Skip to Content
Rule

Regulation of Fuels and Fuel Additives: Changes to Renewable Fuel Standard Program

Action

Final Rule.

Summary

Under the Clean Air Act Section 211(o), as amended by the Energy Independence and Security Act of 2007 (EISA), the Environmental Protection Agency is required to promulgate regulations implementing changes to the Renewable Fuel Standard program. The revised statutory requirements specify the volumes of cellulosic biofuel, biomass-based diesel, advanced biofuel, and total renewable fuel that must be used in transportation fuel. This action finalizes the regulations that implement the requirements of EISA, including the cellulosic, biomass-based diesel, advanced biofuel, and renewable fuel standards that will apply to all gasoline and diesel produced or imported in 2010. The final regulations make a number of changes to the current Renewable Fuel Standard program while retaining many elements of the compliance and trading system already in place. This final rule also implements the revised statutory definitions and criteria, most notably the new greenhouse gas emission thresholds for renewable fuels and new limits on renewable biomass feedstocks. This rulemaking marks the first time that greenhouse gas emission performance is being applied in a regulatory context for a nationwide program. As mandated by the statute, our greenhouse gas emission assessments consider the full lifecycle emission impacts of fuel production from both direct and indirect emissions, including significant emissions from land use changes. In carrying out our lifecycle analysis we have taken steps to ensure that the lifecycle estimates are based on the latest and most up-to-date science. The lifecycle greenhouse gas assessments reflected in this rulemaking represent significant improvements in analysis based on information and data received since the proposal. However, we also recognize that lifecycle GHG assessment of biofuels is an evolving discipline and will continue to revisit our lifecycle analyses in the future as new information becomes available. EPA plans to ask the National Academy of Sciences for assistance as we move forward. Based on current analyses we have determined that ethanol from corn starch will be able to comply with the required greenhouse gas (GHG) threshold for renewable fuel. Similarly, biodiesel can be produced to comply with the 50% threshold for biomass-based diesel, sugarcane with the 50% threshold for advanced biofuel and multiple cellulosic-based fuels with their 60% threshold. Additional fuel pathways have also been determined to comply with their thresholds. The assessment for this rulemaking also indicates the increased use of renewable fuels will have important environmental, energy and economic impacts for our Nation.

 

Table of Contents Back to Top

Tables Back to Top

DATES: Back to Top

This final rule is effective on July 1, 2010, and the percentage standards apply to all gasoline and diesel produced or imported in 2010. The incorporation by reference of certain publications listed in the rule is approved by the Director of the Federal Register as of July 1, 2010.

ADDRESSES: Back to Top

EPA has established a docket for this action under Docket ID No. EPA-HQ-OAR-2005-0161. All documents in the docket are listed in the http://www.regulations.gov Web site. Although listed in the index, some information is not publicly available, e.g., confidential business information (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 through http://www.regulations.gov or in hard copy at the Air and Radiation Docket and Information Center, EPA/DC, EPA West, Room 3334, 1301 Constitution Ave., NW., Washington, DC. The Public Reading Room is open from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal holidays. The telephone number for the Public Reading Room is (202) 566-1744, and the telephone number for the Air Docket is (202) 566-1742.

FOR FURTHER INFORMATION CONTACT: Back to Top

Julia MacAllister, Office of Transportation and Air Quality, Assessment and Standards Division, Environmental Protection Agency, 2000 Traverwood Drive, Ann Arbor, MI 48105; Telephone number: 734-214-4131; Fax number: 734-214-4816; E-mail address: macallister.julia@epa.gov, or Assessment and Standards Division Hotline; telephone number (734) 214-4636; E-mail address asdinfo@epa.gov.

SUPPLEMENTARY INFORMATION: Back to Top

General Information Back to Top

I. Does This Final Rule Apply to Me?

Entities potentially affected by this final rule are those involved with the production, distribution, and sale of transportation fuels, including gasoline and diesel fuel or renewable fuels such as ethanol and biodiesel. Regulated categories include:

Category NAICS1codes SIC2codes Examples of potentially regulated entities
1North American Industry Classification System (NAICS)
2Standard Industrial Classification (SIC) system code.
Industry 324110 2911 Petroleum Refineries.
Industry 325193 2869 Ethyl alcohol manufacturing.
Industry 325199 2869 Other basic organic chemical manufacturing.
Industry 424690 5169 Chemical and allied products merchant wholesalers.
Industry 424710 5171 Petroleum bulk stations and terminals.
Industry 424720 5172 Petroleum and petroleum products merchant wholesalers.
Industry 454319 5989 Other fuel dealers

This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely to be regulated by this final action. This table lists the types of entities that EPA is now aware could potentially be regulated by this final action. Other types of entities not listed in the table could also be regulated. To determine whether your activities would be regulated by this final action, you should carefully examine the applicability criteria in 40 CFR part 80. If you have any questions regarding the applicability of this final action to a particular entity, consult the person listed in the preceding section.

Outline of This Preamble Back to Top

I. Executive Summary

A. Summary of New Provisions of the RFS Program

1. Required Volumes of Renewable Fuel

2. Standards for 2010 and Effective Date for New Requirements

a. 2010 Standards

b. Effective Date

3. Analysis of Lifecycle Greenhouse Gas Emissions and Thresholds for Renewable Fuels

a. Background and Conclusions

b. Fuel Pathways Considered and Key Model Updates Since the Proposal

c. Consideration of Fuel Pathways Not Yet Modeled

4. Compliance with Renewable Biomass Provision

5. EPA-Moderated Transaction System

6. Other Changes to the RFS Program

B. Impacts of Increasing Volume Requirements in the RFS2 Program

II. Description of the Regulatory Provisions

A. Renewable Identification Numbers (RINs)

B. New Eligibility Requirements for Renewable Fuels

1. Changes in Renewable Fuel Definitions

a. Renewable Fuel

b. Advanced Biofuel

c. Cellulosic Biofuel

d. Biomass-Based Diesel

e. Additional Renewable Fuel

f. Cellulosic Diesel

2. Lifecycle GHG Thresholds

3. Renewable Fuel Exempt From 20 Percent GHG Threshold

a. General Background of the Exemption Requirement

b. Definition of Commenced Construction

c. Definition of Facility Boundary

d. Proposed Approaches and Consideration of Comments

i. Comments on the Proposed Basic Approach

ii. Comments on the Expiration of Grandfathered Status

e. Final Grandfathering Provisions

i. Increases in Volume of Renewable Fuel Produced at Grandfathered Facilities Due to Expansion

ii. Replacements of Equipment

iii. Registration, Recordkeeping and Reporting

4. New Renewable Biomass Definition and Land Restrictions

a. Definitions of Terms

i. Planted Crops and Crop Residue

ii. Planted Trees and Tree Residue

iii. Slash and Pre-Commercial Thinnings

iv. Biomass Obtained From Certain Areas at Risk From Wildfire

v. Algae

b. Implementation of Renewable Biomass Requirements

i. Ensuring That RINs Are Generated Only For Fuels Made From Renewable Biomass

ii. Whether RINs Must Be Generated For All Qualifying Renewable Fuel

c. Implementation Approaches for Domestic Renewable Fuel

i. Recordkeeping and Reporting for Feedstocks

ii. Approaches for Foreign Producers of Renewable Fuel

(1) RIN-Generating importers

(2) RIN-Generating foreign producers

iii. Aggregate Compliance Approach for Planted Crops and Crop Residue From Agricultural Land

(1) Analysis of Total Agricultural Land in 2007

(2) Aggregate Agricultural Land Trends Over Time

(3) Aggregate Compliance Determination

d. Treatment of Municipal Solid Waste (MSW)

C. Expanded Registration Process for Producers and Importers

1. Domestic Renewable Fuel Producers

2. Foreign Renewable Fuel Producers

3. Renewable Fuel Importers

4. Process and Timing

D. Generation of RINs

1. Equivalence Values

2. Fuel Pathways and Assignment of D Codes

a. Producers

b. Importers

c. Additional Provisions for Foreign Producers

3. Facilities With Multiple Applicable Pathways

4. Facilities That Co-Process Renewable Biomass and Fossil Fuels

5. Facilities That Process Municipal Solid Waste

6. RINless Biofuel

E. Applicable Standards

1. Calculation of Standards

a. How Are the Standards Calculated?

b. Standards for 2010

2. Treatment of Biomass-Based Diesel in 2009 and 2010

a. Shift in 2009 Biomass-Based Diesel Compliance Demonstration to 2010

b. Treatment of Deficit Carryovers, RIN Rollover, and RIN Valid Life For Adjusted 2010 Biomass-Based Diesel Requirement

3. Future Standards

F. Fuels That Are Subject to the Standards

1. Gasoline

2. Diesel

3. Other Transportation Fuels

G. Renewable Volume Obligations (RVOs)

1. Designation of Obligated Parties

2. Determination of RVOs Corresponding to the Four Standards

3. RINs Eligible To Meet Each RVO

4. Treatment of RFS1 RINs Under RFS2

a. Use of RFS1 RINs To Meet Standards Under RFS2

b. Deficit Carryovers From the RFS1 Program to RFS2

H. Separation of RINs

1. Nonroad

2. Heating Oil and Jet Fuel

3. Exporters

4. Requirement to Transfer RINs With Volume

5. Neat Renewable Fuel and Renewable Fuel Blends Designated as Transportation Fuel, Heating Oil, or Jet Fuel

I. Treatment of Cellulosic Biofuel

1. Cellulosic Biofuel Standard

2. EPA Cellulosic Biofuel Waiver Credits for Cellulosic Biofuel

3. Application of Cellulosic Biofuel Waiver Credits

J. Changes to Recordkeeping and Reporting Requirements

1. Recordkeeping

2. Reporting

3. Additional Requirements for Producers of Renewable Natural Gas, Electricity, and Propane

4. Attest Engagements

K. Production Outlook Reports

L. What Acts Are Prohibited and Who Is Liable for Violations?

III. Other Program Changes

A. The EPA Moderated Transaction System (EMTS)

1. Need for the EPA Moderated Transaction System

2. Implementation of the EPA Moderated Transaction System

3. How EMTS Will Work

4. A Sample EMTS Transaction

B. Upward Delegation of RIN-Separating Responsibilities

C. Small Producer Exemption

D. 20% Rollover Cap

E. Small Refinery and Small Refiner Flexibilities

1. Background—RFS1

a. Small Refinery Exemption

b. Small Refiner Exemption

2. Statutory Options for Extending Relief

3. The DOE Study/DOE Study Results

4. Ability To Grant Relief Beyond 211(o)(9)

5. Congress-Requested Revised DOE Study

6. What We're Finalizing

a. Small Refinery and Small Refiner Temporary Exemptions

b. Case-by-Case Hardship for Small Refineries and Small Refiners

c. Program Review

7. Other Flexibilities Considered for Small Refiners

a. Extensions of the RFS1 Temporary Exemption for Small Refiners

b. Phase-in

c. RIN-Related Flexibilities

F. Retail Dispenser Labeling for Gasoline With Greater Than 10 Percent Ethanol

G. Biodiesel Temperature Standardization

IV. Renewable Fuel Production and Use

A. Overview of Renewable Fuel Volumes

1. Reference Cases

2. Primary Control Case

a. Cellulosic Biofuel

b. Biomass-Based Diesel

c. Other Advanced Biofuel

d. Other Renewable Fuel

3. Additional Control Cases Considered

B. Renewable Fuel Production

1. Corn/Starch Ethanol

a. Historic/Current Production

b. Forecasted Production Under RFS2

2. Imported Ethanol

3. Cellulosic Biofuel

a. Current State of the Industry

b. Setting the 2010 Cellulosic Biofuel Standard

c. Current Production Outlook for 2011 and Beyond

d. Feedstock Availability

i. Urban Waste

ii. Agricultural and Forestry Residues

iii. Dedicated Energy Crops

iv. Summary of Cellulosic Feedstocks for 2022

4. Biodiesel & Renewable Diesel

a. Historic and Projected Production

i. Biodiesel

ii. Renewable Diesel

b. Feedstock Availability

C. Biofuel Distribution

1. Biofuel Shipment to Petroleum Terminals

2. Petroleum Terminal Accommodations

3. Potential Need for Special Blendstocks at Petroleum Terminals for E85

4. Need for Additional E85 Retail Facilities

D. Ethanol Consumption

1. Historic/Current Ethanol Consumption

2. Increased Ethanol Use Under RFS2

a. Projected Gasoline Energy Demand

b. Projected Growth in Flexible Fuel Vehicles

c. Projected Growth in E85 Access

d. Required Increase in E85 Refueling Rates

e. Market Pricing of E85 Versus Gasoline

3. Consideration of >10% Ethanol Blends

V. Lifecycle Analysis of Greenhouse Gas Emissions

A. Introduction

1. Open and Science-Based Approach to EPA's Analysis

2. Addressing Uncertainty

B. Methodology

1. Scope of Analysis

a. Inclusion of Indirect Land Use Change

b. Models Used

c. Scenarios Modeled

2. Biofuel Modeling Framework & Methodology for Lifecycle Analysis Components

a. Feedstock Production

i. Domestic Agricultural Sector Impacts

ii. International Agricultural Sector Impacts

b. Land Use Change

i. Amount of Land Area Converted and Where

ii. Type of Land Converted

iii. GHG Emissions Associated With Conversion

(1) Domestic Emissions

(2) International Emissions

iv. Timeframe of Emission Analysis

v. GTAP and Other Models

c. Feedstock Transport

d. Biofuel Processing

e. Fuel Transportation

f. Vehicle Tailpipe Emissions

3. Petroleum Baseline

C. Threshold Determination and Assignment of Pathways

D. Total GHG Reductions

E. Effects of GHG Emission Reductions and Changes in Global Temperature and Sea Level

VI. How Would the Proposal Impact Criteria and Toxic Pollutant Emissions and Their Associated Effects?

A. Overview of Impacts

B. Fuel Production & Distribution Impacts of the Proposed Program

C. Vehicle and Equipment Emission Impacts of Fuel Program

D. Air Quality Impacts

1. Particulate Matter

a. Current Levels

b. Projected Levels Without RFS2 Volumes

c. Projected Levels With RFS2 Volumes

2. Ozone

a. Current Levels

b. Projected Levels Without RFS2 Volumes

c. Projected Levels With RFS2 Volumes

3. Air Toxics

a. Current Levels

b. Projected Levels

i. Acetaldehyde

ii. Formaldehyde

iii. Ethanol

iv. Benzene

v. 1,3-Butadiene

vi. Acrolein

vii. Population Metrics

4. Nitrogen and Sulfur Deposition

a. Current Levels

b. Projected Levels

E. Health Effects of Criteria and Air Toxics Pollutants

1. Particulate Matter

a. Background

b. Health Effects of PM

2. Ozone

a. Background

b. Health Effects of Ozone

3. NO X and SO X

a. Background

b. Health Effects of NO X

c. Health Effects of SO X

4. Carbon Monoxide

5. Air Toxics

a. Acetaldehyde

b. Acrolein

c. Benzene

d. 1,3-Butadiene

e. Ethanol

f. Formaldehyde

g. Peroxyacetyl Nitrate (PAN)

h. Naphthalene

i. Other Air Toxics

F. Environmental Effects of Criteria and Air Toxic Pollutants

1. Visibility

2. Atmospheric Deposition

3. Plant and Ecosystem Effects of Ozone

4. Environmental Effects of Air Toxics

VII. Impacts on Cost of Renewable Fuels, Gasoline, and Diesel

A. Renewable Fuel Production Costs

1. Ethanol Production Costs

a. Corn Ethanol

b. Cellulosic Ethanol

i. Feedstock Costs

ii. Production Costs for Cellulosic Biofuels

c. Imported Sugarcane Ethanol

2. Biodiesel and Renewable Diesel Production Costs

a. Biodiesel

b. Renewable Diesel

B. Biofuel Distribution Costs

1. Ethanol Distribution Costs

2. Cellulosic Distillate and Renewable Diesel Distribution Costs

3. Biodiesel Distribution Costs

C. Reduced U.S. Refining Demand

D. Total Estimated Cost Impacts

1. Refinery Modeling Methodology

2. Overall Impact on Fuel Cost

VIII. Economic Impacts and Benefits

A. Agricultural and Forestry Impacts

1. Biofuel Volumes Modeled

2. Commodity Price Changes

3. Impacts on U.S. Farm Income

4. Commodity Use Changes

5. U.S. Land Use Changes

6. Impact on U.S. Food Prices

7. International Impacts

B. Energy Security Impacts

1. Implications of Reduced Petroleum Use on U.S. Imports

2. Energy Security Implications

a. Effect of Oil Use on Long-Run Oil Price, U.S. Import Costs, and Economic Output

b. Short-Run Disruption Premium From Expected Costs of Sudden Supply Disruptions

c. Costs of Existing U.S. Energy Security Policies

3. Combining Energy Security and Other Benefits

4. Total Energy Security Benefits

C. Benefits of Reducing GHG Emissions

1. Introduction

2. Derivation of Interim Social Cost of Carbon Values

3. Application of Interim SCC Estimates to GHG Emissions Reductions

D. Criteria Pollutant Health and Environmental Impacts

1. Overview

2. Quantified Human Health Impacts

3. Monetized Impacts

4. What Are the Limitations of the Health Impacts Analysis?

E. Summary of Costs and Benefits

IX. Impacts on Water

A. Background

1. Agriculture and Water Quality

2. Ecological Impacts

3. Impacts to the Gulf of Mexico

B. Upper Mississippi River Basin Analysis

1. SWAT Model

2. AEO 2007 Reference Case

3. Reference Cases and RFS2 Control Case

4. Case Study

5. Sensitivity Analysis

C. Additional Water Issues

1. Chesapeake Bay Watershed

2. Ethanol Production and Distribution

a. Production

b. Distillers Grain With Solubles

c. Ethanol Leaks and Spills From Fueling Stations

3. Biodiesel Plants

4. Water Quantity

5. Drinking Water

X. Public Participation

XI. Statutory and Executive Order Reviews

A. Executive Order 12866: Regulatory Planning and Review

B. Paperwork Reduction Act

C. Regulatory Flexibility Act

1. Overview

2. Background

3. Summary of Potentially Affected Small Entities

4. Reporting, Recordkeeping, and Compliance

5. Related Federal Rules

6. Steps Taken To Minimize the Significant Economic Impact on Small Entities

a. Significant Panel Findings

b. Outreach With Small Entities (and the Panel Process)

c. Panel Recommendations, Proposed Provisions, and Provisions Being Finalized

i. Delay in Standards

ii. Phase-in

iii. RIN-Related Flexibilities

iv. Program Review

v. Extensions of the Temporary Exemption Based on a Study of Small Refinery Impacts

vi. Extensions of the Temporary Exemption Based on Disproportionate Economic Hardship

7. Conclusions

D. Unfunded Mandates Reform Act

E. Executive Order 13132: Federalism

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

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

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

I. National Technology Transfer Advancement Act

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

K. Congressional Review Act

XII. Statutory Provisions and Legal Authority

I. Executive Summary Back to Top

Through this final rule, the U.S. Environmental Protection Agency is revising the National Renewable Fuel Standard program to implement the requirements of the Energy Independence and Security Act of 2007 (EISA). EISA made significant changes to both the structure and the magnitude of the renewable fuel program created by the Energy Policy Act of 2005 (EPAct). The EISA fuel program, hereafter referred to as RFS2, mandates the use of 36 billion gallons of renewable fuel by 2022—a nearly five-fold increase over the highest volume specified by EPAct. EISA also established four separate categories of renewable fuels, each with a separate volume mandate and each with a specific lifecycle greenhouse gas emission threshold. The categories are renewable fuel, advanced biofuel, biomass-based diesel, and cellulosic biofuel. There is a notable increase in the mandate for cellulosic biofuels in particular. EISA increased the cellulosic biofuel mandate to 16 billion gallons by 2022, representing the bulk of the increase in the renewable fuels mandate.

EPA's proposed rule sought comment on a multitude of issues, ranging from how to interpret the new definitions for renewable biomass to the Agency's proposed methodology for conducting the greenhouse gas lifecycle assessments required by EISA. The decisions presented in this final rule are heavily informed by the many public comments we received on the proposed rule. In addition, and as with the proposal, we sought input from a wide variety of stakeholders. The Agency has had multiple meetings and discussions with renewable fuel producers, technology companies, petroleum refiners and importers, agricultural associations, lifecycle experts, environmental groups, vehicle manufacturers, states, gasoline and petroleum marketers, pipeline owners and fuel terminal operators. We also have worked closely with other Federal agencies and in particular with the Departments of Energy and Agriculture.

This section provides an executive summary of the final RFS2 program requirements that EPA is implementing as a result of EISA. The RFS2 program will replace the RFS1 program promulgated on May 1, 2007 (72 FR 23900). [1] Details of the final requirements can be found in Sections II and III, with certain lifecycle aspects detailed in Section V.

This section also provides a summary of EPA's assessment of the environmental and economic impacts of the use of higher renewable fuel volumes. Details of these analyses can be found in Sections IV through IX and in the Regulatory Impact Analysis (RIA).

A. Summary of New Provisions of the RFS Program

Today's notice establishes new regulatory requirements for the RFS program that will be implemented through a new subpart M to 40 CFR part 80. EPA is maintaining several elements of the RFS1 program such as regulations governing the generation, transfer, and use of Renewable Identification Numbers (RINs). At the same time, we are making a number of updates to reflect the changes brought about by EISA

1. Required Volumes of Renewable Fuel

The RFS program is intended to require a minimum volume of renewable fuel to be used each year in the transportation sector. In response to EPAct 2005, under RFS1 the required volume was 4.0 billion gallons in 2006, ramping up to 7.5 billion gallons by 2012. Starting in 2013, the program also required that the total volume of renewable fuel contain at least 250 million gallons of fuel derived from cellulosic biomass.

In response to EISA, today's action makes four primary changes to the volume requirements of the RFS program. First, it substantially increases the required volumes and extends the timeframe over which the volumes ramp up through at least 2022. Second, it divides the total renewable fuel requirement into four separate categories, each with its own volume requirement. Third, it requires, with certain exceptions applicable to existing facilities, that each of these mandated volumes of renewable fuels achieve certain minimum thresholds of GHG emission performance. Fourth, it requires that all renewable fuel be made from feedstocks that meet the new definition of renewable biomass including certain land use restrictions. The volume requirements in EISA are shown in Table I.A.1-1.

BILLING CODE 6560-50-P

BILLING CODE 6560-50-C

As shown in the table, the volume requirements are not exclusive, and generally result in nested requirements. Any renewable fuel that meets the requirement for cellulosic biofuel or biomass-based diesel is also valid for meeting the advanced biofuel requirement. Likewise, any renewable fuel that meets the requirement for advanced biofuel is also valid for meeting the total renewable fuel requirement. See Section V.C for further discussion of which specific types of fuel may qualify for the four categories shown in Table I.A.1-1.

2. Standards for 2010 and Effective Date for New Requirements

While EISA established the renewable fuel volumes shown in Table I.A.1-1, it also requires that the Administrator set the standards based on these volumes each November for the following year based in part on information provided from the Energy Information Agency (EIA). In the case of the cellulosic biofuel standard, section 211(o)(7)(D) of EISA specifically requires that the standard be set based on the volume projected to be available during the following year. If the volume is lower than the level shown in Table I.A.1-1, then EISA allows the Administrator to also lower the advanced biofuel and total renewable fuel standards each year accordingly. Given the implications of these standards and the necessary judgment that can't be reduced to a formula akin to the RFS1 regulations, we believe it is appropriate to set the standards through a notice-and-comment rulemaking process. Thus, for future standards, we intend to issue an NPRM by summer and a final rule by November 30 of each year in order to determine the appropriate standards applicable in the following year. However, in the case of the 2010 standards, we are finalizing them as part of today's action.

a. 2010 Standards

While we proposed that the cellulosic biofuel standard would be set at the EISA-specified level of 100 million gallons for 2010, based on analysis of information available at this time, we no longer believe the full volume can be met. Since the proposal, we have had detailed discussions with over 30 companies that are in the business of developing cellulosic biofuels and cellulosic biofuel technology. Based on these discussions, we have found that many of the projects that served as the basis for the proposal have been put on hold, delayed, or scaled back. At the same time, there have been a number of additional projects that have developed and are moving forward. As discussed in Section IV.B.3, the timing for many of the projects indicates that while few will be able to provide commercial volumes for 2010, an increasing number will come on line in 2011, 2012, and 2013. The success of these projects is then expected to accelerate growth of the cellulosic biofuel industry out into the future. EIA provided us with a projection on October 29, 2009 of 5.04 million gallons (6.5 million ethanol-equivalent gallons) of cellulosic biofuel production for 2010. While our company-by-company assessment varies from EIA's, as described in Section IV.B.3., and actual cellulosic production volume during 2010 will be a function of developments over the course of 2010, we nevertheless believe that 5 million gallons (6.5 million ethanol equivalent) represents a reasonable, yet achievable level for the cellulosic standard for 2010. While this is lower than the level specified in EISA, no change to the advanced biofuel and total renewable fuel standards is warranted. With the inclusion of an energy-based Equivalence Value for biodiesel and renewable diesel, 2010 compliance with the biomass-based diesel standard will be more than enough to ensure compliance with the advanced biofuel standard for 2010.

Today's rule also includes special provisions to account for the 2009 biomass-based diesel volume requirements in EISA. As described in the NPRM, in November 2008 we used the new total renewable fuel volume of 11.1 billion gallons from EISA as the basis for the 2009 total renewable fuel standard that we issued under the RFS1 regulations. [2] While this approach ensured that the total mandated renewable fuel volume required by EISA for 2009 was used, the RFS1 regulatory structure did not provide a mechanism for implementing the 0.5 billion gallon requirement for biomass-based diesel nor the 0.6 billion gallon requirement for advanced biofuel. As we proposed, and as is described in more detail in Section II.E.2, we are addressing this issue in today's rule by combining the 2010 biomass-based diesel requirement of 0.65 billion gallons with the 2009 biomass based diesel requirement of 0.5 billion gallons to require that obligated parties meet a combined 2009/2010 requirement of 1.15 billion gallons by the end of the 2010 compliance year. No similar provisions are required in order to fulfill the 2009 advanced biofuel volume mandate.

The resulting 2010 standards are shown in Table I.A.2-1. These standards represent the fraction of a refiner's or importer's gasoline and diesel volume which must be renewable fuel. Additional discussion of the 2010 standards can be found in Section II.E.1.b.

Table I.A.2-1—Standards for 2010 Back to Top
Cellulosic biofuel 0.004%
Biomass-based diesel 1.10%
Advanced biofuel 0.61%
Renewable fuel 8.25%

b. Effective Date

Under CAA section 211(o) as modified by EISA, EPA is required to revise the RFS1 regulations within one year of enactment, or December 19, 2008. Promulgation by this date would have been consistent with the revised volume requirements shown in Table I.A.1-1 that begin in 2009 for certain categories of renewable fuel. As described in the NPRM, we were not able to promulgate final RFS2 program requirements by December 19, 2008.

Under today's rule, the transition from using the RFS1 regulatory provisions regarding registration, RIN generation, reporting, and recordkeeping to using comparable provisions in this RFS2 rule will occur on July 1, 2010. This is the start of the 1st quarter following completion of the statutorily required 60-day Congressional Review period for such a rulemaking as this. This will provide adequate lead time for all parties to transition to the new regulatory requirements, including additional time to prepare for RFS2 implementation for those entities who may find it helpful, especially those covered by the RFS program for the first time. In addition, making the transition at the end of the quarter will help simplify the recordkeeping and reporting transition to RFS2. To facilitate the volume obligations being based on the full year's gasoline and diesel production, and to enable the smooth transition from the RFS1 to RFS2 regulatory provisions, Renewable Identification Numbers (RINs—which are used in the program for both credit trading and for compliance demonstration) that were generated under the RFS1 regulations will continue to be valid for compliance with the RFS2 obligations. Further discussion of transition issues can be found in Sections II.A and II.G.4, respectively.

According to EISA, the renewable fuel obligations applicable under RFS2 apply on a calendar basis. That is, obligated parties must determine their renewable volume obligations (RVOs) at the end of a calendar year based on the volume of gasoline or diesel fuel they produce during the year, and they must demonstrate compliance with their RVOs in an annual report that is due two months after the end of the calendar year.

For 2010, today's rule will follow this same general approach. The four RFS2 RVOs for each obligated party will be calculated on the basis of all gasoline and diesel produced or imported on and after January 1, 2010, through December 31, 2010. Obligated parties will be required to demonstrate by February 28 of 2011 that they obtained sufficient RINs to satisfy their 2010 RVOs. We believe this is an appropriate approach as it is more consistent with Congress' provisions in EISA for 2010, and there is adequate lead time for the obligated parties to achieve compliance.

The issue for EPA to resolve is how to apply the four volume mandates under EISA for calendar year 2010. These volume mandates are translated into applicable percentages that obligated parties then use to determine their renewable fuel volume obligations based on the gasoline and diesel they produce or import in 2010. There are three basic approaches that EPA has considered, based on comments on the proposal. The first is the approach adopted in this rule—the four RFS2 applicable percentages are determined based on the four volume mandates covered by this rule, and the renewable volume obligation for a refiner or importer will be determined by applying these percentages to the volume of gasoline and diesel fuel they produce during calendar year 2010. Under this approach, there is no separate applicable percentage under RFS1 for 2010, however RINs generated in 2009 and 2010 under RFS1 can be used to meet the four volume obligations for 2010 under the RFS2 regulations. Another option, which was considered and rejected by EPA, is much more complicated—(1) determine an RFS1 applicable percentage based on just the total renewable fuel volume mandate, using the same total volume for renewable fuel as used in the first approach, and require obligated parties to apply that percentage to the gasoline produced from January 1, 2010 until the effective date of the RFS2 regulations, and (2) determine the four RFS2 applicable percentages as discussed above, but require obligated parties to apply them to only the gasoline and diesel in 2010 after the effective date of the RFS2 regulations. Of greater concern than its complexity, the second approach fails to ensure that the total volumes for three of the volume mandates are met for 2010. In effect EPA would be requiring that obligated parties use enough cellulosic biofuel, biomass-based diesel, and advanced biofuel to meet approximately 75% of the total volumes required for these fuels under EISA. While the total volume mandate under EISA for renewable fuel would likely be met, the other three volumes mandates would only be met in part. The final option would involve delaying the RFS2 requirements until January 1, 2011, which would avoid the complexity of the second approach, but would be even less consistent with EISA's requirements.

The approach adopted in this rule is clearly the most consistent with EISA's requirement of four different volume mandates for all of calendar year 2010. In addition, EPA is confident that obligated parties have adequate lead-time to comply with the four volume requirements under the approach adopted in this rule. The volume requirements are achieved by obtaining the appropriate number of RINs from producers of the renewable fuel. The obligated parties do not need lead time for construction or investment purposes, as they are not changing the way they produce gasoline or diesel, do not need to design to install new equipment, or take other actions that require longer lead time. Obtaining the appropriate amount of RINs involves contractual or other arrangements with renewable fuel producers or other holders of RINs. Obligated parties now have experience implementing RFS1, and the actions needed to comply under the RFS2 regulations are a continuation of these kinds of RFS1 activities. In addition, an adequate supply of RINs is expected to be available for compliance by obligated parties. RFS1 RINs have been produced throughout 2009 and continue to be produced since the beginning of 2010. There has been and will be no gap or lag in the production of RINS, as the RFS1 regulations continue in effect and require that renewable fuel producers generate RINs for the renewable fuel they produce. These 2009 and 2010 RFS1 RINs will be available and can be used towards the volume requirements of obligated parties for 2010. These RFS1 RINS combined with the RFS2 RINs that will be generated by renewable fuel producers are expected to provide an adequate supply of RINs to ensure compliance for all of the renewable volume mandates. For further discussion of the expected supply of renewable fuel, see section IV.

In addition, obligated parties have received adequate notice of this obligation. The proposed rule called for obligated parties to meet the full volume mandates for all four volume mandates, and to base their volume obligation on the volume of gasoline and diesel produced starting January 1, 2010. While the RFS2 regulations are not effective until after January 1, 2010, the same full year approach is being taken for the 2010 volumes of gasoline and diesel. Obligated parties have been on notice based on EPA's proposal, discussions with many stakeholders during the rulemaking, the issuance of the final rule itself, and publication of this rule in the Federal Register. As discussed above, there is adequate time for obligated parties to meet their 2010 volume obligations by the spring of 2011.

This approach does not impose any retroactive requirements. The obligation that is imposed under the RFS2 regulations is forward looking—by the spring of 2011, when compliance is determined, obligated parties must satisfy certain volume obligations. These future requirements are calculated in part based on volumes of gasoline and diesel produced prior to the effective date of the RFS2 regulations, but this does not make the RFS2 requirement retroactive in nature. The RFS2 regulations do not change in any way the legal obligations or requirements that apply prior to the effective date of the RFS2 regulations. Instead, the RFS2 requirements impose new requirements that must be met in the future. There is adequate lead time to comply with these RFS2 requirements, and they achieve a result that is more consistent with Congress' goals in establishing 4 volume mandates for calendar year 2010, and for these reasons EPA is adopting this approach for calendar year 2010.

Parties that intend to generate RINs, own and/or transfer them, or use them for compliance purposes after July 1, 2010 will need to register or re-register under the RFS2 provisions and modify their information technology (IT) systems to accommodate the changes we are finalizing today. As described more fully in Section II, these changes include redefining the D code within the RIN that identifies which standard a fuel qualifies for, adding a process for verifying that feedstocks meet the renewable biomass definition, and calculating compliance with four standards instead of one. EPA's registration system is available now for parties to complete the registration process. Further details on this process can be found elsewhere in today's preamble as well as at http://www.epa.gov/otaq/regs/fuels/ fuelsregistration.htm. Parties that produce motor vehicle, nonroad, locomotive, and marine (MVNRLM) diesel fuel but not gasoline will be newly obligated parties and may be establishing IT systems for the RFS program for the first time.

3. Analysis of Lifecycle Greenhouse Gas Emissions and Thresholds for Renewable Fuels

a. Background and Conclusions

A significant aspect of the RFS2 program is the requirement that the lifecycle GHG emissions of a qualifying renewable fuel must be less than the lifecycle GHG emissions of the 2005 baseline average gasoline or diesel fuel that it replaces; four different levels of reductions are required for the four different renewable fuel standards. These lifecycle performance improvement thresholds are listed in Table I.A.3-1. Compliance with each threshold requires a comprehensive evaluation of renewable fuels, as well as the baseline for gasoline and diesel, on the basis of their lifecycle emissions. As mandated by EISA, the greenhouse gas emissions assessments must evaluate the aggregate quantity of greenhouse gas emissions (including direct emissions and significant indirect emissions such as significant emissions form land use changes) related to the full lifecycle, including all stages of fuel and feedstock production, distribution and use by the ultimate consumer.

Table I.A.3-1—Lifecycle GHG Thresholds Specified in EISA Back to Top
[Percent Reduction from Baseline]
aThe 20% criterion generally applies to renewable fuel from new facilities that commenced construction after December 19, 2007.
Renewable fuela 20
Advanced biofuel 50
Biomass-based diesel 50
Cellulosic biofuel 60

It is important to recognize that fuel from the existing capacity of current facilities and the capacity of all new facilities that commenced construction prior to December 19, 2007 (and in some cases prior to December 31, 2009) are exempt, or grandfathered, from the 20% lifecycle requirement for the Renewable Fuel category. Therefore, EPA has in the discussion below emphasized its analysis on those plants and fuels that are likely to be used for compliance with the rule and would be subject to the lifecycle thresholds. Based on the analyses and approach described in Section V of this preamble, EPA is determining that ethanol produced from corn starch at a new facility (or expanded capacity from an existing) using natural gas, biomass or biogas for process energy and using advanced efficient technologies that we expect will be most typical of new production facilities will meet the 20% GHG emission reduction threshold compared to the 2005 baseline gasoline. We are also determining that biobutanol from corn starch meets the 20% threshold. Similarly, EPA is making the determination that biodiesel and renewable diesel from soy oil or waste oils, fats and greases will exceed the 50% GHG threshold for biomass-based diesel compared to the 2005 petroleum diesel baseline. In addition, we have now modeled biodiesel and renewable diesel produced from algal oils as complying with the 50% threshold for biomass-based diesel. EPA is also determining that ethanol from sugarcane complies with the applicable 50% GHG reduction threshold for advanced biofuels. The modeled pathways (feedstock and production technology) for cellulosic ethanol and cellulosic diesel would also comply with the 60% GHG reduction threshold applicable to cellulosic biofuels. As discussed later in section V, there are also other fuels and fuel pathways that we are determining will comply with the GHG thresholds.

Under EISA, EPA is allowed to adjust the GHG reduction thresholds downward by up to 10% if necessary based on lifecycle GHG assessment of biofuels likely to be available. Based on the results summarized above, we are not finalizing any adjustments to the lifecycle GHG thresholds for the four renewable fuel standard categories.

EPA recognizes that as the state of scientific knowledge continues to evolve in this area, the lifecycle GHG assessments for a variety of fuel pathways are likely to be updated. Therefore, while EPA is using its current lifecycle assessments to inform the regulatory determinations for fuel pathways in this final rule, as required by the statute, the Agency is also committing to further reassess these determinations and lifecycle estimates. As part of this ongoing effort, we will ask for the expert advice of the National Academy of Sciences, as well as other experts, and incorporate their advice and any updated information we receive into a new assessment of the lifecycle GHG emissions performance of the biofuels being evaluated in this final rule. EPA will request that the National Academy of Sciences evaluate the approach taken in this rule, the underlying science of lifecycle assessment, and in particular indirect land use change, and make recommendations for subsequent lifecycle GHG assessments on this subject. At this time we are estimating this review by the National Academy of Sciences may take up to two years. As specified by EISA, if EPA revises the analytical methodology for determining lifecycle greenhouse gas emissions, any such revision will apply to renewable fuel from new facilities that commence construction after the effective date of the revision.

b. Fuel Pathways Considered and Key Model Updates Since the Proposal

EPA is making the GHG threshold determination based on a methodology that includes an analysis of the full lifecycle, including significant emissions related to international land-use change. As described in more detail below and in Section V of this preamble, EPA has used the best available models for this purpose, and has incorporated many modifications to its proposed approach based on comments from the public and peer reviewers and developing science. EPA has also quantified the uncertainty associated with significant components of its analyses, including important factors affecting GHG emissions associated with international land use change. As discussed below, EPA has updated and refined its modeling approach since proposal in several important ways, and EPA is confident that its modeling of GHG emissions associated with international land use is comprehensive and provides a reasonable and scientifically robust basis for making the threshold determinations described above. As discussed below, EPA plans to continue to improve upon its analyses, and will update it in the future as appropriate.

Through technical outreach, the peer review process, and the public comment period, EPA received and reviewed a significant amount of data, studies, and information on our proposed lifecycle analysis approach. We incorporated a number of new, updated, and peer-reviewed data sources in our final rulemaking analysis including better satellite data for tracking land use changes and improved assessments of N2O impacts from agriculture. The new and updated data sources are discussed further in this section, and in more detail in Section V.

We also performed dozens of new modeling runs, uncertainty analyses, and sensitivity analyses which are leading to greater confidence in our results. We have updated our analyses in conjunction with, and based on, advice from experts from government, academia, industry, and not for profit institutions.

The new studies, data, and analysis performed for the final rulemaking impacted the lifecycle GHG results for biofuels in a number of different ways. In some cases, updates caused the modeled analysis of lifecycle GHG emissions from biofuels to increase, while other updates caused the modeled emissions to be reduced. Overall, the revisions since our proposed rule have led to a reduction in modeled lifecycle GHG emissions as compared to the values in the proposal. The following highlights the most significant revisions. Section V details all of the changes made and their relative impacts on the results.

Corn Ethanol: The final rule analysis found less overall indirect land use change (less land needed), thereby improving the lifecycle GHG performance of corn ethanol. The main reasons for this decrease are:

  • Based on new studies that show the rate of improvement in crop yields as a function of price, crop yields are now modeled to increase in response to higher crop prices. When higher crop yields are used in the models, less land is needed domestically and globally for crops as biofuels expand.
  • New research available since the proposal indicates that the corn ethanol production co-product, distillers grains and solubles (DGS), is more efficient as an animal feed (meaning less corn is needed for animal feed) than we had assumed in the proposal. Therefore, in our analyses for the final rule, domestic corn exports are not impacted as much by increased biofuel production as they were in the proposal analysis.
  • Improved satellite data allowed us to more finely assess the types of land converted when international land use changes occur, and this more precise assessment led to a lowering of modeled GHG impacts. Based on previous satellite data, the proposal assumed cropland expansion onto grassland would require an amount of pasture to be replaced through deforestation. For the final rulemaking analysis we incorporated improved economic modeling of demand for pasture area and satellite data which indicates that pasture is also likely to expand onto existing grasslands. This reduced the GHG emissions associated with an amount of land use change.

However, we note that not all modeling updates necessarily reduced predicted GHG emissions from land use change. As one example, since the proposal a new version of the GREET model (Version 1.8C) has been released. EPA reviewed the new version and concluded that this was an improvement over the previous GREET release that was used in the proposal analysis (Version 1.8B). Therefore, EPA updated the GHG emission factors for fertilizer production used in our analysis to the values from the new GREET version. This had the result of slightly increasing the GHG emissions associated with fertilizer production and thus slightly increasing the GHG emission impacts of domestic agriculture.

For the final rule, EPA has analyzed a variety of corn ethanol pathways including ethanol made from corn starch using natural gas, coal, and biomass as process energy sources in production facilities utilizing both dry mill and wet mill processes. For corn starch ethanol, we also considered the technology enhancements likely to occur in the future such as the addition of corn oil fractionation or extraction technology, membrane separation technology, combined heat and power and raw starch hydrolysis.

Biobutanol from corn starch: In addition to ethanol from corn starch, for this final rule, we have also analyzed bio-butanol from corn starch. Since the feedstock impacts are the same as for ethanol from corn starch, the assessment for biobutanol reflects the differing impacts due to the production process and energy content of biobutanol compared to that of ethanol.

Soybean Biodiesel: The new information described above for corn ethanol also leads to lower modeled GHG impacts associated with soybean biodiesel. The revised assessment predicts less overall indirect land use change (less land needed) and less impact from the land use changed that does occur (due to updates in types of converted land assumed). In addition, the latest IPCC guidance indicates reduced domestic soybean N2O emissions, and updated USDA and industry data show reductions in biodiesel processing energy use and a higher co-product credit, all of which further reduced the modeled soybean biodiesel lifecycle GHG emissions. This has resulted in a significant improvement in our assessment of the lifecycle performance of soybean biodiesel as compared to the estimate in the proposal.

Biodiesel and Renewable Diesel from Algal Oil and Waste Fats and Greases: In addition to biodiesel from soy oil, biodiesel and renewable diesel from algal oil (should it reach commercial production) and biodiesel from waste oils, fats and greases have been modeled. These feedstock sources have little or no land use impact so the GHG impacts associate with their use in biofuel production are largely the result of energy required to produce the feedstock (in the case of algal oil) and the energy required to turn that feedstock into a biofuel.

Sugarcane Ethanol: Sugarcane ethanol was analyzed considering a range of technologies and assuming alternative pathways for dehydrating the ethanol prior to its use as a biofuel in the U.S. For the final rule, our analysis also shows less overall indirect land use change (less land needed) associated with sugarcane ethanol production. For the proposal, we assumed sugarcane expansion in Brazil would result in cropland expansion into grassland and lost pasture being replaced through deforestation. Based on newly available regional specific data from Brazil, historic trends, and higher resolution satellite data, in the final rule, sugarcane expansion onto grassland is coupled with greater pasture intensification, such that there is less projected impact on forests. Furthermore, new data provided by commenters showed reduced sugarcane ethanol process energy, which also reduced the estimated lifecycle GHG impact of sugarcane ethanol production.

Cellulosic Ethanol: We analyzed cellulosic ethanol production using both biochemical (enzymatic) and thermo-chemical processes with corn stover, switchgrass, and forestry thinnings and waste as feedstocks. For cellulosic diesel, we analyzed production using the Fischer-Tropsch process. For the final rule, we updated the cellulosic ethanol conversion rates based on new data provided by the National Renewable Energy Laboratory (NREL.) As a result of this update, the gallons per ton yields for switchgrass and several other feedstock sources increased in our analysis for the final rule, while the predicted yields from corn residue and several other feedstock sources decreased slightly from the NPRM values. In addition, we also updated our feedstock production yields based on new work conducted by the Pacific Northwest National Laboratory (PNNL). This analysis increased the tons per acre yields for several dedicated energy crops. These updates increased the amount of cellulosic ethanol projected to come from energy crops. While the increase in crop yields and conversion efficiency reduced the GHG emissions associated with cellulosic ethanol, there remains an increased demand for land to grow dedicated energy crops; this land use impact resulted in increased GHG emissions with the net result varying by the type of cellulosic feedstock source.

We note that several of the renewable fuel pathways modeled are still in early stages of development or commercialization and are likely to continue to develop as the industry moves toward commercial production. Therefore, it will be necessary to reanalyze several pathways using updated data and information as the technologies develop. For example, biofuel derived from algae is undergoing wide ranging development. Therefore for now, our algae analyses presume particular processes and energy requirements which will need to be reviewed and updated as this fuel source moves toward commercial production.

For this final rule we have incorporated a statistical analysis of uncertainty about critical variables in our pathway analysis. This uncertainty analysis is explained in detail in Section V and is consistent with the specific recommendations received through our peer review and public comments on the proposal. The uncertainty analysis focused on two aspects of indirect land use change—the types of land converted and the GHG emission associated with different types of land converted. In particular, our uncertainty analysis focused on such specific sources of information as the satellite imaging used to inform our assessment of land use trends and the specific changes in carbon storage expected from a change in land use in each geographic area of the world modeled. We have also performed additional sensitivity analyses including analysis of two yield scenarios for corn and soy beans to assess the impact of changes in yield assumptions.

This uncertainty analysis provides information on both the range of possible outcomes for the parameters analyzed, an estimate of the degree of confidence that the actual result will be within a particular range (in our case, we estimated a 95% confidence interval) and an estimate of the central tendency or midpoint of the GHG performance estimate.

In the proposal, we considered several options for the timeframe over which to measure lifecycle GHG impacts and the possibility of discounting those impacts. Based on peer review recommendations and other comments received, EPA is finalizing its assessments based on an analysis assuming 30 years of continued emission impacts after the program is fully phased in by 2022 and without discounting those impacts.

EPA also notes that it received significant comment on our proposed baseline lifecycle greenhouse gas assessment of gasoline and diesel (“petroleum baseline”). While EPA has made several updates to the petroleum analysis in response to comments (see Section V for further discussion), we are finalizing the approach based on our interpretation of the definition in the Act as requiring that the petroleum baseline represent an average of the gasoline and diesel fuel (whichever is being replaced by the renewable fuel) sold as transportation fuel in 2005.

As discussed in more detail later, the modeling results developed for purposes of the final rule provide a rich and comprehensive base of information for making the threshold determinations. There are numerous modeling runs, reflecting updated inputs to the model, sensitivity analyses, and uncertainty analyses. The results for different scenarios include a range and a best estimate or mid-point. Given the potentially conservative nature of the base crop yield assumption, EPA believes the actual crop yield in 2022 may be above the base yield; however we are not in a position to characterize how much above it might be. To the extent actual yields are higher, the base yield modeling results would underestimate to some degree the actual GHG emissions reductions compared to the baseline.

In making the threshold determinations for this rule, EPA weighed all of the evidence available to it, while placing the greatest weight on the best estimate value for the base yield scenario. In those cases where the best estimate for the base yield scenario exceeds the reduction threshold, EPA judges that there is a good basis to be confident that the threshold will be achieved and is determining that the bio-fuel pathway complies with the applicable threshold. To the extent the midpoint of the scenarios analyzed lies further above a threshold for a particular biofuel pathway, we have increasingly greater confidence that the biofuel exceeds the threshold.

EPA recognizes that certain commenters suggest that there is a very high degree of uncertainty associated in particular with determining international indirect land use changes and their emissions impacts, and because of this EPA should exclude any calculation of international indirect land use changes in its lifecycle analysis. Commenters say EPA should make the threshold determinations based solely on modeling of other sources of lifecycle emissions. In effect, commenters argue that the uncertainty of the modeling associated with international indirect land use change means we should use our modeling results but exclude that part of the results associated with international land use change.

For the reasons discussed above and in more detail in Section V, EPA rejects the view that the modeling relied upon in the final rule, which includes emissions associated with international indirect land use change, is too uncertain to provide a credible and reasonable scientific basis for determining whether the aggregate lifecycle emissions exceed the thresholds. In addition, as discussed elsewhere, the definition of lifecycle emissions includes significant indirect emissions associated with land use change. In deciding whether a bio-fuel pathway meets the threshold, EPA has to consider what it knows about all aspects of the lifecycle emissions, and decide whether there is a valid basis to find that the aggregate lifecycle emissions of the fuel, taking into account significant indirect emissions from land use change meets the threshold. Based on the analyses conducted for this rule, EPA has determined international indirect land use impacts are significant and therefore must be included in threshold compliance assessment.

If the international land use impacts were so uncertain that their impact on lifecycle GHG emissions could not be adequately determined, as claimed by commenters, this does not mean EPA could assume the international land use change emissions are zero, as commenters suggest. High uncertainty would not mean that emissions are small and can be ignored; rather it could mean that we could not tell whether they are large or small. If high uncertainty meant that EPA were not able to determine that indirect emissions from international land use change are small enough that the total lifecycle emissions meet the threshold, then that fuel could not be determined to meet the GHG thresholds of EISA and the fuel would necessarily have to be excluded from the program.

In any case, that is not the situation here as EPA rejects commenters' suggestion and does not agree that the uncertainty over the indirect emissions from land use change is too high to make a reasoned threshold determination. Therefore biofuels with a significant international land use impact are included within this program.

c. Consideration of Fuel Pathways Not Yet Modeled

Not all biofuel pathways have been directly modeled for this rule. For example, while we have modeled cellulosic biofuel produced from corn stover, we have not modeled the specific GHG impact of cellulosic biofuel produced from other crop residues such as wheat straw or rice straw. Today, in addition to finalizing a threshold compliance determination for those pathways we specifically modeled, in some cases, our technical judgment indicates other pathways are likely to be similar enough to modeled pathways that we are also assured these similar pathways qualify. These pathways include fuels produced from the same feedstock and using the same production process but produced in countries other than those modeled. The agricultural sector modeling used for our lifecycle analysis does not predict any soybean biodiesel or corn ethanol will be imported into the U.S., or any imported sugarcane ethanol from production in countries other than Brazil. However, these rules do not prohibit the use in the U.S. of these fuels produced in countries not modeled if they are also expected to comply with the eligibility requirements including meeting the thresholds for GHG performance. Although the GHG emissions of producing these fuels from feedstock grown or biofuel produced in other countries has not been specifically modeled, we do not anticipate their use would impact our conclusions regarding these feedstock pathways. The emissions of producing these fuels in other countries could be slightly higher or lower than what was modeled depending on a number of factors. Our analyses indicate that crop yields for the crops in other countries where these fuels are also most likely to be produced are similar or lower than U.S. values indicating the same or slightly higher GHG impacts. Agricultural sector inputs for the crops in these other countries are roughly the same or lower than the U.S. pointing toward the same or slightly lower GHG impacts. If crop production were to expand due to biofuels in the countries where the models predict these biofuels might additionally be produced would tend to lower our assessment of international indirect impacts but could increase our assessment of the domestic (i.e., the country of origin) land use impacts. EPA believes, because of these offsetting factors along with the small amounts of fuel potentially coming from other countries, that incorporating fuels produced in other countries will not impact our threshold analysis. Therefore, fuels of the same fuel type, produced from the same feedstock using the same fuel production technology as modeled fuel pathways will be assessed the same GHG performance decisions regardless of country of origin. These pathways also include fuels that might be produced from similar feedstock sources to those already modeled and which are expected to have less or no indirect land use change. In such cases, we believe that in order to compete economically in the renewable fuel marketplace such pathways are likely to be at least as energy efficient as those modeled and thus have comparable lifecycle GHG performance. Based on these considerations, we are extending the lifecycle results for the fuel pathways already modeled to 5 broader categories of feedstocks. This extension of lifecycle modeling results is discussed further in Section V.C.

We have established five categories of biofuel feedstock sources under which modeled feedstock sources and feedstock sources similar to those modeled are grouped and qualify on the basis of our existing modeling. These are:

1. Crop residues such as corn stover, wheat straw, rice straw, citrus residue.

2. Forest material including eligible forest thinnings and solid residue remaining from forest product production.

3. Annual cover crops planted on existing crop land such as winter cover crops.

4. Separated food and yard waste including biogenic waste from food processing.

5. Perennial grasses including switchgrass and miscanthus.

The full set of pathways for which we have been able to make a compliance decision are described in Section V.

Threshold determinations for certain other pathways were not possible at this time because sufficient modeling or data is not yet available. In some of these cases, we recognize that a renewable fuel is already being produced from an alternative feedstock. Although we have the data needed for analysis, we did not have sufficient time to complete the necessary lifecycle GHG impact assessment for this final rule. We will model and evaluate additional pathways after this final rule on the basis of current or likely commercial production in the near-term and the status of current analysis at EPA. EPA anticipates modeling grain sorghum ethanol, woody pulp ethanol, and palm oil biodiesel after this final rule and including the determinations in a rulemaking within 6 months. Our analyses project that they will be used in meeting the RFS2 volume standard in the near-term. During the course of the NPRM comment period, EPA received detailed information on these pathways and is currently in the process of analyzing these pathways. We have received comments on several additional feedstock/fuel pathways, including rapeseed/canola, camelina, sweet sorghum, wheat, and mustard seed, and we welcome parties to utilize the petition process described in Section V.C to request EPA to examine additional pathways.

We anticipate there could be additional cases where we currently do not have information on which to base a lifecycle GHG assessment perhaps because we are not yet aware of potential unique plant configurations or operations that could result in greater efficiencies than assumed in our analysis. In many cases, such alternative pathways could have been explicitly modeled as a reasonably straightforward extension of pathways we have modeled if the necessary information had been available. For example, while we have modeled specific enhancements to corn starch ethanol production such as membrane separation or corn oil extraction, there are likely other additional energy saving or co-product pathways available or under development by the industry. It is reasonable to also consider these alternative energy saving or co-product pathways based upon their technical merits. Other current or emerging pathways may require new analysis and modeling for EPA to fully evaluate compliance. For example, fuel pathways with feedstocks or fuel types not yet modeled by EPA may require additional modeling and, it follows, public comment before a determination of compliance can be made.

Therefore, for those fuel pathways that are different than those pathways EPA has listed in today's regulations, EPA is establishing a petition process whereby a party can petition the Agency to consider new pathways for GHG reduction threshold compliance. As described in Section V.C, the petition process is meant for parties with serious intention to move forward with production via the petitioned fuel pathway and who have moved sufficiently forward in the business process to show feasibility of the fuel pathway's implementation. In addition, if the petition addresses a fuel pathway that already has been determined to qualify as one or more types of renewable fuel under RFS (e.g., renewable fuel, or advanced biofuel), the pathway must have the potential to result in qualifying for a renewable fuel type for which it was not previously qualified. Thus, for example, the Agency will not undertake any additional review for a party wishing to get a modified LCA value for a previously approved fuel pathway if the desired new value would not change the overall pathway classification.

The petition must contain all the necessary information on the fuel pathway to allow EPA to effectively assess the lifecycle performance of the new fuel pathway. See Section V.C for a full description. EPA will use the data supplied via the petition and other pertinent data available to the Agency to evaluate whether the information for that fuel pathway, combined with information developed in this rulemaking for other fuel pathways that have been determined to exceed the threshold, is sufficient to allow EPA to evaluate the pathway for a determination of compliance. We expect such a determination would be pathway specific. For some fuel pathways with unique modifications or enhancements to production technologies in pathways otherwise modeled for the regulations listed today, EPA may be able to evaluate the pathway as a reasonably straight-forward extension of our current assessments. In such cases, we would expect to make a decision for that specific pathway without conducting a full rulemaking process. We would expect to evaluate whether the pathway is consistent with the definitions of renewable fuel types in the regulations, generally without going through rulemaking, and issue an approval or disapproval that applies to the petitioner. We anticipate that we will subsequently propose to add the pathway to the regulations. Other current or emerging fuel pathways may require significant new analysis and/or modeling for EPA to conduct an adequate evaluation for a compliance determination (e.g., feedstocks or fuel types not yet included in EPA's assessments for this regulation). For these pathways, EPA would give notice and seek public comment on a compliance determination under the annual rulemaking process established in today's regulations. If we make a technical determination of compliance, then we anticipate the fuel producer will be able to generate RINs for fuel produced under the additional pathway following the next available quarterly update of the EPA Moderated Transaction System (EMTS). EPA will process those petitions as expeditiously as possible for those pathways which are closer to the commercial production stage than others. In all events, parties are expected to begin this process with ample lead time as compared to their commercial start dates. Further discussion of this petition process can be found in Section V.C.

We note again that the continued work of EPA and others is expected to result in improved models and data sources, and that re-analysis based on such updated information could revise these determinations. Any such reassessment that would impact compliance would necessarily go through rulemaking and would only be applicable to production from future facilities after the revised rule was finalized, as required by EISA.

4. Compliance With Renewable Biomass Provision

EISA changed the definition of “renewable fuel” to require that it be made from feedstocks that qualify as “renewable biomass.” EISA's definition of the term “renewable biomass” limits the types of biomass as well as the types of land from which the biomass may be harvested. The definition includes:

  • Planted crops and crop residue from agricultural land cleared prior to December 19, 2007 and actively managed or fallow on that date.
  • Planted trees and tree residue from tree plantations cleared prior to December 19, 2007 and actively managed on that date.
  • Animal waste material and byproducts.
  • Slash and pre-commercial thinnings from non-federal forestlands that are neither old-growth nor listed as critically imperiled or rare by a State Natural Heritage program.
  • Biomass cleared from the vicinity of buildings and other areas at risk of wildfire.
  • Algae.
  • Separated yard waste and food waste.

In today's rule, EPA is finalizing definitions for the many terms included within the definition of renewable biomass. Where possible, EPA has adhered to existing statutory, regulatory or industry definitions for these terms, although in some cases we have altered definitions to conform to EISA's statutory language, to further the goals of EISA, or for ease of program implementation. For example, EPA is defining “agricultural land” from which crops and crop residue can be harvested for RIN-generating renewable fuel production as including cropland, pastureland, and land enrolled in the Conservation Reserve Program. An in-depth discussion of the renewable biomass definitions can be found in Section II.B.4.

In keeping with EISA, under today's final rule, renewable fuel producers may only generate RINs for fuels made from feedstocks meeting the definition of renewable biomass. In order to implement this requirement, we are finalizing three potential mechanisms for domestic and foreign renewable fuel producers to verify that their feedstocks comply with this requirement. The first involves renewable biomass recordkeeping and reporting requirements by renewable fuel producers for their individual facilities. As an alternative to these individual recordkeeping and reporting requirements, the second allows renewable fuel producers to form a consortium to fund an independent third-party to conduct an annual renewable biomass quality-assurance survey, based on a plan approved by EPA. The third is an aggregate compliance approach applicable only to crops and crop residue from the U.S. It utilizes USDA's publicly available agricultural land data as the basis for an EPA determination of compliance with the renewable biomass requirements for these particular feedstocks. This determination will be reviewed annually, and if EPA finds it is no longer warranted, then renewable fuel producers using domestically grown crops and crop residue will be required to conduct individual or consortium-based verification processes to ensure that their feedstocks qualify as renewable biomass. These final provisions are described below, with a more in-depth discussion in Section II.B.4.

For renewable fuel producers using feedstocks other than planted crops or crop residue from agricultural land that do not choose to participate in the third-party survey funded by an industry consortium, the final renewable biomass recordkeeping and reporting provisions require that individual producers obtain documentation about their feedstocks from their feedstock supplier(s) and take the measures necessary to ensure that they know the source of their feedstocks and can demonstrate to EPA that they have complied with the EISA definition of renewable biomass. Specifically, EPA's renewable biomass reporting requirements for producers who generate RINs include a certification on renewable fuel production reports that the feedstock used for each renewable fuel batch meets the definition of renewable biomass. Additionally, producers will be required to include with their quarterly reports a summary of the types and volumes of feedstocks used throughout the quarter, as well as maps of the land from which the feedstocks used in the quarter were harvested. EPA's final renewable biomass recordkeeping provisions require renewable fuel producers to maintain sufficient records to support their claims that their feedstocks meet the definition of renewable biomass, including maps or electronic data identifying the boundaries of the land where the feedstocks were produced, documents tracing the feedstocks from the land to the renewable fuel production facility, other written records from their feedstock suppliers that serve as evidence that the feedstock qualifies as renewable biomass, and for producers using planted trees or tree residue from tree plantations, written records that serve as evidence that the land from which the feedstocks were obtained was cleared prior to December 19, 2007 and actively managed on that date.

Based on USDA's publicly available agricultural land data, EPA is able to establish a baseline of the aggregate amount of U.S. agricultural land (meaning cropland, pastureland and CRP land in the United States) that is available for the production of crops and crop residues for use in renewable fuel production consistent with the definition of renewable biomass. EPA has determined that, in the aggregate this amount of agricultural land (land cleared or cultivated prior to EISA's enactment (December 19, 2007) and actively managed or fallow, and nonforested on that date) is expected to, at least in the near term, be sufficient to support EISA renewable fuel obligations and other foreseeable demands for crop products, without clearing and cultivating additional land. EPA also believes that economic factors will lead farmers to use the “agricultural land” available for crop production under EISA rather than bring new land into crop production. As a result, EPA is deeming renewable fuel producers using domestically-grown crops and crop residue as feedstock to be in compliance with the renewable biomass requirements, and those producers need not comply with the recordkeeping and quarterly reporting requirements as established for the non-crop-based biomass sector. However, EPA will annually review USDA data on lands in agricultural production to determine if these conclusions remain valid. If EPA determines that the 2007 baseline amount of eligible agricultural land has been exceeded, EPA will publish a notice of that finding in the Federal Register. At that point, renewable fuel producers using planted crops or crop residue from agricultural lands would be subject to the same recordkeeping and reporting requirements as other renewable fuel producers.

5. EPA-Moderated Transaction System

We introduced the EPA Moderated Transaction System (EMTS) in the NPRM as a new method for managing the generation of RINs and transactions involving RINs. EMTS is designed to resolve the RIN management issues of RFS1 that lead to widespread RIN errors, many times resulting in invalid RINs and often tedious remedial procedures to resolve those errors. It is also designed to address the added RIN categories, more complex RIN generation requirements, and additional volume of RINs associated with RFS2. Commenters broadly support EMTS and most stated that its use should coincide with the start of RFS2; however, many commenters expressed concerns over having sufficient time to implement the new system. In today's action, we are requiring the use of EMTS for all RFS2 RIN generations and transactions beginning July 1, 2010. EPA has utilized an open process for the development of EMTS since it was first introduced in the NPRM, conducting workshops and webinars, and soliciting stakeholder participation in its evaluation and testing. EPA pledges to work with the regulated community, as a group and individually, to ensure EMTS is successfully implemented. EPA anticipates that with this level of assistance, regulated parties will not experience significant difficulties in transitioning to the new system, and EPA believes that the many benefits of the new system warrant its immediate use.

6. Other Changes to the RFS Program

Today's final rule also makes a number of other changes to the RFS program that are described in more detail in Sections II and III below, including:

  • Grandfathering provisions: Renewable fuel from existing facilities is exempt from the lifecycle GHG emission reduction threshold of 20% up to a baseline volume for that facility that will be established at the time of registration. As discussed in Section II.B.3, the exemption from the 20% GHG threshold applies only to renewable fuel that is produced from facilities which commenced construction on or before December 19, 2007, or in the case of ethanol plants that use natural gas or biodiesel for process heat, on or before December 31, 2009.
  • Renewable fuels produced from municipal solid waste (MSW): The new renewable biomass definition in EISA modified the ability for MSW-derived fuels to qualify under the RFS program by restricting it to “separated yard waste or food waste.” We are finalizing provisions that would allow certain portions of MSW to be included as renewable biomass, provided that reasonable separation has first occurred.
  • Equivalence Values: We are generally maintaining the provisions from RFS1 that the Equivalence Value for each renewable fuel will be based on its energy content in comparison to ethanol, adjusted for renewable content. The cellulosic biofuel, advanced biofuel, and renewable fuel standards can be met with ethanol-equivalent volumes of renewable fuel. However, since the biomass-based diesel standard is a “diesel” standard, its volume must be met on a biodiesel-equivalent energy basis.
  • Cellulosic biofuel waiver credits: If EPA reduces the required volume of cellulosic biofuel according to the waiver provisions in EISA, EPA will offer a number of credits to obligated parties no greater than the reduced cellulosic biofuel standard. These waiver credits are not allowed to be traded or banked for future use, and are only allowed to be used to meet the cellulosic biofuel standard for the year that they are offered. In response to concerns expressed in comments on the proposal, we are implementing certain restrictions on the use of these waiver credits. For example, unlike Cellulosic Biofuel RINs, waiver credits may not be used to meet either the advanced biofuel standard or the total renewable fuel standard. For the 2010 compliance period, since the cellulosic standard is lower than the level otherwise required by EISA, we are making cellulosic waiver credits available to obligated parties for end-of-year compliance should they need them at a price of $1.56 per gallon-RIN.
  • Obligated fuels: EISA expanded the program to cover “transportation fuel”, not just gasoline. Therefore, under RFS2, obligated fuel volumes will include all gasoline and all MVNRLM diesel fuel. Other fuels such as jet fuel and fuel intended for use in ocean-going vessels are not obligated fuels under RFS2. However, renewable fuels used in jet fuel or heating oil are valid for meeting the renewable fuel volume mandates. Similarly, while we are not including natural gas, propane, or electricity used in transportation as obligated fuels at this time, we will allow renewable forms of these fuels to qualify under the program for generating RINs.

B. Impacts of Increasing Volume Requirements in the RFS2 Program

The displacement of gasoline and diesel with renewable fuels has a wide range of environmental and economic impacts. As we describe in Sections IV-IX, we have assessed many of these impacts for the final rule. It is difficult to ascertain how much of these impacts might be due to the natural growth in renewable fuel use due to market forces as crude oil prices rise versus what might be forced by the RFS2 standards. Regardless, these assessments provide important information on the wider public policy considerations related to renewable fuel production and use, climate change, and national energy security. Where possible, we have tried to provide two perspectives on the impacts of the renewable fuel volumes mandated in EISA—both relative to the RFS1 mandated volumes, and relative to a projection from EIA (AEO 2007) of renewable fuel volumes that would have been expected without EISA.

Based on the results of our analyses, when fully phased in by 2022, the increased volume of renewable fuel required by this final rule in comparison to the AEO 2007 forecast would result in 138 million metric tons fewer CO 2-equivalent GHG emissions (annual average over 30 years), the equivalent of removing 27 million vehicles from the road today.

At the same time, increases in emissions of hydrocarbons, nitrogen oxides, particulate matter, and other pollutants are projected to lead to increases in population-weighted annual average ambient PM and ozone concentrations, which in turn are anticipated to lead to up to 245 cases of adult premature mortality. The air quality impacts, however, are highly variable from region to region. Ambient PM 2.5 is likely to increase in areas associated with biofuel production and transport and decrease in other areas; for ozone, many areas of the country will experience increases and a few areas will see decreases. Ethanol concentrations will increase substantially; for the other modeled air toxics there are some localized impacts, but relatively little impact on national average concentrations. We note that the air quality modeling results presented in this final rule do not constitute the “anti-backsliding” analysis required by Clean Air Act section 211(v). EPA will be analyzing air quality impacts of increased renewable fuel use through that study and will promulgate appropriate mitigation measures under section 211(v), separate from this final action.

In addition to air quality, there are also expected to be adverse impacts on both water quality and quantity as the production of biofuels and their feedstocks increase.

In addition to environmental impacts, the increased volumes of renewable fuels required by this final rule are also projected to have a number of other energy and economic impacts. The increased renewable fuel use is estimated to reduce dependence on foreign sources of crude oil, increase domestic sources of energy, and diversify our energy portfolio to help in moving beyond a petroleum-based economy. The increased use of renewable fuels is also expected to have the added benefit of providing an expanded market for agricultural products such as corn and soybeans and open new markets for the development of cellulosic feedstock industries and conversion technologies. Overall, however, we estimate that the renewable fuel standards will result in significant net benefits, ranging between $16 and $29 billion in 2022.

Table I.B-1 summarizes the results of our impacts analyses of the volumes of renewable fuels required by the RFS2 standards in 2022 relative to the AEO2007 reference case and identifies the section where you can find further explanation of it. As we work to implement the requirements of EISA, we will continue to assess these impacts. These are the annual impacts projected in 2022 when the program is fully phased in. Impacts in earlier years would differ but in most cases were not able to be modeled or assessed for this final rule.

Table I.B-1—Impact Summary of the RFS2 Standards in 2022 Relative to the AEO2007 Reference Case (2007 Dollars) Back to Top
Category Impact in 2022 Section discussed
aThe models used to estimate SCC values have not been exercised in a systematic manner that would allow researchers to assess the probability of different values. Therefore, the interim SCC values should not be considered to form a range or distribution of possible or likely values. See Section VIII.D for a complete summary of the interim SCC values.
bSum of Overall Fuel Costs, Energy Security, Monetized Health Impacts, and GHG Impacts (SCC).
Emissions and Air Quality    
GHG Emissions −138 million metric tons V.D.
Non-GHG Emissions (criteria and toxic pollutants) −1% to +10% depending on the pollutant VI.A.
Nationwide Ozone +0.12 ppb population-weighted seasonal max 8 hr average VIII.D.
Nationwide PM 2.5 +0.002 μg/m3population-weighted annual average PM 2.5 VIII.D.
Nationwide Ethanol +0.409 μg/m3population-weighted annual average VI.D.
Other Nationwide Air Toxics −0.0001 to −0.023 μg/m3population-weighted annual average depending on the pollutant VI.D.
PM 2.5-related Premature Mortality 33 to 85 additional cases of adult mortality (estimates vary by study) VIII.D.
Ozone-related Premature Mortality 36 to 160 additional cases of adult mortality (estimates vary by study) VIII.D.
Other Environmental Impacts    
Loadings to the Mississippi River from the Upper Mississippi River Basin Nitrogen: +1,430 million lbs. (1.2%) Phosphorus: +132 million lbs. (0.7%) IX.
Fuel Costs    
Gasoline Costs −2.4¢/gal VII.D.
Diesel Costs −12.1 ¢/gal VII.D.
Overall Fuel Cost −$11.8 Billion VII.D.
Gasoline and Diesel Consumption −13.6 Bgal VII.C.
Food Costs    
Corn +8.2% VIII.A.
Soybeans +10.3% VIII.A.
Food +$10 per capita VIII.A.
Economic Impacts    
Energy Security +$2.6 Billion VIII.B.
Monetized Health Impacts −$0.63 to −$2.2 Billion VIII.D.
GHG Impacts (SCC)a +$0.6 to $12.2 Billion (estimates vary by SCC assumption) VIII.C.
Oil Imports −$41.5 Billion VIII.B
Farm Gate Food +$3.6 Billion VIII.A.
Farm Income +$13 Billion (+36%) VIII.A.
Corn Exports −$57 Million (−8%) VIII.A.
Soybean Exports −$453 Million (−14%) VIII.A.
Total Net Benefitsb +$13 to $26 Billion (estimates vary by SCC assumption) VIII.F.

II. Description of the Regulatory Provisions Back to Top

While EISA made a number of changes to CAA section 211(o) that must be reflected in the RFS program regulations, it left many of the basic program elements intact, including the mechanism for translating national renewable fuel volume requirements into applicable standards for individual obligated parties, requirements for a credit trading program, geographic applicability, treatment of small refineries, and general waiver provisions. As a result, many of the regulatory requirements of the RFS1 program will remain largely or, in some cases, entirely unchanged. These provisions include the distribution of RINs, separation of RINs, use of RINs to demonstrate compliance, provisions for exporters, recordkeeping and reporting, deficit carryovers, and the valid life of RINs.

The primary elements of the RFS program that we are changing to implement the requirements in EISA fall primarily into the following seven areas:

(1) Expansion of the applicable volumes of renewable fuel.

(2) Separation of the volume requirements into four separate categories of renewable fuel, with corresponding changes to the RIN and to the applicable standards.

(3) New definitions of renewable fuel, advanced biofuel, biomass-based diesel, and cellulosic biofuel.

(4) New requirement that renewable fuels meet certain lifecycle emission reduction thresholds.

(5) New definition of renewable biomass from which renewable fuels can be made, including certain land use restrictions.

(6) Expansion of the types of fuels that are subject to the standards to include diesel.

(7) Inclusion of specific types of waivers for different categories of renewable fuels and, in certain circumstances, EPA-generated credits for cellulosic biofuel.

EISA does not change the basic requirement under CAA 211(o) that the RFS program include a credit trading program. In the May 1, 2007 final rulemaking implementing the RFS1 program, we described how we reviewed a variety of approaches to program design in collaboration with various stakeholders. We finally settled on a RIN-based system for compliance and credit purposes as the one which met our goals of being straightforward, maximizing flexibility, ensuring that volumes are verifiable, and maintaining the existing system of fuel distribution and blending. RINs represent the basic framework for ensuring that the statutorily required volumes of renewable fuel are used as transportation fuel in the U.S. Since the RIN-based system generally has been successful in meeting the statutory goals, we are maintaining much of its structure under RFS2.

This section describes the regulatory changes we are finalizing to implement the new EISA provisions. Section III describes other changes to the RFS program that we considered or are finalizing, including an EPA-moderated RIN trading system that provides a context within which all RIN transfers will occur.

A. Renewable Identification Numbers (RINs)

Under RFS2, each RIN will continue to represent one gallon of renewable fuel in the context of demonstrating compliance with Renewable Volume Obligations (RVO), consistent with our approach under RFS1, and the RIN will continue to have unique information similar to the 38 digits in RFS1. However in the EPA Moderated Transaction System (EMTS), RIN detail information will be available but generally hidden during transactions. In general the codes within the RIN will have the same meaning under RFS2 as they do under RFS1, with the exception of the D code which will be expanded to cover the four categories of renewable fuel defined in EISA.

As described in Section I.A.2, the RFS2 regulatory program will go into effect on July 1, 2010, but the 2010 percentage standards issued as part of today's rule will apply to all gasoline and diesel produced or imported on or after January 1, 2010. As a result, some 2010 RINs will be generated under the RFS1 requirements and others will be generated under the RFS2 requirements, but all RINs generated in 2010 will be valid for meeting the 2010 annual standards. Since RFS1 RINs and RFS2 RINs will differ in the meaning of the D codes, we are implementing a mechanism for distinguishing between these two categories of RINs in order to appropriately apply them to the standards. In short, we are requiring the use of D codes under RFS2 that do not overlap the values for the D codes under RFS1. Table II.A-1 describes the D code definitions we are finalizing in today's action.

Table II.A-1—Final D Code Definitions Back to Top
D value Meaning under RFS1 Meaning under RFS2
1 Cellulosic biomass ethanol Not applicable.
2 Any renewable fuel that is not cellulosic biomass ethanol Not applicable.
3 Not applicable Cellulosic biofuel.
4 Not applicable Biomass-based diesel.
5 Not applicable Advanced biofuel.
6 Not applicable Renewable fuel.
7 Not applicable Cellulosic diesel.

Under this approach, D code values of 1 and 2 are only relevant for RINs generated under RFS1, and D code values of 3, 4, 5, 6, and 7 are only relevant for RINs generated under RFS2. As described in Section I.A.2, the RFS1 regulations will apply in January through June of 2010, while the RFS2 regulations will become effective on July 1, 2010. RINs generated under RFS1 regulations in the first three months of 2010 can be used for meeting the four 2010 standards applicable under RFS2. To accomplish this, these RFS1 RINs will be subject to the RFS1/RFS2 transition provisions wherein they will be deemed equivalent to one of the four RFS2 RIN categories using their RR and/or D codes. See Section II.G.4 for further description of how RFS1 RINs will be used to meet standards under RFS2. The determination of which D code will be assigned to a given batch of renewable fuel is described in more detail in Section II.D.2 below.

Table II.A-1 includes one D code corresponding to each of the four renewable fuel categories defined in EISA, and an additional D code of 7 corresponding to the unique, additional type of renewable fuel called cellulosic diesel. As described in the NPRM, a diesel fuel product produced from cellulosic feedstocks that meets the 60% GHG threshold could qualify as either cellulosic biofuel or biomass-based diesel. The NPRM described two possible approaches to this unique category of renewable fuel:

1. Have the producer of the cellulosic diesel designate their fuel up front as either cellulosic biofuel with a D code of 3, or biomass-based diesel with a D code of 4, limiting the subsequent potential in the marketplace for the RIN to be used for just one standard or the other.

2. Have the producer of the cellulosic diesel designate their fuel with a new cellulosic D code of 7, allowing the subsequent use of the RIN in the marketplace interchangeably for either the cellulosic biofuel standard or the biomass-based diesel standard.

We are finalizing the second option. By creating an additional D code of 7 to represent cellulosic diesel RINs, we believe its value in the marketplace will be maximized as it will be priced according to the relative demand for cellulosic biofuel and biomass-based diesel RINs. For instance, if demand for cellulosic biofuel RINs is higher than demand for biomass-based diesel RINs, then cellulosic diesel RINs will be priced as if they are cellulosic biofuel RINs. Not only does this approach benefit producers, but it allows obligated parties the flexibility to apply a RIN with a D code of 7 to either their cellulosic biofuel RVO or their biomass-based diesel RVO, depending on the number of RINs they have acquired to meet these two obligations. It also helps the functionality of the RIN program by helping protect against the potential for artificial RIN shortages in the marketplace for one standard or the other even though sufficient qualifying fuel was produced.

Under RFS2, each batch-RIN generated will continue to uniquely identify not only a specific batch of renewable fuel, but also every gallon-RIN assigned to that batch. Thus the RIN will continue to be defined as follows:

RIN: KYYYYCCCCFFFFFBBBBBRRDSSSSSSSSEEEEEEEE

Where:

K = Code distinguishing assigned RINs from separated RINs

YYYY = Calendar year of production or import

CCCC = Company ID

FFFFF = Facility ID

BBBBB = Batch number

RR = Code identifying the Equivalence Value

D = Code identifying the renewable fuel category

SSSSSSSS = Start of RIN block

EEEEEEEE = End of RIN block

B. New Eligibility Requirements for Renewable Fuels

Aside from the higher volume requirements, most of the substantive changes that EISA makes to the RFS program affect the eligibility of renewable fuels in meeting one of the four volume requirements. Eligibility is determined based on the types of feedstocks that are used, the land that is used to grow feedstocks for renewable fuel production, the processes that are used to convert those feedstocks into fuel, and the lifecycle greenhouse gas (GHG) emissions that are emitted in comparison to the gasoline or diesel that the renewable fuel displaces. This section describes these eligibility criteria and how we are implementing them for the RFS2 program.

1. Changes in Renewable Fuel Definitions

Under the previous Renewable Fuel Standards (RFS1), renewable fuel was defined generally as “any motor vehicle fuel that is used to replace or reduce the quantity of fossil fuel present in a fuel mixture used to fuel a motor vehicle”. The RFS1 definition included motor vehicle fuels produced from biomass material such as grain, starch, fats, greases, oils, and biogas. The definition specifically included cellulosic biomass ethanol, waste derived ethanol, and biodiesel, all of which were defined separately. (See 72 FR 23915).

The definitions of renewable fuels under today's rule (RFS2) are based on the new statutory definition in EISA. Like the previous rules, the definitions in RFS2 include a general definition of renewable fuel, but unlike RFS1, we are including a separate definition of “Renewable Biomass” which identifies the feedstocks from which renewable fuels may be made.

Another difference in the definitions of renewable fuel is that RFS2 contains three subcategories of renewable fuels: (1) Advanced Biofuel, (2) Cellulosic Biofuel and (3) Biomass-Based Diesel. Each must meet threshold levels of reduction of greenhouse gas emissions as discussed in Section II.B.2. The specific definitions and how they differ from RFS1 follow below.

a. Renewable Fuel

“Renewable Fuel” is defined as fuel produced from renewable biomass and that is used to replace or reduce the quantity of fossil fuel present in a transportation fuel. The definition of “Renewable Fuel” now refers to “transportation fuel” rather than referring to motor vehicle fuel. “Transportation fuel” is also defined, and means fuel used in motor vehicles, motor vehicle engines, nonroad vehicles or nonroad engines (except for ocean going vessels). Also renewable fuel now includes heating fuel and jet fuel.

Given that the primary use of electricity, natural gas, and propane is not for fueling vehicles and engines, and the producer generally does not know how it will be used, we cannot require that producers or importers of these fuels generate RINs for all the volumes they produce as we do with other renewable fuels. However, we are allowing fuel producers, importers and end users to include electricity, natural gas, and propane made from renewable biomass as a RIN-generating renewable fuel in RFS only if they can identify the specific quantities of their product which are actually used as a transportation fuel,. This may be possible for some portion of renewable electricity and biogas since many of the affected vehicles and equipment are in centrally-fueled fleets supplied under contract by a particular producer or importer of natural gas or propane. A producer or importer of renewable electricity or biogas who documents the use of his product in a vehicle or engine through a contractual pathway would be allowed to generate RINs to represent that product, if it met the definition of renewable fuel. (This is also discussed in Section II.D.2.a)

b. Advanced Biofuel

“Advanced Biofuel” is a renewable fuel other than ethanol derived from corn starch and for which lifecycle GHG emissions are at least 50% less than the gasoline or diesel fuel it displaces. Advanced biofuel would be assigned a D code of 5 as shown in Table II.A-1.

While “Advanced Biofuel” specifically excludes ethanol derived from corn starch, it includes other types of ethanol derived from renewable biomass, including ethanol made from cellulose, hemicellulose, lignin, sugar or any starch other than corn starch, as long as it meets the 50% GHG emission reduction threshold. Thus, even if corn starch-derived ethanol were made so that it met the 50% GHG reduction threshold, it will still be excluded from being defined as an advanced biofuel. Such ethanol while not an advanced biofuel will still qualify as a renewable fuel for purposes of meeting the standards.

c. Cellulosic Biofuel

Cellulosic biofuel is renewable fuel derived from any cellulose, hemicellulose, or lignin each of which must originate from renewable biomass. It must also achieve a lifecycle GHG emission reduction of at least 60%, compared to the gasoline or diesel fuel it displaces. Cellulosic biofuel is assigned a D code of 3 as shown in Table II.A-1. Cellulosic biofuel in general also qualifies as both “advanced biofuel” and “renewable fuel”.

The definition of cellulosic biofuel for RFS2 is broader in some respects than the RFS1 definition of “cellulosic biomass ethanol”. That definition included only ethanol, whereas the RFS2 definition of cellulosic biofuels includes any biomass-to-liquid fuel such as cellulosic gasoline or diesel in addition to ethanol. The definition of “cellulosic biofuel” in RFS2 differs from RFS1 in another significant way. The RFS1 definition provided that ethanol made at any facility—regardless of whether cellulosic feedstock is used or not—may be defined as cellulosic if at such facility “animal wastes or other waste materials are digested or otherwise used to displace 90% or more of the fossil fuel normally used in the production of ethanol.” This provision was not included in EISA, and therefore does not appear in the definitions pertaining to cellulosic biofuel in the final rule.

d. Biomass-Based Diesel

“Biomass-based diesel” includes both biodiesel (mono-alkyl esters) and non-ester renewable diesel (including cellulosic diesel). The definition of biodiesel is the same very broad definition of “biodiesel” that was in EPAct and in RFS1, and thus, it includes any diesel fuel made from biomass feedstocks. However, EISA added three restrictions. First, EISA requires that such fuel be made from renewable biomass. Second, its lifecycle GHG emissions must be at least 50% less than the diesel fuel it displaces. Third, the statutory definition of “Biomass-based diesel” excludes renewable fuel derived from co-processing biomass with a petroleum feedstock. In our proposed rule, we sought comment on two options for how co-processing could be treated. The first option considered co-processing to occur only if both petroleum and biomass feedstock are processed in the same unit simultaneously. The second option considered co-processing to occur if renewable biomass and petroleum feedstock are processed in the same unit at any time; i.e., either simultaneously or sequentially. Under the second option, if petroleum feedstock was processed in the unit, then no fuel produced from such unit, even from a biomass feedstock, would be deemed to be biomass-based diesel.

We selected the first option to be used in the final rule. Under this approach, a batch of fuel qualifying for the D code of 4 that is produced in a processing unit in which only renewable biomass is the feedstock for such batch, will meet the definition of “Biomass-Based Diesel. Thus, serial batch processing in which 100% vegetable oil is processed one day/week/month and 100% petroleum the next day/week/month could occur without the activity being considered “co-processing.” The resulting products could be blended together, but only the volume produced from vegetable oil will count as biomass-based diesel. We believe this is the most straightforward approach and an appropriate one, given that it would allow RINs to be generated for volumes of fuel meeting the 50% GHG reduction threshold that is derived from renewable biomass, while not providing any credit for fuel derived from petroleum sources. In addition, this approach avoids the need for potentially complex provisions addressing how fuel should be treated when existing or even mothballed petroleum hydrotreating equipment is retrofitted and placed into new service for renewable fuel production or vice versa.

Under today's rule, any fuel that does not satisfy the definition of biomass-based diesel only because it is co-processed with petroleum will still meet the definition of “Advanced Biofuel” provided it meets the 50% GHG threshold and other criteria for the D code of 5. Similarly it will meet the definition of renewable fuel if it meets a GHG emission reduction threshold of 20%. In neither case, however, will it meet the definition of biomass-based diesel.

This restriction is only really an issue for renewable diesel and biodiesel produced via the fatty acid methyl ester (FAME) process. For other forms of biodiesel, it is never made through any sort of co-processing with petroleum. [3] Producers of renewable diesel must therefore specify whether or not they use “co-processing” to produce the fuel in order to determine the correct D code for the RIN.

e. Additional Renewable Fuel

The statutory definition of “additional renewable fuel” specifies fuel produced from renewable biomass that is used to replace or reduce fossil fuels used in heating oil or jet fuel. EISA indicates that EPA may allow for the generation of credits for such additional renewable fuel that will be valid for compliance purposes. Under the RFS program, RINs operate in the role of credits, and RINs are generated when renewable fuel is produced rather than when it is blended. In most cases, however, renewable fuel producers do not know at the time of fuel production (and RIN generation) how their fuel will ultimately be used.

Under RFS1, only RINs assigned to renewable fuel that was blended into motor vehicle fuel (i.e., highway fuel) are valid for compliance purposes. We therefore created special provisions requiring that RINs be retired if they were assigned to renewable fuel that was ultimately blended into nonroad fuel. The new EISA provisions regarding additional renewable fuel make the RFS1 requirement for retiring RINs unnecessary if renewable fuel is blended into heating oil or jet fuel. As a result, we have modified the regulatory requirements to allow RINs assigned to renewable fuel blended into heating oil or jet fuel in addition to highway and nonroad transportation fuels to continue to be valid for compliance purposes. From a regulatory standpoint, there is no difference between renewable fuels used for transportation purposes, versus heating oil and jet fuels.

EISA uses the term “home heating oil” in the definition of “additional renewable fuel.” The statute does not clarify whether the term should be interpreted to refer only to heating oil actually used in homes, or to all fuel of a type that can be used in homes. We note that the term “home heating oil” is typically used in industry in the latter manner, to refer to a type of fuel, rather than a particular use of it, and the term is typically used interchangeably in industry with heating oil, heating fuel, home heating fuel, and other terms depending on the region and market. We believe this broad interpretation based on typical industry usage best serves the goals and purposes of the statute. If EPA interpreted the term to apply only to heating oil actually used in homes, we would necessarily require tracking of individual gallons from production through ultimate use in use in homes in order to determine eligibility of the fuel for RINs. Given the fungible nature of the oil delivery market, this would likely be sufficiently difficult and potentially expensive so as to discourage the generation of RINs for renewable fuels used as home heating oil. This problem would be similar to that which arose under RFS1 for certain renewable fuels (in particular biodiesel) that were produced for the highway diesel market but were also suitable for other markets such as heating oil and non-road applications where it was unclear at the time of fuel production (when RINs are typically generated under the RFS program) whether the fuel would ultimately be eligible to generate RINs. Congress eliminated the complexity with regards to non-road applications in RFS2 by making all fuels used in both motor vehicle and nonroad applications subject to the renewable fuel standard program. We believe it best to interpret the Act so as to also avoid this type of complexity in the heating oil context. Thus, under today's regulations, RINs may be generated for renewable fuel used as “heating oil,” as defined in existing EPA regulations at 80.2(ccc). In addition to simplifying implementation and administration of the Act, this interpretation will best realize the intent of EISA to reduce or replace the use of fossil fuels,

f. Cellulosic Diesel

In the proposed rule, we sought comment on how diesel made from cellulosic feedstocks should be considered. Specifically, a diesel fuel product produced from cellulosic feedstocks that meets the 60% GHG threshold could qualify as either cellulosic biofuel or biomass-based diesel. Based on comments received, and as discussed previously in Section II.A, today's rule requires the cellulosic diesel producer to categorize their product as cellulosic diesel with a D code of 7. It can then be traded in the marketplace and used for compliance with either the biomass-based diesel standard or the cellulosic biofuel standard.

2. Lifecycle GHG Thresholds

As part of the new definitions that EISA creates for cellulosic biofuel, biomass-based diesel, advanced biofuel, and renewable fuel, EISA also sets minimum performance measures or “thresholds” for lifecycle GHG emissions. These thresholds represent the percent reduction in lifecycle GHGs that is estimated to occur when a renewable fuel displaces gasoline or diesel fuel. Table II.B.2-1 lists the thresholds established by EISA.

Table II.B.2-1—Lifecycle GHG Thresholds in EISA Back to Top
[Percent reduction from a 2005 gasoline or diesel baseline]
Renewable fuel 20%
Advanced biofuel 50%
Biomass-based diesel 50%
Cellulosic biofuel 60%

There are also special provisions for each of these thresholds:

Renewable fuel: The 20% threshold only applies to renewable fuel from new facilities that commenced construction after December 19, 2007, with an additional exemption from the 20% threshold for ethanol plants that commenced construction in 2008 or 2009 and are fired with natural gas, biomass, or any combination thereof. Facilities not subject to the 20% threshold are “grandfathered.”See Section II.B.3 below for a complete discussion of grandfathering. Also, EPA can adjust the 20% threshold to as low as 10%, but the adjustment must be the minimum possible, and the resulting threshold must be established at the maximum achievable level based on natural gas fired corn-based ethanol plants.

Advanced biofuel and biomass-based diesel: The 50% threshold can be adjusted to as low as 40%, but the adjustment must be the minimum possible and result in the maximum achievable threshold taking cost into consideration. Also, such adjustments can be made only if it is determined that the 50% threshold is not commercially feasible for fuels made using a variety of feedstocks, technologies, and processes.

Cellulosic biofuel: Similarly to advanced biofuel and biomass-based diesel, the 60% threshold applicable to cellulosic biofuel can be adjusted to as low as 50%, but the adjustment must be the minimum possible and result in the maximum achievable threshold taking cost into consideration. Also, such adjustments can be made only if it is determined that the 60% threshold is not commercially feasible for fuels made using a variety of feedstocks, technologies, and processes.

Our analyses of lifecycle GHG emissions, discussed in detail in Section V, identified a range of fuel pathways that are capable of complying with the GHG performance thresholds for each of these separate fuel standards. Thus, we have determined that the GHG thresholds in Table II.B.2-1 should not be adjusted. Further discussion of this determination can be found in Section V.C.

3. Renewable Fuel Exempt From 20 Percent GHG Threshold

After considering comments received, the Agency has decided to implement the proposed option for interpreting the grandfathering provisions that provide an indefinite exemption from the 20 percent GHG threshold for renewable fuel facilities which have commenced construction prior to December 19, 2007. For these facilities, only the baseline volume of renewable fuel is exempted. For ethanol facilities which commenced construction after that date and which use natural gas, biofuels or a combination thereof, we proposed that such facilities would be “deemed compliant” with the 20 percent GHG threshold. The exemption for such facilities is conditioned on construction being commenced on or before December 31, 2009, and is specific only to facilities which produce ethanol only, per language in EISA. The exemption would continue indefinitely, provided the facility continues to use natural gas and/or biofuel. This section provides the background and summary of the original proposal, and the reasons for the selection of this option.

a. General Background of the Exemption Requirement

EISA amends section 211(o) of the Clean Air Act to provide that renewable fuel produced from new facilities which commenced construction after December 19, 2007 must achieve at least a 20% reduction in lifecycle greenhouse gas emissions compared to baseline lifecycle greenhouse gas emissions. [7] Facilities that commenced construction before December 19, 2007 are “grandfathered” and thereby exempt from the 20% GHG reduction requirement.

For facilities that produce ethanol and for which construction commenced after December 19, 2007, section 210 of EISA states that “for calendar years 2008 and 2009, any ethanol plant that is fired with natural gas, biomass, or any combination thereof is deemed to be in compliance with the 20% threshold.” Since all renewable fuel production facilities that commenced construction prior to the date of EISA enactment are covered by the more general grandfathering provision, this exemption can only apply to those facilities that commenced construction after enactment of EISA, and before the end of 2009. We proposed that the statute be interpreted to mean that fuel from such qualifying facilities, regardless of date of startup of operations, would be exempt from the 20% GHG threshold requirement for the same time period as facilities that commence construction prior to December 19, 2007, provided that such plants commence construction on or before December 31, 2009, complete such construction in a reasonable amount of time, and continue to burn only natural gas, biomass, or a combination thereof. Most commenters generally agreed with our proposal, while other commenters argued that the exemption was only meant to last for a two-year period. As we noted in the NPRM, we believe that it would be a harsh result for investors in these new facilities, and would be generally inconsistent with the energy independence goals of EISA, to interpret the Act such that these facilities would only be guaranteed two years of participation in the RFS2 program. In light of these considerations, we continue to believe that it is an appropriate interpretation of the Act to allow the deemed compliant exemption to continue indefinitely with the limitations we proposed. Therefore we are making final this interpretation in today's rule.

b. Definition of Commenced Construction

In defining “commence” and “construction”, we proposed to use the definitions of “commence” and “begin actual construction” from the Prevention of Significant Deterioration (PSD) regulations, which draws upon definitions in the Clean Air Act. (40 CFR 52.21(b)(9) and (11)). Specifically, under the PSD regulations, “commence” means that the owner or operator has all necessary preconstruction approvals or permits and either has begun a continuous program of actual on-site construction to be completed in a reasonable time, or entered into binding agreements which cannot be cancelled or modified without substantial loss.” Such activities include, but are not limited to, “installation of building supports and foundations, laying underground pipe work and construction of permanent storage structures.” We proposed adding language to the definition that is currently not in the PSD definition with respect to multi-phased projects. We proposed that for multi-phased projects, commencement of construction of one phase does not constitute commencement of construction of any later phase, unless each phase is “mutually dependent” on the other on a physical and chemical basis, rather than economic.

The PSD regulations provide additional conditions beyond addressing what constitutes commencement. Specifically, the regulations require that the owner or operator “did not discontinue construction for a period of 18 months or more and completed construction within a reasonable time.” (40 CFR 52.21(i)(4)(ii)(c)). While “reasonable time” may vary depending on the type of project, we proposed that for RFS2 a reasonable time to complete construction of renewable fuel facilities be no greater than 3 years from initial commencement of construction. We sought comment on this time frame.

Commenters generally agreed with our proposed definition of commenced construction. Some commenters felt that the 3 year time frame was not a “reasonable time” to complete construction in light of the economic difficulties that businesses have been and will likely continue to be facing. We recognize that there have been extreme economic problems in the past year. Based on historical data which show construction of ethanol plants typically take about one year, we believe that the 3-year time frame allows such conditions to be taken into account and that it is an appropriate and fair amount of time to allow for completion. Therefore, we are not extending the amount of time that constitutes “reasonable” to five years as was suggested.

c. Definition of Facility Boundary

We proposed that the grandfathering and deemed compliant exemptions apply to “facilities.” Our proposed definition of this term is similar in some respects to the definition of “building, structure, facility, or installation” contained in the PSD regulations in 40 CFR 52.21. We proposed to modify the definition, however, to focus on the typical renewable fuel plant. We proposed to describe the exempt “facilities” as including all of the activities and equipment associated with the manufacture of renewable fuel which are located on one property and under the control of the same person or persons. Commenters agreed with our proposed definition of “facility” and we are making that definition final today.

d. Proposed Approaches and Consideration of Comments

We proposed one basic approach to the exemption provisions and sought comment on five additional options. The basic approach would provide an indefinite extension of grandfathering and deemed compliant status but with a limitation of the exemption from the 20% GHG threshold to a baseline volume of renewable fuel. The five additional options for which we sought comment were: (1) Expiration of exemption for grandfathered and “deemed compliant” status when facilities undergo sufficient changes to be considered “reconstructed”; (2) Expiration of exemption 15 years after EISA enactment, industry-wide; (3) Expiration of exemption 15 years after EISA enactment with limitation of exemption to baseline volume; (4) “Significant” production components are treated as facilities and grandfathered or deemed compliant status ends when they are replaced; and (5) Indefinite exemption and no limitations placed on baseline volumes.

i. Comments on the Proposed Basic Approach

Generally, commenters supported the basic approach in which the volume of renewable fuel from grandfathered facilities exempt from the 20% GHG reduction threshold would be limited to baseline volume. One commenter objected to the basic approach and argued that the statute's use of the word “new” and the phrase “after December 19, 2007” provided evidence that facilities which commenced construction prior to that date would not ever be subject to the threshold regardless of the volume produced from such facilities. In response, we note first that the statute does not provide a definition of the term “new facilities” for which the 20% GHG threshold applies. We believe that it would be reasonable to include within our interpretation of this term a volume limitation, such that a production plant is considered a new facility to the extent that it produces renewable fuel above baseline capacity. This approach also provides certainty in the marketplace in terms of the volumes of exempt fuel, and a relatively straightforward implementation and enforcement mechanism as compared to some of the other alternatives considered. Furthermore, EPA believes that the Act should not be interpreted as allowing unlimited expansion of exempt facilities for an indefinite time period, with all volumes exempt, as suggested by the commenter. Such an approach would likely lead to a substantial increase in production of fuel that is not subject to any GHG limitations, which EPA does not believe would be consistent with the objectives of the Act.

We solicited comment on whether changes at a facility that resulted in an increase in GHG emissions, such as a change in fuel or feedstock, should terminate the facility's exemption from the 20 percent GHG threshold. Generally, commenters did not support such a provision, pointing out that there are many variations within a plant that cannot be adequately captured in a table of fuel and feedstock pathways as we proposed (see 74 FR 24927). Implementing such a provision would create questions of accounting and tracking that would need to be evaluated on a time-consuming case-by-case basis. For example, if a switch to a different feedstock or production process resulted in less efficiency, facilities may argue that they are increasing energy efficiency elsewhere (e.g. purchasing waste heat instead of burning fuel onsite to generate steam). We would then need to assess such changes to track the net energy change a plant undergoes. Given the added complexity and difficulty in carrying out such an option, we have decided generally not to implement it. There is an exception, however, for “deemed compliant” facilities. These facilities achieve their status in part by being fired only by natural gas or biomass, or a combination thereof. Today's rule provides, as proposed, that these facilities will lose their exemption if they switch to a fuel other than natural gas, biomass, or a combination thereof, since these were conditions that Congress deemed critical to granting them the exemption from the 20% GHG reduction requirement.

We also solicited comment on whether we should allow a 10% tolerance on the baseline volume for which RINs can be generated without complying with the 20% GHG reduction threshold to allow for increases in volume due to debottlenecking. Some favored this concept, while others argued that the tolerance should be set at 20 percent. After considering the comments received, we have decided that a 10% (and 20%) level is not appropriate for this regulation for the following reasons: (1) We have decided to interpret the exemption of the baseline volume of renewable fuel from the 20 percent requirement as extending indefinitely. Any tolerance provided could, therefore, be present in the marketplace for a considerable time period; (2) increases in volume of 10% or greater could be the result of modifications other than debottlenecking. Consistent with the basic approach we are taking today towards interpreting the grandfathering and deemed compliant provisions, we believe that the fuel produced as a result of such modifications comes from “new facilities” within the meaning of the statute, and should be subject to the 20% GHG reduction requirement; (3) we are allowing baseline volume to be based on the maximum capacity that is allowed under state and federal air permits. With respect to the last reason, facilities that have been operating below the capacity allowed in their state permits would be able to claim a baseline volume based on the maximum capacity. As such, these facilities may indeed be able to increase their volume by 10 to 20 percent by virtue of how their baseline volume is defined. We believe this is appropriate, however, since their permits should reflect their design, and the fuel resulting from their original pre-EISA (or pre-2010, for deemed compliant facilities) design should be exempt from the 20% GHG reduction requirement. Nevertheless, we recognize and agree with commenters that some allowances should be made for minor changes brought about by normal maintenance which are consistent with the proper operation of a facility. EPA is not aware of a particular study or analysis that could be used as a basis for picking a tolerance level reflecting this concept, We believe, however, that the value should be relatively small, so as not to encourage plant expansions that are unrelated to debottlenecking. We believe that a 5% tolerance level is consistent with these considerations, and have incorporated that value in today's rule.

ii. Comments on the Expiration of Grandfathered Status

Commenters who supported an expiration of the exemption did so because of concerns that the proposed approach of providing an indefinite exemption would not provide any incentives to bring these plants into compliance with current standards. They also objected to plants being allowed an indefinite period beyond the time period when it could be expected that they would have paid off their investors. The commenters argued that the cost of operation for such plants would be less than competing plants that do have to comply with current standards; as such, commenters opposed to the basic approach felt an indefinite exemption would be a subsidy to plants that will never comply with the 20 percent threshold level. The renewable fuels industry, on the other hand, viewed the options that would set an expiration date (either via cumulative reconstruction, or a 15-year period from date of enactment) as harsh, particularly if the lifecycle analysis results make it costly for existing facilities to meet the 20% threshold. Some also argued that no such temporal limitation appears in the statute.

We considered such comments, but in light of recent lifecycle analyses we conducted in support of this rule we have concluded that many of the current technology corn ethanol plants may find it difficult if not impossible to retrofit existing plants to comply with the 20 percent GHG reduction threshold. In addition, the renewable fuels industry viewed the alternative proposals that would set an expiration date (either via cumulative reconstruction, or a 15-year period from date of enactment) as harsh, particularly if the lifecycle analysis results make it costly for existing facilities to meet the 20% threshold. Given the difficulty of meeting such threshold, owners of such facilities could decide to shut down the plant. Given such implications of meeting the 20 percent threshold level for existing facilities we have chosen not to finalize any expiration date.

e. Final Grandfathering Provisions

For the reasons discussed above, the Agency has decided to proceed with the proposed baseline volume approach, rather than the expiration options. We hold open the possibility, therefore, of revisiting and reproposing the exemption provision in a future rulemaking to take such advances into account. Ending the grandfathering exemption after its usefulness is over would help to streamline the ongoing implementation of the program.

The final approach adopted today is summarized as follows:

i. Increases in volume of renewable fuel produced at grandfathered facilities due to expansion

For facilities that commenced construction prior to December 19, 2007, we are defining the baseline volume of renewable fuel exempt from the 20% GHG threshold requirement to be the maximum volumetric capacity of the facility that is allowed in any applicable state air permit or Federal Title V operating permit. [4] We had proposed in the NPRM that nameplate capacity be defined as permitted capacity, but that if the capacity was not stipulated in any federal, state or local air permit, then the actual peak output should be used. We have decided that since permitted capacity is the limiting condition, by virtue of it being an enforceable limit contained in air permits, that the term “nameplate capacity” is not needed. In addition, we are allowing a 5% tolerance as discussed earlier. Therefore, today's rule defines permitted capacity as 105% of the maximum permissible volume output of renewable fuel allowed under operating conditions specified in all applicable preconstruction, construction and operating permits issued by regulatory authorities (including local, regional, state or a foreign equivalent of a state, and federal permits). If the capacity of a facility is not stipulated in such air permits, then the grandfathered volume is 105% of the maximum annual volume produced for any of the last five calendar years prior to 2008. Volumes greater than this amount which may typically be due to expansions of the facility which occur after December 19, 2007, will be subject to the 20% GHG reduction requirement if the facility wishes to generate RINs for the incremental expanded volume. The increased volume will be considered as if produced from a “new facility” which commenced construction after December 19, 2007. Changes that might occur to the mix of renewable fuels produced within the facility are irrelevant—they remain grandfathered as long as the overall volume falls within the baseline volume. Thus, for example, if an ethanol facility changed its operation to produce butanol, but the baseline volume remained the same, the fuel so produced would be exempt from the 20% GHG reduction requirement.

The baseline volume will be defined as above for deemed compliant facilities (those ethanol facilities fired by natural gas or biomass or a combination thereof that commenced construction after December 19, 2007 but before January 1, 2010) with the exception that if the maximum capacity is not stipulated in air permits, then the exempt volume is the maximum annual peak production during the plant's first three years of operation. In addition, any production volume increase that is attributable to construction which commenced prior to December 31, 2009 would be exempt from the 20% GHG threshold, provided that the facility continued to use natural gas, biomass or a combination thereof for process energy. Because deemed compliant facilities owe their status to the fact that they use natural gas, biomass or a combination thereof for process heat, their status will be lost, and they will be subject to the 20% GHG threshold requirement, at any time that they change to a process energy source other than natural gas and/or biomass. Finally, because EISA limits deemed compliant facilities to ethanol facilities, if there are any changes in the mix of renewable fuels produced by the facility, only the ethanol volume remains grandfathered. We had solicited comment on whether fuels other than ethanol could also be deemed compliant. Based on comments received and additional consideration to this matter, we decided that because the Act does not authorize EPA to allow fuels other than ethanol, the deemed compliant provisions will apply only to facilities producing that fuel.

Volume limitations contained in air permits may be defined in terms of peak hourly production rates or a maximum annual capacity. If they are defined only as maximum hourly production rates, they will need to be converted to an annual rate. Because assumption of a 24-hour per day production over 365 days per year (8,760 production hours) may overstate the maximum annual capacity we are requiring a conversion rate of 95% of the total hours in a year (8,322 production hours) based on typical operating “uptime” of ethanol facilities.

The facility registration process (see Section II.C) will be used to define the baseline volume for individual facilities. Owners and operators must submit information substantiating the permitted capacity of the plant, or the maximum annual peak capacity if the maximum capacity is not stipulated in a federal, state or local air permit, or EPA Title V operating permit. Copies of applicable air permits which stipulate the maximum annual capacity of the plant, must be provided as part of the registration process. Subsequent expansions at a grandfathered facility that results in an increase in volume above the baseline volume will subject the increase in volume to the 20% GHG emission reduction threshold (but not the original baseline volume). Thus, any new expansions will need to be designed to achieve the 20% GHG reduction threshold if the facility wants to generate RINs for that volume. Such determinations will be made on the basis of EPA-defined fuel pathway categories that are deemed to represent such 20% reduction.

EPA enforcement personnel commented that claims for an exemption from the 20% GHG reduction requirement should be made promptly, so that they can be verified with recent supporting information. They were concerned, in particular, that claims for exempt status could be made many years into the future for facilities that may or may not have concluded construction within the required time period, but delayed actual production of renewable fuel due to market conditions or other reasons. EPA believes that this comment has merit, and has included a requirement in Section 80.1450(f) of the final rule for registration of facilities claiming an exemption from the 20% GHG reduction requirement by May 1, 2013. This provision does not require actual fuel production, but simply the filing of registration materials that assert a claim for exempt status. It will benefit both fuel producers, who will likely be able to more readily collect the required information if it is done promptly, and EPA enforcement personnel seeking to verify the information. However, given the potentially significant implications of this requirement for facilities that may qualify for the exemption but miss the registration deadline, the rule also provides that EPA may waive the requirement if it determines that the submission is verifiable to the same extent as a timely-submitted registration.

ii. Replacements of Equipment

If production equipment such as boilers, conveyors, hoppers, storage tanks and other equipment are replaced, it would not be considered construction of a “new facility” under this option of today's final rule—the baseline volume of fuel would continue to be exempt from the 20% GHG threshold. We sought comment on an approach that would require that if coal-fired units are replaced, that the replacement units must be fired with natural gas or biofuel for the product to be eligible for RINs that do not satisfy the 20% GHG threshold. Some commenters supported such an approach. We agreed, however, with other commenters who point out that the language in EISA provides for an indefinite exemption for grandfathered facilities. While we interpret the statute to limit the exemption to the baseline volume of a grandfathered facility, we do not interpret the language to allow EPA to require that replacements of coal fired units be natural gas or biofuel. Thus replacements of coal fired equipment will not affect the facility's grandfathered status.

iii. Registration, Recordkeeping and Reporting

Facility owner/operators will be required to provide evidence and certification of commencement of construction. Such certification will require copies of all applicable air permits that apply to the construction and operation of the facility. Owner/operators must provide annual records of process fuels used on a BTU basis, feedstocks used and product volumes. For facilities that are located outside the United States (including outside the Commonwealth of Puerto Rico, the U.S. Virgin Islands, Guam, American Samoa, and the Commonwealth of the Northern Mariana Islands) owners will be required to provide certification as well. Since the definition of commencement of construction includes having all necessary air permits, we will require that facilities outside the United States certify that such facilities have obtained all necessary permits for construction and operation required by the appropriate national and local environmental agencies.

4. New Renewable Biomass Definition and Land Restrictions

As explained in Section I, EISA lists seven types of feedstock that qualify as “renewable biomass.” EISA limits not only the types of feedstocks that can be used to make renewable fuel, but also the land that these renewable fuel feedstocks may come from. Specifically, EISA's definition of renewable biomass incorporates land restrictions for planted crops and crop residue, planted trees and tree residue, slash and pre-commercial thinnings, and biomass from wildfire areas. EISA prohibits the generation of RINs for renewable fuel made from feedstock that does not meet the definition of renewable biomass, which includes not meeting the associated land restrictions. The following sections describe EPA's interpretation of several key terms related to the definition of renewable biomass, and the approach in today's rule to implementing the renewable biomass requirements.

a. Definitions of Terms

EISA's renewable biomass definition includes a number of terms that require definition. The following sections discuss EPA's definitions for these terms, which were developed with ease of implementation and enforcement in mind. We have made every attempt to define these terms as consistently with other federal statutory and regulatory definitions as well as industry standards as possible, while keeping them workable for purposes of program implementation.

i. Planted Crops and Crop Residue

The first type of renewable biomass described in EISA is planted crops and crop residue harvested from agricultural land cleared or cultivated at any time prior to December 19, 2007, that is either actively managed or fallow, and nonforested. We proposed to interpret the term “planted crops” to include all annual or perennial agricultural crops that may be used as feedstock for renewable fuel, such as grains, oilseeds, and sugarcane, as well as energy crops, such as switchgrass, prairie grass, and other species, providing that they were intentionally applied to the ground by humans either by direct application as seed or nursery stock, or through intentional natural seeding by mature plants left undisturbed for that purpose. We received numerous comments on our proposed definition of “planted crops,” largely in support of our proposed definition. However, some commenters noted that “microcrops,” such as duckweed, a flowering plant typically grown in ponds or tanks, are also being investigated for used as renewable fuel feedstocks. These microcrops are typically grown in a similar manner to algae, but cannot be categorized as algae since they are relatively more complex organisms. EPA's proposed definition would have unintentionally excluded microcrops such as duckweed through the requirement that planted crops be “applied to the ground.” After considering comments received, EPA does not believe that there is any basis under EISA for excluding from the definition of renewable biomass crops such as duckweed that are applied to a tank or pond for growth rather than to the soil. As with other planted crops, these ponds or tanks must be located on existing “agricultural land,” as described below, to qualify as renewable biomass under EISA. Therefore, including such microcrops within the definition of renewable biomass will not result in the direct loss of forestland or other ecologically sensitive land that Congress sought to protect through the land restrictions in the definition of renewable biomass. Doing so will further the objectives of the statute of promoting the development of emerging technologies to produce clean alternatives to petroleum-based fuels, and to further U.S. energy independence.

For these reasons, we are finalizing our proposed definition of “planted crops,” with the inclusion of provisions allowing for the growth of “microcrops” in ponds or tanks that are located on agricultural land. Our final definition also includes a reference to “vegetative propagation,” in which a new plant is produced from an existing vegetative structure, as one means by which planted crops may reproduce, since this is an important method of reproduction for microcrops such as duckweed. The final definition of “planted crops” includes all annual or perennial agricultural crops from existing agricultural land that may be used as feedstock for renewable fuel, such as grains, oilseeds, and sugarcane, as well as energy crops, such as switchgrass, prairie grass, duckweed and other species (but not including algae species or planted trees), providing that they were intentionally applied by humans to the ground, a growth medium, or a pond or tank, either by direct application as seed or plant, or through intentional natural seeding or vegetative propagation by mature plants introduced or left undisturbed for that purpose. We note that because EISA contains specific provisions for planted trees and tree residue from tree plantations, our final definition of planted crops in EISA excludes planted trees, even if they may be considered planted crops under some circumstances.

We proposed that “crop residue” be limited to the residue, such as corn stover and sugarcane bagasse, left over from the harvesting of planted crops. We sought comment on including biomass from agricultural land removed for purposes of invasive species control or fire management. We received many comments supporting the inclusion of biomass removed from agricultural land for purposes of invasive species control and/or fire management. We believe that such biomass is typically removed from agricultural land for the purpose of preserving or enhancing its value in agricultural crop production. It may be removed at the time crops are harvested, post harvest, periodically (e.g., for pastureland) or during extended fallow periods. We agree with the commenters that this material is a form of biomass residue related to crop production, whether or not derived from a crop itself, and, therefore, are modifying the proposed definition of “crop residue” to include it. We also received comments encouraging us to expand the definition of crop residue to include materials left over after the processing of the crop into a useable resource, such as husks, seeds, bagasse and roots. EPA agrees with these comments and has altered the final definition to cover such materials. Based on comments received, our final definition of “crop residue” is the biomass left over from the harvesting or processing of planted crops from existing agricultural land and any biomass removed from existing agricultural land that facilitates crop management (including biomass removed from such lands in relation to invasive species control or fire management), whether or not the biomass includes any portion of a crop or crop plant.

Our proposed regulations restricted planted crops and crop residue to that harvested from existing agricultural land, which, under our proposed definition, includes three land categories—cropland, pastureland, and Conservation Reserve Program (CRP) land. We proposed to define cropland as land used for the production of crops for harvest, including cultivated cropland for row crops or close-grown crops and non-cultivated cropland for horticultural crops. We proposed to define pastureland as land managed primarily for the production of indigenous or introduced forage plants for livestock grazing or hay production, and to prevent succession to other plant types. We also proposed that CRP land, which is administered by USDA's Farm Service Agency, qualify as “agricultural land” under RFS2.

EPA received numerous comments on our proposed definition of existing agricultural land. Generally, commenters were in support of our definition of “cropland” and its inclusion in the definition of existing agricultural land. Additionally, commenters generally did not object to CRP lands or pastureland being included in the definition of agricultural land. Based on our consideration of comments received on the proposed rule, EPA is including cropland, pastureland and CRP land in the definition of existing agricultural land, as proposed.

We sought comment in the proposal on whether rangeland should be included as agricultural land under RFS2. Rangeland is land on which the indigenous or introduced vegetation is predominantly grasses, grass-like plants, forbs or shrubs and which—unlike cropland or pastureland—is predominantly managed as a natural ecosystem. EPA received a number of comments concerning whether rangeland should be included in the definition of existing agricultural land under RFS2. Some commenters urged EPA to expand the definition of existing agricultural land to include rangeland, arguing that rangelands could serve as important sources of renewable fuel feedstocks. Many of these commenters argued that, although it is generally less intensively managed than cropland, rangeland is nonetheless actively managed through control of brush or weed species, among other practices. In contrast, other commenters argued against the inclusion of rangeland, contending that the potential conversion of rangeland into cropland for growing renewable biomass would lead to losses of carbon, soil, water quality, and biodiversity.

Under EISA, renewable biomass includes crops and crop residue from agricultural land cleared or cultivated at any time prior to the enactment of EISA that is either “actively managed of fallow” and nonforested. In determining whether rangeland should be considered existing agricultural land under this provision, EPA must decide if rangeland qualifies as “actively managed or fallow.” EPA believes that the term “actively managed” is best interpreted by reference to the type of material and practices that this provision addresses—namely crops and residue associated with growing crops. We think it is appropriate to inquire whether the type of management involved in a land type is consistent with that which would occur on land where crops are harvested. Thus, while we acknowledge that some types of rangeland are managed to a certain degree, the level of “active management” that is typically associated with land dedicated to growing agricultural crops is far more intensive than the types of management associated with rangeland. For example, rangeland is rarely tilled, fertilized or irrigated as croplands and, to a lesser degree, pasturelands, are. Furthermore, since rangeland encompasses a wide variety of ecosystems, including native grasslands or shrublands, savannas, wetlands, deserts and tundra, including it in the definition of agricultural land would increase the risk that these sensitive ecosystems would become available under EISA for conversion into intensively managed mono-culture cropland. Finally, the conversion of relatively undisturbed rangeland to the production of annual crops could in some cases lead to large releases of GHGs stored in the soil, as well as a loss of biodiversity, both of which would be contrary to EISA's stated goals. For these reasons, EPA is not including rangeland in the definition of “existing agricultural land” in today's final rule.

We proposed to include in our definition of existing agricultural land the requirement that the land was cleared or cultivated prior to December 19, 2007, and that, since December 19, 2007, it has been continuously actively managed (as agricultural land) or fallow, and nonforested. We proposed to interpret the phrase “that is actively managed or fallow, and nonforested” as meaning that land must have been actively managed or fallow, and nonforested, on December 19, 2007, and continuously thereafter in order to qualify for renewable biomass production. We received extensive comments on this interpretation. Many commenters suggested an interpretation of the requirement that agricultural land be “actively managed” to mean that the land had to be “actively managed” at the time EISA was passed on December 17, 2007, such that the amount of land available for biofuel feedstock production was established at that point and would not diminish over time. Other commenters supported our proposed interpretation, which would mean that the amount of land available for biofuel feedstock production could diminish over time if parcels of land cease to be actively managed at any point, thus taking them out of contention for biofuel feedstock cultivation. Some commenters argued that this interpretation is contrary to Congress' intent and the basic premise of the RFS program since, over time, it could lead to a reduction in the amount of renewable biomass available for use as renewable fuel feedstocks, while the statutorily required volumes of renewable fuel increase over time. These commenters further argue that the active management provision should be interpreted as a “snapshot” of agricultural land existing and actively managed on December 19, 2007. Under this interpretation, the land that was cleared or cultivated prior to December 19, 2007 and was actively managed on that date, would be eligible for renewable biomass production indefinitely.

We agree that the goal of the EISA and RFS program, to increase the presence of renewable fuels in transportation fuel, will be better served by interpreting the “actively managed or fallow” requirement in the renewable biomass definition as applying to land actively managed or fallow on December 19, 2007, rather than interpreting this requirement as applying beginning on December 19, 2007 and continuously thereafter. In addition, by simplifying the requirement in this fashion, there will be significantly less burden on regulated parties in ensuring that their feedstocks come from qualifying lands. For these reasons, we are modifying the definition of existing agricultural land so that the “active management” requirement is satisfied for those that were cleared or cultivated and actively managed or fallow, and non-forested on December 19, 2007.

Further, we proposed and are finalizing that “actively managed” means managed for a predetermined outcome as evidenced by any of the following: Sales records for planted crops, crop residue, or livestock; purchasing records for land treatments such as fertilizer, weed control, or reseeding; a written management plan for agricultural purposes; documentation of participation in an agricultural program sponsored by a Federal, state or local government agency; or documentation of land management in accordance with an agricultural certification program. While we received comments indicating that including a definitive checklist of required evidential records would be helpful to have explicitly identified in the regulations, we are not doing so in order to maintain flexibility, as feedstock producers may vary in the types of evidence they can readily obtain to show that their agricultural land was actively managed. We are adding, however, a clarification that the records must be traceable to the land in question. For example, it will not be sufficient to have a receipt for seed purchase if there is not additional evidence indicating that the seed was applied to the land which is claimed as existing agricultural land.

The term “fallow” is generally used to describe cultivated land taken out of production for a finite period of time. We proposed and sought comment on defining fallow to mean agricultural land that is intentionally left idle to regenerate for future agricultural purposes, with no seeding or planting, harvesting, mowing, or treatment during the fallow period. We also proposed and sought comment on requiring documentation of such intent. We received many comments that supported our proposed definition of fallow. We also received comments indicating that EPA should set a time limit for land to qualify as fallow (as opposed to abandoned for agricultural purposes). We have decided not to include a time limit for land to qualify as “fallow” because we understand that agricultural land may be left fallow for many different purposes and for varying amounts of time. Any particular timeframe that EPA might choose for this purpose would be somewhat arbitrary. Further, EISA does not indicate a time limit on the period of time that qualifying land could be fallow, so EPA does not believe that it would be appropriate to do so in its regulations. Therefore, EPA is finalizing its proposed definition of “fallow.”

Finally, in order to define the term “nonforested” as used in the definition of “existing agricultural land,” we proposed first to define the term “forestland” as generally undeveloped land covering a minimum area of one acre upon which the predominant vegetative cover is trees, including land that formerly had such tree cover and that will be regenerated. We also proposed that forestland would not include tree plantations. “Nonforested” land under our proposal would be land that is not forestland.

We received many comments on our proposed definition of forestland. Some commenters urged EPA to broaden the definition of “forestland” to include tree plantations, arguing that plantations are well-accepted as a subset of forestland. Others advocated that EPA should make every effort to distinguish between tree plantations and forestland so as not to run the risk of allowing native forests to be converted into less diverse tree plantations from which trees could be harvested for renewable fuel production. For today's final rule, EPA is including tree plantations as a subset of forestland since it is commonly understood as such throughout the forestry industry. Under EISA, renewable biomass may include “slash and pre-commercial thinnings” from non-federal forestlands, and “planted trees and tree residue” from actively managed tree plantations on non-federal land. One effect under EISA of the modification from the proposed rule to include tree plantations as a subset of forestland is to allow pre-commercial thinnings and slash, in addition to planted trees and tree residue, harvested from tree plantations to serve as qualifying feedstocks for renewable fuel production. EPA believes it is appropriate to include pre-commercial thinnings and slash from actively managed tree plantations as renewable biomass, consistent with the EISA provision allowing harvested trees and tree residue from tree plantations to qualify as renewable biomass. Another effect of including the tree plantations as a kind of forestland is that, since crops and crop residue must come from land that was “non-forested” as of the date of EISA enactment, a tract of land managed as a tree plantation on the date of EISA enactment could not be converted to cropland for the production of feedstock for RIN-generating renewable fuel. EPA believes that this result in keeping with Congressional desire to avoid the conversion of new lands to crop production for renewable fuel production.

Additionally, EPA received comments indicating that, in order to be consistent with existing statutory and/or regulatory definitions of “forestland,” EPA should exclude tree covered areas in intensive agricultural crop production settings, such as fruit orchards, or tree-covered areas in urban settings such as city parks from the definition of forestland. EPA agrees that these types of land cannot be characterized as “forestland,” and is thus excluding them from the definition. EPA's final definition of forestland is “generally undeveloped land covering a minimum of 1 acre upon which the primary vegetative species is trees, including land that formerly had such tree cover and that will be regenerated and tree plantations. Tree covered areas in intensive agricultural crop production settings, such as fruit orchards, or tree-covered areas in urban settings such as city parks, are not considered forestland.”

ii. Planted Trees and Tree Residue

The definition of renewable biomass in EISA includes planted trees and tree residue from actively managed tree plantations on non-federal land cleared at any time prior to December 19, 2007, including land belonging to an Indian tribe or an Indian individual, that is held in trust by the United States or subject to a restriction against alienation imposed by the United States.

We proposed to define the term “planted trees” to include not only trees that were established by human intervention such as planting saplings and artificial seeding, but also trees established from natural seeding by mature trees left undisturbed for such a purpose. Some commenters disagreed with our inclusion of naturally seeded trees in our definition of “planted trees.” They argue that an area which is managed for natural regeneration of trees is more akin to a natural forest than a tree plantation, and that the difference between the two types of land should be clear in order to distinguish between the two and to avoid the effective conversion of natural forests to tree plantations under EISA. EPA agrees that the inclusion of natural reseeding in the definition of “planted trees” would make distinguishing between tree plantations and forests difficult or impossible, thus negating the separate restrictions that Congress placed on the two types of land. On the other hand, EPA believes that trees that are naturally seeded and grown together with hand- or machine-planted trees in a tree plantation should not categorically be excluded from qualifying as renewable biomass. Such natural reseeding may occur after planting the majority of trees in a tree plantation, and may be consistent with the management plan for a tree plantation. EPA has decided, therefore, to modify its proposed definition of “planted tree” to be trees harvested from a tree plantation. The term “tree plantation” is defined as a stand of no less than 1 acre composed primarily of trees established by hand- or machine-planting of a seed or sapling, or by coppice growth from the stump or root of a tree that was hand- or machine-planted.” The net effect is that as long as a tree plantation consists “primarily” of trees that were hand- or machine planted (or derived therefrom, as described below), then all trees from the tree plantation, including those established from natural seeding by mature trees left undisturbed for such a purpose, will qualify as renewable biomass.

We also received a number of comments suggesting that EPA broaden the definition of planted trees to include other methods of tree regeneration, such as coppice (the production of new stems from stumps or roots), that are frequently used in the forestry industry to regenerate tree plantations. EPA believes that “planted” implies direct human intervention, and that allowing stump-growth from the stump or roots of a tree that was hand- or machine-planted is consistent with this concept. Therefore, today's final rule broadens the concept of “planted trees” from a tree plantation to include “a tree established by hand- or machine-planting of a seed or sapling, or by coppice growth from the stump or root of a tree that was hand- or machine-planted.” This new language will appear in the definition of “tree plantation.”

In the NPRM, we proposed to define a “tree plantation” as a stand of no fewer than 100 planted trees of similar age and comprising one or two tree species, or an area managed for growth of such trees covering a minimum of one acre. We received numerous comments on our definition of tree plantation. Several commenters urged EPA to define tree plantation more broadly by using the definition from the Dictionary of Forestry—“a stand composed primarily of trees established by planting or artificial seeding,” However, this definition does not provide sufficiently clear guidelines for determining whether a given parcel of land would be considered a tree plantation rather than a natural forest. Since trees are considered renewable biomass under RFS2 only if they are harvested from tree plantations, we believe that our proposed definition was clearer and more easily applied in the field. Accordingly, EPA has not adopted the definition of this term from the Dictionary of Forestry. Other commenters argued that there is no technical justification for limiting the number of species or number of trees in a plantation, and that many tree plantations include a variety of species. EPA believes that there is merit in these comments. Accordingly, EPA is finalizing a broadened definition of “tree plantation,” by removing the limitations on the number and species of trees. EPA is defining tree plantation as “a stand of no less than 1 acre composed primarily of trees established by hand- or machine-planting of a seed or sapling, or by coppice growth from the stump or root of a tree that was hand- or machine-planted.”

We proposed to apply similar management restrictions to tree plantations as would apply to existing agricultural land and also to interpret the EISA language as requiring that to qualify as renewable biomass for renewable fuel production under RFS2, a tree plantation must have been cleared at any time prior to December 19, 2007, and continuously actively managed since December 19, 2007. Consistent with our final position regarding actively managed existing agricultural land, we are defining the term “actively managed” in the context of tree plantations as managed for a predetermined outcome as evidenced by any of the following that must be traceable to the land in question: Sales records for planted trees or slash; purchasing records for seeds, seedlings, or other nursery stock together with other written documentation connecting the land in question to these purchases; a written management plan for silvicultural purposes; documentation of participation in a silvicultural program sponsored by a Federal, state or local government agency; documentation of land management in accordance with an agricultural or silvicultural product certification program; an agreement for land management consultation with a professional forester that identifies the land in question; or evidence of the existence and ongoing maintenance of a road system or other physical infrastructure designed and maintained for logging use, together with one of the above-mentioned documents. Silvicultural programs such as those of the Forest Stewardship Council, the Sustainable Forestry Initiative, the American Tree Farm System, or USDA are examples of the types of programs that could indicate actively managed tree plantations. As with the definition of “actively managed” as it applies to crops from existing agricultural lands, we received extensive comments on this interpretation. As with our final position for crops from existing agricultural lands, we are interpreting the “active management” requirement for tree plantations to apply on the date of EISA's enactment, December 19, 2007. Those tree plantations that were cleared or cultivated and actively managed on December 19, 2007 are eligible for the production of planted trees, tree residue, slash and pre-commercial thinnings for renewable fuel production.

In lieu of the term “tree residue,” we proposed to use the term “slash” in our regulations as a more descriptive, but otherwise synonymous, term. According to the Dictionary of Forestry (1998, p. 168), a source of commonly understood industry definitions, slash is “the residue, e.g., treetops and branches, left on the ground after logging or accumulating as a result of a storm, fire, girdling, or delimbing.” We also proposed to clarify that slash can include tree bark and can be the result of any natural disaster, including flooding. We received comments in support of this additional inclusion and are expanding the definition of “slash” to include tree bark and residue resulting from natural disaster, including flooding. We received general support for our proposal to substitute our definition of “slash” for “tree residue,” however, several commenters argued that our definition of slash is too narrow to be substituted for “tree residue,” which should include woody residues from saw mills and paper mills that process planted trees from tree plantations. EPA agrees that the term “residue” should include this material. Therefore, EPA is expanding the definition of “tree residue” to include residues from processing planted trees at lumber and paper mills, but is limiting it to the biogenically derived portion of the residues that can be traced back to feedstocks meeting the definition of renewable biomass (i.e. planted trees and tree residue from actively managed tree plantations on non-federal land cleared at any time prior to December 19, 2007). RINs may only be generated for the fraction of fuel produced that represents the biogenic portion of the tree residue, using the procedures described in ASTM test method D-6866. Thus, if the tree residues are mixed with chemicals or other materials during processing at the lumber or paper mills, producers may only generate RINs for the portion of the mixture that is actually derived from planted trees. EPA's final definition of “tree residue” is “slash and any woody residue generated during the processing of planted trees from actively managed tree plantations for use in lumber, paper, furniture or other applications, providing that such woody residue is not mixed with similar residue from trees that do not originate in actively managed tree plantations.

iii. Slash and Pre-Commercial Thinnings

The EISA definition of renewable biomass includes slash and pre-commercial thinnings from non-federal forestlands, including forestlands belonging to an Indian tribe or an Indian individual, that are held in trust by the United States or subject to a restriction against alienation imposed by the United States. However, EISA excludes slash and pre-commercial thinnings from forests or forestlands that are ecological communities with a global or State ranking of critically imperiled, imperiled, or rare pursuant to a State Natural Heritage Program, old growth forest, or late successional forest.

As described in Sec. II.B.4.a.i of this preamble, our definition of “forestland” is generally undeveloped land covering a minimum of 1 acre upon which the primary vegetative species is trees, including land that formerly had such tree cover and that will be regenerated and tree plantations. Tree-covered areas in intensive agricultural crop production settings, such as fruit orchards or tree-covered areas in urban setting such as city parks, are not considered forestland. Also as noted in Sec. III.B.4.a.ii of this preamble, we are adopting the definition of slash listed in the Dictionary of Forestry, with the addition of tree bark and residue resulting from natural disaster, including flooding.

As for “pre-commercial thinnings,” the Dictionary of Forestry defines the act of such thinning as “the removal of trees not for immediate financial return but to reduce stocking to concentrate growth on the more desirable trees.” Because what may now be considered pre-commercial may eventually be saleable as renewable fuel feedstock, we proposed not to include any reference to “financial return” in our definition, but rather to define pre-commercial thinnings as those trees removed from a stand of trees in order to reduce stocking to concentrate growth on more desirable trees. Additionally, we proposed to include diseased trees in the definition of pre-commercial thinnings due to the fact that they can threaten the integrity of an otherwise healthy stand of trees, and their removal can be viewed as reducing stocking to promote the growth of more desirable trees. We sought comment on whether our definition of pre-commercial thinnings should include a maximum diameter and, if so, what the appropriate maximum diameter should be. We received comments on our proposed definition of pre-commercial thinnings that were generally supportive of our proposed definition. Many commenters argued that EPA should not use a maximum tree diameter as a basis for defining pre-commercial thinning as tree diameter varies greatly by forest type and location, making any diameter limitation EPA might set arbitrary. EPA agrees with this assessment. Commenters also argued that pre-commercial thinnings may include other non-tree vegetative material that is removed to promote and improve tree growth. EPA is attempting to utilize standard industry definitions to the extent practicable, and believes that the proposed definition of pre-commercial thinnings, based largely on the Dictionary of Forestry definition with the addition of other vegetative material removed to promote tree growth, is appropriate. Therefore, we are finalizing the proposed definition of “pre-commercial thinnings,” with the addition of the phrase “or other vegetative material that is removed to promote tree growth.”

We proposed that the State Natural Heritage Programs referred to in EISA are those comprising a network associated with NatureServe, a non-profit conservation and research organization. Individual Natural Heritage Programs collect, analyze, and distribute scientific information about the biological diversity found within their jurisdictions. As part of their activities, these programs survey and apply NatureServe's rankings, such as critically imperiled (S1), imperiled (S2), and rare (S3) to species and ecological communities within their respective borders. NatureServe meanwhile uses data gathered by these Natural Heritage Programs to apply its global rankings, such as critically imperiled (G1), imperiled (G2), or vulnerable (the equivalent of the term “rare,” or G3), to species and ecological communities found in multiple States or territories. We proposed and sought comment on prohibiting slash and pre-commercial thinnings from all forest ecological communities with global or State rankings of critically imperiled, imperiled, or vulnerable (“rare” in the case of State rankings) from being used for renewable fuel for which RINs may be generated under RFS2.

We proposed to use data compiled by NatureServe and published in special reports to identify “ecologically sensitive forestland.” The reports listed all forest ecological communities in the U.S. with a global ranking of G1, G2, or G3, or with a State ranking of S1, S2, or S3, and included descriptions of the key geographic and biologic attributes of the referenced ecological community. We proposed that the document be incorporated by reference into the definition of renewable biomass in the final RFS2 regulations (and updated as appropriate through notice and comment rulemaking). The document would identify specific ecological communities from which slash and pre-commercial thinnings could not be used as feedstock for the production of renewable fuel that would qualify for RINs under RFS2. Draft versions of the document containing the global and State rankings were placed in the docket for the proposed rule.

EPA received several comments on our proposed interpretation of EISA's State Natural Heritage Program requirement and the reports listing G1-G3 and S1-S3 ecological communities. Several commenters argued that while EISA authorizes EPA to exclude slash and pre-commercial thinnings from S1-3 and G1 and G2 communities, it does not authorize the exclusion of biomass from G3 communities, which are designated as “vulnerable,” not “critically imperiled, imperiled or rare,” as EISA requires. The commenters further argue that there is little or no environmental benefit to adding G3 communities to the list of lands unavailable for renewable fuel feedstock production, and that their inclusion limits the availability of forest-derived biomass. EPA agrees with these comments, and has drafted today's final rule so as not to specifically exclude from the definition of renewable biomass slash and pre-commercial thinnings from G3-ranked “vulnerable” ecological communities to qualify as renewable biomass for purposes of RFS2. We are interpreting EISA's language to exclude from the definition of renewable biomass any biomass taken from ecological communities in the U.S. with Natural Heritage Programs global ranking of G1 or G2, or with a State ranking of S1, S2, or S3. We are including in today's rulemaking docket (EPA-HQ-OAR-2005-0161) the list of ecological communities fitting this description.

To complete the definition of “ecologically sensitive forestland,” we proposed to include old growth and late successional forestland which is characterized by trees at least 200 years old. We received comments on this proposed definition recommending that EPA not use a single tree age in the define old growth and late-successional forests, as this criterion does not apply to all types of forests. While EPA understands that there are a number of criteria for determining whether a forest is old growth and that the criteria differ depending on the type of forest, for purposes of the RFS2 rule, EPA seeks to use definitive criteria that can be applied by non-professionals. EPA is finalizing the definition of “old growth” as proposed.

iv. Biomass Obtained From Certain Areas at Risk From Wildfire

The EISA definition of renewable biomass includes biomass obtained from the immediate vicinity of buildings and other areas regularly occupied by people, or of public infrastructure, at risk from wildfire. We proposed to clarify in the regulations that “biomass” is organic matter that is available on a renewable or recurring basis, and that it must be obtained from within 200 feet of buildings, campgrounds, and other areas regularly occupied by people, or of public infrastructure, such as utility corridors, bridges, and roadways, in areas at risk of wildfire.

Furthermore, we proposed to define “areas at risk of wildfire” as areas located within—or within one mile of—forestland, tree plantations, or any other generally undeveloped tract of land that is at least one acre in size with substantial vegetative cover. We sought comment on two possible implementation alternatives for identifying areas at risk of wildfire. The first proposed alternative would incorporate into our definition of “areas at risk of wildfire” any communities identified as “communities at risk” and covered by a community wildfire protection plan (CWPP). Communities at risk are defined through a process within the document, “Field Guidance—Identifying and Prioritizing Communities at Risk” (National Association of State Foresters, June 2003). CWPPs are developed in accordance with “Preparing a Community Wildfire Protection Plan—A Handbook for Wildland-Urban Interface Communities” (Society of American Foresters, March 2004) and certified by a State Forester or equivalent. We sought comment on incorporating by reference into the final RFS2 regulations a list of “communities at risk” with an approved CWPP. We also sought comment on a second implementation approach, which would incorporate into our definition of “areas at risk of wildfire” any areas identified as wildland urban interface (WUI) land, or land in which houses meet wildland vegetation or are mixed with vegetation. We noted that SILVIS Lab, in the Department of Forest Ecology and Management and the University of Wisconsin, Madison, has, with funding provided by the U.S. Forest Service, mapped WUI lands based on the 2000 Census and the U.S. Geological Survey National Land Cover Data (NLCD), and we sought comment on how best to use this map.

We received comments on the proposal and on the two proposed alternative options for identifying areas at risk of wildfire. A number of commenters argued that EPA should define “areas at risk of wildfire” using an existing definition of WUI from the Healthy Forests Restoration Act (Pub. L. 108-148). Many commenters recommended that EPA include both lands covered by a CWPP as well as lands meeting the Healthy Forests Restoration Act definition of WUI in order to maximize the amount of land available for biomass feedstock and to encourage the removal of hazardous fuel for wildfires. EPA understands that very few communities that might be eligible for a CWPP actually have one in place, due to the numerous administrative steps that must be taken in order to have a CWPP approved, so the option of defining areas at risk of wildfire exclusively by reference to a list of communities with an approved CWPP would be underinclusive of all lands that a professional forester would consider to be at risk of wildfire. Furthermore, EPA believes that the statutory definition of WUI from the Healthy Forests Restoration Act (Pub. L. 108-148) is too vague using directly in implementing the RFS2 program. If EPA used this WUI definition, individual plots of land would have to be assessed by a professional forester on a case-by-case basis in order to determine if they meet the WUI definition, creating an expensive burden for landowners seeking to sell biomass from their lands as renewable fuel feedstocks.

In light of the comments received and the need for a simple way for landowners and renewable fuel producers to track the status of particular plots of land, for the final rule we are identifying “areas at risk of wildfire” as those areas identified as wildland urban interface. Those areas are depicted and mapped at http://silvis.forest.wisc.edu/Library/WUILibrary.asp. The electronic WUI map is a readily accessible reference tool that was prepared by experts in the field of identifying areas at risk of wildfire, and is thus an ideal reference for purposes of implementing RFS2. EPA has included in the rulemaking docket instructions on using the WUI map to find the status of a plot of land.

v. Algae

EISA specifies that “algae” qualify as renewable biomass. EPA did not propose a definition for this term. A number of commenters have requested clarification, specifically asking whether cyanobacteria (also known as blue-green algae), diatoms, and angiosperms are within the definition. Technically, the term “algae” has recently been defined as “thallophytes (plants lacking roots, stems and leaves) that have chlorophyll a as their primary photosynthetic pigment and lack a sterile covering of cells around the reproductive cells.” [5] Algae are relatively simple organisms that are virtually ubiquitous, occurring in freshwater, brackish water, saltwater, and terrestrial habitats. When present in water, they may be suspended, or grow attached to various substrates. They range in size from unicellular to among the longest living organisms (e.g. sea kelp). There is some disagreement among scientists as to whether cyanobacteria should be considered bacteria or algae. Some consider them to be bacteria because of their cellular organization and biochemistry. However, others find it more significant that they contain chlorophyll a, which differs from the chlorophyll of bacteria which are photosynthetic, and also because free oxygen is liberated in blue-green algal photosynthesis but not in that of the bacteria. [6] EPA believes that it furthers the purposes of EISA to interpret the term “algae” in EISA broadly to include cyanobacteria, since doing so will make available another possible feedstock for renewable fuel production that will further the energy independence and greenhouse gas reduction objectives of the Act. Further, EPA expects that cyanobacteria used in biofuel production would be cultivated, as opposed to harvested, and therefore that there would be no significant impact from use of cyanobacteria for biofuel production on naturally occurring algal populations. Diatoms are generally considered by the scientific community to be algae, [7] and, consistent with this general scientific consensus, EPA interprets the EISA definition of algae to include them. Microcrop angiosperms, however, do not meet the definition of algae, even if they live in an aquatic habitat, since they are relatively more complex organisms than the algae. A discussion of microcrop angiosperms is included above in the discussion of “planted crops and crop residue.”

b. Implementation of Renewable Biomass Requirements

Our proposed approach to the treatment of renewable biomass under RFS2 was intended to define the conditions under which RINs can be generated as well as the conditions under which renewable fuel can be produced or imported without RINs. Our proposed and final approaches to both of these areas are described in more detail below.

i. Ensuring That RINs Are Generated Only for Fuels Made From Renewable Biomass

The effect of adding EISA's definition of renewable biomass to the RFS program is to ensure that renewable fuels are only eligible for the program if made from certain feedstocks, and if some of those feedstocks come from certain types of land. In the context of our regulatory program, this means that RINs could only be generated if it can be established that the feedstock from which the fuel was made meets EISA's definitions of renewable biomass include land restrictions. Otherwise, no RINs could be generated to represent the renewable fuel produced or imported. The EISA language does not distinguish between domestic renewable fuel feedstocks and renewable fuel feedstocks that come from abroad, so our final rule requires similar feedstock affirmation and recordkeeping requirements for both RIN-generating domestic renewable fuel producers and RIN-generating foreign producers or importers.

We acknowledge that incidental contaminants can be introduced into feedstocks during cultivation, transport or processing. It is not EPA's intent that the presence of such contaminants should disqualify the feedstock as renewable biomass. The final regulations therefore stipulate that the term “renewable biomass” includes incidental contaminants related to customary feedstock production and transport that are present in feedstock that otherwise meets the definition if such incidental contaminants are impractical to remove and occur in de minimus levels. By “related to customary feedstock production and transport,” we refer to contaminants related to crop production, such as soil or residues related to fertilizer, pesticide and herbicide applications to crops, as well as contaminants related to feedstock transport, such as nylon rope used to bind feedstock materials. It would also include agricultural contaminants introduced to the feedstock during sorting or shipping, such as miscellaneous sorghum grains present in a load of corn kernels. However, contamination is not related to customary feedstock production and transport, so such feedstocks would not qualify, and in particular, any hazardous waste or toxic chemical contaminant in feedstock would disqualify the feedstock as renewable biomass.

ii. Whether RINs Must Be Generated for All Qualifying Renewable Fuel

Under RFS1, virtually all renewable fuel is required to be assigned a RIN by the producer or importer. This requirement was developed and finalized in the RFS1 rulemaking in order to address stakeholder concerns, particularly from obligated parties, that the number of available RINs should reflect the total volume of renewable fuel used in the transportation sector in the U.S. and facilitate program compliance. EISA has dramatically increased the mandated volumes of renewable fuel that obligated parties must ensure are produced and used in the U.S. At the same time, EISA makes it more difficult for renewable fuel producers to demonstrate that they have fuel that qualifies for RIN generation by restricting qualifying renewable fuel to that made from “renewable biomass.” The inclusion of such restrictions under RFS2 may mean that, in some situations, a renewable fuel producer would prefer to forgo the benefits of RIN generation to avoid the cost of ensuring that its feedstocks qualify for RIN generation. If a sufficient number of renewable fuel producers acted in this way, it could lead to a situation in which not all qualifying fuel is assigned RINs, thus resulting in a shortage of RINs in the market that could force obligated parties into non-compliance even though biofuels are being produced and used. Another possible outcome would be that the demand for and price of RINs would increase significantly, making compliance by obligated parties more costly and difficult than necessary and raising prices for consumers.

With these concerns in mind, EPA proposed to preserve in RFS2 the RFS1 requirement that RINs be generated for all qualifying renewable fuel. We also proposed that renewable fuel producers maintain records showing that they utilized feedstocks made from renewable biomass if they are generating RINs, or, if they are not generating RINs, that they did not use feedstocks that qualify as renewable biomass. However, we considered this matter further, and we realize that the implication of these proposed requirements is that renewable fuel producers would be caught in the untenable position of being forced to participate in the RFS2 program (register, keep records, etc.) even if they are unable to generate RINS because their feedstocks do not meet the definition of renewable biomass. We received many comments on the proposed requirement to generate RINs for all qualifying renewable fuel. Most commenters argued that the requirement to keep records for non-qualifying renewable fuels was excessively onerous and served little purpose for the program.

After considering the comments received, EPA has determined that this requirement would be overly burdensome and unreasonable for producers. The burden stems from the requirement that producers prove that their feedstocks do not qualify if they are not generating RINs. If the data did not exist or could not be obtained, producers could not produce the fuel, even if no RINs would be generated. Thus, for the final rule, EPA is requiring only that producers that do generate RINs have the requisite records (as discussed in section II.B.4.c.i. of this preamble) documenting that their fuel is produced from feedstocks meeting the definition of renewable biomass. Non-RIN generating producers need not maintain any paperwork related to their feedstocks and their origins.

Although EPA is not requiring that RINs be generated for all qualifying renewable fuel, EPA is seeking to avoid situations where biofuels are produced, but RINs are not made available to the market for compliance. EPA received comments requesting that we consider a provision in which any volume of renewable fuel for which RINs were not generated would be an obligated volume for that producer, to serve as a disincentive for those producers who might not generate RINs in order to avoid the RFS program requirements. While EPA is not finalizing this provision in today's rule, we may consider a future rulemaking to promulgate a provision such as this if we find that EISA volumes are not being met due to producers declining to generate RINs for their qualifying renewable fuel. We also note that it is ultimately the availability of qualifying renewable fuel, as determined in part by the number of RINs in the marketplace, that will determine the extent to which EPA should issue a waiver of RFS requirements on the basis of inadequate domestic supply. It is in the interest of renewable fuel producers to avoid a situation where a waiver of the EISA volume requirements appears necessary. EPA encourages renewable fuel producers to generate RINs for all fuel that is made from feedstocks meeting the definition of renewable biomass and that meets the GHG emissions reduction thresholds set out in EISA. Please see section II.D.6 for additional discussion of this issue.

c. Implementation Approaches for Domestic Renewable Fuel

Consistent with RFS1, renewable fuel producers will be responsible for generating Renewable Identification Numbers (RINs) under RFS2. In order to determine whether or not their fuel is eligible for generating RINs, renewable fuel producers will generally need to have at least basic information about the origin of their feedstocks, to ensure they meet the definition of renewable biomass. In the proposal, EPA described and sought comment on several approaches for implementing the land restrictions on renewable biomass contained in EISA.

The proposed approach for ensuring that producers generate RINs properly was that EPA would require that renewable fuel producers obtain documentation about their feedstocks from their feedstock supplier(s) and take the measures necessary to ensure that they know the source of their feedstocks and can demonstrate to EPA that they fall within the EISA definition of renewable biomass. EPA would require renewable fuel producers who generate RINs to affirm on their renewable fuel production reports that the feedstock used for each renewable fuel batch meets the definition of renewable biomass. EPA would also require renewable fuel producers to maintain sufficient records to support these claims. Specifically, we proposed that renewable fuel producers who use planted crops or crop residue from existing agricultural land, or who use planted trees or slash from actively managed tree plantations, would be required to have copies of their feedstock producers' written records that serve as evidence of land being actively managed (or fallow, in the case of agricultural land) since December 2007, such as sales records for planted crops or trees, livestock, crop residue, or slash; a written management plan for agricultural or silvicultural purposes; or, documentation of participation in an agricultural or silvicultural program sponsored by a Federal, state or local government agency. In the case of all other biomass, we proposed to require renewable fuel producers to have, at a minimum, written records from their feedstock supplier that serve as evidence that the feedstock qualifies as renewable biomass.

We sought comment on this approach generally as well as other methods of verifying renewable fuel producers' claims that feedstocks qualify as renewable biomass. EPA received extensive comments on the proposed approach. Many affected parties argued that the proposed approach would pose an unnecessary recordkeeping burden on both feedstock and renewable fuel producers when, in practice, new lands will not be cleared, at least in the near future, for purposes of growing renewable fuel feedstocks. Commenters argued that individual recordkeeping was onerous, when compliance with the renewable biomass requirements could be determined through the use of existing data and third-party programs. Commenters contend that the recordkeeping and feedstock tracking requirements are particularly arduous for corn, soybeans and other agricultural crops that are used as renewable fuel feedstocks due to both the maturity and the highly fungible nature of those feedstock systems. In contrast, other commenters argued that recordkeeping and reporting requirements are necessary to ensure that feedstocks are properly verified as renewable biomass to prevent undesirable impacts on natural ecosystems and wildlife habitat globally.

We also sought comment on the possible use under EISA of non-governmental, third-party verification programs used for certifying and tracking agricultural and forest products from point of origin to point of use both within the U.S. and outside the U.S. We examined third-party organizations that certify specific types of biomass from croplands and organizations that certify forest lands, including the Roundtable on Sustainable Palm Oil, the Basel Criteria for Responsible Soy Production, the Roundtable on Sustainable Biofuels (RSB) and the Better Sugarcane Initiative (BSI). Additionally, we examined the work of the international Soy Working Group, the Brazilian Association of Vegetable Oil Industries (ABIOVE) and Brazil's National Association of Grain Exporters (ANEC), Greenpeace, Verified Sustainable Ethanol initiative, the Sustainable Agriculture Network (SAN), the Forest Stewardship Council (FSC), American Tree Farm program and Sustainable Forestry Initiative (SFI). We proposed not to solely rely on any existing third-party verification program to implement the land restrictions on renewable biomass under RFS2 for several reasons. These programs are limited in the scope of products they certify, the acreage of land certified through third parties in the U.S. covers only a small portion of the total available land estimated to qualify for renewable biomass production under the EISA definition, and none of the existing third-party systems had definitions or criteria that perfectly match the land use definitions and restrictions contained in the EISA definition of renewable biomass.

We received several comments indicating that producers would like to use evidence of their participation in these types of programs to prove that their feedstocks meet the definition of renewable biomass. Others argued that while, at this time, the requirements of third-party programs may not encompass all of the restrictions and requirements of EISA's renewable biomass definition, the programs may alter their criteria in the future to parallel EISA's requirements. EPA agrees that this is a possibility and, in the future, will consider the use of these programs in order to simplify compliance with the renewable biomass requirements. We encourage fuel producers to work to identify changes to such programs that could allow them to be used as a viable compliance option.

In the proposal, EPA also acknowledged that land restrictions contained within the definition of renewable biomass may not, in practice, result in a significant change in agricultural practices, since biomass from nonqualifying lands may still be used for non-fuel (e.g., food) purposes. Therefore, we sought comment on a stakeholder suggestion to establish a baseline level of production of biomass feedstocks such that reporting and recordkeeping requirements would be triggered only when the baseline production levels of feedstocks used for biofuels were exceeded. Additionally, EPA offered as an alternative the use of existing satellite and aerial imagery and mapping software and tools to implement the renewable biomass provisions of EISA. We received numerous comments in support of these options. Commenters argued that USDA collects and maintains ample data on land use that EPA could use to demonstrate that, due to increasing crop yields and other considerations, agricultural land acreage will not expand, at least in the near term, to accommodate the increased renewable fuel obligations of RFS2.

EPA also sought comment on an additional alternative in which EPA would require renewable fuel producers to set up and administer a company-wide quality assurance program that would create an additional level of rigor in the implementation scheme for the EISA land restrictions on renewable biomass. EPA is not finalizing this company-wide quality assurance program approach, but rather, is encouraging the option for an industry-wide quality assurance program, as described in the following section, to be administered.

i. Recordkeeping and Reporting for Feedstocks

After considering the comments we received on the proposed approach, EPA is finalizing reporting and recordkeeping requirements comparable to those in the approach we discussed in the proposed rule for all categories of renewable biomass, with the exception of planted crops and crop residue from agricultural land in the United States, which will be covered by the aggregate compliance approach discussed below in Section II.B.4.c.iii. EPA believes that these requirements on the fuel producer utilizing feedstocks other than crops and crop residue are necessary to ensure that the definition of renewable biomass is being met, and to allow feedstocks to be traced from their original producer to the renewable fuel production facility. Furthermore, we believe that, in most cases, feedstock producers will already have or will be able to easily generate the specified documentation for renewable fuel producers necessary to provide them with adequate assurance that the feedstock in question meets the definition of renewable biomass.

Under today's rule, all renewable fuel producers must maintain written records from their feedstock suppliers for each feedstock purchase that identify the type and amount of feedstocks and where the feedstock was produced and that are sufficient to verify that the feedstock qualifies as renewable biomass. Specifically, renewable fuel producers must maintain maps and/or electronic data identifying the boundaries of the land where the feedstock was produced, product transfer documents (PTDs) or bills of lading tracing the feedstock from that land to the renewable fuel production facility, and other written records that serve as evidence that the feedstock qualifies as renewable biomass. We believe the maps or electronic data can be easily generated using existing Web-based information.

Producers using planted trees and tree residue from tree plantations must maintain additional documentation that serves as evidence that the tree plantation was cleared prior to December 19, 2007, and actively managed as a tree plantation on December 19, 2007. This documentation must consist of the following types of records which must be traceable to the land in question: Sales records for planted trees or slash; purchasing records for fertilizer, weed control, or reseeding, including seeds, seedlings, or other nursery stock together with other written documentation connecting the land in question to these purchases; a written management plan for silvicultural purposes; documentation of participation in a silvicultural program sponsored by a Federal, state or local government agency; or documentation of land management in accordance with a silvicultural product certification program; an agreement for land management consultation with a professional forester that identifies the land in question; or evidence of the existence and ongoing maintenance of a road system or other physical infrastructure designed and maintained for logging use. There are many existing programs, such as those administered by USDA and independent third-party certifiers, that could be used as documentation that verifies that feedstock from certain land qualifies as renewable biomass. For example, many tree plantation owners already participate in a third-party certification program such as FSC or SFI. Written proof of participation by a tract of land in a program of this type on December 19, 2007 would be sufficient to show that a tree plantation was cleared prior to that date and that it was actively managed on that date. The tree plantation owner would need to send copies of this documentation to the renewable fuel producer when supplying them with biomass that will be used as a renewable fuel feedstock.

We anticipate that the recordkeeping requirements will result in renewable fuel producers amending their contracts and modifying their supply chain interactions to satisfy the requirement that producers have documented assurance and proof about their feedstock's origins. Enforcement will rely in part on EPA's review of renewable fuel production reports and attest engagements of renewable fuel producers' records. EPA will also consult other data sources, including any data made available by USDA, and may conduct site visits or inspections of feedstock producers' and suppliers' facilities.

The reporting requirements for renewable biomass in today's final rule include, as proposed, include an affirmation by the renewable fuel producer for each batch of renewable fuel for which they generate RINs that the feedstocks used to produce the batch meet the definition of renewable biomass. Additionally, the final reporting requirements include a quarterly report to be sent to EPA by each renewable fuel producer that includes a summary of the types and volumes of feedstocks used throughout the quarter, as well as electronic data or maps identifying the land from which those feedstocks were harvested. Producers need not provide duplicate maps if purchasing feedstocks multiple times from one plot of land; producers may cross-reference the previously submitted map. Producers will also be required to keep records tracing the feedstocks from the land to the renewable fuel production facility, other written records from their feedstock suppliers that serve as evidence that the feedstock qualifies as renewable biomass, and for producers using planted trees or tree residue from tree plantations, written records that serve as evidence that the land from which the feedstocks were obtained was cleared prior to December 19, 2007 and actively managed on that date. These requirements will apply to renewable fuel producers using feedstocks from foreign sources (unless special approvals are granted in the future, as described below), or from domestic sources, except for planted crops or crop residue (discussed below).

This approach will be integrated into the existing registration, recordkeeping, reporting, and attest engagement procedures for renewable fuel producers. It places the burden of implementation and enforcement on renewable fuel producers rather than bringing feedstock producers and suppliers directly under EPA regulation, minimizing the number of regulated parties under RFS2.

EPA also sought comment on, and is finalizing as an option, an alternative approach in which EPA allows renewable fuel producers and renewable fuel feedstock producers and suppliers to develop a quality assurance program for the renewable fuel production supply chain, similar to the model of the successful Reformulated Gasoline Survey Association. While individual renewable fuel producers may still choose to comply with the individual renewable biomass recordkeeping and reporting requirements rather than participate in a quality assurance program, we believe that this preferred alternative could be less costly than an individual compliance demonstration, and it would add a quality assurance element to RFS2. Those participating renewable fuel producers would be presumed to be in compliance with the renewable biomass requirements unless and until the quality assurance program finds evidence to the contrary. Under today's rule, renewable fuel producers must choose either to comply with the individual renewable biomass recordkeeping and reporting described above, or they must participate in the quality assurance program.

The quality assurance program must be carried out by an independent auditor funded by renewable fuel producers and feedstock suppliers. The program must consist of a verification program for participating renewable fuel producers and renewable feedstock producers and handlers designed to provide independent oversight of the feedstock handling processes that are required to determine if a feedstock meets the definition of renewable biomass. Under this option, a participating renewable fuel producer and its renewable feedstock suppliers and handlers would have to participate in the funding of an organization which arranges to have an independent auditor conduct a program of compliance surveys. The compliance audit must be carried out by an independent auditor pursuant to a detailed survey plan submitted to EPA for approval by November 1 of the year preceding the year in which the alternative compliance program would be implemented. The compliance survey program plan must include a statistically supportable methodology for the survey, the locations of the surveys, the frequency of audits to be included in the survey, and any other elements that EPA determines are necessary to achieve the same level of quality assurance as the individual recordkeeping and reporting requirements included in the RFS2 regulations.

Under this alternative compliance program, the independent auditor would be required to visit participating renewable feedstock producers and suppliers to determine if the biomass they supply to renewable fuel producers meets the definition of renewable biomass. This program would be designed to ensure representative coverage of participating renewable feedstock producers and suppliers. The auditor would generate and report the results of the surveys to EPA each calendar quarter. In addition, where the survey finds improper designations or handling, the renewable fuel producers would be responsible for identifying and addressing the root cause of the problem. The renewable fuel producers would have to take corrective action to retire the appropriate number of invalid RINs depending on the violation. EPA received comments from a number of parties who were supportive of this option as an alternative and less-burdensome way of ensuring that renewable fuel feedstocks meet the definition of renewable biomass. EPA believes this option to be an efficient and effective means of implementing and enforcing the renewable biomass requirements of EISA, and has therefore included it as a compliance option in today's final rule.

ii. Approaches for Foreign Producers of Renewable Fuel

The EISA renewable biomass language does not distinguish between domestic renewable fuel and fuel feedstocks and renewable fuel and fuel and feedstocks that come from abroad. EPA proposed that foreign producers of renewable fuel that is exported to the U.S. be required to meet the same compliance obligations as domestic renewable fuel producers, as well as some additional measure, discussed in Section II.C., designed to facilitate EPA enforcement in other countries. These proposed obligations include facility registration and submittal of independent engineering reviews (described in Section II.C below), and reporting, recordkeeping, and attest engagement requirements. The proposal also would have included for foreign producers the same obligations that domestic producers have for verifying that their feedstock meets the definition of renewable biomass, such as certifying on each renewable fuel production report that their renewable fuel feedstock meets the definition of renewable biomass and working with their feedstock suppliers to ensure that they receive and maintain accurate and sufficient documentation in their records to support their claims.

(1) RIN-Generating Importers

EPA proposed to allow importers to generate RINs for renewable fuel they are importing into the U.S. only if the foreign producer of that renewable fuel had not already done so. Under the proposal, in order to generate RINs, importers would need to obtain information from the registered foreign producers concerning the point of origin of their fuel's feedstock and whether it meets the definition of renewable biomass. Therefore, we proposed that in the event that a batch of foreign-produced renewable fuel does not have RINs accompanying it when it arrives at a U.S. port, an importer must obtain documentation that proves that the fuel's feedstock meets the definition of renewable biomass (as described in Section II.B.4.a. of this preamble) from the fuel's producer, who must have registered with the RFS program and conducted a third-party engineering review. With such documentation, the importer could generate RINs prior to introducing the fuel into commerce in the U.S.

We sought comment on this proposed approach and whether and to what extent the approaches for ensuring compliance with the EISA's land restrictions by foreign renewable fuel producers should differ from the proposed approach for domestic renewable fuel producers. We received comments on the proposed implementation option for importers of foreign renewable fuel. Some argue that the proposed recordkeeping requirements for imported fuel were overly burdensome. On the other hand, others argued that importers, similarly to domestic producers, should be required to obtain information that can serve as evidence that the feedstocks meet the definition of renewable biomass, in order to avoid fraud. Some commenters also argued that importers should be able to generate RINs for fuel imported from foreign producers that are not registered with EPA under the RFS2 program.

For the final rule, EPA is requiring that importers may only generate RINs for renewable fuel if the foreign producer has not already done so. The foreign producers must be registered with EPA under the RFS2 program, and must have conducted an independent engineering review. Furthermore, we are requiring that importers obtain from the foreign producer and maintain in their records written documentation that serves as evidence that the renewable fuel for which they are generating RINs was made from feedstocks meeting the definition of renewable biomass. The foreign producer that originally generated the fuel must ensure that these feedstock records are transferred with each batch of fuel and ultimately reach the RIN-generating importer. A requirement that importers maintain these renewable biomass records is consistent with the renewable biomass recordkeeping requirements imposed on domestic producers of renewable fuel.

(2) RIN-Generating Foreign Producers

Foreign producers that intend to generate RINs would be required to designate renewable fuel intended for export to the U.S. as such, segregate the volume until it reaches the U.S., and post a bond to ensure that penalties can be assessed in the event of a violation, as discussed in Section II.D.2.b. Similarly to domestic producers of renewable fuel, foreign producers must obtain and maintain written documentation from their feedstock providers that can serve as evidence that their feedstocks meet the definition of renewable biomass. Foreign producers may also develop a quality assurance program for their renewable fuel production supply chain, as described above. However, while domestic renewable fuel producers using crops or crop residues may rely on the aggregate compliance approach described below to ensure that their feedstocks are renewable biomass, this approach is not available at this time to foreign renewable fuel producers, as described below.

EPA believes that the renewable biomass recordkeeping provisions are necessary in order for EPA to ensure that RINs are being generated for fuel that meets EISA's definition of renewable fuel. Just as for domestic producers, foreign producers must maintain evidence that the fuel meets the GHG reduction requirements and is made from renewable biomass.

iii. Aggregate Compliance Approach for Planted Crops and Crop Residue From Agricultural Land

In light of the comments received on the proposed renewable biomass recordkeeping requirements and implementation options, EPA sought assistance from USDA in determining whether existing data and data sources might suggest an alternative method for verifying compliance with renewable biomass requirements associated with the use of crops and crop residue for renewable fuel production. Taking into consideration publicly available data on agricultural land available from USDA and USGS as well as expected economic incentives for feedstock producers, EPA has determined that an aggregate compliance approach is appropriate for certain types of renewable biomass, namely planted crops and crop residue from the United States.

Under the aggregate compliance approach, EPA is determining for this rule the total amount of “existing agricultural land” in the U.S. (as defined above in Section II.B.4.a.) at the enactment date of EISA, which is 402 million acres. EPA will monitor total agricultural land annually to determine if national agricultural land acreage increases above this 2007 national aggregate baseline. Feedstocks derived from planted crops and crop residues will be considered to be consistent with the definition of renewable biomass and renewable fuel producers using these feedstocks will not be required to maintain specific renewable biomass records as described below unless and until EPA determines that the 2007 national aggregate baseline is exceeded. If EPA finds that the national aggregate baseline is exceeded, individual recordkeeping and reporting requirements as described below will be triggered for renewable fuel producers using crops and crop residue. We believe that the aggregate approach will fully ensure that the EISA renewable biomass provisions related to crops and crop residue are satisfied, while also easing the burden for certain renewable fuel producers and their feedstock suppliers vis-à-vis verification that their feedstock qualifies as renewable biomass.

As discussed in more detail below, there are five main factors supporting the aggregate compliance approach we are taking for planted crops and crop residue. First, EPA is using data sets that allow us to obtain an appropriately representative estimate of the agricultural lands available under EISA for the production of crops and crop residue as feedstock for renewable fuel production. Second, USDA data indicate an overall trend of agricultural land contraction. These data, together with EPA economic modeling, suggest that 2007 aggregate baseline acreage should be sufficient to support EISA renewable fuel obligations and other foreseeable demands for crop products, at least in the near term, without clearing and cultivating additional land. Third, EPA believes that existing economic factors for feedstock producers favor more efficient utilization practices of existing agricultural land rather than converting non-agricultural lands to crop production. Fourth, if, at any point, EPA finds that the total amount of land in use for the production of crops including crops for grazing and forage is equal or greater than 397 million acres (i.e. within 5 million acres of EPA's established 402 million acre baseline), EPA will conduct further investigations to evaluate whether the presumption built into the aggregate compliance approach remains valid. Lastly, EPA has set up a trigger mechanism that in the event there are more than the baseline amount of acres of cropland, pastureland and CRP land in production, renewable fuel producers will be required to meet the same individual or consortium-based recordkeeping and reporting requirements applicable to RIN-generating renewable fuel producers using other feedstocks. Taken together, these factors give EPA high confidence that the aggregate compliance approach for domestically grown crops and crop residues meets the statutory obligation to ensure feedstock volumes used to meet the renewable fuel requirements also comply with the definition of renewable biomass.

(1) Analysis of Total Agricultural Land in 2007

As described in Section II.B.4.a. above, EPA is defining “existing agricultural land” for purposes of the EISA land use restrictions on crops and crop residue to include cropland, pastureland and CRP land that was cleared and actively managed or fallow and nonforested on the date of EISA enactment. To determine the aggregate total acreage of existing agricultural land for the aggregate compliance approach on the date of EISA enactment, EPA obtained from USDA data representing total cropland (including fallow cropland), pastureland, and CRP land in 2007 from three independently gathered national land use data sources (discussed in further detail below): The Farm Service Agency (FSA) Crop History Data, the USDA Census of Agriculture (2007), and the satellite-based USDA Crop Data Layer (CDL). In addition, CRP acreage is provided by FSA's annually published “Conservation Reserve Program: Summary and Enrollment Statistics.” By definition, the cropland, pastureland, and CRP land included in these data sources for 2007 were cleared or cultivated on the date of EISA enactment (December 19, 2007) and, consistent with the principles set forth in Section II.4.a.i, would be considered “actively managed” or fallow and nonforested on that date. These categories of lands include those from which traditional crops, such as corn, soy, wheat and sorghum, would likely be grown. Therefore quantification of cropland, pastureland, and CRP land from these data sources represents a reasonable assessment of the acreage in the United States that is available under the Act for the production of crops and crop residues that could satisfy the definition of renewable biomass in EISA.

Conservation Reserve Program Data. FSA reports CRP enrollment acreage each year in the publication “Conservation Reserve Program: Summary and Enrollment Statistics.” The CRP program includes the general CRP, the Conservation Reserve Enhancement Program (CREP), and the Farmable Wetlands Program (FWP). The Wetlands Reserve Program (WRP) and Grasslands Reserve Program (GRP) are not under CRP and are not included in the total agricultural land figure in this rulemaking. The 2007 CRP acreage was 36.7 million acres. This is an exact count of acreage within the CRP program in 2007.

Farm Service Agency Crop History Data. The FSA maintains annual records of field-level land use data for all farms enrolled in FSA programs. Almost all national cropland and pastureland is reported through FSA and recorded in this data set. We used the “Cropland” category to determine total agricultural land. Pastureland is reported by farms under the category “Cropland” as cropland used for grazing and forage under the crop type “mixed forage.” Timber land and any grazed native grass was removed from the “Cropland” category, because these land types represent either forestland or rangeland, which are not within the definition of existing agricultural land. CRP lands and other conservation program lands are also reported as cropland. Because GRP and WRP lands are not within the definition of “existing agricultural land” as defined in today's regulations, they were also subtracted from the “Cropland” category total. FSA Crop History Data show that there was 402 million acres of agricultural land, as defined here, in the U.S. in 2007 (See Table II.B.4-1).

Table II.B.4-1—Total U.S. Agricultural Land in 2007 From USDA Data Sources Back to Top
Land category FSA crop history data Agricultural census data
Cropland and Pastureland 365 367
CRP Land 37 37
Total Land 402 404

USDA Census of Agriculture. USDA conducts a full census of the U.S. agricultural sector once every five years. The data are available for the U.S., each of the 50 States, and for each county. The most recent census available is the 2007 Census of Agriculture. For the purpose of this rulemaking, USDA provided EPA total acreage and 95% confidence intervals for the Census category “Total Cropland,” which includes the sub-categories “Harvested cropland,” “Cropland used only for pasture and grazing,” and “Other cropland.” WRP and GRP acreage are included in “Other cropland,” so, for purposes of this rulemaking, they were subtracted from the sub-category number (see above). The analysis excluded the “Permanent rangeland and pasture” category, as the pasture data cannot be separated from rangeland in this category. Total CRP acreage in 2007 was added to “Total cropland.” With these adjustments, the Census of Agriculture showed 404 million acres (95% confidence range 401-406 million acres) of existing agricultural land as defined in today's rule, in the U.S. in 2007 (See Table II.B.4-1).

Crop Data Layer. The USDA National Agricultural Statistics Service (NASS) Crop Data Layer (CDL) is a raster, geo-referenced, crop-specific land cover data layer suitable for use in geographic information systems (GIS) analysis. Based on satellite data, the CDL has a ground resolution of 56 meters and was verified using FSA surveys. The CDL covers 21 major agricultural states for 2007 and therefore cannot be used to determine a 2007 national aggregate agricultural land baseline. There will be full coverage of the 48 contiguous states for 2009, and the CDL can be used for analysis validation purposes during monitoring. From 2010 onward, it coverage of the 48 contiguous states will be dependent on available funding. GIS analyses of the CDL will include all cropland and pastureland data for each state. To ensure that non-pasture grasslands are not included in the final sum, all areas of the “Grassland herbaceous” category from the U.S. Geological National Land Cover Data layer (NLCD) that overlap the CDL layers are removed from the total agricultural land number. Producer and user accuracies [8] are available for the CDL crop categories.

Primary Data Source Selection for Aggregate Compliance Approach. EPA has determined that the FSA Crop History Data will be used as the data set on which the total existing agricultural land baseline will be based for the aggregate compliance approach. The FSA Crop History Data is the only complete data set for 2007 that is collected annually, enabling EPA to monitor agricultural land expansion or contraction from year to year using a consistent data set. The total existing agricultural land value derived from FSA Crop History Data rests within the 95% confidence interval of the 2007 Census of Agriculture and is only 2 million acres less than the Census of Agriculture point estimate. The Census of Agriculture provides slightly fuller coverage than the FSA Crop History Data due to the nature of the data collection; however, given that both data collection systems have consistent and long-standing methodologies, the disparity between the two should remain approximately constant. Therefore, the FSA Crop History Data will provide a consistent data set for analyzing any expansion or contraction of total national agricultural land in the U.S.

During its annual monitoring, EPA will use the FSA Crop History Data and the CDL analyses as a secondary source to validate our annual assessment. In years when the Census of Agriculture is updated, this data will also be used to validate our annual assessment. Other data sources, such as the annual NASS Farms, Land in Farms and Livestock Operations may also be useful as secondary data checks. Lastly, EPA intends to consider, as appropriate, other data sources for the annual monitoring analysis of total agricultural land as new technologies and data sources come online that would improve the accuracy and robustness of annual monitoring.

(2) Aggregate Agricultural Land Trends Over Time

The Census of Agriculture (conducted every five years) shows that U.S. agricultural land has decreased by 44 million acres from 1997 to 2007, indicating an overall decade trend of contraction of agricultural land utilization despite some year-to-year variations that can be seen by reference to the annual FSA Crop History records (See Table II.B.4-2 and Table II.B.4-3). EPA's FASOM modeling results, which model full EISA volumes in 2022, support this contraction trend, indicating that total cropland, pastureland, and CRP land in the U.S. in 2022, under a scenario of full renewable fuel volume as required by EISA, would be less than the 2007 national acreage reported in the FSA Crop History Data (See preamble Section VII and RIA Chapter 5).

Table II.B.4-2—Total Agricultural Land (as Defined in Section II.B.4. a) Counted in the Census of Agriculture From 1997-2007 Back to Top
Census year Total agricultural land (millions of acres)
*2002 data do not include farms with land in FWP or CREP.
2007 404
2002 * 431
1997 * 445
Table II.B.4-3—Total Agricultural Land (as Defined in Section II.B.4. a) Recorded in FSA Crop History Data From 2005-2007 Back to Top
Year Total agricultural land (millions of acres)
2007 402
2006 393
2005 392

(3) Aggregate Compliance Determination

The foundation of the aggregate compliance approach is establishment of a baseline amount of eligible agricultural land that was cleared or cultivated and actively managed or fallow and non-forested on December 19, 2007. Based on USDA-FSA Crop History Data, EPA is establishing a baseline of 402 million acres of U.S. agricultural land, as defined in Section II.B.4.a and based upon the methods described in Section II.B.4.c.iii.(1), that is eligible for production of planted crops and crop residue meeting the EISA definition of renewable biomass. EPA will monitor total U.S. agricultural land annually, using FSA Crop History Data as a primary determinant, but using other data sources for support (See Section II.4.c.iii.(1)). If, at any point, EPA finds that the total land in use for the production of crops, including crops for grazing and forage, is greater than 397 million acres (i.e. within 5 million acres of EPA's established 402 million acre baseline), EPA will conduct further investigations to evaluate whether the presumption built into the aggregate compliance approach remains valid. Additionally, if EPA determines that the data indicates that this 2007 baseline level of eligible agricultural land has been exceeded, EPA will publish in the Federal Register a finding to that effect, and additional requirements will be triggered for renewable fuel producers to verify that they are using planted crops and crop residue from “existing agricultural land” as defined in today's rule as their renewable fuel feedstock. EPA's findings will be published by November 30, at the latest. If in November the 402 million acres baseline is found to be exceeded, then on July 1 of the following year, renewable fuel producers using feedstocks qualifying for this aggregate compliance approach, namely planted crops and crop residue from the United States, will be required to comply with the recordkeeping and reporting requirements applicable to producers using other types of renewable biomass, as described in the previous sections. This includes the option that fuel producers could utilize a third-party consortium to demonstrate compliance.

EPA acknowledges that it is possible that under this approach some of the land available under EISA for crop production on the date of EISA enactment could be retired and other land brought into production, without altering the assessment of the aggregate amount of cropland, pastureland and CRP land. Under EISA, crops or crop residues from the new lands would not qualify as renewable biomass. However, EPA expects such shifts in acreage to be de minimus, as long as the total aggregate amount of agricultural land does not exceed the 2007 national aggregate baseline. EPA expects that new lands are unlikely to be cleared for agricultural purposes for two reasons. First, it can be assumed that most undeveloped land that was not used as agricultural land in 2007 is generally not suitable for agricultural purposes and would serve only marginally well for production of renewable fuel feedstocks. Due to the high costs and significant inputs that would be required to make the non-agricultural land suitable for agricultural purposes, it is highly unlikely that farmers will undertake the effort to “shift” land that is currently non-agricultural into agricultural use. Second, crop yields are projected to increase, reducing the need for farmers to clear new land for agricultural purposes. We believe that this effect is reflected in the overall trend, discussed earlier, of an overall contraction in agricultural land acreage over time.

If EPA determines that the baseline is exceeded, and that individual compliance with the renewable biomass reporting and recordkeeping requirements is triggered, renewable fuel producers using crops and crop residue as a feedstock for renewable fuel would become responsible, beginning July 1 of the following year, for meeting individual recordkeeping and reporting requirements related to renewable biomass verification. These requirements are identical to those that apply to producers using other types of renewable biomass feedstocks, such as planted trees from tree plantations, as described in the previous sections. Renewable fuel producers generating RINs under the RFS2 program would continue to be required to affirm (through EMTS—EPA Moderated Transaction System) for each batch of renewable fuel that their feedstocks meet the definition of renewable biomass. Additionally, producers would send a quarterly report to EPA that includes a summary of the types and volumes of feedstocks used throughout the quarter, as well as electronic data or maps identifying the land from which those feedstocks were harvested.

Furthermore, those RIN-generating renewable fuel producers will be required to obtain and maintain in their files written records from their feedstock suppliers for each feedstock purchase that identify where the feedstocks were produced and that are sufficient to verify that the feedstocks qualify as renewable biomass. This includes maps and/or electronic data identifying the boundaries of the land where the feedstock was produced, PTDs or bills of lading tracing the feedstock from that land to the renewable fuel production facility, and other written records that serve as evidence that the feedstock qualifies as renewable biomass. Finally, producers using planted crops and crop residue must maintain additional documentation that serves as evidence that the agricultural land used to produce the crop or crop residue was cleared or cultivated and actively managed or fallow, and nonforested on December 19, 2007. This documentation must consist of the following types of records which must be traced to the land in question: sales records for planted crops, crop residue, or livestock, purchasing records for land treatments such as fertilizer, weed control, or reseeding or a written agricultural management plan or documentation of participation in an agricultural program sponsored by a Federal, State or local government agency.

Alternatively, if the baseline is exceeded and the requirements are triggered for individual producer verification that their feedstocks are renewable biomass renewable fuel producers may choose to work with other renewable fuel producers as well as feedstock producers and suppliers to develop a quality assurance program for the renewable fuel production supply chain. This quality assurance program would take the place of individual accounting and would consist of an independent third party quality-assurance survey of all participating renewable fuel producers and their feedstock suppliers, completed in accordance with an industry-developed, EPA-approved plan, to ensure that they are utilizing feedstocks that meet the definition of renewable biomass. An in-depth discussion of this industry survey option is included in the previous section.

While the aggregate compliance approach is appropriate for planted crops and crop residues from agricultural land in the United States, due in part to certain additional or different constraints imposed by EISA, the aggregate approach cannot be applied, at this time, to the other types of renewable biomass. Renewable fuel producers utilizing these types of renewable biomass, including planted trees and tree residues from tree plantations, slash and pre-commercial thinnings from non-federal forestland, animal waste, separated yard and food waste, etc., will be subject to the individual reporting and recordkeeping requirements discussed in the previous section.

Additionally, EPA is not finalizing the aggregate compliance approach for foreign producers of renewable fuel. EPA does not, at this time, have sufficient data to make a finding that non-domestically grown crops and crop residues used in renewable fuel production satisfy the definition of renewable biomass. Nevertheless, if, in the future, adequate land use data becomes available to make a finding that, in the aggregate, crops and crop residues used in renewable fuel production in a particular country satisfy the definition of renewable biomass, EPA is willing to consider an aggregate compliance approach for renewable biomass on a country by country basis, in lieu of the individual recordkeeping and reporting requirements.

d. Treatment of Municipal Solid Waste (MSW)

The statutory definition of “renewable biomass” does not include a reference to municipal solid waste (MSW) as did the definition of “cellulosic biomass ethanol” in the Energy Policy Act of 2005 (EPAct), but instead includes “separated yard waste and food waste.”

We solicited comment on whether EPA can and should interpret EISA as including MSW that contains yard and/or food waste within the definition of renewable biomass. On the one hand, the reference in the statutory definition to “separated yard waste and food waste,” and the lack of reference to other components of MSW (such as waste paper and wood waste) suggests that only yard and food wastes physically separated from other waste materials satisfy the definition of renewable biomass. On the other hand, we noted that EISA does not define the term “separated,” and so does not specify the degree of separation required. We also noted that there was some evidence in the Act that Congress did not intend to exclude MSW entirely from the definition of renewable biomass. The definition of “advanced biofuel” includes a list of fuels that are “eligible for consideration” as advanced biofuel, including “ethanol derived from waste material” and biogas “including landfill gas.”

As an initial matter, we note that some materials clearly fall within the definition of “separated yard or food waste.” The statute itself identifies “recycled cooking and trap grease” as one example of separated food waste. An example of separated yard waste is the leaf waste that many municipalities pick up at curbside and keep separate from other components of MSW for mulching or other uses. However, a large quantity of food and yard waste is disposed of together with other household waste as part of MSW. EPA estimates that about 120 million tons of MSW are disposed of annually much of it inextricably mixed with yard and especially food waste. This material offers a potentially reliable, abundant and inexpensive source of feedstock for renewable fuel production which, if used, could reduce the volume of discarded materials sent to landfills and could help achieve both the GHG emissions reductions and energy independence goals of EISA. Thus, EPA believes we should consider under what conditions yard and food waste that is present in MSW can be deemed sufficiently separated from other materials to qualify as renewable biomass.

One commenter stated that it is clear that MSW does not qualify as renewable biomass under EISA, since the 2005 Energy Policy Act explicitly allowed for qualifying renewable fuel to be made from MSW, and EISA has no mention of it. Commenters from the renewable fuel industry generally favored maximum flexibility for the use of MSW in producing qualifying fuels under EISA, offering a variety of arguments based on the statutory text and reasons why it would benefit the environment and the nation's energy policy to do so. They favored either (1) a determination that unsorted MSW can be used as a feedstock for advanced biofuel even if it does not meet the definition of renewable biomass, (2) that the Act be interpreted to include MSW as renewable biomass, or (3) that MSW from which varying amounts of recyclable materials have been removed could qualify as renewable biomass. A consortium of ten environmental groups said that for EISA volume mandates to be met, it is important to take advantage of biomass resources from urban wastes that would otherwise be landfilled. They urged that post-recycling residues (i.e., those wastes that are left over at material recovery facilities after separation and recycling) would fit within the letter and spirit of the definition of renewable biomass.

EPA does not believe that the statute can be reasonably interpreted to allow advanced biofuel to be made from material that does not meet the definition of renewable biomass as suggested in the first approach. The definition of advanced biofuel specifies that it is a form of “renewable fuel,” and renewable fuel is defined in the statute as fuel that is made from renewable biomass. While the definition of advanced biofuel includes a list of materials that “may” be “eligible for consideration” as advanced biofuel, and that list includes “ethanol derived from waste materials” and biogas “including landfill gas,” the fact that the specified items are “eligible for consideration” indicates that they do not necessarily qualify but must meet the definitional requirements—being “renewable fuel” made from renewable biomass and having life cycle greenhouse gas emissions that are at least 50% less than baseline fuel. There is nothing in the statute to suggest that Congress used the term “renewable fuel” in the definition of “advanced biofuel” to have a different meaning than the definition provided in the statute. The result of the commenter's first approach would be that general renewable fuel and cellulosic biofuel would be required to be made from renewable biomass because the definitions of those terms specifically refer to renewable biomass, whereas advanced biofuel and biomass-based diesel would not, because their definitions refer to “renewable fuel” rather than “renewable biomass.” EPA can discern no basis for such a distinction. EPA believes that the Act as a whole is best interpreted as requiring all types of qualifying renewable fuels under EISA to be made from renewable biomass. In this manner the land and feedstock restrictions that Congress deemed important in the context of biofuel production apply to all types of renewable fuels.

EPA also does not agree with the commenter who suggested that the listing in the definition of renewable biomass of “biomass obtained from the immediate vicinity of buildings and other areas regularly occupied by people, or of public infrastructure, at risk from wildfire” should be interpreted to include MSW. It is clear that the term “at risk of wildfire” modifies the entire sentence, and the purpose of the listing is to make the biomass that is removed in wildfire minimization efforts, such as brush and dead woody material, available for renewable fuel production. Such material does not typically include MSW. Had Congress intended to include MSW in the definition of renewable biomass, EPA believes it would have clearly done so, in a manner similar to the approach taken in EPAct.

EPA also does not believe that it would be reasonable to interpret the reference to “separated yard or food waste” to include unsorted MSW. Although MSW contains yard and food waste, such an approach would not give meaning to the word “separated.”

We do believe, however, that yard and food wastes that are part of MSW, and are separated from it, should qualify as renewable biomass. MSW is the logical source from which yard waste and food waste can be separated. As to the degree of separation required, some commenters suggested a simple “post recycling” test be appropriate. They would leave to municipalities and waste handlers a determination of how much waste should be recycled before the residue was used as a feedstock for renewable fuel production. EPA believes that such an approach would not guarantee sufficient “separation” from MSW of materials that are not yard waste or food waste to give meaning to the statutory text. Instead, EPA believes it would be reasonable in the MSW context to interpret the word “separated” in the term “separated yard or food waste” to refer to the degree of separation to the extent that is reasonably practicable. A large amount of material can be, and is, removed from MSW and sold to companies that will recycle the material. EPA believes that the residues remaining after reasonably practicable efforts to remove recyclable materials other than food and yard waste (including paper, cardboard, plastic, textiles, metal and glass) from MSW should qualify as separated yard and food waste. This MSW-derived residue would likely include some amount of residual non-recyclable plastic and rubber of fossil fuel origin, much of it being wrapping and packaging material for food. Since this material cannot be practicably separated from the remaining food and yard waste, EPA believes it is incidental material that is impractical to remove and therefore appropriate to include in the category of separated food and yard waste. In sum, EPA believes that the biogenic portion of the residue remaining after paper, cardboard, plastic, textiles metal and glass have been removed for recycling should qualify as renewable biomass. This interpretation is consistent with the text of the statute, and will promote the productive use of materials that would otherwise be landfilled. It will also further the goals of EISA in promoting energy independence and the reduction of GHG emissions from transportation fuels.

EPA notes there are a variety of recycling methods that can be used, including curbside recycling programs, as well as separation and sorting at a material recovery facility (MRF). For the latter, the sorting could be done by hand or by automated equipment, or by a combination of the two. Sorting by hand is very labor intensive and much slower than using an automated system. In most cases the “by-hand” system produces a slightly cleaner stream, but the high cost of labor usually makes the automated system more cost-effective. Separation via MRFs is generally very efficient and can provide comparable if not better removal of recyclables to that achieved by curbside recycling.

Based on this analysis, today's rule provides that those MSW-derived residues that remain after reasonably practicable separation of recyclable materials other than food and yard waste is renewable biomass. What remains to be addressed is what regulatory mechanisms should be used to ensure the appropriate generation of RINs when separated yard and food waste is used as a feedstock. We are finalizing two methods.

The first method would apply primarily to a small subset of producers who are able to obtain yard and/or food wastes that have been kept separate since waste generation from the MSW waste stream. Examples of such wastes are lawn and leaf waste that have never entered the general MSW waste stream. Typically, such wastes contain incidental amounts of materials such as the plastic twine used to bind twigs together, food wrappers, and other extraneous materials. As with our general approach to the presence of incidental, de minimus contaminants in feedstocks that are unintentionally present and impractical to remove, the presence of such material in separated yard or food waste will not disqualify such wastes as renewable biomass, and the contaminants may be disregarded by producers and importers generating RINs. (See definition of renewable biomass and 80.1426(f)(1).) Waste streams kept separate since generation from MSW that consist of yard waste are expected to be composed almost entirely of woody material or leaves, and therefore will be deemed to be composed of cellulosic materials. Waste streams consisting of food wastes, however, may contain both cellulosic and non-cellulosic materials. For example, a food processing plant may generate both wastes that are primarily starches and sugars (such as carrot and potato peelings, as well as fruits and vegetables that are discarded) as well as corn cobs and other materials that are cellulosic. We will deem waste streams consisting of food waste to be composed entirely of non-cellulosic materials, and qualifying as advanced biofuels, unless the producer demonstrates that some portion of the food waste is cellulosic. The cellulosic portion would then qualify as cellulosic biofuel. The method for quantifying the cellulosic and non-cellulosic portions of the food waste stream is to be described in a written plan which must be submitted to EPA under the registration procedures in 80.1450(b)(vii) for approval and which indicates the location of the facility from which wastes are obtained, how identification and quantification of waste material is to be accomplished, and evidence that the wastes qualify as fully separated yard or food wastes. The producer must also maintain records regarding the source of the feedstock and the amounts obtained.

The second method would involve use as feedstock by a renewable fuel producer of the portion of MSW remaining after reasonably practical separation activities to remove recyclable materials, resulting in a separated MSW-derived residue that qualifies as separated yard and food waste. Today's rule requires that parties that intend to use MSW-derived residue as a feedstock for RIN-generating renewable fuel production ensure that reasonably practical efforts are made to separate recyclable paper, cardboard, textiles, plastics, metal and glass from the MSW, according to a plan that is submitted by the renewable fuel producer and approved by EPA under the registration procedures in 80.1450(b)(viii). In determining whether the plan submittals provide for reasonably practicable separation of recyclables EPA will consider: (1) The extent and nature of recycling that may have occurred prior to receipt of the MSW material by the renewable fuel producer, (2) available recycling technology and practices, and (3) the technology or practices selected by the fuel producer, including an explanation for such selection and reasons why other technologies or practices were not selected. EPA asks that any CBI accompanying a plan or a party's justification for a plan be segregated from the non-CBI portions of the submissions, so as to facilitate disclosure of the non-CBI portion of plan submittals, and approved plans, to interested members of the public.

Producers using this second option, will need to determine what RINs to assign to a fuel that is derived from a variety of materials, including yard waste (largely cellulosic) and food waste (largely starches and sugar), as well as incidental materials remaining after reasonably practical separation efforts such as plastic and rubber of fossil origin. EPA has not yet evaluated the lifecycle greenhouse gas performance of fuel made from such mixed sources of waste, so is unable at this time to assign a D code for such fuel. However, if a producer uses ASTM test method D-6866 on the fuel made from MSW-derived feedstock, it can determine what portion of the rule is of fossil and non-fossil origin. The non-fossil portion of the fuel will likely be largely derived from cellulosic materials (yard waste, textiles, paper, and construction materials), and to a much smaller extent starch-based materials (food wastes). Unfortunately, EPA is not aware of a test method that is able to distinguish between cellulosic- and starch-derived renewable fuel. Under these circumstances, EPA believes that it is appropriate for producers to base RIN assignment on the predominant component and, therefore, to assume that the biogenic portion of their fuel is entirely of cellulosic origin. The non-biogenic portion of the fuel, however, would not qualify for RINs at this time. Thus, in sum, we are providing via the ASTM testing method an opportunity for producers using an MSW-derived feedstock to generate RINs only for the biogenic portion of their renewable fuel. There is no D code for the remaining fossil-derived fraction of the fuel in today's rule nor for the entire volume of renewable fuel produced when using MSW-derived residue as a feedstock. The petition process for assigning such codes in today's rule can be used for such purpose.

Procedures for the use of ASTM Method D-6866 are detailed in 40 CFR 80.1426(f)(9) of today's rule. We solicited comment on this method, and while the context of the discussion of method D-6866 was with respect to using it for gasoline (see 74 FR 24951), the comments we received provided us information on the method itself. Also, commenters were supportive of its use. Fuel producers must either run the ASTM D-6866 method for each batch of fuel produced, or run it on composite samples of the food and yard waste-derived fuel derived from post-recycling MSW residues. Producers will be required at a minimum to take samples of every batch of fuel produced over the course of one month and combine them into a single composite sample. The D-6866 test would then be applied to the composite sample, and the resulting non-fossil derived fraction will be deemed cellulosic biofuel, and applied to all batches of fuel produced in the next month to determine the appropriate number of RINs that must be generated. The producer would be required to recalculate this fraction at least monthly. For the first month, the producer can estimate the non-fossil fraction, and then make a correction as needed in the second month. (The procedure using the ASTM D-6866 method applies not only to the waste-derived fuel discussed here but also to all partially renewable transportation fuels, and is discussed in further detail in Section II.D.4. See also the regulations at § 80.1426(f)(4)).

The procedures for assigning D codes to the fuel produced from such wastes are discussed in further detail in Section II.D.5.

One commenter suggested that biogas from landfills should be treated in the same manner as renewable fuel produced from MSW. EPA agrees with the commenter to a certain extent. The definition of “advanced biofuels” in EISA identifies “Biogas (including landfill gas and sewage waste treatment gas) produced through the conversion of organic matter from renewable biomass” as “eligible for consideration” as an advanced biofuel. However, as with MSW, the statute requires that advanced biofuel be a “renewable fuel” and that such fuel be made from “renewable biomass.” The closest reference within the definition of renewable biomass to landfill material is “separated yard or food waste.” However, in applying the interpretation of “separated” yard and food waste described above for MSW to landfill material, we come to a different result. Landfill material has by design been put out of practical human reach. It has been disposed of in locations, and in a manner, that is designed to be permanent. For example, modern landfills are placed over impermeable liners and sealed with a permanent cap. In addition, the food and yard waste present in a landfill has over time become intermingled with other materials to an extraordinary extent. This occurs in the process of waste collection, shipment, and disposal, and subsequently through waste decay, leaching and movement within the landfill. Additionally, we note that the process of biogas formation in a landfill provides some element of separation, in that it is formed only from the biogenic components of landfill material, including but not strictly limited to food and yard waste. Thus, plastics, metal and glass are effectively “separated” out through the process of biogas formation. As a result of the intermixing of wastes, the fact that biogas is formed only from the biogenic portion of landfill material, and the fact that landfill material is as a practical matter inaccessible for further separation, EPA believes that no further practical separation is possible for landfill material and biogas should be considered as produced from separated yard and food waste for purposes of EISA. Therefore, all biogas from landfills is eligible for RIN generation.

We have considered whether to require biogas producers to use ASTM Method D-6866 to identify the biogenic versus non-biogenic fractions of the fuel. However, as noted above, biogas is not formed from non-biogenic compounds in landfills. (Kaplan, et al., 2009) [9] Thus, no purpose would be solved in using the ASTM method in the biogas context.

C. Expanded Registration Process for Producers and Importers

In order to implement and enforce the new restrictions on qualifying renewable fuel under RFS2, we are revising the registration process for renewable fuel producers and importers. Under the RFS1 program, all producers and importers of renewable fuel who produce or import more than 10,000 gallons of fuel annually must register with EPA's fuels program prior to generating RINs. Renewable fuel producer and importer registration under the RFS1 program consists of filling out two forms: 3520-20A (Fuels Programs Company/Entity Registration), which requires basic contact information for the company and basic business activity information and 3520-20B (Gasoline Programs Facility Registration) or 3520-20B1 (Diesel Programs Facility Registration), which require basic contact information for each facility owned by the producer or importer. More detailed information on the renewable fuel production facility, such as production capacity and process, feedstocks, and products was not required for most producers or importers to generate RINs under RFS1 (producers of cellulosic biomass ethanol and waste-derived ethanol are the exception to this).

Additionally, EPA recommends companies register their renewable fuels or fuel additives under title 40 CFR part 79 as a motor vehicle fuel. In fact, renewable fuels intended for use in motor vehicles will be required to be registered under title 40 CFR part 79 prior to any introduction into commerce. Manufacturers and subsequent parties of fuels and fuel additives not registered under part 79 will be liable for separate penalties under 40 CFR parts 79 and 80 in the event their unregistered product is introduced into commerce for use in a motor vehicle. Further if a registered fuel or fuel additive is used in manner that is not consistent with their product's registration under part 79 the manufacturer and subsequent parties will be liable for penalties under parts 79 and 80. If EPA determines based on the company's registration that they are not producing renewable fuel, the company will not be able to generate RINs and the RINs generated for fuel produced from nonrenewable sources will be invalidated.

Due to the revised definitions of renewable fuel under EISA, we proposed to expand the registration process for renewable fuel producers and importers in order to implement the new program effectively. We received a number of comments that opposed the expanded registration as commenters deemed it overly burdensome, costly and unnecessary. However, EPA is finalizing the proposed expanded registration requirements for the following reasons. The information to be collected through the expanded registration process is essential to generating and assigning a certain category of RIN to a volume of fuel. Additionally, the information collected is essential to determining whether the feedstock used to produce the fuel meets the definition of renewable biomass, whether the lifecycle greenhouse gas emissions of the fuel meets a certain GHG reduction threshold and, in some cases, whether the renewable fuel production facility is considered to be grandfathered into the program. Therefore, we are requiring producers, including foreign producers, and importers that generate RINs to provide us with information on their feedstocks, facilities, and products, in order to implement and enforce the program and have confidence that producers and importers are properly categorizing their fuel and generating RINs. The registration procedures will be integrated with the new EPA Moderated Transaction System, discussed in detail in Section III.A of this preamble.

1. Domestic Renewable Fuel Producers

Information on products, feedstocks, and facilities contained in a producer's registration will be used to verify the validity of RINs generated and their proper categorization as either cellulosic biofuel, biomass-based diesel, advanced biofuel, or other renewable fuel. In addition, producers of renewable fuel from facilities that qualify for the exemption from the 20% GHG reduction threshold (as discussed in Section II.B.3) must provide information that demonstrates when the facility commenced construction, and that establishes the baseline volume of the fuel. For those facilities that would qualify as grandfathered but are not in operation we are allowing until May 1, 2013 to submit and receive approval for a complete facility registration. This provision does not require actual fuel production, but simply the filing of registration materials that assert a claim for exempt status. It will benefit both fuel producers, who will likely be able to more readily collect the required information if it is done promptly, and EPA enforcement personnel seeking to verify the information. However, given the potentially significant implications of this requirement for facilities that may qualify for the exemption but miss the registration deadline, the rule also provides that EPA may waive the requirement if it determines that the submission is verifiable to the same extent as a timely-submitted registration.

With respect to products, we are requiring that producers provide information on the types of renewable fuel and co-products that a facility is capable of producing. With respect to feedstocks, we are requiring producers to provide to EPA a list of all the different feedstocks that a renewable fuel producer's facility is likely to use to convert into renewable fuel. With respect to the producer's facilities, two types of information must be reported to the Agency. First, producers must describe each facility's fuel production processes (e.g., wet mill, dry mill, thermochemical, etc.), and thermal/process energy source(s). Second, in order to determine what production volumes would be grandfathered and thus deemed to be in compliance with the 20% GHG threshold, we are requiring evidence and certification of the facility's qualification under the definition of “commence construction” as well as information necessary to establish its renewable fuel baseline volume per the requirement outlined in Section II.B.3 of this preamble.

EPA proposed to require that renewable fuel producers have a third-party engineering review of their facilities prior to generating RINs under RFS2, and every 3 years thereafter. EPA received comments that the on-site engineering review was overly burdensome, unnecessary and costly. A number of commenters noted that the time allotted for conducting the reviews, between the rule's publication and prior to RIN generation, is not adequate for producers to hire an engineer and conduct the review for all of their facilities. Several commenters requested that on-site licensed engineers be allowed to conduct any necessary facility reviews.

EPA is finalizing the proposed requirement for an on-site engineering review of facilities producing renewable fuel due to the variability of production facilities, the increase in the number of categories of renewable fuels, and the importance of ensuring that RINs are generated in the correct category. Without these engineering reviews, we do not believe it would be possible to implement the RFS2 program in a manner that ensured the requirements of EISA were being fulfilled. Additionally, the engineering review provides a check against fraudulent RIN generation. In order to establish the proper basis for RIN generation, we are requiring that every renewable fuel producer have the on-site engineering review of their facility performed in conjunction with his or her initial registration for the new RFS program. The engineering reviews must be conducted by independent third parties who can maintain impartiality and objectivity in evaluating the facilities and their processes. Additionally, the on-site engineering review must be conducted every three years thereafter to verify that the fuel pathways established in the initial registration are still applicable. These requirements apply unless the renewable fuel producer updates its facility registration information to qualify for a new RIN category (i.e., D code), in which case the review needs to be performed within 60 days of the registration update. Finally, producers are required to submit a copy of their independent engineering review to EPA, for verification and enforcement purposes.

2. Foreign Renewable Fuel Producers

Under RFS1, foreign renewable fuel producers of cellulosic biomass ethanol and waste-derived ethanol may apply to EPA to generate RINs for their own fuel. For RFS2, we proposed that foreign producers of renewable fuel meet the same requirements as domestic producers, including registering information about their feedstocks, facilities, and products, as well as submitting an on-site independent engineering review of their facilities at the time of registration for the program and every three years thereafter. These requirements apply to all foreign renewable fuel producers who plan to export their products to the U.S. as part of the RFS2 program, whether the foreign producer generates RINs for their fuel or an importer does.

Foreign producers, like domestic producers, must also undergo an independent engineering review of their facilities, conducted by an independent third party who is a licensed professional engineer (P.E.), or foreign equivalent who works in the chemical engineering field. The independent third party must provide to EPA documentation of his or her qualifications as part of the engineering review, including proof of appropriate P.E. license or foreign equivalent. The third-party engineering review must be conducted by both foreign producers who plan to generate RINs and those that don't generate RINs but anticipate their fuel will be exported to the United States by an importer who will generate the RINs.

3. Renewable Fuel Importers

We are requiring importers who generate RINs for imported fuel that they receive without RINs may only do so under certain circumstances. If an importer receives fuel without RINs, the importer may only generate RINs for that fuel if they can verify the fuel pathway and that feedstocks use meet the definition of renewable biomass. An importer must rely on his supplier, a foreign renewable fuel producer, to provide documentation to support any claims for their decision to generate RINs. An importer may have an agreement with a foreign renewable fuel producer for the importer to generate RINs if the foreign producer has not done so already. However, the foreign renewable fuel producer must be registered with EPA and must have had a third-party engineering review conducted, as noted above, in order for EPA to be able to verify that the renewable biomass and GHG reduction requirements of EISA are being fulfilled. Section II.D.2.b describes the RIN generating restrictions and requirements for importers under RFS2.

4. Process and Timing

We are making forms for expanded registration for renewable fuel producers and importers, as well as forms for registration of other regulated parties, available electronically with the publication of this final rule. Paper registration forms will only be accepted in exceptional cases. Registration forms must be submitted and accepted by the EPA by July 1, 2010, or 60 days prior to a producer producing or importer importing any renewable fuel, whichever dates come later. If a producer changes its fuel pathway (feedstock, production process, or fuel type) to not listed in his registration information on file with EPA but the change will not incur a change of RIN category for the fuel (i.e., a change in the appropriate D code), the producer must update his registration information within seven (7) days of the change. However, if the fuel producer changes its fuel pathway in a manner that would result in a change in its RIN category (and thus a new D code), such an update would need to be submitted at least 60 days prior to the change, followed by submittal of a complete on-site independent engineering review of the producer's facility also within 60 days of the change. If EPA finds that these deadlines and requirements have not been met, or that a facility's registered profile, dictated by the various parameters for product, process and feedstock, does not reflect actual products produced, processes employed, or feedstocks used, then EPA reserves the right to void, ab initio, any affected RINs generated and may impose significant penalties. For example a newly registered (i.e. not grandfathered) ethanol production facility claims in their registration that they qualify to generate RINs based upon the use of two advanced engineering practices (1) corn oil fractionation and (2) production of wet DGS co-product that is, at a minimum, 35% of its total DGS produced annually. However, during an audit of the producer's records, it is found that of all their DGS produced, less than 15% was wet. In this example, the producer has committed a violation that results in the disqualification of their eligibility to generate RINs; that is, they no longer have an eligible pathway that demonstrates qualification with the 20% GHG threshold requirement for corn ethanol producers. As such any and all RINs produced may be deemed invalid and the producer may be subject to Clean Air Act penalties.

The required independent engineering review as discussed above for domestic and foreign renewable fuel producers is an integral part of the registration process. The agency recognizes, through comments received, that there are significant concerns involving timing necessary and ability to produce a completed engineering review to satisfy registration requirements. Since the publication of the RFS2 NPRM, we have delivered consistently a message stating that advanced planning and preparation was necessary from all parties, EPA and the regulated community inclusive, for successful implementation of this program. In an effort to reduce demand on engineering resources, we are allowing grandfathered facilities an additional six months to submit their engineering review. This will direct the focus of engineering review resources on producers of advanced, cellulosic and biomass based diesel. EPA fully expects these producers of advanced renewable fuels to meet the engineering review requirement; however, if they are having difficulties producing engineer's reports prior to April 1, we ask that they contact us.

D. Generation of RINs

Under RFS2, each RIN will continue to be generated by the producer or importer of the renewable fuel, as in the RFS1 program. In order to determine the number of RINs that must be generated and assigned to a batch of renewable fuel, the actual volume of the batch of renewable fuel must be multiplied by the appropriate Equivalence Value. The producer or importer must also determine the appropriate D code to assign to the RIN to identify which of the four standards the RIN can be used to meet. This section describes these two aspects of the generation of RINs. Other aspects of the generation of RINs, such as the definition of a batch, as well as the assignment of RINs to batches, will remain unchanged from the RFS1 requirements. We received several comments regarding the method for calculating temperature standardization of biodiesel and address this issue in Section III.G.

1. Equivalence Values

For RFS1, we interpreted CAA section 211(o) as allowing us to develop Equivalence Values representing the number of gallons that can be claimed for compliance purposes for every physical gallon of renewable fuel. We described how the use of Equivalence Values adjusted for renewable content and based on energy content in comparison to the energy content of ethanol was consistent with the sections of EPAct that provided extra credit for cellulosic and waste-derived renewable fuels, and the direction that EPA establish “appropriate” credit for biodiesel and renewable fuel volumes in excess of the mandated volumes. We also noted that the use of Equivalence Values based on energy content was an appropriate measure of the extent to which a renewable fuel would replace or reduce the quantity of petroleum or other fossil fuel present in a fuel mixture. EPA stated that these provisions indicated that Congress did not intend to restrict EPA discretion in implementing the program to utilizing a straight volume measurement of gallons. See 72 FR 23918-23920, and 71 FR 55570-55571. The result was an Equivalence Value for ethanol of 1.0, for butanol of 1.3, for biodiesel (mono alkyl ester) of 1.5, and for non-ester renewable diesel of 1.7.

In the NPRM we noted that EISA made a number of changes to CAA section 211(o) that impacted our consideration of Equivalence Values in the context of the RFS2 program. For instance, EISA eliminated the 2.5-to-1 credit for cellulosic biomass ethanol and waste-derived ethanol and replaced this provision with large mandated volumes of cellulosic biofuel and advanced biofuels. EISA also expanded the program to include four separate categories of renewable fuel (cellulosic biofuel, biomass-based diesel, advanced biofuel, and total renewable fuel) and included GHG thresholds in the definitions of each category. Each of these categories of renewable fuel has its own volume requirement, and thus there will exist a guaranteed market for each. As a result of these new requirements, we indicated that there may no longer be a need for additional incentives for certain fuels in the form of Equivalence Values greater than 1.0.

In the NPRM we co-proposed and took comment on two options for Equivalence Values:

1. Equivalence Values would be based on the energy content and renewable content of each renewable fuel in comparison to denatured ethanol, consistent with the approach under RFS1, with the addition that biomass-based diesel standard would be based on energy content in comparison to biodiesel.

2. All liquid renewable fuels would be counted strictly on the basis of their measured volumes, and the Equivalence Values for all renewable fuels would be 1.0 (essentially, Equivalence Values would no longer apply).

In response to the NPRM, some stakeholders pointed to the aforementioned changes brought about by EISA as support for a straight volume approach to Equivalence Values, and argued that it had always been the intent of Congress that the statutory volume mandates be treated as straight volumes. Stakeholders taking this position were generally producers of corn ethanol. However, a broad group of other stakeholders including refiners, biodiesel producers, a broad group of advanced biofuel producers, fuel distributor and States indicated that the first option for an energy-based approach to Equivalence Values was both supported by the statute and necessary to provide for equitable treatment of advanced biofuels. They noted that EISA did not change certain of the statutory provisions EPA looked to for support under RFS1 in establishing Equivalence Values based on relative volumetric energy content in comparison to ethanol. For instance, CAA 211(o) continues to direct EPA to determine an “appropriate” credit for biodiesel, and also directs EPA to determine the “appropriate” amount of credit for renewable fuel use in excess of the required volumes. Had Congress intended to change these provisions they could have easily done so. Moreover, some stakeholders argued that the existence of four standards is not a sufficient reason to eliminate the use of energy-based Equivalence Values for RFS2. The four categories are defined in such a way that a variety of different types of renewable fuel could qualify for each category, such that no single specific type of renewable fuel will have a guaranteed market. For example, the cellulosic biofuel requirement could be met with both cellulosic ethanol or cellulosic diesel. As a result, the existence of four standards under RFS2 does not obviate the value of standardizing for energy content, which provides a level playing field under RFS1 for various types of renewable fuels based on energy content.

Some stakeholders who supported an energy-based approach to Equivalence Values also argued that a straight volume approach would be likely to create a disincentive for the development of new renewable fuels that have a higher energy content than ethanol. For a given mass of feedstock, the volume of renewable fuel that can be produced is roughly inversely proportional to its energy content. For instance, one ton of biomass could be gasified and converted to syngas, which could then be catalytically reformed into either 80 gallons of ethanol (and another 14 gal of other alcohols) or 50 gallons of diesel fuel (and naphtha). [10] If RINs were assigned on a straight volume basis, the producer could maximize the number of RINs he is able to generate and sell by producing ethanol instead of diesel. Thus, even if the market would otherwise lean towards demanding greater volumes of diesel, the greater RIN value for producing ethanol may favor their production instead. However, if the energy-based Equivalence Values were maintained, the producer could assign 1.7 RINs to each gallon of diesel made from biomass in comparison to 1.0 RIN to each gallon of ethanol from biomass, and the total number of RINs generated would be essentially the same for the diesel as it would be for the ethanol. The use of energy-based Equivalence Values could thus provide a level playing field in terms of the RFS program's incentives to produce different types of renewable fuel from the available feedstocks. The market would then be free to choose the most appropriate renewable fuels without any bias imposed by the RFS regulations, and the costs imposed on different types of renewable fuel through the assignment of RINs would be more evenly aligned with the ability of those fuels to power vehicles and engines, and displace fossil fuel-based gasoline or diesel. Since the technologies for producing more energy-dense fuels such as cellulosic diesel are still in the early stages of development, they may benefit from not having to overcome the disincentive in the form of the same Equivalence Value based on straight volume.

Based on our interpretation of EISA as allowing the use of energy-based Equivalence Values, and because we believe it provides a level playing field for the development of different fuels that can displace the use of fossil fuels, and that this approach therefore furthers the energy independence goals of EISA, we are finalizing the energy-based approach to Equivalence Values in today's action. We also note that a large number of companies have already made investments based on the decisions made for RFS1, and using energy-based Equivalence Values will maintain consistency with RFS1 and ease the transition into RFS2. Insofar as renewable fuels with volumetric energy contents higher than ethanol are used, the actual volumes of renewable fuel that are necessary to meet the EISA volume mandates will be smaller than those shown in Table I.A.1-1. The impact on the physical volume will depend on actual volumes of various advanced biofuels produced in the future. The main scenario modeled for this final rule includes a forecast for considerable volumes of relatively high energy diesel fuel made from renewable biomass, and still results in a physical volume mandate of 30.5 billion gallons. The energy-based approach results in the advanced biofuel standard being automatically met during the first few years of the program. For instance, the biomass-based diesel mandated volume for 2010 is 0.65 billion gallons, which will be treated as 0.975 billion gallons (1.5 × 0.65) in the context of meeting the advanced biofuel standard. Since the mandated volume for advanced biofuel in 2010 is 0.95 billion gallons, this requirement is automatically met by compliance with the biomass-based diesel standard.

Although we are finalizing an energy-based approach to Equivalence Values, we believe that Congress intended the biomass-based diesel volume mandate to be treated as diesel volumes rather than as ethanol-equivalent volumes. Since all RINs are generated based on energy equivalency to ethanol, to accomplish this, we have modified the formula for calculating the standard for biomass-based diesel to compensate such that one physical gallon of biomass-based diesel will count as one gallon for purposes of meeting the biomass-based diesel standard, but will be counted based on their Equivalence Value for purposes of meeting the advanced biofuel and total renewable fuel standards. Since it is likely that the statutory volume mandates were based on projections for biodiesel, we have chosen to use the Equivalence Value for biodiesel, 1.5, in this calculation. See Section II.E.1.a for further discussion. Other diesel fuel made from renewable biomass can also qualify as biomass-based diesel (e.g., renewable diesel, cellulosic diesel). But since the variation in energy content between them is relatively small, variation in the total physical volume of biomass-based diesel will likewise be small.

In the NPRM we also proposed that the energy content of denatured ethanol be changed from the 77,550 Btu/gal value used in the RFS1 program to 77,930 Btu/gal (lower heating value). The revised value was intended to provide a more accurate estimate of the energy content of pure ethanol, 76,400 Btu/gal, rather than the rounded value of 76,000 Btu/gal that was used under RFS1. Except for the Renewable Fuels Association who supported this change, most stakeholders did not comment on this proposal. However, based on new provisions in the Food, Conservation, and Energy Act of 2008, we have since determined that the denaturant content of ethanol should be assumed to be 2% rather than the 5% used in the RFS1 program. This additional change results in a denatured ethanol energy content of 77,000 Btu/gal and a renewable content of denatured ethanol of 97.2%. [11] The value of 77,000 Btu/gal will be used to convert biogas and renewable electricity into volumes of renewable fuel under RFS2. This change also affects the formula for calculating Equivalence Values assigned to renewable fuels. The new formula is shown below:

EV = (R/0.972) * (EC/77,000)

Where:

EV = Equivalence Value for the renewable fuel, rounded to the nearest tenth.

R = Renewable content of the renewable fuel. This is a measure of the portion of a renewable fuel that came from a renewable source, expressed as a percent, on an energy basis.

EC = Energy content of the renewable fuel, in Btu per gallon (lower heating value).

Under this new formula, Equivalence Values assigned to specific types of renewable fuel under RFS1 will continue unchanged under RFS2. However, non-ester renewable diesel will be required to have a lower energy content of at least 123,500 Btu/gal in order to qualify for an Equivalence Value of 1.7. A non-ester renewable diesel with a lower energy content would be required to apply for a different Equivalent Value according to the provisions in § 80.1415.

2. Fuel Pathways and Assignment of D Codes

As described in Section II.A, RINs under RFS2 would in general continue to have the same number of digits and code definitions as under RFS1. The one change will be that, while the D code will continue to identify the standard to which the RIN can be applied, it will be modified to have four values corresponding to the four different renewable fuel categories defined in EISA. These four D code values and the corresponding categories are shown in Table II.A-1.

In order to generate RINs for renewable fuel that meets the various eligibility requirements (see Section II.B), a producer or importer must know which D code to assign to those RINs. Following the approach we described in the NPRM, a producer or importer will determine the appropriate D code using a lookup table in the regulations. The lookup table lists various combinations of fuel type, production process, and feedstock, and the producer or importer chooses the appropriate combination representing the fuel he is producing and for which he is generating RINs. Parties generating RINs are required to use the D code specified in the lookup table and are not permitted to use a D code representing a broader renewable fuel category. For example, a party whose fuel qualified as biomass-based diesel could not choose to categorize that fuel as advanced biofuel or general renewable fuel for purposes of RIN generation. [12]

This section describes our approach to the assignment of D codes to RINs for domestic producers, foreign producers, and importers of renewable fuel. Subsequent sections address the generation of RINs in special circumstances, such as when a production facility has multiple applicable combinations of feedstock, fuel type, and production process within a calendar year, production facilities that co-process renewable biomass and fossil fuels, and production facilities for which the lookup table does not provide an applicable D code.

a. Producers

For both domestic and foreign producers of renewable fuel, the lookup table identifies individual fuel “pathways” comprised of unique combinations of the type of renewable fuel being produced, the feedstock used to produce the renewable fuel, and a description of the production process. Each pathway is assigned to one of the D codes on the basis of the revised renewable fuel definitions provided in EISA and our assessment of the GHG lifecycle performance for that pathway. A description of the lifecycle assessment of each fuel pathway and the process we used for determining the associated D code can be found in Section V.

Note that the generation of RINs also requires as a prerequisite that the feedstocks used to make the renewable fuel meet the definition of “renewable biomass” as described in Section II.B.4, including applicable land use restrictions. If a producer is not able to demonstrate that his feedstocks meet the definition of renewable biomass, RINs cannot be generated. However, as noted in Section II.B.4.b.1, feedstocks typically include incidental contaminants. These contaminants may have been intentionally added to promote cultivation (e.g., pesticides, herbicides, fertilizer) or transport (e.g., nylon baling rope). In addition, there may be some incidental contamination of a particular load of feedstocks with co-product during feedstock production, or with other agricultural materials during shipping. For example, there may be incidental corn kernels remaining on some corn cobs used to produce cellulosic biofuel, or some sorghum kernels left in a shipping container that are introduced into a load of corn kernels being shipped to a biofuel production facility. The final regulations clarify that in assigning D codes for renewable fuel, producers and importers should disregard the presence of incidental contaminants in their feedstocks if the incidental contaminants are related to customary feedstock production and transport, and are impractical to remove and occur in de minimus levels.

Through our assessment of the lifecycle GHG impacts of different pathways and the application of the EISA definitions for each of the four categories of renewable fuel, including the GHG thresholds, we have determined that all four categories will have pathways that could be used to meet the Act's volume requirements. For example, ethanol made from corn stover or switchgrass in an enzymatic hydrolysis process will count as cellulosic biofuel. Biodiesel made from waste grease or soybean oil can count as biomass-based diesel. Ethanol made from sugarcane sugar will count as advanced biofuel. Finally, a variety of pathways will count as renewable fuel under the RFS2 program. The complete list of pathways that are valid under our final RFS2 program is discussed in Section V.C and are provided in the regulations at § 80.1426(f).

Producers must choose the appropriate D code from the lookup table in the regulations based on the fuel pathway that describes their facility. The fuel pathway must be specified by the producer in the registration process as described in Section II.C. If there are changes to a producer's facility or feedstock such that their fuel would require a D code that was different from any D code(s) which their existing registration information already allowed, the producer is required to revise its registration information with EPA 30 days prior to changing the applicable D code it uses to generate RINs. Situations in which multiple fuel pathways could apply to a single facility are addressed in Section II.D.3 below.

For producers for whom none of the defined fuel pathways in the lookup table apply, a producer can still generate RINs if he meets the criteria for grandfathered or deemed compliant status as described in Section II.B.3 and his fuel meets the definition of renewable fuel as described in Section II.B.1. In this case he would use a D code of 6 for those RINs generated under the grandfathering or deemed compliant provisions.

A diesel fuel product produced from cellulosic feedstocks that meets the 60% GHG threshold can qualify as either cellulosic biofuel or biomass-based diesel. In the NPRM, we proposed that the producer of such “cellulosic diesel” be required to choose whether to categorize his product as either cellulosic biofuel or biomass-based diesel. However, we requested comment on an alternative approach in which an additional D code would be defined to represent cellulosic diesel allowing the cellulosic diesel RIN to be sold into either market. As described more fully in Section II.A above, we are finalizing this alternative approach in today's final rule. Producers or importers of a fuel that qualifies as both biomass-based diesel and cellulosic biofuel must use a D code of 7 in the RINs they generate, and will thus have the flexibility of marketing such RINs to parties seeking either cellulosic biofuel or biomass-based diesel RINs, depending on market demand. Obligated parties can apply RINs with a D code of 7 to either their cellulosic biofuel or biomass-based diesel RVOs, but not both.

In addition to the above comments, we received comments requesting that the use of biogas as process heat in the production of ethanol, should not be limited to use at the site of renewable fuel production. Specifically, commenters point out that the introduction of gas produced from landfills or animal wastes to fungible pipelines is the only practical manner for most renewable fuel facilities to acquire and use landfill gas, since very few are located adjacent to landfills, or have dedicated pipelines from landfill gas operations to their facilities. [13] The commenters suggested that ethanol plants causing landfill gas to be introduced into a fungible gas pipeline be allowed to claim those volumes. The alternative would be to allow landfill gas that is only used onsite to be counted in establishing the pathway.

We believe that the suggested approach has merit. We agree that it does not make any difference in terms of the beneficial environmental attributes associated with the use of landfill gas whether the displacement of fossil fuel occurs in a fungible natural gas pipeline, or in a specific facility that draws gas volume from that pipeline. In fact, a similar approach is widely used with respect to electricity generated by renewable biomass that is placed into a commercial electricity grid. A party buying the renewable power is credited with doing so in state renewable portfolio programs even though the power from these sources is placed in the fungible grid and the electrons produced by a renewable source may never actually be used by the party purchasing it. In essence these programs assume that the renewable power purchased and introduced into the grid is in fact used by the purchaser, even though all parties acknowledge that use of the actual renewable-derived electrons can never be verified once placed in the fungible grid. We believe that this approach will ultimately further the GHG reduction and energy security goals of RFS2.

Producers may therefore take into account such displacement provided that they demonstrate that a verifiable contractual pathway exists and that such pathway ensures that (1) a specific volume of landfill gas was placed into a commercial pipeline that ultimately serves the transportation fueling facility and (2) that the drawn into this facility from that pipeline matches the volume of landfill gas placed into the pipeline system. Thus facilities using such a fuel pathway may then use an appropriate D code for generation of RINs.

This approach also applies to biogas and electricity made from renewable fuels and which are used for transportation. Producers of such fuel will be able to generate RINs, provided that a contractual pathway exists that provides evidence that specific quantities of the renewable fuel (either biogas or electricity) was purchased and contracted to be delivered to a specific transportation fueling facility. [14] We specify that the pipeline (or transmission line) system must ultimately serve the subject facility. For electricity that is produced by the co-firing of fossil fuels with renewable biomass derived fuels, we are requiring that the resulting electricity is pro-rated to represent only that amount of electricity generated by the qualifying biogas, for the purpose of computing RINs.

We are also providing for those situations in which biogas or renewable electricity is provided directly to the transportation facility, rather than using a commercial distribution system such as pipelines or transmission lines. For both cases—dedicated use and commercial distribution—producers must provide contractual evidence of the production and sale of such fuel, and there are also reporting and recordkeeping requirements to be followed as well.

Presently, there is no D code for electricity that is produced from renewable biomass. The petition process for assigning such codes in today's rule can be used for such purpose.

b. Importers

For imported renewable fuel under RFS2, we are anticipating the importer to be the primary party responsible for generating RINs. However, the foreign producer of renewable fuel can instead elect to generate RINs themselves under certain conditions as described more fully in Section II.D.2.c below. This approach is consistent with the approach under RFS1.

Under RFS1, importers who import more than 10,000 gallons in a calendar year were required to generate RINs for all imported renewable fuel based on its type, except for cases in which the foreign producer generated RINs for cellulosic biomass ethanol or waste-derived ethanol. Due to the new definitions of renewable fuel and renewable biomass in EISA, importers can no longer generate RINs under RFS2 on the basis of fuel type alone. Instead, they must be able to demonstrate that the renewable biomass definition has been met for the renewable fuel they intend to import and for which they will generate RINs. They must also have sufficient information about the feedstock and process used to make the renewable fuel to allow them to identify the appropriate D code from the lookup table for the RINs they generate. Therefore, in order to generate RINs, the importer will be required to obtain this information from a foreign producer. RINs can only be generated if a demonstration is made that the feedstocks used to produce the renewable fuel meet the definition of renewable biomass.

In summary, under today's final rule, importers can import any renewable fuel, but can only generate RINs to represent the imported renewable fuel under the two conditions described below. If these conditions do not apply, the importer can import biofuel but cannot generate RINs to represent that biofuel.

1. The imported renewable fuel is not accompanied by RINs generated by the registered foreign producer

2. The importer obtains from the foreign producer:

—Documentation demonstrating that the renewable biomass definition has been met for the volume of renewable fuel being imported.

—Documentation about the feedstock and production process used to produce the renewable fuel to allow the importer to determine the appropriate D-code designation in the RINs generated.

We are also finalizing additional requirements for foreign producers who either generate RINs or provide documentation to an importer sufficient to allow the importer to generate RINs. As described more fully in the next section, these additional requirements include restrictions on mixing of biofuels in the distribution system as it travels from the foreign producer to the importer.

Finally, EPA is assessing whether additional requirements on foreign-generated fuel may be necessary for situations in which importers are generating RINs for the fuel. Additional requirements may be necessary to ensure that the importers have sufficient information to properly generate the RINs and that EPA has sufficient information to determine whether those RINs have been legitimately generated. EPA will pursue an amendment to the final RFS2 regulations if we find that additional requirements are appropriate and necessary.

c. Additional Provisions for Foreign Producers

In general, we are requiring foreign producers of renewable fuel to meet the same requirements as domestic producers with respect to registration, recordkeeping and reporting, attest engagements, and the transfer of RINs they generate with the batches of renewable fuel that those RINs represent. However, we are also placing additional requirements on foreign producers to ensure that RINs entering the U.S. are valid and that the regulations can be enforced at foreign facilities. These additional requirements are designed to accommodate the more limited access that EPA enforcement personnel have to foreign entities that are regulated parties under RFS2, and also the fact that foreign-produced biofuel intended for export to the U.S. is often mixed with biofuel that will not be exported to the U.S.

Under RFS1, foreign producers had the option of generating RINs for the renewable fuel that they export to the U.S. if they wanted to designate their fuel as cellulosic biomass ethanol or waste-derived ethanol, and thereby take advantage of the additional 1.5 credit value afforded by the 2.5 Equivalence Value for such products. In order to ensure that EPA had the ability to enforce the regulations relating to the generation of RINs from such foreign ethanol producers, the RFS1 regulations specified additional requirements for them, including posting a bond, admitting EPA enforcement personnel, and submitting to third-party engineering reviews of their production process. For RFS2, we are maintaining these additional requirements for foreign producers because EPA enforcement personnel have the same limitations under RFS2 with regard to access to foreign entities that are regulated parties as they did under RFS1.

EISA also creates other unique challenges in the implementation and enforcement of the renewable fuel standards for foreign-produced renewable fuel imported into the U.S. Unlike our other fuels programs, EPA cannot determine whether a particular shipment of renewable fuel is eligible to generate RINs under the new program by testing the fuel itself. Instead, information regarding the feedstock that was used to produce renewable fuel and the process by which it was produced is vital to determining the proper renewable fuel category and RIN type for the imported fuel under the RFS2 program. Thus, whether foreign producers or importers generate RINs, this information must be collected and maintained by the RIN generator.

If a foreign producer generates RINs for renewable fuel that it produces and exports to the U.S., we are requiring that ethanol must be dewatered and denatured by the foreign producer prior to leaving the production facility and prior to the generation of RINs. This is consistent with our definition of renewable fuel in which ethanol that is valid under RFS2 must be denatured. Moreover, the foreign producer is required to strictly segregate a batch of renewable fuel and its associated RINs from all other volumes of renewable fuel as it travels from the foreign producer to the importer. The strict segregation ensures that RINs entering the U.S. appropriately represent the renewable fuel imported into the U.S. both in terms of renewable fuel type and volume.

Several commenters requested that in general the importer be the RIN generator for imported renewable fuel. Since most imported ethanol is currently made in Brazil and is not denatured by the foreign producer, any RINs generated must be generated by the importer. However, to accomplish this, the importer must obtain the appropriate information from a foreign producer regarding compliance with the renewable biomass definition and a description of the associated pathway for the renewable fuel. Under these circumstances, the foreign producer must ensure that the information is transferred along with the renewable fuel through the distribution system until it reaches the importer. The foreign producer's volume of renewable fuel need not be strictly segregated from other volumes in this case, so long as a volume of chemically indistinguishable renewable fuel is tracked through the distribution system from the foreign producer to the importer, and the information needed by the importer to generate RINs follows this same path through the distribution system. Strict segregation of the volume is not necessary in this case, and the importer will determine appropriate number of RINs for the specific volume and type of renewable fuel that he imports.

Finally, if a foreign producer chooses not to participate in the RFS2 program and thus neither generates RINs nor provides information to the importer so that the importer can generate RINs, the foreign producer can still export biofuel to the U.S. However, under these circumstances the biofuel would not be renewable fuel under RFS2, no RINs could be generated by any party, and thus the foreign producer would not be subject to any of the registration, recordkeeping, reporting, or attest engagement requirements.

3. Facilities With Multiple Applicable Pathways

If a given facility's operations can be fully represented by a single pathway, then a single D code taken from the lookup table will be applicable to all RINs generated for fuel produced at that facility. However, we recognize that this will not always be the case. Some facilities use multiple feedstocks at the same time, or switch between different feedstocks over the course of a year. A facility may be modified to produce the same fuel but with a different process, or may be modified to produce a different type of fuel. Any of these situations could result in multiple pathways being applicable to a facility, and thus there may be more than one applicable D code for various RINs generated at the facility.

If more than one pathway applies to a facility within a compliance period, no special steps will need to be taken if the D code is the same for all the applicable pathways. In this case, all RINs generated at the facility will have the same D code regardless. Such a producer with multiple applicable pathways must still describe its feedstock(s), fuel type(s), and production process(es) in its initial registration and annual report to the Agency so that we can verify that the D code used was appropriate.

However, if more than one pathway applies to a facility within a compliance period and these pathways have been assigned different D codes, then the producer must determine which D codes to use when generating RINs. There are a number of different ways that this could occur. For instance, a producer could change feedstocks, production processes, or the type of fuel he produces in the middle of a compliance period. Or, he could use more than one feedstock or produce more than one fuel type simultaneously. The approach we are finalizing for designating D codes for RINs in these cases follows the approach described in the NPRM and is summarized in Table II.D.3-1.

Table II.D.3-1—Approach To Assigning Multiple D Codes for Multiple Applicable Pathways Back to Top
Case/Description Proposed approach
1. The pathway applicable to a facility changes on a specific date, such that one single pathway applies before the date and another single pathway applies on and after the date The applicable D code used in generating RINs must change on the date that the fuel produced changes pathways.
2. One facility produces two or more different types of renewable fuel at the same time The volumes of the different types of renewable fuel should be measured separately, with different D codes applied to the separate volumes.
3. One facility uses two or more different feedstocks at the same time to produce a single type of renewable fuel For any given batch of renewable fuel, the producer should assign the applicable D codes using a ratio (explained below) defined by the amount of each type of feedstock used.

Commenters were generally supportive of this approach to multiple applicable pathways, and as a result we are finalizing it with few modifications from the proposal. Further discussion of the comments we received can be found in Section 3.5.4 of the S&A document.

Following our proposal, cases listed in Table II.D.3-1 will be treated as hierarchical, with Case 2 only being used to address a facility's circumstances if Case 1 is not applicable, and Case 3 only being used to address a facility's circumstances if Case 2 is not applicable. This approach covers all likely cases in which multiple applicable pathways may apply to a renewable fuel producer. Some examples of how Case 2 or 3 would apply are provided in the NPRM.

A facility where two or more different types of feedstock are used to produce a single fuel (such as Case 3 in Table II.D.3-1) will be required to generate two or more separate batch-RINs [15] for a single volume of renewable fuel, and these separate batch-RINs will have different D codes. The D codes will be chosen on the basis of the different pathways as defined in the lookup table in § 80.1426(f). The number of gallon-RINs that will be included in each of the batch-RINs will depend on the relative amount of the different types of feedstocks used by the facility. In the NPRM, we proposed to use the relative energy content of the feedstocks to determine how many gallon-RINs should be assigned to each D code. Commenters generally did not address this aspect of our proposal, and we are finalizing it in today's action. Thus, the useable energy content of each feedstock must be used to divide the total number of gallon-RINs generated for a batch of renewable fuel into two or more groups, each corresponding to a different D code. Several separate batch-RINs can then be generated and assigned to the single volume of renewable fuel. The applicable calculations are given in the regulations at § 80.1426(f)(3).

We proposed several elements of the calculation of the useable energy content of the feedstocks, including the following:

1. Only that fraction of a feedstock which is expected to be converted into renewable fuel by the facility can be counted in the calculation, taking into account facility conversion efficiency.

2. The producer of the renewable fuel is required to designate this fraction once each year for the feedstocks processed by his facility during that year, and to include this information as part of his reporting requirements.

3. Each producer is required to designate the energy content (in Btu/lb) once each year of the portion of each of his feedstocks which is converted into fuel. The producer may determine these values for his own feedstocks, or may use default values provided in the regulations at § 80.1426(f)(7).

4. Each producer is required to determine the total mass of each type of feedstock used by the facility on at least a daily basis.

Based on the paucity of comments we received on this issue, we are finalizing the provisions regarding the calculation of useable energy content of the feedstocks as it was proposed in the NPRM. As described in Section II.J, producers of renewable fuel will be required to submit information in their reports on the feedstocks they used, their production processes, and the type of fuel(s) they produced during the compliance period. This will apply to both domestic producers and foreign producers who export any renewable fuel to the U.S. We will use this information to verify that the D codes used in generating RINs were appropriate.

4. Facilities That Co-Process Renewable Biomass and Fossil Fuels

We expect situations to arise in which a producer uses a renewable feedstock simultaneously with a fossil fuel feedstock, producing a single fuel that is only partially renewable. For instance, biomass might be co-fired with coal in a coal-to-liquids (CTL) process that uses Fischer-Tropsch chemistry to make diesel fuel, biomass and waste plastics might be fed simultaneously into a catalytic or gasification process to make diesel fuel, or vegetable oils could be fed to a hydrotreater along with petroleum to produce a diesel fuel. In these cases, the diesel fuel will be only partially renewable. RINs can be generated in such cases, but must be done in such a way that the number of gallon-RINs corresponds only to the renewable portion of the fuel.

Under RFS1, we created a provision to address the co-processing of “renewable crudes” along with petroleum feedstocks to produce a gasoline or diesel fuel that is partially renewable. See 40 CFR 80.1126(d)(6). However, this provision would not apply in cases where either the renewable feedstock or the fossil fuel feedstock is a gas (e.g., biogas, natural gas) or a solid (e.g., biomass, coal). Therefore, we are eliminating the RFS1 provision applicable only to liquid feedstocks and replacing it with a more comprehensive approach that will apply to liquid, solid, or gaseous feedstocks and any type of conversion process. In this final approach, producers are required to use the relative energy content of their renewable and non-renewable feedstocks to determine the renewable fraction of the fuel that they produce. This fraction in turn is used to determine the number of gallon-RINs that should be generated for each batch. Commenters said little about our proposed methodology to use the relative energy content of the feedstocks, and we are therefore finalizing it largely as proposed.

We also requested comment on allowing renewable fuel producers to use an accepted test method to directly measure the fraction of the fuel that is derived from biomass rather than a fossil fuel feedstock. For instance, ASTM D-6866 is a radiocarbon dating test method that can be used to determine the renewable content of transportation fuel. The use of such a test method can be used in lieu of the calculation of the renewable portion of the fuel based on the relative energy content of the renewable biomass and fossil feedstocks. Commenters generally supported the option of using a radiocarbon dating approach. As a result, we believe it would be appropriate and are finalizing a provision to allow parties that co-process renewable biomass and fossil fuels to choose between using the relative energy in the feedstocks or ASTM D-6866 to determine the number of gallon-RINs that should be generated. Regardless of the approach chosen, the producer will still need to separately verify that the renewable feedstocks meet the definition of renewable biomass.

If a producer chose to use the energy content of the feedstocks, the calculation would be similar to the treatment of renewable fuels with multiple D codes as described in Section II.D.3 above. As shown in the regulations at § 80.1426(f)(3), the producer would determine the renewable fuel volume that would be assigned RINs based on the amount of energy in the renewable feedstock relative to the amount of energy in the fossil feedstock. Only one batch-RIN would be generated for a single volume of fuel produced from both a renewable feedstock and a fossil feedstock, and this one batch-RIN must be based on the contribution that the renewable feedstock makes to the total volume of fuel. The calculation of the relative energy contents includes factors that take into account the conversion efficiency of the plant, and as a result potentially different reaction rates and byproduct formation for the various feedstocks will be accounted for. The relative energy content of the feedstocks is used to adjust the basic calculation of the number of gallon-RINs downward from that calculated on the basis of batch fuel volume and the applicable Equivalence Value. The D code that must be assigned to the RINs is drawn from the lookup table in the regulations as if the feedstock was entirely renewable biomass. Thus, for instance, a coal-to-liquids plant that co-processes some cellulosic biomass to make diesel fuel would be treated as a plant that produces only cellulosic diesel for purposes of identifying the appropriate D code for the fraction of biofuel that qualifies as renewable fuel under EISA.

If a producer chose to use D-6866, he would be required to either apply this test to every batch, or alternatively to take samples of every batch of fuel he produced over the course of one month and combine them into a single composite sample. The D-6866 test would then be applied to the composite sample, and the resulting renewable fraction would be applied to all batches of fuel produced in the next month to determine the appropriate number of RINs that must be generated. For the first month, the producer can estimate the non-fossil fraction, and then make a correction as needed in the second month. The producer would be required to recalculate the renewable fraction every subsequent month. See the regulations at § 80.1426(f)(9).

5. Facilities That Process Municipal Solid Waste

As described in Section II.B.4.d, only the separated yard and food waste of municipal solid waste (MSW) are considered to be renewable biomass and may be used to produce renewable fuels under the RFS2 program. While renewable fuel producers may produce fuel from all organic components of MSW, they may generate RINs for only that portion of MSW that qualifies as renewable biomass. We are providing two methods for determining the appropriate number of RINs to generate for each batch of fuel, depending on whether the feedstock is pure food and yard waste, or separated municipal solid waste, as described in Section II.B.4.d. While not all biogenic material in the separated MSW is cellulosic, the vast majority of it is likely to be in most situations. Specifically, separated municipal solid waste may contain some non-biogenic materials such as plastics that were unable to be recycled due to market conditions. We are requiring producers of renewable fuel made from separated municipal solid waste to use the radiocarbon dating method D-6866 to calculate the biogenic fraction, presumed to be composed of cellulosic materials. Therefore, unless a renewable fuel producer is using MSW streams that are clearly not cellulosic, we anticipate that a D code of either 3 or 7 will be appropriate for such RINs. See the regulations at § 80.1426(f).

6. RINless Biofuel

Under the RFS1 program, all renewable fuel made from renewable feedstocks and used as motor vehicle fuel in the U.S. was assigned RINs. Therefore, aside from the very small amounts of biofuel used in nonroad applications or as heating oil, all renewable fuel produced or imported counted towards the mandated volume goals of the RFS program. Although conventional diesel fuel was not subject to the standards under RFS1, all other motor vehicle fuel fell into two groups: fuel subject to the standards, and fuel for which RINs were generated and was used to meet those standards.

Under RFS2, our approach to compliance with the renewable biomass provision will allow the possibility for some biofuel to be produced without RINs. As described in Section II.B.4 above, we are modifying our approach to compliance with the renewable biomass provision so that renewable fuel producers using feedstocks from domestic planted crops and crop residue will be presumed to meet the renewable biomass provision. Under this “aggregate compliance” approach, these producers will be generating RINs for all their renewable fuel. However, producers who use foreign-grown crops or crop residue or other feedstocks such as planted trees or forestry residues will not be able to take advantage of this aggregate compliance approach. Instead, they will be required to demonstrate that their feedstocks meet the renewable biomass definition, including the associated land use restrictions, before they will be permitted to generate RINs. Absent such a demonstration, these producers can still produce biofuel but will not generate RINs. In addition, fuel producers whose fuel does not qualify as renewable fuel under this program because it does not meet the 20% GHG threshold (and is not grandfathered) can still produce biofuel but will not be allowed to generate RINs. Transportation fuel consumed in the U.S. will therefore be comprised of three groups: fuel subject to the standards (gasoline and diesel), fuel for which RINs are generated and will be used to meet those standards, and RINless biofuel. RINless biofuel will not be covered under any aspect of the RFS2 program, despite the fact that in many cases it will meet the EISA definition of transportation fuel upon blending with gasoline or diesel.

In their comments in response to the NPRM, several refiners suggested that RINless biofuel should be treated as an obligated volume similar to gasoline and diesel, and thus be subject to the standards. Doing so would ensure that all transportation fuels are covered under the RFS2 program, consistent with RFS1. Such an approach would also provide renewable fuel producers with an incentive to demonstrate that their feedstocks meet the renewable biomass definition and thus generate RINs for all the biofuel that they produce. There could be less potential for market manipulation on the part of biofuel producers who might be considering producing RINless biofuel as a means for increasing demand for renewable fuel and RINs.

Nevertheless, we do not believe that it would be appropriate at this time to finalize a requirement that RINless biofuel be considered an obligated fuel subject to the standards. We did not propose such an approach in the NPRM, and as a result many renewable fuel producers who could be affected did not have an opportunity to consider and comment on it. Moreover, the volume of RINless biofuel is likely to be small compared to the volume of renewable fuel with RINs since RINs have value and producers currently have an incentive to generate them. However, if in the future RIN values should fall—for instance, if crude oil prices rise high enough and the market drives up demand for biofuels—the incentive to demonstrate compliance with the renewable biomass definition may decrease and there may be an increase in the volume of RINless biofuel. Under such circumstances it may be appropriate to reconsider whether RINless biofuel should be designated as an obligated volume subject to the standards.

E. Applicable Standards

The renewable fuel standards are expressed as a volume percentage, and are used by each refiner, blender or importer to determine their renewable fuel volume obligations. The applicable percentages are set so that if each regulated party meets the percentages, then the amount of renewable fuel, cellulosic biofuel, biomass-based diesel, and advanced biofuel used will meet the volumes specified in Table I.A.1-1. [16]

The formulas finalized today for use in deriving annual renewable fuel standards are based in part on an estimate of combined gasoline and diesel volumes, for both highway and nonroad uses, for the year in which the standards will apply. The standards will apply to refiners, blenders, and importers of these fuels. As described more fully in Section II.F.3, other producers of transportation fuel, such as producers of natural gas, propane, and electricity from fossil fuels, are not subject to the standards. Since the standards apply to refiners, blenders and importers of gasoline and diesel, these are also the transportation fuels that are used to determine the annual volume obligations of an individual refiner, blender, or importer.

The projected volumes of gasoline and diesel used to calculate the standards will continue to be provided by EIA's Short-Term Energy Outlook (STEO). The standards applicable to a given calendar year will be published by November 30 of the previous year. Gasoline and diesel volumes will continue to be adjusted to account for the required renewable fuel volumes. In addition, gasoline and diesel volumes produced by small refineries and small refiners will be exempt through 2010, and that year's standard is adjusted accordingly, as discussed below.

As discussed in the proposal, four separate standards are required under the RFS2 program, corresponding to the four separate volume requirements shown in Table I.A.1-1. The specific formulas we use to calculate the renewable fuel standards are described below in Section II.E.1.

In order for an obligated party to demonstrate compliance, the percentage standards are converted into the volume of renewable fuel each obligated party is required to satisfy. This volume of renewable fuel is the volume for which the obligated party is responsible under the RFS program, and continues to be referred to as its Renewable Volume Obligation (RVO). Since there are four separate standards under the RFS2 program, there are likewise four separate RVOs applicable to each obligated party. Each standard applies to the sum of all gasoline and diesel produced or imported. Determination of RVOs is discussed in Section II.G.2.

1. Calculation of Standards

a. How Are the Standards Calculated?

The four separate renewable fuel standards are based primarily on (1) the 49-state [17] gasoline and diesel consumption volumes projected by EIA, and (2) the total volume of renewable fuels required by EISA for the coming year. Table I.A.2-1 shows the required overall volumes of four types of renewable fuel specified in EISA. Each renewable fuel standard is expressed as a volume percentage of combined gasoline and diesel sold or introduced into commerce in the U.S., and is used by each obligated party to determine its renewable volume obligation.

Today we are finalizing an approach to setting standards that is based in part on the sum of all gasoline and diesel produced or imported in the 48 contiguous states and Hawaii. An approach we are not adopting but which we discussed in the proposal would have split the standards between those that would be specific to gasoline and those that would be specific to diesel. Though this approach to setting standards would more readily align the RFS obligations with the relative amounts of gasoline and diesel produced or imported by each obligated party, we are not adopting this approach because it relies on projections of the relative amounts of gasoline-displacing and diesel-displacing renewable fuels. These projections would need to be updated every year, and as stated in the proposal, we believe that such an approach would unnecessarily complicate the program.

While the required amount of total renewable fuel for a given year is provided by EISA, the Act requires EPA to base the standards on an EIA estimate of the amount of gasoline and diesel that will be sold or introduced into commerce for that year. As discussed in the proposal, EIA's STEO will continue to be the source for projected gasoline, and now diesel, consumption estimates. In order to achieve the volumes of renewable fuels specified in EISA, the gasoline and diesel volumes used to determine the standard must be the non-renewable portion of the gasoline and diesel pools. Because the STEO volumes include renewable fuel use, we must subtract the total renewable fuel volume from the total gasoline and diesel volume to get total non-renewable gasoline and diesel volumes. The Act also requires EPA to use EIA estimates of renewable fuel volumes; the best estimation of the coming year's renewable fuel consumption is found in Table 8 (U.S. Renewable Energy Supply and Consumption) of the STEO. Additional information on projected renewable fuel use will be included as it becomes available.

As discussed in Section II.D.1, we are finalizing the energy content approach to Equivalence Values for the cellulosic biofuel, advanced biofuel, and total renewable fuel standards. However, the biomass-based diesel standard is based on the volume of biodiesel. In order to align both of these approaches simultaneously, biodiesel will continue to generate 1.5 RINs per gallon as in RFS1, and the biomass-based diesel volume mandate from EISA is then adjusted upward by the same 1.5 factor. The net result is a biomass-based diesel gallon being worth 1.0 gallons toward the biomass-based diesel standard, but 1.5 gallons toward the other standards.

CAA section 211(o) exempts small refineries [18] from the RFS requirements until the 2011 compliance period. In RFS1, we extended this exemption to the few remaining small refiners not already exempted. [19] Small refineries and small refiners will continue to be exempt from the program until 2011 under the new RFS2 regulations. Thus we have excluded their gasoline and diesel volumes from the overall non-renewable gasoline and diesel volumes used to determine the applicable percentages until 2011. As discussed in the proposal, total small refinery and small refiner gasoline production volume is expected to be fairly constant compared to total U.S. transportation fuel production. Thus we estimated small refinery and small refiner gasoline and diesel volumes using a constant percentage of national consumption, as we did in RFS1. Using information from gasoline batch reports submitted to EPA for 2006, EIA data, and input from the California Air Resources Board regarding California small refiners, we estimate that small refinery volumes constitute 11.9% of the gasoline pool, and 15.2% of the diesel pool.

CAA section 211(o) requires that the small refinery adjustment also account for renewable fuels used during the prior year by small refineries that are exempt and do not participate in the RFS2 program. Accounting for this volume of renewable fuel would reduce the total volume of renewable fuel use required of others, and thus directionally would reduce the percentage standards. However, as we discussed in RFS1, the amount of renewable fuel that would qualify, i.e., that was used by exempt small refineries and small refiners but not used as part of the RFS program, is expected to be very small. In fact, these volumes would not significantly change the resulting percentage standards. Whatever renewable fuels small refineries and small refiners blend will be reflected as RINs available in the market; thus there is no need for a separate accounting of their renewable fuel use in the equations used to determine the standards. We proposed and are finalizing this value as zero.

The levels of the percentage standards would be reduced if Alaska or a U.S. territory chooses to participate in the RFS2 program, as gasoline and diesel produced in or imported into that state or territory would then be subject to the standard. Section 211(o) of the Clean Air Act requires that the renewable fuel be consumed in the contiguous 48 states, and any other state or territory that opts-in to the program (Hawaii has subsequently opted in). However, because renewable fuel produced in Alaska or a U.S. territory is unlikely to be transported to the contiguous 48 states or to Hawaii, including their renewable fuel volumes in the calculation of the standard would not serve the purpose intended by section 211(o) of the Clean Air Act of ensuring that the statutorily required renewable fuel volumes are consumed in the 48 contiguous states and any state or territory that opts-in. Therefore, renewable fuels used in Alaska or U.S. territories are not included in the renewable fuel volumes that are subtracted from the total gasoline and diesel volume estimates.

In summary, the total projected non-renewable gasoline and diesel volumes from which the annual standards are calculated are based on EIA projections of gasoline and diesel consumption in the contiguous 48 states and Hawaii, adjusted by constant percentages of 11.9% and 15.2% in 2010 to account for small refinery/refiner gasoline and diesel volumes, respectively, and with built-in correction factors to be used when and if Alaska or a territory opt-in to the program.

The following formulas are used to calculate the percentage standards:

[Graphic not available; view image of printed page]

[Graphic not available; view image of printed page]

[Graphic not available; view image of printed page]

[Graphic not available; view image of printed page]

Where

Std CB,i= The cellulosic biofuel standard for year i, in percent

Std BBD,i= The biomass-based diesel standard (ethanol-equivalent basis) for year i, in percent

Std AB,i= The advanced biofuel standard for year i, in percent

Std RF,i= The renewable fuel standard for year i, in percent

RFV CB,i= Annual volume of cellulosic biofuel required by section 211(o)(2)(B) of the Clean Air Act for year i, in gallons

RFV BBD,i= Annual volume of biomass-based diesel required by section 211(o)(2)(B) of the Clean Air Act for year i, in gallons

RFV AB,i= Annual volume of advanced biofuel required by section 211(o)(2)(B) of the Clean Air Act for year i, in gallons

RFV RF,i= Annual volume of renewable fuel required by section 211(o)(2)(B) of the Clean Air Act for year i, in gallons

G i= Amount of gasoline projected to be used in the 48 contiguous states and Hawaii, in year i, in gallons*

D i= Amount of diesel projected to be used in the 48 contiguous states and Hawaii, in year i, in gallons

RG i= Amount of renewable fuel blended into gasoline that is projected to be consumed in the 48 contiguous states and Hawaii, in year i, in gallons

RD i= Amount of renewable fuel blended into diesel that is projected to be consumed in the 48 contiguous states and Hawaii, in year i, in gallons

GS i= Amount of gasoline projected to be used in Alaska or a U.S. territory in year i if the state or territory opts-in, in gallons*

RGS i= Amount of renewable fuel blended into gasoline that is projected to be consumed in Alaska or a U.S. territory in year i if the state or territory opts-in, in gallons

DS i= Amount of diesel projected to be used in Alaska or a U.S. territory in year i if the state or territory opts-in, in gallons *

RDS i= Amount of renewable fuel blended into diesel that is projected to be consumed in Alaska or a U.S. territory in year i if the state or territory opts-in, in gallons

GE i= The amount of gasoline projected to be produced by exempt small refineries and small refiners in year i, in gallons, in any year they are exempt per §§ 80.1441 and 80.1442, respectively. Equivalent to 0.119*(G i−RG i).

DE i= The amount of diesel projected to be produced by exempt small refineries and small refiners in year i, in gallons, in any year they are exempt per §§ 80.1441 and 80.1442, respectively. Equivalent to 0.152*(D i−RD i).

* Note that these terms for projected volumes of gasoline and diesel use include gasoline and diesel that has been blended with renewable fuel.

b. Standards for 2010

We are finalizing the standards for 2010 in today's action. As explained in Section I.A.2, while the rulemaking is not effective until July 1, 2010, the 2010 standards we are setting are annual standards with compliance demonstrations are due by February 28, 2011.

Under CAA section 211(o)(7)(D)(i), EPA is required to make a determination each year regarding whether the required volumes of cellulosic biofuel for the following year can be produced. For any calendar year for which the projected volume of cellulosic biofuel production is less than the minimum required volume, the projected volume becomes the basis for the cellulosic biofuel standard. In such a case, the statute also indicates that EPA may also lower the required volumes for advanced biofuel and total renewable fuel.

As discussed in Section IV.B., we are utilizing the EIA projection of 5.04 million gallons (6.5 million ethanol equivalent gallons) of cellulosic biofuel as the basis for setting the percentage standard for cellulosic biofuel for 2010. This is lower than the 100 million gallon standard set by EISA that we proposed upholding, but reflects the current state of the industry, as discussed in section V.B. We expect continued growth in the industry in 2011 and beyond. Since the advanced biofuel standard is met by just the biomass-based diesel volume required in 2010, and additional volumes of other advanced biofuels (e.g., sugarcane ethanol) are available as well, no change to the advanced biofuel standard is necessary for 2010. Moreover, given the nested nature of the volume mandates, since no change in the advanced biofuel standard is necessary, the total renewable fuel standard need not be changed either.

Table II.E.1. b-1—Standards for 2010 Back to Top
Percent
Cellulosic biofuel 0.004
Biomass-based diesel 1.10
Advanced biofuel 0.61
Renewable fuel 8.25

2. Treatment of Biomass-Based Diesel in 2009 and 2010

As described in Section I.A.2, the four separate 2010 standards issued in today's rule will apply to all gasoline and diesel produced in 2010. However, EISA included volume mandates for biomass-based diesel, advanced biofuel, and total renewable fuel that applied in 2009. Since the RFS2 program was not effective in 2009 and thus the volume mandates for biomass-based diesel and advanced biofuel were not implemented in 2009, our NPRM proposed a mechanism to ensure that the 2009 biomass-based diesel volume mandate would eventually be met. In today's final rule we are finalizing the proposed approach.

a. Shift in 2009 Biomass-Based Diesel Compliance Demonstration to 2010

Under the RFS1 regulations that applied in 2009, we set the applicable standard for total renewable fuel in November 2008 [20] using the required volume of 11.1 billion gallons specified in the Clean Air Act (as amended by EISA), gasoline volume projections from EIA, and the formula provided in the regulations at § 80.1105(d). The existing RFS1 regulations did not provide a mechanism for requiring the use of 0.5 billion gallons of biomass-based diesel or the 0.6 billion gallons of advanced biofuel mandated by EISA for 2009.

In the NPRM we proposed that the compliance demonstration for the 2009 biomass-based diesel requirement of 0.5 bill gal be extended to 2010. This approach would combine the 0.5 bill gal requirement for 2009 and the 0.65 bill gal requirement for 2010 into a single requirement of 1.15 bill gal for which compliance demonstrations would be made by February 28, 2011. As described in the NPRM, we believe that the deficit carryover provision provides a conceptual mechanism for this approach, since it would have allowed obligated parties to defer compliance with any or all of the 2009 standards until 2010. We are finalizing this approach in today's action. We believe it will ensure that these two year's worth of biomass-based diesel will be used, while providing reasonable lead time for obligated parties. It avoids a transition that fails to have any requirements related to the 2009 biomass-based diesel volume, and instead requires the use of the 2009 volume but achieves this by extending the compliance period by one year. We believe this is a reasonable exercise of our authority under section 211(o)(2) to issue regulations that ensure that the volumes for 2009 are ultimately used, even though we were unable to issue final regulations prior to the 2009 compliance year. We announced our intentions to implement the 2009 and 2010 biomass-based diesel requirements in this manner in the November 2008 Federal Register notice cited previously. We reiterated these intentions in our NPRM. Thus, obligated parties will have had sufficient lead time to acquire a sufficient number of biomass-based diesel RINs by the end of 2010 to comply with the standard based on 1.15 bill gal.

Data available at the time of this writing suggests that approximately 450 million gallons of biodiesel was produced in 2009, thus requiring 700 million gallons to be produced in 2010 to satisfy the combined 2009 and 2010 volume mandates. Information from commenters and other contacts in the biodiesel industry indicate that feedstocks and production facilities will be available in 2010 to produce this volume.

Refiners generally commented that the proposed approach to 2009 and 2010 biomass-based diesel volumes was not appropriate and should not be implemented. They also recommended that the RFS2 program should be made effective on January 1, 2011 with no carryover of any previous-year obligations for biomass-based diesel or any other volume mandate. In contrast, the National Biodiesel Board and several individual biodiesel producers supported the proposed approach, but believed it was insufficient to compel obligated parties to purchase biodiesel in 2009, something they considered critical to the survival of the biodiesel industry. Many of these commenters requested that we conduct an interim rulemaking that would apply to 2009 to implement the EISA mandated volume of 0.5 billion gallons of biomass-based diesel. If the RFS2 program could not be implemented until 2011, they likewise requested that interim measures be taken for 2010 to ensure that the full 1.15 bill gal requirement would be implemented. However, putting in place this new volume requirement without also putting in place EISA's new definition for biomass-based diesel, renewable fuel, and renewable biomass would have raised significant legal and policy issues that would necessarily have required a new proposal with its own public notice and comment process. Because of the significant time required for notice and comment rulemaking, the need to provide industry with adequate lead time for new requirements, and the fact that we were already well into calendar year 2009 at the time the request for an interim rule was received, it was unlikely that any interim rule could have impacted biodiesel demand in 2009. Moreover, Agency resources applied to the interim rulemaking would have been unavailable for development of the final RFS2 rulemaking. Developing an interim rule could have undermined EPA's ability to complete the full RFS2 program regulations in time for 2010 implementation. As a result, we did not pursue an interim rulemaking.

With regard to advanced biofuel, it is not necessary to implement a separate requirement for the 0.6 billion gallon mandate for 2009. Due to the nested nature of the volume requirements and the fact that Equivalence Values will be based on the energy content relative to ethanol, the 0.5 billion gallon requirement for biomass-based diesel will count as 0.75 billion gallons of advanced biofuel, exceeding the requirement of 0.6 billion gallons. Thus compliance with the biomass-based diesel requirement in 2009 automatically results in compliance with the advanced biofuel standard.

All 2009 biodiesel and renewable diesel RINs, identifiable through an RR code of 15 or 17 respectively under the RFS1 regulations, will be valid for showing compliance with the adjusted 2010 biomass-based diesel standard of 1.15 billion gallons. This use of previous year RINs for current year compliance is consistent with our approach to any other standard for any other year and consistent with the flexibility available to any obligated party that carries a deficit from one year to the next. Moreover, it allows an obligated party to acquire sufficient biodiesel and renewable diesel RINs during 2009 to comply with the 0.5 billion gallons requirement, even though their compliance demonstration would not occur until the 2010 compliance period.

We did not reduce the 2009 volume requirement for total renewable fuel by 0.5 billion gallons to account for the fact that we intended to move the compliance demonstration for this volume has been moved to the 2010 compliance period. Instead, we are allowing 2009 biodiesel and renewable diesel RINs to be used for compliance purposes for both the 2009 total renewable fuel standard as well as the 2010 adjusted biomass-based diesel standard (but not for the 2010 advanced biofuel or total renewable fuel standards). To accomplish this, we proposed in the NPRM that an obligated party would add up the 2009 biodiesel and renewable diesel RINs that he used for 2009 compliance with the RFS1 standard for total renewable fuel, and reduce his 2010 biomass-based diesel obligation by this amount. Thus, 2009 biodiesel and renewable diesel RINs are essentially used twice. Any remaining 2010 biomass-based diesel obligation would need to be covered either with 2009 biodiesel and renewable diesel RINs that were not used for compliance in 2009 or with 2010 biomass-based diesel RINs. We are finalizing this approach in today's notice.

b. Treatment of Deficit Carryovers, RIN Rollover, and RIN Valid Life for Adjusted 2010 Biomass-Based Diesel Requirement

Our transition approach for biomass-based diesel is conceptually similar, but not identical, to the statutory deficit carryover provision. In a typical deficit carryover situation, an obligated party can carry forward any amount of a current-year deficit to the following year. In the absence of any modifications to the deficit carryover provisions for our biomass-based diesel transition provisions, then, an obligated party that did not fully comply with the 2010 biomass-based diesel requirement of 1.15 billion gallons could carry a deficit of any amount into 2011. As described in the NPRM, we believe that the deficit carryover provisions should be modified in the context of the transition biomass-based diesel approach to more closely represent what would have occurred if we had been able to implement the 0.5 bill gal requirement in 2009. Specifically, we are prohibiting obligated parties from carrying over a biomass-based diesel deficit into 2011 larger than that based on the 0.65 bill gal volume requirement for 2010. This is the amount that would have been permitted had we been able to implement the biomass-based diesel requirements in 2009. In practice, this means that deficit carryovers from 2010 into 2011 for biomass-based diesel cannot not exceed 57% (0.65/1.15) of an obligated party's 2010 RVO. This approach also helps to ensure a minimum volume mandate for companies producing biomass-based diesel each year.

Similarly, in the absence of any modifications to the provisions regarding valid life of RINs, 2008 biodiesel and renewable diesel RINs could not be used for compliance in 2010 with the adjusted biomass-based diesel standard, despite the fact that the 2010 standard includes the 2009 requirement for which 2008 RINs should be valid. The National Biodiesel Board opposed this approach on the basis that the use of 2008 RINs for 2010 compliance demonstrations violated the 2-year valid life limit for RINs. However, since the 2010 compliance demonstration will include the obligation that would have applied in 2009, and 2008 RINs would be valid for 2009 compliance, we are allowing excess 2008 biodiesel and renewable diesel RINs that were not used for compliance purposes in 2008 to be used for compliance purposes in 2009 or 2010.

As described in Section III.D, we are requiring the 20% RIN rollover cap to apply in all years, and separately for all four standards. However, consistent with our approach to deficit carryovers, we believe that an additional constraint is warranted in the application of the rollover cap to the biomass-based diesel obligation in the 2010 compliance year to more closely represent what would have occurred if we had been able to implement the 0.5 bill gal requirement in 2009. Specifically, we are limiting the use of excess 2008 RINs to 20% of the statutory 2009 requirement of 0.5 bill gal. This is equivalent to 0.1 bill gal (20% of 0.5 bill gal), or 8.7% of the combined 2009/2010 obligation of 1.15 bill gal (0.1/1.15). Thus, obligated parties will be allowed to use excess 2008 and 2009 biodiesel and renewable diesel RINs for compliance with the 2010 combined standard of 1.15 bill gal, so long as the sum of all previous-year RINs (2008 plus 2009 RINs) does not exceed 20% of their 2010 obligation, and the 2008 RINs do not exceed 8.7% of their 2010 obligation.

Under RFS1, RINs are generated when renewable fuel is produced, but if the fuel is ultimately used for purposes other than as motor vehicle fuel the RINs must generally be retired. Under EISA, however, RINs generated for renewable fuel that is ultimately used for nonroad purposes, heating oil, or jet fuel are valid for compliance purposes. To more closely align our transition approach for biomass-based diesel to what could have occurred if we had issued the RFS2 standards prior to 2009, we are allowing 2009 RINs that are retired because they are ultimately used for nonroad, heating oil or jet fuel purposes to be valid for compliance with the 2010 standards. Such RINs can be reinstated by the retiring party in 2010.

3. Future Standards

The statutorily-prescribed phase-in period ends in 2012 for biomass-based diesel and in 2022 for cellulosic biofuel, advanced biofuel, and total renewable fuel. Beyond these years, EISA requires EPA to determine the applicable volumes based on a review of the implementation of the program up to that time, and an analysis of a wide variety of factors such as the impact of the production of renewable fuels on the environment, energy security, infrastructure, costs, and other factors. For these future standards, EPA must promulgate rules establishing the applicable volumes no later than 14 months before the first year for which such applicable volumes would apply. For biomass-based diesel, this would mean that final rules would need to be issued by October 31, 2011 for application starting on January 1, 2013. In today's rulemaking, we are not suggesting any specific volume requirements for biomass-based diesel for 2013 and beyond that would be appropriate under the statutory criteria that we must consider. Likewise, we are not suggesting any specific volume requirements for the other three renewable fuel categories for 2023 and beyond. However, the statute requires that the biomass-based diesel volume in 2013 and beyond must be no less than 1.0 billion gallons, and that advanced biofuels in 2023 and beyond must represent at a minimum the same percentage of total renewable fuel as it does in 2022. These provisions will be implemented as part of an annual standard-setting process.

F. Fuels That Are Subject to the Standards

Under RFS1, producers and importers of gasoline are obligated parties subject to the standards—any party that produces or imports only diesel fuel is not subject to the standards. EISA changes this provision by expanding the RFS program in general to include all transportation fuel. As discussed above, however, section 211(o)(3) continues to require EPA to determine which refiners, blenders, and importers are treated as subject to the standard. As described further in Section II.G below, under this rule, the sum of all highway and nonroad gasoline and diesel fuel produced or imported within a calendar year will be the basis on which the RVOs are calculated. This section provides our final definition of gasoline and diesel for the purposes of the RFS2 program.

1. Gasoline

As with the RFS1 rule, the volume of gasoline used in calculating the RVO under RFS2 will continue to include all finished gasoline (reformulated gasoline (RFG) and conventional gasoline (CG)) produced or imported for use in the contiguous United States or Hawaii, as well as all unfinished gasoline that becomes finished gasoline upon the addition of oxygenate blended downstream from the refinery or importer. This includes both unfinished reformulated gasoline, called “reformulated gasoline blendstock for oxygenate blending,” or “RBOB,” and unfinished conventional gasoline designed for downstream oxygenate blending (e.g., sub-octane conventional gasoline), called “CBOB.” The volume of any other unfinished gasoline or blendstock, (such as butane or naphtha produced in a refinery) or exported gasoline, will not be included in the obligated volume, except where the blendstock is combined with other blendstock or gasoline to produce finished gasoline, RBOB, or CBOB. Where a blendstock is blended with other blendstock to produce finished gasoline, RBOB, or CBOB, the total volume of the gasoline blend will be included in the volume used to determine the blender's renewable fuels obligation. Where a blendstock is added to finished gasoline, only the volume of the blendstock will be included, since the finished gasoline would have been included in the compliance determinations of the refiner or importer of the gasoline. For purposes of this preamble, the various gasoline products described above that we are including in a party's obligated volume are collectively called “gasoline.”

Also consistent with the RFS1 program, we are continuing the exclusion of any volume of renewable fuel contained in gasoline from the volume of gasoline used to determine the renewable fuels obligations. This exclusion applies to any renewable fuels that are blended into gasoline at a refinery, contained in imported gasoline, or added at a downstream location. Thus, for example, any ethanol added to RBOB or CBOB at a refinery's rack or terminal downstream from the refinery or importer will be excluded from the volume of gasoline used by the refiner or importer to determine the obligation. This is consistent with how the standard itself is calculated—EPA determines the applicable percentage by comparing the overall projected volume of gasoline used to the overall renewable fuel volume that is specified in the statute, and EPA excludes ethanol and other renewable fuels that are blended into the gasoline in determining the overall projected volume of gasoline. When an obligated party determines their RVO by applying the applicable percentage to the amount of gasoline they produce or import, it is consistent to also exclude ethanol and other renewable fuel blends from the calculation of the volume of gasoline produced.

As with the RFS1 rule, Gasoline Treated as Blendstock (GTAB) will continue to be treated as a blendstock under the RFS2 program, and thus will not count towards a party's renewable fuel obligation. Where the GTAB is blended with other blendstock (other than renewable fuel) to produce gasoline, the total volume of the gasoline blend, including the GTAB, will be included in the volume of gasoline used to determine the renewable fuel obligation. Where GTAB is blended with renewable fuel to produce gasoline, only the GTAB volume will be included in the volume of gasoline used to determine the renewable fuel obligation. Where the GTAB is blended with finished gasoline, only the GTAB volume will be included in the volume of gasoline used to determine the renewable fuel obligation.

2. Diesel

EISA expanded the RFS program to include transportation fuels other than gasoline, thus both highway and nonroad diesel must be used in calculating a party's RVO. Any party that produces or imports petroleum-based diesel fuel that is designated as motor vehicle, nonroad, locomotive, and marine diesel fuel (MVNRLM) (or any subcategory of MVNRLM) will be required to include the volume of that diesel fuel in the determination of its RVO under the RFS2 rule. Diesel fuel includes any distillate fuel that meets the definition of MVNRLM diesel fuel as it has already been defined in the regulations at § 80.2(qqq), including any subcategories such as MV (motor vehicle diesel fuel produced for use in highway diesel engines and vehicles), NRLM (diesel fuel produced for use in nonroad, locomotive, and marine diesel engines and equipment/vessels), NR (diesel fuel produced for use in nonroad engines and equipment), and LM (diesel fuel produced for use in locomotives and marine diesel engines and vessels). [21] Transportation fuels meeting the definition of MVNRLM will be used to calculate the RVOs, and refiners, blenders, or importers of MVNRLM will be treated as obligated parties. As such, diesel fuel that is designated as heating oil, jet fuel, or any designation other than MVNRLM or a subcategory of MVNRLM, will not be subject to the applicable percentage standard and will not be used to calculate the RVOs. [22] We requested comment on the idea that any diesel fuel not meeting these requirements, such as distillate or residual fuel intended solely for use in ocean-going vessels, would not be used to calculate the RVOs.

One commenter expressed support for including heating oil and jet fuel into the RIN program, but not to subject these fuels to the RVO mandate. The commenter stated that fluctuating weather conditions make it hard to predict with any reliability the volumes of heating oil that will be used in a given year. Another commenter stated that it supports the extension of the RFS program to transportation fuels, including diesel and nonroad fuels.

With respect to fuels for use in ocean-going vessels, EISA specifies that “transportation fuels” do not include such fuels. We are interpreting that “fuels for use in ocean-going vessels” means residual or distillate fuels other than MVNRLM intended to be used to power large ocean-going vessels (e.g., those vessels that are powered by Category 3 (C3), and some Category 2 (C2), marine engines and that operate internationally). Thus, fuel for use in ocean-going vessels, or that an obligated party can verify as having been used in an ocean-going vessel, will be excluded from the renewable fuel standards. Also, in the context of the recently finalized fuel standards for C3 marine vessels, this would mean that fuel meeting the 1,000 ppm fuel sulfur standard would not be considered obligated volume, while all MVNRLM diesel fuel would.

3. Other Transportation Fuels

Transportation fuels other than gasoline or MVNRLM diesel fuel (natural gas, propane, and electricity) will not be used to calculate the RVOs of any obligated party. We believe this is a reasonable way to implement the obligations of 211(o)(3) because the volumes are small and the producers cannot readily differentiate the small portion used in the transportation sector from the large portion used in other sectors (in fact, the producer may have no knowledge of its ultimate use). We will reconsider this approach if and when these volumes grow. At the same time, it is clear that these fuels can be used as transportation fuel, and under certain circumstances, producers of such “other transportation fuels” may generate RINs as a producer or importer of a renewable fuel. See Section II.D.2.a for further discussion of other RIN-generating fuels.

G. Renewable Volume Obligations (RVOs)

Under RFS1, each obligated party was required to determine its RVO based on the applicable percentage standard and its annual gasoline volume. The RVO represented the volume of renewable fuel that the obligated party was required to ensure was used in the U.S. in a given calendar year. Obligated parties were required to meet their RVO through the accumulation of RINs which represent the amount of renewable fuel used as motor vehicle fuel that was sold or introduced into commerce within the U.S. Each gallon-RIN counted as one gallon of renewable fuel for compliance purposes.

We are maintaining this approach to compliance under the RFS2 program. However, one primary difference between RFS1 and the new RFS2 program in terms of demonstrating compliance is that each obligated party now has four RVOs instead of one (through 2012) or two (starting in 2013) under the RFS1 program. Also, as discussed above, RVOs are now calculated based on production or importation of both gasoline and diesel fuels, rather than gasoline alone.

By acquiring RINs and applying them to their RVOs, obligated parties are deemed to have satisfied their obligation to cause the renewable fuel represented by the RINs to be consumed as transportation fuel in highway or nonroad vehicles or engines. Obligated parties are not required to physically blend the renewable fuel into gasoline or diesel fuel themselves. The accumulation of RINs will continue to be the means through which each obligated party shows compliance with its RVOs and thus with the renewable fuel standards.

If an obligated party acquires more RINs than it needs to meet its RVOs, then in general it can retain the excess RINs for use in complying with its RVOs in the following year (subject to the 20% rollover cap discussed in Section III.D) or transfer the excess RINs to another party. If, alternatively, an obligated party has not acquired sufficient RINs to meet its RVOs, then under certain conditions it can carry a deficit into the next year.

This section describes our approach to the calculation of RVOs under RFS2 and the RINs that are valid for demonstrating compliance with those RVOs. This includes a description of the special treatment that must be applied to RFS1 RINs used for compliance purposes under RFS2, since RINs generated under RFS1 regulations are not exactly the same as those generated in under RFS2.

1. Designation of Obligated Parties

In the NPRM, we proposed to continue to designate obligated parties under the RFS2 program as they were designated under RFS1, with the addition of diesel fuel producers and importers. Regarding gasoline producers and importers, we proposed that obligated parties who are subject to the standard would be those that produce or import finished gasoline (RFG and conventional) or unfinished gasoline that becomes finished gasoline upon the addition of an oxygenate blended downstream from the refinery or importer. Unfinished gasoline would include reformulated gasoline blendstock for oxygenate blending (RBOB), and conventional gasoline blendstock designed for downstream oxygenate blending (CBOB) which is generally sub-octane conventional gasoline. The volume of any other unfinished gasoline or blendstock, such as butane, would not be included in the volume used to determine the RVO, except where the blendstock was combined with other blendstock or finished gasoline to produce finished gasoline, RBOB, or CBOB. Thus, parties downstream of a refinery or importer would only be obligated parties to the degree that they use non-renewable blendstocks to make finished gasoline, RBOB, CBOB, or diesel fuel.

We also took comment on two alternative approaches to the designation of obligated parties:

—Elimination of RBOB and CBOB from the list of fuels that are subject to the standard, such that a party's RVO would be based only on the non-renewable volume of finished gasoline or diesel that he produces or imports, thereby moving a portion of the obligation to downstream blenders of renewable fuels into RBOB and CBOB.

—Moving the obligations for all gasoline and diesel downstream of refineries and importers to parties who supply finished transportation fuels to retail outlets or to wholesale purchaser-consumer facilities.

These alternative approaches have the potential to more evenly align a party's access to RINs with that party's obligations under the RFS2 program. As described more fully in the NPRM, we considered these alternatives because of market conditions that had changed since the RFS1 program began. For instance, obligated parties who have excess RINs have been observed to retain rather than sell them to ensure they have a sufficient number for the next year's compliance. This was most likely to occur with major integrated refiners who operate gasoline marketing operations and thus have direct access to RINs for ethanol blended into their gasoline. Refiners whose operations are focused primarily on producing refined products with less marketing do not have such direct access to RINs and could potentially find it difficult to acquire a sufficient number for compliance despite the fact that the total nationwide volume of renewable fuel meets or exceeds the standard. The result might be a higher price for RINs (and fuel) in the marketplace than would be expected under a more liquid RIN market. For similar reasons, we also took comment on possible changes to the requirement that RINs be transferred with volume through the distribution system as discussed more fully in Section II.H.4.

In response to the NPRM, stakeholders differed significantly on whether EPA should implement one of these alternative approaches. For instance, while some refiners expressed support for moving the obligations to downstream parties such as blenders, terminals, and/or wholesale purchaser-consumers, other refiners preferred to maintain the current approach. Blenders and other downstream parties generally expressed opposition to a change in the designation of obligated parties, citing the additional burden of demonstrating compliance with the standard especially for small businesses. They also pointed to the need to implement new systems for determining and reporting compliance, the short leadtime for doing so, and the fewer resources that smaller downstream companies have to manage such work in comparison to the much larger refiners. Finally, they pointed to the additional complexity that would be added to the RFS program beyond that which is necessary to carry out the renewable fuels mandate under CAA section 211(o).

When the RFS1 regulations were drafted, the obligations were placed on the relatively small number of refiners and importers rather than on the relatively large number of downstream blenders and terminals in order to minimize the number of regulated parties and keep the program simple. However, with the expanded RFS2 mandates, essentially all downstream blenders and terminals are now regulated parties under RFS2 since essentially all gasoline will be blended with ethanol. Thus the rationale in RFS1 for placing the obligation on just the upstream refiners and importers is no longer valid. Nevertheless, based on the comments we received, we do not believe that the concerns expressed warrant a change in the designation of obligated parties for the RFS2 program at this time. We continue to believe that the market will provide opportunities for parties who are in need of RINs to acquire them from parties who have excess. Refiners who market considerably less gasoline or diesel than they produce can establish contracts with splash blenders to purchase RINs. Such refiners can also purchase ethanol from producers directly, separate the RINs, and then sell the ethanol without RINs to blenders. Since the RFS program is based upon ownership of RINs rather than custody of volume, refiners need never take custody of the ethanol in order to separate RINs from volumes that they own. Moreover, a change in the designation of obligated parties would result in a significant change in the number of obligated parties and the movement of RINs, changes that could disrupt the operation of the RFS program during the transition from RFS1 to RFS2.

We will continue to evaluate the functionality of the RIN market. Should we determine that the RIN market is not operating as intended, driving up prices for obligated parties and fuel prices for consumers, we will consider revisiting this provision in future regulatory efforts.

In the NPRM we also took comment on several other possible ways to help ensure that obligated parties can demonstrate compliance. For instance, one alternative approach would have left our proposed definitions for obligated parties in place, but would have added a regulatory requirement that any party who blends ethanol into RBOB or CBOB must transfer the RINs associated with the ethanol to the original producer of the RBOB or CBOB. Stakeholders generally opposed this change, agreeing with our assessment that it would be extremely difficult to implement given that RBOB and CBOB are often transferred between multiple parties prior to ethanol blending. As a result, a regulatory requirement for RIN transfers back to the original producer would have necessitated an additional tracking requirement for RBOB and CBOB so that the blender would know the identity of the original producer. It would also be difficult to ensure that RINs representing the specific category of renewable fuel blended were transferred to the producer of the RBOB or CBOB, given the fungible nature of RINs assigned to batches of renewable fuel. For these reasons, we have not finalized this alternative approach.

Another alternative approach on which we took comment would have allowed use of RINs that expire without being used for compliance by an obligated party to be used to reduce the nationwide volume of renewable fuel required in the following year. This alternative approach could have helped to prevent the hoarding of RINs from driving up demand for renewable fuel. However, it would also effectively alter the valid life limit for RINs. Comments from stakeholders did not change our position that such an approach is not warranted at this time, and thus we have not finalized it.

2. Determination of RVOs Corresponding to the Four Standards

In order for an obligated party to demonstrate compliance, the percentage standards described in Section II.E.1 which are applicable to all obligated parties must be converted into the volumes of renewable fuel each obligated party is required to satisfy. These volumes of renewable fuel are the volumes for which the obligated party is responsible under the RFS program, and are referred to here as its RVO. Under RFS2, each obligated party will need to acquire sufficient RINs each year to meet each of the four RVOs corresponding to the four renewable fuel standards.

The calculation of the RVOs under RFS2 follows the same format as the formulas in the RFS1 regulations at § 80.1107(a), with one modification. The standards for a particular compliance year must be multiplied by the sum of the gasoline and diesel volume produced or imported by an obligated party in that year rather than only the gasoline volume as under the RFS1 program. [23] To the degree that an obligated party did not demonstrate full compliance with its RVOs for the previous year, the shortfall will be included as a deficit carryover in the calculation. CAA section 211(o)(5) only permits a deficit carryover from one year to the next if the obligated party achieves full compliance with each of its RVOs including the deficit carryover in the second year. Thus deficit carryovers cannot occur two years in succession for any of the four individual standards. They can, however, occur as frequently as every other year for a given obligated party for each standard.

Note that a party that produces only diesel fuel will have an obligation for all four standards even though he will not have the opportunity to blend ethanol into his own gasoline. Likewise, a party that produces only gasoline will have an obligation for all four standards even though he will not have an opportunity to blend biomass-based diesel into his own diesel fuel.

3. RINs Eligible To Meet Each RVO

Under RFS1, all RINs had the same compliance value and thus it did not matter what the RR or D code was for a given RIN when using that RIN to meet the total renewable fuel standard. In contrast, under RFS2 only RINs with specified D codes can be used to meet each of the four standards.

As described in Section I.A.1, the volume requirements in EISA are generally nested within one another, so that any fuel that satisfies the advanced biofuel requirement also satisfies the total renewable fuel requirement, and fuel that meets either the cellulosic biofuel or the biomass-based diesel requirements also satisfies the advanced biofuel requirement. As a result, the RINs that can be used to meet the four standards are likewise nested. Using the D codes defined in Table II.A-1, the RFS2 RINs that can be used to meet each of the four standards are shown in Table II.G.3-1. RFS1 RINs generated in 2010 and identified by a D code of 1 or 2 can also be applied to these standards using the protocol described in Section II.G.4 below.

Table II.G.3-1—RINs That Can Be Used To Meet Each Standard Back to Top
Standard Obligation Allowable D codes
Cellulosic biofuel RVO CB 3 and 7.
Biomass-based diesel RVO BBD 4 and 7.
Advanced biofuel RVO AB 3, 4, 5, and 7.
Renewable fuel RVO RF 3, 4, 5, 6, and 7.

The nested nature of the four standards also means that in some cases we must allow the same RIN to be used to meet more than one standard in the same year. Thus, for instance, a RIN with a D code of 3 can be used to meet three of the four standards, while a RIN with a D code of 5 can be used to meet both the advanced biofuel and total renewable fuel standards. However, a D code of 6 can only be used to meet the renewable fuel standard. Consistent with our proposal, we are continuing to prohibit the use of a single RIN for compliance purposes in more than one year or by more than one party. [24]

4. Treatment of RFS1 RINs Under RFS2

As described in the introduction to this section, we are implementing a number of changes to the RFS program as a result of the requirements in EISA. These changes will go into effect on July 1, 2010 and, among other things, will affect the conditions under which RINs are generated and their applicability to each of the four standards. As a result, RINs generated in 2010 under these RFS2 regulations will not be exactly the same as RINs generated under RFS1 regulations. Given the valid RIN life that allows a RIN to be used in the year generated or the year after, we must address circumstances in which excess 2009 RINs are used for compliance purposes in 2010. Also, since RINs generated in January through June of 2010 will be generated under RFS1 regulations, we must provide a means for them to be used to meet the annual 2010 RFS2 standards. Finally, we must address deficit carryovers from 2009 to 2010, since the total renewable fuel standards in these two years will be defined differently.

a. Use of RFS1 RINs To Meet Standards Under RFS2

In 2009 and the first three months of 2010, the RFS1 regulations will continue to apply and thus producers will not be required to demonstrate that their renewable fuel is made from renewable biomass as defined by EISA, nor that their combination of fuel type, feedstock, and process meets the GHG thresholds specified in EISA. Moreover, there is no practical way to determine after the fact if RINs generated under RFS1 regulations meet any of these criteria. However, we believe that the vast majority of RFS1 RINs generated in 2009 and the first two months of 2010 will in fact meet the RFS2 requirements. First, while ethanol made from corn must meet a 20% GHG threshold under RFS2 if produced by a facility that commenced construction after December 19, 2007, facilities that were already built or had commenced construction as of December 19, 2007 are exempt from this requirement. Essentially all ethanol produced in 2009 and the first three months of 2010 will meet the prerequisites for this exemption. Second, it is unlikely that renewable fuels produced in 2009 or the first three months of 2010 will have been made from feedstocks that do not meet the new renewable biomass definition. It is very unlikely that new land would have been cleared or cultivated since December 19, 2007 for use in growing crops for renewable fuel production, and thus the land use restrictions associated with the renewable biomass definition will very likely be met. Finally, the text of section 211(o)(5) states that a “credit generated under this paragraph shall be valid to show compliance for the 12 months as of the date of generation,” and EISA did not change this provision and did not specify any particular transition protocol to follow. A straightforward interpretation of this provision is to allow RFS1 RINs generated in 2009 and early 2010 to be valid to show compliance for the annual 2010 obligations.

The separate definitions for cellulosic biofuel and biomass-based diesel require GHG thresholds of 60% and 50%, respectively. While we do not have a mechanism in place to determine if these thresholds have been met for RFS1 RINs generated in 2009 or early 2010, any shortfall in GHG performance for this one transition period is unlikely to have a significant impact on long-term GHG benefits of the program. Few stakeholders commented on our proposed treatment of RFS1 RINs under RFS2. Of those that did, most supported our proposed approach to the use of RFS1 RINs to meet RFS2 obligations. Based on our belief that it is critical to the smooth operation of the program that excess 2009 RINs be allowed to be used for compliance purposes in 2010, we are allowing RFS1 RINs that were generated in 2009 or 2010 representing cellulosic biomass ethanol to be valid for use in satisfying the 2010 cellulosic biofuel standard. Likewise, we are allowing RFS1 RINs that were generated in 2009 or 2010 representing biodiesel and renewable diesel to be valid for use in satisfying the 2010 biomass-based diesel standard.

Consistent with our proposal, we have used information contained in the RR and D codes of RFS1 RINs to determine how those RINs should be treated under RFS2. The RR code is used to identify the Equivalence Value of each renewable fuel, and under RFS1 these Equivalence Values are unique to specific types of renewable fuel. For instance, biodiesel (mono alkyl ester) has an Equivalence Value of 1.5, and non-ester renewable diesel has an Equivalence Value of 1.7, and both of these fuels may be valid for meeting the biomass-based diesel standard under RFS2. Likewise, RINs generated for cellulosic biomass ethanol under RFS1 regulations must be identified with a D code of 1, and these fuels will be valid for meeting the cellulosic biofuel standard under RFS2. Our final treatment of RFS1 RINs for compliance under RFS2 is shown in Table II.G.4.a-1.

Table II.G.4. a-1—Treatment of RFS1 RINs for RFS2 Compliance Purposes Back to Top
RINs generated under RFS1a Treatment under RFS2b
a See RFS1 RIN code definitions at § 80.1125.
b See RFS2 RIN code definitions at § 80.1425.
Any RIN with D code of 2 and RR code of 15 or 17 Equivalent to RFS2 RINs with D code of 4.
All other RINs with D code of 2 Equivalent to RFS2 RINs with D code of 6.
Any RIN with D code of 1 Equivalent to RFS2 RINs with D code of 3.

b. Deficit Carryovers From the RFS1 Program to RFS2

The calculation of RVOs in 2010 under the RFS2 regulations will be somewhat different than the calculation of RVOs in 2009 under RFS1. In particular, 2009 RVOs were based on gasoline production only, while 2010 RVOs will be based on volumes of gasoline and diesel. As a result, 2010 compliance demonstrations that include a deficit carried over from 2009 will combine obligations calculated on two different bases.

We do not believe that deficits carried over from 2009 to 2010 will undermine the goals of the program in requiring specific volumes of renewable fuel to be used each year. Although RVOs in 2009 and 2010 will be calculated differently, obligated parties must acquire sufficient RINs in 2010 to cover any deficit carried over from 2009 in addition to that portion of their 2010 obligation which is based on their 2010 gasoline and diesel production. As a result, the 2009 nationwide volume requirement of 11.1 billion gallons of renewable fuel will be consumed over the two year period concluding at the end of 2010. Thus, we are not implementing any special treatment for deficits carried over from 2009 to 2010.

A deficit carried over from 2009 to 2010 will only affect a party's total renewable fuel obligation in 2010, as the 2009 obligation is for total renewable fuel use, not a subcategory. The RVOs for biomass-based diesel or advanced biofuel will not be affected, as they do not have parallel obligations in 2009 under RFS1. [25]

H. Separation of RINs

As we proposed in the NPRM, we are requiring the RFS1 provisions regarding the separation of RINs from volumes of renewable fuel to be retained for RFS2. However, the modifications in EISA required changes to the treatment of RINs associated with nonroad renewable fuel and renewable fuels used in heating oil and jet fuel. Our approach to the separation of RINs by exporters must also be modified to account for the fact that there would be four categories of renewable fuel under RFS2.

1. Nonroad

Under RFS1, RINs associated with renewable fuels used in nonroad vehicles and engines downstream of the renewable fuel producer were required to be retired by the party who owned the renewable fuel at the time of blending. This provision derived from the EPAct definition of renewable fuel which was limited to fuel used to replace fossil fuel used in a motor vehicle. However, EISA expands the definition of renewable fuel, and ties it to the definition of transportation fuel which is defined as any “fuel for use in motor vehicles, motor vehicle engines, nonroad vehicles, or nonroad engines (except for ocean-going vessels).” To implement these changes, the RFS2 program eliminates the RFS1 RIN retirement requirement for renewable fuels used in nonroad applications, with the exception of RINs associated with renewable fuels used in ocean-going vessels.

Since RINs have a valid life of two years, the NPRM proposed that a 2009 RFS1 RIN that is retired because the renewable fuel associated with it was used in nonroad vehicles or engines could be reinstated in 2010 for use in compliance with the 2010 standards. Stakeholders supported this approach, and we are finalizing it in today's action.

2. Heating Oil and Jet Fuel

EISA defines “additional renewable fuel” as “fuel that is produced from renewable biomass and that is used to replace or reduce the quantity of fossil fuel present in home heating oil or jet fuel.” [26] While we are not requiring fossil-based heating oil and jet fuel to be included in the fuel used by a refiner or importer to calculate their RVOs, we are allowing renewable fuels used as or in heating oil and jet fuel to generate RINs. Similarly, RINs associated with a renewable fuel, such as biodiesel, that is blended into heating oil will continue to be valid for compliance purposes. See also discussion in Section II.B.1.e.

3. Exporters

Under RFS1, exporters were assigned an RVO representing the volume of renewable fuel that was exported, and they were required to separate all RINs that were assigned to fuel that was exported. Since there was only one standard, there was only one possible RVO applicable to exporters.

Under RFS2, there are four possible RVOs corresponding to the four categories of renewable fuel (cellulosic biofuel, biomass-based diesel, advanced biofuel, and total renewable fuel). However, given the fungible nature of the RIN system and the fact that an assigned RIN transferred with a volume of renewable fuel may not be the same RIN that was originally generated to represent that volume, RINs from different fuel types can accompany volumes. Thus, there may be no way for an exporter to determine from an assigned RIN which of the four categories applies to an exported volume. In order to determine its RVOs, the only information available to the exporter may be the type of renewable fuel that he is exporting.

However, if an exporter knows, or has reason to know, that the renewable fuel that it is exporting is either cellulosic biofuel or advanced biofuel, we are requiring the exporter to determine an RVO for the exported fuel based upon these fuel types. For instance, if an exporter purchases cellulosic biofuel or advanced biofuel directly from a producer or if the fuel has been segregated from other fuels, we would expect the exporter to know or have reason to know the type of fuel that it is exporting. Another example of when we would expect an exporter to know or have reason to know that the fuel that it is exporting is cellulosic or advanced biofuel would be if the commercial documents that accompany the purchase or sale of the renewable fuel identify the product as cellulosic or advanced biofuel.

EPA recognizes that in many situations, exporters will not know or have reason to know which of the four categories of renewable fuel apply to the exported fuel. If this is the case, we are requiring exporters to follow the approach proposed in the NPRM. Exported volumes of biodiesel (mono alkyl esters) and renewable diesel must be used to determine the exporter's RVO for biomass-based diesel. For all other types of renewable fuel, the most likely category is general renewable fuel. Thus, we are requiring that all renewable fuels be used to determine the exporter's RVO for total renewable fuel. Our final approach is provided at § 80.1430.

In the NPRM we took comment on an alternative approach in which the total nationwide volumes required in each year (see Table I.A.1-1) would be used to apportion specific types of renewable fuel into each of the four categories. For example, exported ethanol may have originally been produced from cellulose to meet the cellulosic biofuel requirement, from corn to meet the total renewable fuel requirement, or may have been imported as advanced biofuel. If ethanol were exported, we could divide the exported volume into three RVOs for cellulosic biofuel, advanced biofuel, and total renewable fuel using the same proportions represented by the national volume requirements for that year. However, as described in the NPRM, we believe that this alternative approach would have added considerable complexity to the compliance determinations for exporters without necessarily adding more precision. Given the expected small volumes of exported renewable fuel, we continue to believe that this added complexity is not warranted at this time.

As described above, exporters must separate any RINs assigned to renewable fuel that they export. However, since RINs are fungible and the owner of a batch of renewable fuel has the flexibility to assign between zero and 2.5 gallon-RINs to each gallon, we have made this flexibility explicit for exporters. Thus, an exporter can separate up to 2.5 gallon-RINs for each gallon of renewable fuel that he exports. While the exporter is not required to retain these separated RINs for use in complying with his RVOs calculated on the basis of the exported volumes, this would be the most straightforward approach and would ensure that the exporter has sufficient RINs to comply. However, we are aware of some exporters who sell RINs that they separate as a source of revenue, with the intention to purchase replacement RINs on the open RIN market later in the year to comply with their RVOs. At this time we are not aware of such activities resulting in noncompliance, and thus the RFS2 regulations promulgated today will continue to allow this. However, we may revisit this issue in the future if there is evidence that exporters are failing to comply because they are selling RINs that they separate from exported volumes.

4. Requirement To Transfer RINs With Volume

In the NPRM, we proposed that the approach to RIN transfers established under RFS1—that RINs generated by renewable fuel producers and importers must be assigned to batches of renewable fuel and transferred along with those batches—be continued under RFS2. However, given the higher volumes required under RFS2 and the resulting expansion in the number of regulated parties, we also took comment on two alternative approaches to RIN transfers. Along with the alternative approaches for designation of obligated parties as described in Section II.G.1 above, a change to the requirement to transfer RINs with batches had the potential to more evenly align a party's access to RINs with that party's obligations under the RFS2 program. Nevertheless, for the reasons described below, we have determined that it would not be appropriate to implement these alternative approaches at this time.

In the first alternative approach, we would have removed the restriction established under the RFS1 rule requiring that RINs be assigned to batches of renewable fuel and transferred with those batches. Instead, renewable fuel producers could have sold RINs (with a K code of 2 rather than 1) separately from volumes of renewable fuel to any party.

In the second alternative approach, producers and importers of renewable fuels would be required to separate and transfer the RIN, but only to an obligated party. This “direct transfer” approach would require renewable fuel producers to transfer RINs with renewable fuel for all transactions with obligated parties, and sell all other RINs directly to obligated parties on a quarterly basis for any renewable fuel volumes that were not sold directly to obligated parties. Any RINs not sold in this way would be required to be offered for sale to any obligated party through a public auction. Only renewable fuel producers, importers, and obligated parties would be allowed to own RINs.

Many renewable fuel producers supported the concept of allowing them to separate the RINs from renewable fuel that they produce. They generally argued in favor of a free market approach to RINs in which there would be no restrictions on whom they could sell RINs to, or in what timeframe. The direct transfer approach was unnecessary, they argued, since the market would compel them to sell all RINs they generated, and all RINs would eventually end up in the hands of the obligated parties that need them. However, other renewable fuel producers opposed any change to the requirement that RINs be assigned to volumes of renewable and transferred with those volumes through the distribution system. They argued that the system established under RFS1 has proven to work and it would create an unwarranted burden to require producers to modify their IT systems for RFS2.

Marketers and distributors were generally opposed to our proposed alternative approaches to RIN transfers. Moreover, SIGMA and NACS, as in the RFS1 rulemaking process, recommended that RINs not be generated by producers at all, but rather by the party that blends renewable fuel into gasoline or diesel, or uses renewable fuel in its neat form as a transportation fuel.

Obligated parties generally opposed any change to the RFS1 requirement that RINs be assigned to volumes of renewable fuel by the producer or importer, and transferred with volumes through the distribution system. They reiterated their concern, first raised in the RFS1 rulemaking, that a free market approach would place them at greater risk of market manipulation by renewable fuel producers. Moreover, while generally expressing support for the concept of a direct transfer approach, they also expressed doubt that the auctions could be regulated in such a way as to ensure that RIN generators could not withhold RINs from the market by such means as failing to adequately advertise the time and location of an auction, by setting the selling price too high, by specifying a minimum number of bids before selling, by conducting auctions infrequently, by having unduly short bidding windows, etc. These concerns were exacerbated by the nested standards required by EISA, under which many obligated parties have expressed concern about being able to acquire sufficient RINs for compliance.

Given the significant challenges associated with a change to the requirement that RINs be transferred with volume and the opposing views among stakeholders, we are not making any change in today's final rule.

5. Neat Renewable Fuel and Renewable Fuel Blends Designated as Transportation Fuel, Heating Oil, or Jet Fuel

Under RFS1, RINs must, with limited exceptions, be separated by an obligated party taking ownership of the renewable fuel, or by a party that blends renewable fuel with gasoline or diesel. In addition, a party that designates neat renewable fuel as motor vehicle fuel may separate RINs associated with that fuel if the fuel is in fact used in that manner without further blending. One exception to these provisions is that biodiesel blends in which diesel constitutes less than 20 volume percent are ineligible for RIN separation by a blender. While EPA understands that in the vast majority of cases, biodiesel is blended with diesel in concentrations of 80 volume percent or less, there may be instances in which biodiesel is blended with diesel in concentrations of more than 80 percent biodiesel, but the blender is prohibited from separating RINs under the RFS1 regulations.

Thus, in order to account for situations in which biodiesel blends of 81 percent or greater may be used as transportation fuel, heating oil, or jet fuel without ever having been owned by an obligated party, EPA proposed, and is finalizing a change to the applicability of the RIN separation provisions for RFS2. Section 80.1429(b)(4) will allow for separation of RINs for neat renewable fuel or blends of renewable fuel and diesel fuel that the party designates as transportation fuel, heating oil, or jet fuel, provided the neat renewable fuel or blend is used in the designated form, without further blending, as transportation fuel, heating oil, or jet fuel. Those parties that blend renewable fuel with gasoline or diesel fuel (in a blend containing 80 percent or less biodiesel) must separate RINs pursuant to § 80.1429(b)(2).

Thus, for example, if a party intends to separate RINs from a volume of B85, the party must designate the blend for use as transportation fuel, heating oil, or jet fuel and the blend must be used in its designated form without further blending. The party is also required to maintain records of this designation pursuant to § 80.1454(b)(5). Finally, the party is required to comply with the proposed PTD requirements in § 80.1453(a)(11)(iv), which serve to notify downstream parties that the volume of fuel has been designated for use as transportation fuel, heating oil, or jet fuel, and must be used in that designated form without further blending. Parties may separate RINs at the time they comply with the designation and PTD requirements, and do not need to physically track ultimate fuel use.

I. Treatment of Cellulosic Biofuel

1. Cellulosic Biofuel Standard

EISA requires that the Administrator set the cellulosic biofuel standard each November for the next year based on the lesser of the volume specified in the Act or the projected volume of cellulosic biofuel production based on EIA estimates for that year. In the event that the projected volume is less than the amount required in the Act, EPA may also reduce the applicable volume of the total renewable fuel and advanced biofuels requirement by the same or a lesser volume. We will examine EIA's projected volumes and other available data including the required production outlook reports discussed in Section II.K to decide the appropriate standard for the following year. The outlook reports from all renewable fuel producers will assist EPA in determining what the cellulosic biofuel standard should be and if the total renewable fuel and/or advanced biofuel standards should be adjusted. For years where EPA determines that the projected volume of cellulosic biofuels is not sufficient to meet the levels in EISA we will consider the availability of other advanced biofuels in deciding whether to lower the advanced biofuel standard as well.

In determining whether the advanced biofuel and/or total renewable fuel volume requirements should also be adjusted downward in the event that projected volumes of cellulosic biofuel fall short of the statutorily required volumes, we believe it may be appropriate to allow excess advanced biofuels to make up some or all of the shortfall in cellulosic biofuel. For instance, if we determined that sufficient biomass-based diesel was available, we could decide that the required volume of advanced biofuel need not be lowered, or that it should be lowered to a smaller degree than the required cellulosic biofuel volume. Thus, the Act requires EPA to examine the total and advanced renewable fuel standards and volumes in the event of a cellulosic volume waiver. EPA will look at projections for each year on an individual yearly basis to determine if the standards should be adjusted. EPA believes that since the standards are nested and the total and advanced renewable fuel volume mandates are met in part by the cellulosic volume mandate, Congress gave EPA the flexibility to lower the required total and advanced volumes, but Congress also wanted to encourage the development of advanced renewable fuels as well and allow in appropriate circumstances for the use of those fuels in the event they can meet that year's required volumes that would have been met by the cellulosic mandate.

2. EPA Cellulosic Biofuel Waiver Credits for Cellulosic Biofuel

Whenever EPA sets the cellulosic biofuel standard at a level lower than that required in EISA, but greater than zero, EPA is required to provide a number of cellulosic credits for sale that is no more than the volume used to set the standard. Congress also specified the price for such credits: Adjusted for inflation, they must be offered at the price of the higher of 25 cents per gallon or the amount by which $3.00 per gallon exceeds the average wholesale price of a gallon of gasoline in the United States. The inflation adjustment will be for years after 2008. The inflation adjustment will be based on the standard US inflation measure Consumer Price Index for All Urban Consumers (CPI-U) for All Items expenditure category as provided by the Bureau of Labor Statistics. [27]

Congress afforded the Agency considerable flexibility in implementing the system of cellulosic biofuel credits. EISA states EPA; “shall include such provisions, including limiting the credits' uses and useful life, as the Administrator deems appropriate to assist market liquidity and transparency, to provide appropriate certainty for regulated entities and renewable fuel producers, and to limit any potential misuse of cellulosic biofuel credits to reduce the use of other renewable fuels, and for such other purposes as the Administrator determines will help achieve the goals of this subsection.”

We have fashioned a number of limitations on the use of cellulosic that reflect these considerations. Specifically, the credits will be called “Cellulosic Biofuel Waiver Credits” (or “waiver credits”) so that there is no confusion with RINs or allowances used in the acid rain program. Such waiver credits will only be available for the current compliance year for which we have waived some portion of the cellulosic biofuel standard, they will only be available to obligated parties, and they will be nontransferable and nonrefundable. Further, obligated parties may only purchase waiver credits up to the level of their cellulosic biofuel RVO less the number of cellulosic biofuel RINs that they own. A company owning cellulosic biofuel RINs and cellulosic waiver credits may use both types of credits if desired to meet their RVOs, but unlike RINs obligated parties will not be able to carry waiver credits over to the next calendar year. Obligated parties may not use waiver credits to meet a prior year deficit obligation. These restrictions help ensure that waiver credits are not overutilized at the expense of actual renewable volume.

In the NPRM, EPA proposed that the credits could be usable for the advanced and total renewable standards similarly to cellulosic biofuel RINs. Several commenters stated this provision could displace advanced and total renewable fuel that was actually produced which would be against the intent of the Act, and that unlike RINs a company should only be permitted to use waiver credits to meet its cellulosic biofuel obligation. We agree, and are limiting the use of waiver credits for compliance with only a company's cellulosic biofuel RVO.

In the event the total volume of conventional gasoline and diesel fuel produced or imported in the country exceeds the projections used to set the standard, companies will still be able to purchase waiver credits up to their cellulosic volume obligation. When setting a reduced cellulosic biofuel standard EPA makes a determination that the cellulosic volume specified in EISA will not be met and that determination is not based on how much nonrenewable motor fuel will be produced. EPA sets the standard based on the volumes in the Act and a projection of gasoline production to ensure the obligation is broken up most equitably. EPA believes that Congress wanted all obligated parties to have equal access to the waiver credits in the event of the waiver and did not want obligated parties to incur a deficit due to the timing of when they purchased waiver credits.

Cellulosic Biofuel Waiver Credits, in the event of a waiver, will be offered in a generic format rather than a serialized format, like RINs. Waiver credits can be purchased using procedures defined by the EPA, and at the time that an obligated party submits its annual compliance demonstration to the EPA and establishes that it owns insufficient cellulosic biofuel RINs to meet its cellulosic biofuel RVO. EPA will define these procedures with the U.S. Treasury before the end of the first annual compliance period. EPA will publish these procedures with the obligated party annual compliance report template. EPA will provide the forms necessary to purchase the credits. EPA intends to provide options for obligated parties to use Pay.Gov or if desired to mail payment to the U.S. Treasury.

The wholesale price of gasoline used by EPA in setting the price of the waiver credits will be based on the average monthly bulk (refinery gate) price of gasoline using data from the most recent twelve months of data from EIA available to EPA at the time it develops the cellulosic biofuel standard. [28] EPA will use refinery gate price, U.S. Total Gasoline Bulk Sales (Price) by Refiners from EIA in calculating the average, since it is the price most reflective of what most obligated parties are selling their fuel. EPA will use the most recent twelve months of data provided by EIA to develop an average price on actual volumes produced in the year prior to the compliance year. In order to provide regulatory certainty, we will set the waiver credits price for the following year each November when and if we set a cellulosic biofuel standard for the following year that is based on achieving a lower volume of cellulosic biofuel use than is specified in EISA.

For the 2010 compliance period, since the cellulosic standard is lower than the level otherwise required by EISA, we are also making cellulosic waiver credits available to obligated parties for end-of-year compliance should they need them at a price of $1.56 per gallon-RIN.” The price for the 2011 compliance period, if necessary will be set when we announce the 2011 cellulosic biofuel standard.

3. Application of Cellulosic Biofuel Waiver Credits

While the credit provisions of section 202(e) of EISA ensure that there is a predictable upper limit to the price that cellulosic biofuel producers can charge for a gallon of cellulosic biofuel and its assigned RIN, there may be circumstances in which this provision has other unintended consequences. This could occur in situations where the cost of total renewable fuel RINs exceeds the cost of the cellulosic waiver credits. To prevent this, we sought comment on and are finalizing an additional restriction: An obligated party may only purchase waiver credits from the EPA to the degree that it establishes it owns insufficient cellulosic biofuel RINs to meet its cellulosic biofuel RVO. This approach forces obligated parties to apply all their cellulosic biofuel RINs to their cellulosic biofuel RVO before applying any waiver credits to their cellulosic biofuel RVO.

Even with this restriction the approach in the NPRM might not have operated as intended. For instance, if the combination of cellulosic biofuel volume price and RIN price were to become low compared to that for general renewable fuel, a small number of obligated parties could have purchased more cellulosic biofuel than they need to meet their cellulosic biofuel RVOs and could have used the additional cellulosic biofuel RINs to meet their advanced biofuel and total renewable fuel RVOs. Other obligated parties would then have had no access to cellulosic biofuel volume nor cellulosic biofuel RINs, and would have been forced to purchase waiver credits from the EPA. This situation would have had the net effect of waiver credits replacing advanced biofuels and/or general renewable fuel rather than cellulosic biofuel. Based on comments received on the NPRM, EPA is placing the additional restriction of only allowing the waiver credits to count towards the cellulosic biofuel standard and not the advanced or renewable fuel standards.

Moreover, under certain conditions it may be possible for the market price of general renewable fuel RINs to be significantly higher than the market price of cellulosic biofuel RINs, as the latter is limited in the market by the price of EPA-generated waiver credits according to the statutory formula described in Section II.I.2 above. Under some conditions, this could result in a competitive disadvantage for cellulosic biofuel in comparison to corn ethanol, for example. For instance, if gasoline prices at the pump are significantly higher than ethanol production costs, while at the same time corn-ethanol production costs are lower than cellulosic ethanol production costs, profit margins for corn-ethanol producers will be larger than for cellulosic ethanol producers. Under these conditions, while obligated parties may still purchase cellulosic ethanol volume and its associated RINs rather than waiver credits, cellulosic ethanol producers will realize lower profits than corn-ethanol producers due to the upper limit placed on the price of cellulosic biofuel RINs through the pricing formula for waiver credits. For a newly forming and growing cellulosic biofuel industry, this competitive disadvantage could make it more difficult for investors to secure funding for new projects, threatening the ability of the industry to reach the statutorily mandated volumes.

Finally, in the NPRM we sought comment on a “dual RIN” approach to cellulosic biofuel. In this approach, both cellulosic biofuel RINs (with a D code of 3) and waiver credits would have only been applied to an obligated party's cellulosic biofuel RVO, but producers of cellulosic biofuel would also generate an additional RIN representing advanced biofuel (with a D code of 5). The producer would have only been required to transfer the advanced biofuel RIN with a batch of cellulosic biofuel, and could retain the cellulosic biofuel RIN for separate sale to any party. [29] The cellulosic biofuel and its attached advanced biofuel RIN would then have competed directly with other advanced biofuel and its attached advanced biofuel RIN, while the separate cellulosic biofuel RIN would have an independent market value that would have been effectively limited by the pricing formula for waiver credits as described in Section II.I.2. However, this approach would have been a more significant deviation from the RIN generation and transfer program structure that was developed cooperatively with stakeholders during RFS1. It would have provided cellulosic biofuel producers with significantly more control over the sale and price of cellulosic biofuel RINs, which was one of the primary concerns of obligated parties during the development of RFS1. Therefore, EPA is treating the transfer of cellulosic RINs in the same manner as the other required volumes.

J. Changes to Recordkeeping and Reporting Requirements

1. Recordkeeping

Recordkeeping, including product transfer documents (PTDs), will support the enforcement of the use of RINs for compliance purposes. Parties are afforded significant freedom with regard to the form that PTDs take. Product codes may be used as long as they are understood by all parties, but they may not be used for transfers to truck carriers or to retailers or wholesale purchaser-consumers. Parties must keep copies of all PTDs they generate and receive, as well as copies of all reports submitted to EPA and all records related to the sale, purchase, brokering or transfer or RINs, for five (5) years. Parties must keep copies of records that relate to program flexibilities, such as small business-oriented provisions. Upon request, parties are responsible for providing their records to the Administrator or the Administrator's authorized representative. We reserve the right to request to receive documents in a format that we can read and use.

In Section III.A. of this preamble, we describe an EPA-Moderated Transaction System (EMTS) for RINs. The new system allows for “real-time” recording of transactions involving RINs.

2. Reporting

Producers and importers who generate or take ownership of RINs shall submit RIN Transaction Reports [30] and/or RIN Generation Reports quarterly. Renewable fuel exporters and obligated parties shall submit their RIN Transaction Reports quarterly, and RIN owners shall submit their RIN Transaction Reports quarterly. EMTS will be used by all parties to record “real time” generation of RINs and transactions involving RINs starting July 1, 2010. “Real time” means recordation within five (5) business days of generation or any transaction involving a RIN.

Quarterly reports are to be submitted on the following schedule. Quarterly reports include RIN Activity Reports and, with EMTS, simplified reporting and certification of the RIN Generation and RIN Transaction Reports.

Table II.J-1—Quarterly Reporting Schedule Back to Top
Quarter covered by report Due date for report
January-March May 31.
April-June August 31.
July-September November 30.
October-December February 28.

Annual reports (covering January through December) would continue to be due on February 28. The only annual report is the Obligated Party Annual Compliance Report. [31]

Simplified, secure reporting is currently available through our Central Data Exchange (CDX). CDX permits us to accept reports that are electronically signed and certified by the submitter in a secure and robustly encrypted fashion. Using CDX eliminates the need for wet ink signatures and reduces the reporting burden on regulated parties. EMTS will also make use of the CDX environment.

Due to the criteria that renewable fuel producers and importers must meet in order to generate RINs under RFS2, and due to the fact that renewable fuel producers and importers must have documentation about whether their feedstock(s) meets the definition of “renewable biomass,” we proposed several changes to the RIN Generation Report. [32] We proposed to make the report a more general report on renewable fuel production in order to capture information on all batches of renewable fuel, whether or not RINs are generated for them. This final rule adopts the proposed approach. All renewable fuel producers and importers above 10,000 gallons per year must report to EPA on each batch of their fuel and indicate whether or not RINs are generated for the batch. If RINs are generated, the producer or importer is required to certify that his feedstock meets the definition of “renewable biomass.” If RINs are not generated, the producer or importer must state the reason for not generating RINs, such as they have documentation that states that their feedstock did not meet the definition of “renewable biomass,” or the fuel pathway used to produce the fuel was such that the fuel did not qualify to generate RINs as a renewable fuel. For each batch of renewable fuel produced, we require information about the types and volumes of feedstock used and the types and volumes of co-products produced, as well as information about the process or processes used. This information is necessary to confirm that the producer or importer assigned the appropriate D code to their fuel and that the D code was consistent with their registration information. In this final rule, we adopt the approach set forth in the notice of proposed rulemaking.

In addition, we proposed two changes for the RIN Transaction Report. [33] First, for reports of RINs assigned to a volume of renewable fuel, the volume of renewable fuel must be reported. Second, RIN price information must be submitted for transactions involving both separated RINs and RINs assigned to a renewable volume. This information was not collected under RFS1, but because we believe this information has great programmatic value to EPA, we proposed to collect it for RFS2. As we explained in the proposed rule, price information may help us to anticipate and appropriately react to market disruptions and other compliance challenges, will be beneficial when setting future renewable standards, and will provide additional insight into the market when assessing potential waivers. Our incomplete knowledge regarding RIN pricing for RFS1 adversely affected our ability to assess the general health and direction of the market and overall liquidity of RINs. Because we believe the inclusion of price information in reports will be beneficial to both EPA and to regulated parties, this final rule includes that information element in reports, as well as incorporating it as part of the “real time” transactional information collected via EMTS.

3. Additional Requirements for Producers of Renewable Natural Gas, Electricity, and Propane

In addition to the general reporting requirement listed above, we are requiring an additional item of reporting for producers of renewable natural gas, electricity, and propane who choose to generate and assign RINs. While producers of renewable natural gas, electricity, and propane who generate and assign RINs are responsible for filing the same reports as other producers of RIN-generating renewable fuels, we are requiring that additional reporting for these producers support the actual use of their products in the transportation sector. We believe that one simple way to achieve this may be to add a requirement that producers of renewable natural gas, electricity, and propane add the name of the purchaser (e.g., the name of the wholesale purchaser-consumer (WPC) or fleet) to their RIN generation reports and then maintain appropriate records that further identify the purchaser and the details of the transaction. We are not requiring that a purchaser who is either a WPC or an end user would have to register under this scenario, unless that party engages in other activities requiring registration under this program.

4. Attest Engagements

The purpose of an attest engagement is to receive third party verification of information reported to EPA. An attest engagement, which is similar to a financial audit, is conducted by a Certified Public Accountant (CPA) or Certified Independent Auditor (CIA) following agreed-upon procedures. We have found the information in attest engagements submitted under RFS1 to be extremely valuable as a compliance monitoring tool. The approach adopted in this final rule is identical to the approach adopted under the RFS1 program, [34] although the universe of obligated parties and renewable fuels producers is broader under this final rule for RFS2.

As with the RFS1 program, an attest engagement must be conducted by an individual who is a Certified Public Accountant (CPA) or Certified Internal Auditor (CIA), who is independent of the party whose records are being reviewed, and who will follow agreed-upon procedures to determine whether underlying records, reported items, and transactions agree. The CPA or CIA will generate a report as to their findings.

We have received numerous questions and comments related to how attest engagements apply to foreign companies and whether or not a foreign accountant may perform the required agreed-upon procedures. EPA will accept an attest engagement performed by a foreign accountant who holds an equivalent credential to an American CPA or CIA. A written explanation as to the foreign accountant's qualifications and the equivalency of the credential must accompany the attest engagement.

Producers of renewable fuels, obligated parties, exporters, and any party who owns RINs must arrange for an annual attest engagement. The attest engagement report for any given year must be submitted to EPA by no later than May 31 of the following year. Section 80.1464 of the regulations specifies the attest engagement procedures to be followed.

K. Production Outlook Reports

Under this program we are requiring the submission, starting in 2010, of annual production outlook reports from all domestic renewable fuel producers, foreign renewable fuel producers who register to generate RINs, and importers of renewable fuels. These production outlook reports will be similar in nature to the pre-compliance reports required under the Highway and Nonroad Diesel programs. These reports will contain information about existing and planned production capacity, long-range plans, and feedstocks and production processes to be used at each production facility. For expanded production capacity that is planned or underway at each existing facility, or new production facilities that are planned or underway, the progress reports will require information on: (1) Strategic planning; (2) Planning and front-end engineering; (3) Detailed engineering and permitting; (4) Procurement and construction; (5) Commissioning and startup; (6) Projected volumes; (7) Contracts currently in place (feedstocks, sales, delivery, etc.); and (8) Whether or not feedstocks have been purchased. The first five project phases are described in EPA's June 2002 Highway Diesel Progress Review report (EPA document number EPA420-R-02-016, located at: www.epa.gov/otaq/regs/hd2007/420r02016.pdf). In the proposed rule, we asked for comment on the first five project phases, and whether or not they were appropriate for renewable fuels production. We also proposed additional phases in order to provide better specificity for ascertaining industry status. EPA plans to use this information in order to provide annual summary reports regarding such planned capacity.

The full list of requirements for the production outlook reports is provided in the regulations at § 80.1449. The information submitted in the reports will be used to evaluate the progress that the industry is making towards the renewable fuels volume goals mandated by EISA. They will help EPA set the annual cellulosic biofuel standard and consider whether waivers would be appropriate with respect to the advanced biofuel, biomass-based diesel, and total renewable fuel standards (see Section II.I of this preamble for more discussion on this). Production outlook reports will be due annually by March 31 (except that for the year 2010, the report will be due September 1) and each annual report must provide projected information, including any updated information from the previous year's report.

As mentioned in the preamble to the proposed rule, EPA currently receives data on projected flexible-fuel vehicle (FFV) sales and conversions from vehicle manufacturers. These are helpful in providing EPA with information regarding the potential market for renewable fuels. We requested comment on whether we should require the annual submission of data to facilitate our evaluation of the ability of the distribution system to deliver the projected volumes of biofuels to petroleum terminals that are needed to meet the RFS2 standards, the extent to which such information is already publicly available or can be purchased from a proprietary source, and the extent to which such publicly available or purchasable data would be sufficient for EPA to make its determination. We further requested comment on the parties that should be required to report to EPA, and data requirements. We believe that publicly available information on E15, E85, and other refueling facilities is sufficient for us to make a determination about the adequacy of such facilities to support the projected volumes that would be used to satisfy the RFS2 standards. Therefore, we are not finalizing such a requirement.

While we understand that the types of projections we request in the Outlook Reports could be somewhat speculative in nature, we believe that the projections will provide us with the most reliable information possible to inform the annual RFS standards and waiver considerations. Further, we believe this information will be more useful to us than other public information that is released in other contexts (e.g., announcements for marketing purposes). As mentioned above in Section II.I, we believe that we can use this information to supplement other available information (such as volume projections from EIA) to help set the standard for the following year. Specifically, it will provide more accurate information for setting the cellulosic biofuel and biomass-based diesel standards, and any adjustments to the advanced biofuel and total renewable fuel standards.

We received comments that both support and oppose the Production Outlook Reports, or some element of them. One commenter stated that EPA provided no reasonable explanation to require the information being requested for the reports; the commenter further stated that such information is not needed to assist parties to come into compliance. Another commenter stated that the renewable fuels industry cannot confidently project what will happen in 2010, or even 2020, because there are too many unknowns, no previous history of renewable fuels mandates, and no sense of continued tax rebate. The commenter suggested that until the industry operates for a few years under the RFS2 carve-outs and the issues on the tax rebates for renewables are resolved, the industry cannot develop a meaningful outlook forecast. The commenter further suggested that EPA instead hire a consultant who can look at the big picture and provide a more meaningful evaluation than could the individual members of the biofuels industry. However, as discussed above, while these reports will have their limitations, we believe they will provide the best and most up to date information available for us to use in setting the standards and considering any waiver requests. We will of course also look to other publicly available information, and may consider using contractors to help out in this regard, but it cannot replace the need for the production outlook report data.

A commenter noted that this provision is similar to reports required under the diesel program. The commenter further stated that if the required information can be captured by EMTS, the commenter fully supports this requirement. However, the commenter stated that it is opposed to some of the required elements of the reports for planned expanded or new production (strategic planning, planning and front-end engineering, detailed engineering and permitting, procurement and construction, and commissioning and start-up); these are an aspect of financial planning that the commenter believes EPA has no jurisdiction over and cannot derive basis from EISA in any form regardless of interpretation. As explained above, this information will be used by EPA to inform us for setting the standards on an annual basis and in responding to any waiver petitions. It will not be used to assess compliance with the program. The other provisions for registration, recordkeeping and reporting serve that purpose.

Another commenter stated that the reports should be required, but that EPA should not rely too heavily upon the data (particularly for new biofuel technologies). Some commenters noted that they believe that requiring Production Outlook Reports is duplicative in nature and/or a burden to the industry. These commenters also believe that EPA already receives such information through the reporting that currently exists, and that EPA could also obtain this information from DOE's Energy Information Administration (EIA) and the National Biodiesel Board (NBB). Other commenters expressed concern over reporting such confidential and strategic information (even as confidential business information (CBI)), and that information out to 2022 seems excessive and useless; and that the reports should be limited to just domestic and foreign producers of renewable fuels but not importers (as they tend to import renewable fuels based on variable economic conditions and will not likely have the ability to reliably predict their future import volumes). The information that currently exists from other sources is current and historical information. For the purposes of setting future standards, we need to have information on future plans and projections. We understand that reality will always be different from the projections, but they will still give us the best possible source of information. Furthermore, by having projections five years out into the future, and then obtaining new reports every year, we will be able to assess the trends in the data and reports to better utilize them over time.

Some commenters have expressed concern that the information required for Production Outlook Reports is not needed, won't provide useful information because it is speculative, or asks for information that could be sensitive/confidential. However, we continue to believe that such information is essential to our annual cellulosic biofuel standard setting, and consideration of whether waivers should be provided for other standards. All information submitted to EPA will be treated as confidential business information (CBI), and if used by EPA in a regulatory context will only be reported out in very general terms. As with our Diesel Pre-compliance Reports, we fully expect that the information will be somewhat speculative in the early reports, and we will weight it accordingly. As the program progresses, however, information submitted for the reports will continue to improve. We believe that any information, whether speculative or concrete, will be helpful for the purposes described above. Thus we are finalizing Production Outlook Reports, and the required elements at § 80.1449.

L. What Acts Are Prohibited and Who Is Liable for Violations?

The prohibition and liability provisions under this rule are similar to those of the RFS1 program and other fuels programs in 40 CFR part 80. The rule identifies certain prohibited acts, such as a failure to acquire sufficient RINs to meet a party's RVOs, producing or importing a renewable fuel that is not assigned a proper RIN category (or D Code), improperly assigning RINs to renewable fuel that was not produced with renewable biomass, failing to assign RINs to qualifying fuel, or creating or transferring invalid RINs. Any person subject to a prohibition is liable for violating that prohibition. Thus, for example, an obligated party is liable if the party failed to acquire sufficient RINs to meet its RVO. A party who produces or imports renewable fuels is liable for a failure to assign proper RINs to qualifying batches of renewable fuel produced or imported. Any party, including an obligated party, is liable for transferring a RIN that was not properly identified.

In addition, any person who is subject to an affirmative requirement under this program is liable for a failure to comply with the requirement. For example, an obligated party is liable for a failure to comply with the annual compliance reporting requirements. A renewable fuel producer or importer is liable for a failure to comply with the applicable batch reporting requirements. Any party subject to recordkeeping or product transfer document (PTD) requirements is liable for a failure to comply with these requirements. Like other EPA fuels programs, this rule provides that a party who causes another party to violate a prohibition or fail to comply with a requirement may also be found liable for the violation.

EPAct amended the penalty and injunction provisions in section 211(d) of the Clean Air Act to apply to violations of the renewable fuels requirements in section 211(o). Accordingly, any person who violates any prohibition or requirement of this rule is subject to civil penalties of up to $37,500 per day and per each individual violation, plus the amount of any economic benefit or savings resulting from each violation. Under this rule, a failure to acquire sufficient RINs to meet a party's renewable fuels obligation constitutes a separate day of violation for each day the violation occurred during the annual averaging period.

As discussed above, the regulations prohibit any party from creating or transferring invalid RINs. These invalid RIN provisions apply regardless of the good faith belief of a party that the RINs are valid. These enforcement provisions are necessary to ensure the RFS2 program goals are not compromised by illegal conduct in the creation and transfer of RINs.

As in other motor vehicle fuel credit programs, the regulations address the consequences if an obligated party is found to have used invalid RINs to demonstrate compliance with its RVO. In this situation, the obligated party that used the invalid RINs will be required to deduct any invalid RINs from its compliance calculations. An obligated party is liable for violating the standard if the remaining number of valid RINs was insufficient to meet its RVO, and the obligated party might be subject to monetary penalties if it used invalid RINs in its compliance demonstration. In determining what penalty is appropriate, if any, we would consider a number of factors, including whether the obligated party did in fact procure sufficient valid RINs to cover the deficit created by the invalid RINs, and whether the purchaser was indeed a good faith purchaser based on an investigation of the RIN transfer. A penalty might include both the economic benefit of using invalid RINs and/or a gravity component.

Although an obligated party is liable under our proposed program for a violation if it used invalid RINs for compliance purposes, we would normally look first to the generator or seller of the invalid RINs both for payment of penalty and to procure sufficient valid RINs to offset the invalid RINs. However, if, for example, that party was out of business, then attention would turn to the obligated party who would have to obtain sufficient valid RINs to offset the invalid RINs.

III. Other Program Changes Back to Top

In addition to the regulatory changes we are finalizing today in response to comments received on the proposed rule and EISA (which are designed to implement the provisions of RFS2), there are a number of other changes to the RFS program that we are making. We believe that these changes will increase flexibility, simplify compliance, or address RIN transfer issues that have arisen since the start of the RFS1 program. Throughout the rulemaking process, we also investigated impacts on small businesses and we are finalizing provisions to address the impacts of the program on them.

A. The EPA Moderated Transaction System (EMTS)

The EPA Moderated Transaction System (EMTS) emerged as a result of our experiences with and lessons learned from implementing RFS1. Recognizing that the addition of significant volumes of renewable fuels and expansion of renewable fuel categories were adding complexity to an already stressed system, EMTS was introduced as a new approach for managing RINs in our NPRM. We received broad acceptance of the EMTS concept in the public comments as well as support for its expeditious implementation. This section describes the need for EMTS, implementation of EMTS, and an explanation of how EMTS will work. By implementing EMTS, we believe that we will be able to greatly reduce RIN-related errors while efficiently and accurately managing the universe of RINs. EMTS will save considerable time and resources for both industry and EPA. This is most evident considering that the system virtually eliminates multiple sources of administrative errors, resulting in a reduction of costs and effort expended to correct and regenerate product transfer documents, documentation and recordkeeping, and resubmitting reports to EPA. Use of EMTS will result in fewer report resubmissions and easier reporting for industry, while leaving fewer reports to be processed by EPA. Industry will spend less time and effort validating the RINs they procure with greater assurance and confidence in the RIN market. EPA will spend less time tracking down invalid RINs and working with regulated parties on complex remedial actions. This is possible because EMTS removes management of the 38-digit RIN from the hands of the reporting community. At the same time, EPA and the reporting community will be working with a standardized system, reducing stresses and development costs on IT systems.

We received comments suggesting that EPA remove the attest engagement requirements and certain recordkeeping requirements due to the use of EMTS. While we believe that EMTS will simplify and reduce burdens on the regulated community, it is important to point out that EMTS is strictly a RIN tracking and managing tool designed to facilitate reporting under the Renewable Fuel Standard program. Product transfer documents are the commercial documents used to memorialize transactions of RINs between a buyer and a seller in the market. The EMTS will rely on references to these documents, which can take many forms, but it is not capable of replacing those documents. Attest engagements are used to verify that the records required to be kept by regulated parties, including information retained by a regulated party as well as information reported to EPA such as laboratory test results, contracts between renewable fuel/RIN buyers and sellers, feedstock documentation, etc. is correctly maintained or reported. The information reported via EMTS is but a subset of the information required to be maintained in a regulated party's records, and both PTDs and attest engagements are necessary to ensure that the information collected and tracked in EMTS concurs with actual events.

1. Need for the EPA Moderated Transaction System

In implementing RFS1, we found that the 38-digit standardized RINs proved to be confusing to many parties in the distribution chain. Parties made various errors in generating and using RINs. For example, parties transposed digits within the RIN and incorrectly referenced volume numbering. Also, parties created alphanumeric RINs, despite the fact that RINs were supposed to consist of all numbers.

Once an error is made within a RIN, the error propagates throughout the distribution system. Correcting an error can require significant time and resources and usually involves many steps. Not only must reports to EPA be corrected, underlying records and reports reflecting RIN transactions must also be located and corrected to reflect discovery of an error. Because reporting related to RIN transactions under RFS1 was only on a quarterly basis, a RIN error could exist for several months before being discovered.

Incorrect RINs are invalid RINs. If parties in the distribution system cannot track down and correct errors in a timely manner, then all downstream parties that traded the invalid RIN are in violation. Because RINs are the basic unit of compliance for the RFS program, it is important that parties have confidence when generating and using them.

All parties in the RFS1 and the RFS2 regulated community are required to use RINs. Under RFS2, we foresee that regulated party community will substantially expand. Newer regulated parties of an already complex system necessitate EMTS. These parties include renewable fuel producers and importers, obligated parties, exporters, and other RIN owners; (typically marketers of renewable fuels and blenders). Under RFS1, all RINs were used to comply with a single standard. With RFS2, there are four standards. RINs must be generated to identify one of the fuel categories: cellulosic biofuel, cellulosic diesel, biomass-based diesel, advanced biofuel, and renewable fuels (e.g., corn ethanol). (For a more detailed discussion of RINs, see Section II.A of this preamble.) The different types of RINs will be managed in the EMTS.

2. Implementation of the EPA Moderated Transaction System

We proposed that EMTS would be an opt-in for the calendar year 2010 and mandatory for calendar year 2011. We received many comments strongly supporting EMTS implementation with the start of the RFS2 program to ensure confidence and simplicity in an increasingly complex program. We also received comments that EMTS implementation with RFS2 is necessary so industry would not have to create a new system to handle RFS2 RINs for 2010 and then move to EMTS for 2011 while still handling RFS1 RINs. Potentially, three RIN transaction systems would exist during transition from RFS1 to RFS2 if EMTS could not be implemented with the start of the RFS2 program. EPA agrees that this three system issue would be an undue burden to industry as it would require industry to create two systems within a 12 month period. EMTS development started with the introduction of the NPRM, and has been in beta testing since early November with a select group of different industry stakeholders. Industry feedback has been overwhelmingly strong for the implementation of EMTS with the start of RFS2. With this final rule, EPA decided that EMTS will start on the same date when RFS2 RINs are required to be generated. In addition, to ensure that parties will have enough time to incorporate RFS2 and EMTS requirements into private RIN tracking systems, the generation of RFS2 RINs will begin on July 1, 2010. Therefore, all RFS regulated parties are required to use EMTS starting July 1, 2010.

RIN transactions are required to be verified and certified on a quarterly basis. EMTS will provide summaries for parties to verify, report, and certify transactions to EPA through the fuels reporting system, DCFuels. Additional information may be required to be added to the EMTS provided summary. This additional certification step allows parties to verification that the information sent to EMTS is accurate. However, parties may choose to review their data by checking their EMTS account at anytime.

With EMTS, RIN transactions are required to be verified and certified on a quarterly basis. EMTS will provide summaries for parties to verify, report, and certify transactions to EPA through the fuels reporting system, DCFuels. Additional information may be required to be added to the EMTS provided report. This additional certification step allows parties to verify that the information sent to EMTS is accurate. However, parties may choose to review their data by checking their EMTS account at any time.

3. How EMTS Will Work

EMTS will be a closed, EPA-moderated system that provides a mechanism for screening RINs and a structured environment for conducting RIN transactions. “Screening” of RINs means that parties can have greater confidence that the RINs they handle are genuine. Although screening cannot remove all human error, we believe it can remove most of it.

We received comments opposing the 3 day time window for reporting transactions to the EMTS. One commenter requested 7 days from the event for sellers to report a transaction and 7 days after that for the buyer to accept the transaction. In order for this to be a “real time” system, we must require that the information comes in a timely manner. One commenter requested 10 days from the event to send information to EMTS. EPA has concluded that five days, or a business week, is an appropriate amount of time for both parties to receive or provide necessary documentation in order to interact with EMTS accurately and timely. “Real time” will be defined as within five (5) business days of a reportable event (e.g., generation and assignment of RINs, transfer of RINs).

Parties who use EMTS must first register with EPA in accordance with the RFS2 registration program described in Section II.C of this preamble. Parties will also have to create an account (i.e., register) via EPA's Central Data Exchange (CDX), as users will access EMTS via CDX. CDX is a secure and central electronic portal through which parties may submit compliance reports. Parties must establish an account with EMTS by July 1, 2010 or 60 days prior to engaging in any transaction involving RINs, whichever is later. Once registration occurs, individual accounts will be established within EMTS and the system will enable a party to submit transactions based on their registration information.

In EMTS, the screening and assignment of RINs will be made at the logical point, i.e., the point when RINs are generated through production or importation of renewable fuel. A renewable producer will electronically submit, in “real time,” a volume of renewable fuel produced or imported, as well as a number of the RINs generated and assigned. EMTS will automatically screen each batch and either reject the information or allow RINs created in the RIN generator's account as one of the five types of RINs.

We received comments supporting the RFS1 approach that allows producers and importers to generate RINs at the renewable fuel point of sale. EPA realizes that this is an industry practice and this flexibility will still be allowed for RIN generators, but only if applied consistently.

After RINs have entered the system, parties may then trade them based on agreements outside of EMTS. One major advantage of EMTS, over the RFS1 system, is that the system will simplify trading by allowing RINs to be traded generically. Only some specifying information will be needed to trade RINs, such as RIN quantity, fuel type, RIN assignment, RIN year, RIN price or price per gallon. The unique identification of the RIN will exist within EMTS, but parties engaging in RIN transactions will no longer have to worry about incorrectly recording or using 38-digit RIN numbers. The actual items of transactional information covered under RFS2 are very similar to those reported under RFS1. The RIN price is one of the new pieces of transactional information required to be submitted under RFS2.

We received several adverse comments strongly opposing the collection of price information due to Confidential Business Information (CBI) concerns, other services being able to provide this information, marketplace delays and undue stress on the EMTS from disagreements in RIN price. We received one comment strongly supporting EPA collecting this information. EPA decided that the price information has great programmatic value because it will help us anticipate and appropriately react to market disruptions and other compliance challenges, assess and develop responses to potential waivers, and assist in setting future renewable fuel standards. In addition, EPA decided that highly summarized price information (e.g., the average price of RINs traded nationwide) may be valuable to regulated parties, as well, and may help them to anticipate and avoid market disruptions. Also, EPA will not require the matching of the exact RIN price to alleviate the burden of resubmission due to price mistakes. However, the price information must be accurate and rounded to the nearest cent (U.S. Dollar) at the time of sending the transactional information to EMTS.

We received one comment requesting publication of security precautions taken by EPA to protect EMTS from attacks. EPA cannot provide security information to the public because providing such information may create security vulnerabilities. However, EMTS will be compliant with the appropriate security requirements for all federal agency information technology systems.

Also as with RFS1, there is no “good faith” provision to RIN ownership. An underlying principle of RIN ownership is still one of “buyer beware” and RINs may be prohibited from use at any time if they are found to be invalid. Because of the “buyer beware” aspect, we will offer the option for a buyer to accept or reject RINs from specific RIN generators or from classes of RIN generators.

4. A Sample EMTS Transaction

This sample illustrates how two parties may trade RINs in EMTS:

(1) Seller logs into EMTS and posts a sale of 10,000 RINs to Buyer at X price. For this example, assume the RINs were generated in 2010 and were assigned to 10,000 gallons of “Renewable fuel (D=6)”. Seller's RIN account for “Renewable fuel (D=6)” is put into a “pending” status of 10,000 with the posting of the sale to Buyer. Buyer receives automatic notification of the pending transaction.

(2) Buyer logs into EMTS. Buyer sees the sale transaction pending. Assuming it is correct, Buyer accepts it. Upon acceptance, Buyer's RIN account for “Renewable fuel (D=6)” RINs is automatically increased by 10,000 2010 assigned RINs sold at X price.

(3) After Seller has posted the sale and Buyer has accepted it, EMTS automatically notifies both Buyer and Seller that the transaction has been fully completed.

Under EMTS, the seller will always have to initiate any transaction. The specific amount of RINs are put into a pending status when the seller posts the sale. The buyer must confirm the sale in order to have the RINs transferred to the buyer's account. Transactions will always be limited to available RINs. Notification will automatically be sent to both the buyer and the seller upon completion of the transaction. EPA considers any sale or transfer as complete upon acknowledgement by the buyer. We will also allow buyers to submit their acknowledgement prior to a seller initiating the transaction. However, these buy transactions will not initiate any RINs being put into a pending status from a seller's account. Instead, the buy transactions will be queued and checked periodically to see if a “sell” transaction was posted by the seller. If a buy is posted without a matching sell transaction, then the seller will be notified that a buy transaction is pending. Both buy and sell transactions must be matched within a set number of days from the submission date or they will expire. Transactions will expire 7 days after the submission of the file. Since both parties are required to submit information within 5 days, we allow the full 5 days to expire plus 2 days in the case of late submissions.

In summary, the advantage to implementing EMTS is that parties may engage in RIN transactions with a high degree of confidence, errors will be virtually eliminated, and everyone engaging in RIN transactions will have a simplified environment in which to work, which should minimize the level of resources needed for implementation.

B. Upward Delegation of RIN-Separating Responsibilities

Since the start of the RFS program on September 1, 2007, there have been a number of instances in which a party who receives RINs with a volume of renewable fuel is required to either separate or retire those RINs, but views the recordkeeping and reporting requirements under the RFS program as an unnecessary burden. Such circumstances typically might involve a renewable fuel blender, a party that uses renewable fuel in its neat form, or a party that uses renewable fuel in a non-highway application and is therefore required to retire the RINs (under RFS1) associated with the volume. In some of these cases, the affected party may purchase and/or use only small volumes of renewable fuel and, absent the RFS program, would be subject to few (if any other) EPA regulations governing fuels.

This situation will become more prevalent with the RFS2 rule, as EISA added diesel fuel to the RFS program. With the RFS1 rule, small blenders (generally farmers and other parties that use nonroad diesel fuel) blending small amounts of biodiesel were not covered under the rule as EPAct mandated renewable fuel blending for highway gasoline only. EISA mandates certain amounts of renewable fuels to be blended into all transportation fuels—which includes highway and nonroad diesel fuel. Thus, parties that were not regulated under the RFS1 rule who only blend a small amount of renewable fuel (and, as mentioned above, are generally not subject to EPA fuels regulations) will now be regulated by the RFS program.

Consequently, we believe it is appropriate, and thus we are finalizing as proposed, to permit blenders who only blend a small amount of renewable fuel to allow the party directly upstream to separate RINs on their behalf. Such a provision is consistent with the fact that the RFS program already allows marketers of renewable fuels to assign more RINs to some of their sold product and no RINs to the rest of their sold product. We believe that this provision will eliminate undue burden on small parties who would otherwise not be regulated by this program. This provision is solely for the case of blenders who blend and trade less than 125,000 total gallons of renewable fuel per year (i.e., a company that blends 100,000 gallons and trades another 100,000 gallons would not be able to use this provision) and is available to any blender who must separate RINs from a volume of renewable fuel under § 80.1429(b)(2).

We requested comment in the NPRM on this concept, the 125,000 gallon threshold, and appropriate documentation to authorize this upward delegation. In general, those that commented on this provision support the idea of upward delegation for small blenders, though one commenter stated that EPA should not allow small entities to delegate their RIN-related responsibilities upward. Those commenters that support the upward delegation provision stated that it should be limited to small blenders only and should only be for delegating to the party directly upstream. A few commenters stated that they believe the 125,000 gallon threshold is appropriate; while others commented that it should be higher. We believe that the 125,000 gallon limit strikes the correct balance between providing relief to small blenders, while still ensuring that non-obligated parties cannot unduly influence the RIN market.

We did not receive any comments on appropriate documentation, however a couple commenters suggested that we retain the proposed annual authorization between the blender and the party directly upstream, as well as allowing a small blender to enter into arrangements with multiple suppliers on a transaction-by-transaction basis. Please see Chapter 5 of the Summary and Analysis of Comments Document for more discussion on the comments received and our responses to those comments.

We are also finalizing, as stated in the preamble to the proposed rule, that for upstream delegation, both parties must sign a quarterly written statement (which must be included with the reporting party's reports) authorizing the upward delegation. Copies of these statements must be retained as records by both parties. The supplier would then be allowed to retain ownership of RINs assigned to a volume of renewable fuel when that volume is transferred, under the condition that the RINs be separated or retired concurrently with the transfer of the volume. This statement would apply to all volumes of renewable fuel transferred between the two parties. Thus, the two parties would enter into a contract stating that the supplier has RIN-separation responsibilities for all transferred volumes between the two parties, and no additional permissions from the small blender would be needed for any volumes transferred. A blender may enter into such an agreement with as many parties as they wish.

C. Small Producer Exemption

Under the RFS1 rule, parties who produce or import less than 10,000 gallons of renewable fuel in a year are not required to generate RINs for that volume, and are not required to register with the EPA if they do not take ownership of RINs generated by other parties. These producers and importers are also exempt from registration, reporting, recordkeeping, and attest engagement requirements. In the preamble to the proposed rule, we requested comment on whether or not this 10,000 gallon threshold was appropriate. One commenter suggested that we retain the 10,000 gallon threshold as-is. Another commenter supported the concept of less burdensome requirements for small producers, but suggested that these entities should, at a minimum, be required to generate RINs for all qualifying renewables. We are maintaining this exemption under the RFS2 rule for parties who produce or import less than 10,000 gallons of renewable fuel per year.

In addition to the permanent exemption for those producers and importers who produce or import less than 10,000 gallons of renewable fuel per year, we are also finalizing a temporary exemption for renewable fuel producers who produce less than 125,000 gallons of renewable fuel each year from new production facilities. These producers are not required to generate and assign RINs to batches of renewable fuel for a period of up to three years, beginning with the calendar year in which the production facility produces its first gallon of renewable fuel. Such producers are also exempt from registration, reporting, recordkeeping, and attest engagement requirements as long as they do not own RINs or voluntarily generate and assign RINs. This provision is intended to allow pilot and demonstration plants of new renewable fuel technologies to focus on developing the technology and obtaining financing during these early stages of their development without having to comply with the RFS2 regulations.

D. 20% Rollover Cap

EISA does not change the language in CAA section 211(o)(5) stating that renewable fuel credits must be valid for showing compliance for 12 months as of the date of generation. As discussed in the RFS1 final rulemaking, we interpreted the statute such that credits would represent renewable fuel volumes in excess of what an obligated party needs to meet their annual compliance obligation. Given that the renewable fuel standard is an annual standard, obligated parties determine compliance shortly after the end of the year, and credits would be identified at that time. In the context of our RIN-based program, we have accomplished the statute's objective by allowing RINs to be used to show compliance for the year in which the renewable fuel was produced and its associated RIN first generated, or for the following year. RINs not used for compliance purposes in the year in which they were generated will by definition be in excess of the RINs needed by obligated parties in that year, making excess RINs equivalent to the credits referred to in section 211(o)(5). Excess RINs are valid for compliance purposes in the year following the one in which they initially came into existence. RINs not used within their valid life will thereafter cease to be valid for compliance purposes.

In the RFS1 final rulemaking, we also discussed the potential “rollover” of excess RINs over multiple years. This can occur in situations wherein the total number of RINs generated each year for a number of years in a row exceeds the number of RINs required under the RFS program for those years. The excess RINs generated in one year could be used to show compliance in the next year, leading to the generation of new excess RINs in the next year, causing the total number of excess RINs in the market to accumulate over multiple years despite the limit on RIN life. When renewable fuel volumes are being produced that exceed the RFS2 standards, the rollover issue could undermine the ability of a limit on credit life to guarantee an ongoing market for renewable fuels.

To implement EISA's restriction on the life of credits and address the rollover issue, the RFS1 final rulemaking implemented a 20% cap on the amount of an obligated party's RVO that can be met using previous-year RINs. Thus each obligated party is required to use current-year RINs to meet at least 80% of its RVO, with a maximum of 20% being derived from previous-year RINs. Any previous-year RINs that an obligated party may have that are in excess of the 20% cap can be traded to other obligated parties that need them. If the previous-year RINs in excess of the 20% cap are not used by any obligated party for compliance, they will thereafter cease to be valid for compliance purposes.

As described in the NPRM, EISA does not modify the statutory provisions regarding credit life, and the volume changes by EISA also do not change at least the possibility of large rollovers of RINs for individual obligated parties. As a result we proposed to maintain the regulatory requirement for a 20% rollover cap under the new RFS2 program, and to apply this cap separately to all four RVOs under RFS2. However, we took comment on changing the level of the cap to some alternative value lower or higher than 20%.

A lower cap could provide a greater incentive for parties with excess RINs to sell them rather than hold onto them, increasing the availability of RINs for parties that need them for compliance purposes. But a lower cap would also reduce flexibility for obligated parties attempting to minimize the costs of compliance with increasing annual volume requirements, particularly if there are concerns that the RIN market may be tighter in the future than it is currently.

Conversely, the increasing annual volume requirements in EISA make it less likely that renewable fuel producers will overcomply, and as a result it is less likely that there will be an excess of RINs in the market. Under these circumstances, there is little opportunity for RINs to build up in the market, and the rollover cap would have less of an impact on the market as a whole. Thus a higher cap might be warranted. However, while a higher cap would create greater flexibility for some obligated parties, it could also create disruptions in the RIN market as parties with excess RINs would have a greater opportunity to hold onto them rather than sell them. Parties without direct access to RINs through the purchase and blending of renewable fuels would be placed at a competitive disadvantage in comparison to parties with excess RINs. In the extreme, removal of the cap entirely would allow obligated parties to roll over up to one year's worth of their obligations indefinitely.

In general, commenters on the NPRM reiterated the positions that they raised during development of the RFS1 program. While one renewable fuel producer requested that the rollover cap be left at 20%, most producers requested that the rollover cap be reduced to 0%, such that compliance with the standards applicable in a given year could only be demonstrated using RINs generated in that year. In contrast, refiners requested that the rollover cap be either eliminated, such that any number of previous year RINs could be used for current year compliance, or at least raised to 40 or 50 percent. Small refiners requested that the cap be raised for small refiners only to accommodate the competitive disadvantage with respect to the RIN market that they believe they experience in comparison to larger refiners.

Based on the comments received, we believe that the 20% level continues to provide the appropriate balance between, on the one hand, allowing legitimate RIN carryovers and protecting against potential supply shortfalls that could limit the availability of RINs, and on the other hand ensuring an annual demand for renewable fuels as envisioned by EISA. Therefore, we are continuing the 20% rollover cap for obligated parties for the RFS program.

E. Small Refinery and Small Refiner Flexibilities

This section discusses flexibilities for small refineries and small refiners for the RFS2 rule. As explained in the discussion of our compliance with the Regulatory Flexibility Act below in Section XI.C and in the Final Regulatory Flexibility Analysis in Chapter 7 of the RIA, we considered the impacts of the RFS2 regulations on small businesses (small refiners). Most of our analysis of small business impacts was performed as a part of the work of the Small Business Advocacy Review Panel (SBAR Panel, or “the Panel”) convened by EPA for this rule, pursuant to the Regulatory Flexibility Act as amended by the Small Business Regulatory Enforcement Fairness Act of 1996 (SBREFA). The Final Report of the Panel is available in the rulemaking docket. For the SBREFA process, we conducted outreach, fact-finding, and analysis of the potential impacts of our regulations on small business refiners.

1. Background—RFS1

a. Small Refinery Exemption

CAA section 211(o)(9), enacted as part of EPAct, provides a temporary exemption to small refineries (those refineries with a crude throughput of no more than 75,000 barrels of crude per day, as defined in section 211(o)(1)(K)) through December 31, 2010. [35] Accordingly, the RFS1 program regulations exempt gasoline produced by small refineries from the renewable fuels standard (unless the exemption was waived), see 40 CFR 80.1141. EISA did not alter the small refinery exemption in any way.

b. Small Refiner Exemption

As mentioned above, EPAct granted a temporary exemption from the RFS program to small refineries through December 31, 2010. In the RFS1 final rule, we exercised our discretion under section 211(o)(3)(B) and extended this temporary exemption to the few remaining small refiners that met the Small Business Administration's (SBA) definition of a small business (1,500 employees or less company-wide) but did not meet the EPAct small refinery definition as noted above.

2. Statutory Options for Extending Relief

There are two provisions in section 211(o)(9) that allow for an extension of the temporary exemption for small refineries beyond December 31, 2010.

One provision involves a study by the Department of Energy (DOE) concerning whether compliance with the renewable fuel requirements would impose disproportionate economic hardship on small refineries, and would grant an automatic extension of at least two years for small refineries that DOE determines would be subject to such disproportionate hardship (per section 211(o)(9)(A)(ii)). If the DOE study determines that such hardship exists, then section 211(o)(9)(A)(ii) (which was retained in EISA) provides that EPA shall extend the exemption for a period of at least two years.

The second provision, at section 211(o)(9)(B), authorizes EPA to grant an extension for a small refinery based upon disproportionate economic hardship, on a case-by-case basis. A small refinery may, at any time, petition EPA for an extension of the small refinery exemption on the basis of disproportionate economic hardship. EPA is to consult with DOE and consider the findings of the DOE small refinery study in evaluating such petitions. These petitions may be filed at any time, and EPA has discretion to determine the length of any exemption that may be granted in response.

3. The DOE Study/DOE Study Results

As discussed above, EPAct required that DOE perform a study by December 31, 2008 on the impact of the renewable fuel requirements on small refineries (section 211(o)(9)(A)(ii)(I)), and whether or not the requirements would impose a disproportionate economic hardship on these refineries. In the small refinery study, “EPACT 2005 Section 1501 Small Refineries Exemption Study,” DOE's finding was that there is no reason to believe that any small refinery would be disproportionately harmed by inclusion in the proposed RFS2 program. This finding was based on the fact that there appeared to be no shortage of RINs available under RFS1, and EISA has provided flexibility through waiver authority (per section 211(o)(7)). Further, in the case of the cellulosic biofuel standard, cellulosic biofuel allowances can be provided from EPA at prices established in EISA (see regulation section 80.1456). DOE thus determined that small refineries would not be subject to disproportionate economic hardship under the proposed RFS2 program, and that the exemption should not, on the basis of the study, be extended for small refineries (including those small refiners who own refineries meeting the small refinery definition) beyond December 31, 2010. DOE noted in the study that, if circumstances were to change and/or the RIN market were to become non-competitive or illiquid, individual small refineries have the ability to petition EPA for an extension of their small refinery exemption (pursuant to Section 211(o)(9)(B)).

4. Ability To Grant Relief Beyond 211(o)(9)

The SBREFA panel made a number of recommendations for regulatory relief and additional flexibility for small refineries and small refiners. These are described in the Final Panel Report (located in the rulemaking docket), and summarized below. During the development of this final rule, we again evaluated the various options recommended by the Panel and also comments on the proposed rule. We also consulted the small refinery study prepared by DOE.

As described in the Final Panel Report, EPA early-on identified limitations on its authority to issue additional flexibility and exemptions to small refineries. In section 211(o)(9) Congress specifically addressed the issue of an extension of time for compliance for small refineries, temporarily exempting them from renewable fuel obligations through December 31, 2010. As discussed above, the statute also includes two specific provisions describing the basis and manner in which further extensions of this exemption can be provided. In the RFS1 rulemaking, EPA considered whether it should provide additional relief to the limited number of small refiners who were not covered by the small refinery provision, by providing them a temporary exemption consistent with that provided by Congress for small refineries. EPA exercised its discretion under section 211(o)(3) and provided such relief. Thus, in RFS1, EPA did not modify the relief provided by Congress for small refineries, but did exercise its discretion to provide the same relief specified by statute to a few additional parties.

In RFS2 we are faced with a different issue—the extent to which EPA should provide additional relief to small refineries beyond the relief specified by statute, and whether it should provide such further relief to small refiners as well. There is considerable overlap between entities that are small refineries and those that are small refiners. Providing additional relief just to small refiners would, therefore, also extend additional relief to at least a number of small refineries. Congress spoke directly to the relief that EPA may provide for small refineries, including those small refineries operated by small refiners, and limited that relief to a blanket exemption through December 31, 2010, with additional extensions if the criteria specified by Congress are met. EPA believes that an additional or different extension, relying on a more general provision in section 211(o)(3) would be inconsistent with Congressional intent. Further, we do not believe that the statute allows us the discretion to give relief to small refiners only—as this would result in a subset of small refineries (those that also qualify as small refiners) receiving relief that is greater than the relief already given to all small refineries under EISA.

EPA also notes that the criteria specified by statute for providing a further compliance extension to small refineries is a demonstration of “disproportionate economic hardship.” The statute provides that such hardship can be identified through the DOE study, or in individual petitions submitted to the Agency. However, the DOE study has concluded that no disproportionate economic hardship exists, at least under current conditions and for the foreseeable future under RFS2. Therefore, absent further information that may be provided through the petition process, there does not currently appear to be a basis under the statute for granting further compliance extensions to small refineries. If DOE revises its study and comes to a different conclusion, EPA can revisit this issue.

5. Congress-Requested Revised DOE Study

In their written comments, as well as in discussions we had with them on the proposed rule, small refiners indicated that they did not believe that EPA should rely on the results of the DOE small refinery study to inform any decisions on small refiner provisions. Small refiners generally commented that they believe that the study was flawed and that the conclusions of the study were reached without adequate analysis of, or outreach with, small refineries (as the majority of the small refiners own refineries that meet the Congressional small refinery definition). One commenter stated that such a limited investigation into the impact on small refineries could not have resulted in any in-depth analysis on the economic impacts of the program on these entities. Another commenter stated that it believes that DOE should be directed to reopen and reassess the small refinery study be June 30, 2010, as suggested by the Senate Appropriations Committee.

We are aware that there have been expressions of concern from Congress regarding the DOE Study. Specifically, in Senate Report 111-45, the Senate Appropriations Committee “directed [DOE] to reopen and reassess the Small Refineries Exemption Study by June 30, 2010,” noting a number of factors that the Committee intended that DOE consider in the revised study. The Final Conference Report 111-278 to the Energy & Water Development Appropriations Act (H.R. 3183), referenced the language in the Senate Report, noting that the conferees “support the study requested by the Senate on RFS and expect the Department to undertake the requested economic review.” At the present time, however, the DOE study has not been revised. If DOE prepares a revised study and the revised study finds that there is a disproportionate economic impact, we will revisit the exemption extension at that point in accordance with section 211(o)(9)(A)(ii).

6. What We're Finalizing

a. Small Refinery and Small Refiner Temporary Exemptions

As mentioned above, the RFS1 program regulations exempt gasoline produced by small refineries from the renewable fuels standard through December 31, 2010 (at 40 CFR 80.1141), per EPAct. As EISA did not alter the small refinery exemption in any way, we are retaining this small refinery temporary exemption in the RFS2 program without change (except for the fact that all transportation fuel produced by small refineries will be exempt, as EISA also covers diesel and nonroad fuels).

Likewise, as we extended under RFS1 the small refinery temporary exemption to the few remaining small refiners that met the Small Business Administration's (SBA) definition of a small business (1,500 employees or less company-wide), we are also finalizing a continuation of the small refiner temporary exemption through December 31, 2010.

b. Case-by-Case Hardship for Small Refineries and Small Refiners

As discussed in Section III.E.2, EPAct also authorizes EPA to grant an extension for a small refinery based upon disproportionate economic hardship, on a case-by-case basis. We believe that these avenues of relief can and should be fully explored by small refiners who are covered by the small refinery provision. In addition, we believe that it is appropriate to allow petitions to EPA for an extension of the temporary exemption based on disproportionate economic hardship for those small refiners who are not covered by the small refinery provision (again, per our discretion under section 211(o)(3)(B)); this would ensure that all small refiners have the same relief available to them as small refineries do. Thus, we are finalizing a hardship provision for small refineries in the RFS2 program, that any small refinery may apply for a case-by-case hardship at any time on the basis of disproportionate economic hardship per CAA section 211(o)(9)(B). We are also finalizing a case-by-case hardship provision for those small refiners that do not operate small refineries using our discretion under CAA section 211(o)(3)(B). This provision will allow those small refiners that do not operate small refineries to apply for the same kind of hardship extension as a small refinery. In evaluating applications for this hardship provision EPA will take into consideration information gathered from annual reports and RIN system progress updates, as recommended by the SBAR Panel, as well as information provided by the petitioner and through consultation with DOE.

c. Program Review

During the SBREFA process, the small refiner Small Entity Representatives (SERs) also requested that EPA perform an annual program review, to begin one year before small refiners are required to comply with the program, to provide information on RIN system progress. As mentioned in the preamble to the proposed rule, we were concerned that such a review could lead to some redundancy with the notice of the applicable RFS standards that EPA will publish in the Federal Register annually, and this annual process will inevitably include an evaluation of the projected availability of renewable fuels. Nevertheless, some Panel members commented that they believe a program review could be beneficial to small entities in providing them some insight to the RFS program's progress and alleviate some uncertainty regarding the RIN system. As we will be publishing a Federal Register notice annually, the Panel recommended, and we proposed, that an update of RIN system progress (e.g., RIN trading, publicly-available information on RIN availability, etc.) be included in this annual notice.

Based on comments received on the proposed rule, we believe that such information could be helpful to industry, especially to small businesses to help aid the proper functioning of the RIN market, especially in the first years of the program. However, during the development of the final rule, it became evident that there could be instances where we would want to report out RIN system information on a more frequent basis than just once a year. Thus we are finalizing that we will periodically report out elements of RIN system progress; but such information will be reported via other means (e.g., the RFS Web site (http://www.epa.gov/otaq/renewablefuels/index.htm), EMTS homepage, etc.).

7. Other Flexibilities Considered for Small Refiners

During the SBREFA process, and in their comments on the proposed rule, small refiners informed us that they would need to rely heavily on RINs and/or make capital improvements to comply with the RFS2 requirements. These refiners raised concerns about the RIN program itself, uncertainty (with the required renewable fuel volumes, RIN availability, and costs), the desire for an annual RIN system review, and the difficulty in raising capital and competing for engineering resources to make capital improvements.

The Panel recommended that EPA consider the issues raised by the small refiner SERs and discussions had by the Panel itself, and that EPA should consider comments on flexibility alternatives that would help to mitigate negative impacts on small businesses to the extent allowable by the Clean Air Act. A summary of further recommendations of the Panel are discussed in Section XI.C of this preamble, and a full discussion of the regulatory alternatives discussed and recommended by the Panel can be found in the SBREFA Final Panel Report. Also, a complete discussion of comments received on the proposed rule regarding small refinery and small refiner flexibilities can be found in Chapter 5 of the Summary and Analysis of Comments document.

a. Extensions of the RFS1 Temporary Exemption for Small Refiners

As previously stated, the RFS1 program regulations provide small refiners who operate small refineries, as well as those small refiners who do not operate small refineries, with a temporary exemption from the standards through December 31, 2010. This provided an exemption for small refineries (and small refiners) for the first five years of the RFS program. Small refiner SERs suggested that an additional temporary exemption for the RFS2 program would be beneficial to them in meeting the RFS standards as increased by Congress in EISA. The Panel recommended that EPA propose a delay in the effective date of the standards until 2014 (for a total of eight years) for small entities, to the extent allowed by the statute.

During the development of both the Final Panel Report and the proposed rule, we evaluated various options for small refiners, including an additional temporary exemption for small refiners from the required RFS2 standards. As discussed above, we concluded that we do not have the statutory authority to provide such extensions through means other than those specified in the statute. Thus, further extensions will be as a result of any revised DOE study, or in response to a petition, pursuant to the authorities specified in section 211(o)(9).

We proposed to continue the temporary exemption finalized in RFS1—through December 31, 2010. Commenters that oppose an extension of the temporary exemption generally stated that an extension is not warranted, and some commenters expressed concerns about allowing provisions for small refiners. One commenter also stated that it believes that the small refinery exemption should not be extended and that the small refiner exemption should be eliminated completely. Two commenters supported the continuation of the exemption through December 31, 2010 only, and one stated that it does not support an extension as it believes that all parties have been well aware of the passage of EISA and small refineries and small refiners should have been striving to achieve compliance by the end of 2010. Two commenters also expressed views that the exemption should not have been offered to small refiners in RFS1 as this was not provided by EPAct, and that an extension of the exemption should not be finalized for small refineries at all. The commenters further commented that an economic hardship provision was included in EPAct, and any exemption extension should be limited to such cases, and only to the specific small refinery (not small refiner) that has petitioned for such an extension.

Commenters supporting an extension of the exemption commented that they believe that the statutes (EPAct and EISA) do not prohibit EPA from providing relief to regulated small entities on which the rule will have a significant economic impact, and that such a delay could lessen the burden on these entities. One commenter stated that it believes EPA denied or ignored much of the relief recommended by the Panel in the proposal. Another commenter stated that it believes EPA's concerns regarding the legal authority are unsustainable considering EPA's past exercises of discretion under the RFS1 program, and with the discretion afforded to EPA under section 211(o) of the CAA. Some commenters requested a delay until 2014 for small refiners. One additional commenter expressed support for an extension of the small refinery exemption only, and that these small refineries should be granted a permanent exemption.

During the development of this final rule, we again evaluated the various options recommended by the Panel, the legality of offering an extension of the exemption to small refiners only, and also comments on the proposed rule. Specifically in the case of an extension of the exemption for small refiners, we also consulted the small refinery study prepared by DOE, as the statute directs us to use this as a basis for providing an additional two year exemption. As discussed above in Sections III.E.4 and 5, we do not believe that we can provide an extension of the exemption considering the outcome of the DOE small refinery study, which did not find that there was a disproportionate economic hardship. Further, we do not believe that the statute allows us the discretion to give relief to a subset of small refineries (those that also qualify as small refiners) that is greater than the relief already given to all small refineries under EPAct. However, it is important to recognize that the 211(o)(9) small refinery provision does allow for extensions beyond December 31, 2010, as discussed above in Section III.E.2. Thus, refiners may apply for individual hardship relief.

b. Phase-in

The small refiner SERs suggested that a phase-in of the obligations applicable to small refiners would be beneficial for compliance, such that small refiners would comply by gradually meeting the standards on an incremental basis over a period of time, after which point they would comply fully with the RFS2 standards. However we stated in the NPRM that we had serious concerns about our legal authority to provide such a phase-in. CAA section 211(o)(3)(B) states that the renewable fuel obligation shall “consist of a single applicable percentage that applies to all categories of persons specified” as obligated parties. A phase-in approach would essentially result in different applicable percentages being applied to different obligated parties. Further, such a phase-in approach would provide more relief to small refineries operated by small refiners than that provided under the statutory small refinery provisions.

Some commenters stated that they believe that EPA has the ability to consider a phase-in of the standards for small refiners. One commenter suggested that a temporary phase-in could help lessen the burden of regulation on small entities and promote compliance. Another commenter stated that it believes EPA's legal concerns regarding a phase-in are unsustainable considering EPA's past exercises of discretion under the RFS1 program and with the discretion afforded to EPA under section 211(o) of the CAA.

After considering the comments on this issue, EPA continues to believe that allowing a phase-in of regulatory requirements for small refineries and/or small refiners would be inconsistent with the statute, for the reasons mentioned above. Any individual entities that are experiencing hardship that could justify a phase-in of the standards have the ability to petition EPA for individualized relief. Therefore we are not including a phase-in of standards for small refiners in today's rule.

c. RIN-Related Flexibilities

The small refiner SERs requested that the RFS2 rule contain provisions for small refiners related to the RIN system, such as flexibilities in the RIN rollover cap percentage and allowing small refiners only to use RINs interchangeably. In the RFS1 rule, up to 20% of a previous year's RINs may be “rolled over” and used for compliance in the following year. In the preamble to the proposed rule, we discussed the concept of allowing for flexibilities in the rollover cap, such as a higher RIN rollover cap for small refiners for some period of time or for at least some of the four standards. As the rollover cap is the means through which we are implementing the limited credit lifetime provisions in section 211(o) of the CAA, and therefore cannot simply be eliminated, we requested comment on the concept of increasing the RIN rollover cap percentage for small refiners and an appropriate level of that percentage. In response to the Panel's recommendation, we also sought comment on allowing small refiners to use the four types of RINs interchangeably.

In their comments on the proposed rule, one small refiner commented that, in regards to small refiners' concerns about RIN pricing and availability, there is no mechanism in the rule to address the possibility that the RIN market will not be viable. The commenter further suggested that more “durable” RINs are needed for small refiners that can be carried over from year to year, to alleviate some of the potentially market volatility for renewable fuels. Another commenter suggested that RINs should be interchangeable for small refiners, or alternatively, some mechanism should be implemented to ensure that RIN prices are affordable for small refiners. Further, with regard to interchangeable RINs, one commenter stated that small refiners do not have the staff or systems to manage and account for four different categories of RINs and rural small refiners will suffer economic hardship and disadvantage because of the unavailability of biofuels. The commenter also requested an increase in the rollover cap to 50% for small refiners.

We are not finalizing additional RIN-related flexibilities for small refiners in today's action. As highlighted in the NPRM, we continue to believe that the concept of interchangeable RINs for small refiners only fails to require the four different standards mandated by Congress (e.g., conventional biofuel could not be used instead of cellulosic biofuel or biomass-based diesel), and is not consistent with section 211(o) of the Clean Air Act. Essentially, it would circumvent the explicit direction of Congress in EISA to require that the four RFS2 standards be met separately. Further, given the findings from the DOE study that small refineries (and thus, most small refiners) do not currently face disproportionate economic hardship, and are not expected to do so as RFS2 is implemented, we do not believe that a basis exists to justify providing small refiners with a larger rollover cap than other regulated entities. Thus, small refiners will be held to the same RIN rollover cap as other obligated parties.

F. Retail Dispenser Labeling for Gasoline With Greater Than 10 Percent Ethanol

We proposed labeling requirements for fuel dispensers that handle greater than 10 volume percent ethanol blends which included the following text: For use only in flexible-fuel vehicles, May damage non-flexible-fuel vehicles, Federal law prohibits use in non-flexible-fuel vehicles. This proposal was primarily meant to help address concerns about the potential misfueling of non-flex-fuel vehicles with E85, in light of the anticipated increase in E85 sales volumes in response to the RFS2 program. All ethanol blends above 10 volume percent were included due to the increasing industry focus on ethanol blender pumps that are designed to dispense a variety of ethanol blends (e.g., E30, and E40) for use in flex-fuel vehicles.

Commenters stated that EPA should undertake additional analysis of the potential impacts from misfueling and what preventative measures might be appropriate before finalizing labeling requirements for >E10 blends. They also stated that EPA should coordinate any such labeling provisions with those already in place by the Federal Trade Commission. EPA is also currently evaluating a petition to allow the use of up to 15 volume percent ethanol in non-flex fuel vehicles. One potential result of this evaluation might be for EPA to grant a partial waiver that is applicable only for a subset of the current vehicle population. Under such an approach, a label for E15 fuel dispensers would be needed that identifies what vehicles are approved to use E15.

Based on the public comments and the fact that EPA has not completed its evaluation of the E15 waiver petition, we believe that it is appropriate to defer finalizing labeling requirements for >E10 blends at this time. This will afford us the opportunity to complete our analysis of what measures might be appropriate to prevent misfueling with >E10 blends before this may become a concern in the context of the RFS2 program.

G. Biodiesel Temperature Standardization

The volume of a batch of renewable fuel can change under extreme changes in temperature. The volume of a batch of renewable fuel can experience expansion as the temperature increases, or can experience contraction as temperature decreases. The Agency requires temperature standardization of renewable fuels at 60° Fahrenheit (°F) so renewable fuel volumes are accounted for on a uniform and consistent basis over the entire fuels industry. In the May 1, 2007 Renewable Fuels Standard (RFS) final rule the Agency required biodiesel temperature standardization to be completed as follows:

V s,b= V a,b× (−0.0008008 × T + 1.0480)

Where

V s,b= Standard Volume of biodiesel at 60 degrees F, in gallons;

V a,b= Actual volume of biodiesel, in gallons;

T = Actual temperature of batch, in degrees F.

This equation was based on data from a published research paper by Tate et al. [36] Members of the petroleum industry have indicated that the current biodiesel temperature standardization equation in the regulations provides different results than that commonly used by both the petroleum and biodiesel industry for commercial trading of biodiesel. These commercial values are either based on American Petroleum Institute (API) tables for petroleum products or on empirical values from industry measurements at common temperatures and pressures observed in bulk fuel facilities. The difference between RIN calculated volumes and commercial sales volumes has created confusion within the record keeping system of both the petroleum and biodiesel industry.

In the RFS2 proposed rule, the Agency proposed the temperature standardization of biodiesel remain unchanged from the RFS1 requirements. [37] The Agency received comments from Archer Daniels Midland Company (ADM), World Energy Alternatives, Marathon Petroleum Company (Marathon) and the National Biodiesel Board (NBB) to revise the biodiesel temperature standardization equation.

Both ADM and NBB agreed on the necessity for biodiesel temperature standardization at 60 °F. ADM and NBB commented on several empirical calculations which have been developed specific to biodiesel temperature standardization since the 2007 RFS1 final rule. These include a 2004 data set developed by the Minnesota Department of Commerce and the Renewable Energy Group and updated in 2008; information embedded in the European Biodiesel Specification EN 14214; and information from the Alberta Research Council. The table below provides values from NBB for 1000 gallons of biodiesel standardized to a temperature at 60 °F for these empirical calculations, along with the current EPA equation, and the American Petroleum Institute (API) Refined Products Table 6.

Table III.G-1—NBB Comparison of Biodiesel Temperature Standardization Calculations to 60 °F for 1000 gallons of Biodiesel at 90 °F Back to Top
Gallons
2007 EPA Biodiesel Formula 975.28
2008 Minnesota (Hedman) data 986.270
API Refined Products Table 6 (biodiesel density @ 7.359) 986.625
Alberta Research Council 986.238
EN 14214 data 986.401
2004 Minnesota Renewable Energy Group data 986.830

As illustrated by the results from the above table, the values for the various biodiesel temperature standardization empirical calculations are within 1 gallon of agreement of each other for a 1000 gallon biodiesel batch, except for the current biodiesel temperature standardization equation in the regulations.

To ensure consistency in RIN generation, ADM commented EPA should adopt only one biodiesel temperature standardization calculation. ADM commented that all biodiesel temperature standardization calculations developed, including the API Refined Products Table 6, are in very close agreement with each other and the differences between them all are insignificant. They further commented the API Refined Products Table 6 has provided a uniform measurement of volume for years for the entire liquid fuels industry. Thus, ADM believes the API Refined Products Table 6 should be adopted for biodiesel to be consistent with the calculation of sales volumes. Finally ADM comments adoption of the API Refined Products Table 6 would allow for easier verification within the marketplace, eliminate the need for calculating one volume for sales and trades and another for RINs, and prevents the entire distribution network from facing the financial burden of reprogramming existing meters that already are based on the API Refined Products Table 6.

NBB commented that earlier surveys from its members indicate a fifty-fifty split between members using the API Refined Products Table 6 or some variation of the current EPA biodiesel formula for biodiesel temperature standardization. Some NBB members indicated that the API Refined Products Table 6 was more commonly used by the petroleum industry and embedded into the meters, pumps and accounting systems of the petroleum industry. Companies already using the API Refined Products Table 6 would have a reduction in required paperwork with RIN generation and tracking because already existing commercial documents could serve that purpose and they thus could eliminate or reduce their current dual tracking system. Other NBB members have already embedded the current EPA biodiesel equation within their accounting and sales systems and would like to continue using that type of biodiesel temperature standardization approach rather than the API Refined Products Table 6. The NBB recommended EPA revise its current equation in the regulations to the 2008 Hedman biodiesel temperature standardization equation. Thus, NBB commented EPA should provide flexibility to their members by allowing the use of either the API Refined Products Table 6 or the use of a biodiesel temperature standardization equation.

Marathon commented the regulations allow for the standardization of volume for other renewable fuels to be determined by an appropriate formula commonly accepted by the industry which may be reviewed by the EPA for appropriateness. They recommended that EPA extend this courtesy to biodiesel.

The Agency acknowledges that the current biodiesel temperature standardization equation is likely not correct for biodiesel temperature standardization at ambient temperatures observed in the fuel distribution system. Based on the comments received, the Agency is amending the regulations to allow for two ways for biodiesel temperature standardization: (1) The American Petroleum Institute Refined Products Table 6B, as referenced in ASTM D1250-08, entitled, “Standard Guide for Use of the Petroleum Measurement Tables”, and (2) a biodiesel temperature standardization equation that utilizes the 2008 data generated by the Minnesota Department of Commerce and the Renewable Energy Group. These two methods for biodiesel temperature standardization are within one gallon of agreement of each other for a 1000 gallon biodiesel batch and thus in very close agreement. Both ADM and NBB acknowledged that the differences between these two methods are insignificant and the resulting corrected volumes from these two methods of calculation are within accuracy tolerances of any metered measurement. Thus, the Agency believes the allowance of both of these methods for biodiesel temperature standardization will increase flexibility while still providing for a consistent generation and accounting of biodiesel RINs over the entire fuel delivery system.

IV. Renewable Fuel Production and Use Back to Top

An assessment of the impacts of increased volumes of renewable fuel must begin with an analysis of the kind of renewable fuels that could be used, the types and locations of their feedstocks, the fuel volumes that could be produced by a given feedstock, and any challenges associated with their use. This section provides an assessment of the potential feedstocks and renewable fuels that could be used to meet the Energy Independence and Security Act (EISA) and the rationale behind our projections of various fuel types to represent the control cases for analysis purposes. As new technologies, feedstocks, and fuels continue to develop on a daily basis, markets may appear differently from our projections. Although actual volumes and feedstocks may differ, we believe the projections made for our control cases are within the range of possible predictions for which the standards are met and allow for an assessment of the potential impacts of the increases in renewable fuel volumes that meet the requirements of EISA.

A. Overview of Renewable Fuel Volumes

EISA mandates the use of increasing volumes of renewable fuel. To assess the impacts of this increase in renewable fuel volume from business-as-usual (what is likely to have occurred without EISA), we have established reference and control cases from which subsequent analyses are based. The reference cases are projections of renewable fuel volumes without the enactment of EISA and are described in Section IV.A.1. The control cases are projections of the volumes and types of renewable fuel that might be used in the future to comply with the EISA volume mandates. For the NPRM we had focused on one primary control case (see Section IV.A.2) whereas for the final rule we have expanded the analysis to include two additional sensitivity cases (see Section IV.A.3). Based on the public comments received as well as new information, we have updated the primary control case volumes from the NPRM to reflect what we believe could be a more likely set of volumes to analyze. We assume in each of the cases the same ethanol-equivalence basis as was used in the RFS1 rulemaking to meet the standard. Volumes are listed in tables for this section in both straight-gallons and ethanol-equivalent gallons (i.e., times 1.5 for biodiesel or 1.7 for cellulosic diesel and renewable diesel). The volumes included in this section are for 2022. For intermediate years, refer to Section 1.2 of the RIA.

1. Reference Cases

Our primary reference case renewable fuel volumes are based on the Energy Information Administration's (EIA) Annual Energy Outlook (AEO) 2007 reference case projections. [38] While AEO 2007 is not as up-to-date as AEO 2008 or AEO 2009, we chose to use AEO 2007 because later versions of AEO already include the impact of increased renewable fuel volumes under EISA as well as fuel economy improvements under CAFE as required in EISA, whereas AEO 2007 did not.

For the final rule we have also assessed a number of the impacts relative to a reference case assuming the mandated renewable fuel volumes under RFS1 from the Energy Policy Act of 2005 (EPAct). This allows for a more complete assessment of the impacts of the EISA volume mandates, especially when combined with the impacts assessment conducted for the RFS1 rulemaking (though many factors have changed since then). Table IV.A.1-1 summarizes the 2022 renewable fuel volumes for the AEO 2007 and the RFS1 reference cases (listed in both straight volumes and ethanol-equivalent volumes).

Table IV.A.1-1—Reference Case Renewable Fuel Volumes in 2022 Back to Top
Source/volume type Advanced biofuel Non-advanced biofuel Total renewable fuel
Cellulosic biofuel Biomass-based diesel a Other advanced biofuel Corn ethanol
Cellulosic ethanol c FAME biodiesel b Imported ethanol
[Billion gallons]
a Biomass-Based Diesel could include FAME biodiesel, cellulosic diesel, and non-co-processed renewable diesel.
b Only fatty acid methyl ester (FAME) biodiesel volumes were considered.
c Under the RFS1 reference case, we assumed the 250-million gallon cellulosic standard set by EPAct would be met primarily by corn ethanol plants utilizing 90% biomass for energy, thus actual production of cellulosic biofuel is zero. AEO 2007 reference case assumes actual production of cellulosic biofuel and therefore assumed to be 0.25 billion gallons.
AEO 2007 Straight Volume 0.25 0.38 0.64 12.29 13.56
AEO 2007 Ethanol-Equivalent 0.25 0.58 0.64 12.29 13.76
RFS 1 Straight Volume 0.00 0.30 0.00 7.05 7.35
RFS 1 Ethanol-Equivalent 0.00 0.45 0.00 7.05 7.50

2. Primary Control Case

Our assessment of the renewable fuel volumes required to meet EISA necessitates establishing a primary set of fuel types and volumes on which to base our assessment of the impacts of the new standards. EISA contains four broad categories: cellulosic biofuel, biomass-based diesel, total advanced biofuel, and total renewable fuel. As these categories could be met with a wide variety of fuel choices, in order to assess the impacts of increased volumes of renewable fuel, we projected a set of reasonable renewable fuel volumes based on our projection of fuels that could come to market.

Although actual volumes and feedstocks will be different, we believe the projections made for our control cases are within the range of possible predictions for which the standards are met and allow for an assessment of the potential impacts of increased volumes of renewable fuel. Table IV.A.2-1 summarizes the fuel types used for the primary control case and their corresponding volumes for the year 2022.

Table IV.A.2-1—Primary Control Case Projected Renewable Fuel Volumes in 2022 Back to Top
Volume type Advanced biofuel Non-advanced biofuel Total renewable fuel
Cellulosic biofuel Biomass-based diesela Other advanced biofuel Corn ethanol
Cellulosic ethanol Cellulosic dieselb FAMEcbiodiesel NCRDd Other biodiesele Imported ethanol
[Billion gallons]
aBiomass-Based Diesel could include FAME biodiesel, cellulosic diesel, and non-co-processed renewable diesel.
bCellulosic Diesel includes at least 1.96 billion gallons (3.33 billion ethanol-equivalent gallons) from Fischer-Tropsch Biomass-to-Liquids (BTL) processes based on EIA's forecast and an additional 4.56 billion gallons (7.75 billion ethanol-equivalent gallons) from this or other types of cellulosic diesel processes.
cFatty acid methyl ester (FAME) biodiesel.
dNon-Co-processed Renewable Diesel (NCRD).
eOther Biodiesel is biodiesel that could be produced in addition to the amount needed to meet the biomass-based diesel standard.
Straight Volume 4.92 6.52 0.85 0.15 0.82 2.24 15.00 30.50
Ethanol-Equivalent 4.92 11.08 1.28 0.26 1.23 2.24 15.00 36.00

The following subsections detail our rationale for projecting the amount and type of fuels needed to meet EISA as shown in Table IV.A.2-1. For cellulosic biofuel we have assumed that by 2022 on a straight-volume basis about half would come from cellulosic ethanol and the other half from cellulosic diesel. On an ethanol-equivalent volume basis, cellulosic diesel would make up almost 70% of the 16 billion gallons cellulosic biofuel standard. Biomass-based diesel is assumed to be comprised of a majority of fatty-acid methyl ester (FAME) biodiesel and a smaller portion of non-co-processed renewable diesel. The portion of the advanced biofuel category not met by cellulosic biofuel and biomass-based diesel is assumed to come mainly from imported sugarcane ethanol with a smaller amount from additional biodiesel sources. The total renewable fuel volume not required to be comprised of advanced biofuels is assumed to be met with corn ethanol with small amounts of other grain starches and waste sugars.

The main difference between the volumes used for the NPRM and the volumes used for the FRM is the inclusion of cellulosic diesel for the FRM. The NPRM made the simplifying assumption that the cellulosic biofuel standard would be met entirely with cellulosic ethanol. However, due to growing interest and recent developments in hydrocarbon-based or so-called “drop-in” renewable fuels as well as butanol, and marketplace challenges for consuming high volumes of ethanol, we have included projections of more non-ethanol renewables in our primary control case for the final rule. [39] In the future, this could include various forms of “green hydrocarbons” (i.e., cellulosic gasoline, diesel and jet) and higher alcohols, but for analysis purposes, we have modeled it as cellulosic diesel fuel. We describe these fuels in greater detail in Section IV.B-D. We have also included some algae-derived biofuels in our FRM analyses given the large interest and potential for such fuels. We have continued to assume zero volume for renewable fuels or blendstocks such as biogas, jatropha, palm, imported cellulosic biofuel, and other alcohols or ethers in our control cases. Although we have not included these renewable fuels and blendstocks in our impact analyses, it is important to note that they can still be counted under our program if they meet the lifecycle thresholds and definitions for renewable biomass, and recent information suggests that some of them may be likely.

a. Cellulosic Biofuel

As discussed in our NPRM, whether cellulosic biofuel is ethanol will depend on a number of factors, including production costs, the form of tax subsidies, credit programs, and factors influencing the blending of biofuel into the fuel pool. It will also depend on the relative demand for gasoline and diesel fuel. As a result of our analyses on ethanol consumption (see Section IV.D) and continual tracking of the industry's interest in hydrocarbon-based renewables (see Section IV.B), we have decided to analyze a cellulosic biofuel standard made up of both cellulosic ethanol and cellulosic diesel fuels.

For assessing the impacts of the RFS2 standards, we used AEO 2009 (April release) cellulosic ethanol volumes (4.92 billion gallons), as well as the cellulosic biomass-to-liquids (BTL) diesel volumes (1.96 billion gallons) using Fischer-Tropsch (FT) processes. We consider BTL diesel from FT processes as a subset of cellulosic diesel. In order to reach a total of 16 billion ethanol-equivalent gallons, we assumed that an additional 4.56 billion gallons of cellulosic diesel could be produced from other cellulosic diesel processes. Refer to Section 1.2 of the RIA for more discussion.

b. Biomass-Based Diesel

Biomass-based diesel can include fatty acid methyl ester (FAME) biodiesel, renewable diesel (RD) that has not been co-processed with a petroleum feedstock, as well as cellulosic diesel. Although cellulosic diesel could potentially contribute to the biomass-based diesel category, we have assumed for our analyses that the fuel produced through Fischer-Tropsch (F-T) or other processes and its corresponding feedstocks (cellulosic biomass) are already accounted for in the cellulosic biofuel category discussed previously in Section IV.A.2.a.

FAME and RD processes can both utilize vegetable oils, rendered fats, and greases, and thus will generally compete for the same feedstock pool. We have based RD volumes on our forecast of industry plans, and expect these plants to use rendered fats as feedstock. Most biodiesel plants now have the capability to use vegetable or animal fats as feedstock, and thus our analysis assumes biodiesel will be made from a mix of inputs, depending on local availability, economics, and season. Refer to Section 1.1 of the RIA for more detail on FAME and RD feedstocks

Renewable diesel production can be further classified as co-processed or non-co-processed, depending on whether the renewable material is mixed with petroleum during the hydrotreating operations. EISA specifically forbids co-processed RD from being counted as biomass-based diesel, but it can still count toward the total advanced biofuel requirement. At this time, based on current industry plans, we expect most, if not all, RD will be non-co-processed (that is, non-refinery operations).

Perhaps the feedstock with the greatest potential for providing large volumes of oil for the production of biomass-based diesel is algae. However, several technical hurdles do still exist. Specifically, more efficient harvesting, dewatering, and lipid extraction methods are needed to lower costs to a level competitive with other feedstocks. For all three control cases, we have chosen to include 100 million gallons of algae-based biodiesel by 2022. We believe this is reasonable given several announcements from the algae industry about their production plans. [40] Although algae to biofuel companies can focus on producing algae oil for traditional biodiesel production, several companies are alternatively using algae for producing ethanol or crude oil for gasoline or diesel which could also help contribute to the advanced biofuel mandate. For more detail on algae as a feedstock, refer to Section 1.1 of the RIA.

During the comment period, we received information from stakeholders on alternative biodiesel feedstocks such as camelina and pennycress, to name a few. These feedstocks are currently being researched due to their potential for lower agricultural inputs and higher oil yields than traditional vegetable oil feedstocks as well as their use in additional crop rotations (i.e., winter cover crops) on a given area of land. We acknowledge that as we learn more about the challenges and benefits to the use of newer feedstocks, these could be used in the future towards meeting the biomass-based diesel standard under the RFS2 program provided they meet the lifecycle thresholds and definitions for renewable biomass. For the purpose of our impacts analysis, however, we have chosen not to include these feedstocks in our analyses at this time.

c. Other Advanced Biofuel

As defined in EISA, advanced biofuel includes the cellulosic biofuel and biomass-based diesel categories that were mentioned in Sections IV.A.2.a and IV.A.2.b above. However, EISA requires greater volumes of advanced biofuel than just the volumes required of these fuels. It is entirely possible that greater volumes of cellulosic biofuel and biomass-based diesel than required by EISA could be produced in the future. Our control case assumes that the cellulosic biofuel volumes will not exceed those required under EISA. We do assume, however, that additional biodiesel than that needed to meet the biomass-based diesel volume will be used to meet the total advanced biofuel volume. Despite additional volumes assumed from biodiesel, to fully meet the total advanced biofuel volume required under EISA, other types of advanced biofuel are necessary through 2022.

We have assumed for our control case that the most likely sources of advanced fuel other than cellulosic biofuel and biomass-based diesel would be from imported sugarcane ethanol and perhaps limited amounts of co-processed renewable diesel. Our assessment of international fuel ethanol production and demand indicate that anywhere from 3.8-4.2 Bgal of sugarcane ethanol from Brazil could be available for export by 2020/2022. If this volume were to be made available to the U.S., then there would be sufficient volume to meet the advanced biofuel standard. To calculate the amount of imported ethanol needed to meet the EISA advanced biofuel standards, we assumed it would make up the difference not met by cellulosic biofuel, biomass-based diesel and additional biodiesel categories (see Table IV.A.2-1). The amount of imported ethanol required by 2022 is approximately 2.2 Bgal.

As discussed in the NPRM, other potential advanced biofuels could include for example, U.S. domestically produced sugarcane ethanol, biobutanol, and biogas. While we have not chosen to reflect these fuels in our control case, they can still be counted under our program assuming they meet the lifecycle thresholds and other definitions under the program.

d. Other Renewable Fuel

The remaining portion of total renewable fuel not met with advanced biofuel was assumed to come from corn-based ethanol (including small amounts from other grains and waste sugars). EISA effectively sets a limit for participation in the RFS program of 15 Bgal of corn ethanol, and we are assuming for our analysis that sufficient corn ethanol will be produced to meet the 15-Bgal limit that either meets the 20% GHG threshold or is grandfathered. It should be noted, however, that there is no specific “corn-ethanol” mandated volume, and that any advanced biofuel produced above and beyond what is required for the advanced biofuel requirements could reduce the amount of corn ethanol needed to meet the total renewable fuel standard. This occurs in our projections during the earlier years (2010-2015) in which we project that some fuels could compete favorably with corn ethanol (e.g., biodiesel and imported ethanol). Refer to Section 1.2 of the RIA for more details on interim years. Beginning around 2016, fuels qualifying as advanced biofuels likely will be devoted to meeting the increasingly stringent volume mandates for advanced biofuel. It is also important to note that more than 15 Bgal of corn ethanol could be produced and RINs generated for that volume under the RFS2 regulations. However, obligated parties would not be required to purchase more than 15 Bgal worth of non-advanced biofuel RINs, e.g. corn ethanol RINs.

3. Additional Control Cases Considered

Since there is significant uncertainty surrounding what fuels will be produced to meet the 16 billion gallon cellulosic biofuel standard, we have decided to investigate two other sensitivity cases for our cost and emission impact analyses conducted for the rule. The first case, we refer to as the “low-ethanol” control case and assume only 250 million gallons of cellulosic ethanol (from AEO 2007 reference case). The rest of the 16 billion gallon cellulosic biofuel standard is made up of cellulosic diesel as shown in Table IV.A.3-1. The second case, we refer to as the “high-ethanol” control case and assume the entire 16 billion gallon cellulosic biofuel standard is met with cellulosic ethanol, also shown in Table IV.A.3-1.

Table IV.A.3-1—Control Case Projected Renewable Fuel Volumes in 2022 Back to Top
Case/volume type Advanced biofuel Non-advanced biofuel Total renewable fuel
Cellulosic biofuel Biomass-based diesela Other advanced biofuel Corn ethanol
Cellulosic ethanol Cellulosic dieselb FAMEcbiodiesel NCRDd Other biodiesele Imported ethanol
[Billion gallons]
aBiomass-Based Diesel could include FAME biodiesel, cellulosic diesel, and non-co-processed renewable diesel.
bCellulosic Diesel includes 1.96 billion gallons (3.33 ethanol-equivalent billion gallons) from Fischer-Tropsch Biomass-to-Liquids (BTL) processes and 7.30 billion gallons (12.42 ethanol-equivalent billion gallons) from other types of cellulosic diesel processes for the Low-Ethanol case and zero cellulosic diesel in the High-Ethanol Case.
cFatty acid methyl ester (FAME) biodiesel.
dNon-Co-processed Renewable Diesel (NCRD).
eOther Biodiesel is biodiesel that could be produced in addition to the amount needed to meet the biomass-based diesel standard.
Low-Ethanol Straight Volume 0.25 9.26 0.85 0.15 0.82 2.24 15.00 28.57
Low-Ethanol Ethanol-Equivalent 0.25 15.75 1.28 0.26 1.23 2.24 15.00 36.00
High-Ethanol Straight Volume 16.00 0.00 0.85 0.15 0.82 2.24 15.00 35.06
High-Ethanol Ethanol-Equivalent 16.00 0.00 1.28 0.26 1.23 2.24 15.00 36.00

In comparison, our primary control case described in Section IV.A.2, could be considered a “mid-ethanol” control case, as the cellulosic ethanol and diesel volumes analyzed are in between the low-ethanol and high-ethanol cases described in this section. We believe the addition of these sensitivity cases is useful in understanding the potential impacts of the renewable fuels standards. Refer to Section 1.2 of the RIA for more detail on three control cases analyzed as part of this rule.

B. Renewable Fuel Production

1. Corn/Starch Ethanol

The majority of domestic biofuel production currently comes from plants processing corn and other similarly processed grains in the Midwest. However, there are a handful of plants located outside the Corn Belt and a few plants processing simple sugars from food or beverage waste. In this section, we summarize the present state of the corn/starch ethanol industry and discuss how we expect things to change in the future under the RFS2 program.

a. Historic/Current Production

The United States is currently the largest ethanol producer in the world. In 2008, the U.S. produced nine billion gallons of fuel ethanol for domestic consumption, the majority of which came from locally grown corn. [41] The nation is currently on track for producing over 10 billion gallons by the end of 2009. [42] Although the U.S. ethanol industry has been in existence since the 1970s, it has rapidly expanded in recent years due to the phase-out of methyl tertiary butyl ether (MTBE), elevated crude oil prices, state mandates and tax incentives, the introduction of the Federal Volume Ethanol Excise Tax Credit (VEETC), [43] the implementation of the existing RFS1 program, [44] and the new volume requirements established under EISA. As shown in Figure IV.B.1-1, U.S. ethanol production has grown exponentially over the past decade.

Asof November 2009 there were 180 corn/starch ethanol plants operating in the U.S. with a combined production capacity of approximately 12 billion gallons per year. [46] This does not include idled ethanol plants, discussed later in this subsection. The majority of today's ethanol production (91.5% by volume) comes from 155 plants relying exclusively on corn. Another 8.3% comes from 18 plants processing a blend of corn and/or similarly processed grains (milo, wheat, or barley). The remainder comes from seven small plants processing waste beverages or other waste sugars and starches.

Of the 173 plants processing corn and/or other similarly processed grains, 162 utilize dry-milling technologies and the remaining 11 plants rely on wet-milling processes. Dry mill ethanol plants grind the entire kernel and generally produce only one primary co-product: distillers' grains with solubles (DGS). The co-product is sold wet (WDGS) or dried (DDGS) to the agricultural market as animal feed. However, there are a growing number of plants using front-end fractionation to produce food-grade corn oil or back-end extraction to produce fuel-grade corn oil for the biodiesel industry. A company called GreenShift has corn oil extraction facilities located at five ethanol plants in Michigan, Indiana, New York and Wisconsin. [47] Collectively, these facilities are designed to extract in excess of 7.3 million gallons of corn oil per year. Primafuel Solutions is another company offering corn oil extraction technologies to make existing ethanol plants more sustainable. For more information on corn oil extraction and other advanced technologies being pursued by today's corn ethanol industry, refer to Section 1.4.1 of the RIA.

In contrast to dry mill plants, wet mill facilities separate the kernel prior to processing into its component parts (germ, fiber, protein, and starch) and in turn produce other co-products (usually gluten feed, gluten meal, and food-grade corn oil) in addition to DGS. Wet mill plants are generally more costly to build but are larger in size on average. [48] As such, 11.4% of the current grain ethanol production comes from the 11 previously mentioned wet mill facilities.

The remaining seven ethanol plants process waste beverages or waste sugars/starches and operate differently than their grain-based counterparts. These small production facilities do not require milling and operate simpler enzymatic fermentation processes.

Ethanol production is a relatively resource-intensive process that requires the use of water, electricity, and steam. Steam needed to heat the process is generally produced on-site or by other dedicated boilers. [49] The ethanol industry relies primarily on natural gas. Of today's 180 ethanol production facilities, an estimated 151 burn natural gas [50] (exclusively), three burn a combination of natural gas and biomass, one burns natural gas and coal (although natural gas is the primary fuel), one burns a combination of natural gas, landfill biogas and wood, and two burn natural gas and syrup from the process. We are aware of 17 plants that burn coal as their primary fuel and one that burns a combination of coal and biomass. [51] Our research suggests that three corn ethanol plants rely on a combination of waste heat and natural gas and one plant does not have a boiler and relies solely on waste heat from a nearby power plant. Overall, our research suggests that 27 plants currently utilize cogeneration or combined heat and power (CHP) technology, although others may exist. [52] CHP is a mechanism for improving overall plant efficiency. Whether owned by the ethanol facility, their local utility, or a third party, CHP facilities produce their own electricity and use the waste heat from power production for process steam, reducing the energy intensity of ethanol production. [53]

During the ethanol fermentation process, large amounts of carbon dioxide (CO 2) gas are released. In some plants the CO 2 is vented into the atmosphere, but where local markets exist, it is captured, purified, and sold to the food processing industry for use in carbonated beverages and flash-freezing applications. We are currently aware of 40 fuel ethanol plants that recover CO 2 or have facilities in place to do so. According to Airgas, a leading gas distributor, the U.S. ethanol industry currently recovers 2 to 2.5 million tons of CO 2 per year which translates to about 5-7% of all the CO 2 produced by the industry. [54]

Since the majority of ethanol is made from corn, it is no surprise that most of the plants are located in the Midwest near the Corn Belt. Of today's 180 ethanol production facilities, 163 are located in the 15 states comprising PADD 2. For a map of the government's Petroleum Administration for Defense Districts or PADDs, refer to Figure IV.B.1-2.

As a region, PADD 2 accounts for over 94% (or 11.3 billion gallons) of today's estimated ethanol production capacity, followed by PADD 3 (2.4%), PADDs 4 and 1 (each with 1.3%) and PADD 5 (0.8%). For more information on today's ethanol plant locations, refer to Section 1.5.1 of the RIA.

The U.S. ethanol industry is currently comprised of a mixture of company-owned plants and locally-owned farmer cooperatives (co-ops). The majority of today's ethanol production facilities are company-owned, and on average these plants are larger in size than farmer-owned co-ops. Accordingly, these facilities account for about 80% of today's online ethanol production capacity. [55] Furthermore, nearly 30% of the total domestic product comes from 40 plants owned by just three different companies—POET Biorefining, Archer Daniels Midland (ADM), and Valero Renewables. Valero entered the ethanol industry in March of 2009 when it acquired seven ethanol plants from former ethanol giant, Verasun. The oil company currently has agreements in place to purchase three more ethanol plants that would bring the company's ethanol production capacity to 1.1 billion gallons per year. [56] However, ethanol plants are much smaller than petroleum refineries. Valero's smallest petroleum refinery in Ardmore, OK has about twice the throughput of all its ethanol plants combined. [57] Still, as obligated parties under RFS1 and RFS2, the refining industry continues to show increased interest in biofuels. Suncor and Murphy Oil recently joined Valero as the second and third oil companies to purchase idled U.S. ethanol plants. Many refiners are also supporting the development of cellulosic biofuels and algae-based biodiesel.

b. Forecasted Production Under RFS2

As highlighted earlier, domestic ethanol production is projected to grow to over 10 billion gallons in 2009. And with over 12 billion gallons of capacity online as of November 2009, ethanol production should continue to grow in 2010, provided plants continue to produce at or above today's production levels. In addition, despite current market conditions (i.e., poor ethanol margins), the ethanol industry is expected to grow in the future under the RFS2 program. Although there is not a set corn ethanol requirement, EISA allows for 15 billion gallons of the 36-billion gallon renewable fuel standard to be met by conventional biofuels. We expect that corn ethanol will fulfill this requirement, provided it is more cost competitive than imported ethanol or cellulosic biofuel in the marketplace.

In addition to the 180 aforementioned corn/starch ethanol plants currently online, 27 plants are presently idled. [58] Some of these are smaller ethanol plants that have been idled for quite some time, whereas others are in a more temporary “hot idle” mode, ready to be restarted. In response to the economic downturn, a number of ethanol producers have idled production, halted construction projects, sold off plants and even filed for Chapter 11 bankruptcy protection. Some corn ethanol companies have exited the industry all together (e.g., Verasun) whereas others are using bankruptcy as a means to protect themselves from creditors as they restructure their finances with the goal of becoming sustainable.

Crude oil prices are expected to increase in the future making corn ethanol more economically viable. According to EIA's AEO 2009, crude oil prices are projected to increase from about $80/barrel (today's price) to $116/barrel by 2022. [59] As oil and gas prices rebound, we expect that the biofuels industry will as well. Since our April 2009 industry assessment used for the NPRM, at least nine corn ethanol plants have come back online.

For analysis purposes, we assumed that all 27 idled corn/starch ethanol plants would resume operations by 2022 under the RFS2 program. We also assumed that a total of 11 new ethanol plants and two expansion projects currently under construction or in advanced stages of planning would come online. [60] This includes two large dry mill expansion projects currently underway at existing ADM wet mill plants and two planned combination corn/cellulosic ethanol plants that received funding from DOE. While several of these projects are delayed or on hold at the moment, we expect that these facilities (or comparable replacement projects) would eventually come online to get the nation to approximately 15 billion gallons of corn ethanol production capacity.

Almost 100% of conventional ethanol plant growth is expected to come from facilities processing corn or other similarly processed grains. And not surprisingly, the majority of growth (approximately 70% by volume) is expected to originate from PADD 2. However, growth is expected to occur in all PADDs. With the exception of one facility, [61] all new corn/grain ethanol plants are expected to utilize dry milling technologies and the majority of new production is expected to come from plants burning natural gas. However, we anticipate that two manure biogas plants, [62] one biomass-fired plant, and two coal-fired ethanol plants will be added to the mix. [63] Of these new and returning idled plants, we're aware of five facilities currently planning to use CHP technology, bringing the U.S. total to 32.

The above predictions are based on the industry's current near-term production plans. However, we anticipate additional growth in advanced ethanol production technologies under the RFS2 program. Forecasted fuel prices are projected to drive corn ethanol producers to transition from conventional boiler fuels to biomass feedstocks. In addition, fossil fuel/electricity prices will likely drive a number of ethanol producers to pursue CHP technology. For more on our projected 2022 utilization of these technologies under the RFS2 program, refer to Section 1.5.1.3 of the RIA.

2. Imported Ethanol

As discussed in the proposal, ethanol imports have traditionally played a relatively small role in the U.S. transportation fuel market due to historically low crude prices and the tariff on imported ethanol. Between years 2000 and 2008, the volume of ethanol imported into the U.S. has ranged from 46-720 million gallons per year. So far this year, from January through November 2009, imported ethanol has only reached 197 million gallons. [64] As the data show, the volume of imported ethanol can fluctuate greatly.

In the past, the majority of volume has originated from countries that are part of the Caribbean Basin Initiative. Direct Brazilian imports have also made up a sizeable portion of total ethanol imported into the U.S. However, recently there have been relatively small amounts of direct imports of ethanol from Brazil. [65] This indicates that current market conditions have made importing Brazilian ethanol directly to the U.S. uneconomical. Part of the reason for this decline in imports is the cessation of the duty drawback that became effective on October 1, 2008, but also changes in world sugar prices. [66]

It is difficult to project the potential volume of future ethanol imports to the U.S. based purely on historical data. Rather, it is necessary to assess future import potential by analyzing the major players for foreign ethanol production and consumption. In 2008, the top three fuel ethanol producers were the U.S., Brazil, and the European Union (EU), producing 9.0, 6.5, and 0.7 billion gallons, respectively. [67] Consumption of fuel ethanol is also dominated by the United States and Brazil with approximately 9.6 and 4.9 billion gallons consumed in each country, respectively. 68 69 The EU consumed approximately 0.9 billion gallons of fuel ethanol in 2008. [70]

In our assessment of foreign ethanol production and consumption, we analyzed the following countries or group of countries: Brazil, the EU, Japan, India, and China. Our analyses indicate that Brazil would likely be the only nation able to supply any meaningful amount of ethanol to the U.S. in the future. Depending on whether the mandates and goals of the EU, Japan, India, and China are enacted or met in the future, it is likely that this group of countries would consume any growth in their own production and be net importers of ethanol, thus competing with the U.S. for Brazilian ethanol exports.

Due to uncertainties in the future demand for ethanol domestically and internationally, uncertainties in the actual investments made in the Brazilian ethanol industry, as well as uncertainties in future sugar prices, there appears to be a wide range of Brazilian production and domestic consumption estimates. The most current and complete estimates indicate that total Brazilian ethanol exports will likely reach 3.8-4.2 billion gallons by 2022. 71 72 73 As this volume of ethanol export is available to countries around the world, only a portion of this will be available exclusively to the United States. If the balance of the EISA advanced biofuel requirement not met with cellulosic biofuel and biomass-based diesel were to be met with imported sugarcane ethanol alone, it would require about 2.2 billion gallons (see Table IV.A.2-1), or approximately 55% of total Brazilian ethanol export estimates. This is aggressive, yet within the bounds of reason, therefore, we have made this simplifying assumption for the purposes of further analysis.

Generally speaking, Brazilian ethanol exporters will seek routes to countries with the lowest costs for transportation, taxes, and tariffs. With respect to the U.S., the most likely route is through the Caribbean Basin Initiative (CBI). [74] Brazilian ethanol entering the U.S. through CBI countries is not currently subject to the 54 cent/gal imported ethanol tariff and yet receives the 45 cent/gal ethanol blender credit. In addition to the U.S., other countries also have similar tariffs on imported ethanol. Refer to Section 1.5.2 of the RIA for more details. Due to the economic incentive of transporting ethanol through the CBI, we expect the majority of the tariff rate quota (TRQ) to be met or exceeded, perhaps 90% or more. The TRQ is set each year as 7% of the total domestic ethanol consumed in the prior year. If we assume that 90% of the TRQ is met and that total domestic ethanol (corn and cellulosic ethanol) consumed in 2021 was 19.2 Bgal (under the primary control case), then approximately 1.21 Bgal of ethanol could enter the U.S. through CBI countries in 2022. The rest of the Brazilian ethanol exports not entering the CBI will compete on the open market with the rest of the world demanding some portion of direct Brazilian ethanol. To meet our advanced biofuel standard, we assumed 1.03 Bgal of sugarcane ethanol would be imported directly to the U.S. in 2022.

3. Cellulosic Biofuel

The majority of the biofuel currently produced in the United States comes from plants processing first-generation feedstocks like corn, plant oils, sugarcane, etc. Non-edible cellulosic feedstocks have the potential to greatly expand biofuel production, both volumetrically and geographically. Research and development on cellulosic biofuel technologies has exploded over the last few years, and plants to commercialize a number of these technologies are already beginning to materialize. The $1.01/gallon tax credit for cellulosic biofuel that was introduced in the 2008 Farm Bill and recently became effective, is also offering much incentive to this developing industry. In addition to today's RFS2 program which sets aggressive goals for cellulosic biofuel production, the Department of Energy (DOE), Department of Agriculture (USDA), Department of Defense (DOD) and state agencies are helping to spur industry growth.

a. Current State of the Industry

There are a growing number of biofuel producers, biotechnology companies, universities and research institutes, start-up companies as well as refiners investigating cellulosic biofuel production. The industry is currently pursuing a wide range of feedstocks, conversion technologies and fuels. There is much optimism surrounding the long-term viability of cellulosic ethanol and other alcohols for gasoline blending. There is also great promise and growing interest in synthetic hydrocarbons like gasoline, diesel and jet fuel as “drop in” petroleum replacements. Some companies intend to start by processing corn or sugarcane and then transition to cellulosic feedstocks while others are focusing entirely on cellulosic materials. Regardless, cellulosic biofuel production is beginning to materialize.

We are currently aware of over 35 small pilot- and demonstration-level plants operating in North America. However, the main focus at these facilities is research and development, not commercial production. Most of the plants are rated at less than 250,000 gallons per year and that's if they were operated at capacity. Most only operate intermittently for the purpose of demonstrating that the technologies can be used to produce transportation fuels. The industry as a whole is still working to increase efficiency, improve yields, reduce costs and prove to the public, as well as investors, that cellulosic biofuel is both technologically and economically feasible.

As mentioned above, a variety of feedstocks are being investigated for cellulosic biofuel production. There is a great deal of interest in urban waste (MSW and C&D debris) because it is virtually free and abundant in many parts of the country, including large metropolitan areas where the bulk of fuel is consumed. There is also a lot of interest in agricultural residues (corn stover, rice and other cereal straws) and wood (forest thinnings, wood chips, pulp and paper mill waste and yard waste). However, researchers are still working to find viable harvesting and storage solutions. Others are investigating the possibility of growing dedicated energy crops for cellulosic biofuel production, e.g., switchgrass, energy cane, sorghum, poplar, miscanthus and other fast-growing trees. While these crops have tremendous potential, many are starting with the feedstocks that are available today with the mentality that once the industry has proven itself, it will be easier to secure growing contracts and start producing energy crops. For more information on cellulosic feedstock availability, refer to preamble Section IV.B.3.d and Section 1.1.2 of the RIA.

The industry is also pursuing a number of different cellulosic conversion technologies and biofuels. Most of the technologies fall into one of two categories: biochemical or thermochemical. Biochemical conversion involves the use of acids and/or enzymes to hydrolyze cellulosic materials into fermentable sugars and lignin. Thermochemical conversion involves the use of heat to convert biomass into synthesis gas or pyrolysis oil for upgrading. A third technology pathway is emerging that involves the use of catalysts to depolymerize or reform the feedstocks into fuel. The technologies currently being considered are capable of producing cellulosic alcohols or hydrocarbons for the transportation fuel market. Many companies are also researching the potential of co-firing biomass to produce plant energy in addition to biofuels. For a more in-depth discussion on cellulosic technologies, refer to Section 1.4.3 of the RIA.

b. Setting the 2010 Cellulosic Biofuel Standard

The Energy Independence and Security Act (EISA) set aggressive cellulosic biofuel targets beginning with 100 million gallons in 2010. However, EISA also supplied EPA with cellulosic biofuel waiver authority. For any calendar year in which the projected cellulosic biofuel production is less than the minimum applicable volume, EPA can reduce the standard based on the volume expected to be available that year. EPA is required to set the annual cellulosic standard by November 30th each year and should consider the annual estimate made by EIA by October 31st of each year. We are setting the 2010 standard as part of this final rule.

Setting the cellulosic biofuel standard for 2010 represents a unique challenge. As discussed above, the industry is currently characterized by a wide range of companies mostly focused on research, development, demonstration, and financing their developing technologies. In addition, while we are finalizing a requirement that producers and importers of renewable fuel provide us with production outlook reports detailing future supply estimates (refer to § 80.1449), we do not have the benefit of this valuable cellulosic supply information for setting the 2010 standard. Finally, since today's cellulosic biofuel production potential is relatively small, and the number of potential producers few (as described in more detail below), the overall volume for 2010 can be heavily influenced by new developments, either positive or negative associated with even a single company, which can be very difficult to predict. This is evidenced by the magnitude of changes in cellulosic biofuel projections and the potential suppliers of these fuels since the proposal.

In the proposal, we did a preliminary assessment of the cellulosic biofuel industry to arrive at the conclusion that it was possible to uphold the 100 million gallon standard in 2010 based on anticipated production. At the time of our April 2009 NPRM assessment, we were aware of a handful of small pilot and demonstration plants that could help meet the 2010 standard, but the largest volume contributions were expected to come from Cello Energy and Range Fuels.

Cello Energy had just started up a 20 million gallon per year (MGY) cellulosic diesel plant in Bay Minette, AL. EPA staff visited the facility twice in 2009 to confirm that the first-of-its-kind commercial plant was mechanically complete and poised to produce cellulosic biofuel. It was assumed that start-up operations would go as planned and that the facility would be operating at full capacity by the end of 2009 and that three more 50 MGY cellulosic diesel plants planned for the Southeast could be brought online by the end of 2010.

At the time of our assessment, we were also anticipating cellulosic biofuel production from Range Fuels' first commercial-scale plant in Soperton, GA. The company received a $76 million grant from DOE to help build a 40 MGY wood-based ethanol plant and they broke ground in November 2007. In January 2009, Range was awarded an $80 million loan guarantee from USDA. [75] With the addition of this latest capital, the company seemed well on its way to completing construction of its first 10 MGY phase by the end of 2009 and beginning production in 2010.

Since our April 2009 industry assessment there have been a number of changes and delays in production plans due to technological, contractual, financial and other reasons. Cello Energy and Range Fuels have delayed or reduced their production plans for 2010. Some of the small plants expected to come online in 2010 have pushed back production to the 2011-2012 timeframe, e.g., Clearfuels Technology, Fulcrum River Biofuels, and ZeaChem. Alltech/Ecofin and RSE Pulp & Chemical, two companies that were awarded DOE funding back in 2008 to build small-scale biorefineries appear to be permanently on hold or off the table. In addition, Bell Bio-Energy, a company that received DOD funding has since abandoned plans to produce cellulosic diesel from MSW at U.S. military bases. [76]

At the same time, there has also been an explosion of new companies, new business relationships, and new advances in the cellulosic biofuel industry. Keeping track of all of them is a challenge in and of it self as the situation can change on a daily basis. EIA recently provided EPA with their first cellulosic biofuel supply estimate required under CAA section 211(o)(7)(D)(i). In a letter to the Administrator dated October 29, 2009, they arrived at a 5.04 million gallon estimate for 2010 based on publicly available information and assumptions made with respect production capacity utilization. [77] A summary of the plants they considered is shown below in Table IV.B.3-1.

Table IV.B.3-1—EIA's Projected Cellulosic Biofuel Plant Production Capacities for 2010 Back to Top
Online Company Location Product Capacity (million gallons) Expected utilization (%) Production (million gallons)3
Notes: 1. Cello Energy is assigned a 10-percent utilization factor as they have not been able to run on a continuous basis long enough to apply for a Synthetic Minor Operating Permit or produce significant amounts of fuel during 2009. 2. It is estimated that only half the 2010 projected capacity (10 million gallons per year) will be a qualified fuel. 3. The production from these facilities in 2009 is not surveyed by EIA or EPA.
2007 KL Process Design Upton, WY Ethanol 1.5 10 0.15
2008 Verenium Jennings, LA Ethanol 1.4 10 0.14
2008 Terrabon Bryan, TX Bio-Crude 0.93 10 0.09
2010 Zeachem Boardman, OR Ethanol 1.5 10 0.15
2010 Cello Energy Bay Minette, AL Diesel 20.0 101 2.00
2010 Range Fuels Soperton, GA Ethanol 5.02 50 2.5
Total 30.35 5.04

In addition to receiving EIA's information and coordinating with them and other offices in DOE, we have initiated meetings and conversations with over 30 up-and-coming advanced biofuel companies to verify publicly available information, obtain confidential business information, and better assess the near-term cellulosic biofuel production potential for use in setting the 2010 standard. What we have found is that the cellulosic biofuel landscape has continued to evolve. Based on information obtained, not only do we project significantly different production volumes on a company-by-company basis, but the list of potential producers of cellulosic biofuel in 2010 is also significantly different than that identified by EIA.

Overall, our industry assessment suggests that it is difficult to rely on commercial production from small pilot or demonstration-level plants. The primary purpose of these facilities is to prove that a technology works and demonstrate to investors that the process is capable of being scaled up to support a larger commercial plant. Small plants are cheaper to build to demonstrate technology than larger plants, but the operating costs ($/gal) are higher due to their small scale. As a result, it's not economical for most of these facilities to operate continuously. Most of these plants are regularly shut down and restarted as needed as part of the research and development process. Due to their intermittent nature, most of these plants operate at a fraction of their rated capacity, some less than the 10% utilization rate assumed by EIA. In addition, few companies plan on making their biofuel available for commercial sale.

However, there are at least two cellulosic biofuel companies currently operating demonstration plants in the U.S. and Canada that could produce fuel commercially in 2010. The first is KL Energy Corporation, a company we considered for the NPRM with a 1.5 MGY cellulosic ethanol plant in Upton, WY. This plant was considered by EIA and is included in Table IV.B.3-1. The second is Iogen's cellulosic ethanol plant in Ottawa, Canada with a 0.5 MGY capacity. Iogen's commercial demonstration plant was referenced by EIA as a potential foreign source for cellulosic biofuel but was not included in their final table. In addition to these online demonstration plants, there are three additional companies not on EIA's list that are currently building demonstration-level cellulosic biofuel plants in North America that are scheduled to come online in 2010. This includes DuPont Danisco Cellulosic Ethanol and Fiberight, companies building demonstration plants in the U.S. and Enerkem, a company building a demonstration plant in Canada. Cello Energy's plant in Bay Minette, AL continues to offer additional potential for cellulosic biofuel in 2010. And finally, Dynamotive, a company that currently has two biomass-based pyrolysis oil production plants in Canada is another potential source of cellulosic biofuel in 2010. All seven aforementioned companies are discussed in greater detail below along with Range Fuels.

KL Energy Corporation (KL Energy), through its majority-owned Western Biomass Energy, LLC (WBE) located in Upton, WY, is designed to convert wood products and wood waste products into ethanol. Since the end of construction in September 2007, equipment commissioning and process revisions continued until the October 2009 startup. The plant was built as a 1.5 MGY demonstration plant and was designed to both facilitate research and operate commercially. It is KL Energy's intent that WBE's future use will involve the production and sale of small but commercial-quality volumes of ethanol and lignin co-product. The company's current 2010 goal is for WBE to generate RINs under the RFS2 program. [78]

Iogen is responsible for opening the first commercial demonstration cellulosic ethanol plant in North America. Iogen's plant located in Ottawa, Canada has been producing cellulosic ethanol from wheat straw since 2004. Like KL Energy, Iogen has slowly been ramping up production at its 0.5 MGY plant. According to the company's Web site, they produced approximately 24,000 gallons in 2004 and 34,000 gallons in 2005. Production dropped dramatically in 2006 and 2007 but came back strong with 55,000 gallons in 2008. Iogen recently produced over 150,000 gallons of ethanol from the demonstration plant in 2009. Iogen also recently became the first cellulosic ethanol producer to sell its advanced biofuel at a retail service station in Canada. Their cellulosic ethanol was blended to make E10 available for sale to consumers at an Ottawa Shell station. Iogen also recently announced plans to build its first commercial scale plant in Prince Albert, Saskatchewan in the 2011/2012 timeframe. Based on the company's location and operating status, Iogen certainly has the potential to participate in the RFS2 program. However, at this time, we are not expecting them to import any cellulosic ethanol into the U.S. in 2010. [79]

DuPont Danisco Cellulosic Ethanol, LLC (DDCE), a joint venture between DuPont and Danisco, is another potential source for cellulosic biofuel in 2010. DDCE received funding from the State of Tennessee and the University of Tennessee to build a small 0.25 MGY demonstration plant in Vonore, TN to pursue switchgrass-to-ethanol production. According to DDCE, construction commenced in October 2008 and the plant is now mechanically complete and undergoing start-up operations. The facility is scheduled to come online by the end of January and the company hopes to operate at or around 50% of production capacity in 2010. According to the DDCE, the objective in Vonore is to validate processes and data for commercial scale-up, not to make profits. However, the company does plan to sell the cellulosic ethanol it produces. [80]

Enerkem is another company pursuing cellulosic ethanol production. The Canadian-based company was recently announced as a recipient of a joint $50 million grant from DOE and USDA to build a 10 MGY woody biomass-to-ethanol plant in Pontotoc, MS. [81] The U.S. plant is not scheduled to come online until 2012, but Enerkem is currently building a 1.3 MGY demonstration plant in Westbury, Quebec. According to the company, plant construction in Westbury started in October 2007 and the facility is currently scheduled to come online around the middle of 2010. While it's unclear at this time whether the cellulosic ethanol produced will be exported to the United States, Enerkem has expressed interest in selling its fuel commercially. [82]

Additional cellulosic biofuel could come from Fiberight, LLC (Fiberight) in 2010. We recently became aware of this start-up company and contacted them to learn more about their process and cellulosic biofuel production plans. According to Fiberight, they have been operating a pilot-scale facility in Lawrenceville, VA for three years. They have developed a proprietary process that not only fractionates MSW but biologically converts the non-recyclable portion into cellulosic ethanol and biochemicals. Fiberight recently purchased a shut down corn ethanol plant in Blairstown, IA and plans to convert it to become MSW-to-ethanol capable. According to the company, construction is currently underway and the goal is to bring the 2 MGY demonstration plant online by February or March, 2010. If the plant starts up according to plan, the company intends on making cellulosic ethanol commercially available in 2010 and generating RINS under the RFS2 program. Fiberight's long-term goal is to expand the Blairstown plant to a 5-8 MGY capacity and build other small commercial plants around the country that could convert MSW into fuel. [83]

Cello Energy, a company considered in the proposal, continues to be another viable source for cellulosic biofuel in 2010. Despite recent legal issues which have constrained the company's capital, Cello Energy is still pursuing cellulosic diesel production. According to the company, they are currently working to resolve materials handling and processing issues that surfaced when they attempted to scale up production to 20 MGY from a previously operated demonstration plant. As of November 2009, they were waiting for new equipment to be ordered and installed which they hoped would allow for operations to be restarted as early as February or March, 2010. Cello's other planned commercial facilities are currently on hold until the Bay Minette plant is operational. [84]

Another potential supplier of cellulosic biofuel is Dynamotive Energy Systems (Dynamotive) headquartered in Vancouver, Canada. Dynamotive currently has two plants in West Lorne and Guelph, Ontario that produce biomass-based pyrolysis oil (also known as “BioOil”) for industrial applications. The BioOil production capacity between the two plants is estimated at around 9 MGY, but both plants are currently operating at a fraction of their rated capacity. [85] However, according to a recent press release, Dynamotive has contracts in place to supply a U.S.-based client with at least nine shipments of BioOil in 2010. If Dynamotive's BioOil is used as heating oil or upgraded to transportation fuel, it could potentially count towards meeting the cellulosic biofuel standard in 2010.

As for the Range Fuels plant, construction of phase one in Soperton, GA is about 85% complete, with start-up planned for mid-2010. However, there have been some changes to the scope of the project that will limit the amount of cellulosic biofuel that can be produced in 2010. The initial capacity has been reduced from 10 to 4 million gallons per year. In addition, since they plan to start up the plant using a methanol catalyst they are not expected to produce qualifying renewable fuel in 2010. During phase two of their project, currently slated for mid-2012, Range plans to expand production at the Soperton plant and transition from a methanol to a mixed alcohol catalyst. This will allow for a greater alcohol production potential as well as a greater cellulosic biofuel production potential. [86]

Overall, our most recent industry assessment suggests that there could potentially be over 30 MGY of cellulosic biofuel production capacity online by the end of 2010. [87] However, since most of the plants are still under construction today, the amount of cellulosic biofuel produced in 2010 will be contingent upon when and if these plants come online and whether the projects get delayed due to funding or other reasons. In addition, based on our discussions with the developing industry, it is clear that we cannot count on demonstration plants to produce at or near capacity in 2010, or in their first few years of operation for that matter. The amount of cellulosic biofuel actually realized will depend on whether the process works, the efficiency of the process, and how regularly the plant is run. As mentioned earlier, most small plants, including commercial demonstration plants, are not operated continuously. As such, we cannot base the standard on these plants running at capacity—at least until the industry develops further and proves that such rates are achievable. We currently estimate that production from first-of-its kind plants could be somewhere in the 25-50% range in 2010. Together, the implementation timelines and anticipated production levels of the plants described above brings the cellulosic biofuel supply estimate to somewhere in the 6-13 million gallon range for 2010.

In addition, it is unclear how much we can rely on Canadian plants for cellulosic biofuel in 2010. Although we currently receive some conventional biofuel imports from Canada and many of the aforementioned Canadian companies have U.S. markets in mind, the country also has its own renewable fuel initiatives that could keep much of the cellulosic biofuel produced from coming to the United States, e.g., Iogen. Finally, it's unclear whether all fuel produced by these facilities will qualify as cellulosic biofuel under the RFS2 program. Several of the companies are producing fuels or using feedstocks which may not in fact qualify as cellulosic biofuel once we receive their detailed registration information. Factoring in these considerations, the cellulosic biofuel potential from the six more likely companies described above could result in several different production scenarios in the neighborhood of the recent EIA estimate. We believe this estimate of 5 million gallons or 6.5 ethanol-equivalent million gallons represents a reasonable yet achievable level for the cellulosic biofuel standard in 2010 considering the degree of uncertainty involved with setting the standard for the first year. As mentioned earlier, we believe standard setting will be easier in future years once the industry matures, we start receiving production outlook reports and there is less uncertainty regarding feasibility of cellulosic biofuel production.

c. Current Production Outlook for 2011 and Beyond

Since the proposal, we have also learned about a number of other cellulosic biofuel projects in addition to those described above. This includes commercial U.S. production plans by Coskata, Enerkem and Vercipia. However, production isn't slated to begin until 2011 or later and the same is true for most of the other larger plants we're aware of that are currently under development. Nonetheless, while cellulosic biofuel production in 2010 may be limited, it is remarkable how much progress the industry has made in such a short time, and there is a tremendous growth opportunity for cellulosic biofuels over the next several years.

Most of the cellulosic biofuel companies we've talked to are in different stages of proving their technologies. Regardless of where they are at, many have fallen behind their original commercialization schedules. As with any new technology, there have been delays associated with scaling up capacity, i.e., bugs to work out going from pilot to demonstration to commercialization. However, most are saying it's not the technologies that are delaying commercialization, it is lack of available funding. Obtaining capital has been very challenging given the current recession and the banking sector's financial difficulties. This is especially true for start-up companies that do not have access to capital through existing investors, plant profits, etc. From what we understand, banks are looking for cellulosic companies to be able to show that their plants are easily “scalable” or expandable to commercial size. Many are only considering companies that have built plants to one-tenth of commercial scale and have logged many hours of continuous operation.

The government is currently trying to help in this area. To date, the Department of Energy (DOE) and the Department of Agriculture (USDA) have allocated over $720 million in federal funding to help build pilot and demonstration-scale biorefineries employing advanced technologies in the United States. The largest installment from Recovery Act funding was recently announced on December 4, 2009 and includes funding for a series of larger commercial demonstration plants including cellulosic ethanol projects by Enerkem and INEOS New Planet BioEnergy, LLC. DOE has also issued grants to help fund some of the first commercial cellulosic biofuel plants. Current recipients include Abengoa Bioenergy, BlueFire Ethanol [88] and POET Biorefining in addition to Range Fuels. DOE and USDA are also issuing loan guarantees to help support the up-and-coming cellulosic biofuels industry and funding research and development. Many states are also providing assistance. For more information on government support for biofuels, refer to Section 1.5.3.3 of the RIA.

The refining industry is also helping to fund cellulosic biofuel R&D efforts and some of the first commercial plants. Many of the major oil companies have invested in advanced second-generation biofuels over the past 12-18 months. A few refiners (e.g., BP and Shell) have even entered into joint ventures to become cellulosic biofuel producers. General Motors and other vehicle/engine manufacturers are also providing financial support to help with research and development.

A summary of some of the cellulosic biofuel companies with near-term commercialization plans in North America is provided in Table IV.B.3-2. The capacities presented represent maximum annual average throughput based on each company's current production plans. However, as noted, capacity does not necessarily translate to production. Actual production of cellulosic biofuel will likely be well below capacity, especially in the early years of production. We will continue to track these companies and the cellulosic biofuel industry as a whole throughout the duration of the RFS2 program. In addition, we will continue to collaborate with EIA in annual standard setting. A more detailed discussion of the plants corresponding to these company estimates is provided in Section 1.5.3 of the RIA.

BILLING CODE 6560-50-P

BILLING CODE 6560-50-C

d. Feedstock Availability

A wide variety of feedstocks can be used for cellulosic biofuel production, including: Agricultural residues, forestry biomass, certain renewable portions of municipal solid waste and construction and demolition waste (i.e., separated food, yard and incidental, and post-recycled paper and wood waste as discussed in Section II.B.4) and energy crops. These feedstocks are currently much more difficult to convert into biofuel than traditional corn/starch crops or at least require new and different processes because of the more complex structure of cellulosic material.

To determine the likely cellulosic feedstocks for production of 16 billion gallons cellulosic biofuel by 2022, we analyzed the data and results from various sources. Sources include agricultural modeling from the Forestry Agriculture Sector Optimization Model (FASOM) to determine the most economical volume of agriculture residues, energy crops, and forestry resources (see Section VIII for more details on the FASOM) used to meet the standard. We supplemented these estimates with feedstock assessment estimates for the biomass portions of municipal solid waste and construction and demolition waste. [89]

The following subsections describe the availability of various cellulosic feedstocks and the estimated amounts from each feedstock needed to meet the EISA requirement of 16 Bgal of cellulosic biofuel by 2022. Refer to Section IV.B.2.c.iv for the summarized results of the types and volumes of cellulosic feedstocks chosen based on our analyses.

i. Urban Waste

Cellulosic feedstocks available at the lowest cost to the ethanol producer will likely be chosen first. This suggests that urban waste which is already being gathered today and incurs a fee for its disposal may be among the first to be used. Urban wastes are used in a variety of ways. Most commonly, wastes are ground into mulch, dumped into land-fills, or incinerated. We describe two components of urban waste, municipal solid waste (MSW) and construction and demolition (C&D) debris, below.

MSW consists of paper, glass, metals, plastics, wood, yard trimmings, food scraps, rubber, leather, textiles, etc. The portion of MSW that can qualify as renewable biomass under the program is discussed in Section II.B.4.d. The bulk of the biogenic portion of MSW that can be converted into biofuel is cellulosic material such as wood, yard trimmings, paper, and much of food wastes. Paper made up approximately 31% of the total MSW generated in 2008. [90] Although recycling/recovery rates are increasing over time, there appears to still be a large fraction of biogenic material that ends up unused and in land-fills. C&D debris is typically not available in wood waste assessments, although some have estimated this feedstock based on population. Utilization of such feedstocks could help generate energy or biofuels for transportation. However, despite various assessments on urban waste resources, there is still a general lack of reliable data on delivered prices, issues of quality (potential for contamination), and lack of understanding of potential competition with other alternative uses (e.g., recycling, burning for electricity).

We estimated that a total of 44.5 million dry tons of MSW (wood, yard trimmings, paper, and food waste) and C&D wood waste could be available for producing biofuels after factoring in several assumptions, e.g., percent contamination, percent recovered or combusted for other uses, and percent moisture. 91 92 Between the proposal and this final rule, we have updated the assumptions noted above based on newer reports. It should be noted, however, that our estimates of urban waste availability have not changed significantly between the proposal and the final rule. We assumed that approximately 26 million dry tons (of the total 44.5 million dry tons) could be used to produce biofuels. However, many areas of the U.S. (e.g., much of the Rocky Mountains) have such sparse resources that an MSW and C&D cellulosic facility would not likely be justifiable. We did assume that in areas with other cellulosic feedstocks (forest and agricultural residue), that the MSW would be used even if the MSW could not justify the installation of a plant on its own. Therefore, we have estimated that urban waste could help contribute to the production of approximately 2.3 ethanol-equivalent billion gallons of fuel. [93] Note that some processes are likely to also process other portions of MSW (e.g., plastics, rubbers) into fuel, but we have only accounted for the portion expected to qualify as renewable fuel and produce RINs.

In addition to MSW and C&D waste generated from normal day-to-day activities, there is also potential for renewable biomass to be generated from natural disasters. This includes diseased trees, other woody debris, and C&D debris. For instance, Hurricane Katrina was estimated to have damaged approximately 320 million large trees. [94] Katrina also generated over 100 million tons of residential debris, not including the commercial sector. Much of this waste would likely be disposed of and therefore go unused. Collection of this material for the generation of biofuel could be a better alternative use for this waste. While we acknowledge this material could provide a large source in the short-term, natural disasters are highly variable, making it hard to predict amounts of material available in the future. Thus, for our analyses we have not included natural disaster renewable biomass in our estimates.

ii. Agricultural and Forestry Residues

The next category of feedstocks chosen will likely be those that are readily produced but have not yet been commercially collected. This includes both agricultural and forestry residues.

Agricultural residues are expected to play an important role early on in the development of the cellulosic ethanol industry due to the fact that they are already being grown. Agricultural crop residues are biomass that remains in the field after the harvest of agricultural crops. The most common residues are corn stover (the stalks, leaves, and/or cobs) and straw from wheat, rice, barley, and oats. These U.S. crops and others produce more than 500 million tons of residues each year, although only a fraction can be used for fuel and/or energy production due to sustainability and conservation constraints. [95] Crop residues can be found all over the United States, but are primarily concentrated in the Midwest since corn stover accounts for half of all available agricultural residues.

Agricultural residues play an important role in maintaining and improving soil quality, protecting the soil surface from water and wind erosion, helping to maintain nutrient levels, and protecting water quality. Thus, collection and removal of agricultural residues raise concerns about the potential for increased erosion, reduced crop productivity, depletion of soil carbon and nutrients, and water pollution. Sustainable removal rates for agricultural residues have been estimated in various studies, many showing tremendous variability due to local differences in soil and erosion conditions, soil type, landscape (slope), tillage practices, crop rotation managements, and the use of cover crops. One of the most recent studies by top experts in the field shows that under current rotation and tillage practices, about 30% of corn stover (about 59 million metric tons) produced in the U.S. could be collected, taking into consideration erosion, soil moisture concerns, and nutrient replacement costs. [96] The same study shows that if farmers convert to no-till corn management and total stover production does not change, then approximately 50% of stover (100 million metric tons) could be collected without causing erosion to exceed the tolerable soil loss. This study, however, did not consider possible soil carbon loss which other studies indicate may be a greater constraint to environmentally sustainable feedstock harvest than that needed to control water and wind erosion. [97] Experts agree that additional studies are needed to further evaluate how soil carbon and other factors affect sustainable removal rates. Despite unclear guidelines for sustainable removal rates due to the uncertainties explained above, our agricultural modeling analysis assumes that no stover is removable on conventional tilled lands, 35% of stover is removable on conservation tilled lands, and 50% is removable on no-till lands. In general, these removal guidelines are appropriate only for the Midwest, where the majority of corn is currently grown.

As already noted, removal rates will vary by region due to local differences. Given the current understanding of sustainable removal rates, we believe that such assumptions are reasonably justified. Based on our research, we also note that calculating residue maintenance requirements for the amount of biomass that must remain on the land to ensure soil quality is another approach for modeling sustainable residue collection quantities. This approach would likely be more accurate for all landscapes as site-specific conditions such as soil type, topography, etc. could be taken into account. This would prevent site-specific soil erosion and soil quality concerns that would inevitably exist when using average values for residue removal rates across all soils and landscapes. At the time of our analyses, however, we had limited data on which to accurately apply this approach and therefore assumed the removal guidelines based on tillage practices.

Our agricultural modeling (FASOM) suggests that corn stover will make up the majority of agricultural residues used by 2022 to meet the EISA cellulosic biofuel standard (4.9 ethanol-equivalent Bgal). [98] Smaller contributions are expected to come from other crop residues including sugarcane bagasse (0.6 ethanol-equivalent Bgal), wheat residues (0.1 ethanol-equivalent Bgal), and sweet sorghum pulp (0.1 ethanol-equivalent Bgal). [99]

The U.S. also has vast amounts of forest resources that could potentially provide feedstock for the production of cellulosic biofuel. One of the major sources of woody biomass could come from logging residues. The U.S. timber industry harvests over 235 million dry tons annually and produces large volumes of non-merchantable wood and residues during the process. [100] Logging residues are produced in conventional harvest operations, forest management activities, and clearing operations. In 2004, these operations generated approximately 67 million dry tons of forest residues that were left uncollected at harvest sites. [101] Other feedstocks include those from other removal residues, thinnings from timberland, and primary mill residues.

For the NPRM, FASOM was not able to model forestry biomass as a potential feedstock. As a result, we relied on USDA-Forest Service (FS) for information on the forestry sector at the time. For the final rule, we were able to incorporate the forestry sector model in FASOM. EISA does not allow forestry material from national forests and virgin forests that could be used to produce biofuels to count towards the renewable fuels requirement under EISA. Therefore, our modeling of forestry biomass excluded such material. The FASOM model estimated that approximately 0.1 ethanol-equivalent billion gallons would be produced from forestry biomass to meet EISA.

iii. Dedicated Energy Crops

While urban waste, agricultural residues and forest residues will likely be the first feedstocks used in the production of cellulosic biofuel, there may be limitations to their use due to land availability and sustainable removal rates. Energy crops which are not yet grown commercially but have the potential for high yields and a series of environmental benefits could help provide additional feedstocks in the future. Dedicated energy crops are plant species grown specifically for energy purposes. Various perennial plants have been researched as potential dedicated feedstocks, including switchgrass, mixed prairie grasses, hybrid poplar, miscanthus, energy cane, energy sorghum, and willow trees. Refer to Section 1.1.2.2 of the RIA for more information on the benefits and challenges with using dedicated energy crops.

In addition to estimating the extent that agricultural residues might contribute to cellulosic ethanol production, FASOM also estimated the contribution that energy crops might provide (7.9 ethanol-equivalent Bgal). [102] FASOM covers all cropland and pastureland in production in the 48 contiguous United States. For the NPRM, FASOM did not contain all categories of grassland and rangeland captured in USDA's Major Land Use data sets. For the final rule, FASOM accounts for all major land categories, including forestland and rangeland. All crop production, including dedicated energy crops, takes place on cropland. Land categories that can be converted to cropland production include cropland pasture, forest pasture, and forestland. More detail can be found in Chapter VIII of this preamble. Furthermore, we constrained FASOM to be consistent with the 2008 Farm Bill and assumed 32 million acres would stay in Conservation Reserve Program (CRP). [103] Other models, such as USDA's Regional Environment and Agriculture Programming (REAP) model and University of Tennessee's POLYSYS model, have shown that the use of energy crops to meet EISA could be significant, similar to our FASOM modeling results for the final rule. [104]

iv. Summary of Cellulosic Feedstocks for 2022

Table IV.B.3-3 summarizes our internal estimate of the types of cellulosic feedstocks projected to be used and their corresponding volume contribution to 16 billion gallons cellulosic biofuel by 2022 for the purposes of our impacts assessment. The majority of feedstock is projected to come from dedicated energy crops. Other feedstocks include agricultural residues, forestry biomass, and urban waste.

Table IV.B.3-3—Cellulosic Feedstocks Assumed To Meet EISA in 2022105 Back to Top
Feedstock Volume (ethanol-equivalent Bgal)
Agricultural Residues 5.7
Corn Stover 4.9
Sugarcane Bagasse 0.6
Wheat Residue 0.1
Sweet Sorghum Pulp 0.1
Forestry Biomass 0.1
Urban Waste 2.3
Dedicated Energy Crops (Switchgrass) 7.9
Total 16.0

4. Biodiesel & Renewable Diesel

Biodiesel and renewable diesel are replacements for petroleum diesel that are made from plant or animal fats. Biodiesel consists of fatty acid methyl esters (FAME) and can be used in low-concentration blends in most types of diesel engines and other combustion equipment with no modifications. The term renewable diesel covers fuels made by hydrotreating plant or animal fats in processes similar to those used in refining petroleum. Renewable diesel is chemically analogous to blendstocks already used in petroleum diesel, thus its use can be transparent and its blend level essentially unlimited. The goal of both biodiesel and renewable diesel conversion processes is to change the properties of a variety of feedstocks to more closely match those of petroleum diesel (such as its density, viscosity, and storage stability) for which the engines have been designed. The definition of biodiesel given in applicable regulations is sufficiently broad to be inclusive of both fuels. [106] However, the EISA stipulates that renewable diesel that is co-processed with petroleum diesel cannot be counted as biomass-based diesel for purposes of complying with the RFS2 volume requirements. [107]

In general, plant and animal oils are valuable commodities with many uses other than transportation fuel. Therefore we expect the primary limiting factor in the supply of both biodiesel and renewable diesel to be feedstock availability and price. Expansion of their market volumes is dependent on being able to compete on price with the petroleum diesel they are displacing, which will depend largely on continuation of current subsidies and other incentives.

Other biomass-based diesel fuel processes are at various stages of development, but due to uncertainty on production timelines, we didn't include these fuels in the biomass-based diesel impact assessments.

a. Historic and Projected Production

i. Biodiesel

As of November 2009, the aggregate production capacity of biodiesel plants in the U.S. was estimated at 2.8 billion gallons per year across approximately 191 facilities. [108] (However, at the time of this writing it is anticipated that capacity utilization will be approximately 17% for calendar year 2009.) Biodiesel plants exist in nearly all states, with the largest density of plants in the Midwest and Southeast where agricultural feedstocks are most plentiful.

Table IV.B.4-1 gives data on U.S. biodiesel production and use for recent years, including net domestic use after accounting for imports and exports. The figures suggest that the industry has grown out of proportion with actual biodiesel demand. Reasons for this include various state incentives to build plants, along with state and federal incentives to blend biodiesel, which have given rise to an optimistic industry outlook over the past several years. Since the cost of capital is relatively low for the biodiesel production process (typically four to six percent of the total per-gallon cost), this industry developed along a path of more small, privately-owned plants in comparison to the ethanol industry, with median size less than 10 million gallons/yr. [109] These small plants, with relatively low costs other than feedstock, have generally been able to survive producing well below their nameplate capacities.

Table IV.B.4-1—Summary of U.S. Biodiesel Production and Use Back to Top
Year Domestic production capacity Domestic total production Apparent capacity utilization (percent) Net domestic biodiesel use Net domestic use as percent of production
[Million gallons]110
2004 245 28 11 27 96
2005 395 91 23 91 100
2006 792 250 32 261 104
2007 1,809 490 27 358 73
2008 2,610 776 30 413 53
2009 2,806 475 (est.) 17 296 (est.) 62

Someof this industry capacity may not be dedicated specifically to fuel production, instead being used to make oleochemical feedstocks for further conversion into products such as surfactants, lubricants, and soaps. These products do not show up in renewable fuel sales figures.

During 2004-2006, demand for biodiesel grew rapidly, but the trend of increasing sales was quickly surpassed by construction and start-up of new plants Since then, periods of high commodity prices followed by reduced demand for transportation fuel during the economic downturn have caused additional strain on the industry beyond the overcapacity situation. Biodiesel producers were able to find additional markets overseas, and a significant portion of the 2007 and 2008 production was exported to Europe where fuel prices and additional tax subsidies helped offset high feedstock costs. However, the EU enacted a tariff to protect domestic producers early in 2009, after which exports dropped to a small fraction of production. [111] We understand there may be some additional export markets developing within North America, but given the uncertainty at this time, we do not account for any biodiesel exports in our projections.

To perform our impacts analyses for this rule, it was necessary to forecast the state of the biodiesel industry in the timeframe of the fully-phased-in RFS. In general, this consisted of reducing the industry capacity to be much closer to 1.67 billion gallons per year by 2022 (based on the volume requirements to meet the standard; see Section IV.A.2). This was accomplished by considering as screening factors the current production and sales incentives in each state as well as each plant's primary feedstock type and whether it was BQ-9000 certified. [112] Going forward producers will compete for feedstocks and markets may consolidate. During this period the number of operating plants is expected to shrink, with surviving plants utilizing feedstock segregation and pre-treatment capabilities, giving them flexibility to process any mix of feedstocks available in their area. By the end of this period we project a mix of large regional plants and some smaller plants taking advantage of local market niches, with an overall average capacity utilization around 85%. Table IV.B.4-2 summarizes this forecast. See Section 1.5.4 of the RIA for more details.

Table IV.B.4-2—Summary of Projected Biodiesel Industry Characterization Used in Our Analyses113 Back to Top
2008 2022
Total production capacity on-line (million gal/yr) 2,610 1,968
Number of operating plants 176 121
Median plant size (million gal/yr) 5 5
Total biodiesel production (million gal) 776 1,670
Average plant utilization 0.30 0.85

ii. Renewable Diesel

Renewable dieselis a fuel (or blendstock) produced from animal fats, vegetable oils, and waste greases using chemical processes similar to those employed in petroleum hydrotreating. These processes remove oxygen and saturate olefins, converting the triglycerides and fatty acids into paraffins. Renewable diesel typically has higher cetane, lower nitrogen, and lower aromatics than petroleum diesel fuel, while also meeting stringent sulfur standards.

As a result of the oxygen and olefins in the feedstock being removed, renewable diesel has storage, stability, and shipping properties equivalent to petroleum diesel. This allows renewable diesel fuel to be shipped in existing petroleum pipelines used for transporting fuels, thus avoiding a significant issue with distribution of biodiesel. For more on fuel distribution, refer to Section IV.C.

Considering that this industry is still in development and that there are no long-term projections of production volume, we base our volume estimate of 150 MMgal/yr primarily on recent industry project announcements involving proven technology. Due to the current status of tax incentives, we project all of this fuel will be produced at stand-alone facilities.

b. Feedstock Availability

Publically available industry information along with agricultural commodity modeling we have done for this rule (see Section VIII.A) suggests that the three largest sources of feedstock for biodiesel will be rendered animal fats, soy oil, and corn oil extracted from dry mill ethanol facilities. Renewable diesel plants are expected to use solely animal fats due to the fact that these feedstocks are cheaper than vegetable oils and the process can handle them without issue. Comments we have received from a large rendering company suggest there will be adequate fats and greases feedstocks to supply biofuels as well as other historical uses. Table IV.B.4-3 summarizes the feedstock types, process types, and volumes projected to be used in 2022 for biodiesel and renewable diesel. More details on feedstock sources and volumes are presented in Section 1.1.3 of the RIA.

Table IV.B.4-3—Summary of Projected Biodiesel and Renewable Diesel Feedstock Use in 2022 Back to Top
Feedstock type Base catalyzed biodiesel Acid-pretreatment biodiesel Renewable diesel
[MMgal]
Virgin vegetable oil 660
Corn oil from ethanol production 680
Rendered animal fats and greases 230 150
Algae oil or other advanced source 100

C. Biofuel Distribution

The current motor fuel distribution infrastructure has been optimized to facilitate the movement of petroleum-based fuels. Consequently, there are very efficient pipeline-terminal networks that move large volumes of petroleum-based fuels from production/import centers on the Gulf Coast and the Northeast into the heartland of the country. In contrast, most biofuel is produced in the heartland of the country and needs to be shipped to the coasts, flowing roughly in the opposite direction of petroleum-based fuels. In addition, while some renewable fuels such as hydrocarbons may be transparent to the distribution system, the physical/chemical nature of other renewable fuels may limit the extent to which they can be shipped/stored fungibly with petroleum-based fuels. The vast majority of biofuels are currently shipped by rail, barge and tank truck to petroleum terminals. All biofuels are currently blended with petroleum-based fuels prior to use. [114] Most biofuel blends can be used in conventional vehicles. However, E85 can only be used in flex-fuel vehicles, requires specially constructed retail dispensing/storage equipment, and may require special blendstocks at terminals. These factors limit the ability of biofuels to utilize the existing petroleum fuel distribution infrastructure. Hence, the distribution of renewable fuels raises unique concerns and in many instances requires the addition of new transportation, storage, blending, and retail equipment.

1. Biofuel Shipment to Petroleum Terminals

Ethanol currently is not commonly shipped by pipeline because it can cause stress corrosion cracking in pipeline walls and its affinity for water and solvency can result in product contamination concerns. A short gasoline pipeline in Florida is currently shipping batches of ethanol, and other more extensive pipeline systems have feasibility studies underway. [115] Thus, existing petroleum pipelines in some areas of the country may play an increasing role in the shipment of ethanol. Evaluations are also currently underway regarding the feasibility of constructing a new dedicated ethanol pipeline from the Midwest to the East coast. We expect that cellulosic distillate fuels will not have materials compatibility issues with the existing petroleum fuel distribution infrastructure. Thus, there may be more opportunity for cellulosic distillate fuel to be shipped by pipeline. However, the location of both ethanol and cellulosic distillate production facilities relative to the origination points for existing petroleum pipelines will be a limiting factor regarding the extent to which pipelines can be used.

Our analysis of the shipment of ethanol and cellulosic distillate fuels to petroleum terminals is based on the projections of the location of biofuel production facilities and end use areas contained in the NPRM. We assume that the majority of ethanol and cellulosic distillate fuel would be produced in the Midwest, and that both fuels would be shipped to petroleum terminals in a similar fashion (by rail, barge, and tank truck). To the extent which new biofuel production facilities are more dispersed than projected in the NPRM, there may be more opportunity for both fuels to be used closer to their point of manufacture. This potential benefit would primarily apply to cellulosic ethanol and distillate production facilities given that such facilities have yet to be constructed, whereas most corn-ethanol production facilities have already been constructed in the Midwest.

Biodiesel is currently not typically shipped by pipeline due to concerns that it may contaminate jet fuel that is shipped on the same pipeline and potential incompatibility with pipeline gaskets and seals. Kinder Morgan's Plantation pipeline is currently shipping B5 blends on segments of its system that do not handle jet fuel. The shipment of biodiesel by pipeline may become more widespread and might be expanded to systems that handle jet fuel. However, the relatively small production volumes from individual biodiesel plants and the widespread location of such production facilities will tend to limit the extent to which biodiesel may be shipped by pipeline.

Due to the uncertainties regarding the extent to which pipelines might participate in the transportation of biofuels in the future, we assumed that biofuels will continue to be transported by rail, barge, and truck to petroleum terminals as the vast majority of biofuel volumes are today. To the extent that pipelines do play an increasing role in the distribution of ethanol, this may improve reliability in supply and reduce distribution costs. Apart from increased shipment by pipeline, biofuel distribution, and in particular ethanol distribution can be further optimized primarily through the expanded use of unit trains. [116] We anticipate that the vast majority of ethanol and cellulosic distillate facilities will be sized to facilitate unit train service. [117] We do not expect that biodiesel facilities will be of sufficient size to justify shipment by unit train. In the NPRM, we projected that unit train receipt facilities would be located at petroleum terminals and existing rail terminals. Based on industry input regarding the logistical hurdles in locating unit train receipt facilities at petroleum/existing rail terminals, we expect that such facilities will be constructed on dedicated property with rail access that is as close to petroleum terminals as practicable. [118]

Shipment of biofuels by manifest rail to existing rail terminals will continue to be an important means of supplying biofuels to distant markets where the volume of the production facility and/or the local demand is not sufficient to justify shipment by unit train. [119] Shipments by barge will also play an important role in those instances where production and demand centers have water access and in some cases as the final link from a unit train receipt facility to a petroleum terminal. Direct shipment by tank truck from production facilities to petroleum terminals will also continue for shipment over distances shorter than 200 miles.

We project that most biofuel volumes shipped by rail will be delivered to petroleum terminals by tank truck. [120] We expect that this will always be the case for manifest rail shipments. In the NPRM, we projected that trans-loading of biofuels from rail cars to tank trucks would be an interim measure until biofuel storage tanks were constructed. [121] Based on industry input, we now expect trans-loading will be a long-term means of transferring manifest rail car shipments of biofuels received at existing rail terminals to tank trucks for delivery to petroleum terminals. We also anticipate that trans-loading will be used at some unit train receipt facilities, although we expect that most of these facilities will install biofuel storage tanks from which tank trucks will be filled for delivery to petroleum terminals. Imported biofuels will typically be received and be further distributed by tank truck from petroleum terminals that already have receipt facilities for waterborne fuel shipments.

We anticipate that the deployment of the necessary distribution infrastructure to accommodate the shipment of biofuels to petroleum terminals is achievable. [122] We believe that construction of the requisite rail cars, barges, tank trucks, tank truck and rail/barge/truck receipt facilities is within the reach of corresponding construction firms. [123] Although shipment of biofuels by rail represents a major fraction of all biofuel ton-miles, it is projected to account for approximately 0.4% of all rail freight by 2022. Many improvements to the freight rail system will be required in the next 15 years to keep pace with the large increase in the overall freight demand. Given the broad importance to the U.S. economy of meeting the anticipated increase in freight rail demand, and the substantial resources that seem likely to be focused on this cause, we believe that overall freight rail capacity would not be a limiting factor to the successful implementation of the biofuel requirements under EISA.

2. Petroleum Terminal Accommodations

Terminals will need to install additional storage capacity to accommodate the volume of biofuels that we anticipate will be used in response to the RFS2 standards. Petroleum terminals will also need to install truck receipt facilities for biofuels and equipment to blend biofuels into petroleum-based fuels. Upgrades to barge receipt facilities to handle deliveries of biofuels may also be needed at petroleum terminals with water access. Biodiesel storage and blending facilities will need to be insulated/heated in cold climates to prevent biodiesel from gelling. [124] Questions have been raised about the ability of some terminals to install the needed storage capacity due to space constraints and difficulties in securing permits. [125] Overall demand for fuel used in motor vehicles is expected to remain relatively constant through 2022. Thus, much of the increased demand for biofuel storage could be accommodated by modifying storage tanks previously used for the gasoline and petroleum-based diesel fuels that would displaced by biofuels. The areas served by existing terminals also often overlap. In such cases, one terminal might be space constrained while another serving the same area may be able to install the additional capacity to meet the increase in demand. In cases where it is impossible for existing terminals to expand their storage capacity due to a lack of adjacent available land or difficulties in securing the necessary permits, new satellite storage or new separate terminal facilities may be needed for additional storage of biofuels. However, we believe that there would be few such situations.

In the NPRM, we stated the current EPA policy that the RFG and anti-dumping regulations currently require certified gasoline to be blended with denatured ethanol to produce E85. We also stated that if terminal operators add blendstocks to finished gasoline for use in manufacturing E85, the terminal operator would need to register as a refiner with EPA and meet all applicable standards for refiners. Commenters questioned these statements. As we are not taking any action in this final rule with respect to policies surrounding E85, we will consider these comments outside the context of this rule.

3. Potential Need for Special Blendstocks at Petroleum Terminals for E85

ASTM International is considering a proposal to lower the minimum ethanol concentration in E85 to facilitate meeting ASTM minimum volatility specifications in cold climates and when only low vapor pressure gasoline is available at terminals. [126] Commenters have stated that the current proposal to lower the minimum ethanol concentration to 68 volume percent may not be sufficient for this purpose. ASTM International may consider an additional proposal to further decrease the minimum ethanol concentration. Absent such an adjustment, a high-vapor pressure petroleum-based blendstock such as butane would need to be supplied to most petroleum terminals to produce E85 that meets minimum volatility specifications. In such a case, butane would need to be transported by tank truck from petroleum refineries to terminals and storage and blending equipment would be needed at petroleum terminals. [127]

Instead of lowering the minimum ethanol concentration of E85, some stakeholders are discussing establishing a new high-ethanol blend for use in flex-fuel vehicles. Such a fuel would have a minimum ethanol concentration that would be sufficient to allow minimum volatility specifications to be satisfied while using finished gasoline that is already available at petroleum terminals. [128] E85 would continue to be marketed in addition to this new fuel for use in flex-fuel vehicles when E85 minimum volatility considerations could be satisfied.

We believe that industry will resolve the concerns over the ability to meet the minimum volatility needed for high-ethanol blends used in flex-fuel vehicles in a manner that will not necessitate the use of high-vapor pressure blendstocks in their manufacture. Nevertheless, petroleum terminals may find it advantageous to blend butane into E85 because of the low cost of butane relative to gasoline provided that the cost benefit outweighs the associated butane distribution costs. [129]

4. Need for Additional E85 Retail Facilities

The number of additional E85 retail facilities needed to consume the volume of ethanol used under EISA varies substantially depending on the control case. Under our primary mid-ethanol scenario, we estimate that by 2022 an additional 19,765 E85 retail facilities would be needed relative to the AEO reference case to enable the consumption of the ethanol that we project would be used in E85. [130] Under the high-ethanol scenario, we estimate that an additional 23,809 E85 facilities would be needed and that 4,500 E85 facilities that would otherwise be in place would need to be upgraded to include more E85 dispensers by 2022. Whereas under the low-ethanol volume scenario, we project that 11,677 additional E85 facilities would be needed by 2022.

On average, approximately 1,520 additional E85 facilities will be needed each year from 2010 through 2022 under our primary scenario. Under the high and low ethanol scenarios, an additional 1,820 and 900 E85 retail facilities per year respectively would be needed. Under the high ethanol case and to a lesser extent under the primary case, this represents an aggressive timeline for the addition of new E85 facilities given that there are approximately 2,000 E85 retail facilities in service today. Nevertheless, we believe the addition of these new E85 facilities may be possible for the industries that manufacture and install E85 retail equipment. Underwriters Laboratories requires that E85 refueling dispenser systems must be certified as complete units. [131] To date, no complete E85 dispenser systems have been certified by UL. We understand that all the fuel dispenser components with the exception of the hoses that connect to the refueling nozzle have successfully passed the necessary testing. There does not appear to be a technical difficulty in finding hoses that can pass the required testing. Therefore, we anticipate this situation will be resolved once the demand for new E85 facilities is demonstrated. Hence, we believe that the current lack of a UL certification for complete E85 dispenser systems will not impede the installation of the additional E85 facilities that we projected will be needed.

Petroleum retailers expressed concerns about their ability to bear the cost installing the needed E85 refueling equipment given that most retailers are small businesses and have limited capital resources. They also expressed concern regarding their ability to discount the price of E85 relative to E10 sufficiently to persuade flexible fuel vehicle owners to choose E85 given the lower energy density of ethanol. Today's rule does not contain a requirement for retailers to carry E85. We understand that retailers will only install E85 facilities if they can be assured of sufficient E85 throughput to recover their capital costs. The current projections regarding the future cost of gasoline relative to ethanol indicate that it may be possible to price E85 in a competitive fashion to E10. Thus, demand for E85 may be sufficient to encourage retailers to install the needed E85 refueling facilities.

D. Ethanol Consumption

1. Historic/Current Ethanol Consumption

Ethanol and ethanol-gasoline blends have a long history as automotive fuels. In fact, the well-known Model-T was capable of running on both ethanol and gasoline. [132] However, inexpensive crude oil prices kept ethanol from making a significant presence in the transportation sector until the end of the 20th century. Over the past decade, ethanol use has grown rapidly due to oxygenated fuel requirements, MTBE bans, tax incentives, state mandates, the first federal renewable fuels standard (“RFS1”), and rising crude oil prices. Although the cost of crude has come down since reaching record levels in 2008, uncertainty surrounding pricing and the environmental implications of fossil fuels continue to drive ethanol use.

A record 9.5 billion gallons of ethanol were blended into U.S. gasoline in 2008 and EIA is forecasting additional growth in the years to come. [133] According to their recently released Short-Term Energy Outlook (STEO), EIA is forecasting 0.7 million barrels of daily ethanol use in 2009, which equates to 10.7 billion gallons. The October 2009 STEO projects that total ethanol usage (domestic production plus imports) will reach 12.1 billion gallons by 2010. [134]

The National Petrochemical and Refiners Association (NPRA) estimates that ethanol is currently blended into about 75 percent of all gasoline sold in the United States. [135] The vast majority is blended as E10 or 10 volume percent ethanol, although a small amount is blended as E85 for use in flexible fuel vehicles (FFVs).

Complete saturation of the gasoline market with E10 is referred to as the ethanol “blend wall.” The height of the blend wall in any given year is directly related to gasoline demand. In AEO 2009, EIA projects that gasoline demand will peak around 2013 and then start to taper off due to vehicle fuel economy improvements. Based on the primary ethanol growth scenario we're forecasting under today's RFS2 program, the nation is expected to hit the 14-15 billion gallon blend wall by around 2014 (refer ahead to Figure IV.D.2-1), although it could be sooner if gasoline demand is lower than expected. It could also be lower if projected volumes of non-ethanol renewables do not materialize and ethanol usage is higher than expected.

Over the years there have been several policy attempts to increase FFV sales including Corporate Average Fuel Economy (CAFE) credits and government fleet alternative-fuel vehicle requirements. As a result, there are an estimated 8 million FFVs on the road today, up from just over 7 million in 2008. While this is not insignificant in terms of growth, FFVs continue to make up less than 4 percent of the total gasoline vehicle fleet. In addition, E85 is only currently offered at about 1 percent of gas stations nationwide. Ethanol consumption is currently limited by the number of FFVs on the road and the number of E85 outlets or, more specifically, the number of FFVs with access to E85. Still many FFV owners with access to E85 are not choosing it because it is currently priced almost 40 cents per gallon higher than conventional gasoline on an energy equivalent basis. [136] According to EIA, only 12 million gallons of E85 were consumed in 2008. [137]

To meet today's RFS2 requirements we are going to need to see growth in FFV and E85 infrastructure as well as changes in retail pricing and consumer behavior. However, the amount of change needed is proportional to the amount of ethanol observed under the RFS2 program. As explained in Section IV.A, EPA expects total ethanol demand could be anywhere from 17.5 to 33.2 billion gallons in 2022, depending on the amount of non-ethanol cellulosic biofuels that are realized. The low-ethanol case would require only moderate changes in FFV/E85 infrastructure and refueling whereas the high-ethanol case would require very dramatic changes and likely a mandate. For the final rule, we have chosen to focus our impact analyses on the primary mid-ethanol case of 22.2 billion gallons. A discussion of how this volume of ethanol could be consumed in 2022 with expanded FFV/E85 infrastructure is presented below. As expected, the infrastructure changes required under this FRM scenario are less extreme than those highlighted in the proposal based on a predominant ethanol world (34.2 billion gallons of ethanol). However, there are additional technological, logistical and financial barriers that will need to be overcome with respect to commercialization of BTL and non-ethanol cellulosic biofuels. For more on cellulosic diesel technologies, distribution impacts, and production costs, refer to Sections 1.4, 1.6 and 4.1 of the RIA.

2. Increased Ethanol Use Under RFS2

Under the primary ethanol growth scenario considered as part of today's rule, ethanol consumption will need to be about three times higher than RFS1 levels, more than twice as much as today's levels, and 9 billion gallons higher than the ethanol predicted to occur in 2022 absent RFS2 (according to AEO 2007). To get to 22.2 billion gallons of ethanol use according to the potential ramp-up described in Section 1.2 of the RIA, the nation is predicted to hit the blend wall in 2014 as shown below in Figure IV.D.2-1.

As shown above, we are anticipating almost 14 billion gallons of non-ethanol advanced biofuels under today's RFS2 program. But overall, ethanol is expected to continue to be the nation's primary biofuel with over 22 billion gallons in 2022. To get beyond the blend wall and consume more than 14-15 billion gallons of ethanol, we are going to need to see increases in the number FFVs on the road, the number of E85 retailers, and the FFV E85 refueling frequency.

It is possible that conventional gasoline (E0) could continue to co-exist with E10 and E85 for quite some time. However, for analysis purposes, we have assumed that E10 would replace E0 as expeditiously as possible and that all subsequent ethanol growth would come from E85. Furthermore, we assumed that no ethanol consumption would come from the mid-level ethanol blends (e.g., E15) under our primary control case since they are not currently approved for use in non-FFVs. However, as a sensitivity analysis, we have examined the impacts that E15 would have on ethanol consumption (refer to Section IV.D.3).

a. Projected Gasoline Energy Demand

The maximum amount of ethanol our country is capable of consuming in any given year is a function of the total gasoline energy demanded by the transportation sector. Our nation's gasoline energy demand is dependent on the number of gasoline-powered vehicles on the road, their average fuel economy, vehicle miles traveled (VMT), and driving patterns. For analysis purposes, we relied on the gasoline energy projections provided by EIA in the AEO 2009 final release. [138] AEO 2009 takes the fuel economy improvements set by EISA into consideration and also assumes a slight dieselization of the light-duty vehicle fleet. [139] It also takes the recession's impacts on driving patterns into consideration. The result is a 25% reduction in the projected 2022 gasoline energy demand from AEO 2007 (a pre-EISA world) to AEO 2009. [140] EIA essentially has total gasoline energy demand (petroleum-based gasoline plus ethanol) flattening out, and even slightly decreasing, as we move into the future.

b. Projected Growth in Flexible Fuel Vehicles

Over one million FFVs were sold in both 2008 and 2009 according to EPA certification data. Despite the recession and current state of the auto industry, automakers are incorporating more and more FFVs into their light-duty production plans. While the FFV system (i.e., fuel tank, sensor, delivery system, etc.) used to be an option on some vehicles, most automakers are moving in the direction of converting entire product lines over to E85-capable systems. Still, the number of FFVs that will be manufactured and purchased in future years is uncertain.

To measure the impacts of increased volumes of renewable fuel, we considered three different FFV production scenarios that might correspond to the three biofuel control cases analyzed for the final rule. For all three cases, we assumed that total light-duty vehicle sales would follow AEO 2009 trends. The latest EIA report suggests lower than average sales in 2008-2013 (less than 16 million vehicles per year) before rebounding and growing to over 17 million vehicles by 2019. [141] These vehicle projections are consistent with EPA's recently proposed Light-Duty Vehicle GHG Rule. [142]

Although we assumed total vehicle and car/truck sales would be the same in all three cases, we assumed varying levels of FFV production. For our low-ethanol control case, we assumed steady business-as-usual FFV growth according to AEO 2009 predictions. [143] For our primary mid-ethanol control case, we assumed increased FFV sales under the presumption that GM, Ford and Chrysler (referred to hereafter as the “Detroit 3”) would follow through with their commitment to produce 50% FFVs by 2012. Despite the current state of the economy and the hardships facing the auto industry (GM and Chrysler filed for bankruptcy earlier this year), the Detroit 3 appear to still be moving forward with their voluntary FFV commitment. [144] Under our primary control case, we assumed that non-domestic FFVs sales would track around 2%, consistent with today's production/plans. [145] Finally, for our high-ethanol control case, we assumed a theoretical 80% FFV mandate based on the Open Fuel Standard Act of 2009 that was reintroduced in Congress on March 12, 2009. [146] Given today's reduced vehicle sales and gasoline demand, we believe a mandate would be the only viable means for consuming 32.2 billion gallons of ethanol in 2022.

Under our primary mid-ethanol control case, total FFV sales are estimated at just over 4 million vehicles per year in 2017 and beyond. This is less aggressive than the assumptions made in the NPRM. At that time, we were expecting more cellulosic ethanol which could justify higher FFV production assumptions. We assumed that not only would the Detroit 3 fulfill their 50% by 2012 FFV production commitment, non-domestic automakers might follow suit and produce 25% FFV in 2017 and beyond. We also assumed that annual light-duty vehicle sales would continue around the historical 16 million vehicle mark resulting in 6 million FFVs in 2017 and beyond.

Based on our revised vehicle/FFV production assumptions coupled with vehicle survival rates, VMT, and fuel economy estimates applied in the recently proposed Light-Duty Vehicle GHG Rule, the maximum percentage of fuel (gasoline/ethanol mix) that could feasibly be consumed by FFVs in 2022 would be about 20% (down from 30% in the NPRM). For more information on our FFV production assumptions and fuel fraction calculations, refer to Section 1.7.2 of the RIA.

c. Projected Growth in E85 Access

According to the National Ethanol Vehicle Coalition (NEVC), there are currently 2,100 gas stations offering E85 in 44 states plus the District of Columbia. [147] While this represents significant industry growth, it still only translates to 1.3% of U.S. retail stations nationwide carrying the fuel. [148] As a result, most FFV owners clearly do not have reasonable access to E85. For our FFV/E85 analysis, we have defined “reasonable access” as one-in-four pumps offering E85 in a given area. [149] Accordingly, just over 5% of the nation currently has reasonable access to E85, up from 4% in 2008 (based on a mid-year NEVC pump estimate). [150]

There are a number of states promoting E85 usage by offering FFV/E85 awareness programs and/or retail pump incentives. A growing number of states are also offering infrastructure grants to help expand E85 availability. Currently, 10 Midwest states have adopted a progressive Energy Security and Climate Stewardship Platform. [151] The platform includes a Regional Biofuels Promotion Plan with a goal of making E85 available at one third of all stations by 2025. In addition, the American Recovery and Reinvestment Act of 2009 (ARRA or Recovery Act) recently increased the existing federal income tax credit from $30,000 or 30% of the total cost of improvements to $100,000 or 50% of the total cost of needed alternative fuel equipment and dispensing improvements. [152]

Given the growing number of subsidies, it is clear that E85 infrastructure will continue to expand in the future. However, like FFVs, we expect that E85 station growth will be somewhat proportional to the amount of ethanol realized under the RFS2 program. As such, we analyzed three different E85 growth scenarios for the final rule that could correspond to the three different RFS2 control cases. As an upper bound for our high-ethanol control case, we maintained the 70% access assumption we applied for the NPRM. This is roughly equivalent to all urban areas in the United States offering reasonable (one-in-four-station) access to E85. [153] For our other control cases we assumed access to E85 would be lower with the logic that retail stations (the majority of which are independently owned and operated and net around $30,000 per year) would not invest in more E85 infrastructure than what was necessary to meet the RFS2 requirements. For our primary mid-ethanol control case we assumed reasonable access would grow from 4% in 2008 to 60% in 2022 and for our low-ethanol control case we assumed that access would only grow to 40% by 2022. As discussed in Section IV.C, we believe these E85 growth scenarios are possible based on our assessment of distribution infrastructure capabilities.

d. Required Increase in E85 Refueling Rates

As mentioned earlier, there were just over 7 million FFVs on the road in 2008. If all FFVs refueled on E85 100% of the time, this would translate to about 8.3 billion gallons of E85 use. [154] However, E85 usage was only around 12 million gallons in 2008. [155] This means that, on average, FFV owners were only tapping into about 0.15% of their vehicles' E85/ethanol usage potential last year. Assuming that only 4% of the nation had reasonable one-in-four access to E85 in 2008 (as discussed above), this equates to an estimated 4% E85 refueling frequency for those FFVs that had reasonable access to the fuel.

There are several reasons behind today's low E85 refueling frequency. For starters, many FFV owners may not know they are driving a vehicle that is capable of handling E85. As mentioned earlier, more and more automakers are starting to produce FFVs by engine/product line, e.g., all 2008 Chevy Impalas are FFVs. [156] Consequently, consumers (especially brand loyal consumers) may inadvertently buy a flexible fuel vehicle without making a conscious decision to do so. And without effective consumer awareness programs in place, these FFV owners may never think to refuel on E85. In addition, FFV owners with reasonable access to E85 and knowledge of their vehicle's E85 capabilities may still not choose to refuel on E85. They may feel inconvenienced by the increased refueling requirements. Based on its lower energy density, FFV owners will need to stop to refuel 21% more often when filling up on E85 over E10 (and likewise, 24% more often when refueling on E85 over conventional gasoline). [157] In addition, some FFV owners may be deterred from refueling on E85 out of fear of reduced vehicle performance or just plain unfamiliarity with the new motor vehicle fuel. However, as we move into the future, we believe the biggest determinant will be price—whether E85 is priced competitively with gasoline based on its reduced energy density (discussed in more detail in the subsection that follows).

To comply with the RFS2 program and consume 22.2 billion gallons of ethanol by 2022 (under our primary ethanol control case), not only would we need more FFVs and more E85 retailers, we would need to see a significant increase in the current FFV E85 refueling frequency. Based on the FFV and retail assumptions described above in subsections (b) and (c), our analysis suggests that FFV owners with reasonable access to E85 would need to fill up on it as often as 58% of the time, a significant increase from today's estimated 4% refueling frequency. In order for this to be possible, there will need to be an improvement in the current E85/gasoline price relationship.

e. Market Pricing of E85 Versus Gasoline

According to an online fuel price survey, E85 is currently priced almost 40 cents per gallon or about 15% lower than regular grade conventional gasoline. [158] But this is still about 30 cents per gallon higher than conventional gasoline on an energy-equivalent basis. To increase our nation's E85 refueling frequency to the levels described above, E85 needs to be priced competitively with (if not lower than) conventional gasoline based on its reduced energy content, increased time spent at the pump, and limited availability. Overall, we estimate that E85 would need to be priced about 25% lower than E10 at retail in 2022 in order for it to make sense to consumers.

However, ultimately it comes down to what refiners are willing to pay for ethanol blended as E85. The more ethanol you try to blend as E85, the more devalued ethanol becomes as a gasoline blendstock. Changes to state and Federal excise tax structures could help promote ethanol blending as E85. Similarly, high crude prices make E85 look more attractive. According to EIA's AEO 2009, crude oil prices are expected to increase from about $80 per barrel (today's price) to $116/barrel by 2022. [159] Based on our retail cost calculations, ethanol would have to be priced around $2/gallon or less in order to be attractive to refiners for E85 blending in 2022. According to the DTN Ethanol Center, the current rack price for ethanol is around $2.20/gallon. [160] However, as explained in Section 4.4 of the RIA, we project that the average ethanol delivered price will come down in the future under the RFS2 program. Therefore, while gasoline refiners and markets will always have a greater profit margin selling ethanol in low-level blends to consumers based on volume, they should be able to maintain a profit selling it as E85 based on energy content in the future.

Once the nation gets past the blend wall, more ethanol will need to be blended as E85 and less as E10. FFV owners who were formerly refueling on gasoline will need to start filling up on E85. Under our primary control case, we expect that 12.9 billion gallons of ethanol would be blended as E10 and 9.3 billion gallons would be blended as E85 to reach the 22.2 billion gallons in 2022. For more on our ethanol consumption feasibility and retail cost calculations, including discussion of the other two control cases, refer to Section 1.7 of the RIA.

3. Consideration of >10% Ethanol Blends

On March 6, 2009, Growth Energy and 54 ethanol manufacturers submitted an application for a waiver of the prohibition of the introduction into commerce of certain fuels and fuel additives set forth in section 211(f) of the Act. This application seeks a waiver for ethanol-gasoline blends of up to 15 percent ethanol by volume. [161] On April 21, 2009, EPA issued a Federal Register notice announcing receipt of the Growth Energy waiver application and soliciting comment on all aspects of it. [162] On May 20, 2009, EPA issued an additional Federal Register notice extending the public comment period by an additional 60 days. [163] The comment period ended on July 20, 2009, and EPA is now evaluating the waiver application and considering the comments which were submitted.

In a letter dated November 30, 2009, EPA notified the applicant that, because crucial vehicle durability information being developed by the Department of Energy would not be available until mid-2010, EPA would be delaying its decision on the application until a sufficient amount of this information could be included in its analysis so that the most scientifically supportable decision could be made. [164] As the current Growth Energy waiver application is still under review, EPA believes it is appropriate to address aspects of the mid-level blend waiver in its decision announcement on the waiver application as opposed to dealing with the comments and evaluation of the potential waiver in the preamble of today's final rule.

Although EPA has yet to make a waiver decision, since its approval could have a significant impact on our analyses that are based on the use of E85, as a sensitivity analysis, we have evaluated the impacts that E15 could have on ethanol consumption feasibility. More specifically, we have assessed the impacts of a partial waiver for newer technology vehicles consistent with the direction of EPA's November 30, 2009 letter. We assumed that E10 would need to continue to co-exist for legacy and non-road equipment based on consumer demand regardless of any waiver decision. For analysis purposes, we assumed E10 would be marketed as premium-grade gasoline (the universal fuel), E15 would be marketed as regular-grade gasoline (to maximize ethanol throughput) and, like today, midgrade would be blended from the two fuels to make a 12.5 vol% blend (E12.5). In addition, we assumed that some E15-capable vehicles would continue to choose E10 or E12.5 based on our knowledge of today's premium and midgrade sales. [165]

In the event of a partial waiver, it is unclear how long it would take for E15 to be fully deployed or whether it would ever be available nationwide. For analysis purposes, we assumed that E15 would be fully phased in and available at all retail stations nationwide by the time the nation hit the blend wall, or around 2014 for our primary control case shown in Figure IV.D.3-1.

As modeled, a partial waiver for E15 could increase the ethanol consumption potential from conventional vehicles to about 19 billion gallons. Under our primary control case (shown in Figure IV.D.3-1), E15 could postpone the blend wall by up to five years, or to 2019. Although E15 would fall short of meeting the RFS2 requirements under this scenario, it could provide interim relief while the county ramps up non-ethanol cellulosic biofuel production and expands E85/FFV infrastructure. Under our high-ethanol control case, a partial waiver for E15 could eliminate the need for FFV or E85 infrastructure mandates. Under our low-ethanol control case, E15 could eliminate the need for additional FFV/E85 infrastructure all together. For more information, refer to Section 1.7.6 of the RIA.

V. Lifecycle Analysis of Greenhouse Gas Emissions Back to Top

A. Introduction

As recognized earlier in this preamble, a significant aspect of the RFS2 program is the requirement that a fuel meet a specific lifecycle greenhouse gas (GHG) emissions threshold for compliance for each of four types of renewable fuels. This section describes the methodology used by EPA to determine the lifecycle GHG emissions of biofuels, and the petroleum-based transportation fuels that they replace. EPA recognizes that this aspect of the RFS2 regulatory program has received particular attention and comment throughout the public comment period. Therefore, this section also will describe the enhancements made to our approach in conducting the lifecycle analysis for the final rule. This section will highlight areas where we have incorporated new scientific data that has become available since the proposal as well as the approach the Agency has taken to recognize and quantify, where appropriate, the uncertainty inherent in this analysis.

1. Open and Science-Based Approach to EPA's Analysis

Throughout the development of EPA's lifecycle analysis, the Agency has employed a collaborative, transparent, and science-based approach. EPA's lifecycle methodology, as developed for the RFS2 proposal, required breaking new scientific ground and using analytical tools in new ways. The work was generally recognized as state of the art and an advance on lifecycle thinking, specifically regarding the indirect impacts of biofuels.

However, the complexity and uncertainty inherent in this work made it extremely important that we seek the advice and input of a broad group of stakeholders. In order to maximize stakeholder outreach opportunities, the comment period for the proposed rule was extended to 120 days. In addition to this formal comment period, EPA made multiple efforts to solicit public and expert feedback on our approach. Beginning early in the NPRM process and continuing throughout the development of this final rule, EPA held hundreds of meetings with stakeholders, including government, academia, industry, and non-profit organizations, to gather expert technical input. Our work was also informed heavily by consultation with other federal agencies. For example, we have relied on the expert advice of USDA and DOE, as well as incorporating the most recent inputs and models provided by these Agencies. Dialogue with the State of California and the European Union on their parallel, on-going efforts in GHG lifecycle analysis also helped inform EPA's methodology. As described below, formal technical exchanges and an independent, formal peer review of the methodology were also significant components of the Agency's outreach. A key result of our outreach effort has been awareness of new studies and data that have been incorporated into our final rule analysis.

Technology Exchanges: Immediately following publication of the proposed rule, EPA held a two-day public workshop focused specifically on lifecycle analysis to assure full understanding of the analyses conducted, the issues addressed, and the options discussed. The workshop featured EPA presentations on each component of the methodology as well as presentations and discussions by stakeholders from the renewable fuel community, federal agencies, universities, and environmental groups. The Agency also took advantage of opportunities to meet in the field with key, affected stakeholders. For example, the Agency was able to twice participate in meetings and tours in Iowa hosted by the local renewable fuel and agricultural community. As described in this section, one of the many outcomes of these meetings was an improved understanding of agricultural and biofuel production practices.

As indicated in the proposal, our lifecycle results were particularly impacted by assumptions about land use patterns and emissions in Brazil. During the public comment process we were able to update and refine these assumptions, including the incorporation of new, improved sources of data based on Brazil-specific data and programs. In addition, the Agency received more recent trends on Brazilian crop productivity, areas of crop expansion, and regional differences in costs of crop production and land availability. Lastly, we received new information on efforts to curb deforestation allowing the Agency to better predict this impact through 2022.

Peer Review: To ensure the Agency made its decisions for this final rule on the best science available, EPA conducted a formal, independent peer review of key components of the analysis. The reviews were conducted following the Office of Management and Budget's peer review guidance that ensures consistent, independent government-wide implementation of peer review, and according to EPA's longstanding and rigorous peer review policies. In accordance with these guidelines, EPA used independent, third-party contractors to select highly qualified peer reviewers. The reviewers selected are leading experts in their respective fields, including lifecycle assessment, economic modeling, remote sensing imagery, biofuel technologies, soil science, agricultural economics, and climate science. They were asked to evaluate four key components of EPA's methodology: (1) Land use modeling, specifically the use of satellite data and EPA's proposed land conversion GHG emission factors; (2) methods to account for the variable timing of GHG emissions; (3) GHG emissions from foreign crop production (both the modeling and data used); and (4) how the models EPA relied upon are used together to provide overall lifecycle estimates.

The advice and information received through this peer review are reflected throughout this section. EPA's use of higher resolution satellite data is one example of a direct outcome of the peer review, as is the Agency's decision to retain its reliance upon this data. The reviewers also provided recommendations that have helped to inform the larger methodological decisions presented in this final rule. For example, the reviewers in general supported the importance of assessing indirect land use change and determined that EPA used the best available tools and approaches for this work. However, the review also recognized that no existing model comprehensively simulates the direct and indirect effects of biofuel production both domestically and internationally, and therefore model development is still evolving. The uncertainty associated with estimating indirect impacts and the difficulty in developing precise results also were reflected in the comments. In the long term, this peer review will help focus EPA's ongoing lifecycle analysis work as well as our future interactions with the National Academy of Science and other experts.

Altogether, the many and extensive public comments we received to the rule docket, the numerous meetings, workshops and technical exchanges, and the scientific peer review have all been instrumental to EPA's ability to advance our analysis between proposal and final and to develop the methodological and regulatory approach described in this section.

2. Addressing Uncertainty

The peer review, the public comments we have received, and the analysis conducted for the proposal and updated here for the final rule, indicate that it is important to take into account indirect emissions when looking at lifecycle emissions from biofuels. It is clear that, especially when considering commodity feedstocks, including the market interactions of biofuel demand on feedstock and agricultural markets is a more accurate representation of the impacts of an increase in biofuels production on GHG emissions than if these market interactions are not considered.

However, it is also clear that there are significant uncertainties associated with these estimates, particularly with regard to indirect land use change and the use of economic models to project future market interactions. Reviewers highlighted the uncertainty associated with our lifecycle GHG analysis and pointed to the inherent uncertainty of the economic modeling.

In the proposal, we asked for comment on whether and how to conduct an uncertainty analysis to help quantify the magnitude of this uncertainty and its relative impact on the resulting lifecycle emissions estimates. The results of the peer review, and the feedback we have received from the comment process, supported the value of conducting such an analysis. Therefore, working closely with other government agencies as well as incorporating feedback from experts who commented on the rule, we have quantified the uncertainty associated specifically with the international indirect land use change emissions associated with increased biofuel production.

Although there is uncertainty in all portions of the lifecycle modeling, we focused our uncertainty analysis on the factors that are the most uncertain and have the biggest impact on the results. For example, the energy and GHG emissions used by a natural gas-fired ethanol plant to produce one gallon of ethanol can be calculated through direct observations, though this will vary somewhat between individual facilities. The indirect domestic emissions are also fairly well understood, however these results are sensitive to a number of key assumptions (e.g., current and future corn yields). The indirect, international emissions are the component of our analysis with the highest level of uncertainty. For example, identifying what type of land is converted internationally and the emissions associated with this land conversion are critical issues that have a large impact on the GHG emissions estimates.

Therefore, we focused our efforts on the international indirect land use change emissions and worked to manage the uncertainty around those impacts in three ways: (1) Getting the best information possible and updating our analysis to narrow the uncertainty, (2) performing sensitivity analysis around key factors to test the impact on the results, and (3) establishing reasonable ranges of uncertainty and using probability distributions within these ranges in threshold assessment. The following sections outline how we have incorporated these three approaches into our analysis.

EPA recognizes that as the state of scientific knowledge continues to evolve in this area, the lifecycle GHG assessments for a variety of fuel pathways will continue to change. Therefore, while EPA is using its current lifecycle assessments to inform the regulatory determinations for fuel pathways in this final rule, as required by the statute, the Agency is also committing to further reassess these determinations and lifecycle estimates. As part of this ongoing effort, we will ask for the expert advice of the National Academy of Sciences, as well as other experts, and incorporate their advice and any updated information we receive into a new assessment of the lifecycle GHG emissions performance of the biofuels being evaluated in this final rule. EPA will request that the National Academy of Sciences over the next two years evaluate the approach taken in this rule, the underlying science of lifecycle assessment, and in particular indirect land use change, and make recommendations for subsequent rulemakings on this subject. This new assessment could result in new determinations of threshold compliance compared to those included in this rule that would apply to future production (from plants that are constructed after each subsequent rule).

B. Methodology

The regulatory purpose of this analysis is to determine which biofuels (both domestic and imported) qualify for the four different GHG reduction thresholds and renewable fuel categories established in EISA (see Section I.A). This threshold assessment compares the lifecycle emissions of a particular biofuel against the lifecycle emissions of the petroleum-based fuel it is replacing (e.g., ethanol replacing gasoline or biodiesel replacing diesel). This section discusses the Agency's approach both for assessing the lifecycle GHG emissions from biofuels as well as for the petroleum-based fuels replaced by the biofuels.

As described in detail below, EPA has received a number of comments on the different pieces of this analysis and has thoroughly considered those comments as well as feedback from our peer review process. In each section below we will discuss comments received and how they impacted our analysis.

1. Scope of Analysis

As stated in the proposal, the definition of lifecycle GHG emissions established by Congress in EISA is critical to establishing the scope of our analysis. Congress specified that:

The term “lifecycle greenhouse gas emissions” means the aggregate quantity of greenhouse gas emissions (including direct emissions and significant indirect emissions such as significant emissions from land use changes), as determined by the Administrator, related to the full fuel lifecycle, including all stages of fuel and feedstock production and distribution, from feedstock generation or extraction through the distribution and delivery and use of the finished fuel to the ultimate consumer, where the mass values for all greenhouse gases are adjusted to account for their relative global warming potential. [166]

This definition forms the basis of defining the goal and scope of our lifecycle GHG analysis and in determining to what extent changes should be made to the analytical approach outlined in our proposed rulemaking.

a. Inclusion of Indirect Land Use Change

EPA notes that it received significant comment on including international indirect emissions in its lifecycle calculations. Most of the comments suggested that the science of international indirect land use change was too new, or that the uncertainty involved was too great, to be included in a regulatory analysis. EPA continues to believe that compliance with the EISA mandate—determining “the aggregate GHG emissions related to the full fuel lifecycle, including both direct emissions and significant indirect emissions such as land use changes”—makes it necessary to assess those direct and significant indirect impacts that occur not just within the United States, but also those that occur in other countries.

Some commenters strongly supported EPA's proposal to include significant GHG emissions that occur overseas and are related to the lifecycle of renewable fuels or baseline fuels used in the United States. These commenters agreed that the text of the statute supports EPA's proposed approach, and that the alternative of ignoring such emissions would result in grossly inaccurate assessments, and would be inconsistent with the international nature of GHG pollution and the fact that overseas emissions have domestic impacts.

Other commenters argued that the presumption against extraterritorial application of domestic laws carries with it the presumption that Congress is concerned with domestic effects and domestic impacts only. They assert further that Congress intended to benefit domestic agriculture through EISA enactment, and that the statute's ambiguous terms should not be interpreted in a manner that could harm domestic agriculture in general or, for one commenter, the biodiesel industry in particular. Although considering international emissions in its analyses could result in different implications under the statute for various fuels and fuel pathways as compared to ignoring these emissions, EPA believes that this is precisely the outcome that Congress intended. Implementation of EISA will undoubtedly benefit the domestic agricultural sector as a whole, with some components benefiting more than others depending in part on the lifecycle GHG emissions associated with the products to be made from individual feedstocks. If Congress had sought to promote all biofuel production without regard to GHG emissions related to the full lifecycle of those fuels, it would not have specified GHG reduction thresholds for each category of renewable fuel for which volume targets are specified in the Act.

It is also important to note that including international indirect emissions in EPA's lifecycle analysis does not exercise regulatory authority over activities that occur solely outside the U.S., nor does it raise questions of extra-territorial jurisdiction. EPA's regulatory action involves an assessment of products either produced in the U.S. or imported into the U.S. EPA is simply assessing whether the use of these products in the U.S. satisfies requirements under EISA for the use of designated volumes of renewable fuel, cellulosic biofuel, biomass-based diesel, and advanced biofuel. Considering international emissions in determining the lifecycle GHG emissions of the domestically-produced or imported fuel does not change the fact that the actual regulation of the product involves its use solely inside the U.S.

A number of commenters pointed to the text and structure of the definition of “lifecycle greenhouse gas emissions” to argue that EPA either is not authorized to consider GHG emissions related to international land use change, or that it is not required to do so. One commenter suggested that the reference in the definition of “lifecycle greenhouse gas emissions” to “all stages” of the lifecycle “from” feedstock generation “through” use of the fuel by the ultimate consumer does not include indirect emissions that result from decisions to place more land in acreage overseas for such non-fuel purposes as cattle feed. Another commenter stated that EPA's approach does not give any meaning to the terms “significant” and “fuel lifecycle” in the definition, but instead focuses on the words such as “full” to arrive at an expansive meaning. This commenter also noted the lack of any specific reference to international considerations in Section 211(o), as opposed to other provisions in the CAA, such as Section 115.

EPA believes that a complete analysis of the aggregate GHG emissions related to the full lifecycle of renewable fuels includes the significant indirect emissions from international land use change that are predicted to result from increased domestic use of agricultural feedstocks to produce renewable fuel. The statute specifically directs EPA to include in its analyses significant indirect emissions such as significant emissions from land use changes. EPA has not ignored either the terms “significant” or “life cycle.” It is clear from EPA's assessments that the modeled indirect emissions from land use changes are “significant” in terms of their relationship to total GHG emissions for given fuel pathways. Therefore, they are appropriately considered in the total GHG emissions profile for the fuels in question. EPA has not ignored the term “life cycle.” The entire approach used by EPA is directed to fully analyzing emissions related to the complete lifecycle of renewable and baseline fuels.

Although the definition of lifecycle greenhouse gas emissions in Section 211(o) does not specifically mention international emissions, it would be inconsistent with the intent of this section of the amended Act to exclude them. A large variety of activities outside the U.S. play a major part in the full fuel lifecycle of both baseline (gasoline and diesel fuel used as transportation fuel in 2005) and renewable fuels. For example, several stages of the lifecycle process for gasoline and diesel can occur overseas, including extraction and delivery of imported crude oil, and for imported gasoline and diesel products, emissions associated with refining and distribution of the finished product to the U.S. For imported renewable fuel, all of the emissions associated with feedstock production and distribution, fuel processing, and delivery of the finished renewable fuel to the U.S. occur overseas. The definition of lifecycle GHG emissions makes it clear that EPA is to determine the aggregate emissions related to the “full” fuel lifecycle, including “all stages of fuel and feedstock production and distribution.” Thus, EPA could not, as a legal matter, ignore those parts of a fuel lifecycle that occur overseas.

Drawing a distinction between GHG emissions that occur inside the U.S. as compared to emissions that occur outside the U.S. would result in a lifecycle analysis that bears no apparent relationship to the purpose of this provision. The purpose of the thresholds in EISA is to require the use of renewable fuels that achieve reductions in GHG emissions compared to the baseline. Ignoring international emissions, a large part of the GHG emission associated with the different fuels, would result in a GHG analysis that bears no relationship to the real world emissions impact of transportation fuels. The baseline would be significantly understated, given the large amount of imported crude and imported finished gasoline and diesel used in 2005. Likewise, the emissions estimates for imported renewable fuel would be grossly reduced in comparison to the aggregate emissions estimates for fuels made domestically with domestically-grown feedstocks, simply because the impacts of domestically produced fuels occurred within the U.S. EPA does not believe that Congress intended such a result.

Excluding international impacts means large percentages of GHG emissions would be ignored. This would take place in a context where the global warming impact of emissions is irrespective of where the emissions occur. If the purpose of thresholds is to achieve some reduction in GHG emissions in order to help address climate change, then ignoring emissions outside our borders interferes with the ability to achieve this objective. Such an approach would essentially undermine the purpose of the provision, and would be an arbitrary interpretation of the broadly phrased text used by Congress.

One commenter stated that matters that could appropriately be considered part of a food lifecycle (new land clearing for overseas grain production as a result of decreased U.S. grain exports) should not be considered part of a renewable fuel lifecycle. However, the suggested approach would mean that EPA would fail to account for the significant indirect emissions that relate to renewable fuel production. EPA believes this would be counter to Congressional intent. Although a life cycle analysis of foreign food production may also take into account a given land use change, that does not mean that the same land use change should not be considered in evaluating its ultimate cause, which may be renewable fuel production in the United States.

Some comments asserted that significant GHG gas emissions from international land use change should not be considered if the only available models for doing so are not generally accepted or valid considering economics or science, or where the approach is new and untested, or where the data are faulty and EPA models unrealistic scenarios. As described in this rulemaking, EPA has used the best available models and substantially modified key inputs to those models to reflect comments by peer reviewers, the public, and emerging science. EPA has also modeled additional scenarios from those described in the NPRM. EPA recognizes that uncertainty exists with respect to the results, and has attempted to quantify the range of uncertainty. While EPA agrees that application of the models it has used in the context of assessing GHG emissions represents changes from previous biofuel lifecycle modeling, EPA disagrees that it has used faulty data, modeled unrealistic scenarios, or that its approach is otherwise scientifically indefensible. Although the results of modeling GHG emissions associated with international land use change are uncertain, EPA has attempted to quantify that uncertainty and is now in a better position to consider the uncertainty inherent in its approach.

One commenter asserted that by considering international land use changes, EPA is seeking to penalize domestic renewable fuel producers for impacts over which they have no control. In response, EPA disagrees that it is seeking to penalize anyone at all. EPA is simply attempting to account for all GHG emissions related to the full fuel lifecycle. Domestic renewable fuel producers may have no direct control over land use changes that occur overseas as a result of renewable fuel production and use here, but their choice of feedstock can and does influence oversees activities, and EPA believes it is appropriate to consider the GHG emissions from those activities in its analyses.

Some commenters noted that a finding of causation is built into the definitions of “indirect effects” in the Endangered Species Act and the National Environmental Policy Act, and that EPA should interpret the reference to “indirect emissions' in EISA as requiring similar findings of causation. Specifically, they argue that for EPA to count GHG emissions from international land use change in its assessments, EPA must find that renewable fuel production “caused” the land use change. In response, without addressing the commenter's claims regarding the requirements of NEPA or the ESA, EPA notes that Congress has specified in Section 211(o) the required causal link between a fuel and indirect emissions. The indirect emissions must be “related to” the full fuel lifecycle. EPA believes that it has demonstrated this link through its modeling efforts. Specifically, the models predict that increased demand for feedstocks to produce renewable fuel that satisfies EISA mandates will likely result in international land use change. Such change is, then, “related to” the full fuel lifecycle of these fuels. EPA does not believe that the statute requires EPA to wait until these effects occur to establish the required linkage, but instead believes that it is authorized to use predictive models to demonstrate likely results.

The term “related to” is generally interpreted broadly as meaning to have a connection to or refer to a matter. To determine whether an indirect emission has the appropriate connection to the full fuel lifecycle, we must look at both the objectives of this provision as well as the nature of the relationship. EPA has used a suite of global models to project a variety of agricultural impacts of the RFS program, including changes in the types of crops and number of acres planted world-wide. These shifts in the agricultural market are a direct consequence of the increased demand for biofuels in the U.S. This increased demand diverts biofuel feedstocks from other competing uses, and also increases the price of the feedstock, thus spurring additional international production. Our analysis uses country-specific information to determine the amount, location, and type of land use change that would occur to meet these changes in production patterns. The linkages of these changes to increased U.S. biofuel demand in our analysis are generally close, and are not extended or overly complex.

Overall, EPA is confident that it is appropriate to consider indirect emissions, including those from both domestic and international land use changes, as “related to” the full fuel lifecycle, based on the results of our modeling. These results form a reasonable technical basis for the linkage between the full fuel lifecycle of transportation fuels and indirect emissions, as well as for the determination that these emissions are significant. EPA believes that while uncertainty in the resulting aggregate GHG estimates should be taken into consideration, it would be inappropriate to exclude indirect emissions estimates from this analysis. The use of reasonable estimates of these kinds of indirect emissions allows EPA to conduct a reasoned evaluation of total GHG impacts, which is needed to promote the objectives of this provision, as compared to ignoring or not accounting for these indirect emissions.

EPA understands that including international indirect land use change is a key decision and that there is significant uncertainty associated with it. That is why we have taken an approach that quantifies that uncertainty and presents the weight of currently available evidence in making our threshold determinations.

b. Models Used

As described in the proposal, to estimate lifecycle indirect impacts of biofuel production requires the use of economic modeling to determine the market impacts of using agricultural commodity feedstocks for biofuels. The use of economic models and the uncertainty of those models to accurately predict future agricultural sector scenarios was one of the main comments we received on our analysis. While the comments and specifically the peer review supported our need to use economic models to incorporate and measure indirect impacts of biofuel production, they also highlighted the uncertainty with that modeling approach, especially in projecting out to the future.

However, it is important to note that while there are many factors that impact the uncertainty in predicting total land used for crop production, making accurate predictions of many of these factors are not relevant to our analysis. For example different assumptions about economic growth rates, weather, and exchange rates will all impact future agricultural projections including amount of land use for crops. However, we are interested only in the difference between two biofuel scenarios holding all other changes constant. So the absolute values and projections for crops and other variables in the model projections are not as important as the difference the model is projecting due to an increase in biofuels production. This limits the uncertainty of using the economic models for our analysis.

Furthermore, one of the key uncertainties associated with our agricultural sector economic modeling that has the biggest impact on land use change results is the assumptions around crop yields. As discussed in Section V.A.2, we are conducting sensitivity analysis around different yield assumptions in our analysis.

Therefore, because of the fact that we are only using the economic models to determine the difference between two projected scenarios and the fact that we are conducting sensitivity analysis around the yield assumptions we feel it is appropriate and acceptable to use economic models in our analysis of determining GHG thresholds in our final rule analysis.

As was the case in the proposed analysis, to estimate the changes in the domestic agricultural sector (e.g., changes in crop acres resulting from increased demand for biofuel feedstock or changes in the number of livestock due to higher corn prices) and their associated emissions, EPA uses the Forestry and Agricultural Sector Optimization Model (FASOM), developed by Texas A&M University and others. To estimate the impacts of biofuels feedstock production on international agricultural and livestock production, we used the integrated Food and Agricultural Policy and Research Institute international models, as maintained by the Center for Agricultural and Rural Development (FAPRI-CARD) at Iowa State University.

One of the main comments we received on our choice of models was the issue of transparency. Several comments were concerned that the results of EPA's modeling efforts can not be duplicated outside the experts who developed the models and conducted the analysis used by EPA in the proposal. Upon the release of the proposal, EPA requested comment on the use of these various models. EPA conducted a number of measures to gather comments, including the public comment period upon release of the NPRM analysis, holding a public workshop on the lifecycle methodology, and conducting a peer review of the lifecycle methodology. Specifically, one of the major tasks of the peer review of EPA's lifecycle GHG methodology was to review and comment on the use of the various models and their linkages. The response we received through the peer review is supportive of our use of the FASOM and FAPRI-CARD models, affirming that they are the strong and appropriate tools for the task of estimating land use changes stemming from agricultural economic impacts due to changes in biofuel policy.

In addition, in an effort to garner as useful comments as possible and to be as transparent as possible about the modeling process, EPA supplied in the docket technical documents for the FASOM and FAPRI-CARD models, the output received by EPA from each model, and the models themselves such that the public and commenters could learn and examine how each model operates.

Building upon the support for the use of the FASOM and FAPRI-CARD models, a number of important enhancements were made to both models in response to comments received through the public comment system and through the peer review, and in consultation with various experts on domestic and international agronomics. These enhancements include updated substitution rates of corn and soybean meal for distillers grains (DG) based on recent scientific research by Argonne National Laboratory, the addition of a corn oil from the dry mill ethanol extraction process as a source of biodiesel, the full incorporation of FASOM's forestry model that dynamically interacts with the agriculture sector model in the U.S., as well as the addition of a Brazil regional model to the FAPRI-CARD modeling system. All of these enhancements are discussed in more detail below and in the RIA (Chapter 2 and 5). In addition to the model enhancements we also conducted a sensitivity analysis on yields as part of our final rule analysis. These updates to our modeling and the sensitivity analysis was done in response to public comments specifically asking for this to add transparency to the modeling and modeling results.

We also received comments on the combined use of FASOM and FAPRI-CARD. Several comments and peer reviewers questioned the benefit of using two agricultural sector models. Specifically reviewers pointed to some of the inconsistencies in the FASOM and FAPRI-CARD domestic results. For the final rule analysis we worked to reconcile the two model results. We apply the same set of scenarios and key input assumptions in both models. For example, both models were updated to apply consistent treatment of DGs in domestic livestock feed replacement and consistent assumptions regarding DG export.

Some reviewers questioned the benefits of using FASOM and suggested we rely entirely on the FAPRI-CARD model for the analysis. However, we continue to believe there are benefits to the use of FASOM. Specifically, the fact that FASOM has domestic land use change interactions between crop, pasture, and forest integrated into the modeling is an advantage over using the domestic FAPRI-CARD model that only tracks cropland.

c. Scenarios Modeled

As was done for the proposal, to quantify the lifecycle GHG emissions associated with the increase in renewable fuel mandated by EISA, we compared the differences in total GHG emissions between two future volume scenarios in our economic models. For each individual biofuel, we analyzed the incremental GHG emission impacts of increasing the volume of that fuel to the total mix of biofuels needed to meet the EISA requirements. Rather than focus on the impacts associated with a specific gallon of fuel and tracking inputs and outputs across different lifecycle stages, we determined the overall aggregate impacts across sectors of the economy in response to a given volume change in the amount of biofuel produced.

Volume Scenarios: The two future scenarios considered included a “business as usual” volume of a particular renewable fuel based on what would likely be in the fuel pool in 2022 without EISA, as predicted by the Energy Information Agency's Annual Energy Outlook (AEO) for 2007 (which took into account the economic and policy factors in existence in 2007 before EISA). The second scenario assumed a higher volume of renewable fuels as mandated by EISA for 2022.

We project our analysis and economic modeling through the life of the program. We then consider the impacts of an increase of biofuels on the agricultural sector in 2022 as the basis for our threshold analysis. This was an area that we received numerous comments on highlighting that this approach adds uncertainty to our results because we are projecting uncertain technology and other changes out into the future. One of the recommendations was to base the lifecycle GHG assessments on a near term time frame and update the analysis every few years to capture actual technology changes.

We continue to focus our final rule analyses on 2022 results for two main reasons. First, it would require an extremely complex assessment and administratively difficult implementation program to track how biofuel production might continuously change from month to month or year to year. Instead, it seems appropriate that each biofuel be assessed a level of GHG performance that is constant over the implementation of this rule, allowing fuel providers to anticipate how these GHG performance assessments should affect their production plans. Second, it is appropriate to focus on 2022, the final year of ramp up in the required volumes of renewable fuel as this year. Assessment in this year allows the complete fuel volumes specified in EISA to be incorporated. This also allows for the complete implementation of technology changes and updates that were made to improve or modeling efforts. For example, the inclusion of price induced yield increases and the efficiency gains of DGs replacement are phased in over time. Furthermore, these changes are in part driven by the changes in earlier years of increased biofuel use.

Crop Yield Scenarios: EPA received numerous comments to the effect that we should consider a case in our economic models with higher yields that what were projected for the proposed rule analysis. There are many factors that go into the economic modeling but the yield assumptions for different crops has one of the biggest impacts on land use and land use change. Therefore, for this analysis we ran a base yield case and a high yield case. This will provide two distinct model results for key parameters like total amount of land converted by crop by country.

EPA's base yield projections are derived from extrapolating through 2022 long-term historical U.S. corn yields from 1985 to 2009. This estimate, 183 bushels/acre for corn and 48 bushels/acre for soybeans, is consistent with USDA's method of projecting future crop yields. During the public comment process we learned that numerous technical advancements— including better farm practices, seed hybridization and genetic modification—have led to more rapid gains in yields since 1995. In addition, commenters, including many leading seed companies, provided data supporting more rapid improvements in future yields. For example, commenters pointed to recent advancements in seed development (including genetic modification) and the general accumulation of knowledge of how to develop and bring to market seed varieties—factors that would allow for a greater rate of development of seed varieties requiring fewer inputs such as fertilizer and pest management applications. This new information would suggest that the base yield may be a conservative estimate of future yields in the U.S. Therefore, in coordination with USDA experts, EPA has developed for this final rule a high yield case scenario of 230 bushels/acre for corn and 60 bushels/acre for soybeans. These figures represent the 99% upper bound confidence limit of variability in historical U.S. yields. This high yield case represents a feasible high yield scenario for the purpose of a sensitivity test of the impact on the results of higher yields.

Feedback we received indicated that corn and soybean yields respond in tandem and that a high yield corn case would also imply a higher yield for soybeans as well. The high yield case is therefore based on higher yield corn and soybeans in the U.S. as well as in the major corn and soybean producing countries around the world. For international yields, it is reasonable to assume the same percent increases from the baseline yield assumptions could occur as we are estimating for the U.S. Thus in the case of corn, 230 bushels per acre is approximately 25% higher than the U.S. baseline yield of 183 bushels per acre in 2022. This same 25% increase in yield can be expected for the top corn producers in the rest of the world by 2022, as justified improvements in seed varieties and, perhaps even more so than in the case of the U.S., improvements in farming practices which can take more full advantage of the seed varieties' potential. For example, seeds can be more readily developed to perform well in the particular regions of these countries and can be coupled with much improved farming practices as farmers move away from historical practices such as saving seeds from their crop for use the next year and better understand the economic advantages of modern farming practices. So the high yield scenarios would not have the same absolute yield values in other countries as the U.S. but would have the same percent increase.

While we modeled a high yield scenario for this analysis we continue to rely primarily on the base yield estimates in our assessments of different biofuel lifecycle GHG emissions recognizing that the base yields could be conservative. The reasons outlined above could lead to higher rates of yield growth in the future, however, there are mitigating factors that could limit this yield growth or potentially cause reductions in yield growth rates. For example, the water requirements for both increased corn farming and ethanol production could lead to future water constraints that may in some regions limit yield growth potential. Furthermore, one of the long term impacts of potential global climate change could be a reduction in agricultural output of different impacted regions around the world, including the U.S. This could also serve to reduce yield growth. As with many aspects of this lifecycle modeling, as the science and data evolves on crop yields, the Agency will update its factors accordingly.

2. Biofuel Modeling Framework & Methodology for Lifecycle Analysis Components

As discussed above, to account for the direct and indirect emissions of biofuel production required the use of agricultural sector economic models. The results of these models were combined with other data sources to generate lifecycle GHG emissions for the different fuels. The basic modeling framework involved the following steps and modeling tools.

To estimate the changes in the domestic agricultural sector we used FASOM, developed by Texas A&M University and others. FASOM is a partial equilibrium economic model of the U.S. forest and agricultural sectors that tracks over 2,000 production possibilities for field crops, livestock, and biofuels for private lands in the contiguous United States. Because FASOM captures the impacts of all crop production, not just biofuel feedstock, we are able to use it to determine secondary agricultural sector impacts, such as crop shifting and reduced demand due to higher prices.

The output of the FASOM analysis includes changes in total domestic agricultural sector fertilizer and energy use. These are calculated based on the inputs required for all the different crops modeled and changes in the amounts of the different crops produced due to increased biofuel production. FASOM output also includes changes in the number and type of livestock produced. These changes are due to the changes in animal feed prices and make-up due to the increase in biofuel production. The FASOM output changes in fertilizer, energy use, and livestock are combined with GHG emission factors from those sources to generate biofuel lifecycle impacts. The GHG emission factors for fuel and fertilizer production come from the Greenhouse gases, Regulated Emissions, and Energy use in Transportation (GREET) spreadsheet analysis tool developed by Argonne National Laboratories, and livestock GHG emission factors are from IPCC guidance.

To estimate the domestic impacts of N 2 O emissions from fertilizer application, we used the DAYCENT model developed by Colorado State University. The DAYCENT model simulates plant-soil systems and is capable of simulating detailed daily soil water and temperature dynamics and trace gas fluxes (CH 4, N 2 O, and NO X). DAYCENT model results for N 2 O emissions from different crop and land use changes were combined with FASOM output to generate overall domestic N 2 O emissions.

FASOM output also provides changes in total land use required for agriculture and land use shifting between crops, and interactions with pasture, and forestry. This output is combined with emission factors from land use change to generate domestic land use change GHG emissions from increased biofuel production.

To estimate the impacts of biofuels feedstock production on international agricultural and livestock production, we used the integrated FAPRI-CARD international models, developed by Iowa State University. These worldwide agricultural sector economic models capture the biological, technical, and economic relationships among key variables within a particular commodity and across commodities.

The output of the FAPRI-CARD model included changes in crop acres and livestock production by type by country globally. Unlike FASOM, the FAPRI-CARD output did not include changes in fertilizer or energy use or have land type interactions built in. These were developed outside the FAPRI-CARD model and combined with the FAPRI-CARD output to generate GHG emission impacts.

Crop input data by crop and country was developed and combined with the FAPRI-CARD output crop acreage change data to generate overall changes in fertilizer and energy use. These fertilizer and energy changes along with the FAPRI-CARD output livestock changes were then converted to GHG emissions based on the same basic approach used for domestic sources, which involves combining with emission factors from GREET and IPCC.

International land use change emissions were determined based on combining FAPRI-CARD output of crop acreage change with satellite data to determine types of land impacted by the projected crop changes and then applying emission factors of different land use conversions to generate GHG impacts.

Additional modeling and data sources used to determine the GHG emissions of other stages in the biofuel lifecycle include studies and data on the distance and modes of transport needed to ship feedstock from the field to the biofuel processing facility and the finished biofuel from the facility to end use. These distances and modes are used to develop amount and type of energy used for transport which is combined with GREET factors to generate GHG emissions. We also calculate energy use needed in the biofuel processing facility from industry sources, reports, and process modeling. This energy use is combined with emissions factors from GREET to develop GHG impacts of the biofuel production process

The following sections outline how the modeling tools and methodology discussed above were used in conducting the analysis for the different lifecycle stages of biofuel production, including changes made since the proposal. Lifecycle stages discussed include feedstock production, land use change, feedstock and fuel transport, biofuel production, and vehicle end use. The modeling of the petroleum fuels baseline is discussed in Section V.B.3.

a. Feedstock Production

Our analysis addresses the lifecycle GHG emissions from feedstock production by capturing both the direct and indirect impacts of growing corn, soybeans, and other renewable fuel feedstocks. For both domestic and international agricultural feedstock production, we analyzed four main sources of GHG emissions: agricultural inputs (e.g., fertilizer and energy use), fertilizer N 2 O, livestock, and rice methane. (Emissions related to land use change are discussed in the next section).

i. Domestic Agricultural Sector Impacts

Agricultural Sector Inputs: The proposal analysis calculated GHG emissions from domestic agriculture fertilizer and energy use and production change by applying rates of energy and fertilizer use by crop by region to the FASOM acreage data and then multiplying by default factors for GHG emissions from GREET. Fuel use emissions from GREET include both the upstream emissions associated with production of the fuel and downstream combustion emissions.

In general commenters supported this approach as it captures all indirect impacts of agricultural sector emissions and not just those associated with the specific biofuel crop in question. However, we did receive comments as part of our Model Linkages Peer Review that the input data for some crops may be overestimating GHG emissions. Specifically, the commenter highlighted that N 2 O emissions from domestic hay production seemed to be over estimated. As part of the final rule analysis EPA confirmed that input data was being used correctly, however, the hay N 2 O emissions in the proposal may have been overestimated based on the approach used in the proposal to generate N 2 O emissions from nitrogen fixing crops. This has been updated for the final rule analysis as discussed in the next section which resulted in lower emissions from nitrogen fixing crops.

Other comments indicated that we should be using the most up to date data for our calculations of GHG emissions. Since the proposal there has been a new release of the GREET model (Version 1.8C). EPA reviewed the new version and concluded that this was an improvement over the previous GREET release that was used in the proposal analysis (Version 1.8B). Therefore, EPA updated the GHG emission factors for fertilizer production used in our analysis to the values from the new GREET version. This had the result of slightly increasing the GHG emissions associated with fertilizer production and thus slightly increasing the GHG emission impacts of domestic agriculture.

As was the case in the proposal, we held the rates of domestic fertilizer application constant over time. This is true for both of our yield scenarios considered as well as for price induced yield increases. This constant rate of application is justified based on USDA data indicating that crops are becoming more efficient in their uptake of fertilizer such that higher yields can be achieved based on the same per acre fertilizer application rates.

N 2 O Emissions: The proposal analysis calculated N 2 O emissions from domestic fertilizer application and nitrogen fixing crops based on the amount of fertilizer used and different regional factors to represent the percent of nitrogen (N) fertilizer applied that result in N 2 O emissions. The proposal analysis N 2 O factors were based on existing DAYCENT modeling that was developed using the 1996 IPCC guidance for calculating N 2 O emissions from fertilizer applications and nitrogen fixing crops. We identified in the proposal that this was an area we would be updating for the final rule based on new analysis from Colorado State University using the DAYCENT model. This update was not available at time of proposal.

We received a number of comments on our proposal results indicating that the N 2 O emissions were overestimated from soybean and other legume production (e.g., nitrogen fixing hay) in our analysis. The main issue is that because the N 2 O emission factors used in the proposal were based on the 1996 IPCC guidance for N 2 O accounting they were overestimating N 2 O emissions from nitrogen fixing crops. As an update in 2006, IPCC guidance was changed such that biological nitrogen fixation was removed as a direct source of N 2 O because of the lack of evidence of significant emissions arising from the fixation process itself. IPCC concluded that the N 2 O emissions induced by the growth of legume crops/forages may be estimated solely as a function of the above-ground and below-ground nitrogen inputs from crop/forage residue. This change effectively reduces the N 2 O emissions from nitrogen fixing crops like soybeans and nitrogen fixing hay from the 1996 to 2006 IPCC guidance.

Therefore, as part of the update to new N 2 O emission factors from DAYCENT used for our final rule analysis we have updated to the 2006 IPCC guidance which reduces the N 2 O emissions from soybean production. This has the effect of reducing lifecycle GHG emissions for soybean biodiesel production. When we model corn expansion as would result from increased production of corn-based ethanol, one of the impacts is that the increase in corn acres displaces some acres otherwise planted to soy beans. Since the GHG emissions impact of this change in land use considers the N 2 O emissions benefit from the displaced soy, the result of this lower soy bean N 2 O assessment means that the benefits for soy displacement are less, directionally increasing the net GHG emissions for corn expansion.

We also received comments on our approach that we should use IPCC factors directly as opposed to relying on DAYCENT modeling. The difference is that IPCC provides default factors by crop by country, while DAYCENT models N 2 O emissions by crop but also by region within the US, accounting for different soil types and weather factors. For the final rule we still rely on the DAYCENT modeling results as we believe them to be more accurate. For example, the National Greenhouse Gas Inventory as reported annually by the US to the Framework Convention on Climate Change uses the DAYCENT model to determine N 2 O emissions from domestic fertilizer use as opposed to using default IPCC factors as the DAYCENT modeling is recognized to be a more accurate approach.

Livestock Emissions: GHG emissions from livestock have two main sources: enteric fermentation and manure management. For the proposal, enteric fermentation methane emissions were determined by applying IPCC default factors for different livestock types to herd values as calculated by FASOM to get GHG emissions. Comments we received on this approach were that the default IPCC factors do not account for the beneficial use of distiller grains (DGs) as animal feed. Use of DGs has been shown to decrease methane produced from enteric fermentation if replacing corn as animal feed. This is due to the fact that the DGs are a more efficient feed source. Consistent with our assumptions regarding the efficiency of DGs as an animal feed in our agricultural sector modeling, we have also included the enteric fermentation methane reductions of DGs use in our final rule analysis. The reduction amount was based on default factors in GREET that calculated this reduction based on the same Argonne report used to determine DGs feed replacement efficiency discussed in Section V.B.2.b.i. This resulted in a reduction in the lifecycle GHG emissions for corn ethanol compared to the proposal assumptions. More detail on the enteric fermentation methane reductions of DGs use can be found in Chapter 2 of the RIA.

The proposal analysis also included the methane and N 2 O emissions of livestock manure management based on IPCC default factors for emissions from the different types of livestock and management methods combined with FASOM results for livestock changes. We received comments that this was a good approach as it quantifies the indirect impacts of emissions associated with biofuel production. The same approach was used for the final rule analysis.

Methane from Rice: For the proposal, methane emissions from rice production were calculated by taking the FASOM output predicted changes in rice acres, resulting from the increase in biofuel production, and multiplying by default methane emission factors from IPCC to generate GHG impacts. We received comments that this was a good approach as it quantifies the indirect impacts of emissions associated with biofuel production. The same approach was used for the final rule analysis.

ii. International Agricultural Sector Impacts

Agricultural Sector Inputs: For the proposal we determined international fertilizer and energy use emissions based on applying input data collected by the Food and Agriculture Organization (FAO) of the United Nations and the International Energy Agency (IEA) to the FAPRI-CARD crop output data and then applied GREET defaults for converting those inputs to GHG emissions.

As part of our public comment and peer review process we had this component of our analysis specifically peer reviewed. The main comment we received was to update our input data with newer data sources. Therefore, for the final rule analysis we updated fertilizer and pesticide consumption projections from the incorporation of updates made by the FAO to its Fertistat and FAOStat datasets, as well as the incorporation of more up-to-date fertilizer consumption statistics provided by a recent International Fertilizer Institute (IFA) report. This update had varying impacts on the amount of fertilizer used on different crops in different countries but in general increased the amount of fertilizer assumed and thus international agriculture lifecycle GHG emissions from fertilizer use for all biofuels.

Another comment from the peer review was that we should include lime use for some of the key crops modeled in our analysis. Lime use was not included in the proposal because of lack of international data on lime use by crop. Excluding lime used is an underestimate of international agriculture GHG emissions. For our final rule analysis we included lime use for sugarcane production in Brazil based on information received from Brazilian agricultural experts provided as part of the comment process. This led to an increase in GHG emissions from sugarcane farming. We did not include lime use for other crops in the final rule analysis because of lack of other data sources for other crops.

Other comments we received on our approach were that we were potentially underestimating GHG emissions from international agriculture energy use. Our proposal based international agriculture energy use on factors from the International Energy Agency (IEA) that included all energy use for agriculture that we divided by all agricultural sector land by country to get a GHG emission per acre for each country considered. The comment raised the issue that by using all agricultural land this includes pasture land that would not have the same energy input as crop production. Effectively, higher energy use from crop production was getting averaged with lower energy use for pasture and then this lower number was applied only to crop production. We specifically asked as part of our peer review for guidance and comment on our international agriculture energy use calculation. We did not receive significant comments or data to suggest that we change our approach and reviewers generally agreed we were using the best data available. Furthermore, the energy use values represent all agriculture including forestry and fishing which could in some countries be overestimating energy use for crop production. So for our final rule analysis we used the same approach as for the proposal to calculate international agriculture energy use GHG emissions.

We also received comments on the applicability of applying GREET defaults for fuel and fertilizer production to international fuel and fertilizer use to generate GHG emissions. The comments noted that GREET factors are developed for domestic US conditions and would not necessarily apply internationally. Specifically on the issue of nitrogen fertilizer production, the comments indicated that nitrogen fertilizer production internationally could rely on coal as a fuel source as opposed to natural gas used in the US, which would cause international GHG emissions associated with fertilizer production and hence biofuel production to be underestimated in our analysis. This was also an area we asked peer reviewers for comment and guidance. The peer review response generally supported our approach and did not offer suggestions for other data sources. So for our final rule analysis we used the same approach as for the proposal and applied GREET defaults to calculate international fertilizer production GHG emissions.

As was the case in the proposal and for domestic agriculture, we held the rates of international fertilizer application constant over time. This is true for both of our yield scenarios considered as well as for price induced yield increases. This was an area that was specifically addressed in our peer review of International Agricultural Greenhouse Gas Emissions and Factors. The reviewers supported the approach we have taken, for example indicating that generally crop production as a unit of fertilizer application has increased over time, therefore, crop yields have increased with the same or lower fertilizer applications.

N 2 O Emissions: For the proposal we included N 2 O emissions from fertilizer application by applying IPCC default factors for different crops in different countries. We use IPCC default factors because we do not have the same level of regional factors like we do in the US from the DAYCENT model. The IPCC guidance has emission factors for four sources of N 2 O emissions from crops, Direct N 2 O Emissions from Synthetic Fertilizer Application, Indirect N 2 O Emissions from Synthetic Fertilizer Application, Direct Emissions from Crop Residues, and Indirect Emissions from Crop Residues. The proposal did not include N 2 O emissions from the Direct and Indirect Emissions from Crop Residues for cotton, palm oil, rapeseed, sugar beet, sugarcane, or sunflower. These were not included for these crops because default crop-specific IPCC factors used in the calculation were not available.

Comments from our peer review process suggested that we include proxy emissions from these crops based on similar crop types that do have default factors. Therefore, for our final rule analysis we have included crop residue N 2 O emissions from sugarcane production based on perennial grass as a proxy. Perennial grass is chosen as a proxy based on input from N 2 O modeling experts. This change results in an increase in N 2 O emissions from sugarcane and therefore sugarcane ethanol production compared to the proposal.

Livestock Emissions: Similar to domestic livestock impacts, enteric fermentation and manure management GHG emissions were included in our proposal analysis. The proposal calculated international livestock GHG impacts based on activity data provided by the FAPRI-CARD model (e.g., number and type of livestock by country) multiplied by IPCC default factors for GHG emissions.

Based on the peer review of the methodology used for the proposal it was determined that the calculations for manure management did not include emissions from soil application. These emissions were included for our final rule analysis but do not cause a significant change in the livestock GHG emission results.

Rice Emissions: To estimate rice emission impacts internationally, the proposal used the FAPRI-CARD model to predict changes in international rice production as a result of the increase in biofuels demand in the U.S. We then applied IPCC default factors by country to these predicted changes in rice acres to generate GHG emissions. We received comments that this was a good approach as it quantifies the indirect impacts of emissions associated with biofuel production. The same approach was used for the final rule analysis.

b. Land Use Change

The following sections discuss our final rulemaking assessment of GHG emissions associated with land use changes that occur domestically and internationally as a result of the increase in renewable fuels demand in the U.S. There are four main methodology questions addressed both domestically and internationally:

  • Amount of Land Converted and Where.
  • Type of Land Converted.
  • GHG Emissions Associated with Conversion.
  • Timeframe of Emission Analysis.

Each of those methodology components are discussed as are the comments we received as part of the comment and peer review process. We also outline in addition to our main FASOM and FAPRI-CARD approach a general equilibrium modeling approaches and its results.

i. Amount of Land Area Converted and Where

Based on a number of modeling changes made to the FASOM and FAPRI-CARD models since the NPRM, the amount of land use change resulting from an increase in biofuel demand in the U.S. is significantly lower in this FRM analysis for most renewable fuels. Many of the changes made were a direct result of comments received through the notice-and-comment period, comments received from the peer-reviewers, or as a result of incorporating new science that has become available since the analysis was conducted in the proposal. Some of the key changes that had the largest impact on the land use change estimates are included in this section. For additional information, see Chapter 2 of the RIA.

As discussed in the NPRM, one of the key factors in determining the amount of new land needed to meet an increase in biofuel demand is the treatment of co-products of ethanol and biodiesel production. We received many comments on this topic, particularly on the amount of corn and soybean meal a pound of DGS, the byproduct of dry mill grain ethanol production, can replace in animal feed. For the final rule, we predict that distiller grains will be absorbed by livestock more efficiently over time. We updated the displacement rate assumptions in the FASOM and FAPRI-CARD models based on comments we received and on the recent research conducted by Argonne National Laboratory and others. [167] According to this research, one pound of DGS replaces more than a pound of corn and/or soybean meal in beef and dairy rations, in part because cattle fed DGS show faster weight gain and increased milk production compared to those fed a traditional diet. While this study represents a significant increase over current DGS replacement rates, we believe it is reasonable to assume that improvements will be made in the use and efficiency of DGS over time as the DGS market matures, the quality and consistency of DGS improves, and as livestock producers learn to optimize DGS feed rations. As a result of this modification, less land is needed to replace the amount of corn diverted to ethanol production. Additional details on the DGS assumptions are included in Chapters 2 and 5 of the RIA.

A second factor that can have a significant impact on the amount of land that may be converted as a result of increasing biofuel demand are changes in crop yields over time. As discussed in the NPRM, our proposal based domestic yields on USDA projections for both the reference case and the control case. As discussed in Section V.B.1.c, for this FRM we have also included scenarios that use higher yield projections in both the reference case and the control case. However, in the NPRM we also requested comment on whether the higher prices caused by an increased in demand for biofuels would increase future yield projections in the policy case beyond the yield trends in the reference case (sometimes referred to as “price induced yields”), or whether these price induced yields would be offset by the reduction in yields associated with expanding production onto new marginal acres (sometimes referred to as extensification). Based on the comments we received, along with additional historical trend analysis conducted by FAPRI-CARD, the international agricultural modeling framework now incorporates a price induced yield component. [168] The new yield adjustments are partially offset by the extensification factor, however, the combined impact is that fewer new acres are needed for agricultural production to meet world agricultural demands.

One additional change we made to the yield assumptions was to update the FASOM model with new analysis by Pacific Northwest National Laboratories (PNNL) on switchgrass yields. [169] We included this new data for two reasons. First, we received several comments that our assumptions on switchgrass yields were too low, based on more recent field work. In addition, for out NPRM analysis, we did not have data for switchgrass yields in certain regions of the US. Therefore, the PNNL data helped to fill a pre-existing data gap. As a result of these updates, less land is needed per gallon of switchgrass ethanol produced. Additional details on switchgrass yields and other agricultural sector modeling assumptions are included in RIA Chapter 5.

One of the major changes made to the FAPRI-CARD model between the NPRM and FRM includes the more detailed representation of Brazil through a new integrated module. The Brazil module was developed by Iowa State with input from Brazilian agricultural sector experts and we believe it is an improvement over the approach used in the proposal. In the NPRM, we requested additional data for countries outside the U.S. We received comments encouraging us to use regional and country specific data where it was available. We also received comments encouraging us to take into account the available supply of abandoned pastureland in Brazil as a potential source of new crop land. The new Brazil module addresses these comments. Since the Brazil module contains data specific to six regions, this additional level of details allows FAPRI-CARD to more accurately capture real-world responses to higher agricultural prices. For example, double cropping (the practice of planting a winter crop of corn or wheat on existing crop acres) is a common practice in Brazil. Increased double cropping is feasible in response to higher agricultural prices, which increases total production without increasing land use conversion. The new Brazil module also explicitly accounts for changes in pasture acres, therefore accounting for the competition between crop and pasture acres. Furthermore, the Brazil module explicitly models livestock intensification, the practice of increasing the number of heads of cattle per acre of land in response to higher commodity prices or increased demand for land.

In addition to modifying how pasture acres are treated in Brazil, we also improved the methodology for calculating pasture acreage changes in other countries. We received several comments through the public comment period and peer reviewers supporting a better analysis of the interaction between crops, pasture, and livestock. In the NPRM, although we accounted for GHG emissions from livestock production (e.g., manure management), we did not explicitly account for GHG emissions from changes in pasture demand. In response to comments received, our new methodology accounts for changes in pasture area resulting from livestock fluctuations and therefore captures the link between livestock and land used for grazing. Based on regional pasture stocking rates (livestock per acre), we now calculate the amount of land used for livestock grazing. The regional stocking rates were determined with data on livestock populations from the UN Food and Agricultural Organization (FAO) and data on pasture area measured with agricultural inventory and satellite-derived land cover data. As a result of this change, in countries where livestock numbers decrease, less land is needed for pasture. Therefore, unneeded pasture acres are available for crop land or allowed to revert to their natural state. In countries where livestock numbers increase, more land is needed for pasture, which can be added on abandoned cropland or unused grassland, or it can result in deforestation. We believe this new methodology provides a more realistic assessment of land use changes, especially in regions where livestock populations are changing significantly. For additional information on the pasture replacement methodology, see RIA Chapter 2.

Although the total amount of land use conversion is lower in the FRM analysis compared to the NPRM analysis, the regional distribution of this land use change has shifted. Due to the many changes made in response to comments associated with agriculture and livestock markets, Brazil is now much more responsive to changes in world biofuel and agricultural product demand. As a result, a larger portion of the projected land use change occurs in Brazil compared to the NPRM analysis. Additional details on the geographical location of land use change are included in Chapter 2 of the RIA.

ii. Type of Land Converted

Based on a number of improvements in our analysis, the types of land affected by biofuel-induced tend to be less carbon intensive compared to the NPRM. Therefore, the net effect of our revisions to this part of our analysis significantly reduced land use change GHG emissions. The updated FAPRI-CARD Brazil model, discussed in the previous section, showed more pasture expansion in the Amazon which increased land use change emissions. However, the most important revisions to this part of our international analysis, in terms of their net effect on GHG emissions, were improvements that we made in our modeling of the interactions between livestock, pasture, crops and unused, or underutilized, grasslands globally. In the NPRM we made the broad assumption that international crop expansion would necessarily displace pasture, which would require an equivalent amount of pasture to expand into forests and shrublands. In the FRM analysis as discussed in the previous section, we have linked international changes in livestock production with changes in pasture area to allow for pasture abandonment in regions where livestock production decreases as a result of biofuel production. We also incorporated the ability of pasture to expand onto unused, or underutilized, grasslands and savannas which on a global basis reduced the amount of forest conversion compared to the proposal. These revisions, as well as a quantitative uncertainty assessment, are discussed in this section.

In the same way that the amount and location of land use change is important, the type of land converted is also a critical determinant of the magnitude of the GHG emissions impacts associated with biofuel production. For example, the conversion of rainforest to agriculture results in a much larger GHG release than conversion of grassland. In the proposed rule analysis we used two approaches, based on the best available information to us at the time, to evaluate the types of land that would be affected domestically and internationally. Domestically, we used the FASOM model, which simulates rental rates for different types of land (e.g., forest, pasture, crop) and chooses the land uses that would produce the highest net returns. Internationally, we used the FAPRI-CARD/Winrock analysis whereby historical land conversion trends, as evaluated with satellite imagery, are used to determine what types of land are affected by agricultural land use changes in each country or sub-region.

In the proposed rule we also explained several other options to determine what types of land will be affected by biofuel-induced land use changes, such as the use of general equilibrium models. EPA specifically sought expert peer review input and public comment on our approach and all of the analytical options for this part of the lifecycle assessment. The expert peer reviewers agreed that EPA's approach was scientifically justifiable, but they highlighted problematic areas and suggested important revisions to improve our analysis. The public comments received on this issue expressed a wide range of views regarding EPA's approach. In general, the commenters that objected to our analytical approach raised similar concerns as the peer reviewers, such as the need for more data validation and uncertainty assessment. As discussed below, we made significant improvements to our analysis based on the recommendations and comments we received. Based on the peer reviewers agreement that our general approach is scientifically justifiable, and in light of the significant improvements made, we think that our approach represents the best available analysis of the types of land affected by biofuel-induced land use changes. We did consider a range of other analytical options, but based on all of the information considered and the requirements for this analysis, we did not find any alternative approaches that are superior at this time. As part of periodic updates to the lifecycle analysis, we will continue to consider ways to improve this part of our analysis, as well as the merits of alternate approaches.

Domestic: In response to comments received, we made two major improvements to the FASOM model for the final rulemaking. As discussed in the NPRM and supported by comments, we were able to include the forestry sector into the FASOM analysis. Only the agricultural sector of FASOM was analyzed for the NPRM, due to the fact that the forestry sector component was undergoing model modifications. For this FRM analysis, we were able to use the fully integrated forestry and agricultural sector model, thereby capturing the interaction between agricultural land and forests in the U.S. In addition, the inclusion of the forestry model allows us to explicitly model the land use change impacts of the competing demand for cellulosic ethanol from agricultural sources with cellulosic ethanol from logging and mill residues. As a result of this modification, the FRM analysis includes some land use conversion from forests into agriculture in the U.S. as a result of the increased demand for renewable fuels.

The second major modification we made in response to comments was the disaggregation of different types of land included in FASOM. In the proposed rulemaking, the FASOM model included three major categories of land: cropland, pasture, and acres enrolled in the Conservation Reserve Program (CRP). Although this categorization allowed for a detailed regional analysis of land used to grow crops, acres used for livestock production were not fully captured. We received comments requesting a more detailed breakdown of land types in order to capture the interaction between livestock, pasture, and cropland. Therefore, the FASOM model now includes rangeland, pasture and forest land that can be used for grazing. Since we also received comments that we should take into account the potential for idle land to be used for other purposes such as the production of cellulosic ethanol, FASOM now accounts for the amount of land within each category that is either idle or used for production.

These two major modifications to the FASOM model now allow us to explicitly track land transfers between various land categories in the U.S. As a result, we can more accurately capture the GHG impacts of different types of land use changes domestically. More detail and results of the FASOM model can be found in Section V.B.1.b of the preamble.

International: The proposed rule included a detailed description of the FAPRI-CARD/Winrock approach used to determine the type of land affected internationally. This approach uses satellite data depicting recent land conversion trends in conjunction with economic projections from the FAPRI-CARD model (an economic model of global agricultural markets) to determine the type of land converted internationally. In the proposed rule we described areas of uncertainty in this approach, illustrated the uncertainty with sensitivity analyses, and discussed other potential approaches for this analysis. To encourage expert and stakeholder feedback, EPA specifically invited comment on this issue, held public hearings and workshops, and sponsored an independent peer-review, all of which specifically highlighted this part of our analysis for feedback. While there were a wide range of views expressed in these forums, the feedback received by the Agency generally supported the FAPRI-CARD/Winrock approach as appropriate for this analysis. For example, all five experts that peer reviewed EPA's use of satellite imagery agreed that it is scientifically justifiable to use historic remote sensing data in conjunction with agricultural sector models to evaluate and project land use change emissions associated with biofuel production. Additionally, the peer reviewers and public commenters highlighted problematic areas and suggested revisions to improve our analysis. Below, we describe the key revisions that were implemented which have significantly improved our analysis based on the feedback received.

FAPRI-CARD/Satellite Data Approach: As described above in Section V.B.1.b, the FAPRI-CARD model was used to determine the amount of land use change in each country/region in response to increased biofuel production. Because the FAPRI-CARD model does not provide information about what type of land is converted to crop production or pasture, we worked with Winrock International to evaluate the types of land that would be affected internationally. Winrock is a global nonprofit organization with years of experience in the development and application of the IPCC agricultural forestry and other land use (AFOLU) guidance. For the proposed rule, we used satellite data from 2001-2004 to provide a breakdown of the types of land converted to crop production. A key strength of this approach is that satellite information is based on empirical observations which can be verified and statistically tested for accuracy. Furthermore, it is reasonable to assume that recent land use change decisions have been driven largely by economics, and, as such, recent patterns will continue in the future, absent major economic or land use regime shifts caused, for example, by changes in government policies.

As discussed above, all five of the expert peer reviewers that reviewed our use of satellite imagery for this analysis agreed that our general approach was scientifically justifiable. However, all of the peer reviewers qualified that statement by describing relevant uncertainties and highlighting revisions that would improve our analysis. Some of the public commenters supported EPA's use of satellite imagery, while other expressed concern. In general, both sets of public commenters—those in favor and opposed—outlined the same criticisms and suggestions as the expert peer reviewers. Among the many valuable suggestions for satellite data analysis provided in the expert peer reviews and public comments, several major recommendations emerged: EPA should use the most recent satellite data set that covers a period of at least 5 years; EPA should use higher resolution satellite imagery; EPA's analysis should consider a wider range of land categories; EPA should improve it's analysis of the interaction between cropland, pasture and unused or underutilized land; and EPA's analysis should include thorough data validation and a full assessment of uncertainty. Below, we describe these and other recommendations and how we addressed each of them to improve our analysis. Based on the peer reviewers agreement that our general approach is scientifically justifiable, and in light of the significant improvements made, we think that our approach represents the best available analysis of the types of land affected internationally.

One of the fundamental improvements in this analysis since the proposed rule is that it now provides global coverage. The analysis for the proposed rule included satellite imagery for 6 land categories in 314 regions across 35 of the most important countries, with a weighted average applied to the rest of the world. We have since completed a global satellite data analysis including 9 land categories in over 750 distinct regions across 160 countries. This was an analytical improvement that we committed to do in the proposed rule. As described below, the other major analytical enhancements were conducted in response to the many technical recommendations that we received as part of the peer review and public comment process.

All of the expert peer reviewers agreed that the version 4 MODIS data set used in the proposed rule, which covers 2001-2004 with one square-kilometer (1km) spatial resolution, was appropriate for our analysis given the goals of the study at the time. However, almost all of the reviewers strongly recommended using a data set covering a longer time period. The reviewers argued that the 3-year time period from 2001-2004 was too short to capture the often gradual, or sequential, cropland expansion that has been observed in the tropics. The short time period may also show unusual or temporary trends in land use caused by short-term policy changes or market influences. The reviewers suggested that remote sensing observations covering 5-10 years would be adequate to address these problems. The reviewers also recommended that remote sensing observations should be as recent as possible in order to capture current land use change drivers and patterns (e.g., political systems, infrastructure, and protected areas). To use the best available data and respond to the peer reviewers' recommendations, the analysis was updated to include the most recent MODIS data set, version 5, which covers the time period 2001-2007. MODIS land cover products are not available for years prior to 2001, so it is not currently possible to analyze a time period longer than six years (i.e., 2001-2007) with a single, or consistent, data set. Thus, consistent with the peer review recommendations, we are now using the most recent global data set which covers at least 5 years. There are other advantages to using the version 5 MODIS data, such as improved spatial resolution, and robust data validation, which are discussed below.

There was strong agreement among the peer reviewers that higher resolution satellite imagery would be an important improvement over the 1-km resolution data used in the proposed rule analysis. Higher spatial resolution is especially useful in categorizing highly fragmented landscapes. One of the reviewers hypothesized that land use change driven by biofuel production would likely involve large parcels of land, and thus 1-km resolution may be sufficient. However, all of the reviewers agreed that higher resolution data would be preferable. A number of the peer reviewers specifically said that the version 5 MODIS data set, with 500 meter resolution, would be adequate. With four-times higher spatial resolution than version 4, the peer reviewers anticipated that the 500m imagery would classify less area of “mixed class” land, thus providing a more detailed representation of the land in that category. Consistent with the peer reviewer's recommendations and with our goal to use the best available information, our analysis was updated with the higher resolution version 5 MODIS data.

Related to the issue of spatial resolution, the peer review experts were asked whether they would recommend augmenting our global analysis with even higher resolution data for specific regions where there is a high degree of agricultural land use change. All of the peer reviews agreed that this type of analysis would be worthwhile. In response to this recommendation, we analyzed select geographic regions (e.g., Brazil, India) with the higher resolution 30m Landsat data set covering 2000-2005. The Landsat data set does not currently provide global coverage, thus it was not an option for use in the full analysis; instead, it was used as a way to check/validate the appropriateness of the version 5 MODIS imagery. In general, the higher resolution data showed similar land use change patterns as the MODIS data. The results of this analysis are discussed further in Chapter 2 of the RIA.

Another issue that we invited comments on was the re-classification of the MODIS data from 17 land cover categories into 6 aggregated categories (e.g., open and closed shrubland were both re-classified as shrubland). The category aggregation was intended to remove unnecessary complexity from the analysis. All five expert reviewers agreed that the methodology used to re-classify land cover categories using International Geosphere-Biosphere Programme (IGBP) land definitions was sound; however, the reviewers recommended inclusion of more than 6 aggregated land categories. The reviewers specifically recommended the addition cropland/natural vegetation mosaic, permanent wetlands, and barren or sparsely vegetated land, all of which are now included in our analysis. Consistent with these recommendations, there are 9 aggregate land categories in our revised analysis: barren, cropland, excluded (e.g., urban, ice, water bodies), forest, grassland, mixed (i.e., cropland/natural vegetation mosaic), savanna, shrubland and wetland. These land cover categories capture all significant types of land affected by agricultural land use changes. As described below in Section V.B.2.b.iii, we also estimated carbon sequestrations for all of these land categories. The impact of adding these land categories to our analysis is discussed further in RIA Chapter 2.

Another important addition to our analysis was consideration of the types of land affected by changes in pasture area, and the interaction of pasture land with cropland. In the proposed rule, we made a broad assumption that the total land area used for pasture would stay the same in each country or region. Thus, in the proposed rule, we assumed that any crop expansion onto pasture would necessarily require an equal amount of pasture to be replaced on forest or shrubland. We received a large number of comments questioning these assumptions, and the expert peer reviewers encouraged us to develop a better representation of the interactions between cropland and pasture land. As described above in Section V.B.2.6.i, the results from the FAPRI-CARD model are now used to determine pasture area changes in each country or region. In regions where we project that pasture and crop area both increase, the land types affected by pasture expansion are determined using the same analysis used for crop expansion. This new approach accounts for the ability of pasture to expand on to previously unused, or underutilized, grasslands and savanna. In regions where we project that crop and pasture area will change in opposite directions (e.g., crop area increases and pasture decreases) we assume that crops will expand onto abandoned pasture, and vice versa. Our analysis also now accounts for carbon sequestration resulting from crop or pasture abandonment. We used our satellite analysis, which shows the dominant ecosystems and land cover types in each region, to determine which types of ecosystems would grow back on abandoned agricultural lands in each region. More information about our analysis of pasture and abandoned agricultural land are provided in RIA Chapter 2.

A sub-set of the expert peer reviewers recommended combining the historic satellite imagery with other information on land use change drivers (e.g., transportation infrastructure, poverty rates, opportunity costs) as an additional means to estimate the types of land affected. Consideration of these types of information could potentially address two conceptual issues with the use of satellite imagery in this analysis: First, biofuel-induced land use change could affect different types of land than the generic agricultural expansion captured by the historic data; and second, future land use change patterns may differ from historic patterns. Our concerns with the first issue are allayed to some degree by one of the peer reviewers who observed, “While it is theoretically possible that the changes in land use resulting from biofuel production occur in ecosystems or regions that would not be the ones affected by other drivers, this doesn't appear very likely.” [170] Furthermore, the economic drivers of land use change are to a large degree captured by the economic models that are used in our analysis. For example, the FAPRI-CARD model considers economic drivers in its projections of where and how much crop production will change as a result of specifically biofuel-induced changes. The second issue is also addressed to some degree by the FAPRI-CARD model which includes baseline forecasts of future international agricultural, economic and demographic conditions. Furthermore, as discussed above, we used the most recently available satellite data sets in order to capture the most current land use change drivers. Thus, while we think that these issues are currently addressed to a scientifically justifiable degree for the purposes of this analysis, we recognize that these are areas for future investigation, and we have tried to capture the uncertainty from these factors in uncertainty and sensitivity analyses as described below.

While EPA has made significant improvements to the methodology in response to peer review comments, the use of satellite data for forecasting land use changes is a key area of uncertainty in the analysis. To facilitate substantive comments on the impact of uncertainty in international land use changes, and how to address the uncertainty, the proposed rule highlighted areas of uncertainty and included multiple sensitivity analyses. For example, we presented a range of lifecycle results assuming at the high-end that all land conversion caused deforestation and at the low-end that biofuels would cause no deforestation. Further, EPA sought input on this issue in public hearings and workshops, and expert feedback through the independent peer review. The feedback we received, both from experts and the public, overwhelmingly supported a more systematic analysis of the uncertainty in using satellite data to project biofuel-induced land use change patterns. Additionally, commenters recommended more data validation, especially regarding the satellite imagery. To respond to these comments, we incorporated satellite imagery validation and conducted a Monte Carlo analysis of the MODIS satellite data using assessments provided by NASA to quantitatively evaluate the uncertainty in our application of satellite imagery.

One benefit of using the MODIS data set is that it is routinely and extensively validated by NASA's MODIS land validation team. NASA uses several validation techniques for quality assurance and to develop uncertainty information for its products. NASA's primary validation technique includes comparing the satellite classifications to data collected through field and aircraft surveys, and other satellite data sensors. The accuracy of the version 5 MODIS land cover product was assessed over a significant set of international locations, including roughly 1,900 sample site clusters covering close to 150 million square kilometers. The results of these validation efforts are summarized in a “confusion matrix” which compares the satellite's land classifications with the actual land types observed on the ground. We used this information to assess the accuracy and systematic biases in the published MODIS data. In general, the validation process found that MODIS version 5 was quite accurate at distinguishing forest from cropland or grassland. However, the satellite was more likely; for example, to confuse savanna and shrubland because these land types can look quite similar from space.

Using the data validation information from NASA about which types of land MODIS tends to confuse which each other, our Monte Carlo analysis was able to account for systematic misclassifications in the MODIS data set. Therefore, part of the Monte Carlo analysis can be viewed as a way to correct and reduce the inaccuracies in the MODIS data. After this correction is performed, the uncertainty in the satellite data is no longer solely a function of the accuracy of the satellite. Instead, the sizes of the standard errors for each classification are also a function of the sample sizes in the data validation exercise. For example, if NASA validated every pixel on Earth, the corrected data set would be 100% accurate, even if the original satellite data were only 50% accurate. Similarly, although NASA reports that the overall accuracy of the MODIS version 5 land cover data set is approximately 75%, the standard errors after the Monte Carlo procedure are less than 5% for each aggregate land category. These standard errors were used to quantify the uncertainty added by the satellite data used in our analysis. This procedure and the results are described in more detail in Chapter 2 of the RIA.

It should be noted that our assessment of satellite data uncertainty did not try to fully quantify the uncertainty of using historical data to make future projections about the types of land that would be affected internationally. As noted above, we think it is reasonable to assume that in general, recent land use change patterns will continue in the future absent major economic or land use regime shifts caused, for example, by changes in government policies. Thus, our uncertainty assessment provides a reasonable estimate of the variability in land use change patterns absent any fundamental shifts in the factors that affect land use patterns. However, our uncertainty assessment does not attempt to fully quantify the probability of major shifts in land use regimes, such as the implementation of effective international policies to curb deforestation.

Some of the peer reviewers recommended a satellite imagery analysis approach known as change detection, instead of the “differencing” approach used in the Winrock analysis. However, there was disagreement among the peer reviewers on this point, with one peer reviewer saying that thematic differencing between land cover maps generated for two specific dates, as conducted in this study, provides the best approach for detecting and analyzing land use pattern changes globally. In general terms, the differencing method employed by Winrock compared global land cover maps from 2001 and 2007 to evaluate the pattern of land use change during this period. Thus, the differencing method shows all of the land that changed categories, as well as all of the land that stayed the same over this period. For change detection, instead of using comprehensive land cover maps, the data set only shows land categories that changed. One advantage of change detection is that it is better suited to capture the sequential nature of land use changes, e.g., a forest could be converted to savanna, then grassland and then cropland. The differencing method that we employed lends itself more readily to comprehensive global analysis, data validation, and uncertainty assessment. Given the timeframe and priorities for our analysis, we think that the differencing method provides the best approach available at this time. However, we will continue to consider alternative analytical techniques, such as change detection, for use as part of periodic updates to this analysis.

Some of the peer reviewers recommended additional alternative technical approaches for satellite data and land use change analysis. For example, some of the reviewers recommended the use of satellite imagery to identify specific crop-types and rotations, and one reviewer suggested that EPA develop a new interactive spatial model. The Summary and Analysis of Comments document includes discussion of these and other technical comments and recommendations that are not covered here.

iii. GHG Emissions Associated With Conversion

(1) Domestic Emissions

GHG emissions impacts due to domestic land use change are based on GHG emissions the FASOM model generates in association with land type conversions projected in the model. In the proposed rule analysis, estimates of land use change emissions were limited to conversion between different types of agricultural land (e.g., cropland, fallow cropland, pasture). The analysis did not allow for the addition of new domestic agricultural land.

In response to feedback EPA received during the public comment period and based on commitments EPA made in the NPRM, several changes and additions have augmented the analysis of domestic land use change GHG emissions since the proposed rule analysis. The addition of the forest land types and the interaction between cropland, pastureland, forestland, and developed land to the FASOM model provides a more complete emissions profile due to domestic land use change (see Section V.B.4.b.ii). We have updated soil carbon accounting based on new available data. Lastly, the methodology now captures GHG emission streams over time associated with discrete land use changes.

For agricultural soils, FASOM models GHG emissions associated with changes in crop production acreage and with changes in crop type produced. FASOM generates soil carbon factors for cropland and pasture according to IPCC Agriculture, Forestry, and Other Land Use (AFOLU) Guidelines. In the proposed rule, we committed to updating FASOM soil carbon accounting for agriculture. Per our commitment, we have updated FASOM soil carbon accounting for cropland and pasture using the latest DAYCENT modeling from Colorado State University.

In the proposed rule, EPA committed to incorporate the forestry sector and the GHG emission impacts due to the land use interactions between the domestic agricultural and forestry sectors into the FASOM analysis. We received comment supporting the incorporation of the forestry sector. By including the forestry sector in the FASOM domestic model (see Section V.B.4.b.ii), we have incorporated GHG emission impacts associated with change in forest above-ground and below-ground biomass, forest soil carbon stocks, forest management practices (e.g. timber harvest cycles), and forest products and product emission streams over time. Forest carbon accounting in FASOM is based on the FORCARB developed by the U.S. Forest Service and on data derived largely from the U.S. Forest Service RPA modeling system.

With the changes to FASOM discussed above, we also updated the final calculation method of domestic land use change GHG emissions to account for FASOM's cumulative assessment of GHG emissions and the continuous (rather than discrete) nature of soil carbon and forest product emissions. For each category of agricultural and forestry land use emissions, we calculated the mean cumulative emissions from the initial year of FASOM modeling (2000) to 2022. Changes in agricultural and forest soil carbon and forest products have a stream of GHG emissions associated with them in addition to the initial pulse associate with a discrete instance or year of land use change. For each of these categories FASOM calculates the emissions over time associated with the mean land use change over a year. We included in total domestic land use change emissions the annualized emission streams associated with all agricultural soil, forest soil, and forest product changes included in the mean cumulative emissions (2000-2022) for 30 years after 2022.

(2) International Emissions

Based on input from the expert peer review and public comments, we incorporated new data sources and made other methodological improvements in our estimates of GHG emissions from international land conversions. Some of these modifications increased land use change GHG emissions compared to the NPRM, such as the consideration of carbon releases from drained peat soils. Other modifications, such as more conservative foregone sequestration estimates, tended to decrease land use change GHG emissions. For example, our estimates of emissions per acre of deforestation in Brazil tended to increase because of improved data on forest biomass carbon stocks in that region. However, for example, our deforestation estimates in China decreased, in part because of new data on foregone forest sequestration. The net effect of the revisions varied depending on the location and types of land use changes in each biofuel scenario. The major changes to this part of our analysis, including a quantitative uncertainty assessment, are discussed in this section.

To determine the GHG emissions impacts of international land use changes, we followed the 2006 IPCC Agriculture, Forestry, and Other Land Use (AFOLU) Guidelines. [171] We worked with Winrock, which has years of experience developing and implementing the IPCC guidelines, to estimate land conversion emissions factors, including changes in biomass carbon stocks, soil carbon stocks, non-CO 2 emissions from clearing with fire and foregone forest sequestration (i.e., lost future growth in vegetation and soil carbon). In addition to seeking comment on our analysis in the proposed rule, EPA organized public hearings and workshops, and an expert peer review specifically eliciting feedback on this part of the lifecycle analysis. All of the expert peer reviewers generally felt that our analysis followed IPCC guidelines and was scientifically justifiable; however, they did make several suggestions of new data sources and recommended areas that could benefit from additional clarification. Based on the detailed comments we received, we worked with Winrock to make a number of important revisions, which have significantly improved this part of our analysis.

The proposed rule analysis included land conversion emissions factors for 5 land categories in 314 regions across 35 of the most important countries, with a weighted average applied to the rest of the world. We augmented this analysis to provide global coverage, including emissions factors for 10 land categories in over 750 regions across 160 countries. Other significant improvements included incorporation of new data sources, emissions factors for peat soil drainage, sequestration factors for abandoned agricultural land, and a full uncertainty assessment considering every data input.

Another significant improvement in our analysis was incorporation of higher resolution soil carbon data. One of the expert peer reviewers commented that the weakest part of EPA's international emissions factor analysis for the proposed rule was the global soil carbon map that was used because of its coarse resolution. To address this comment, we incorporated the new Harmonized World Soil Database, released in March 2009. This dataset provides one square kilometer spatial resolution, which is a major improvement compared to the proposed rule analysis. This dataset also includes an updated soil map of China that the peer reviewers recommended. Using this updated soil carbon data, the change in soil carbon following conversion of natural land to annual crop production was estimated following the 2006 IPCC guidelines. When land is plowed in preparation for crop production the soil loses carbon over time until a new equilibrium is established. To calculate soil carbon emissions the IPCC approach considers both tillage practices and agricultural inputs. Some of the peer reviewers expressed concern with our annual soil carbon change estimates, which assumed a constant rate of change over 20 years. However, for analytical timeframes greater than 20 years, such as used in our lifecycle analysis, the peer reviewers agreed that the our approach was scientifically justifiable. More information about soil carbon stock estimates is available in Chapter 2 of the RIA.

The expert peer reviewers generally agreed that EPA's estimate of forest carbon stocks followed IPCC guidelines and used the best available data. They did, however, recommend that the analysis could be updated with improved forest biomass maps as they become available. Consistent with these suggestions, we incorporated improved forest biomass maps for regions where they were available. More information about the specific data sources used is available in RIA Chapter 2.

In addition to estimating forest carbon stocks for each region, EPA's analysis also includes estimates of annual forest carbon uptake. When a forest is cleared the future carbon uptake from the forest is lost; this is known as foregone forest sequestration. In the proposed rule, to estimate annual forgone forest sequestration, we used IPCC default data for the growth rates of forests greater than 20 years old. The expert peer reviewers noted that these estimates could be refined with more detailed information from the scientific literature. Many of the public commenters were also concerned that EPA's approach overestimated foregone sequestration because it did not adequately account for natural disturbances, such as fires and disease. To address these comments, our analysis has been updated with peer reviewed studies of long-term growth rates for both tropical and temperate forests. These estimates are based on long-term records (i.e., monitoring stations in old-growth forests for the tropics and multi-decadal inventory comparisons for the temperate regions) and reflect all losses/gains over time. These studies show that the old-growth forests in the tropics that many once assumed to be in “steady state” (i.e., carbon gains equal losses) are in fact still gaining carbon. In summary, our analysis now includes more conservative foregone forest sequestration estimates that account for natural gains and losses over time. More information about these estimates is provided in RIA Chapter 2.

Another consideration when estimating GHG emissions resulting from deforestation is that some of the wood from the cleared forest can be harvested and used in wooden products, such as a table, that retain biogenic carbon for a long period of time. Some commenters argued that consideration of the use of harvested wood in products would decrease land use change emissions and reduce the impacts of biofuel production. As part of analysis for the proposed rule, we investigated the share of cleared forest biomass that is typically used in harvested wood products (HWP). However, we did not account for this factor in the proposed rule after it was determined that HWP would have a very small impact on the magnitude of land use change emissions. A number of commenters expressed concern that we did not account for HWP, and they argued that HWP would be more significant than we had determined. However, in response to specific questions on this topic, all of the expert peer reviewers agreed that EPA had properly accounted for HWP and other factors (e.g., land filling) that could prevent or delay emissions from land clearing. One of the peer reviewers noted that forests converted to croplands are generally driven by interests unrelated to timber, and thus the trees are simply burned and exceptions are probably of minor importance. To study this issue further, we looked at FAO timber volume estimates for 111 developing countries, and published literature on the share of harvested timber used in wood products and the oxidation period for wood products, such as wood-based panels and other industrial roundwood. Consistent with the peer reviewers' statements, our analysis concluded that even in countries with high rates of harvested timber utilization, such as Indonesia, a very small share of harvested forest biomass would be sequestered in HWP for longer than 30 years. The details of our HWP analysis are discussed further in RIA Chapter 2. This is an area for further work, but based on our analysis, and the feedback from expert commenters, we do not expect that consideration of HWP would have a significant impact on the magnitude of GHG emissions from international deforestation in our analysis. Furthermore, the range of outcomes from consideration of HWP is indirectly captured in our assessment of forest carbon stock uncertainty, which is described below.

The land conversion emissions estimates used in our analysis consider the carbon stored in crop biomass. In the proposed rule, we used the IPCC default biomass sequestration factor of 5 metric tons of carbon per hectare for annual crops, and applied this value to all crops globally. The final rule analysis now distinguishes between annual and perennial crops, with separate sequestration estimates for sugarcane and oil palm determined from the scientific literature. The peer reviewers suggested approaches to refine our biomass carbon estimates for different types of annual crops, e.g., for corn versus soybeans. However, we determined that adding crop-specific biomass sequestration estimates would have a very small impact on our results, because in general annual cropland carbon stocks range only from 3 to 7 tons per hectare and the average would likely be very close to the IPCC default factor currently applied. This is an area for future work, but we are confident that it would have very small impact. Furthermore, the range of potential outcomes is captured in the uncertainty analysis described below.

Other issues that were covered in the expert peer review and public comments included EPA's carbon stock estimates for grasslands, savanna, shrublands and wetlands, and our assumptions about which regions use fire to clear land prior to agricultural expansion. There is less data available for these parameters relative to some of the other issues discussed above, e.g., forest carbon stocks. Therefore, we worked to use expert judgment to derive global estimates for these parameters. In general, the peer reviewers thought that EPA's approach to these issues was reasonable and scientifically justifiable. Some of the peer reviewers recommended more resource-intensive techniques to refine some of our estimates. For example, regarding the issue of clearing with fire, one of the peer reviewers suggested that we could review fire events in the historical satellite data to estimate where fire is most commonly used. We carefully considered these suggestions, but did not make significant revisions to our analysis of these issues. Our review concluded that given the timeframe and goals of our analysis, the approach used in the proposed rule was most appropriate. We recognize that these are areas for future work, and we will consider new data as part of periodic updates. Furthermore, our uncertainty analysis, described below, considered the fact that these are areas where less data is available.

Other improvements in our analysis included the addition of emissions from peat soil drainage in Indonesia and Malaysia, and sequestration factors for abandoned agricultural land. Consistent with the expert peer reviewers' recommendations, we considered a number of recent studies to estimate average carbon emissions when peat soils are drained in Indonesia and Malaysia (the countries where peat soil is sometimes drained in preparation for new agricultural production). To estimate annual sequestration on abandoned agricultural land we used our foregone sequestration estimates and other data from IPCC. More information about these estimates is available in RIA Chapter 2.

As discussed in Section V.A.2, the uncertainty of land use change emissions is an important consideration in EPA's threshold determinations as part of this rulemaking. We conducted a full assessment of the uncertainty in international land use change emissions factors consistent with 2006 IPCC guidance. [172] This analysis considers the uncertainty in the every parameter used in our emissions factor estimates. Standard deviations for each parameter were estimated based on the quality and quantity of the underlying data. For example, in our analysis the standard errors (as a percent of the mean) tend to be smallest for forest carbon stocks in Brazil, because a large amount of high quality/resolution data was considered to estimate that parameter. Standard errors are largest for parameters that were estimated by scaling other data, or applying IPCC defaults, e.g., savanna carbon stocks in Yemen. More detail about our estimate of parameter uncertainty is available in RIA Chapter 2.

Following IPCC guidance, the uncertainties in the individual parameters of an emission factor can be combined using either error propagation methods (IPCC Tier 1) or Monte Carlo simulation (IPCC Tier 2). We used the Tier 2 Monte Carlo simulation method for this analysis. Monte Carlo is a method for analyzing uncertainty propagation by randomly sampling from the probability distributions of model parameters, calculating the results of the model from each sample, and characterizing the probability of the outcomes. An important consideration for Monte Carlo analysis is the treatment of correlation, or dependencies, among parameter errors. Strong positive correlation among parameter errors will result in greater overall uncertainty. As a simplified example, if the errors in our forest carbon stock estimates are positively correlated, then if we are overestimating forest carbon in one region we are likely overestimating forest carbon in every region. We worked with Winrock to estimate the degree of correlation among variables—both the correlation of one variable across space as well as the correlation of one variable to any others used in the analysis. This was done by considering dependencies in the underlying data used to estimate each parameter. For example, our forest carbon stock estimates are correlated across Russia because they were derived from one biomass map covering Russia. However, forest carbon stocks in Russia are not correlated with China, because they were derived from separate biomass maps. This partial correlation approach tended to reduce the overall uncertainty associated with GHG emissions factor data.

The information about the uncertainty in each parameter and the degree of correlation across parameters was utilized in Monte Carlo analysis to determine the overall uncertainty in our emissions factor estimates. We used the Monte Carlo simulation to combine the emissions factor and satellite data uncertainty for every biofuel scenario analyzed. Uncertainty ranges varied across scenarios depending on the types and locations of land use changes. For example, based on the sources of uncertainty analyzed, the 95% confidence range for land use change emissions (as a percent of the mean) was −27% to +32% for base yield corn ethanol in 2022, and −56% to +76% for base yield soy biodiesel in 2022. [173] More details about this uncertainty analysis are provided in RIA Chapter 2.

iv. Timeframe of Emission Analysis

Based on input from the expert peer review and public comments, EPA has chosen to analyze lifecycle GHG emissions using a 30 year time period, over which emissions are not discounted, i.e., a zero discount rate is applied to future emissions. The input we received and the reasons for our use of this approach are described in this section.

As required by EISA, EPA must determine whether biofuels reduce GHG emissions by the required percentage relative to the 2005 petroleum baseline. In the proposal the Agency discussed a number of accounting methods for capturing the full stream of GHG emissions and benefits over time. When accounting for the time profile of lifecycle GHG emissions, two important assumptions to consider are: (1) The time period considered and (2) the discount rate (which could be zero) applied to future emissions streams. At the time of proposal, EPA requested public comment on the choice of time frames and discounting approaches for purposes of estimating lifecycle GHG emissions. Also, as part of the peer review process, EPA requested comment from expert peer reviewers on the choice of the appropriate time frames and discount rates for the RFS2 analysis. Below is a summary of the comments we received on these issues and how we address them in our analytical approach.

Time Period for Analysis: In the proposed rule, EPA highlighted two time periods, 30 years and 100 years, for consideration in our lifecycle analysis. The Agency discussed the relative advantages of these, and other, time periods. In addition, the Agency sought comment on whether it is appropriate to split the time period for GHG emissions assessment based upon how long the biofuel would be produced (i.e., the “project” period) and the time period for which there would likely be GHG emissions changes (i.e., the “impact” period). To encourage expert and public comments on these issues, EPA held public hearings and workshops and sponsored an expert peer review specifically focused on this topic. The expert input and comments that we received included many valuable points which guided our decisions about which time frame should be the focus of our analysis. Below we summarize some of the key arguments made by the peer reviewers and commenters, and how these arguments factored into our choice of analytical approach.

The expert peer reviewers discussed a number of justifiable time periods ranging from 13 to 100 years for assessing lifecycle GHG emissions. A subset of the reviewers said that EPA's analysis should be restricted to 2010-2022 based on the years specified in EISA, because these reviewers argued that EPA should not assume that biofuel production will continue beyond 2022 at the RFS2 levels. The reviewers said that longer time frames, such as 100 years, were only appropriate if the Agency used positive discount rates to value future emissions. Almost all of the peer reviewers said that a time frame of 20 to 30 years would be a reasonable timeframe for assessing lifecycle GHG emissions. They gave several reasons for why a short time period is appropriate: This time frame is the average life of a typical biofuel production facility; future emissions are less certain and more difficult to value, so the analysis should be confined insofar as possible to the foreseeable future; and a near-term time horizon is consistent with the latest climate science that indicates that relatively deep reductions of heat-trapping gasses are needed to avoid catastrophic changes due to a warming climate. The peer reviewers suggested that while there is no unassailable basis for choosing a precise timeframe the expected average lifetime of a biofuel production facility is the “most sensible anchor” for the choice of a timeframe.

There was support in the public comments for both the 30 year and 100 year time frames. A number of public commenters supported the use of a 30 year time period, or less, and made arguments similar to those of the expert peer reviewers. They argued that shorter time periods give more weight to the known, more immediate, effects of biofuel production and that use of longer time periods gives more weight to activities that are much more uncertain, and that the 100 year timeframe is inappropriate because it is much longer than the life of individual biofuel plants.

On the issue of whether to split the time period for GHG emissions analysis into the “project and “impact” periods, there was little support for the use of a split time frame for evaluating lifecycle GHG emissions by the peer reviewers or in the public comments. The peer reviewers thought that it would be difficult to find a scientific basis for determining the length of the two different time horizons. Also, splitting the time horizon would necessitate consideration of the land use changes following the end of the project time horizon such as land reversion. However, the majority of expert peer reviewers did not think it was appropriate to attribute potential land reversions, following the project time frame, to a biofuel's lifecycle.

Based upon the comments discussed above, EPA has decided to use a 30 year frame for assessing the lifecycle GHG emissions. There are several reasons why the 30 year time frame was chosen. The full life of a typical biofuel plant seems reasonable as a basis for the timeframe for assessing the GHG emissions impacts of a biofuel, because it provides a guideline for how long we can expect biofuels to be produced from a particular entity using a specific processing technology. Also, the 30 year time frame focuses on GHG emissions impacts that are more near term and, hence, more certain. We also determined that longer time periods were less appropriate because the peer reviewers recommended that they should only be used in conjunction with positive discount rates; but, for the reasons discussed below, we are using a zero discount rate in our analysis. In addition, the 30 year time frame is consistent with responses of the peer reviewers that EPA should not split the time periods for analysis, or include potential land reversions following the project time period in the biofuel lifecycle.

Discounting: In the RFS2 Proposal, EPA highlighted two principal options for discounting the lifecycle GHG emission streams from biofuels over time. The first involved the use of a 2% discount rate using the 100 year time horizon for assessing lifecycle GHG emissions streams. The second option involved using a 30 year time horizon for examining lifecycle GHG emissions impacts. In the 30 year case, each GHG emission is treated equally through time, which implicitly assumes a zero discount rate to GHG lifecycle emissions streams. The issue of whether to discount lifecycle GHG emissions was raised as a topic that EPA sought comment on in both the peer review process and in public comments.

EPA received numerous comments on the issue of whether the Agency should be discounting lifecycle GHG emissions through time. While many of peer reviewers thought that current GHG emissions reductions should be more strongly weighted than future reductions, the peer reviewers were in general agreement that a discount rate should only be applied to a monetary unit, rather than a physical unit, such as GHG emissions. Public commenters suggested that discounting is an essential part of long term cost benefit analysis but it is not necessary in the context of the physical aggregation of lifecycle GHG emissions called for in the EISA. Further, public commenters expressed concerns that any discount rate chosen by the Agency would be based upon relatively arbitrary criteria.

After considering the comments on discounting from the peer review and the public, EPA has decided not to discount (i.e., use a 0% discount rate) GHG emissions due to the many issues associated with applying an economic concept to a physical parameter. First, it is unclear whether EISA intended lifecycle GHG emissions to be converted into a metric whose underpinnings rest on principals of economic valuation. A more literal interpretation of EISA is that EPA should consider only physical GHG emissions. Second, even if the principle of tying GHG emissions to economic valuation approaches were to be accepted, there would still be the problem that there is a lack of consensus in the scientific community about the best way to translate GHG emissions into a proxy for economic damages. Also, there is a lack of consensus as to the appropriate discount rate to apply to GHG lifecycle emissions streams through time. Finally, since EPA has decided to base threshold assessments of lifecycle GHG emissions on a 30 year time frame, the issue of whether to discount GHG emissions is not as significant as if the EPA had chosen the 100 year time frame to assess GHG emissions impacts. More discussion of discount rates and their impact on the lifecycle results can be found in Chapter 2 of the RIA.

v. GTAP and Other Models

Although we have used the partial equilibrium (PE) models FASOM and FAPRI-CARD as the primary tools for evaluating whether individual biofuels meet the GHG thresholds, as part of the peer review process, we explicitly requested input on whether general equilibrium (GE) models should be used. None of the comments recommended using a GE model as the sole tool for estimating GHG emissions, given the limited details on the agricultural sector contained in most GE models. The peer reviewers generally supported the use of the FASOM and FAPRI-CARD models for our GHG analysis given the need for additional detail offered in the PE models, however several comments suggested incorporating GE models into the analysis.

Given these recommendations, we opted to use the GTAP model to inform the range of potential GHG emissions associated with land use change resulting from an increase in renewable fuels. As discussed in the NPRM, there are several advantages to using GTAP. As a general equilibrium model, GTAP captures the interaction between different markets (e.g., agriculture and energy) in different regions. It is distinctive in estimating the complex international land use change through trade linkages. In addition, GTAP explicitly models land-use conversion decisions, as well as land management intensification. Most importantly, in contrast to other models, GTAP is designed with the framework of predicting the amount and types of land needed in a region to meet demands for both food and fuel production. The GTAP framework also allows predictions to be made about the types of land available in the region to meet the needed demands, since it explicitly represents different types of land cover within each Agro-Ecological Zone.

Like the peer reviewers, we felt that some of the drawbacks of the GTAP model prevent us from using GTAP as the sole model for estimating GHG emissions from biofuels. As discussed in the NPRM, GTAP does not utilize unmanaged cropland, nor is it able to capture the long-run baseline issues (e.g., the state of the economy in 2022). For our analysis, the GTAP model was most valuable for providing another estimate of the quantity and type of land conversion resulting from an increase in corn ethanol and biodiesel given the competition for land and other inputs from other sectors of the economy. These results were therefore considered as part of the weight of evidence when determining whether corn ethanol or biodiesel met the GHG thresholds.

The quantity of total acres converted to crop land projected by FAPRI-CARD were within the range of values projected by GTAP when normalized on a per BTU basis, although there were differences in the regional distribution of these changes. The land use changes projected by GTAP were smaller than land use changes predicted by FAPRI-CARD, which is primarily due to several important differences in the modeling frameworks. First, the GTAP model incorporates a more optimistic view of intensification options by which higher prices induced by renewable fuels results in higher yields, not just for corn, but also for other displaced crops. Second, the demands for other uses of land are explicitly captured in GTAP. Therefore, when land is withdrawn from these uses, the prices of these products rise and provide a certain amount of “push-back” on the conversion of land to crops from pasture or forest. Third, none of the peer-reviewed versions of GTAP currently contain unmanaged cropland, thereby omitting additional sources of land. Finally, the GTAP model also predicted larger increases in forest conversion than the FAPRI-CARD/Winrock analysis, in part because the GTAP model includes only three types of land (i.e., crops, pasture, forest). As discussed in the FAPRI-CARD/Winrock section, there are many other categories of land which may be converted to pasture and crop land.

As with all economic models, GTAP results are sensitive to certain key parameter values. One advantage of this framework is that it offers a readily usable approach to Systematic Sensitivity Analysis (SSA) using efficient sampling techniques. We have exploited this tool in order to develop a set of 95% confidence intervals around the projected land use changes. Several key parameters were identified that have a significant impact on the land use change projections, including the yield elasticity (i.e., the change in yield that results from a change in that commodity's price), the elasticity of transformation of land supply (i.e., the measure of how easily land can be converted between forest, pasture, and crop land), and the elasticity of transformation of crop land (i.e., the measure of how easily land can be converted between crops). Although the confidence intervals are relatively large, in most cases the ranges do not bracket zero. Therefore, we conclude that the impacts of the corn ethanol and soybean biodiesel mandates on land use change are statistically significant. These confidence intervals also bracket the FAPRI-CARD results. Additional information on the GTAP results is discussed in RIA Chapter 2.

c. Feedstock Transport

To estimate the GHG impacts of transporting corn from the field to an ethanol production facility and transporting the co-product DDGS from the ethanol facility to the point of use, we used the method described in the proposed rule. We also did not change our estimates for the transport of cellulosic biofuel feedstock and biomass-based diesel feedstock.

For sugarcane transport, we received the comment that the GREET defaults used to estimate the energy consumption and associated GHG emissions do not all reflect current industry practices. To address this concern, we reviewed the current literature on sugarcane transport and updated our assumptions on the distance sugarcane travels by truck from the field to ethanol production facilities as well as the payload and fuel economy of those trucks. We incorporated these revised inputs into an updated version of the GREET model (Version 1.8c) in order to estimate the GHG impacts of sugarcane transport. More details on these updates can be found in Chapter 2 of the RIA.

In the proposal, we discussed updating our analysis to incorporate the results of a recent study detailing biofuel production locations and modes of transport. This study, conducted by Oak Ridge National Laboratory, modeled the transportation of ethanol from production or import facilities to petroleum blending terminals. Since the study did not explicitly address the transport of biofuel feedstocks, we did not implement the results for this part of the analysis. However, we did incorporate the results into our assessment of the GHG impacts of fuel transportation. We will continue to examine whether our feedstock transport estimates could be significantly improved by implementing more detailed information on the location of biofuel production facilities.

We also discussed updating the transportation modes and distances assumed for corn and DDGS to account for the secondary or indirect transportation impacts. For example, decreases in exports will reduce overall domestic agricultural commodity transport and emissions but will increase transportation of commodities internationally. We did not implement these secondary transportation impacts in this final rule. While we do not anticipate that such impacts would significantly change the lifecycle analysis, we plan to continue to look at this issue and consider incorporating them in the future.

d. Biofuel Processing

For the proposal the GHG emissions from renewable fuel production were calculated by multiplying the Btus of the different types of energy inputs at biofuel process plants by emissions factors for combustion of those fuel sources. The Btu of energy input was determined based on analysis of the industry and specific work done as part of the NPRM. The emission factors for the different fuel types are from GREET and were based on assumed carbon contents of the different process fuels. The emissions from producing electricity in the U.S. were also taken from GREET and represent average U.S. grid electricity production emissions.

We received comments on our approach and updated the analysis of GHG emissions from biofuel process for the final rule specifically regarding process energy use and the treatment of co-products.

Process Energy Use: For the final rule we updated each of our biofuel pathways to include the latest data available on process energy use. For the proposal, one of the key sources of information on energy use for corn ethanol production was a study from the University of Illinois at Chicago Energy Resource Center. Between proposal and final rule, the study was updated, therefore, we incorporated the results of the updated study in our corn ethanol pathways process energy use for the final rule. We also updated corn ethanol production energy use for different technologies in the final rule based on feedback from industry technology providers as part of the public comment period. The main difference between proposal and final corn ethanol energy use values was a slight increase in energy use for the corn ethanol fractionation process, based on feedback from industry technology providers.

For the proposal we based biodiesel processing energy on a process model developed by USDA-ARS to simulate biodiesel production from the Fatty Acid Methyl Ester (FAME) transesterification process. We received a number of comments from stakeholders that the energy balance for biodiesel production was overestimating energy use and should be updated. During the comment period USDA updated their energy balance for biodiesel production to incorporate a different biodiesel dehydration process based on a system which has resulted in a decrease in energy requirements. This change was reflected in the energy use values for biodiesel assumed in our final rule analysis which resulted in reduced GHG impacts from the biodiesel production process.

In addition, for the final rule we have included an analysis of algae oil production for biodiesel based on ASPEN process modeling from NREL. [174] The analysis is for two major cultivation pathways (open pond and photobioreactors) for a facility that can be feasibly commercialized in the future, represented by a “2022” target production. We coupled the algae oil production process (which includes cultivation, harvesting, and extraction) with the biodiesel production energy use from virgin oils energy use model under the assumption that algae oil is similar enough to that of virgin oil.

For the cellulosic biofuel pathways, we updated our final rule energy consumption assumptions on process modeling also completed by NREL. For the NPRM, NREL estimated energy use for the biochemical enzymatic process to ethanol route in the near future (2010) and future (2015 and 2022). 175 176 177 As there are multiple processing pathways for cellulosic biofuel, we have expanded the analysis for the FRM to also include thermochemical processes (Mixed-Alcohols route and Fischer-Tropsch to diesel route) for plants which assume woody biomass as its feedstock.

Under the imported sugarcane ethanol cases we updated process energy use assumptions to reflect anticipated increases in electricity production for 2022 based on recent literature and comments to the proposal. One major change was assuming the potential use of trash (tops and leaves of sugarcane) collection in future facilities to generate additional electricity. The NPRM had only assumed the use of bagasse for electricity generation. Based on comments received, we are also assuming marginal electricity production (i.e., natural gas) instead of average electricity mix in Brazil which is mainly hydroelectricity. This approach assumes surplus electricity will likely displace electricity which is normally dispatched last, in this case typically natural gas based electricity. The result of this change is a greater credit for displacing marginal grid electricity and thus a lower GHG emissions profile for imported sugarcane ethanol than that assumed in the NPRM. We also received public comment that there are differences in the types of process fuel e.g. used in the dehydration process for ethanol. While using heavier fuels such as diesel or bunker fuel tends to increase the imported sugarcane ethanol emissions profile, the overall impact was small enough that lifecycle results did not change dramatically.

Co-Products: In response to comments received, we included corn oil fractionation and extraction as a potential source of renewable fuels for this final rulemaking. Based on research of various corn ethanol plant technologies, corn oil as a co-product from dry mill corn ethanol plants can be used as an additional biodiesel feedstock source (see Section VII.A.2 for additional information). Dry mill corn ethanol plants have two different technological methods to withdraw corn oil during the ethanol production process. The fractionation process withdraws corn oil before the production of the DGS co-product. The resulting product is food-grade corn oil. The extraction process withdraws corn oil after the production of the DGS co-product, resulting in corn oil that is only suitable for use as a biodiesel feedstock.

Based on cost projections outlined in Section VII.A, it is estimated that by 2022, 70% of dry mill ethanol plants will conduct extraction, 20% will conduct fractionation, and that 10% will choose to do neither. These parameters have been incorporated into the FASOM and FAPRI-CARD models for the final rulemaking analysis, allowing for corn oil from extraction as a major biodiesel feedstock.

Glycerin is a co-product of biodiesel production. Our proposal analysis did not assume any credit for this glycerin product. The assumption for the proposal was that by 2022 the market for glycerin would be saturated due to the large increase in biodiesel production in both the US and abroad and the glycerin would therefore be a waste product. We received a number of comments that we should be factoring in a co-product credit for glycerin as there would be some valuable use for this product in the market. Based on these comments we have included for the final rule analysis that glycerin would displace residual oil as a fuel source on an energy equivalent basis. This is based on the assumption that the glycerin market would still be saturated in 2022 and that glycerin produced from biodiesel would not displace any additional petroleum glycerin production. However, the biodiesel glycerin would not be a waste and a low value use would be to use the glycerin as a fuel source. The fuel source assumed to be replaced by the glycerin is residual oil. This inclusion of a co-product credit for glycerin reduces the overall GHG impact of biodiesel compared to the proposal analysis.

e. Fuel Transportation

For the proposed rule, we estimated the GHG impacts associated with the transportation and distribution of domestic and imported ethanol and biomass-based diesel using GREET defaults. We have upgraded to the most recent version of GREET (Version 1.8c) for our transportation analysis in the final rule. [178] We made several other updates to the method we utilized in the proposed rule. These updates are described here and in more detail in Chapter 2 of the RIA.

In the proposal, we noted our intention to incorporate the results of a recent study by Oak Ridge National Laboratory (ORNL) into our transportation analysis for the final rule. The ORNL study models the transportation of ethanol from refineries or import facilities to the petroleum blending terminals by domestic truck, marine, and rail distribution systems. We used ORNL's transportation projections for 2022 under the EISA policy scenario to update our estimates of the GHG impacts associated with the transportation of corn, cellulosic, and sugarcane ethanol. Since the study did not address the distribution of ethanol from petroleum blending terminals to refueling stations, we continu