumption standards (including those for light-duty vehicles) have been in place for decades as tailpipe standards. The agencies find no reasonable basis in the comments or elsewhere to change fundamentally from this longstanding approach.

Although the final standards do not account for life cycle emissions, the agencies have estimated the upstream emission impact of reducing fuel consumption for heavy-duty vehicles. As shown in Section VII and VIII, these upstream emission reductions are significant and worth estimating, even with some uncertainty. However, this analysis would not be a sufficient basis for inclusion in the standards themselves.

(a) Challenges for Addressing Life Cycle Emissions With Vehicle Standards

Commenters supporting accounting for life cycle emissions generally did so in the context of one or more specific technologies. However, the agencies cannot accurately address life-cycle emissions on a technology specific basis at this time for two reasons:

  • We lack data to address each technology, and see no path to selectively apply a life cycle analysis to some technologies, but not to others.
  • Actual life cycle emissions are dependent on factors outside the scope of the rulemaking that may change in the future.

With respect to the first reason, even if we were able to accurately and fully account for life cycle impacts of one technology (such as weight reduction), this would not allow us to address life cycle emissions for other technologies. For example, how would the agencies address potential differences in life cycle emissions for shifting from a manual transmission to and AMT, or the life cycle emissions of aerodynamic fairings? If we cannot factor in life cycle impacts for all technologies, how would we do it for weight reductions? Given the complexity of these rules and the number of different technologies involved, we see no way to treat the technologies equitably. Commenters do not provide the information necessary to address this challenge, nor are the agencies aware of such information.

The second reason is just as problematic. This rulemaking is setting standards for vehicles under specific statutory provisions. It is not regulating manufacturing processes, distribution practices, or the locations of manufacturing facilities. And yet each of these factors could impact life cycle emissions. So while we could take a snapshot of life cycle emissions at this point in time for specific manufacturers, it may or may not have any relation to life cycle emissions in 2027, or for other manufacturers. Consider, for example, two component manufacturers: One that produces its components near the vehicle assembly plant, and relies on natural gas to power its factory; and a second that is located overseas and relies on coal-fired power. How would the agencies equitably (or even non-arbitrarily) factor in these differences without regulating these processes? To the extent commenters provided any information on life cycle impacts, they did not address this challenge.

(b) Need for Life Cycle Consideration in the Standards

The agencies acknowledge that a full and accurate accounting of life cycle emissions (if it were possible) could potentially make the Phase 2 program marginally better. However, we do not agree that this is an issue of fundamental importance. While some commenters submitted estimates of the importance of life cycle emissions for light-duty vehicles, life cycle emissions are less important for heavy-duty vehicles. Consider, for example, the difference between a passenger car and a heavy-duty tractor. If the passenger car achieves 40 mile per gallon and travels 150,000 miles in its life, it would consume less than 4,000 gallons of fuel in its life. On the other hand, a tractor that achieves 8 miles per gallon and travels 1,000,000 miles would consume 125,000 gallons of fuel in its life, or more than 30 times the fuel of the passenger car. Commenters provide no basis to assume the energy consumption associated with tractor production would be 30 times that of the production of a passenger car.

(4) Amendments to the Phase 1 Program

The agencies are revising some test procedures and compliance provisions used for Phase 1. These changes are described in Section XII. This includes both amendments specific to Phase 1, as well as amendments that apply more broadly than Phase 1, such as the revisions to the delegated assembly provisions. As a drafting matter, EPA notes that we are moving the GHG standards for Class 2b and 3 pickups and vans from 40 CFR 1037.104 to 40 CFR 86.1819-14.

NHTSA is also amending 49 CFR part 535 to make technical corrections to its Phase 1 program to better align with EPA's compliance approach, standards and CO2 performance results. In general, these changes are intended to improve the regulatory experience for regulated Start Printed Page 73529parties and also reduce agency administrative burden. More specifically, NHTSA is changing the rounding of its standards and performance values to have more significant digits. Increasing the number of significant digits for values used for compliance with NHTSA standards reduces differences in credits generated and overall credit balances for the EPA and NHTSA programs. NHTSA is also removing the petitioning process for off-road vehicles, clarifying requirements for the documentation needed for submitting innovative technology requests in accordance with 40 CFR 1037.610 and 49 CFR 535.7, and adding further detail to requirements for submitting credit allocation plans as specified in 49 CFR 535.9. Finally, NHTSA is adding the same recordkeeping requirements that EPA currently requires to facilitate in-use compliance inspections. These changes are intended to improve the regulatory experience for regulated parties and also reduce agency administrative burden.

The agencies received few comments on these changes, with most supporting the proposed changes or suggesting improvements. These comments as well as the few comments opposing any of these changes are discussed in Section XII and in the RTC.

(5) Other Amendments to EPA Regulations

EPA is finalizing certain other changes to regulations that we proposed, which are not directly related to the HD Phase 1 or Phase 2 programs, as detailed in Section XIII. For these amendments, there are no corresponding changes in NHTSA regulations. Some of these amendments relate directly to heavy-duty highway engines, but not to the GHG programs. Others relate to nonroad engines. This latter category reflects the regulatory structure EPA uses for its mobile source regulations, in which regulatory provisions applying broadly to different types of mobile sources are codified in common regulatory parts such as 40 CFR part 1068. This approach creates a broad regulatory structure that regulates highway and nonroad engines, vehicles, and equipment collectively in a common program. Thus, it is appropriate to include some amendments to nonroad regulations in addition to the changes applicable only for highway engines and vehicles.

Except as noted below, the agencies received relatively few significant comments on these issues. All comments are discussed in more detail in Section XIII and in the RTC. One area, for which we did receive significant comment was the issue of competition vehicles. As described in Section XIII, EPA is not finalizing the proposed clarification related to highway vehicles used for competition.

(a) Standards for Engines Installed In Glider Kits

EPA regulations currently allow used pre-2013 engines to be installed into new glider kits without meeting currently applicable standards. As described in Section XIII.B, EPA is amending its regulations to allow only engines that have been certified to meet standards for the model year in which the glider vehicle is assembled (i.e. current model year engine standards) to be installed in new glider kits, with certain exceptions. First, engines certified to earlier MY standards that are identical to the current model year standards may be used. Second, engines still within their useful life (and certain similar engines) may be used. Note that this would not allow use of the pre-2002 engines that are currently being used in most glider vehicles because they all would be outside of the 10-year useful life period. Finally, the interim small manufacturer allowance for glider vehicles will also apply for the engines used in the exempted glider kits. Comments on this issue are summarized and addressed in Section XIII.B and in RTC Section 14.2.

(b) Nonconformance Penalty Process Changes

Nonconformance penalties (NCPs) are monetary penalties established by regulation that allow a vehicle or engine manufacturer to sell engines that do not meet the emission standards. Manufacturers unable to comply with the applicable standard pay penalties, which are assessed on a per-engine basis.

On September 5, 2012, EPA adopted final NCPs for heavy heavy-duty diesel engines that could be used by manufacturers of heavy-duty diesel engines unable to meet the current oxides of nitrogen (NOX) emission standard. On December 11, 2013 the U.S. Court of Appeals for the District of Columbia Circuit issued an opinion vacating that Final Rule. It issued its mandate for this decision on April 16, 2014, ending the availability of the NCPs for the current NOX standard, as well as vacating certain amendments to the NCP regulations due to concerns about inadequate notice. In particular, the amendments revise the text explaining how EPA determines when NCP should be made available. In the Phase 2 NPRM, EPA re-proposed most of these amendments to provide fuller notice and additional opportunity for public comment. As discussed in Section XIII, although EPA received one comment opposing these amendments, they are being finalized as proposed.

(c) Updates to Heavy-Duty Engine Manufacturer In-Use Testing Requirements

EPA and manufacturers have gained substantial experience with in-use testing over the last four or five years. This has led to important insights in ways that the test protocol can be adjusted to be more effective. We are accordingly making changes to the regulations in 40 CFR part 86, subparts N and T.

(d) Extension of Certain 40 CFR Part 1068 Provisions to Highway Vehicles and Engines

As part of the Phase 1 GHG standards, we applied the exemption and importation provisions from 40 CFR part 1068, subparts C and D, to heavy-duty highway engines and vehicles. We also specified that the defect reporting provisions of 40 CFR 1068.501 were optional. In an earlier rulemaking, we applied the selective enforcement auditing under 40 CFR part 1068, subpart E (75 FR 22896, April 30, 2010). We are adopting the rest of 40 CFR part 1068 for heavy-duty highway engines and vehicles, with certain exceptions and special provisions.

As described above, we are applying all the general compliance provisions of 40 CFR part 1068 to heavy-duty engines and vehicles subject to 40 CFR parts 1036 and 1037. We are also applying the recall provisions and the hearing procedures from 40 CFR part 1068 for highway motorcycles and for all vehicles subject to standards under 40 CFR part 86, subpart S.

EPA is updating and consolidating the regulations related to formal and informal hearings in 40 CFR part 1068, subpart G. This will allow us to rely on a single set of regulations for all the different categories of vehicles, engines, and equipment that are subject to emission standards. We also made an effort to write these regulations for improved readability.

We are also making a number of changes to part 1068 to correct errors, to add clarification, and to make adjustments based on lessons learned from implementing these regulatory provisions.

(e) Amendments to Engine and Vehicle Test Procedures in 40 CFR Parts 1065 and 1066

EPA is making several changes to our engine testing procedures specified in Start Printed Page 7353040 CFR part 1065. None of these changes will significantly impact the stringency of any standards.

(f) Amendments Related to Marine Diesel Engines in 40 CFR Parts 1042 and 1043

EPA's emission standards and certification requirements for marine diesel engines under the Clean Air Act and the act to Prevent Pollution from Ships are identified in 40 CFR parts 1042 and 1043, respectively. EPA is amending these regulations with respect to continuous NOX monitoring and auxiliary engines, as well as making several other minor revisions.

(g) Amendments Related to Locomotives in 40 CFR Part 1033

EPA's emission standards and certification requirements for locomotives under the Clean Air Act are identified in 40 CFR part 1033. EPA is making several minor revisions to these regulations.

(6) Other Amendments to NHTSA Regulations

NHTSA proposed to amend 49 CFR parts 512 and 537 to allow manufacturers to submit required compliance data for the Corporate Average Fuel Economy (CAFE) program electronically, rather than submitting some reports to NHTSA via paper and CDs and some reports to EPA through its VERIFY database system. NHTSA is not finalizing this proposal in this rulemaking and will consider electronic submission for CAFE reports in a future action.

II. Vehicle Simulation and Separate Engine Standards for Tractors and Vocational Chassis

A. Introduction

This Section II. describes two regulatory program elements that are common among tractors and vocational chassis. In contrast, Sections III and V respectively describe the regulatory program elements that are unique to tractors and to vocational chassis. The common elements described here are the vehicle simulation approach to vehicle certification and the separate standards for engines. Section II.B discusses the reasons for this Phase 2 regulatory approach; namely, requiring vehicle simulation for tractor and vocational chassis certification, maintaining separate engine standards, and expanding and updating their related mandatory and optional test procedures. Section II.C discusses in detail the evolution and final version of the vehicle simulation computer program, which is called the Greenhouse gas Emissions Model or “GEM.” Section II.C also discusses the evolution and final versions of the test procedures for determining the GEM inputs that are common for tractors and vocational chassis. Section II.D discusses in detail the separate engine standards for GHGs and fuel efficiency and their requisite test procedures.

In this final action, the agencies have built on the success of the Phase 1 GEM-based approach for the certification of tractors and vocational chassis. To better recognize the real-world impact of vehicle technologies, we have expanded the number of required and optional vehicle inputs into GEM. Inputting these additional details into GEM results in more accurate representations of vehicle performance and greater opportunities to demonstrate reductions in CO2 emissions and fuel consumption. We are also finalizing revisions to the vehicle driving patterns that are programmed into GEM to better reflect real-world vehicle operation and the emissions reductions that result from applying GHG and fuel efficiency technologies to vehicles. As a result of these revisions, the final GEM-based vehicle certification approach necessitates new testing of engines and testing of some other vehicle components to generate the additional GEM inputs for Phase 2. More detail is provided in Section II.C.

Based on our assessments of the technological feasibility; cost effectiveness; requisite lead times for implementing new and additional tractor and vocational vehicle technologies; and based on comments we received in response to our notice of proposed rulemaking and in response to our more recent notice of additional data availability, the agencies are finalizing steadily increasing stringencies of the CO2 and fuel consumption standards for tractors and vocational chassis for vehicle model years 2021, 2024 and 2027. See Section I or Sections III and V respectively for these numerical standards for tractors and vocational chassis. As part of our analytical process for determining the numerical values of these standards, the agencies utilized GEM. Using GEM as an integral part of our own standard-setting process helps ensure consistency between our technology assessments and the GEM-based certification process that we require for compliance with the Phase 2 standards. Our utilization of GEM in our standard-setting process is described further in Section II.C.

For Phase 2 we are finalizing, as proposed, the same Phase 1 certification approach for all of the GHG and fuel efficiency separate engine standards for those engines installed in tractors and vocational chassis. For the separate engine standards, we will continue to require the Phase 1 engine dynamometer certification test procedures, which were adopted substantially from EPA's existing heavy-duty engine emissions test procedures. In this action we are finalizing, as proposed, revisions to the weighting factors of the tractor engine 13-mode steady-state test cycle (i.e., the Supplemental Engine Test cycle or “SET”). The SET is required for determining tractor engine CO2 emissions and fuel consumption. Consistent with the rationale we presented in our proposal and consistent with comments we received, these revised SET weighting factors better reflect the lower engine speed operation of modern engines, which frequently occurs at tractor cruise speeds. We used these revised weighting factors as part of our engine technology assessments of both current engine technology (i.e., our “baseline engine” technology) and future engine technology.

Based on our assessments of the technological feasibility; cost effectiveness; requisite lead times for implementing new and additional engine technologies; and based on comments we received in response to our notice of proposed rulemaking and in response to our more recent notice of additional data availability, the agencies are finalizing steadily increasing stringencies of the CO2 and fuel consumption separate engine standards for engine model years 2021, 2024 and 2027. In addition, for each of these model years, EPA is maintaining the Phase 1 separate engine standards for CH4 and N2 O emissions—both at their Phase 1 numeric values. While EPA is not finalizing at this time more stringent N2 O emissions standards, as originally proposed, EPA may soon revisit these separate engine N2 O standards in a future rulemaking. All of the final Phase 2 separate engine standards are presented in Section II.D, along with our related assessments.

B. Phase 2 Regulatory Structure

As proposed, in this final action the agencies have built on the success of the Phase 1 GEM-based approach for the certification of tractors and vocational chassis, while also maintaining the Phase 1 separate engine standards approach to engine certification. While the regulatory structures of both Phase 1 and Phase 2 are quite similar, there are a number of new elements for Phase 2. Note that we are not applying these new Start Printed Page 73531Phase 2 elements for compliance with the Phase 1 standards.

These modifications for Phase 2 are consistent with the agencies' Phase 1 commitments to consider a range of regulatory approaches during the development of future regulatory efforts (76 FR 57133), especially for vehicles not already subject to full vehicle chassis dynamometer testing. For example, we committed to consider a more sophisticated approach to vehicle testing to more completely capture the complex interactions within the total vehicle, including the engine and powertrain performance. We also committed to consider the potential for full vehicle certification of complete tractors and vocational chassis using a chassis dynamometer test procedure. We also considered chassis dynamometer testing of complete tractors and vocational chassis as a complementary approach for validating a more complex vehicle simulation approach. We committed to consider the potential for a regulatory program for some of the trailers hauled by tractors. After considering these various approaches, the agencies proposed a structure in which regulated tractor and vocational chassis manufacturers would additionally enter engine and powertrain-related inputs into GEM, which was not part of in Phase 1.

The basic structure in the proposal was widely supported by commenters, although some commenters supported changing certain aspects. Some commenters suggested revising GEM to recognize additional technologies, such as tire pressure monitoring systems and electronic controls that decrease fuel consumption while a vehicle is coasting. To the extent that the agencies were able to collect and receive sufficient data to support such revisions in GEM, these changes were made. See Section II.C. for details. For determining certain GEM inputs, some commenters suggested more cost-effective test procedures for separate engine and transmission testing, compared to the engine-plus-transmission powertrain test procedure that the agencies proposed. In collaboration with researchers at engine manufacturer test laboratories, at Oak Ridge National Laboratory and at Southwest Research Institute, the agencies completed a number of laboratory evaluations of these suggested test procedures.[131] Based on these results, which were made available to the public for a 30-day comment period in the NODA, the agencies are finalizing these more cost-effective test procedures as options, in addition to the powertrain test procedure we proposed. We note that we are also finalizing some of these more cost-effective test procedures, the cycle average approach for all vehicle cycles, as optional for the testing of “pre-transmission” hybrids. In response to our request for comment, some commenters expressed support for a so-called, “cycle-average” approach for generating engine map data for input into GEM. This approach facilitates an accurate recognition of an engine's transient performance. The agencies further refined this approach, and we made detailed information on this approach available in the NODA.[132] Based on comments, we are finalizing this approach as mandatory for mapping engines over GEM's transient cycle, and we are allowing this approach as optional for GEM's 55 mph and 65 mph cycles.

Some commenters expressed concern about GEM and our proposed tractor standards appropriately accounting for the performance of powertrain technologies installed in some of the largest specialty tractors. We have addressed this concern by finalizing a new “heavy-haul” tractor sub-category, with a unique payload and vehicle masses in GEM, which result in a unique set of numeric standards for these vehicles. This is explained in detail in Section III.D. Other commenters expressed concern about the greater complexity of GEM's additional inputs and the appropriateness of our proposed vocational chassis standards, as applied to certain custom-built vocational chassis. We have addressed these concerns by finalizing a limited number of optional custom chassis standards, tailored according to a vocational chassis' final application (e.g., school bus, refuse truck, cement mixer, etc.). To address the concerns about GEM's complexity for these specialty vehicles, these optional custom chassis standards require a smaller number of GEM inputs. This is explained in detail in Section V.D.

Some vehicle manufacturers did not support the agencies finalizing separate engine standards. However, as described below, the agencies continue to believe that separate engine standards are necessary and appropriate. Thus, the agencies are finalizing the basic rule structure that was proposed, but with a number of refinements.

For trailer manufacturers, which will be subject to first-time standards under Phase 2, we will apply the standards using a GEM-based certification, but to do so without actually running GEM. More specifically, based on the agencies' analysis of the results of running GEM many times and varying GEM's trailer configurations, the agencies have developed a simple equation that replicates GEM results, based on inputting certain trailer values into the equation. Use of the equation, rather than full GEM, should significantly facilitate trailer certification. As described in Chapter 2.10.5 of the RIA, the equation has a nearly perfect correlation with GEM, so that they can be used instead of GEM, without impacting stringency. This is a result of the relative simplicity of the trailer inputs as compared to the tractor and vocational vehicle inputs.

(1) Other Structures Considered

To follow-up on the commitment to consider other approaches, the agencies spent significant time and resources before the proposal in evaluating six different options for demonstrating compliance with the proposed Phase 2 standards as shown in Figure II.1

Start Printed Page 73532

As shown in Figure II.1 these six options include:

1. Full vehicle simulation, where vehicle inputs are entered into simulation software.

2. Vehicle simulation, supplemented with separate engine standards.

3. Controllers-in-the-loop simulation, where an actual electronic transmission controller module (TCM) and an actual engine controller module (ECM) are tested in hardware.

4. Engine-in-the-loop simulation, with or without a TCM, where at least the engine is tested in hardware.

5. Vehicle simulation with powertrain-in-the-loop, where the engine and transmission are tested in hardware. One variation involves an engine standard.

6. Full vehicle chassis dynamometer testing.

The agencies evaluated these options in terms of the capital investment required of regulated manufacturers to conduct the testing and/or simulation, the cost per test, the accuracy of the simulation, and the challenges of validating the results. Other considerations included the representativeness compared to the real world behavior, maintaining existing Phase 1 certification approaches that are known to work well, enhancing the Phase 1 approaches that could use improvements, the alignment of test procedures for determining GHG and non-GHG emissions compliance, and the potential to circumvent the intent of the test procedures. The agencies presented our evaluations in the proposal, and we received comments on some of these approaches, and these comments were considered carefully in our evaluations for this final action. Notably, in this final action we are adopting a combination of these options, where some are mandatory and others are optional for certification via GEM. We have concluded that this combination of these options strikes an optimal balance between their costs, accuracy with respect to real-world performance, and robustness for ensuring compliance. In this section we present our evaluation and rationale for finalizing these Phase 2 certification approaches.

Chassis dynamometer testing (Option 6) is used extensively in the development and certification of light-duty vehicles. It also is used in Phase 1 to certify complete Class 2b/3 pickups and vans, as well as to certify certain incomplete vehicles (at the manufacturer's option). The agencies considered chassis dynamometer testing more broadly as a heavy-duty fuel efficiency and GHG certification option because chassis dynamometer testing has the ability to evaluate a vehicle's performance in a manner that most closely resembles the vehicle's in-use performance. Nearly all of the fuel efficiency technologies can be evaluated simultaneously on a chassis dynamometer, including the vehicle systems' interactions that depend on the behavior of the engine, transmission, and other vehicle electronic controllers. One challenge associated with the application of wide-spread heavy-duty chassis testing is the small number of heavy-duty chassis test sites that are available in North America. As discussed in RIA Chapter 3, the agencies were only able to locate 11 heavy-duty chassis test sites. However, more recently we have seen an increased interest in building new sites since issuing the Phase 1 Final Rule. For example, EPA is currently building a heavy-duty chassis dynamometer with the ability to test up to 80,000 pound vehicles at the National Vehicle and Fuel Emissions Laboratory in Ann Arbor, Michigan.

Nevertheless, the agencies continue to be concerned about requiring a chassis test procedure for certifying tractors or vocational chassis due to the initial cost of a new test facility and the large number of heavy duty tractor and vocational chassis variants that could require testing. We have also concluded that for heavy-duty tractors and vocational chassis, there can be increased test-to-test variability under chassis dynamometer test conditions, versus other approaches. First, the agencies recognize that such testing Start Printed Page 73533requires expensive, specialized equipment that is not widely available. The agencies estimate that it would vary from about $1.3 to $4.0 million per new test site depending on existing facilities.[133] In addition, the large number of heavy-duty vehicle configurations would require significant amounts of testing to cover the sector. For example, for Phase 1 tractor manufacturers typically certified several thousand variants of one single tractor model. Finally, EPA's evaluation of heavy-duty chassis dynamometer testing has shown that the variation of chassis test results is greater than light-duty testing, up to 3 percent worse, based on our sponsored testing at Southwest Research Institute.[134] The agencies' research identified a number of unique sources of test-to-test variability in HD chassis dynamometer testing versus other types of testing (described next). These unique sources include variations in HD tire performance and tire temperature and pressure stability; variations in human driver performance; and variations in the test facilities' heating, ventilation and air conditioning system affecting emissions after-treatment performance (e.g., increased fuel consumption to maintain after-treatment temperature) and engine accessory power (e.g., engine fan clutching). Although the agencies are not requiring chassis dynamometer certification of tractors and vocational chassis, we believe such an approach could potentially be appropriate in the future for some heavy duty vehicles if more test facilities become available and if the agencies are able to address the large number of vehicle variants that might require testing and the unique sources of test-to-test variability. Note, as discussed in Section II.C.(4) we are finalizing a manufacturer-run complete tractor heavy-duty chassis dynamometer test program for monitoring relative trends fuel efficiency and for comparing those trends to the trends indicated via GEM simulation. While the agencies did not receive significant comment on the appropriateness of full vehicle heavy-duty chassis dynamometer testing for certification, the agencies did receive significant, mostly negative, comment on the costs versus benefits of a manufacturer-run complete tractor heavy-duty chassis dynamometer test program for data collection. These comments and our responses are detailed in Section II.C.(4).

Another option considered for certification involves testing a vehicle's powertrain in a modified engine dynamometer test facility, which is part of option 5 shown in Figure II.1. In this case the engine and transmission are installed together in a laboratory test facility, and a dynamometer is connected to the output shaft of the transmission. GEM or an equivalent vehicle simulation computer program is then used to control the dynamometer to simulate vehicle speeds and loads. The step-by-step test procedure considered for this option was initially developed as an option for hybrid powertrain testing for Phase 1. We are not finalizing this approach as mandatory, but we are allowing this as an option for manufacturers to generate powertrain inputs for use in GEM. For Phase 2 we generally require this test procedure for evaluating hybrid powertrains for inputs into GEM, but there are certain exceptions where engine-only test procedures may be used to certify hybrids via GEM (e.g., pre-transmission hybrids).

A key advantage of the powertrain test approach is that it directly measures the effectiveness of the engine, the transmission, and the integration of these two components. Engines and transmissions are particularly challenging to simulate within a computer program like GEM because the engines and transmissions installed in vehicles today are actively and interactively controlled by their own sophisticated electronic controls; namely the ECM and TCM.

We believe that the capital investment impact on manufacturers for powertrain testing is reasonable; especially for those who already have heavy-duty engine dynamometer test facilities. We have found that, in general, medium-duty powertrains can be tested in heavy-duty engine test cells. EPA has successfully completed such a test facility conversion at the National Vehicle and Fuel Emissions Laboratory in Ann Arbor, Michigan. Southwest Research Institute (SwRI) in San Antonio, Texas has completed a similar test cell conversion. Oak Ridge National Laboratory in Oak Ridge, Tennessee has been operating a recently constructed heavy heavy-duty powertrain dynamometer facility, and EPA currently has an interagency agreement with DOE to fund EPA powertrain testing at ORNL. The results from this testing were published for a 30-day comment period, as part of the NODA.[135] Eaton Corporation has been operating a heavy-duty powertrain test cell and has provided the agencies with valuable test results and other comments.[136] PACCAR recently constructed and began operation of a powertrain test cell that includes engine, transmission and axle test capabilities.[137] EPA also contracted SwRI to evaluate North America's capabilities (as of 2014) for powertrain testing in the heavy-duty sector and the cost of installing a new powertrain cell that meets agency requirements.[138] Results from this 2014 survey indicated that one supplier (Eaton) already had this capability. We estimate that the upgrade costs to an existing engine test facility are on the order of $1.2 million, and a new test facility in an existing building are on the order of $1.9 million. We also estimate that current powertrain test cells that could be upgraded to measure CO2 emissions would cost approximately $600,000. For manufacturers or suppliers wishing to contract out such testing, SwRI estimated that a cost of $150,000 would provide about one month of powertrain testing services. Once a powertrain test cell is fully operational, we estimate that for a nominal powertrain family (i.e. one engine family tested with one transmission family), the cost for powertrain installation, testing, and data analysis would be about $70,000 in calendar year 2016, in 2016 dollars. Since the NPRM in July 2015, the agencies and other stakeholders have completed significant new work toward refining the powertrain test procedure itself, and these results confirm the robustness of this approach. The agencies regulations provide details of the final powertrain test procedure. See 40 CFR 1037.550.

Furthermore, the agencies have worked with key transmission suppliers to develop an approach to define transmission families. Coupled with the agencies' existing definitions of engine families (40 CFR 1036.230 and 1037.230), we are finalizing powertrain family definitions in 40 CFR 1037.231 and axle and transmission families in 40 CFR 1037.232.

Even though there is conclusive evidence that powertrain testing is a Start Printed Page 73534technically robust and cost-effective approach to evaluating the CO2 and fuel consumption performance of powertrains, and even though there has been a clear trend toward manufacturers and other test laboratories recognizing the benefits and investing in new powertrain testing facilities, the agencies also received significant negative comment regarding the sheer amount of powertrain testing that could be required to certify the large number of unique configurations (i.e., unique combinations of engines and transmissions). While the agencies proposed to allow manufacturers to group powertrains in powertrain families, as defined by the EPA in 40 CFR 1037.231, requiring powertrain testing broadly would still likely require a large number of tests. To address these concerns, while at the same time achieving most of the advantages of powertrain testing, the agencies are also finalizing some mandatory and optional test procedures to separately evaluate engine transient performance (via the mandatory “cycle-average” approach for the transient cycle) and transmission efficiency performance. While neither of these test procedures capture the optimized shift logic and other benefits of deep integration of the engine and transmission controllers, which only powertrain testing can capture, these separate test procedures do capture the remaining benefits of powertrain testing. The advantage of these separate tests is that their results can be mixed and matched within GEM to represent many more combinations of engines and transmissions than a comparable number of powertrain tests. For example, separately testing three parent engines that each have two child ratings and separately efficiency testing three transmissions that each have three major calibrations requires the equivalent test time of testing 6 powertrains, but without requiring the use of a powertrain test facility. More importantly, the results of these 6 tests can be combined within GEM to certify at least 27 different powertrain families, which would otherwise have required 27 powertrain tests—more than a four-fold increase in costs. This example clearly shows how cost-effective a vehicle simulation approach to vehicle certification can be.

Another regulatory structure option considered by the agencies was engine-only testing over the GEM duty cycles over a range of simulated vehicle configurations, which is part of Option 4 in Figure II.1. This is essentially a “cycle-average approach,” which would use GEM to generate engine duty cycles by simulating a range of transmissions and other vehicle variations. These engine-level duty cycles would then be programmed into a separate controller of a dynamometer connected to an engine's output shaft. The agencies requested comment on this approach, and based on continued research that has been conducted since the proposal, and based on comments we received in response to the NODA, we are finalizing this approach as mandatory for determining the GEM inputs that characterize an engine's transient engine performance within GEM over the ARB Transient duty cycle. We are also finalizing this approach as optional for characterizing the more steady-state engine operation in GEM over the 55 mph and 65 mph duty cycles with road grade, in lieu of steady-state engine mapping for these two cycles. We are also finalizing this approach as an option for certifying pre-transmission hybrids, in lieu of powertrain testing. We are calling this approach the “cycle-average” approach, which generates a cycle-average engine fuel map that is input into GEM. This map simulates an engine family's performance over a given vehicle drive cycle, for the full range of vehicles into which that engine could be installed. Unlike the chassis dynamometer or powertrain dynamometer approaches, which could have significant test facility construction or modification costs, this engine-only approach necessitates little capital investment because engine manufacturers already have engine test facilities to both develop engines and to certify engines to meet both EPA's non-GHG standards and the agencies' Phase 1 fuel efficiency and GHG separate engine standards. This option has received significant attention since our notice of proposed rulemaking. EPA and others have published peer reviewed journal articles demonstrating the efficacy of this approach,[139 140] and the agencies have received significant comments on both the information we presented in the proposal and in the NODA. Comments have been predominantly supportive, and the comments we received tended to focus on ideas for further minor refinements of this test procedure.[136 141 142 143 144 145] At this time the agencies believe that the wealth of experimental data supporting the robustness and cost-effectiveness of the cycle-average approach, supports the agencies' decision to finalize this test procedure as mandatory for the determination of the transient performance of engines for use in GEM (i.e., over the ARB Transient Cycle).

The agencies also considered simulating the engine, transmission, and vehicle using a computer program; while having the actual transmission electronic controller connected to the computer running the vehicle simulation program, which is part of Option 3 in Figure II.1. The output of the simulation would be an engine cycle that would be used to test the engine in an engine test facility. Just as in the cycle-average approach, this procedure would not require significant capital investment in new test facilities. An additional benefit of this approach would be that the actual transmission controller would be determining the transmission gear shift points during the test, without a transmission manufacturer having to reveal their proprietary transmission control logic. This approach comes with some significant technical challenges, however. The computer model would have to become more complex and tailored to each new transmission and controller to make sure that the controller would operate properly when it is connected to a computer instead of an actual transmission. Some examples of the transmission specific requirements would be simulating all the Controller Area Network (CAN) communication to and from the transmission controller and the specific sensor responses both through simulation and hardware. Each vehicle manufacturer would have to be Start Printed Page 73535responsible for connecting the transmission controller to the computer, which would require a detailed verification process to ensure it is operating properly while it is in fact disconnected from a real transmission. Determining full compliance with this test procedure would be a significant challenge for the regulatory agencies because the agencies would have to be able to replicate each of the manufacturer's unique interfaces between the transmission controller and computer running GEM. The agencies did not receive any significant comments on this approach, presumably because commenters focused on the more viable options of powertrain testing and the cycle-average engine mapping approach. And because of the significant challenges noted above, the agencies did not pursue this option further between the time of proposal and this final action. However, should this approach receive more research attention in the future, such that the concerns noted above are sufficiently addressed, the agencies could consider allowing this certification approach as an option, within the context of a separate future rulemaking.

Finally, the agencies considered full vehicle simulation plus separate engine standards (Option 2 in Figure II.1), which is the required approach being finalized for Phase 2. This approach is discussed in more detail in the following sections. It should be noted before concluding this subsection that the agencies do provide a regulatory path for manufacturers to apply for approval of alternative test methods that are different than those the agencies specify. See 40 CFR part 1065, subpart A. Therefore, even though we have not finalized some of the certification approaches and test procedures that we investigated, our conclusions about these procedures do not prevent a manufacturer from seeking agency approval of any of these procedures or any other alternative procedures.

(2) Final Phase 2 Regulatory Structure

Under the final Phase 2 structure, tractor and vocational chassis manufacturers will be required to provide engine, transmission, drive axle(s) and tire inputs into GEM (as well as the inputs already required under Phase 1). For Phase 1, GEM used fixed default values for all of these, which limited the types of technologies that could be recognized by GEM to show compliance with the standards. We are expanding GEM to account for a wider range of technological improvements that would otherwise need to be recognized through the more cumbersome off-cycle crediting approach in Phase 1. Additional technologies that will now be recognized in GEM also include lightweight thermoplastic materials, automatic tire inflation systems, tire pressure monitoring systems, advanced cruise control systems, electronic vehicle coasting controls, engine stop-start idle reduction systems, automatic engine shutdown systems, hybrids, and axle configurations that decrease the number of drive axles. The agencies are also continuing separate engine standards. As described below, we see advantages to having both engine-based and vehicle-based standards. Moreover, the advantages described here for full vehicle simulation do not necessarily correspond to disadvantages for engine testing or vice versa.

(a) Advantages of Vehicle Simulation

The agencies' primary purpose in developing fuel efficiency and GHG emissions standards is to increase the use of vehicle technologies that improve fuel efficiency and decrease GHG emissions. Under the Phase 1 tractor and vocational chassis standards, there is no regulatory incentive for vehicle manufacturers to consider adopting new engine, transmission or axle technologies because GEM was not configured to recognize these technologies uniquely, leaving off-cycle credits as the only regulatory mechanism to recognize these technologies' benefits. By recognizing such technologies in GEM under Phase 2, the agencies will be creating a direct regulatory incentive to improve engine, transmission, and axle technologies to improve fuel efficiency and decrease GHG emissions. In its 2014 report, NAS also recognized the benefits of full vehicle simulation and recommended that the Phase 2 rules incorporate such an approach.160

The new Phase 2 approach will create three new specific regulatory incentives. First, vehicle manufacturers will have an incentive to use the most efficient engines. Since GEM will no longer use the agency default engine in simulation, manufacturers will have their own engines recognized in GEM. Under Phase 1, engine manufacturers have a regulatory incentive to design efficient engines, but vehicle manufacturers do not have a similar regulatory incentive to use the most efficient engines in their vehicles. Second, the new Phase 2 approach will create incentives for both engine and vehicle manufacturers to design engines and vehicles to work together to ensure that engines actually operate as much as possible near their most efficient points. This is because Phase 2 GEM will require the vehicle manufacturers to input specific transmission, axle, and tire characteristics, thus recognizing powertrain optimization, such as engine down-speeding, and different transmission architectures and technologies, such as automated manual transmissions, automatic transmissions, and different numbers of transmission gears, transmission gear ratios, axle ratios and tire revolutions per mile. No matter how well designed, all engines have speed and load operation points with differing fuel efficiency and GHG emissions. The speed and load point with the best fuel efficiency (i.e., peak thermal efficiency) is commonly known as the engine's “sweet spot.” The more frequently an engine operates near its sweet spot, the better the vehicle's fuel efficiency will be. In Phase 1, a vehicle manufacturer receives no regulatory credit under GEM for designing its vehicle to operate closer to its engine's sweet spot because Phase 1 GEM does not model the specific engine, transmission, axle, or tire revolutions per mile of the vehicle. Third, this approach will recognize improvements to the overall efficiency of the drivetrain, including the axle. The new version of GEM will recognize the benefits of different integrated axle technologies including axle lubricants (via an optional axle efficiency test), and technologies that reduce axle losses such as by enabling three-axle vehicles to deliver power to only one rear axle. This is accomplished through the simulation of axle disconnect technology (see Chapter 4.5 of the RIA). The new version of GEM also will be able to recognize the benefits of reducing energy losses within a transmission, via an optional transmission efficiency test.

In addition to providing regulatory incentives to use more fuel efficient technologies, expanding GEM to recognize engine and other powertrain component improvements will provide important flexibility to vehicle manufacturers. Providing flexibility to effectively trade engine and other powertrain component improvements against the other vehicle improvements that are recognized in GEM will allow vehicle manufacturers to better optimize their vehicles to achieve the lowest cost for specific customers. Because of the improvements in GEM, GEM will recognize this deeper level of vehicle optimization. Vehicle manufacturers could use this flexibility to reduce overall compliance costs and/or address special applications where certain vehicle technologies are not preferred or Start Printed Page 73536practical. The agencies considered in Phase 1 allowing the exchange of emission certification credits generated relative to the separate brake-specific engine standards and credits generated relative to the vehicle standards. However, we did not allow this in Phase 1 due in part to concerns about the equivalency of credits generated relative to different standards, with different units of measure and different test procedures. The Phase 2 approach eliminates these concerns because engine and other vehicle component improvements will be evaluated relative to the same vehicle standard in GEM. This also means that under the Phase 2 approach there is no need to consider allowing emissions credit trading between engine-generated and vehicle-generated credits because vehicle manufacturers are directly credited by the combination of engine and vehicle technologies they choose to install in each vehicle. Therefore, this approach eliminates one of the concerns about continuing separate engine standards, which was that a separate engine standard and a full vehicle standard were somehow mutually exclusive. That is not the case. In fact, in the next section we describe how we are continuing the separate engine standard along with recognizing engine performance at the vehicle level. The agencies acknowledge that maintaining a separate engine standard will limit flexibility in cases where a vehicle manufacturer wanted to use less efficient engines and make up for them using more efficient vehicle technologies. However, as described below, we see important advantages to maintaining a separate engine standard, and we believe they more than justify the reduced flexibility. Furthermore, in response to comments about some specialized vocational custom chassis, the agencies are finalizing a limited number of optional standards that would be met using a somewhat simplified version of GEM. Specifically, in this simplified version of GEM, which is only applicable as an option for certain custom chassis applications, the GEM inputs for the engine, transmission gears, gear ratios, gear efficiency; axle ratio, axle efficiency; and tire revolutions per mile are all fixed to default values. This simplification allows the option of certifying these custom chassis without penalty for utilizing less efficient engines, transmissions, or axles. This flexibility also addresses a comment the agencies received from Cummins that the inclusion of the specific engine in GEM limits the flexibility provided by the separate engine standards' emissions averaging, banking and trading program. Cummins explained that certain applications like emergency vehicles, cement mixers and recreational vehicles oftentimes require higher-performance, less-efficient, engines, which are credit using engines under the ABT program of the separate engine standards. Because these particular vehicle applications have few other cost-effective and practical vehicle-level technologies with which to offset their use of less efficient engines, the main Phase 2 vocational chassis standards that require engine and other powertrain inputs into GEM (i.e., the standards for other than custom chassis vocational vehicles) could be particularly challenging for these applications. However, the optional custom chassis standards solves this issue for custom chassis applications. This approach solves two issues. First, it provides a means toward certification for these custom chassis applications, without penalty for using the engines they need. Second, this approach maintains the flexibility intended by the separate engine standards' averaging, banking and trading program since these custom chassis applications would still be using certified engines.

One disadvantage of recognizing engines and transmission in GEM is that it will increase complexity for the vehicle standards. For example, vehicle manufacturers will be required to conduct additional engine tests and to generate additional GEM inputs for compliance purposes. However, we believe that most of the burden associated with this increased complexity will be an infrequent burden of engine testing and updating information systems to track these inputs. Furthermore, the agencies are requiring that engine manufacturers certify their respective GEM inputs; namely, their own engine maps. Because there are a relatively small number of heavy-duty engine manufacturers who will be responsible for generating and complying with their declared engine maps for GEM, the overall engine testing burden to the heavy-duty vehicle industry is small. With this approach, the large number of vocational chassis manufacturers will not have to conduct any engine testing.

Another potential disadvantage to GEM-based vehicle certification is that because GEM measures performance over specific duty cycles intended to represent average operation of vehicles in-use, this approach might also create an incentive to optimize powertrains and drivetrains for the best GEM performance rather than the best in-use performance for a particular application. This is always a concern when selecting duty cycles for certification, and so is not an issue unique to GEM. There will always be instances, however infrequent, where specific vehicle applications will operate differently than the duty cycles used for certification. The question is would these differences force manufacturers to optimize vehicles to the certification duty cycles in a way that decreases fuel efficiency and increases GHG emissions in-use? We believe that the certification duty cycles will not create a disincentive for manufacturers to properly optimize vehicles for customer fuel efficiency. First, the impact of the certification duty cycles versus any other real-world cycle will be relatively small because they affect only a small fraction of all vehicle technologies. Second, the emission averaging and fleet average provisions mean that the regulations will not require all vehicles to meet the standards. Vehicles exceeding a standard over the duty cycles because they are optimized for different in-use operation can be offset by other vehicles that perform better over the certification duty cycles. Third, vehicle manufacturers also have the ability to lower such a vehicle's measured GHG emissions by adding technology that would improve fuel efficiency both over the certification duty cycles and in-use (and to be potentially eligible to generate off-cycle credits in doing so). These standards are not intended to be at a stringency where manufacturers will be expected to apply all technologies to all vehicles. Thus, there should be technologies available to add to vehicle configurations that initially fail to meet the Phase 2 standards. Fourth, we are further sub-categorizing the vocational vehicle segment compared to Phase 1, tripling the number of subcategories within this segment from three to nine. These nine subcategories will divide each of the three Phase 1 weight categories into three additional vehicle speed categories. Each of the three speed categories will have unique duty cycle weighting factors to recognize that different vocational chassis are configured for different vehicle speed applications. This further subdivision better recognizes technologies' performance under the conditions for which the vocational chassis was configured to operate. This also decreases the potential of the certification duty cycles to encourage manufacturers to configure vocational chassis differently than the optimum configuration for specific customers' applications. Similarly, for the tractor Start Printed Page 73537category we are finalizing a new “heavy-haul” category to recognize the greater payload and vehicle mass of these tractors, as well as their limitations to effectively utilize some technologies like aerodynamic technologies. These new categories help minimize differences between GEM simulation and real-world operation. Finally, we are also recognizing seven specific vocational vehicle applications under the optional custom chassis vocational vehicle standards.

Another disadvantage of our full vehicle simulation approach is the potential requirement for engine manufacturers to disclose information to vehicle manufacturers who install their engines that engine manufacturers might consider to be proprietary. Under this approach, vehicle manufacturers may need to know some additional details about engine performance long before production, both for compliance planning purposes, as well as for the actual submission of applications for certification. Moreover, vehicle manufacturers will need to know details about the engine's performance that are generally not publicly available—specifically the detailed steady-state fuel consumption map of an engine. Some commenters expressed significant concern about the Phase 2 program forcing the disclosure of proprietary steady-state engine performance information to business competitors; especially prior to an engine being introduced into commerce. It can be argued that a sufficiently detailed steady-state engine map, such as the one required for input into GEM, can reveal proprietary engine design elements such as intake air, turbo-charger, and exhaust system design; exhaust gas recirculation strategies; fuel injection strategies; and exhaust after-treatment thermal management strategies. Conversely, the agencies also received comments requesting that all GEM inputs be made public, as a matter of transparency and public interest.

It is unclear at this point whether such information is truly proprietary. In accordance with Federal statutes, EPA does not release information from certification applications (or other compliance reports) that we determine to be Confidential Business Information (CBI) under 40 CFR part 2. Consistent with section 114(c) of the CAA, EPA does not consider emission test results to be CBI after introduction into commerce of the certified engine or vehicle. However, we have generally treated test results as protected before a product's introduction into commerce date. EPA has not yet made a final CBI determination for Phase 1 or Phase 2 GEM inputs. Nevertheless, at this time we expect to continue our current policy of non-disclosure prior to introduction into commerce, but we consider it likely that we would ultimately not treat any test results or other GEM inputs as CBI after the introduction into commerce date, as identified by the manufacturer.

To further address the specific concern about the Phase 2 program forcing the disclosure of proprietary steady-state engine maps to business competitors, especially prior to an engine being introduced into commerce, the agencies are finalizing an option for engine manufacturers to certify only “cycle average” engine maps over the 55-mph and 65-mph GEM cycles and separately mandating the cycle average approach for use over the ARB Transient cycle. See Section II.B. above. The advantage to this approach is that each data point of a cycle average map represents the average emissions over an entire cycle. Therefore, the cycle average engine map approach does not reveal any potentially proprietary information about an engine's performance at a particular steady-state point of operation.

(b) Advantages of Separate Engine Standards

For engines installed in tractors and vocational vehicle chassis, we are maintaining separate engine standards for fuel consumption and GHG emissions in Phase 2 for both spark-ignition (SI, generally but not exclusively gasoline-fueled) and compression-ignition (CI, generally but not exclusively diesel-fueled) engines. Moreover, we are adopting a sequence of new more stringent engine standards for CI engines for engine model years 2021, 2024 and 2027. While the vehicle standards alone are intended to provide sufficient incentive for improvements in engine efficiency, we continue to see important advantages to maintaining separate engine standards for both SI and CI engines. The agencies believe the advantages described below are critical to fully achieve the goals of the EPA and NHTSA standards.

First, EPA has a robust compliance program based on separate engine testing. For the Phase 1 standards, we applied the existing criteria pollutant compliance program to ensure that engine efficiency in actual use reflected the improvements manufacturers claimed during certification. With engine-based standards, it is straightforward to hold engine manufacturers accountable by testing in-use engines in an engine dynamometer laboratory. If the engines exceed the standards, manufacturers can be required to correct the problem or perform other remedial actions. Without separate engine standards in Phase 2, addressing in-use compliance would be more subjective. Having clearly defined compliance responsibilities is important to both the agencies and to the manufacturers.

Second, engine standards for CO2 and fuel efficiency force engine manufacturers to optimize engines for both fuel efficiency and control of non-CO2 emissions at the same engine operating points. This is of special concern for NOX emissions, given the strong counter-dependency between engine-out NOX emissions and fuel consumption. By requiring engine manufacturers to comply with both NOX and CO2 standards using the same test procedures, the agencies ensure that manufacturers include technologies that can be optimized for both, rather than alternate, calibrations that would trade NOX emissions against fuel consumption, depending how the engine or vehicle is tested. In the past, when there was no CO2 engine standard and no steady-state NOX standard, some manufacturers chose this dual calibration approach instead of investing in technology that would allow them to simultaneously reduce both CO2 and NOX.

It is worth noting that these first two advantages foster fair competition within the marketplace. In this respect, the separate engine standards help assure manufacturers that their competitors are not taking advantage of regulatory ambiguity. The agencies believe that the absence of separate engine standards would leave open the opportunity for a manufacturer to choose a high-risk compliance strategy by gaming the NOX-CO2 tradeoff. Manufacturer concerns that competitors might take advantage of this can create a dilemma for those who wish to fully comply, but also perceive shareholder pressure to choose a high-risk compliance strategy to maintain market share.

Finally, the existence of meaningful separate engine standards allows the agencies to exempt certain vehicles from some or all of the vehicle standards and requirements without forgoing the engine improvements. A good example of this is the off-road vehicle exemption in 40 CFR 1037.631 and 49 CFR 535.3, which exempts vehicles “intended to be used extensively in off-road environments” from the vehicle requirements. The engines used in such vehicles must still meet the engine standards of 40 CFR 1036.108 and 49 CFR 535.5(d). The agencies see no Start Printed Page 73538reason why efficient engines cannot be used in such vehicles. However, without separate engine standards, there would be no way to require the engines to be efficient. The engine standards provide a similar benefit with respect to the custom chassis program discussed in Section V.

In the past there has been some confusion about the Phase 1 separate engine standards somehow preventing the recognition of engine-vehicle optimization that vehicle manufacturers perform to minimize a vehicle's overall fuel consumption. It was not the existence of separate engine standards that prevented recognition of this optimization. Rather it was that the agencies did not allow manufacturers to enter inputs into GEM that characterized unique engine performance. For Phase 2 we are requiring that manufacturers input such data because we intend for GEM to recognize this engine-vehicle optimization. The continuation of separate engine standards in Phase 2 does not undermine in any way the recognition of this optimization in GEM.

C. Phase 2 GEM and Vehicle Component Test Procedures 146

GEM was originally created for the certification of tractors and vocational vehicle chassis to the agencies' Phase 1 CO2 and fuel efficiency standards. See 76 FR 57116, 57146, and 57156-57157. For Phase 2 the agencies proposed a number of modifications to GEM, and based on public comments in response to the agencies' proposed modifications, the agencies have further refined these modifications for this final action.

In Phase 1 the agencies adopted a regulatory structure where regulated entities are required to use GEM to simulate and certify tractors and vocational vehicle chassis. This computer program is provided free of charge for unlimited use, and the program may be downloaded by anyone from EPA's Web site: http://www3.epa.gov/​otaq/​climate/​gem.htm. GEM mathematically combines the results of a number of performance tests of certain vehicle components, along with other pre-determined vehicle attributes and driving patterns to determine a vehicle's characteristic levels of fuel consumption and CO2 emissions, for certification purposes. For Phase 1, the required inputs to GEM for tractors include vehicle aerodynamics information, tire rolling resistance, and whether or not a vehicle is equipped with certain lightweight high-strength steel or aluminum components, a tamper-proof speed limiter, or tamper-proof idle reduction technologies. For Phase 1, the sole input for vocational vehicles is tire rolling resistance. For Phase 1, the computer program's inputs did not include engine test results or attributes related to a vehicle's powertrain; namely, its transmission, drive axle(s), or tire revolutions per mile. Instead, for Phase 1 the agencies specified generic engine and powertrain attributes within GEM. For Phase 1 these are fixed and cannot be changed in GEM.[147]

Similar to other vehicle simulation computer programs, GEM combines various vehicle inputs with known physical laws and justified assumptions to predict vehicle performance for a given period of vehicle operation. GEM represents this information numerically, and this information is integrated as a function of time to calculate CO2 emissions and fuel consumption. Some of the justified assumptions in GEM include average energy losses due to friction between moving parts of a vehicle's powertrain; the logical behavior of an average driver shifting from one transmission gear to the next; and speed limit assumptions such as 55 miles per hour for urban highway driving and 65 miles per hour for rural interstate highway driving. The sequence of the GEM vehicle simulation can be visualized by imagining a human driver initially sitting in a parked running tractor or vocational vehicle. The driver then proceeds to drive the vehicle over a prescribed route that includes three distinct patterns of driving: Stop-and-go city driving, urban highway driving, and rural interstate highway driving. The driver then exits the highway and brings the vehicle to a stop, with the engine still running at idle. This concludes the vehicle simulation sequence.

Over each of the three driving patterns or “duty cycles,” GEM simulates the driver's behavior of pressing the accelerator, coasting, or applying the brakes. GEM also simulates how the engine operates as the gears in the vehicle's transmission are shifted and how the vehicle's weight, aerodynamics, and tires resist the forward motion of the vehicle. GEM combines the driver behavior over the duty cycles with the various vehicle inputs and other assumptions to determine how much fuel must be consumed to move the vehicle forward at each point during the simulation. For Phase 2 the agencies added the effect of road grade. In GEM the effect of road grade on fuel consumption is simulated by increasing fuel consumption uphill, by the amount of fuel consumed by the engine to provide the power needed to raise the mass of the vehicle and its payload against the force of Earth's gravity—while at the same time maintaining the duty cycle's vehicle speed. Downhill road grades are simulated by decreasing the engine's fuel consumption, by the amount of power returned to the vehicle by it moving in the same direction as Earth's gravity. To maintain vehicle speed downhill, simulated brakes are sometimes applied, and the energy lost due to braking results in a certain amount of fuel consumption as well. For each of the three duty cycles, GEM totals the amount of fuel consumed and then divides that amount by the product of the miles travelled and tons of payload carried. The tons of payload carried are specified by the agencies for each vehicle type and weight class, and these cannot be changed in GEM.

In addition to determining fuel consumption over these duty cycles, for Phase 2, GEM calculates a vehicle's fuel consumption rate when it is stopped in traffic with the driver still operating the vehicle (i.e., “drive idle”) and when the vehicle is stopped and parked with the engine still running (i.e., “parked idle”). For each regulatory subcategory of tractor and vocational vehicle (e.g., sleeper cab tractor, day cab tractor, light heavy-duty urban vocational vehicle, heavy heavy-duty regional vocational vehicle, etc.), GEM applies the agencies' prescribed weighting factors to each of the three duty cycles and to each of the two idle fuel consumption rates to represent the fraction of city driving, urban highway driving, rural highway driving, drive idle, and parked idle that is typical of each subcategory. After combining the weighted results of all the cycles and idle fuel rates, GEM then outputs a single composite result for the vehicle, expressed as both fuel consumed in gallon per 1,000 ton-miles (for NHTSA standards) and an equivalent amount of CO2 emitted in grams per ton-mile (for EPA standards). These are the vehicle's GEM results that are used along with other information to demonstrate that a vehicle certificate holder (e.g., a vehicle manufacturer) complies with the applicable standards. This other information includes the annual sales volume of the vehicle family, plus information on emissions credits that may be generated or used as Start Printed Page 73539part of that vehicle family's certification.

For Phase 1 GEM's tractor inputs include vehicle aerodynamics information, tire rolling resistance, and whether or not a vehicle is equipped with lightweight materials, a tamper-proof speed limiter, or tamper-proof idle reduction technologies. Other vehicle and engine characteristics in GEM were fixed as defaults that cannot be altered by the user. These defaults included tabulated data of engine fuel rate as a function of engine speed and torque (i.e., “engine fuel maps”), transmissions, axle ratios, and vehicle payloads. For tractors, Phase 1 GEM simulates a tractor pulling a standard trailer. For vocational vehicles, Phase 1 GEM includes a fixed aerodynamic drag coefficient and vehicle frontal area.

For Phase 2 new inputs are required and other new inputs are allowed as options. These include the outputs of new test procedures to “map” an engine to generate steady-state and transient, cycle-average, engine fuel rate inputs to represent the actual engine in a vehicle. As described in detail in RIA Chapter 4, certification to the Phase 2 standards will require entering new inputs into GEM to describe the vehicle's transmission type and its number of gears and gear ratios. Manufacturers must also enter attributes that describe the vehicle's drive axle(s) type, axle ratio and tire revolutions per mile. We are also finalizing a number of options to conduct additional component testing for the purpose of replacing some of the agencies' “default values” in GEM with inputs that are based on component testing. These include optional axle and transmission power loss test procedures. We are also finalizing an optional powertrain test procedure that would replace both the required engine mapping and the agencies' default values for a transmission and its automated shift strategy. We are also finalizing an option to generate cycle-average maps for the 55 mph and 65 mph cycles in GEM. In addition, we have made a number of improvements to the aerodynamic coast-down test procedures and associated aerodynamic data analysis techniques. While these aerodynamic test and data analysis improvements are primarily intended for tractors, for Phase 2 we are providing a streamlined off-cycle credit pathway for vocational vehicle aerodynamic performance to be recognized in GEM.

As proposed, we are finalizing a significantly expanded number of technologies that are recognized in GEM. These include recognizing lightweight thermoplastic materials, automatic tire inflation systems, advanced cruise control systems, workday idle reduction systems, and axle configurations that decrease the number of drive axles. In response to comments and data submitted to the agencies on the Phase 2 proposal we are also finalizing inputs related to tire pressure monitoring systems and advanced electronically controlled vehicle coast systems.

Although GEM is similar in concept to a number of other commercially available vehicle simulation computer programs, the applicability of GEM is unique. First, GEM was designed exclusively for manufacturers and regulated entities to certify tractor and vocational vehicle chassis to the agencies' fuel consumption and CO2 emissions standards. For GEM to be effective for this purpose, the inputs to GEM include only information related to certain vehicle components and attributes that significantly impact vehicle fuel efficiency and CO2 emissions. For example, these include vehicle aerodynamics, tire rolling resistance, and powertrain component information. On the other hand, other attributes such as those related to a vehicle's suspension, frame strength, or interior features are not included, where these otherwise might be included in other commercially available vehicle simulation programs that are used for other purposes. Furthermore, the simulated payload, driver behavior and duty cycles in GEM cannot be changed. Keeping these values constant helps to ensure that all vehicles are simulated and certified in the same way. However, these fixed attributes in GEM largely preclude GEM from being of much use as a research tool for exploring the effects of payload, driver behavior and different duty cycles.

Similar to Phase 1, GEM for Phase 2 is available free of charge for unlimited use, and the GEM source code is open source. That is, the programming source code of GEM is freely available upon request for anyone to examine, manipulate, and generally use without restriction. In contrast, commercially available vehicle simulation programs are generally not free and open source. Additional details of GEM are included in Chapter 4 of the RIA.

GEM is a computer software program, and like all other software development processes the agencies periodically released a number of developmental versions of the GEM software for others to review and test during the Phase 2 rulemaking process. This type of user testing significantly helps the agencies detect and fix any problems or “bugs” in the GEM software.

As part of Phase 1, the agencies conducted a peer review of GEM version 1.0, which was the version released for the Phase 1 proposal.[148 149] In response to this peer review and to comments from stakeholders, EPA made changes to the version of GEM released with the Phase 1 final rule. Updates to the Phase 1 GEM were also made via Technical Amendments.[150] The current version of Phase 1 GEM is v2.0.1, which is the version applicable for the Phase 1 standards.[150] As part of the development of GEM for Phase 2, both a formal peer review [149] and a series of expert reviews were conducted.[151 152 153 154]

The agencies have provided numerous opportunities for comment on GEM, and its iterative development. Shortly after the Phase 2 proposal's publication in July 2015 (and before the end of the public comment period), the agencies received comments on GEM. Based on these early comments, the agencies made minor revisions to fix a few bugs in GEM and in August 2015 released an updated version of GEM to the public for additional comment, which also included new information on GEM road grade profiles. The agencies also extended the public comment period on the proposal, which provided at least 30 days for public comment on this slightly updated version of GEM.[153] Then, in response to comments submitted at the close of the comment period, in early January 2016 Start Printed Page 73540the agencies released a “debugging” version of GEM to a wide range of expert reviewers.[152] The agencies provided one month for expert reviewers to provide informal feedback for debugging purposes.[152] Because the changes for this debugging version mostly added new features to make GEM easier to use for certifying via optional test procedures, like the powertrain test, there were only minor changes to the way that GEM performed. In the March 2016 NODA, the agencies included another developmental version of GEM [153] for public comment and provided 30 days for public comment. Based on the NREL report, which was also released as part of the NODA for public comment, the NODA version of GEM contained updated weighting factors of the duty cycles and idle cycles.[155] Therefore, the outputs of GEM for a given vehicle configuration changed because these duty cycle weighting factors changed, but there were only minor updates to how the individual technologies were simulated in GEM. Based on comments received on the NODA, the agencies made minor changes to GEM and released another debugging version in May 2016 to manufacturers, NGOs, suppliers, and CARB staff.[154] The most significant change to GEM for the May 2016 version was that 0.5 miles of flat road was added to the beginning and end of the 55 mph and 65 mph drive cycles in response to concerns raised by manufacturers.[156] This change did not change the way that GEM worked, but it did change GEM results because of the change in the duty cycles. This change was made to better align GEM simulation with real-world engine operation. The agencies provided the expert reviewers with at least a 3-week period in which to review GEM and provide feedback. Details on the history of the comments the agencies received and the history of the agencies responses leading to these multiple releases of GEM can be found in Section II.C.(1). The following list summarizes the changes in GEM in response to those comments and data submitted to the agencies in response to the Phase 2 proposal, NODA and other GEM releases:

  • Revised road grade profiles for 55- and 65-mph cruise cycles, only minor changes since August 2015.
  • Revised idle cycles for vocational vehicles with new vocational cycle weightings, weightings released for public comment in NODA.
  • Made changes to the input file structures. Examples includes additions of columns for axle configuration (“6×2,” “6×4,” “6×4D,” “4×2”), and additions of a few more technology improvement inputs, such as “Neutral Idle,” “Start/Stop,” and “Automatic Engine Shutdown.” These were minor changes, all were in NODA version of GEM.
  • Made changes to the output file structures. Examples include an option to allow the user to select an output of detailed results on average speed, average work at the input and output of the transmission, and the numbers of shifts for each cycle (e.g., 55 mph cycle, 65 mph cycle and the ARB Transient cycle). These were minor changes, all were in NODA version of GEM.
  • Added an input file for optional axle power losses (function of axle output speed and torque) and replaced a single axle efficiency value with lookup table of power loss. These were minor changes to streamline the use of GEM, all were in NODA version of GEM.
  • Modified engine torque response to be more realistic, with a fast response region scaled by engine displacement, and a slower torque response in the turbo-charger's highly boosted region. These were minor changes, all were in NODA version of GEM.
  • Added least-squares regression models to interpret cycle-average fuel maps for all cycles. These were minor changes to streamline the use of GEM, all were in NODA version of GEM.
  • Added different fuel properties according to 40 CFR 1036.530. This was a fix to align GEM with regulations.
  • Improved shift strategy based on testing data and comments received. These were minor changes, all were in NODA version of GEM.
  • Added scaling factors for transmission loss and inertia, per regulatory subcategory. These were minor changes, all were in NODA version of GEM.
  • Added optional input table for transmission power loss data. These were minor changes to streamline the use of GEM, all were in NODA version of GEM.
  • Added minimum torque converter lock-up gear user input for automatic transmissions. This was a minor change to streamline the use of GEM, this change was in the NODA version of GEM.
  • Revised the default transmission power loss tables, based on test data. This was a minor change to streamline the use of GEM, this change was in the NODA version of GEM.
  • Added neutral idle and start/stop effects idle portions of the ARB Transient cycle. These were minor changes, all were in NODA version of GEM
  • Adjusted shift and torque converter lockup strategy. This was a minor change to streamline the use of GEM, this change was in the NODA version of GEM.

Notwithstanding these numerous opportunities for public comment (as well as many informal opportunities via individual meetings), some commenters maintained that they still had not received sufficient notice to provide informed comment because each proposal represented too much of a “moving target.” [157 158 159] The agencies disagree. Even at proposal, Phase 2 GEM provided nearly all of the essential features of the version we are promulgating in final form. These include: (1) The reconfiguration of the engine, transmission, and axle sub-models to reflect additional designs and to receive manufacturer inputs; and (2) the addition of road grade and idle cycles for vocational vehicles, along with revised weighting factors. Moreover, the changes the agencies have made to GEM in response to public comment indicates that those comments were highly informed by the proposal. The agencies thus do not accept the contention that commenters were not afforded sufficient information to provide meaningful comment on GEM.

(1) Description of Modifications to GEM From Phase 1 to Phase 2

As explained above, GEM is a computer program that was originally developed by EPA specifically for manufacturers to use to certify to the Phase 1 tractor and vocational chassis standards. GEM mathematically combines the results of vehicle component test procedures with other vehicle attributes to determine a vehicle's certified levels of fuel consumption and CO2 emissions. Again as explained above, for Phase 1 the required inputs to GEM include vehicle aerodynamics information, tire rolling resistance, and whether or not a vehicle is equipped with certain lightweight Start Printed Page 73541high-strength steel or aluminum components, a tamper-proof speed limiter, or tamper-proof idle reduction technologies for tractors. The vocational vehicle inputs to GEM for Phase 1 only included tire rolling resistance. For Phase 1 GEM's inputs did not include engine test results or attributes related to a vehicle's powertrain; namely, its transmission, drive axle(s), or loaded tire radius. Instead, for Phase 1 the agencies specified a generic engine and powertrain within GEM, and for Phase 1 these cannot be changed in GEM.

For this rulemaking, GEM has been modified as proposed and validated against a set of experimental data that represent over 130 unique vehicle variants conducted at powertrain and chassis dynamometers with the manufacturers' provided transmission shifting tables. In addition, GEM has been validated against different types of tests when the EPA transmission default auto-shift strategy is used, which includes powertrain dynamometer tests and two truck tests running in a real-world driving route. Detailed comparisons can be seen in Chapter 4 of the RIA. As noted above, the agencies believe that this new version of GEM is an accurate and cost-effective alternative to measuring fuel consumption and CO2 over a chassis dynamometer test procedure. Again as noted earlier, some of the key modifications will require additional vehicle component test procedures (both mandatory and optional) to generate additional GEM inputs. The results of which will provide additional inputs into GEM. These include a new required engine test procedure to provide engine fuel consumption inputs into GEM. We proposed to measure fuel consumption as a matrix of steady-state points, but also sought comment on a newly developed engine test procedure that captures transient engine performance for use in GEM. We are specifying a combination of these procedures for the final rule—steady-state fuel maps for the highway cruise simulations, and cycle-average maps for transient simulations. As an option, cycle average maps could be also used for the highway cruise simulation as well. See Chapter 3 of the RIA for additional discussion of the fuel mapping procedures. We are also requiring inputs that describe the vehicle's transmission type, and its number of gears and gear ratios. We are allowing an optional powertrain test procedure that would provide inputs to override the agencies' simulated engine and transmission in GEM. In addition, in response to comments, we will also allow manufacturers to measure transmission efficiency in the form of the power loss tables to replace the default values in GEM. We are finalizing the proposed requirement to input a description of the vehicle's drive axle(s), including its type (e.g., 6×4 or 6×2) and axle ratio. We are also finalizing the optional axle efficiency test procedure for which we sought comment. This would allow manufacturers to override the agencies' simulated axle in GEM. Chapter 4 of the RIA details all of these GEM related input changes.

As noted above, we are significantly expanding the number of technologies that are recognized in GEM. These include recognizing lightweight thermoplastic materials, automatic tire inflation systems, advanced cruise control systems, engine stop-start idle reduction systems, and axle configurations that decrease the number of drive axles. To better reflect real-world operation, we are also revising the vehicle simulation computer program's urban and rural highway duty cycles to include changes in road grade, and including a new duty cycle to capture the performance of technologies that reduce the amount of time a vehicle's engine is at idle during a workday. Finally, to better recognize that vocational vehicle powertrains are configured for particular applications, we are further subdividing the vocational chassis category into three different vehicle speed categories, where GEM weights the individual duty cycles' results of each of the speed categories differently. Section 4.2 of the RIA details all these modifications. The following sub-sections provide further details on some of these key modifications to GEM.

(a) Simulating Engines for Vehicle Certification

Before describing the Phase 2 approach, this section first reviews how engines are simulated for vehicle certification in Phase 1. As noted earlier, GEM for Phase 1 simulates the same generic engine for any vehicle in a given regulatory subcategory with a data table of steady-state engine fuel consumption mass rates (g/s) versus a series of steady-state engine output shaft speeds (revolutions per minute, rpm) and loads (torque, N·m). This data table is also sometimes called a “fuel map” or an “engine map,” although the term “engine map” can mean other kinds of data in different contexts. The engine speeds in this map range from idle to maximum governed speed and the loads range from engine motoring (negative load) to the maximum load of an engine. When GEM executes a simulation over a vehicle duty cycle, this data table is linearly interpolated to find a corresponding fuel consumption mass rate at each engine speed and load that is demanded by the simulated vehicle operating over the duty cycle. The fuel consumption mass rate of the engine is then integrated over each duty cycle in GEM to arrive at the total mass of fuel consumed for the specific vehicle and duty cycle. Under Phase 1, manufacturers were not allowed to input their own engine fuel maps to represent their specific engines in the vehicle being simulated in GEM. Because GEM was programmed with fixed engine fuel maps for Phase 1 that all manufacturers had to use, the tables themselves did not have to exactly represent how an actual engine might operate over these three different duty cycles.

In contrast, for Phase 2 we are requiring manufacturers to generate their own engine fuel maps to represent each of their engine families in GEM. This Phase 2 approach is consistent with the 2014 NAS Phase 2 First Report recommendation.[160] To investigate this approach, before proposal we examined the results from 28 individual engine dynamometer tests. Three different engines were used to generate this data, and these engines were produced by two different engine manufacturers. One engine was tested at three different power ratings (13 liters at 410, 450 & 475 bhp) and one engine was tested at two ratings (6.7 liters at 240 and 300 bhp), and other engine with one rating (15 liters 455 bhp) service classes. For each engine and rating the steady-state engine dynamometer test procedure was conducted to generate an engine fuel map to represent that particular engine in GEM. Next, with GEM, we simulated various vehicles in which the engine could be installed. For each of the GEM duty cycles we are using, namely the urban local (ARB Transient), urban highway with road grade (55 mph), and rural highway with road grade (65 mph) duty cycles, we determined the GEM result for each vehicle configuration, and we saved the engine output shaft speed and torque information that GEM created to interpolate the steady-state engine map for each vehicle configuration We then had this same engine output shaft speed and torque information programmed into an engine dynamometer controller, and we had each engine perform the same duty cycles that GEM demanded of the Start Printed Page 73542simulated version of the engine. We then compared the GEM results based on GEM's linear interpolation of the engine maps to the measured engine dynamometer results. We concluded that for the 55 mph and 65 mph duty cycles, GEM's interpolation of the steady-state data tables was sufficiently accurate versus the measured results. This is an outcome one would reasonably expect because even with changes in road grade, the 55 mph and 65 mph duty cycles do not demand rapid changes in engine speed or load. The 55 mph and 65 mph duty cycles are nearly steady-state, as far as engine operation is concerned, just like the engine maps themselves. However, for the ARB Transient cycle, we observed a consistent bias when using the steady-state maps, where GEM consistently under-predicted fuel consumption and CO2 emissions. This low bias over the 28 engine tests ranged from 4.2 percent low to 7.8 percent low. The mean was 5.9 percent low and the 90th percentile value was 7.1 percent low. These observations are consistent with the fact that engines generally operate less efficiently under transient conditions than under steady-state conditions.

A number of reasons explain this consistent trend. For example, under rapidly changing (i.e. transient) engine conditions, it is generally more challenging to program an engine electronic controller to respond with optimum fuel injection rate and timing, exhaust gas recirculation valve position, variable nozzle turbocharger vane position and other set points than under steady-state conditions. Transient heat and mass transfer within the intake, exhaust, and combustion chambers also tend to increase turbulence and enhance energy loss to engine coolant during transient operation. In many cases during cold transient operation, the thermal management is triggered in order to maintain optimal performance of selective catalytic reduction devices for a diesel engine. Furthermore, because exhaust emissions control is more challenging under transient engine operation, engineering tradeoffs sometimes need to be made between fuel efficiency and transient criteria pollutant emissions control. Special calibrations are typically also required to control smoke and manage exhaust temperatures during transient operation for a transient cycle.

To account for these effects in GEM, the agencies have developed and are finalizing a test procedure called “cycle average” mapping to account for this transient behavior (40 CFR 1036.540). Detailed analyses and presentation of the test procedure was published in two peer-reviewed journal articles.[139,140] A number of commenters likewise suggested this approach. Additionally, progress has been made on further improving this test procedure since publication, based on a large number of engine dynamometer tests conducted by a variety of laboratory test facilities.[161] Since the proposal, further refinement of the numerical schemes used for interpreting cycle average engine fuel map was also completed. The engine dynamometer tests include a Cummins medium duty ISB engine, a Navistar heavy duty N13 engine, a Volvo heavy duty D13 engine, and a Cummins heavy duty ISX engine. All testing results indicated that the new test procedure works well for the transient ARB cycle.[162] In addition, Cummins in their NODA comments (see the following paragraph) provided additional data supporting this approach with their ISL 450 bhp rating engine. This data corroborated earlier data showing good agreement between engine dynamometer tests and the cycle average engine mapping approach.[163]

EPA solicited comment on the cycle average approach at proposal. 80 FR 40193. EPA also specifically provided notice and a 30-day opportunity for public comment on the possibility of requiring use of the cycle average mapping approach for the ARB Transient cycle. This was included in the version of GEM that was made available for public comment as part of the NODA [153] . In response, many comments were received on the cycle average approach. These include comments from Cummins [163] and Volvo.[164] Cummins was very supportive of the cycle average approach and also supported applying this approach to the 55 mph and 65 mph cruise cycles in GEM. Volvo expressed some concern over having enough time to fully evaluate this approach. The agencies believe that one of the reasons that Volvo expressed concern over having enough time to evaluate this approach is because Volvo initially declined working with the agencies to collaboratively refine this approach. At the same time, a number of Volvo's competitors chose to actively coordinate laboratory testing and technical analysis to contribute to the development of this approach. We believe these other manufacturers gained a deeper understanding of the approach earlier than Volvo because they invested time and resources to make technical contributions at earlier point in time. Nevertheless, the agencies fully welcome and appreciate Volvo's more recent active involvement in reviewing the cycle average approach and for making a number of productive suggestions for further refinement.

While the agencies are finalizing the cycle average engine mapping test procedure as mandatory for the ARB Transient cycle, for the 55 mph and 65 mph GEM drive cycles, the agencies are finalizing the same steady-state mapping procedure that the agencies originally proposed. The only difference is that we are finalizing about 85 unique steady-state map points, versus the about 143 points that were proposed. See 40 CFR 1036.535 for details. We are adopting a lower number of points because many of the originally proposed points were specified for use with the ARB Transient cycle.139 Again, as an option, the cycle average mapping test procedure also may be used for these two cruise speed cycles, in lieu of the steady-state mapping procedure.

(b) Simulating Human Driver Behavior and Transmissions for Vehicle Certification

GEM for Phase 1 simulates the same generic human driver behavior and manual transmission shifting patterns for all vehicles. The simulated driver responds to changes in the target vehicle speed of the duty cycles by changing the simulated positions of the vehicle's accelerator pedal, brake pedal, clutch pedal, and gear shift lever. For simplicity, in Phase 1 the GEM driver shifted at pre-specified vehicle speeds and the manual transmission was simulated as an ideal transmission that did not have any delay time (i.e., torque interruption) between gear shifts and did not have any energy losses associated with clutch slip during gear shifts.

In GEM for Phase 2 we are allowing manufacturers to select one of four types of transmissions to represent the transmission in the vehicle they are certifying: Manual transmission (MT), automated manual transmission (AMT), automatic transmission (AT) and dual clutch transmission (DCT). For Phase 2 the agencies proposed unique transmission shifting patters to Start Printed Page 73543represent the different types of automated transmissions. These shifting patterns over the steady state cruise cycles has been further modified from the proposed version to be more realistic with respect to slight variations in vehicle speed due to road grade. In particular, when going downhill, the simulated vehicle is now allowed to exceed the speed target by 3 mph before the brakes are applied. In the proposed version, the driver model applied the brakes much sooner to prevent the vehicle from exceeding the speed target. This change allows the vehicle to carry additional momentum into the next hill, much the same as real drivers would.

In the final version of GEM, the driver behavior and the different transmission types are simulated in the same basic manner as in Phase 1, but each transmission type features unique transmission responses that match the transmission responses we measured during vehicle testing of these three transmission types. In general the transmission gear shifting strategy for all of the transmissions is designed to shift the transmission so that it is in the most efficient gear for the current vehicle demand, while staying within certain limits to prevent unrealistically high frequency shifting (i.e., to prevent “short-shifting”). Some examples of these limits are torque reserve limits (which vary as function of engine speed), minimum time-in-gear and minimum fuel efficiency benefit to shift to the next gear. Some of the differences between the transmission types include a driver “double-clutching” during gear shifts of the manual transmission only, and “power shifts” and torque converter torque multiplication, slip, and lock-up in automatic transmissions only. Refer to Chapter 4 of the RIA for a more detailed description of these different simulated driver behaviors and transmission types.

Prior to the proposal, we considered an alternative approach where transmission manufacturers would provide vehicle manufacturers with detailed information about their automated transmissions' proprietary shift strategies for representation in GEM. NAS also recommended this approach.[165] The advantages of this approach would include a more realistic representation of a transmission in GEM and potentially the recognition of additional fuel efficiency improving strategies to achieve additional fuel consumption and CO2 emissions reductions. However, there are a number of technical and compliance disadvantages of this approach. One disadvantage is that it would require the disclosure of proprietary information because some vehicle manufacturers produce their own transmissions and also use other suppliers' transmissions. There are technical challenges too. For example, some transmission manufacturers have upwards of 40 different shift strategies programmed into their transmission controllers. Depending on in-use driving conditions, some of which are not simulated in GEM (e.g., changing payloads, changing tire traction) a transmission controller can change its shift strategy. Representing dynamic switching between multiple proprietary shift strategies would be extremely complex to simulate in GEM. Furthermore, if the agencies were to require transmission manufacturers to provide shift strategy inputs for use in GEM, then the agencies would have to devise a compliance strategy to monitor in-use shift strategies, including a driver behavior model that could be implemented as part of an in-use shift strategy confirmatory test. This too would be very complex. If manufacturers were subject to in-use compliance requirements of their transmission shift strategies, this could lead to restricting the use of certain shift strategies in the heavy-duty sector, which would in turn potentially lead to sub-optimal vehicle configurations that do not improve fuel efficiency or adequately serve the wide range of customer needs; especially in the vocational vehicle segment. For example, if the agencies were to restrict the use of more aggressive and less fuel efficient in-use shift strategies that are used only under heavy loads and steep grades, then certain vehicle applications would need to compensate for this loss of capability through the installation of over-sized and over-powered engines that are subsequently poorly matched and less efficient under lighter load conditions. Therefore, as a policy consideration to preserve vehicle configuration choice and to preserve the full capability of heavy-duty vehicles today, the agencies are intentionally not allowing transmission manufacturers to submit detailed proprietary shift strategy information to vehicle manufacturers to input into GEM. The agencies are finalizing as proposed that vehicle manufacturers can choose from among several transmission types that the agencies have already developed, validated, and programmed into GEM. The vehicle manufacturers will then enter into GEM their particular transmission's number of gears and gear ratios, optionally together with power loss tables representing their transmission's gear friction, pumping and spin losses. If a manufacturer chooses to use the optional powertrain test procedure, however, then the agencies' transmission types in GEM would be overridden by the actual data collected during the powertrain test, which would recognize the transmission's unique shift strategy. (Presumably, vehicle manufacturers will choose to use the optional powertrain test procedure only if their actual transmission shift strategy is more efficient compared to its respective default shift strategy simulated by GEM.)

(c) Simulating Axles for Vehicle Certification

In GEM for Phase 1 the axle ratio of the primary drive axle and the energy losses assumed in the simulated axle itself were the same for all vehicles. For Phase 2 the vehicle manufacturer will be required to input into GEM the axle ratio of the primary drive axle. This input will recognize the design to operate the engine at a particular engine speed when the transmission is operating in its highest transmission gear; especially for the 55 mph and 65 mph duty cycles in GEM. This input facilitates GEM's recognition of vehicle designs that take advantage of operating the engine at the lowest possible engine speeds. This is commonly known as “engine down-speeding,” and the general rule-of-thumb for heavy-duty engines is that for every 100 rpm decrease in engine speed, there can be about a 1 percent decrease in fuel consumption and CO2 emissions. Therefore, it is important that GEM allow this value to be input by the vehicle manufacturer. Axle ratio is also straightforward to verify during any in-use compliance audit. UCS and ACEEE commented that engine down-speeding should be recognized in the agencies' separate engine standards, rather than in the vehicle standard. The agencies disagree with this because recognizing down-speeding at the vehicle level ensures that the powertrain configuration in-use, in the real world, will lead to the engine operating at lower speeds. In contrast, the engine speeds specified in the separate engine standards' test procedures are based on the engine's maximum torque versus speed curve (i.e., lug curve) and not on the configuration of the powertrain to Start Printed Page 73544which the engine is attached in a vehicle. This means that even if a manufacturer manipulated the engine's lug curve such that the separate engine standards' test procedure led to the engine operating at lower speeds during certification, that same engine could be installed in a vehicle with a powertrain configured for the engine to operate at higher engine speeds. Therefore, recognizing down-speeding within GEM, at the vehicle level, best ensures that the agencies' test procedures and standards lead to real-world engine down-speeding in-use.

We proposed to use a fixed axle ratio energy efficiency of 95.5 percent at all speeds and loads, but requested comment on whether this pre-specified efficiency is reasonable. 80 FR 40185. In general, commenters stated that the efficiency of the axle actually varies as a function of axle ratio, axle speed, and axle input torque. Therefore, we have modified GEM to accept an input data table of power loss as a function of axle speed and axle torque. The modified version of GEM subsequently interpolates this table over each of the duty cycles to represent a more realistic axle efficiency at each point of each duty cycle. The agencies specify a default axle efficiency table in GEM for any manufacturer to use. We are also finalizing an optional axle power loss test procedure that requires the use of a dynamometer test facility (40 CFR 1037.560). With this optional test procedure, a manufacturer can create an axle efficiency table for use in lieu of the EPA default table. We requested comment on this test procedure in the proposal, and we received supportive comments. Refer to 40 CFR 1037.560 of the Phase 2 regulations, which contain this test procedure.

Moreover, the final regulations allow the manufacturers to develop analytical methods to derive axle efficiency tables for untested axle configurations, based on testing of similar axles. This would be similar to the analytically derived CO2 emission calculations allowed for pickups and vans. However, manufacturers would be required to obtain prior approval from the agencies before using analytically derived values. In addition, the agencies could conduct confirmatory testing or require a selective enforcement audit for any axle configuration. See 40 CFR 1037.235.

In addition to requiring the primary drive axle ratio input into GEM (and an option to input an actual axle power loss data table), we are requiring that the vehicle manufacturer input into GEM whether one or two drive axles are driven by the engine. When a heavy-duty vehicle is equipped with two rear axles where both are driven by the engine, this is called a “6×4” configuration. “6” refers to the total number of wheel hubs on the vehicle. In the 6×4 configuration there are two front wheel hubs for the two steer wheels and tires plus four rear wheel hubs for the four rear wheels and tires (or more commonly four sets of rear dual wheels and tires). “4” refers to the number of wheel hubs driven by the engine. These are the two rear axles that have two wheel hubs each. Compared to a 6×4 configuration, a 6×2 configuration decreases axle energy loss due to friction and oil churning in two driven axles, by driving only one axle. The decrease in fuel consumption and CO2 emissions associated with a 6×2 versus 6×4 axle configuration can be in the range of 2.5 percent depending on specific axles, which is modeled by the power loss table.[166] Therefore, in the Phase 2 version of GEM, if a manufacturer simulates a 6×2 axle configuration using the default axle efficiencies, GEM decreases the overall GEM result roughly by 2.5 percent on average through the power loss table. Note that GEM will similarly decrease the overall GEM result by 2.5 percent for a 4×2 tractor or Class 8 vocational chassis configuration if it has only two wheel hubs driven. If a manufacturer does not use the default efficiencies, the benefit of 6×2 and 4×2 configurations will be reflected directly in its input tables. Note that the Phase 2 version of GEM does not have an option to simulate more than two drive axles or configurations where the front axle(s) are driven or where there are more than two rear axles. The regulations specify that such vehicles are to be simulated as 6×4 vehicles in GEM. This is consistent with how the standards were developed and the agencies believe this approach will provide the appropriate incentive for manufacturers to apply the same fuel saving technologies to these vehicles, as they would to their conventional 6×4 vehicles. Moreover, because these configurations are manufactured for specialized vehicles that require extra traction for off-road applications, they have very low sales volume and any increased fuel consumption and CO2 emissions from them are not significant in comparison to the overall reductions of the Phase 2 program. Note that 40 CFR 1037.631 (for off-road vocational vehicles), which is being continued from the Phase 1 program, exempts many of these vehicles from the vehicle standards because they are limited mechanically to low-speed operation.

(d) Simulating Accessories for Vehicle Certification

The agencies proposed to continue the approach from Phase 1 whereby GEM uses a fixed power consumption value to simulate the fuel consumed for powering accessories such as steering pumps and alternators. 80 FR 40186. The final rule continues the Phase 1 approach, as proposed. However, Phase 2 GEM provides an option to provide a GEM input reflecting technology improvement inputs for the accessory loads. This allows the manufacturers to receive credit for those technologies that are not modeled in GEM. Manufacturers seeking credit for those technologies that are not modeled in GEM would generally follow the off-cycle credit program procedures in 40 CFR 1037.610.

(e) Aerodynamics in GEM for Tractor, Vocational Vehicle, and Trailer Certification

Phase 2 GEM simulates aerodynamic drag in using Cd A (the product of the drag coefficient and frontal area of the vehicle) rather than a drag coefficient (Cd). For tractors and trailers we will continue to use an aerodynamic bin approach similar to the one that exists in Phase 1 today, although the actual Phase 2 bins are being revised to reflect new test procedures and our projections for more aerodynamic tractors and trailers in the future. This approach allows manufacturers to determine Cd A (or delta-Cd A in the case of trailers) from coastdown testing, scale wind tunnel testing and/or computational fluid dynamics modeling. It requires tractor manufacturers (but not trailer manufacturers) to conduct a certain minimum amount of coast-down vehicle testing to validate their methods. The regulations also provide an alternate path for trailer manufacturers to rely on testing performed by component suppliers. See 40 CFR 1037.

The results of these tests determine into which bin a tractor or trailer is assigned. GEM uses the aerodynamic drag coefficient applicable to the bin, which is the same for all tractors (or trailers) within a given bin. This approach helps to account for limits in the repeatability of aerodynamic testing and it creates a compliance margin since any test result which keeps the vehicle in the same aerodynamic bin is considered compliant. For Phase 2 we are establishing new boundary values for the bins themselves and we are adding two additional tractor bins in order to recognize further advances in Start Printed Page 73545aerodynamic drag reduction beyond what was recognized in Phase 1. Furthermore, while Phase 1 GEM used predefined frontal areas for tractors where the manufacturers input only a Cd value, manufacturers will use a measured drag area (Cd A) value for each tractor configuration for Phase 2. See 40 CFR 1037.525. The agencies do not project that vocational vehicles will need to improve their aerodynamic performance to comply with the Phase 2 vocational chassis standards. However, the agencies are providing features in GEM for vocational vehicles to receive credit for improving the aerodynamics of vocational vehicles (see 40 CFR 1037.520(m)).

In addition to these changes, we are making a number of aerodynamic drag test procedure improvements. One improvement is to update the “standard trailer” that is prescribed for use during aerodynamic drag testing of a tractor. Using the Cd A from such testing means the standard trailer would also be the hypothetical trailer modeled in GEM to represent a trailer paired with the tractor in actual use.[167] In Phase 1, a non-aerodynamic 53-foot long box-shaped dry van trailer was specified as the standard trailer for tractor aerodynamic testing (see 40 CFR 1037.501(g)). For Phase 2 we are modifying this standard trailer for tractor testing to make it more similar to the trailers we expect to be produced during the Phase 2 timeframe. More specifically, we are prescribing the installation of aerodynamic trailer skirts (and low rolling resistance tires as applied in Phase 1) on the standard trailer, as discussed in further in Section III.E.2. As explained more fully in Sections III and IV, the agencies believe that tractor-trailer pairings will be optimized aerodynamically to a significant extent in-use (such as using high-roof cabs when pulling box trailers), and that this real-world optimization should be reflected in the certification testing. We are also revising the test procedures to better account for average wind yaw angle to reflect the true impact of aerodynamic features on the in-use fuel consumption and CO2 emissions of tractors, again as discussed in more detail in Section III below. Refer to the test procedures in 40 CFR 1037.525 through 1037.527 for further details of these aerodynamic test procedures.

For trailer certification, the agencies use GEM in a different way than it is used for tractor certification. As described in Section IV, the agencies developed a simple equation to replicate GEM performance. The trailer standards are based on this equation, and trailer manufacturers use this GEM-based equation for certification. The only technologies recognized by this GEM-based equation for trailer certification are aerodynamic technologies, tire technologies (including tire rolling resistance and tire pressure systems), and weight reduction. Note that since the purpose of this equation is to replicate GEM performance, it can be considered as simply another form of the model using a different input interface. Thus, for simplicity, the remainder of this Section II.C. sometimes discusses GEM as being used for trailers, without regard to how manufacturers will actually input GEM variables. As with all of the standards in Phase 2, compliance is measured consistent with the same test methods used by the agencies to establish the standard.

Similar to tractor certification, trailer manufacturers will use data from aerodynamic testing (e.g., coastdown testing, scale wind tunnel testing, computational fluid dynamics modeling, or possibly aerodynamic component testing) with the equation.[168] As part of the protocol for generating these inputs, the agencies are specifying the configuration of a reference tractor for conducting trailer testing. Refer to Section IV of this Preamble and to 40 CFR 1037.501 of the regulations for details on the reference tractor configuration for trailer test procedures.

Finally, GEM has been modified to accept an optional delta Cd A value for vocational chassis, to simulate aerodynamic improvements relative to pre-specified baseline defined in Chapter 4 of RIA. For example, a manufacturer that demonstrates that adding side skirts to a box truck reduces its Cd A by 0.2 m2 could input that value into GEM for box trucks that include those skirts. See 40 CFR 1037.520(m).

(f) Tires and Tire Inflation Systems for Truck and Trailer Certification

For GEM in Phase 1 tractor and vocational chassis manufacturers input the tire rolling resistance of steer and drive tires directly into GEM. The agencies prescribed an internationally recognized tire rolling resistance test procedure, ISO 28580, for determining the tire rolling resistance value that is input into GEM, as described in 40 CFR 1037.520(c). For Phase 2 we will continue this same approach and the use of ISO 28580, and we are expanding these requirements to trailer tires as well.

In addition to tire rolling resistance, Phase 2 vehicle manufacturers will enter into GEM the tire manufacturer's specified revolutions per distance directly (revs/mile) for the vehicle's drive tires. This value is commonly reported by tire manufacturers already so that vehicle speedometers can be adjusted appropriately. This input value is needed so that GEM can accurately convert simulated vehicle speed into axle speed, transmission speed, and ultimately engine speed.

For tractors and trailers, we proposed to allow manufacturers to specify whether or not an automatic tire inflation system (ATIS) is installed. 80 FR 40187. Based on comments and as discussed further in Sections III, IV, and V, in the Phase 2 final rule we are adopting provisions that allow manufacturers of tractors, trailers, and vocational vehicle chassis to input a percent decrease in overall fuel consumption and CO2 emissions into GEM if the vehicle includes either an ATIS or a tire pressure monitoring system (TPMS). The value that can be input depends on whether a TPMS or ATIS is deployed. See 40 CFR 1037.520.

(g) Weight Reduction for Tractor, Vocational Chassis and Trailer Certification

Phase 2 GEM continues the weight reduction recognition approach in Phase 1, where the agencies prescribe fixed weight reductions, or “deltas,” for using certain lightweight materials for certain vehicle components. In Phase 1 the agencies published a list of weight reductions for using high-strength steel and aluminum materials on a part by part basis. For Phase 2 we use updated values for high-strength steel and aluminum parts for tractors and for trailers and we have scaled these values for use in certifying the different weight classes of vocational chassis. In addition we use a similar part by part weight reduction list for tractor parts made from thermoplastic material. We proposed to assign a fixed weight increase to natural gas fueled vehicles to reflect the weight increase of natural gas fuel tanks versus gasoline or diesel tanks, but we are not finalizing that provision based on comments. 80 FR 40187. Commenters opposing this provision generally noted that the proposed provision was not consistent with how the agencies were treating other technologies. We agree that Start Printed Page 73546natural gas vehicles should be treated consistently with other technologies and so are not adopting the proposed provision.

For tractors, we will continue the same mathematical approach in GEM to assign 1/3 of a total weight decrease to a payload increase and 2/3 of the total weight decrease to a vehicle mass decrease. For Phase 1, these ratios were based on the average frequency that a tractor operates at its gross combined weight rating. We will also use these ratios for trailers in Phase 2. For vocational chassis, for which Phase 1 did not address weight reduction, we will assign 1/2 of a total weight decrease to a payload increase and 1/2 of the total weight decrease to a vehicle mass decrease.

(h) GEM Duty Cycles for Tractor, Vocational Chassis and Trailer Certification

In Phase 1, there are three GEM vehicle duty cycles that represent stop-and-go city driving (ARB Transient), urban highway driving (55 mph), and rural interstate highway driving (65 mph). In Phase 1 these cycles were time-based. That is, they were specified as a function of simulated time and the duty cycles ended once the specified time elapsed in simulation. The agencies proposed to continue to use these three drive cycles in Phase 2, but with some revisions. 80 FR 40187. We are finalizing revisions similar but not identical to those that were proposed. First, GEM will simulate these cycles on a distance-based specification, rather than on a time-based specification. A distance-based specification ensures that even if a vehicle in simulation does not always achieve the target vehicle speed, the vehicle will have to continue in simulation for a longer period to complete the duty cycle. This ensures that vehicles are evaluated over the complete distance of the duty cycle and not just the portion of the duty cycle that a vehicle completes in a given time period. A distance-based duty cycle specification also facilitates a straightforward specification of road grade as a function of distance along the duty cycle. As noted in above, for Phase 2, the agencies have enhanced the 55 mph and 65 mph duty cycles by adding representative road grade to exercise the simulated vehicle's engine, transmission, axle, and tires in a more realistic way. A flat road grade profile over a constant speed test does not properly simulate a transmission with respect to shifting gears, and may have the unintended consequence of enabling underpowered vehicles or excessively down-sped drivetrains to generate credits, when in actuality the engine does not remain down-sped in-use when the vehicle encounters road grades. The road grade profile being finalized is the same hill and valley profile for both the 55 mph and 65 mph duty cycles, and is based on statistical analysis of the United States' national distribution of road grades. Although the final profile is different than that proposed, the agencies provided notice of the analysis that was used to generate the final profile.[169] In written comments, we received in-use engine data from some manufacturers, and based on this information we made minor adjustments to the road grade to ensure that engines simulated in GEM operated similarly to that reported in the in-use engine data submitted to us. See Section III.E.(2)(b) of this document and Chapter 3.4.2.1 of the RIA for more details on development of the road grade profile. We believe that the enhancement of the 55 mph and 65 mph duty cycles with road grade is consistent with the NAS recommendation regarding road grade.[170]

(i) Workday Idle Operation for Vocational Chassis Certification

In the Phase 1 program, reduction in idle emissions was recognized only for sleeper cab tractors, and only with respect to hoteling idle, where a driver needs power to operate heating, ventilation, air conditioning and other electrical equipment in order to use the sleeper cab to eat, rest, or conduct other business. As described in Section V, GEM for Phase 2 will recognize technologies that reduce workday idle emissions, such as automatic stop-start systems, daytime parked idle automatic engine shutdown systems, and transmissions that either automatically or inherently shift to neutral at idle while in drive. Many vocational vehicle applications operate on patterns implicating workday idle cycles, and the agencies use test procedures in GEM to account specifically for these cycles and potential idle controls. GEM will recognize these idle controls in two ways. For technologies like neutral-idle transmissions and stop-start systems that address idle that occurs during vehicle operation when the vehicle is stopped at a stop light, GEM will interpolate lower fuel rates from the engine map during the idle portions of the ARB Transient and during a separate GEM “drive idle cycle.” For technologies like start-stop and auto-shutdown that eliminate some of the idle that occurs when a vehicle is stopped or parked, GEM will assign a value of zero fuel rate during a separate GEM “parked idle cycle.” The idle cycles will be weighted along with the 65 mph, 55 mph, and ARB Transient duty cycles, according to the new vocational chassis duty cycle weighting factors. These weighting factors are different for each of the three vocational chassis speed categories for Phase 2. For tractors, only neutral idle and hotel idle will be addressed in GEM.

(2) Experimental Validation of GEM

The core simulation algorithms in GEM have not changed significantly since the proposal. Most of the changes since proposal focused on streamlining how manufacturers input data into GEM; revising to the drive cycles in GEM; and updating how GEM weights these different drive cycles to determine a composite fuel consumption value. These changes did not alter the fundamental way that GEM simulates varying vehicle “road load” and how GEM converts vehicle speed to engine speed and then interpolates engine maps to determine vehicle fuel consumption and CO2 emissions.

Refinements to GEM since the time of proposal that did alter GEM's simulation performance include modifying the default transmissions' shift strategies and their power losses. Another key refinement was cycle average mapping engines for simulation of the ARB Transient cycle. Each time the agencies made such modifications to GEM, GEM's correlation to the agencies collection of laboratory-generated engine and vehicle data was checked. Potential refinements to GEM were accepted if GEM's correlation was improved versus this set of experimental data. If potential refinements resulted in GEM's correlation to the experimental data Start Printed Page 73547becoming worse, those potential changes were rejected. Chapter 4.3.2 of the RIA details the GEM validation that was performed to determine if potential changes to GEM should be accepted or rejected. The first step of the validation process involves simulating vehicles in GEM using engine fuel maps and transmission shifting strategies obtained from manufacturers and comparing GEM results to experiments conducted with the same engines and transmissions. This first step re-validates all of the non-powertrain elements of GEM, which were already validated in Phase 1. The second step is to use GEM's default transmissions' shift strategies in simulation [171] and then compare GEM results to powertrain tests of several transmissions. The only difference between the first and second step is the shifting strategy and powertrain energy loss assumptions. This step facilitates tuning of GEM's default transmission models so that they correlate well to a variety of real transmissions. The third step is to compare GEM simulations to real-world in-use recorded data from actual vehicles. This is the most challenging step because the experimental data includes real-world effects of wind, road grade, and driver behavior in traffic. The most important element of this third step is not absolute correlation, but rather, relative correlation, which demonstrates that when a technology is added to a real vehicle, the relative improvement in the real world is simulated in GEM with a high degree of correlation.

In the first validation step, the agencies compared GEM to over 130 vehicle variants, consistent with the recommendation made by the NAS in their Phase 2-First Report.[172] As described in Chapter 4 of the RIA, good agreement was observed between GEM simulations and test data over a wide range of vehicles. In general, the model simulations agreed with experimental test results within ±5 percent on an absolute basis. As pointed out in Chapter 4.3.2 of the RIA, relative accuracy is more relevant to the intent of this rulemaking, which is to accelerate the adoption of additional fuel efficiency improving technologies. Consistent with the intent of this rulemaking, all of the numeric standards for tractors, trailers and vocational chassis are derived from running GEM first with Phase 1 “baseline” technology packages and then with various Phase 2 technology packages. The differences between these GEM results are examined to determine final stringencies. In other words, the agencies used the same final version of GEM to establish the numeric standards as will be used by manufacturers to demonstrate compliance. Therefore, it is most important that GEM accurately reflects relative changes in emissions for each added technology. In other words, for vehicle certification purposes it is less important that GEM's absolute value of the fuel consumption or CO2 emissions be accurate compared to laboratory testing of the same vehicle. The ultimate purpose of GEM is to evaluate changes or additions in technology, and compliance is demonstrated on a relative basis to the numeric standards that were also derived from GEM. Nevertheless, the agencies concluded that the absolute accuracy of GEM is generally within ±5 percent, as shown in Figure II.2 2. Chapter 4.3.2 of the RIA shows that relative accuracy is even better, ±2-3 percent.

Start Printed Page 73548

In addition to this successful validation against experimental results, the agencies have also conducted a peer review of the GEM source code. This peer review has been submitted to Docket number EPA-HQ-OAR-2014-0827.

The second validation step was to repeat the first step's GEM simulations with the agencies' default transmission shift strategies.171 It was expected that GEM's absolute accuracy would decrease because these shift strategies were tuned for best average performance and for a particular transmission. Nevertheless, it was shown that relative accuracy did not suffer; therefore, the agencies deemed the GEM default shift strategies acceptable for GEM certification purposes. Further details of this validation step are presented in Chapter 4.3.2.3 of the RIA and in a SwRI final report.162

As explained above and in Chapter 4.3.2.3 of the RIA, it is challenging to achieve absolute correlation between any computer simulation and real-world vehicle operation. Therefore, the agencies focused on relative comparisons. Following the SAE standard procedure SAE J1321 “Type II,” two trucks have been tested and these real-world results were compared to GEM simulations. In summary, the relative comparisons between GEM simulations and the real-world testing of trucks showed a 2.4 percent difference. The details of this testing and correlation analysis is presented in Chapter 4.3.2.3 of the RIA.

In conclusion, the agencies completed a number of validation steps to ensure that GEM demonstrates a reasonable degree of absolute accuracy, but more importantly a high degree of relative accuracy, versus both laboratory and real-world experimental data.

(3) Supplements to GEM Simulation

As in Phase 1, for most tractors and vocational vehicles, compliance with the Phase 2 g/ton-mile vehicle standards could be evaluated by directly comparing the GEM result to the standard. However, in Phase 1, manufacturers incorporating innovative or advanced technologies could apply improvement factors to lower the GEM result before comparing to the standard.[173] For example, a manufacturer incorporating a launch-assist mild hybrid that was pre-approved for a 5 percent benefit would apply a 0.95 improvement factor to its GEM results for such vehicles. In this example, a GEM result of 300 g/ton-mile will be reduced to 285 g/ton-mile.

For Phase 2, the agencies largely continue the existing Phase 1 innovative technology approach, but we name it “off-cycle” to better reflect its purpose.

(a) Off-Cycle Technology Procedures

In Phase 1 the agencies adopted an emissions credit generating opportunity that applied to new and innovative technologies that reduce fuel consumption and CO2 emissions, which were not in common use with heavy-duty vehicles before model year 2010 and are not reflected over the test procedures or GEM (i.e., the benefits are “off-cycle”). See 76 FR 57253. As was the case in the development of Phase 1, the agencies continue this approach for technologies and concepts with CO2 emissions and fuel consumption reduction potential that might not be adequately captured over the Phase 2 duty cycles or are not inputs to GEM. Note, however, that the agencies now refer to these technologies as off-cycle rather than innovative. Comments were generally supportive of continuing this provision. See Section I.C(1)(c) of this document and Section 1 of the RTC for more discussion of innovative and off-cycle technologies.

We recognize that the Phase 1 testing burden associated with the innovative technology credit provisions discouraged some manufacturers from applying. To streamline recognition of many technologies, default values have been integrated directly into GEM. For example, automatic tire inflation systems have fixed default values, and such technologies are now recognized through a post-simulation adjustment approach, discussed in Chapter 4 of the RIA. This is similar to the technology “pick list” from our light-duty programs. See 77 FR 62833-62835 (October 15, 2012). If manufacturers wish to receive additional credit beyond these fixed values, then the off-cycle technology credit provisions provide a regulatory path toward that additional recognition.

Beyond the additional technologies that the agencies have added to GEM, the agencies also believe there are several emerging technologies that are being developed today, but will not be accounted for in GEM because we do not have enough information about these technologies to assign fixed values to them in GEM. Any credits for these technologies will need to be based on the off-cycle technology credit generation provisions. These require the assessment of real-world fuel consumption and GHG reductions that can be measured with verifiable test methods using representative operating conditions typical of the engine or vehicle application.

As in Phase 1, the agencies continue to provide two paths for approval of the test procedure to measure the CO2 emissions and fuel consumption reductions of an off-cycle technology used in the HD tractor. See 40 CFR 1037.610 and 49 CFR 535.7. The first path does not require a public approval process of the test method. A manufacturer can use “pre-approved” test methods for HD vehicles including the A-to-B chassis testing, powertrain testing or on-road testing. A manufacturer may also use any developed test procedure which has known quantifiable benefits. A test plan detailing the testing methodology is required to be approved by the agencies prior to collecting any test data. The agencies will also continue the second path which includes a public approval process of any testing method which could have uncertain benefits (i.e., an unknown usage rate for a technology). Furthermore, the agencies are modifying our provisions to better clarify the documentation required to be submitted for approval aligning them with provisions in 40 CFR 86.1869-12, and NHTSA separately prohibits credits from technologies addressed by any of its crash avoidance safety rulemakings (i.e., congestion management systems).

Sections III and V separately describe tractor and vocational vehicle technologies, respectively, that the agencies anticipate may qualify for these off-cycle credit provisions.

(4) Production Vehicle Testing for Comparison to GEM

As described in Section III.E.(2)(j), The agencies are requiring tractor manufacturers to annually chassis test five production vehicles over the GEM cycles to verify that relative reductions simulated in GEM are being achieved in production. See 40 CFR 1037.665. We do not expect absolute correlation between GEM results and chassis testing. GEM makes many simplifying assumptions that do not compromise its usefulness for certification, but do cause it to produce emission rates different from what would be measured during a chassis dynamometer test. Given the limits of correlation possible between GEM and chassis testing, we would not expect such testing to accurately reflect whether a vehicle was compliant with the GEM standards. Therefore, we are not applying GHG compliance liability to such testing. Rather, this testing will be for data collection and informational purposes only. The agencies will continue to evaluate in-use compliance Start Printed Page 73549by verifying GEM inputs and testing in-use engines. (Note that NTE standards for criteria pollutants may apply for some portion of the test cycles.)

(5) Use of GEM in Establishing the Phase 2 Numerical Standards

As in Phase 1, the agencies are setting specific numerical standards against which tractors and vocational vehicles will be certified using GEM (box trailers will use a GEM-based equation, and some trailers and custom chassis vocational vehicles may optionally use a non-GEM certification path). Although these standards are performance-based standards, which do not specifically require the use of any particular technologies,[174] the agencies established these standards by evaluating specific vehicle technology packages using the final version of Phase 2 GEM. We note that that this means the final numerical standards are not directly comparable to the proposed standards, which were based on an intermediate version of GEM, rather than on the final version.

(a) Relation to In-Use Emissions

The purpose of this rulemaking is to achieve in-use emission and fuel consumption reductions by requiring manufacturers to demonstrate that they meet the promulgated emission standards. Thus, it is important that GEM simulations be reasonably representative of in-use operation. Testing that is unrepresentative of actual in-use operation does not necessarily tell us anything about whether any emission reductions occur. However, we recognize that certain simplifications are necessary for practical simulations. In the past, EPA has addressed this issue by including in our testing regulations a process by which EPA can work with manufacturers to adjust test procedures to make them more representative of in-use operation. For engine testing, this provision is in 40 CFR 1065.10(c)(1), where EPA requires manufacturers to notify us in cases in which they determine that the specified test procedures would result in measurements that do not represent in-use operation.

Although we are not adopting an equivalent provision for GEM at this time, we expect similar principles to apply. To the extent that GEM fails to represent in-use emission, we would expect to work with manufacturers to address the issue—under the existing regulations where possible, or by promulgating a new rulemaking.

We recognize that many compromises must be made between the practicality of testing/simulation and the matching of in-use operation. We have considered many aspects of the test procedures in this respect for the engines, vehicles, and emission controls of which we are currently aware. We have concluded that the procedures will generally result in emission simulations that are sufficiently representative of in-use emissions, even though not all in-use operation will occur during simulation. Nevertheless, we have identified several areas that deserve some additional discussion.

GEM is structured to simulate a single vehicle weight (curb weight plus payload) per regulatory subcategory. However, we know that actual in-use weights will rarely be exactly the same as the simulated weights. Nevertheless, since the representativeness of the simulated weights (or lack thereof) is being fully considered in the setting of the standards, there would be no need to modify the procedures to account for different curb weights or payloads.

GEM simulates vehicle emissions over three drive cycles plus two idle cycles, and weights the cycle results based on the type of vehicle being certified. These cycles and weightings reflect fleet average driving patterns and the agencies do not expect them to fully match driving patterns for individual vehicles. Thus, we would generally not consider GEM's cycles as unrepresentative for vehicles with different in-use driving patterns. However, if new information became available that demonstrated that GEM's cycles somehow did not reflect fleet average driving patterns, the agencies would consider such information in the context of the principles of representative testing, described above.

Finally, GEM includes default values for axle and transmission efficiency derived from baseline technologies. However, we generally expect manufacturers to use more efficient axles and transmissions for Phase 2 vehicles. As noted above, based on comments, the agencies are allowing manufacturers to optionally input measured efficiencies to better represent these more efficient technologies. We would not consider GEM unrepresentative if manufacturers chose to use the default values rather than measure these efficiencies directly.

(b) Relation to Powertrain Testing

As already noted, GEM correlates very well with powertrain testing. To the extent they differ, it would be expected to be primarily related to how transmission performance is modeled in GEM. Although GEM includes a sophisticated model of transmissions, it cannot represent a transmission better than a powertrain test of the same transmission. Thus, the agencies consider powertrain testing to be as good as or better than GEM run using engine-only fuel maps; hence the provision in the final rules allowing results from powertrain testing to be used as a GEM input.

In some respects, powertrain testing can be considered to be a reference method for this rulemaking. Because manufacturers have the option to perform powertrain testing instead of engine-only fuel mapping, the stringency of the final standards can be traced to powertrain testing. In other words, methods that can be shown to be equivalent to powertrain testing can be considered to be consistent with the testing that was used as the basis of the final Phase 2 standards.

In a related context, it may be useful in the future to consider equivalency to powertrain testing as an appropriate criterion for evaluating changes to GEM to address new technologies. Consider, for example, a new technology that is not represented in GEM, but that is reflected in powertrain testing. The agencies could determine that it would be appropriate to modify GEM to reflect the technology rather than to require manufacturers to perform powertrain testing. In such a case, the agencies would not consider the modification to GEM to impact the effective stringency of the Phase 2 standards because the new version of GEM would be equivalent to performing powertrain testing.

D. Engine Test Procedures and Engine Standards

In addition to the Phase 1 GEM-based vehicle certification of tractors and vocational chassis, the agencies also set Phase 1 separate CO2 and fuel efficiency standards for the engines installed in tractors and vocational chassis. EPA also set Phase 1 separate engine standards for capping methane (CH4) and nitrous oxide (N2 O) emissions (essentially capping emissions at current emission levels). Compliance with all of these Phase 1 separate engine standards is demonstrated by measuring these emissions during an engine dynamometer test procedure. For Phase 1 the agencies use the same test procedure specified for EPA's existing heavy-duty engine emissions standards (e.g., NOX and PM standards). These Phase 1 engine standards are specified in terms of brake-specific (g/bhp-hr) fuel, CO2, CH4 and N2 O emissions limits. Since the test procedure already Start Printed Page 73550specified how to measure fuel consumption, CO2 and CH4, few changes were needed to utilize the test procedure for Phase 1, the most notable change being a modification specifying how to measure N2 O.

There are some differences in how these non-GHG test procedures are applied in Phase 1 and Phase 2. In EPA's non-GHG engine emissions standards, heavy-duty engines must meet brake-specific standards for emissions of total oxides of nitrogen (NOX), particulate mass (PM), non-methane hydrocarbon (NMHC), and carbon monoxide (CO). These standards must be met by all engines both over a 13-mode steady-state duty cycle called the “Supplemental Emissions Test” (SET) [175] and over a composite of a cold-start and a hot-start transient duty cycle called the “Federal Test Procedure” (FTP). In contrast, for Phase 1 the agencies require that engines specifically installed in tractors meet fuel efficiency and CO2 standards over only the SET but not the composite FTP. This requirement was intended to reflect that tractor engines typically operate near steady-state conditions versus transient conditions. See 76 FR 57159. For Phase 2 the agencies are finalizing, as proposed, slight changes to the 13-modes' weighting factors to better reflect in-use engine operation. These weighting factors apply only for determining SET fuel consumption and CO2 emissions. No changes are being made to the weighting factors for EPA's non-GHG emission standards. The agencies adopted the converse for engines installed in vocational vehicles. That is, these engines must meet fuel efficiency and CO2 standards over the composite FTP but not the SET. This requirement was intended to reflect that vocational vehicle engines typically operate under transient conditions versus steady-state conditions (76 FR 57178). For both tractor and vocational vehicle engines in Phase 1, EPA set CH4 and N2 O emissions cap standards over the composite FTP only and not over the SET duty cycle. See Section II.D. for details on this final action's engine test procedures for Phase 2.

In response to the agencies' proposed engine standards, we received a number of public comments. The agencies considered those comments, and the following list summarizes key changes we've made in response, and more detailed descriptions of these changes are presented in Chapter 2.7 of the RIA:

  • Recalculated the SET baseline using the new Phase 2 SET weighting factors.
  • Recalculated the FTP baseline, based on MY 2016 FTP certification data from Cummins, DTNA, Volvo, Navistar, Hino, Isuzu, Ford, GM and FCA. These included HHD, MHD, and LHD engines.
  • Projected how manufacturers would modify maximum fuel rates as a function of speed to strategically relocate SET mode points to achieve lowest SET results.
  • Projected a higher market penetration of WHR in 2027, versus what we proposed.
  • Decreased our projected impact of engine technology dis-synergies by increasing the magnitude of our so-called “dis-synergy factors;” accounting for these changes by increasing the research and development costs needed for this additional optimization.

The following section first describes the engine test procedures used to certify engines to the Phase 2 separate engine standards. Sections that follow describe the Phase 2 CO2, N2 O and CH4 separate engine standards and their feasibility.

(1) Engine Test Procedures

(a) SET Cycle Weighting

The SET cycle was adopted by EPA in 2000 and modified in 2005 from a discrete-mode test to a ramped-modal cycle to broadly cover the most significant part of the speed and torque map for heavy-duty engines, defined by three non-idle speeds and three relative torques. The low speed is called the “A speed,” the intermediate speed is called the “B speed,” and the high speed is called the “C speed.” As is shown in Table II-1, the SET cumulatively weights these three speeds at 23 percent, 39 percent, and 23 percent.

Table II-1—SET Modes Weighting Factor in Phase 1

Speed, % LoadWeighting factor in Phase 1 (%)
Idle15
A, 1008
B, 5010
B, 7510
A, 505
A, 755
A, 255
B, 1009
B, 2510
C, 1008
C, 255
C, 755
C, 505
Total100
Cumulative A Speed23
Cumulative B Speed39
Cumulative C Speed23

The C speed is typically in the range of 1800 rpm for current heavy heavy-duty engine designs. However, it is becoming much less common for engines to operate at such a high speeds in real-world driving conditions, and especially not during cruise vehicle speeds in the 55 to 65 mph vehicle speed range. This trend has been corroborated by engine manufacturers' in-use data that has been submitted to the agencies in comments and presented at technical conferences.[176] Thus, although the current SET represents highway operation better than the FTP cycle, it could be improved by adjusting its weighting factors to better reflect modern trends in in-use engine operation. Furthermore, the most recent trends indicate that manufacturers are configuring drivetrains to operate engines at speeds down to a range of 1050-1200 rpm at a vehicle speed of 65 mph.

To address this trend toward in-use engine down-speeding, the agencies are finalizing as proposed refined SET weighting factors for the Phase 2 CO2 emission and fuel consumption standards. The new SET mode weightings move most of the C weighting to “A” speed, as shown in Table II-2. To better align with in-use data, these changes also include a reduction of the idle speed weighting factor. These new mode weightings do not apply to criteria pollutants or to the Phase 1 CO2 emission and fuel consumption standards.

Table II-2—New SET Modes Weighting Factor in Phase 2

Speed/% loadWeighting factor in Phase 2 (%)
Idle12
A, 1009
B, 5010
B, 7510
A, 5012
A, 7512
A, 2512
B, 1009
Start Printed Page 73551
B, 259
C, 1002
C, 251
C, 751
C, 501
Total100
Total A Speed45
Total B Speed38
Total C Speed5

(b) Engine Test Provisions for SET, FTP, and Engine Mapping for GEM Inputs

Although GEM does not apply directly to engine certification, Phase 2 will require engine manufacturers to generate and certify full load and motoring torque curves and engine fuel rate maps for input into GEM for tractor and vocational chassis manufacturers to demonstrate compliance to their respective standards. The full load and motoring torque curve procedures were previously defined in 40 CFR part 1065, and these are already required for non-GHG emissions certification. The Phase 2 final default test procedure for generating an engine map for GEM's 55 mph and 65 mph drive cycles is the “steady-state” mapping procedure. However, the agencies are finalizing an option for manufacturers to use the “cycle average” mapping procedure for GEM's 55 mph and 65 mph drive cycles. The test procedure for generating an engine map for GEM's ARB Transient drive cycle is the “cycle-average” mapping procedure, and the agencies are not finalizing any other mapping options for the ARB Transient drive cycle. Note that if an engine manufacturer elects to conduct powertrain testing to generate inputs for GEM, then steady-state and cycle-average engine maps would not be required for those GEM vehicle configurations to which the powertrain test inputs would apply. The steady-state and cycle-average test procedures are specified in 40 CFR parts 1036 and 1065. The technical and confidential business information motivations for finalizing these test procedures are explained in II. B. (2), along with a summary of comments we received.

One important consideration is the need to correct measured fuel consumption rates for the carbon and energy content of the test fuel. As proposed, we will continue the Phase 1 approach, which is specified in 40 CFR 1036.530. We are specifying a similar approach to GEM fuel maps in Phase 2.

As proposed, the agencies are requiring that engine manufacturers certify fuel maps for GEM, as part of their certification to the engine standards. However, there were a number of manufacturer comments strongly questioning the particular proposed requirement that engine manufacturers provide these maps to vehicle manufacturers starting in MY 2020 for the certification of vehicles commercially marketed as MY 2021 vehicles in calendar year 2020. This is a normal engine and vehicle manufacturing process, where many vehicles may be produced with engines having an earlier model year than the commercial model year of the vehicle. For example, we expect that some MY 2021 vehicles will be produced with MY 2020 engines. Thus, we proposed to require engine manufacturers to begin providing GEM fuel maps for MY 2020 engines so that vehicle manufacturers could run GEM to certify MY 2021 vehicles with MY 2020 engines. EMA and some of its members commented that MY 2020 engines should not be subject to Phase 2 requirements, based on NHTSA's statutory 4-year lead-time requirement and because the potential higher fuel consumption of MY 2020 (i.e., Phase 1) engine maps could force vehicle manufacturers to install additional technologies that were not projected by the agencies for compliance. The agencies considered these comments along with the potential cost savings for manufacturers to align the timing of both their engines' and vehicle's Phase 2 product plans and certification paths. The agencies also considered how this situation would repeat in MY 2024 and MY 2027 and possibly with future standards as well. Based on these considerations, we have decided that it would be more appropriate to harmonize the engine and vehicle standards, starting in MY 2021 so that vehicle manufacturers will not need fuel maps for 2020 engines. Thus, we are not finalizing the requirement to provide fuel maps for MY 2020 engines. However, we are requiring fuel maps for all MY 2021 engines, even those (e.g., small businesses) for which the Phase 2 engine and vehicle standards have been delayed. See 40 CFR 1036.150.

The current engine test procedures also require the development of regeneration emission rate and frequency factors to determine infrequent regeneration adjustment factors (IRAFs) that account for the emission changes for criteria pollutants during an exhaust emissions control system regeneration event. In Phase 1 the agencies adopted provisions to exclude CO2 emissions and fuel consumption due to regeneration. However, for Phase 2, we are requiring the inclusion of CO2 emissions and fuel consumption due to regeneration over the FTP and SET (RMC) cycles, as determined using the IRAF provisions in 40 CFR 1065.680. While some commenters opposed this because of its potential impact on stringency, we do not believe this will significantly impact the stringency of these standards because manufacturers have already made great progress in reducing the frequency and impact of regeneration emissions since 2007. Rather, the agencies are including IRAF CO2 emissions for Phase 2 to prevent these emissions from increasing in the future to the point where they would otherwise become significant. Manufacturers qualitatively acknowledged the likely already small and decreasing magnitude of IRAF CO2 emissions in their comments. For example, EMA stated, “the rates of infrequent regenerations have been going down since the adoption of the Phase 1 standards” and that IRAF “contributions are minor.” Nevertheless, we believe it is prudent to begin accounting for regeneration emissions to discourage manufacturers from adopting criteria emissions compliance strategies that could reverse this trend. Manufacturers expressed concern about the additional test burden, but the only additional requirement would be to measure and report CO2 emissions for the same tests they are already performing to determine IRAFs for other pollutants.

At the time of the proposal, we did not specifically adjust baseline levels to include additional IRAF emissions because we believed them to be negligible and decreasing. Commenters opposing this proposed provision provided no data to dispute this belief. We continue to believe that regeneration strategies can be engineered to maintain these negligible rates. Thus, we do not believe they are of fundamental significance for our baselines in the FRM. Highway operation includes enough high temperature operation to make active regenerations unnecessary. Furthermore, recent improvements in exhaust after-treatment catalyst formulations and exhaust temperature thermal management strategies, such as intake air throttling, minimize CO2 IRAF impacts during non-highway operation, where active regeneration might be required. Finally, as is discussed in Section II.D.(2), recent significant Start Printed Page 73552efficiency improvements over the FTP cycle suggest that FTP emissions may actually be even lower than we have estimated in our updated FTP baselines, which would provide additional margin for manufacturers to manage any minor CO2 IRAF impacts that may occur.

We are not including fuel consumption due to after-treatment regeneration in the creation of fuel maps used in GEM for vehicle compliance. We believe that the IRAF requirements for the separate SET and FTP engine standards, along with market forces that already exist to minimize regeneration events, will create sufficient incentives to reduce fuel consumption during regeneration over the entire fuel map.

(c) Powertrain Testing

The agencies are finalizing a powertrain test option to afford a robust mechanism to quantify the benefits of CO2 reducing technologies that are a part of the powertrain (conventional or hybrid), that are not captured in the GEM simulation. Among these technologies are integrated engine and transmission control and hybrid systems. We are finalizing a number of improvements to the test procedure in 40 CFR 1037.550. As proposed we are finalizing the requirement for Phase 2 hybrid powertrains to mapped using this powertrain test method. The agencies are also finalizing modifications to 40 CFR 1037.550 to separate out the hybrid specific testing protocols.

To limit the amount of testing under this rule, powertrains can be divided into families and are tested in a limited number of simulated vehicles that will cover the range of vehicles in which the powertrain will be used. A matrix of 8 to 9 tests will be needed per vehicle cycle, to enable the use of the powertrain results broadly across all the vehicles in which the powertrain will be installed. The individual tests differ by the vehicle that is being simulated during the test. These are discussed in detail in Chapter 3.6 of the RIA.

(i) Powertrain Test Procedure

The agencies are expanding upon the test procedures defined 40 CFR 1037.550 for Phase 1 hybrid vehicles. The Phase 2 expansion will migrate the current Phase 1 test procedure to a new 40 CFR 1037.555 and will modify the current test procedure in 40 CFR 1037.550, allowing its use for Phase 2 only. The Phase 2 modifications relative to 40 CFR 1037.550 include the addition of the rotating inertia of the driveline and tires, and the axle efficiency. This revised procedure also requires that each of the powertrain components be cooled so that the temperature of each of the components is kept in the normal operation range. We are extending the powertrain procedure to PHEV powertrains.

Powertrain testing contains many of the same requirements as engine dynamometer testing. The main differences are where the test article connects to the dynamometer and the software that is used to command the dynamometer and operator demand setpoints. The powertrain procedure finalized in Phase 2 allows for the dynamometer(s) to be connected to the powertrain either upstream of the drive axle or at the wheel hubs. The output of the transmission is upstream of the drive axle for conventional powertrains. In addition to the transmission, a hydraulic pump or an electric motor in the case of a series hybrid may be located upstream of the drive axle for hybrid powertrains. If optional testing with the wheel hub is used, two dynamometers will be needed, one at each hub. Beyond these points, the only other difference between powertrain testing and engine testing is that for powertrains, the dynamometer and throttle setpoints are not set by fixed speed and torque targets prescribed by the cycle, but are calculated in real time by the vehicle model. The powertrain test procedure requires a forward calculating vehicle model, thus the output of the model is the dynamometer speed setpoints. The vehicle model calculates the speed target using the measured torque at the previous time step, the simulated brake force from the driver model, and the vehicle parameters (tire rolling resistance, drag area, vehicle mass, rotating mass, and axle efficiency). The operator demand that is used to change the torque from the engine is controlled such that the powertrain follows the vehicle speed target for the cycle instead of being controlled to match the torque or speed setpoints of the cycle. The emission measurement procedures and calculations are identical to engine testing.

(ii) Engine Test Procedures for Replicating Powertrain Tests

As described in Section II.B.(2)(b), the agencies are finalizing the proposed powertrain test option to quantify the benefits of CO2-reducing powertrain technologies. This option is very similar to the cycle average mapping approach, although these powertrain test results would be used to override both the engine and transmission (and possibly axle) simulation portions of GEM, not just the engine fuel map. The agencies are requiring that any manufacturer choosing to use this option also measure engine speed and engine torque during the powertrain test so that the engine's performance during the powertrain test could be replicated in a non-powertrain engine test cell. Manufacturers would be required to measure or calculate, using good engineering judgment, the engine shaft output torque, which would be close-coupled to the transmission input shaft during a powertrain test. Subsequent engine testing then could be conducted using the normal part 1065 engine test procedures as specified in 40 CFR 1037.551, and g/bhp-hr CO2 results could be compared to the levels the manufacturer reported during certification. Such testing could apply for both confirmatory and selective enforcement audit (SEA) testing. This would simplify both the certification and SEA testing.

As proposed, engine manufacturers certifying powertrain performance (instead of or in addition to the multi-point fuel maps) will be held responsible for powertrain test results. If the engine manufacturer does not certify powertrain performance and instead certifies only the steady-state and/or cycle-average fuel maps, it will held responsible for fuel map performance rather than the powertrain test results. Engine manufacturers certifying both will be responsible for both.

Some commenters objected to the potential liability for such engine-only tests. However, it appears they do not understand our intent. This provision states clearly that this approach could be used only where “the test engine's operation represents the engine operation observed in the powertrain test.” Also, since the manufacturers perform all SEA testing themselves, this would be an option for the manufacturer rather than something imposed by EPA. Thus, this concern should be limited to the narrow circumstance in which EPA performs confirmatory engine testing of an engine that was certified using powertrain testing, follows the manufacturer's specified engine test cycle, and ensures that the test accurately represents the engine's performance during the powertrain test. However, it is not clear why this would be problematic. It is entirely reasonable to assume that testing the engine in this way would result in equivalent emission results. To the extent manufacturer concerns remain, each manufacturer would be free to certify their engines based on engine-only fuel maps rather than powertrain testing.

(d) CO2 From Urea SCR Systems

For diesel engines utilizing urea SCR emission control systems for NOXStart Printed Page 73553reduction, the agencies will allow, but not require, correction of the final engine (and powertrain) fuel maps to account for the contribution of CO2 from the urea injected into the exhaust. This urea typically contributes 0.2 to 0.5 percent of the total CO2 emissions measured from the engine, and up to 1 percent at certain map points. Since current urea production methods use gaseous CO2 captured from the atmosphere (along with NH3), CO2 emissions from urea consumption does not represent a net carbon emission. This adjustment is necessary so that fuel maps developed from CO2 measurements will be consistent with fuel maps from direct measurements of fuel flow rates. This adjustment is also necessary to fully align EPA's CO2 standards with NHTSA's fuel consumption standards. Failing to account for urea CO2 tailpipe emissions would result in reporting higher fuel consumption than what was actually consumed. Thus, we are only allowing this correction for emission tests where CO2 emissions are determined from direct measurement of CO2 and not from fuel flow measurement, which would not be impacted by CO2 from urea.

We note that this correction will be voluntary for manufacturers, and we expect that some manufacturers may determine that the correction is too small to be of concern. The agencies will use this correction for CO2 measurements with any engines for which the engine manufacturer applied the correction for its fuel maps during certification.

We are not allowing this correction for engine test results with respect to the engine CO2 standards. Both the Phase 1 standards and the new standards for CO2 from diesel engines are based on test results that included CO2 from urea. In other words, these standards are consistent with using a test procedure that does not correct for CO2 from urea.

(2) Engine Standards for CO2 and Fuel Consumption

We are largely maintaining the existing Phase 1 regulatory structure for engine standards, which had separate standards for spark-ignition engines (such as gasoline engines) and compression-ignition engines (such as diesel engines), and for HHD, MHD and LHD engines, but we are changing how these standards will apply to alternative fuel engines as described in Section XII.A.2.

Phase 1 applied different test cycles depending on whether the engine is used for tractors, vocational vehicles, or both, and we are continuing this approach. Tractor engines are subject to standards over the SET, while vocational engines are subject to standards over the FTP. Table II-3 shows the Phase 1 standards for diesel engines.

Table II-3—Phase 1 MY 2017 Diesel Engine CO2 and Fuel Consumption Standards

UnitsHHD SETMHD SETHHD FTPMHD FTPLHD FTP
g/bhp-hr460487555576576
gal/100 bhp-hr4.51874.78395.45195.65825.6582

In the Phase 2 proposal we assumed that these numeric values of the Phase 1 standards were the baselines for Phase 2. We applied our technology assessments to these baselines to arrive at the Phase 2 standards for MY 2021, MY 2024 and MY 2027. In other words, for the Phase 2 proposal we projected that starting in MY 2017 engines would, on average, just meet the Phase 1 standards and not over-comply. However, based on comments we received on how to consistently apply our new SET weighting factors in our analysis and based on recent MY 2016 engine certification data, we are updating our Phase 2 baseline assumptions for both the SET and FTP.

First, with respect to the SET, in the proposal we compared our proposed Phase 2 standards, which are based on these new Phase 2 weighting factors, to the Phase 1 numeric standards, which are based on the current Phase 1 weighting factors. Because we continue to use the same 13-mode brake specific CO2 and fuel consumption numeric values we used for the proposal to represent the performance of a MY 2017 baseline engine, we are not projecting a different technology level in the baseline. Rather, this is simply correcting an “apples-to-oranges” comparison from the proposal by applying the Phase 2 weighting factors to the MY 2017 baseline engine. This was pointed out to us by UCS, ICCT and EDF in their public comments. While this did not impact our technology effectiveness or cost analyses, it did impact the numeric value of our baseline to which we reference the effectiveness of applying technologies to the 13 individual modes of the SET. Because the revised SET weighting factors result in somewhat lower brake specific CO2 and fuel consumption numeric results for the composite baseline SET value, this correction, in turn, lowers the numerical values of the final Phase 2 SET standards. Making this particular update did not result in a change to the relative stringency of the final Phase 2 numeric engine standards (relative to MY 2017 baseline performance), but our updated feasibility analysis did; see Section II.D.(2)(a) below).

Second, the agencies made adjustments to the FTP baselines, but these adjustments were not made because of a calculation error. Rather, MY 2016 FTP certification data showed an unexpected step-change improvement in engine fuel consumption and CO2 emissions. These data were not available at the time of proposal, so the agencies relied upon the MY 2017 Phase 1 standard as a baseline. EDF publicly commented in response to the NODA that the more recent certification data revealed this new step-change. MY 2016 certification data submitted to the agencies [177] as well as to ARB [178] show that many engines from many manufacturers already not only achieve the Phase 1 FTP standards, but some were also below the MY 2027 standards proposed for Phase 2. This was not the case for the SET, where most manufacturers are still not yet complying with the MY 2017 Phase 1 SET standards. In view of this situation for the FTP, the agencies are adjusting the Phase 2 FTP baseline to reflect this shift. The underlying reasons for this shift are mostly related to manufacturers optimizing their SCR thermal management strategy over the FTP in ways that we (mistakenly) thought they already had in MY 2010 (i.e., the Phase 1 baseline). As background, the FTP includes a cold-start, a hot-start and significant time spent at engine idle. During these portions of the FTP, the NOX SCR system can cool down and lose NOX reducing efficiency. One simplistic strategy to maintain SCR temperature is to inefficiently consume additional fuel, such that the fuel energy is lost to the Start Printed Page 73554exhaust system in the form of heat. There are more sophisticated strategies to maintain SCR temperature, however, but these apparently required additional time from MY 2010 for research, development and refinement. In updating these baseline values, the agencies did consider the concerns raised by manufacturers about the potential impact of IRAFs on baseline emissions.

As just noted, at the time of Phase 1 we had not realized that these improvements were not already in the Phase 1 baseline. These include optimizing the use of an intake throttle to decrease excess intake air at idle and SCR catalyst reformulation to maintain SCR efficiency at lower temperatures. Based on this information, which was provided to the agencies by engine manufacturers, but only after we specifically requested this information, the agencies concluded that in Phase 1 we did not account for how much further these kinds of improvements could still impact FTP fuel consumption. Conversely, only by reviewing the new MY 2016 certification data did we realize how little SCR thermal management optimization actually occurred for the engine model years that we used to establish the Phase 1 baseline—namely MY 2009 and MY 2010 engines. Because we never accounted for this kind of improvement in our Phase 2 proposal's stringency analysis for meeting the Phase 2 proposed FTP standards, this baseline shift does not alter our projected effectiveness and market adoption rates from the proposal. Therefore, we continue to apply the same improvements that we proposed, but we apply them to the updated FTP baseline. See Section II.D.(5) for a discussion on how this impacts carry-over of Phase 1 emission credits.

Table II-4 shows the Phase 2 diesel engine final CO2 baseline emissions. Note that the gasoline engine CO2 baseline for Phase 2 is the same as the Phase 1 HD gasoline FTP standard, 627 g/bhp-hr. More detailed analyses on these Phase 2 baseline values of tractor and vocational vehicles can be found in Chapter 2.7.4 of RIA.

Table II-4—Phase 2 Diesel Engine Final CO2 and Fuel Consumption Baseline Emissions

UnitsHHD SETMHD SETHHD FTPMHD FTPLHD FTP
g/bhp-hr455481525558576
gal/100 bhp-hr4.46954.72505.15725.48135.6582

As described below, the agencies are adopting standards for new compression-ignition engines for Phase 2, commencing in MY 2021, that will require additional reductions in CO2 emissions and fuel consumption beyond the Phase 2 baselines. The agencies are not adopting new CO2 or fuel consumption engine standards for new heavy-duty gasoline engines. Note, however, that we are projecting some small improvement in gasoline engine performance that will be recognized over the vehicle cycles (that is, reflected in the stringency of certain of the vocational vehicle standards). See Section V.B.2.a below.

For diesel engines to be installed in Class 7 and 8 combination tractors, the agencies are adopting the SET standards shown in Table II-5.[179] The MY 2027 SET standards for engines installed in tractors will require engine manufacturers to achieve, on average, a 5.1 percent reduction in fuel consumption and CO2 emissions beyond the Phase 2 baselines. We are also adopting SET standards in MY 2021 and MY 2024 that will require tractor engine manufacturers to achieve, on average, 1.8 percent and 4.2 percent reductions in fuel consumption and CO2 emissions, respectively, beyond the Phase 2 baselines.

Table II-5—Phase 2 Heavy-Duty Tractor Engine Standards for Engines 180 Over the SET Cycle

Model yearStandardHeavy heavy-dutyMedium heavy-duty
2021-2023CO2 (g/bhp-hr) Fuel Consumption (gallon/100 bhp-hr)447 4.3910473 4.6464
2024-2026CO2 (g/bhp-hr) Fuel Consumption (gallon/100 bhp-hr)436 4.2829461 4.5285
2027 and LaterCO2 (g/bhp-hr) Fuel Consumption (gallon/100 bhp-hr)432 4.2436457 4.4892

For diesel engines to be installed in vocational chassis, the agencies are adopting the FTP standards shown in Table II-6. The MY 2027 FTP standards for engines installed in vocational chassis will require engine manufacturers to achieve, on average, a 4.2 percent reduction in fuel consumption and CO2 emissions beyond the Phase 2 baselines. We are also adopting FTP standards in MY 2021 and MY 2024 that will require vocational chassis engine manufacturers to achieve, on average, 2.3 percent and 3.6 percent reductions in fuel consumption and CO2 emissions, respectively, beyond the Phase 2 baselines.Start Printed Page 73555

Table II-6—Vocational Diesel (CI) Engine Standards Over the Heavy-Duty FTP Cycle

Model yearStandardHeavy heavy-duty 181Medium heavy-duty diesel 181Light heavy-duty diesel 182
2021-2023CO2 (g/bhp-hr) Fuel Consumption (gallon/100 bhp-hr)513 5.0393545 5.3536563 5.5305
2024-2026CO2 (g/bhp-hr) Fuel Consumption (gallon/100 bhp-hr)506 4.9705538 5.2849555 5.4519
2027 and LaterCO2 (g/bhp-hr) Fuel Consumption (gallon/100 bhp-hr)503 4.9411535 5.2554552 5.4224

(a) Feasibility of the Diesel (Compression-Ignition) Engine Standards

In this section, the agencies discuss our assessment of the feasibility of the engine standards and the extent to which they conform to our respective statutory authorities and responsibilities. More details on the technologies discussed here can be found in RIA Chapter 2.3. The feasibility of these standards is further discussed in RIA Chapter 2.7 for tractor and vocational vehicle engines. While the projected technologies are discussed here separately, as is discussed at the beginning of this Section II.D, the agencies also accounted for dis-synergies between technologies. Note that Section II.D.(2)(e) discusses the potential for some manufacturers to achieve greater emission reductions by introducing new engine platforms, and how and why these reductions are reflected in the tractor and vocational vehicle standards.

Based on the technology analysis described below, the agencies project that a technology path exists that will allow engine manufacturers to meet the final Phase 2 standards by 2027, and to meet the MY 2021 and 2024 standards. The agencies also project that these manufacturers will be able to meet these standards at a reasonable cost and without adverse impacts on in-use reliability.

In general, engine performance for CO2 emissions and fuel consumption can be improved by improving the internal combustion process and by reducing energy losses. More specifically, the agencies have identified the following key means by which fuel efficiency can be improved:

  • Combustion optimization
  • Turbocharger design and optimization
  • Engine friction and other parasitic loss reduction
  • Exhaust after-treatment pressure drop reduction
  • Intake air and exhaust system pressure drop reduction (including EGR system)
  • Engine down-sizing to improve core engine efficiency
  • Engine down-speeding over the SET, and in-use, by lug curve shape optimization
  • Waste heat recovery system installation and optimization
  • Physics model based electronic controls for transient performance optimization

The agencies are gradually phasing in the separate engine standards from 2021 through 2027 so that manufacturers can gradually introduce these technology improvements. For most of these, the agencies project manufacturers could begin applying these technologies to about 45-50 percent of their heavy-duty engines by 2021, 90-95 percent by 2024, and ultimately apply them to 100 percent of their heavy-duty engines by 2027. However, for some of these improvements (such as waste heat recovery and engine downsizing) we project lower application rates in the Phase 2 time frame. This phase-in structure is consistent with the normal manner in which manufacturers introduce new technology to manage limited R&D budgets as well as to allow them to work with fleets to fully evaluate in-use reliability before a technology is applied fleet-wide. The agencies believe the phase-in schedule will allow manufacturers to complete these normal processes. See RIA 2.3.9.

Based on our technology assessment described below, the engine standards appear to be consistent with the agencies' respective statutory authorities. All of the technologies with high penetration rates above 50 percent have already been demonstrated to some extent in the field or in research laboratories, although some development work remains to be completed. We note that our feasibility analysis for these engine standards is not based on projecting 100 percent application for any technology until 2027. We believe that projecting less than 100 percent application is appropriate and gives us additional confidence that the 2021 and 2024 MY standards are feasible.

Because this analysis considers reductions from engines meeting the Phase 1 standards, it assumes manufacturers will continue to include the same compliance margins as in Phase 1. In other words, a manufacturer currently declaring FCLs 10 g/bhp-hr above its measured emission rates (in order to account for production and test-to-test variability) will continue to do the same in Phase 2. Both the costs and benefits are determined relative to these baselines, and so are reflective of these compliance margins.

The agencies have carefully considered the costs of applying these technologies, which are summarized in Section II.D.(2)(d). These costs appear to be reasonable on both a per engine basis, and when considering payback periods.[183] The engine technologies are discussed in more detail below. Readers are encouraged to see the RIA Chapter 2.7 for additional details (and underlying references) about our feasibility analysis.

(i) Combustion Optimization

Although manufacturers are making significant improvements in combustion to meet the Phase 1 engine standards, the agencies project that even more improvement is possible after 2018. For example, improvements to fuel injection systems will allow more flexible fuel injection capability with higher injection pressure, which can provide more opportunities to improve engine fuel efficiency. Further optimization of piston bowls and injector tips will also improve engine performance and fuel efficiency. We project that a reduction of up to 1.0 percent is feasible in the 2024 model year through the use of Start Printed Page 73556these technologies, although it will likely apply to only 95 percent of engines until 2027.

Another important area of potential improvement is advanced engine control incorporating model based calibration to reduce losses of control during transient operation. Improvements in computing power and speed will make it possible to use much more sophisticated algorithms that are more predictive than today's controls. Because such controls are only beneficial during transient operation, they will reduce emissions over the FTP cycle, over the ARB Transient cycle's cycle-average mapping procedure, and during in-use operation, but this technology will not reduce emissions over the SET cycle or over the steady-state engine mapping procedure. Thus, the agencies are projecting model based control reductions only for vocational engines' FTP standards and for projecting improvements captured by the cycle-average mapping over the ARB Transient cycle. Although this control concept is not currently available and is still under development, we project model based controls achieving a 2 percent improvement in transient emissions. Based on model based controls already in widespread use in engine laboratories for the calibration of simpler controllers and based on recent model based control development under the DOE SuperTruck partnership (e.g., DTNA's SuperTruck engine's model based controls), we project that such controls could be in limited production for some engine models by 2021. We believe that some vocational chassis applications would particularly benefit from these controls in-use (e.g., urban applications with significant in-use transient operation). Therefore, we project that a modest amount of engine models will have these controls by MY 2021. We also project that manufacturers will learn more from the in-use operation of these technology leading engines, and manufacturers will be able to improve these controls even further, such that they would additionally benefit other vocational applications, such as multi-purpose and regional applications. By 2027, we project that 40 percent of all vocational diesel engines will incorporate model-based controls at a 2 percent level of effectiveness.

(ii) Turbocharging System

Many advanced turbocharger technologies can be brought into production in the time frame between 2021 and 2027, and some of them are already in production, such as mechanical or electric turbo-compounding, more efficient variable geometry turbines, and Detroit Diesel's patented asymmetric turbocharger. A turbo-compound system, like those installed on some of Volvo's EURO VI compliant diesels and on some of DTNA's current U.S. offerings (supplied to DTNA by a division of Cummins), extracts energy from the exhaust to provide additional power. Mechanical turbo-compounding includes a power turbine located downstream of the turbine which in turn is connected to the crankshaft to supply additional power. On-highway demonstrations of this technology began in the early 1980s. It was used first in heavy duty production in the U.S. by Detroit Diesel for their DD15 and DD16 engines and reportedly provided a 3 to 5 percent fuel consumption reduction. Results are duty cycle dependent, and require significant time at high load to realize an in-use fuel efficiency improvement. Lightly loaded vehicles on flat roads or at low vehicle speeds can expect little or no benefit. Volvo reports two to four percent fuel consumption improvement in line haul applications.[184] Because of turbo-compound technology's drive cycle dependent effectiveness, the agencies are only projecting a market penetration of 10 percent for all tractor engines, at slightly less than 2 percent effectiveness over the SET. The agencies are considering turbo-compound to be mutually exclusive with WHR because both technologies seek to extract additional usable work from the same waste heat and are unlikely to be used together.

(iii) Engine Friction and Parasitic Losses

The friction associated with each moving part in an engine results in a small loss of engine power. For example, frictional losses occur at bearings, in the valve train, and at the piston ring-cylinder interface. Taken together such losses represent a measurable fraction of all energy lost in an engine. For Phase 1, the agencies projected a 1-2 percent reduction in fuel consumption due to friction reduction. However, new information leads us to project that an additional 1.4 percent reduction is possible for some engines by 2021 and all engines by 2027. These reductions are possible due to improvements in bearing materials, lubricants, and new accessory designs such as variable-speed pumps.

(iv) After-Treatment Optimization

All heavy duty diesel engine manufacturers are already using diesel particulate filters (DPFs) to reduce particulate matter (PM) and selective catalytic reduction (SCR) to reduce NOX emissions. The agencies see two areas in which improved after-treatment systems can also result in lower fuel consumption. First, increased SCR efficiency could allow re-optimization of combustion for better fuel consumption because the SCR would be capable of reducing higher engine-out NOX emissions. We don't expect this to be significant, however. Manufacturers already optimize the DEF (urea) consumption and fuel consumption to achieve the lowest cost of operation; taking into account fuel consumption, DEF consumption and the prices of fuel and DEF. Therefore, if manufacturers re-optimized significantly for fuel consumption, it is possible that this would lead to higher net operating costs. This scenario is highly dependent upon fuel and DEF prices, so projecting this technology path is uncertain. Second, improved designs could reduce backpressure on the engine to lower pumping losses. If manufacturers have opportunities to lower backpressure within the size constraints of the vehicle, the agencies project that manufacturers will opt to lower after-treatment back pressure. The agencies project the combined impact of these improvements would be 0.6 percent over the SET.

Note that this improvement is independent of cold-start improvements made recently by some manufacturers with respect to vocational engines. Thus, the changes being made to the FTP baseline engines do not reduce the likelihood of the benefits of re-optimizing after-treatment projected here.

(v) Engine Intake and Exhaust Systems

Various high efficiency air handling for both intake air and exhaust systems could be produced in the 2020 and 2024 time frame. To maximize the efficiency of such processes, induction systems may be improved by manufacturing more efficiently designed flow paths (including those associated with air cleaners, chambers, conduit, mass air flow sensors and intake manifolds) and by designing such systems for improved thermal control. Improved turbocharging and air handling systems will likely include higher efficiency EGR systems and intercoolers that reduce frictional pressure losses while maximizing the ability to thermally control induction air and EGR. EGR systems that often rely upon an adverse pressure gradient (exhaust manifold pressures greater than intake manifold pressures) must be reconsidered and their adverse pressure gradients Start Printed Page 73557minimized. Other components that offer opportunities for improved flow efficiency include cylinder heads, ports and exhaust manifolds to further reduce pumping losses by about 1 percent over the SET.

(vi) Engine Downsizing and Down Speeding

Proper sizing of an engine is an important component of optimizing a vehicle for best fuel consumption. This Phase 2 rule will require reductions in road load due to aerodynamic resistance, tire rolling resistance and weight, which will result in a drop in the vehicle power demand for most operation. This drop moves the engine operating points down to a lower load zone, which can move the engine away from operating near its peak thermal efficiency (a.k.a. the “sweet spot”). Engine downsizing combined with engine down speeding can allow the engine to move back to higher loads and a lower speed zone, thus achieving better fuel efficiency in the real world. However, because of the way engines are tested, little of the benefit of engine downsizing would be detected during engine testing (if power density remains the same) because the engine test cycles are de-normalized based on the full torque curve. Thus, the separate engine standards are not the appropriate standards for recognizing the benefits of engine downsizing. Nevertheless, we project that some small benefit can be measured over the engine test cycles depending on the characteristics of the engine fuel map and how the SET points are determined as a function of the engine's lug curve.

After the proposal we received comments recommending that we should recognize some level of engine down speeding within the separate engine standards. Based on this comment and some additional confidential business information that we received, we believe that engine lug curve reshaping to optimize the locations of the 13-mode points is a way that manufacturers can demonstrate some degree of engine down-speeding over the engine test. As pointed out in Chapter 2.3.8 and 2.7.5 of the RIA, down speeding via lug curve reshaping alone can provide SET reductions in the range of 0.4 percent depending on the engine map characteristics.

(vii) Waste Heat Recovery

More than 40 percent of all energy loss in an engine is lost as heat to the exhaust and engine coolant. For many years, manufacturers have been using turbochargers to convert some of this waste heat in the exhaust into usable mechanical power that is then used to compress the intake air. Manufacturers have also been developing a Rankine cycle-based system to extract additional heat energy from the engine. Such systems are often called waste heat recovery (WHR) systems. The possible sources of waste heat energy include the exhaust, recirculated exhaust gases, compressed charge air, and engine coolant. The basic approach with WHR is to use waste heat from one or more of these sources to evaporate a working fluid, which is passed through a turbine or equivalent expander to create mechanical or electrical power, then re-condensed.

For the proposal, the agencies projected that by 2027, 15 percent of tractor engines would employ WHR systems with an effectiveness of better than three percent. We received many comments on this projection, which are discussed briefly below and in more detail in the RTC. In particular, we note that some of the comments included confidential data related to systems not yet on the market. After carefully considering all of these comments, we have revised our projections to increase the effectiveness, decrease costs, and project higher adoption rates than we proposed.

Prior to the Phase 1 Final Rule, the NAS estimated the potential for WHR to reduce fuel consumption by up to 10 percent.[185] However, the agencies do not believe such levels will be achievable within the Phase 2 time frame. There currently are no commercially available WHR systems for diesel engines, although research prototype systems are being tested by some manufacturers. American Trucking Association, Navistar, DTNA, OOIDA, Volvo, and UPS commented that because WHR is still in the prototype stage, it should not be assumed for setting the stringency of the tractor engine standards. Many of these commenters pointed to the additional design and development efforts that will be needed to reduce cost, improve packaging, reduce weight, develop controls, select an appropriate working fluid, implement expected OBD diagnostics, and achieve the necessary reliability and durability. Some stated that the technology has not been thoroughly tested or asked that more real-world data be collected before setting standards based on WHR. Some of these commenters provided confidential business information pertaining to their analysis of WHR system component costs, failure modes, and projected warranty cost information.

Alternatively, a number of commenters including Cummins, ICCT, CARB, ACEEE, EDF, Honeywell, ARB and others stated that the agencies should increase the assumed application rate of WHR in the final rule and the overall stringency of the engine standards. They argued the agencies' WHR technology assessment was outdated and too conservative, the fuel savings and GHG reduction estimation for WHR were too low, and the agencies' cost estimates were based on older WHR systems where costs were confounded with hybrid component costs and that these have since been improved upon. In addition, the agencies received CBI information supporting the arguments of some of these commenters.

Cummins stated the agencies underestimated the commercial viability of WHR and that we overstated the development challenges and timing in the NPRM. They said WHR can provide a 4 to 5 percent improvement in fuel consumption on tractor drive cycles and that WHR would be commercially viable and available in production as early as 2020 and will exceed the agencies' estimates for market penetration over the period of the rule. According to Cummins, the reliability of their WHR system has improved with each generation of the technology and they have developed a smaller system footprint, improved integration with the engine and vehicle and a low-GWP working fluid, resulting in a much more compact and integrated system. They added that their system would be evaluated in extended customer testing by the end of 2015, and that results of that experience will inform further technology development and product engineering leading to expected commercial product availability in the 2020 timeframe. Furthermore, they said multiple product development cycles over the implementation timeframe of the rule would provide opportunities for further development for reduced cost and improved performance and reliability.

Some commenters, including EDF, said the agencies' assumed design had little in common with the latest designs planned for production. They cited several publications, including the NAS 21st Century Truck Program report #3 and stated WHR effectiveness is much higher than the agencies estimated. Gentham cited an ICCT study saying that up to a 12 percent fuel consumption reduction from a 2010 baseline engine is possible with the application of advanced engine technologies and WHR.Start Printed Page 73558

The agencies recognize that much work remains to be done, but we are providing significant lead time to bring WHR to market. Based on our assessment of each manufacturer's work to date, we are confident that a commercially-viable WHR capable of reducing fuel consumption by over three percent will be available in the 2021 to 2024 time frame. Concerns about the system's cost and complexity may remain high enough to limit the use of such systems in this time frame. Moreover, packaging constraints and lower effectiveness under transient conditions will likely limit the application of WHR systems to line-haul tractors. Refer to RIA Chapter 2.3.9 for a detailed description of these systems and their applicability. For our analysis of the engine standards, the agencies project that WHR with the Rankine technology could be used on 1 percent of tractor engines by 2021, on 5 percent by 2024, and 25 percent by 2027, with nearly all being used on sleeper cabs. We project this sharper increase in market adoption in the 2027 timeframe because we have noted that most technology adoption rate curves follow an S-shape: Slow initial adoption, then more rapid adoption, and then a leveling off as the market saturates (not always at 100 percent).[186] We assumed an S-shape curve for WHR adoption, where we project a steeper rise in market adoption in and around the 2027 timeframe. Given our averaging, banking and trading program flexibilities and that manufacturers may choose from a range of other technologies, we believe that manufacturers will be able to meet the 2027 standards, which we based on a 25 percent WHR adoption in tractor engines. Although we project these as steps, it is more likely that manufacturers will try to gradually increase the WHR adoption in MY 2025 and MY 2026 from the 5 percent in 2024 to generate emission credits to smooth the transition to the 2027 standards.

Commenters opposing the agencies' WHR projections argued that the real-world GHG and fuel consumption savings will be less than in prototype systems. DTNA said a heat rejection increase of 30 percent to 40 percent with WHR systems will require larger radiators, resulting in more aerodynamic drag and lower fuel savings from WHR systems. DTNA cited a Volvo study showing a 2 percent loss of efficiency with the larger frontal areas needed to accommodate heat rejection from WHR systems. Daimler stated effectiveness may be lower than expected since there is large drop off in fuel savings when the tractor is not operating on a steady state cycle and the real world performance of WHR systems will be hurt by transient response issues. Daimler and ACEEE said the energy available from exhaust and other waste heat sources could diminish as tractor aerodynamics improve, thus lowering the expected fuel savings from WHR. Daimler said because of this, WHR estimated fuel savings was overestimated by the agencies. Navistar said WHR working fluids will have a significant GHG impact based on their high global warming potential. They commented that fuel and GHG reductions will be lower in the real world with the re-weighting of the RMC which results in lower engine load, and thus lower available waste heat. However, none of these commenters have access to the full range of data available to the agencies, which includes CBI.

It is important to note that the net cost and effectiveness of future WHR systems depends on the sources of waste heat. Systems that extract heat from EGR gases may provide the side benefit of reducing the size of EGR coolers or eliminating them altogether. To the extent that WHR systems use exhaust heat, they increase the overall cooling system heat rejection requirement and likely require larger radiators. This could have negative impacts on cooling fan power needs and vehicle aerodynamics. Limited engine compartment space under the hood could leave insufficient room for additional radiator size increasing. Many of these issues disappear if exhaust waste heat is not recovered from the tailpipe and brought under the hood for conversion to mechanical work. In fact, it is projected that if a WHR system only utilizes heat that was originally within the engine compartment (e.g., EGR cooler heat, coolant heat, oil heat, etc.), then any conversion of that heat to mechanical heat actually reduces the heat rejection demand under the hood; potentially leading to smaller radiators and lower frontal area, which would actually lead toward improved aerodynamic performance. Refer to RIA Chapter 2.3.9 for more discussion.

Several commenters stated that costs are highly uncertain for WHR technology, but argued that the agencies' assumption of a $10,523 cost in 2027 are likely significantly lower than reality. Volvo estimated a cost of $21,700 for WHR systems. Volvo said that in addition to hardware cost being underestimated, the agencies had not properly accounted for other costs such as the R&D needed to bring the technology into production within a vehicle. Volvo said they would lose $17,920 per unit R&D alone, excluding other costs such as materials and administrative expenses. Daimler said that costs almost always inflate as the complexity of real world requirements drive up need for more robust designs, sensors, controls, control hardware, and complete vehicle integration. They added that development costs will be large and must be amortized over limited volumes. Furthermore, OOIDA said the industry experience with such complex systems is that maintenance, repair, and down-time cost can be much greater than the initial purchase cost. ATA and OOIDA said that potential downtime associated with an unproven technology is a significant concern for the industry.

On the other hand, some commenters argued that the agencies had actually overestimated WHR costs in the proposal. These commenters generally argued that engineering improvements to the WHR systems that will go into production in the Phase 2 time frame would lower costs, in particular by reducing components. The agencies largely agree with these commenters and we have revised our analysis to reflect these cost savings. See RIA 2.11.2.15 for additional discussion.

(viii) Technology Packages for Diesel Engines Installed in Tractors

This Section (a)(viii) describes technology packages that the agencies project could be applied to Phase 1 tractor engines to meet the Phase 2 SET separate engine standards. Section II.D.(2)(e) also describes additional improvements that the agencies project some engine manufacturers will be able to apply to their engines.

We received comments on the tractor engine standards in response to the proposal and in response to the NODA. These comments can be grouped into two general themes. One theme expressed by ARB, non-governmental environmentally focused organizations, Cummins and some technology suppliers like Honeywell, recommended higher engine stringencies, up to 10-15 percent in some comments. Another theme, generally expressed by vertically integrated engine and vehicle manufacturers supported either no Phase 2 engine standards at all, or they supported the proposal's standards, but none of these commenters supported standards that were more stringent than what we proposed. An example of the contrast between these two themes can be shown in one report submitted to the docket and another submission rebutting the statements made in the Start Printed Page 73559report. The report was submitted to the agencies by the Environmental Defense Fund (EDF).[187] On the other hand, four vertically integrated engine and vehicle manufacturers, DTNA, Navistar, Paccar, and Volvo, submitted a rebuttal to EDF's findings.[188] Some of these individual vehicle manufacturers also provided their own comments on EDF's report.[189 190] Cummins also provided comments and recommended stringencies somewhere between EDF's recommendations and the integrated manufacturers' rebuttal. Cummins recommended achieving reductions by 2030 in the range of 9-15 percent. CARB's recommendation from their comments [191] is 7.1 percent in 2024.

The agencies carefully considered this wide range of views, and based on the best data available, the agencies modified some of our technology projections between the proposal and the final rule.

Table II-5 lists our projected technologies together with our projected effectiveness and market adoption rates for tractor engines. The reduction values shown as ”SET reduction” are relative to our Phase 2 baseline values, as shown in Table II-7. It should be pointed out that the reductions in Table II-7 are based on the Phase 2 final SET weighting factors, shown in Table II-2. RIA Chapter 2.7.5 details the reasoning supporting our projection of improvements attributable to this fleet average technology package.

Table II-7—Projected Tractor Engine Technologies and Reduction

SET modeSET weighted reduction (%) 2020-2027Market penetration (2021) (%)Market penetration (2024) (%)Market penetration (2027) (%)
Turbo compound with clutch1.951010
WHR (Rankine cycle)3.61525
Parasitic/Friction (Cyl Kits, pumps, FIE), lubrication1.54595100
After-treatment (lower dP)0.63095100
EGR/Intake & exhaust manifolds/Turbo/VVT/Ports1.14595100
Combustion/FI/Control1.14595100
Downsizing0.3102030
Overall reductions (%)
Weighted reduction (%)1.74.04.8
Down speeding optimization on SET0.10.20.3
Total % reduction1.84.25.1

The weighted reductions shown in this table have been combined using the “Π-formula,” which has been augmented to account for technology dis-synergies that occur when combining multiple technologies. A 0.85 dis-synergy factor was used for 2021, and a 0.90 dis-synergy factor was used for 2024 and 2027.[192] RIA Chapter 2.7.4 provides details on the “Π-formula” and an explanation for how the dis-synergy factors were determined. Some commenters argued that use of a single dis-synergy factor for all technologies is inappropriate. While we agree that it would be preferable to have a more detailed analysis of the dis-synergy between each pair or group of technologies, we do not have the information necessary to conduct such an analysis. In the absence of such information, the simple single value approach is a reasonable approximation. Moreover, we note that the degree of dis-synergy is sufficiently small to make the impact of any errors on the resulting standards negligible.

Figure II.3 2018 HHD Figure II.4 are the samples of the HHD engine fuel maps used for the agencies' MY 2018 baseline engine and MY 2027 sleeper cab engine for tractors. As can be seen from these two figures, the torque curve shapes are different. This is because engine down speeding optimization for the SET is taken into consideration, where the engine peak torque is increased and the engine speed is shifted to lower speed. All maps used by GEM for all vehicles are shown in Chapter 2.7 of the RIA.

Start Printed Page 73560

(ix) Technology Packages for Diesel Engines Installed in Vocational Vehicles

For diesel engines (and other compression-ignition engines) used in vocational vehicles, the MY 2021 standards will require engine manufacturers to achieve, on average, a 2.3 percent reduction in fuel consumption and CO2 emissions beyond the Phase 2 FTP baselines. Beginning in MY 2024, the agencies are requiring a 3.6 percent reduction in fuel consumption and CO2 emissions beyond the Phase 2 FTP baselines for all diesel engines including LHD, MHD, and HHD, and beginning in MY 2027 this increases to 4.2 percent, on average. The agencies have based these FTP standards on the performance of reduced parasitic and friction losses, improved after-treatment, combustion optimization, superchargers and variable geometry turbochargers, physics model-based controls, improved EGR pressure drop, and variable valve timing (only in LHD and MHD engines). Start Printed Page 73561The percent reduction for the MY 2021, MY 2024, and MY 2027 standards is based on the combination of technology effectiveness and the respective market adoption rates projected.

Most of the potential engine technologies discussed previously for tractor engines can also be applied to vocational engines. However, neither of the waste heat technologies, Rankine cycle nor turbo-compound, are likely to be applied to vocational engines because they are less effective under transient operation, which is weighted more heavily for all of the vocational sub-categories. Given the projected cost and complexity of such systems, we believe that for the Phase 2 time frame manufacturers will focus their WHR development work on tractor applications (which will have better payback for operators), rather than on vocational applications. In addition, the benefits due to engine downsizing, which can be realized in some tractor engines, may not be realized at all in in the vocational sector, again because this control technology produces few benefits under transient operation.

One of the most effective technologies for vocational engines is the optimization of transient controls with physics model based control, which would replace current look-up table based controls. These are described more in detail in Chapter 2.3 of the RIA. We project that more advanced transient controls, including different levels of model based control, discussed in Chapter 2.3 of the RIA, would continue to progress and become more broadly applicable throughout the Phase 2 timeframe.

Other effective technologies include parasitic load/friction reduction, as well as improvements to combustion, air handling systems, turbochargers, and after-treatment systems. Table II-8 below lists those potential technologies together with the agencies' projected market penetration rates for vocational engines. Again, similar to tractor engines, the technology reduction and market penetration rates are estimated by combining manufacturer-submitted confidential business information, together with estimates reflecting the agencies' judgment, which is informed by historical trends in the market adoption of other fuel efficiency improving technologies. The reduction values shown as “percent reduction” are relative to the Phase 2 FTP baselines, which are shown in Table II-3. The overall reductions combine the technology reduction values with their market adoption rates. The same set of the dis-synergy factors as the tractor are used for MY 2021, 2024, and 2027.

Table II-8—Projected Vocational Engine Technologies and Reduction

TechnologyPercent reduction 2020-2027Market penetration 2021 (%)Market penetration 2024 (%)Market penetration 2027 (%)
Model based control2.0253040
Parasitic/Friction1.56090100
EGR/Air/VVT/Turbo1.06090100
Improved AT0.53060100
Combustion Optimization1.06090100
Weighted reduction (%)-L/M/HHD2.33.64.2

Figure II.5 is a sample of a 2018 baseline engine fuel map for a MHD vocational engine.

Start Printed Page 73562

(x) Summary of the Agencies' Analysis of the Feasibility of the Diesel Engine Standards

The HD Phase 2 standards are based on projected adoption rates for technologies that the agencies regard as the maximum feasible for purposes of EISA section 32902 (k) and appropriate under CAA section 202(a) based on the technologies discussed above and in RIA Chapter 2. The agencies believe these technologies can be adopted at the estimated rates for these standards within the lead time provided, as discussed in RIA Chapter 2.7. The 2021 and 2024 MY standards are phase-in standards on the path to the 2027 MY standards, and these earlier standards were developed using less aggressive application rates and therefore have lower technology package costs than the 2027 MY standards.

As described in Section II.D.(2)(d) below, the costs to comply with these standards are estimated to range from $275 to $1,579 per engine. This is slightly higher than the costs for Phase 1, which were estimated to be $234 to $1,091 per engine. Although the agencies did not separately determine fuel savings or emission reductions due to the engine standards apart from the vehicle program, it is expected that the fuel savings will be significantly larger than these costs, and the emission reductions will be roughly proportional to the technology costs when compared to the corresponding vehicle program reductions and costs. Thus, we regard these standards as cost-effective. This is true even without considering payback period. The phase-in 2021 and 2024 MY standards are less stringent and less costly than the 2027 MY standards. Given that the agencies believe these standards are technologically feasible, are highly cost effective, and highly cost effective when accounting for the fuel savings, and have no apparent adverse potential impacts (e.g., there are no projected negative impacts on safety or vehicle utility), they appear to represent a reasonable choice under section 202(a) of the CAA and the maximum feasible under NHTSA's EISA authority at 49 U.S.C. 32902(k)(2).

(b) Basis for Continuing the Phase 1 Spark-Ignited Engine Standard

For gasoline vocational engines, we are not adopting more stringent engine standards. Today most SI-powered vocational vehicles are sold as incomplete vehicles by a vertically integrated chassis manufacturer, where the incomplete chassis shares most of the same technology as equivalent complete pickups or vans, including the powertrain. Another, even less common way that SI-powered vocational vehicles are built is by a non-integrated chassis manufacturer purchasing an engine from a company that also produces complete and/or incomplete HD pickup trucks and vans. Gasoline engines used in vocational vehicles are generally the same engines as are used in the complete HD pickups and vans in the Class 2b and 3 weight categories, although the operational demands of vocational vehicles often require use of the largest, most powerful SI engines, so that some engines fitted in complete pickups and vans are not appropriate for use in vocational vehicles. Given the relatively small sales volumes for gasoline-fueled vocational vehicles, manufacturers typically cannot afford to invest significantly in developing separate technology for these engines.

The agencies received many comments suggesting that technologies be applied to increase the stringency of the SI engine standard. These comments were essentially misplaced, since the agencies already had premised the Phase 1 SI MY 2016 FTP engine standards on 100 percent adoption of these technologies. The commenters thus did not identify any additional engine technologies that the agencies did not already consider and account for in setting the MY 2016 FTP engine standard. Therefore, the Phase 1 SI engine FTP standard for these engines will remain in place. However, as noted above, projected engine improvements are being reflected in the stringency of the vehicle standard for the vehicle in which the engine will be installed. In part this is because the GEM cycles result in very different engine operation than what occurs when an engine is run over the engine FTP cycle. We believe that certain technologies will show a fuel consumption and CO2 emissions reduction during GEM cycles that do not occur over the engine FTP. We received comments on engine technologies that can be recognized over the GEM vehicle cycles. As a result, the Phase 2 gasoline-fueled vocational vehicle standards are predicated on adoption of advanced engine friction reduction and cylinder deactivation. To the extent any SI engines do not incorporate the projected engine technologies, manufacturers of SI-powered vocational vehicles would need to achieve equivalent reductions from some other vehicle technology to meet the vehicle standards. See Section V.C of this Preamble for a description of how we applied these technologies to develop the vocational vehicle standards. See Section VI.C of this Preamble for a description of the SI engine technologies that have been considered in developing the HD pickup truck and van standards.

(c) Engine Improvements Projected for Vehicles Over the GEM Duty Cycles

As part of the certification process for the Phase 2 vehicle standards, tractor and vocational vehicle manufacturers will need to represent their vehicles' actual engines in GEM. Although the vehicle standards recognize the same engine technologies as the separate engine standards, each have different test procedures for demonstrating compliance. As explained earlier in Section II.D.(1), compliance with the tractor separate engine standards is determined from a composite of the Supplemental Engine Test (SET) procedure's 13 steady-state operating points. Compliance with the vocational vehicle separate engine standards is determined over the Federal Test Procedure's (FTP) transient engine duty cycle. In contrast, compliance with the vehicle standards is determined using GEM, which calculates composite results over a combination of 55 mph, 65 mph, ARB Transient and idle vehicle cycles. Each of these duty cycles emphasize different engine operating points; therefore, they can each recognize certain technologies differently. Hence, these engine improvements can be readily recognized in GEM and appropriately reflected in the stringency of the vehicle standards. It is important to note, however, that the tractor vehicle standards presented in Section III project that some (but not all) tractor engines will achieve greater reductions than required by the engine standards. This was reflected in the agencies' feasibility analysis using projected engine fuel maps that represent engines having fuel efficiency better than what is required by the engine standards. Similarly, the vocational vehicle standards in presented in Section V project that the average vocational engine will achieve greater reductions than required by the engine standards. These additional reductions are recognized by GEM and are reflected in the stringency of the respective vehicle standards.

Our first step in aligning our engine technology assessment at both the engine and vehicle levels was to separately identify how each technology impacts performance at each of the 13 individual test points of the SET steady-state engine duty cycle. For example, engine friction reduction technology is expected to have the greatest impact at the highest engine speeds, where frictional energy losses are the greatest. Start Printed Page 73563As another example, turbocharger technology is generally optimized for best efficiency at steady-state cruise vehicle speed. For an engine, this is near its lower peak-torque speed and at a moderately high load that still offers sufficient torque reserve to climb modest road grades without frequent transmission gear shifting. The agencies also considered the combination of certain technologies causing dis-synergies with respect to engine efficiency at each of these test points. See RIA Chapter 2.3 and 2.7 for further details. Chapter 2.8 and 2.9 of the RIA details how the engine fuel maps are created for both tractor and vocational vehicles used for GEM as the default engine fuel maps.

(d) Engine Technology Package Costs for Tractor and Vocational Engines (and Vehicles)

As described in Chapters 2 and 7 of the RIA, the agencies estimated costs for each of the engine technologies discussed here. All costs are presented relative to engines projected to at least comply with the model year 2017 standards—i.e., relative to our Phase 2 baseline engines. Note that we are not presenting any costs for gasoline engines (SI engines) in this section because we are not changing the SI engine standards. However, we are including a cost for additional engine technology as part of the vocational vehicle analysis in Section V.C.2.(e) (and appropriately so, since those engine improvements are reflected in the stringency of the vocational vehicle standard).

Our engine cost estimates include a separate analysis of the incremental part costs, research and development activities, and additional equipment. Our general approach used elsewhere in this action (for HD pickup trucks, gasoline engines, Class 7 and 8 tractors, and Class 2b-8 vocational vehicles) estimates a direct manufacturing cost for a part and marks it up based on a factor to account for indirect costs. See also 75 FR 25376. We believe that approach is appropriate when compliance with the standards is achieved generally by installing new parts and systems purchased from a supplier. In such a case, the supplier is conducting the bulk of the research and development on the new parts and systems and including those costs in the purchase price paid by the original equipment manufacturer. Consequently, the indirect costs incurred by the original equipment manufacturer need not reflect significant cost to cover research and development since the bulk of that effort is already completed. For the MHD and HHD diesel engine segment, however, the agencies believe that OEMs will incur costs not associated with the purchase of parts or systems from suppliers or even the production of the parts and systems, but rather the development of the new technology by the original equipment manufacturer itself. Therefore, the agencies have directly estimated additional indirect costs to account for these development costs. The agencies used the same approach in the Phase 1 HD rule. EPA commonly uses this approach in cases where significant investments in research and development can lead to an emission control approach that requires no new hardware. For example, combustion optimization may significantly reduce emissions and cost a manufacturer millions of dollars to develop but would lead to an engine that is no more expensive to produce. Using a bill of materials approach would suggest that the cost of the emissions control was zero reflecting no new hardware and ignoring the millions of dollars spent to develop the improved combustion system. Details of the cost analysis are included in the RIA Chapter 2.7. To reiterate, we have used this different approach because the MHD and HHD diesel engines are expected to comply in part via technology changes that are not reflected in new hardware but rather reflect knowledge gained through laboratory and real world testing that allows for improvements in control system calibrations—changes that are more difficult to reflect through direct costs with indirect cost multipliers. Note that these engines are also expected to incur new hardware costs as shown in Table II-9 through Table II-12. EPA also developed the incremental piece cost for the components to meet each of the 2021 and 2024 standards. The costs shown in Table II-13 include a low complexity ICM of 1.15 and assume the flat-portion of the learning curve is applicable to each technology.

(i) Tractor Engine Package Costs

Table II-9—MY 2021 Tractor Diesel Engine Component Costs Inclusive of Indirect Cost Markups and Adoption Rates

[2013$]

Medium HDHeavy HD
After-treatment system (improved effectiveness SCR, dosing, DPF)$7$7
Valve Actuation8484
Cylinder Head (flow optimized, increased firing pressure, improved thermal management)33
Turbocharger (improved efficiency)99
Turbo Compounding5151
EGR Cooler (improved efficiency)22
Water Pump (optimized, variable vane, variable speed)4444
Oil Pump (optimized)22
Fuel Pump (higher working pressure, increased efficiency, improved pressure regulation)22
Fuel Rail (higher working pressure)55
Fuel Injector (optimized, improved multiple event control, higher working pressure)55
Piston (reduced friction skirt, ring and pin)11
Valve train (reduced friction, roller tappet)3939
Waste Heat Recovery7171
“Right sized” engine−41−41
Total284284
Note: “Right sized” diesel engine is a smaller, less costly engine than the engine it replaces.
Start Printed Page 73564

Table II-10—MY 2024 Tractor Diesel Engine Component Costs Inclusive of Indirect Cost Markups and Adoption Rates

[2013$]

Medium HDHeavy HD
After-treatment system (improved effectiveness SCR, dosing, DPF)$14$14
Valve Actuation169169
Cylinder Head (flow optimized, increased firing pressure, improved thermal management)66
Turbocharger (improved efficiency)1717
Turbo Compounding9393
EGR Cooler (improved efficiency)33
Water Pump (optimized, variable vane, variable speed)8585
Oil Pump (optimized)44
Fuel Pump (higher working pressure, increased efficiency, improved pressure regulation)44
Fuel Rail (higher working pressure)99
Fuel Injector (optimized, improved multiple event control, higher working pressure)1010
Piston (reduced friction skirt, ring and pin)33
Valve train (reduced friction, roller tappet)7777
Waste Heat Recovery298298
“Right sized” engine−82−82
Total712712
Note: “Right sized” diesel engine is a smaller, less costly engine than the engine it replaces.

Table II-11—MY 2027 Tractor Diesel Engine Component Costs Inclusive of Indirect Cost Markups and Adoption Rates

[2013$]

Medium HDHeavy HD
After-treatment system (improved effectiveness SCR, dosing, DPF)$15$15
Valve Actuation172172
Cylinder Head (flow optimized, increased firing pressure, improved thermal management)66
Turbocharger (improved efficiency)1717
Turbo Compounding8989
EGR Cooler (improved efficiency)33
Water Pump (optimized, variable vane, variable speed)8585
Oil Pump (optimized)44
Fuel Pump (higher working pressure, increased efficiency, improved pressure regulation)44
Fuel Rail (higher working pressure)99
Fuel Injector (optimized, improved multiple event control, higher working pressure)1010
Piston (reduced friction skirt, ring and pin)33
Valve train (reduced friction, roller tappet)7777
Waste Heat Recovery1,2081,208
“Right sized” engine−123−123
Total1,5791,579
Note: “Right sized” diesel engine is a smaller, less costly engine than the engine it replaces.

(ii) Vocational Diesel Engine Package Costs

Table II-12—MY 2021 Vocational Diesel Engine Component Costs Inclusive of Indirect Cost Markups and Adoption Rates

[2013$]

Light HDMedium HDHeavy HD
After-treatment system (improved effectiveness SCR, dosing, DPF)$8$8$8
Valve Actuation939393
Cylinder Head (flow optimized, increased firing pressure, improved thermal management)633
Turbocharger (improved efficiency)101010
EGR Cooler (improved efficiency)222
Water Pump (optimized, variable vane, variable speed)585858
Oil Pump (optimized)333
Fuel Pump (higher working pressure, increased efficiency, improved pressure regulation)333
Fuel Rail (higher working pressure)866
Fuel Injector (optimized, improved multiple event control, higher working pressure)866
Piston (reduced friction skirt, ring and pin)111
Valve train (reduced friction, roller tappet)705252
Model Based Controls292929
Total298275275
Start Printed Page 73565

Table II-13—MY 2024 Vocational Diesel Engine Component Costs Inclusive of Indirect Cost Markups and Adoption Rates

[2013$]

Light HDMedium HDHeavy HD
After-treatment system (improved effectiveness SCR, dosing, DPF)$14$14$14
Valve Actuation160160160
Cylinder Head (flow optimized, increased firing pressure, improved thermal management)1066
Turbocharger (improved efficiency)161616
EGR Cooler (improved efficiency)333
Water Pump (optimized, variable vane, variable speed)818181
Oil Pump (optimized)444
Fuel Pump (higher working pressure, increased efficiency, improved pressure regulation)444
Fuel Rail (higher working pressure)1199
Fuel Injector (optimized, improved multiple event control, higher working pressure)131010
Piston (reduced friction skirt, ring and pin)222
Valve train (reduced friction, roller tappet)977373
Model Based Controls323232
Total446413413

Table II-14—MY 2027 Vocational Diesel Engine Component Costs Inclusive of Indirect Cost Markups and Adoption Rates

[2013$]

Light HDMedium HDHeavy HD
After-treatment system (improved effectiveness SCR, dosing, DPF)$15$15$15
Valve Actuation172172172
Cylinder Head (flow optimized, increased firing pressure, improved thermal management)1066
Turbocharger (improved efficiency)171717
EGR Cooler (improved efficiency)333
Water Pump (optimized, variable vane, variable speed)858585
Oil Pump (optimized)444
Fuel Pump (higher working pressure, increased efficiency, improved pressure regulation)444
Fuel Rail (higher working pressure)1199
Fuel Injector (optimized, improved multiple event control, higher working pressure)141010
Piston (reduced friction skirt, ring and pin)333
Valve train (reduced friction, roller tappet)1027777
Model Based Controls414141
Total481446446

(e) Feasibility of Additional Engine Improvements

While the agencies' technological feasibility analysis for the engine standards focuses on what is achievable for existing engine platforms, we recognize that it could be possible to achieve greater reductions by designing entirely new engine platforms. Unlike existing platforms, which are limited with respect to peak cylinder pressures (precluding certain efficiency improvements), new platforms can be designed to have higher cylinder pressure than today's engines. New designs are also better able to incorporate recent improvements in materials and manufacturing, as well as other technological developments. Considered together, it is likely that a new engine platform could be about 2 percent better than engines using older platforms. Moreover, the agencies have seen CBI data that suggests improvement of more than 3 percent are possible. However, because designing and producing a new engine platform requires hundreds of millions of dollars in capital investment and significant lead time for research and development, it would not be appropriate to project that each engine manufacturer could complete a complete redesign of all of its engines within the Phase 2 time frame. Unlike light-duty, heavy-duty sales volumes are not large enough to support short redesign cycles. As a result, it can take 20 years for a manufacturer to generate the necessary return on the investment associated with an engine redesign. Forcing a manufacturer to redesign its engines prematurely could easily result in significant financial strain on a company.

On the other hand, how far the various manufacturers are into their design cycles suggests that one or more manufacturers will probably introduce a new engine platform during the Phase 2 time frame. This would not enable other engine manufacturers to meet more stringent standards, and thus it would not be an appropriate basis to justify more stringent engine standards (and certainly not engine standards reflecting 100 percent use of technologies premised on existence of new platforms). However, the availability of some more efficient engines on the market will provide the opportunity for vehicle manufacturers to lower their average fuel consumption as measured by GEM. Vehicle manufacturers can use a mix of newer and older engine designs to achieve an average engine performance significantly better than what is required by the engine standards. Thus, the vehicle standards can reflect engine platform improvements (which are amenable to measurement in GEM), without necessarily forcing each manufacturer to achieve these additional reductions, Start Printed Page 73566which may be achievable only for new engine platforms.

As discussed in Section III.D.(1)(b)(i), the agencies project that at least one engine manufacturer (and possibly more) will have completed a redesign for tractor engines by 2027. Accordingly, we project that 50 percent of tractor engines in 2027 will be redesigned engines and be 1.6 percent more efficient than required by the engine standards, so the average engine would be 0.8 percent better. However, we could have projected the same overall improvement by projecting 25 percent of engine getting 3.2 percent better. Based on the CBI information available to us, we believe projecting a 0.8 percent improvement is reasonable, but may be somewhat conservative.

Adding this 0.8 percent improvement to the 5.1 percent reduction required by the standards means we project the average 2027 tractor engine would be 5.9 percent better than Phase 1. Because engine improvements for tractors are applied separately for day cabs and sleeper cabs in the vehicle program, we estimated separate improvements for them here. Specifically, we project a 5.4 percent reduction for day cabs and a 6.4 percent reduction in fuel consumption in sleeper cabs beyond Phase 1. It is important to also note that manufacturers that do not achieve this level would be able to make up for the difference by applying one of the many other tractor vehicle technologies to a greater extent than we project, or to achieve greater reductions by optimizing technology efficiency further. We are not including the cost of developing these new engines in our cost analysis because we believe these engines are going to be developed due to market forces (i.e., the new platform, already contemplated) rather than due to this rulemaking.

We are making a similar new engine platform projection for vocational vehicles. This is because many of tractor and vocational engines, such as HHD, would likely share the same engine hardware with the exception of WHR. In addition, the model based control discussed in Chapter 2.3 of the RIA could integrate engines better with transmissions on the vehicle side. We believe manufacturers will first focus their efforts on improving tractor engines but still believe that the 2027 vocational engine will be significantly better than required by the engine standards.

(3) EPA Engine Standards for N2 O

EPA will continue to apply the Phase 1 N2 O engine standard of 0.10 g/bhp-hr and a 0.02 g/bhp-hr default deterioration factor to the Phase 2 program. EPA adopted the cap standard for N2 O as an engine-based standard because the agency believes that emissions of this GHG are technologically related solely to the engine, fuel, and emissions after-treatment systems, and the agency is not aware of any influence of vehicle-based technologies on these emissions. Note that NHTSA did not adopt standards for N2 O because these emissions do not impact fuel consumption in a significant way.

In the proposal we considered reducing both the standard and deterioration factor to 0.05 and 0.01 g/bhp-hr respectively because engines certified in model year 2014 were generally meeting the proposed standard. We also explained the process behind N2 O formation in urea SCR after-treatment systems and how that process could be optimized to elicit additional N2 O reductions. 80 FR 40203. While we have seen some reductions and a few increases in engine family certified N2 O levels across the 2014, 2015, and 2016 model years, the majority have remained unchanged.

While we still believe that further optimization of SCR systems is possible to reduce N2 O emissions, as demonstrated for some engine families, we do not know to what extent further optimization can be achieved given the tradeoffs required to meet the Phase 2 CO2 standards. These tradeoffs potentially include advancing fuel injection timing to reduce CO2 emissions resulting in an increase in NOX emissions at the engine outlet before the after-treatment, increasing the needed NOX reduction efficiency of the SCR system. We will continue to assess N2 O emissions as SCR technology evolves and CO2 emission reductions phase in, and we will revisit the standard at a later date to further control N2 O emission. This will likely be included in the upcoming rule to consider more stringent NOX standards.

Start Printed Page 73567

Start Printed Page 73568

(4) EPA Engine Standards for Methane

EPA will continue to apply the Phase 1 methane engine standards to the Phase 2 program. EPA adopted the cap standards for CH4 (along with N2 O standards) as engine-based standards because the agency believes that emissions of this GHG are technologically related solely to the engine, fuel, and emissions after-treatment systems, and the agency is not aware of any influence of vehicle-based technologies on these emissions. We are applying these cap standards against the FTP duty-cycle because the FTP cycle is the most stringent with respect to emissions of these pollutants and we do not believe that a reduction is stringency from the current Phase 1 standards is warranted. Note that NHTSA did not adopt standards for CH4 (or N2 O) because these emissions do not impact fuel consumption in a significant way.

EPA continues to believe that manufacturers of most engine technologies will be able to comply with the Phase 1 CH4 standard with no technological improvements. We note that we are not aware of any new technologies that would have allowed us to adopt more stringent standards at this time.

(5) Compliance Provisions and Flexibilities for Engine Standards

The agencies are continuing most of the Phase 1 compliance provisions and flexibilities for the Phase 2 engine standards.

(a) Averaging, Banking, and Trading

The agencies' general approach to averaging is discussed in Section I. We did not propose to offer any new or special credits to engine manufacturers to comply with any of the separate engine standards. Except for early credits, the agencies are retaining all Phase 1 credit flexibilities and limitations to continue for use in the Phase 2 engine program.

As discussed below and as proposed, EPA is changing the useful life for LHD engines for GHG emissions from the current 10 years/110,000 miles to 15 years/150,000 miles to be consistent with the useful life of criteria pollutants recently updated in EPA's Tier 3 rule. In order to ensure that banked credits maintain their value in the transition from Phase 1 to Phase 2, EPA and NHTSA are adopting the proposed adjustment factor of 1.36 (i.e., 150,000 mile ÷ 110,000 miles) for credits that are carried forward from Phase 1 to the MY 2021 and later Phase 2 standards. Without this adjustment factor the change in useful life would have effectively resulted in a discount of banked credits that are carried forward from Phase 1 to Phase 2, which is not the intent of the change in the useful life. See Sections V and VI for additional discussion of similar adjustments of vehicle-based credits.

Finally, the agencies are limiting the carryover of certain Phase 1 engine credits into the Phase 2 program. As described in Section II.D.(2) the agencies made adjustments to the FTP baselines, to address the unexpected step-change improvement in engine fuel consumption and CO2 emissions. The underlying reasons for this shift are mostly related to manufacturers optimizing their SCR thermal management strategy over the FTP in ways that we (mistakenly) thought they already had in MY 2010 (i.e., the Phase 1 baseline). At the time of Phase 1 we had not realized that these improvements were not already in the Phase 1 baseline. This issue does not apply for SET emissions, and thus only significantly impacts engines certified Start Printed Page 73569exclusively to the FTP standards (rather than both FTP and SET standards). To prevent manufacturers from diluting the Phase 2 engine program with credits generated relative to this incorrect baseline, we are not allowing engine credits generated against the Phase 1 FTP standards to be carried over into the Phase 2 program.

(b) Changing Global Warming Potential (GWP) Values in the Credit Program for CH4 and N2 O

The Phase 1 rule included a compliance flexibility that allowed heavy-duty manufacturers and conversion companies to comply with the respective methane or nitrous oxide standards by means of over-complying with CO2 standards (40 CFR 1036.705(d)). The heavy-duty rules allow averaging only between vehicles or engines of the same designated type (referred to as an “averaging set” in the rules). Specifically, the Phase 1 heavy-duty rulemaking added a CO2 credits program which allowed heavy-duty engine manufacturers to average and bank emission credits to comply with the methane and nitrous oxide requirements after adjusting the CO2 emission credits based on the relative GWP equivalents. To establish the GWP equivalents used by the CO2 credits program, the Phase 1 rule incorporated the IPCC Fourth Assessment Report GWP values of 25 for CH4 and 298 for N2 O, which are assessed over a 100 year lifetime.

EPA will continue this provision for Phase 2. However, since the Phase 1 rule was finalized, a new IPCC report has been released (the Fifth Assessment Report), with new GWP estimates. This caused us to look again at the relative GWP equivalency of methane and nitrous oxide and to seek comment on whether the methane and nitrous oxide GWPs used to establish the equivalency value for the CO2 Credit program should be updated to those established by IPCC in its Fifth Assessment Report. 80 FR 40206. The Fifth Assessment Report provides four 100 year GWP values for methane ranging from 28 to 36 and two 100 year GWP values for nitrous oxide, either 265 or 298.

EPA is updating the GWP value to convert CO2 credits for use against the methane standard. We are using a GWP of 34 for the value of methane reductions relative to CO2 reductions. (The GWP remains 298 for N2 O). The use of this new methane GWP will not begin until MY 2021, when the Phase 2 engine standards begin. This provides sufficient lead time for both the agencies and manufacturers to update systems, and also ensures that manufacturers would be able make any necessary design changes. The choice of when to commence use of this GWP value for our engines standards does not prejudice the choice of other GWP values for use in regulations and other purposes in the near term. Further discussion is found in Section XI.D.2.a.

(c) In-Use Compliance and Useful Life

Consistent with section 202(a)(1) and 202(d) of the CAA, for Phase 1, EPA established in-use standards for heavy-duty engines. Based on our assessment of testing variability and other relevant factors, we established in-use standards by adding a 3 percent adjustment factor to the full useful life CO2 emissions and fuel consumption results measured in the EPA certification process to address measurement variability inherent in comparing results among different laboratories and different engines. See 40 CFR part 1036. The agencies are not changing this for Phase 2 SET and FTP engine standard compliance.

In Phase 1, EPA set the useful life for engines and vehicles with respect to GHG emissions equal to the respective useful life periods for criteria pollutants. In April 2014, as part of the Tier 3 light-duty vehicle final rule, EPA extended the regulatory useful life period for criteria pollutants to 150,000 miles or 15 years, whichever comes first, for Class 2b and 3 pickup trucks and vans and some light-duty trucks (79 FR 23414, April 28, 2014). As proposed, EPA is applying the same useful life of 150,000 miles or 15 years for the Phase 2 GHG standards for engines primarily intended for use in vocational vehicles with a GVWR at or below 19,500 lbs. NHTSA will use the same useful life values as EPA for all heavy-duty vehicles.

As proposed, we will continue the regulatory allowance in 40 CFR 1036.150(g) that allows engine manufacturers to use assigned deterioration factors (DFs) for most engines without performing their own durability emission tests or engineering analysis. However, the engines will still be required to meet the standards in actual use without regard to whether the manufacturer used the assigned DFs. This allowance is being continued as an interim provision and may be discontinued for later phases of standards as more information becomes known. Manufacturers are allowed to use an assigned additive DF of 0.0 g/bhp-hr for CO2 emissions from any conventional engine (i.e., an engine not including advanced or off-cycle technologies). Upon request, we could allow the assigned DF for CO2 emissions from engines including advanced or off-cycle technologies, but only if we determine that it would be consistent with good engineering judgment. We believe that we have enough information about in-use CO2 emissions from conventional engines to conclude that they will not increase as the engines age. However, we lack such information about the more advanced technologies. For technologies such as WHR that are considered advanced in the context of Phase 1, but would be treated as a more ordinary technology by the end of Phase 2, we plan to work with manufacturers to determine if using the assigned zero DF would be appropriate.

(d) Alternate CO2 Standards

In the Phase 1 rulemaking, the agencies allowed certification to alternate CO2 engine standards in model years 2014 through 2016. This flexibility was intended to address the special case of needed lead time to implement new standards for a previously unregulated pollutant. Since that special case does not apply for Phase 2, we are not adopting a similar flexibility in this rulemaking.

(e) Approach to Standards and Compliance Provisions for Natural Gas Engines

EPA is also making certain clarifying changes to its rules regarding classification of natural gas engines. This relates to standards for all emissions, both greenhouse gases and criteria pollutants. These clarifying changes are intended to reflect the status quo, and therefore should not have any associated costs.

EPA emission standards have always applied differently for gasoline-fueled and diesel-fueled engines. The regulations in 40 CFR part 86 implement these distinctions by dividing engines into Otto-cycle and Diesel-cycle technologies. This approach led EPA to categorize natural gas engines according to their design history. A diesel engine converted to run on natural gas was classified as a diesel-cycle engine; a gasoline engine converted to run on natural gas was classified as an Otto-cycle engine.

The Phase 1 rule described our plan to transition to a different approach, consistent with EPA's non-road programs, in which we divide engines into compression-ignition and spark-ignition technologies based only on the thermodynamic operating characteristics of the engines.[193] However, the Phase 1 rule included a provision allowing us to continue with Start Printed Page 73570the historic approach on an interim basis.

Under the existing EPA regulatory definitions of “compression-ignition” and “spark-ignition,” a natural gas engine would generally be considered compression-ignition if it operates with lean air-fuel mixtures and uses a pilot injection of diesel fuel to initiate combustion, and would generally be considered spark-ignition if it operates with stoichiometric air-fuel mixtures and uses a spark plug to initiate combustion.

EPA's basic premise here is that natural gas engines performing similar in-use functions as diesel engines should be subject to similar regulatory requirements. The compression-ignition emission standards and testing requirements reflect the operating characteristics for the full range of heavy-duty vehicles, including substantial operation in long-haul service characteristic of tractors. The spark-ignition emission standards and testing requirements do not include some of those provisions related to use in long-haul service or other applications where diesel engines predominate, such as steady-state testing, Not-to-Exceed standards, and extended useful life. We believe it would be inappropriate to apply the spark-ignition standards and requirements to natural gas engines that are being used in applications mostly served by diesel engines today. We therefore proposed to replace the interim provision described above with a differentiated approach to certification of natural gas engines across all of the EPA standards—for both GHGs and criteria pollutants. 80 FR 40207. Under the proposed amendment, we would require manufacturers to divide all their natural gas engines into primary intended service classes, as we already require for compression-ignition engines, whether or not the engine has features that otherwise could (in theory) result in classification as SI under the current rules. We proposed that any natural gas engine qualifying as a medium heavy-duty engine (19,500 to 33,000 lbs. GVWR) or a heavy heavy-duty engine (over 33,000 lbs. GVWR) would be subject to all the emission standards and other requirements that apply to compression-ignition engines. However, based on comments, we are finalizing this change only for heavy heavy-duty engines. Commenters identified medium heavy-duty applications in which SI alternative fuel engines compete significantly with gasoline engines, which is not consistent with the premise of the proposal. Thus, we are not finalizing the proposed change for medium heavy-duty engines.

Table II-15 describes the provisions that apply differently for compression-ignition and spark-ignition engines:

Table II-15—Regulatory Provisions That Are Different for Compression-Ignition and Spark-Ignition Engines

ProvisionCompression-ignitionSpark-ignition
Transient duty cycle40 CFR part 86, Appendix I, paragraph (f)(2) cycle; divide by 1.12 to de-normalize40 CFR part 86, Appendix I, paragraph (f)(1) cycle.
Ramped-modal test (SET)yesno.
NTE standardsyesno.
Smoke standardyesno.
Manufacturer-run in-use testingyesno.
ABT—pollutantsNOX, PMNOX, NMHC.
ABT—transient conversion factor6.56.3.
ABT—averaging setSeparate averaging sets for light, medium, and heavy HDDEOne averaging set for all SI engines.
Useful life110,000 miles for light HDDE, a 185,000 miles for medium HDDE, 435,000 miles for heavy HDDE110,000 miles. a
Warranty50,000 miles for light HDDE, 100,000 miles for medium HDDE, 100,000 miles for heavy HDDE50,000 miles.
Detailed AECD descriptionyesno.
Test engine selectionhighest injected fuel volumemost likely to exceed emission standards.
Note:
a As proposed, useful life for light heavy-duty diesel and spark ignition engines is being increased to 150,000 miles for GHG emissions, but remains at 110,000 for criteria pollutant emissions.

The onboard diagnostic requirements already differentiate requirements by fuel type, so there is no need for those provisions to change based on the considerations of this section.

We are not aware of any currently certified engines that will change from compression-ignition to spark-ignition under this approach. Nonetheless, because these proposed changes could result in a change in standards for engines currently under development, we believe it is appropriate to provide additional lead time. We will therefore continue to apply the existing interim provision through model year 2020.[194] Starting in model year 2021, all the provisions will apply as described above for heavy heavy-duty engines. Manufacturers will not be permitted to certify any engine families using carryover emission data if a particular engine model switched from compression-ignition to spark-ignition, or vice versa. However, as noted above, in practice these vehicles are already being certified as CI engines, so we view these changes as clarifications ratifying the current status quo.

These provisions will apply equally to engines fueled by any fuel other than gasoline or ethanol, should such engines be produced in the future. Given the current and historic market for vehicles above 33,000 lbs. GVWR, the agencies believe any alternative-fueled vehicles in this weight range will be competing primarily with diesel vehicles and should be subject to the same requirements as them. See Sections XI and XII for additional discussion of natural gas fueled engines.Start Printed Page 73571

(f) Crankcase Emissions From Natural Gas Engines

EPA proposed to require that all natural gas-fueled engines have closed crankcases, rather than continuing the provision that allows venting to the atmosphere all crankcase emissions from all compression-ignition engines. 80 FR 40208. However, EPA is not finalizing the proposed requirement at this time.

Open crankcases have been allowed as long as these vented crankcase emissions are measured and accounted for as part of an engine's tailpipe emissions. This allowance has historically been in place to address the technical limitations related to recirculating diesel-fueled engines' crankcase emissions, which have high PM emissions, back into the engine's air intake. High PM emissions vented into the intake of an engine can foul turbocharger compressors and after cooler heat exchangers. In contrast, historically EPA has mandated closed crankcase technology on all gasoline fueled engines and all natural gas spark-ignition engines.[195] The inherently low PM emissions from these engines posed no technical barrier to a closed crankcase mandate. However, after considering the comments on this issue, we now believe that there are practical reasons why we should not close natural gas crankcases without also requiring closed crankcases for other compression-ignition engines. Because current natural gas engines are generally produced from diesel engine designs that are not designed to operate with closed crankcases, we have concerns that sealing the crankcase on the natural gas versions will require substantial development effort, and the seals may not function properly. Thus, we expect to update our regulations for crankcase emissions from all compression ignition engines at the same time in a future rulemaking.

(g) Compliance Margins

Some commenters suggested that the agencies should apply a compliance margin to confirmatory and SEA test results to account for variability of engine maps and emission tests. However, EPA's past practice has been to base the standards on technology projections that assume manufacturers will apply compliance margins to their test results for certification. In other words, they design their products to have emissions below the standards by some small margin so that test-to-test or lab-to-lab variability would not cause them to exceed any applicable standards. Consequently, EPA has typically not set standards precisely at the lowest levels achievable, but rather at slightly higher levels—expecting manufacturers to target the lower levels to provide compliance margins for themselves. The agencies have applied this approach to the Phase 2 standards. Thus, the feasibility and cost analyses reflect the expectation that manufacturers will target lower values to provide compliance margins.

The agencies have also improved the engine test procedures and compliance provisions to reduce the agencies' and the manufacturers' uncertainty of engine test results. For example, in the agencies' confirmatory test procedures we are requiring that the agencies use the average of at least three tests (i.e., the arithmetic mean of a sample size of at least three test results) for determining the values of confirmatory test results for any GEM engine fuel maps. We are only doing this for GEM engine fuel maps because these are relatively new tests, compared to Phase 1 testing or EPA's other emissions standards. Therefore, this provision does not apply to any other emissions testing. For all other emissions testing besides GEM engine fuel maps the agencies' maintain our usual convention of utilizing a sample size of one for confirmatory testing. For GEM engine fuel mapping this at least triples the test burden for the agencies to conduct confirmatory testing, but it also decreases confirmatory test result uncertainty by at least 42 percent.[196] Based on improvements like this one, and others described in Section 1.4 of the RTC, we believe that SET, FTP and GEM's steady-state, cycle-average and powertrain test results will have an overall uncertainty of +/−1.0 percent. To further protect against falsely high emissions results or false failures due to this remaining level of test procedure uncertainty, we have included a +1 percent compliance margin into our stringency analyses of the engine standards and the GEM fuel map inputs used to determine the tractor and vocational vehicle standards. In other words we set Phase 2 engine and vehicle standards 1 percent less stringent than if we had not considered this test procedure uncertainty.

In addition to the test procedure improvements and the +1 percent margin we incorporated into our standards, the agencies are also committed to a process of continuous improvement of test procedures to further reduce test result uncertainty. To contribute to this effort, in mid-2016 EPA committed $250,000 to fund research to further evaluate individual sources of engine mapping test procedure uncertainty. This work will occur at SwRI. Should the results of this work or other similar future work indicate test procedure improvements that would further reduce test result uncertainty, the agencies will incorporate these improvements through appropriate guidance or through technical amendments to the regulations via a notice and comment rulemaking. If we determine in the future through the SwRI work or other work that such improvements eliminate the need to require the agencies to conduct triplicate confirmatory testing of GEM engine fuel maps, we will promulgate technical amendments to the regulations to remove this requirement. If we determine in the future through the SwRI work or other work that the +1.0 percent we factored into our stringency analysis was inappropriately low or high, we will promulgate technical amendments to the regulations to address any inappropriate impact this +1.0 percent had on the stringency of the engine and vehicle standards.[197] In addition, whenever the agencies deter