Mandatory Reporting of Greenhouse Gases: Additional Sources of Fluorinated GHGs

Federal RegisterApr 12, 2010

Ask Donna

What actually matters in this document.

Text

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Part 98

[EPA-HQ-OAR-2009-0927; FRL-9130-7]

RIN 2060-AQ00

Mandatory Reporting of Greenhouse Gases: Additional Sources of Fluorinated GHGs

AGENCY:

Environmental Protection Agency (EPA).

ACTION:

Proposed rule.

SUMMARY:

EPA is revising and supplementing its initial proposed actions to require reporting of fluorinated greenhouse gas (fluorinated GHG) emissions from certain source categories. Specifically, EPA is revising and supplementing its initial proposal to require reporting of fluorinated GHG emissions from electronics manufacturing, production of fluorinated gases, and use of electrical transmission and distribution equipment. EPA is also proposing to require such reporting from manufacture or refurbishment of electrical equipment and import and export of pre-charged equipment and closed cell foams. This proposed rule would not require control of greenhouse gases; rather it would require only that sources above certain threshold levels monitor and report emissions.

DATES:

Comments must be received on or before June 11, 2010. There will be a public hearing from 9 a.m. to 12 noon on April 20, 2010 at 1310 L St., NW., Room 152, Washington, DC 20005.

ADDRESSES:

Submit your comments, identified by docket ID EPA-HQ-OAR-2009-0927 by one of the following methods:

•

Federal eRulemaking Portal: http://www.regulations.gov.

Follow the online instructions for submitting comments.

•

E-mail: GHGReportingFGHG@epa.gov.

•

Fax:

(202) 566-1741.

•

Mail:

EPA Docket Center, Attention Docket OAR-2009-0927, Mail code 2822T, 1200 Pennsylvania Avenue, NW., Washington, DC 20460.

•

Hand/Courier Delivery:

EPA Docket Center, Public Reading Room, Room 3334, EPA West Building, Attention Docket OAR-2009-0927, 1301 Constitution Avenue, NW., Washington, DC 20004. Such deliveries are only accepted during the Docket's normal hours of operation, and special arrangements should be made for deliveries of boxed information.

Instructions:

Direct your comments to Docket ID No. EPA-HQ-OAR-2009-0927. EPA's policy is that all comments received will be included in the public docket without change and may be made available online at

http://www.regulations.gov,

including any personal information provided, unless the comment includes information claimed to be CBI or other information whose disclosure is restricted by statute. Do not submit information that you consider to be CBI or otherwise protected through

http://www.regulations.gov

or e-mail. The

http://www.regulations.gov

Web site is an “anonymous access” system, which means EPA will not know your identity or contact information unless you provide it in the body of your comment. If you send an e-mail comment directly to EPA without going through

http://www.regulations.gov

your e-mail address will be automatically captured and included as part of the comment that is placed in the public docket and made available on the Internet. If you submit an electronic comment, EPA recommends that you include your name and other contact information in the body of your comment and with any disk or CD-ROM you submit. If EPA cannot read your comment due to technical difficulties and cannot contact you for clarification, EPA may not be able to consider your comment. Electronic files should avoid the use of special characters, any form of encryption, and be free of any defects or viruses.

Docket:

All documents in the docket are listed in the

http://www.regulations.gov

index. Although listed in the index, some information is not publicly available,

e.g.,

CBI or other information whose disclosure is restricted by statute. Certain other material, such as copyrighted material, will be publicly available only in hard copy. Publicly available docket materials are available either electronically in

http://www.regulations.gov

or in hard copy at EPA's Docket Center, Public Reading Room, EPA West Building, Room 3334, 1301 Constitution Ave., NW., Washington, DC 20004. This Docket Facility 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:

Carole Cook, Climate Change Division, Office of Atmospheric Programs (MC-6207J), Environmental Protection Agency, 1200 Pennsylvania Ave., NW., Washington, DC 20460; telephone number: (202) 343-9263; fax number: (202) 343-2342; e-mail address:

GHGReportingRule@epa.gov.

For technical information contact the Greenhouse Gas Reporting Rule e-mail:

ghgmrr@epa.gov.

To obtain information about the public hearings or to register to speak at the hearings, please go to

http://www.epa.gov/climatechange/emissions/ghgrulemaking.html.

SUPPLEMENTARY INFORMATION:

Additional Information on Submitting Comments: To expedite review of your comments by Agency staff, you are encouraged to send a separate copy of your comments, in addition to the copy you submit to the official docket, to Carole Cook, U.S. EPA, Office of Atmospheric Programs, Climate Change Division, Mail Code 6207-J, Washington, DC 20460, telephone (202) 343-9263, e-mail

GHGReportingRule@epa.gov.

As indicated above, although EPA previously proposed a version of some parts of this rule, that proposal has not become final. This proposal partly supplements and partly replaces that initial proposal. Comments on the initial proposal will be considered only to the extent they remain relevant. To ensure that their comments on newly proposed or re-proposed provisions are considered, parties should submit or re-submit them at this time.

Regulated Entities.

The Administrator determined that this action is subject to the provisions of Clean Air Act (CAA) section 307(d).

See

CAA section 307(d)(1)(V) (the provisions of section 307(d) apply to “such other actions as the Administrator may determine.”). This is a proposed regulation. If finalized, these regulations would affect owners or operators of electronics manufacturing facilities, fluorinated gas production facilities, electric power systems, and electrical equipment manufacturing facilities, as well as importers and exporters of pre-charged equipment and closed-cell foams. Regulated categories and entities would include those listed in Table 1 of this preamble:

Table 1—Examples of Affected Entities by Category

Category

NAICS

Examples of affected facilities

Electronics Manufacturing

334111

Microcomputers manufacturing facilities.

334413

Semiconductor, photovoltaic (solid-state) device manufacturing facilities.

334419

LCD unit screens manufacturing facilities.

334419

MEMS manufacturing facilities.

Fluorinated GHG Production

325120

Industrial gases manufacturing facilities.

Electrical Equipment Use

221121

Electric bulk power transmission and control facilities.

Electrical Equipment Manufacture or Refurbishment

33531

Power transmission and distribution switchgear and specialty transformers manufacturing facilities.

Importers and Exporters of Pre-charged Equipment and Closed-Cell Foams

423730

Air-conditioning equipment (except room units) merchant wholesalers.

333415

Air-conditioning equipment (except motor vehicle) manufacturing.

423620

Air-conditioners, room, merchant wholesalers.

443111

Household Appliance Stores.

326150

Polyurethane foam products manufacturing.

335313

Circuit breakers, power, manufacturing.

423610

Circuit breakers merchant wholesalers.

Table 1 of this preamble is not intended to be exhaustive, but rather provides a guide for readers regarding facilities likely to be affected by this action. Table 1 lists the types of facilities that EPA is now aware could be potentially affected by the reporting requirements. Other types of facilities and companies not listed in the table could also be subject to reporting requirements. To determine whether you are affected by this action, you should carefully examine the applicability criteria found in 40 CFR part 98, subpart A and the relevant criteria in the proposed subparts related to electronics manufacturing facilities, fluorinated gas production facilities, electrical equipment use, electrical equipment manufacturing or refurbishment facilities, and importers and exporters of pre-charged equipment and closed-cell foams. If you have questions regarding the applicability of this action to a particular facility, consult the person listed in the preceding

FOR FURTHER INFORMATION CONTACT

section.

Many facilities that would be affected by the proposed rule have GHG emissions from multiple source categories listed in 40 CFR part 98 or in this proposed rule. Table 2 of this preamble has been developed as a guide to help potential reporters in the source categories subject to the proposed rule identify the source categories (by subpart) that they may need to (1) consider in their facility applicability determination, and/or (2) include in their reporting. The table should only be seen as a guide. Additional subparts in 40 CFR part 98 may be relevant for a given reporter. Similarly, not all listed subparts are relevant for all reporters.

Table 2—Source Categories and Relevant Subparts

Source category (and main applicable subpart)

Subparts recommended for review to determine

applicability

Electricity Generation

Electrical Equipment Use.

Electronics Manufacturing

General Stationary Fuel Combustion.

Fluorinated GHG Production

General Stationary Fuel Combustion. Suppliers of Industrial Greenhouse Gases.

Electrical Equipment Use

General Stationary Fuel Combustion.

Imports and Exports of Fluorinated GHGs Inside Pre-charged Equipment and Closed-Cell Foams

Suppliers of Industrial Greenhouse Gases.

Sulfur Hexafluoride and PFCs from Electrical Equipment Manufacture and Refurbishment.

Electrical Equipment Manufacture or Refurbishment

General Stationary Fuel Combustion

Imports and Exports of Fluorinated GHGs Inside Pre-charged Equipment and Closed-Cell Foams.

Acronyms and Abbreviations.

The following acronyms and abbreviations are used in this document.

ASTM American Society for Testing and Materials

BAMM Best Available Monitoring Methods

CAA Clean Air Act

CARB California Air Resources Board

CBI confidential business information

CFC chlorofluorocarbon

CFR Code of Federal Regulations

carbon dioxide

e CO

-equivalent

EIA Economic Impact Analysis

EO Executive Order

EPA U.S. Environmental Protection Agency

FERC Federal Energy Regulatory Commission

F-GHG fluorinated greenhouse gas

FTIR fourier transform infrared (spectroscopy)

FID flame ionization detector

GC gas chromatography

GHG greenhouse gas

GWP global warming potential

HCFC hydrochlorofluorocarbon

HFC hydrofluorocarbon

HFE hydrofluoroether

HTF heat transfer fluid

ICR information collection request

IPCC Intergovernmental Panel on Climate Change

kg kilograms

LCD liquid crystal displays

MEMS microelectromechanical devices

MMTCO

e million metric tons carbon dioxide equivalent

MRR mandatory greenhouse gas reporting rule

MS mass spectrometry

O nitrous oxide

NACAA National Association of Clean Air Agencies

NAICS North American Industry Classification System

NERC North American Energy Reliability Corporation

NESHAP National Emissions Standard for Hazardous Air Pollutants

nitrogen trifluoride

NMR nuclear magnetic resonance

NSPS New Source Performance Standards

OMB Office of Management and Budget

PFC perfluorocarbon

PSD Prevention of Significant Deterioration

PV photovoltaic cells

QA quality assurance

QA/QC quality assurance/quality control

R&D research and development

RFA Regulatory Flexibility Act

RGGI Regional Greenhouse Gas Initiative

RIA Regulatory Impact Analysis

SSM startup, shutdown, and malfunction

sulfur hexafluoride

TCR The Climate Registry

TSD technical support document

U.S. United States

UMRA Unfunded Mandates Reform Act of 1995

VOC volatile organic compound(s)

WCI Western Climate Initiative

Table of Contents

I. Background

A. Organization of This Preamble

B. Background on the Proposed Rule

C. Legal Authority

D. Relationship to other Federal, State and Regional Programs

II. Summary of and Rationale for the Reporting, Recordkeeping and Verification Requirements for Specific Source Categories

A. Electronics Manufacturing

B. Fluorinated Gas Production

C. Electric Transmission and Distribution Equipment Use

D. Imports and Exports of Fluorinated GHGs inside pre-charged equipment and closed-cell foams

E. Electrical Equipment Manufacture or Refurbishment

F. Subpart A Revisions

III. Economic Impacts on the Rule

A. How were compliance costs estimated?

B. What are the costs of the rule?

C. What are the economic impacts of the rule?

D. What are the impacts of the rule on small businesses?

E. What are the benefits of the rule for society?

IV. Statutory and Executive Order Reviews

A. Executive Order 12866: Regulatory Planning and Review

B. Paperwork Reduction Act

C. Regulatory Flexibility Act (RFA)

D. Unfunded Mandates Reform Act (UMRA)

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 that Significantly Affect Energy Supply, Distribution, or Use

I. National Technology Transfer and Advancement Act

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

I. Background

A. Organization of This Preamble

This preamble is broken into several large sections, as detailed above in the Table of Contents. The paragraphs below describe the layout of the preamble and provide a brief summary of each section.

The first section of this preamble contains the basic background information about the origin of this proposed rule, including a brief discussion of the initial proposed requirements for electronics, fluorinated gas production, and use of electrical transmission and distribution equipment. This section also discusses EPA's use of our legal authority under the CAA to collect the proposed data, and the benefits of collecting the data.

The second section of this preamble provides a brief summary of, and rationale for, the key design elements on which EPA is seeking comment today for each subpart. Depending on the subpart, this section may include EPA's rationale for (i) the definition of the source category, (ii) selection of reporting threshold, (iii) selection of proposed reporting and monitoring methods, (iv) selection of procedures for estimating missing data, (v) selection of data reporting requirements, and (vi) selection of records that must be retained. EPA describes the proposed options for each design element, as well as the other options considered. Throughout this discussion, EPA highlights specific issues on which we solicit comment. Please refer to the specific source category of interest for more details.

The third section provides the summary of the cost impacts, economic impacts, and benefits of this proposed rule from the Economic Analysis. Finally, the last section discusses the various statutory and executive order requirements applicable to this proposed rulemaking.

B. Background on the Proposed Rule

The Final Mandatory GHG Reporting Rule (Final MRR), (40 CFR part 98) was signed by EPA Administrator Lisa Jackson on September 22, 2009 and published in the

Federal Register

on October 30, 2009 (74 FR 56260). The Final MRR, which became effective on December 29, 2009, included reporting of GHGs from the facilities and suppliers that EPA determined should be included to appropriately respond to the direction in the 2008 Consolidated Appropriations Act.

These source categories capture approximately 85 percent of U.S. GHG emissions through reporting by direct emitters as well as suppliers of fossil fuels and industrial gases.

Consolidated Appropriations Act, 2008, Public Law 110-161, 121 Stat. 1844, 2128.

In the April 2009 proposed mandatory GHG reporting rule, the electronics, fluorinated GHG production, and electrical equipment use source categories were included as subparts I, L, and DD. In addition, EPA requested comment on requiring reporting under subpart OO of the quantities of fluorinated GHGs imported and exported inside pre-charged equipment and foams. EPA received a number of lengthy, detailed comments regarding proposed subparts I and L, several comments regarding the definition of “facility” under subpart DD, and several comments regarding a reporting requirement for imports and exports of fluorinated GHGs contained inside pre-charged equipment and foams. These comments, which are described in more detail in the discussions of the individual source categories below, raised concerns about the costs and technical feasibility of implementing subparts I and L as initially proposed, requested clarification of how “facility” should be interpreted under subpart DD, and both favored and opposed a requirement to report imports of fluorinated GHGs contained in imported and exported pre-charged equipment and closed-cell foams.

EPA recognized the concerns raised by stakeholders, and decided not to finalize subparts I, L, and DD with the Final MRR, but instead to re-propose significant pieces of these subparts. For subparts I and L this proposal incorporates a number of changes including, but not limited to, the addition of different methodologies that provide improved emissions coverage at a lower cost burden to facilities as compared to the initial proposal. Where aspects of the initial proposals for subparts I and L are retained in this proposal, such as in the basic mass-balance methodology for subpart L (as an option for some facilities) and in many of the equations for subpart I, today's proposal adds more flexibility in how and how frequently the underlying data are gathered. In addition, EPA is proposing requirements to report emissions from manufacture or refurbishment of electrical equipment and to report the quantities of fluorinated GHGs imported and exported inside pre-charged equipment and foams.

We believe the monitoring approaches proposed in this action, which combine direct measurement and facility-specific calculations, effectively balance

accuracy and costs, and that they are warranted even though the rule does not contain any emissions reduction requirements. As we stated in the Final MRR, the data collected by the rule are expected to be used in analyzing and developing a range of potential CAA GHG policies and programs. A consistent and accurate data set is crucial to serve this intended purpose.

Under this proposed rule, facilities not already reporting but required to report under this rule would begin data collection in 2011 following the methods outlined in the proposed rule and would submit data to EPA by March 31, 2012. As is the case under the Final MRR, facilities would have the option to use Best Available Monitoring Methods (BAMM) for the first quarter of the first reporting year for the source categories included in this proposed rule. Thus, for these source categories, facilities could use BAMM through March 31, 2011.

C. Legal Authority

EPA is proposing this rule under its existing CAA authority, specifically authorities provided in CAA section 114. As discussed further below and in “Mandatory Greenhouse Gas Reporting Rule: EPA's Response to Public Comments, Legal Issues” (available in EPA-HQ-OAR-2008-0508), EPA is not citing the FY 2008 Consolidated Appropriations Act as the statutory basis for this action. While that law required that EPA spend no less than $3.5 million on a rule requiring the mandatory reporting of GHG emissions, it is the CAA, not the Appropriations Act, that EPA is citing as the authority to gather the information proposed by this rule.

As stated in the Final MRR, CAA section 114 provides EPA broad authority to require the information proposed by this rule because such data would inform and are relevant to EPA's carrying out a wide variety of CAA provisions. As discussed in the initial proposed rule (74 FR 16448, April 10, 2009), CAA section 114(a)(1) authorizes the Administrator to require emissions sources, persons subject to the CAA, or persons whom the Administrator believes may have necessary information to monitor and report emissions and provide such other information the Administrator requests for the purposes of carrying out any provision of the CAA. EPA notes that while climate change legislation approved by the U.S. House of Representatives, and pending in the U.S. Senate, would provide EPA additional authority for a GHG registry similar to this proposed rule, and would do so for purposes of that pending legislation, this proposed rule is authorized by, and the information being gathered by this proposed rule is relevant to implementing, the existing CAA. EPA expects, however, that the information collected by this proposed rule would also prove useful to legislative efforts to address GHG emissions.

For further information about EPA's legal authority,

see

the proposed and Final MRR.

D. Relationship to Other Federal, State and Regional Programs

In developing this proposed rule, EPA reviewed monitoring methods included in international guidance (

e.g.,

Intergovernmental Panel on Climate Change), as well as Federal voluntary programs (

e.g.,

EPA PFC Reduction/Climate Partnership for the Semiconductor Industry and the U.S. Department of Energy Voluntary Reporting of Greenhouse Gases Program (1605(b) of the Energy Policy Act), corporate protocols (

e.g.,

World Resources Institute and World Business Council for Sustainable Development GHG Protocol) and industry guidance (

e.g.,

2006 ISMI Guideline for Environmental Characterization of Semiconductor Process Equipment).

EPA also reviewed State reporting programs (

e.g.,

California and New Mexico) and Regional partnerships (

e.g.,

Regional Greenhouse Gas Initiative, Western Climate Initiative, The Climate Registry). These are important programs that not only led the way in reporting of GHG emissions before the Federal government acted but also assist in quantifying the GHG reductions achieved by various policies. Many of these programs collect different or additional data as compared to this proposed rule. For example, State programs may establish lower thresholds for reporting, request information on areas not addressed in EPA's reporting rule, or include different data elements to support other programs (

e.g.,

offsets). For further discussion on the relationship of this proposed rule to other programs, please refer to the preamble to the Final MRR.

II. Summary of and Rationale for the Reporting, Recordkeeping and Verification Requirements for Specific Source Categories

A. Electronics Manufacturing

1. Overview of Reporting Requirements

Electronics manufacturing includes, but is not limited to, the manufacture of semiconductors, liquid crystal displays (LCDs), micro-electro-mechanical systems (MEMS), and photovoltaic cells (PV). The electronics industry uses multiple long-lived fluorinated greenhouse gases (fluorinated GHGs) such as perfluorocarbons (PFCs), hydrofluorocarbons (HFCs), sulfur hexafluoride (SF

), and nitrogen trifluoride (NF

), as well as nitrous oxide (N

O). This proposed rule would apply to electronics manufacturing facilities where emissions from electronics manufacturing processes such as plasma etching, chemical vapor deposition, chamber cleaning, and heat transfer fluid use as well as stationary fuel combustion units equal or exceed 25,000 metric tons of CO

e per year.

In this action, we are proposing methods to estimate emissions from cleaning and etching for semiconductor, LCD, MEMS, and PV manufacture and also methods for estimating N

O emissions from chemical vapor deposition and other manufacturing processes such as chamber cleaning. We are also clarifying methods for estimating emissions from heat transfer fluids. And lastly, we are proposing methods for reporting controlled emissions from abatement systems.

As discussed further below, EPA is proposing that uncontrolled emissions be used for purposes of determining whether a facility's emissions are equal to or greater than 25,000 mtCO

2. Major Changes Since Initial Rule Proposed

In the initial proposal for electronics manufacturing, we included the following provisions for reporting emissions from electronics manufacture: (1) A capacity-based threshold for semiconductors, LCDs, and MEMS facilities and an emissions-based threshold for PV facilities; (2) methods for estimating fluorinated GHG emissions from etching and cleaning; (3) methods for estimating N

O emissions during etching and cleaning; (4) methods for verifying destruction or removal efficiency (DRE) of abatement systems; and (5) methods for estimating emissions from heat transfer fluids.

As noted in the preamble to the Final MRR, we received a number of lengthy, detailed comments regarding the electronics manufacturing subpart. In total, we received comments from approximately 10 entities on the proposed rule regarding electronics manufacture. The commenters generally opposed the proposed reporting requirements for large semiconductor facilities and stated that excessive monitoring and reporting were required. For example, commenters asserted that they do not currently collect the data required to report using an IPCC Tier 3

approach, and that to collect such data would entail significant burden and capital costs. In most cases, commenters provided alternative approaches to each of the reporting requirements.

We have carefully reviewed the comments, issues, and suggestions raised by stakeholders regarding electronics manufacturing. In response, we are revising our initial proposal and are proposing the following reporting provisions for electronics manufacture: (1) A single emissions-based reporting threshold for all semiconductor, LCD, MEMS, and PV facilities; (2) modified methods for estimating emissions from cleaning and etching activities for semiconductor facilities and other electronics facilities including those that manufacture LCDs, MEMS, and PVs; (3) modified methods for estimating facility N

O emissions; (4) clarified methods for estimating emissions from heat transfer fluids; and (5) revised methods for reporting controlled emissions from abatement systems.

In the paragraphs below, we summarize the main provisions included in the initial proposal for reporting emissions from electronics manufacturing and we briefly summarize the major changes that are being proposed today. For more detailed information on the initial proposal,

see

the electronics manufacturing section of EPA's proposed MRR (74 FR 16448, April 10, 2009).

Reporting Threshold.

In the initial proposal, we proposed a capacity-based threshold, requiring those facilities with emissions equal to or greater than the thresholds to report their GHG emissions. We proposed production capacity-based thresholds of 1,080 m

, 1,020 m

, and 236,000 m

of substrate for semiconductor, MEMS, and LCD manufacturing facilities, respectively. The capacity-based threshold proposed were equivalent to 25,000 mtCO

e using the IPCC 2006 Tier 1 default factors and assumed no abatement. Where IPCC 2006 Tier 1 default emission factors were unavailable (

i.e.,

MEMS), the emission factor was estimated based on relevant IPCC Tier 1 emission factors for semiconductor production. Due to a lack of information on use and emissions of fluorinated GHGs for PV manufacture, we proposed an emissions-based threshold of 25,000 mtCO

e for those facilities. We proposed to use a capacity-based threshold based on the published capacities of facilities, as opposed to an emissions-based threshold, where possible, because we believed that it simplified the applicability determination.

Several commenters stated that the proposed capacity-based threshold created ambiguity. For example, one commenter noted that it was unclear how production capacity was defined as actual manufacturing levels could fluctuate year by year. In response to these comments, we are now proposing a single emissions-based threshold equal to or greater than 25,000 metric tons of CO

e per year for electronics manufacturing facilities. We have concluded that a single emissions-based threshold will simplify the applicability determination and that by applying the method for determining whether the threshold is met, a facility will be able to quickly determine whether they must report under this rule.

Estimating Emissions from Cleaning and Etching Processes.

With respect to estimating emissions from chamber cleaning and etching, in our initial proposal, we outlined two different methods; one method for relatively large semiconductor facilities, and another method for all other semiconductor facilities and LCD, MEMS, and PV facilities required to report. We defined large semiconductor facilities as those facilities with annual capacities of greater than 10,500 m

silicon (equivalent to 29 out of 175 total semiconductor manufacturing facilities). For large semiconductor facilities we proposed an approach based on the IPCC Tier 3 method that required the use of company-specific data for (1) gas consumption, (2) gas utilization,

(3) by-product formation

, and (4) DRE for all emissions abatement processes at the facility. As we stated in the initial proposal, we had concluded that large semiconductor facilities were already using Tier 3 methods and/or had the necessary data readily available either in-house or from suppliers to apply the highest Tier method. For smaller semiconductor facilities and LCD, MEMS, and PV facilities, we proposed an approach based on the IPCC Tier 2b method, which required using default emission factors for process utilization, by-product formation, and site-specific DRE measurements.

For purposes of electronics manufacturing, we are using the term “gas utilization” to describe the fraction of input N

O or fluorinated GHG converted to other substances during the etching, deposition, and/or chamber/wafer cleaning processes. Gas utilization is expressed as a rate or factor for specific manufacturing processes. “Utilization” should not be confused with “use;” “use” refers to gas consumption or the quantity of gas fed into process at an electronics manufacturing facility.

For purposes of electronics manufacturing, “by-product formation” is the quantity of fluorinated GHGs created during electronics manufacturing processes. Fluorinated GHG by-products may also be formed by abatement devices.

Comments received in response to our initial proposal stated that the 2006 IPCC Tier 3 method would be overly burdensome for semiconductor manufacturers and that process-specific emission factors do not exist for many tools and processes. The commenters noted that most semiconductor facilities do not track gas consumption by tool or process-type and that currently, only one large semiconductor company uses the Tier 3 method. Generally, commenters requested the use of the 2006 IPCC Tier 2b method.

In response to these comments, we are now proposing the use of a “Refined Method” for estimating these emissions from semiconductor facilities. Our revised methodology includes a simpler approach to estimating emissions from cleaning and etching as compared to the Tier 3 method that was initially proposed for larger semiconductor facilities. To this end, we estimate that our proposed methodology will result in a reduction in burden compared to the Tier 3 method for those facilities previously defined as large semiconductor facilities, and an improvement in accuracy of the emissions estimate as compared to the 2006 IPCC Tier 2b method. Furthermore, since we anticipate that all semiconductor facilities already have, or have ready access to, the information required by this proposed methodology, we are also proposing to require all semiconductor facilities required to report to estimate emissions using the Refined Method. We have concluded the method we are proposing is the most appropriate method taking into account both the cost to the reporter as well as accuracy of emissions achieved.

For LCD, MEMS, and PV facilities, in this action we are proposing to require an approach based on a slightly modified 2006 IPCC Tier 2b method which would include (1) gas-and facility-specific heel factors (consistent with the requirements we are proposing for semiconductor facilities), (2) gas consumption apportioned to 2006 IPCC Tier 2b process categories (

i.e.

clean and etch), (3) default factors consistent with the 2006 IPCC Tier 2b factors, and (4) methods for reporting controlled emissions from abatement systems (as proposed below). The main difference between the method proposed in this revised proposal and in the initial proposal is the addition of a gas-and facility-specific heel factor to determine overall gas consumption. We did not receive any comments on the Tier 2b method that we proposed for LCD, MEMS, and PV facilities in our initial proposal. We are proposing to add the requirement of gas-and-facility specific

heel factors based on comments received from semiconductor facilities in response to the initial proposal. It is our understanding that LCD, MEMS, and PV facilities have the data required to develop a gas-and-facility specific heel factors and that it can be implemented with minimal burden.

Estimating Facility N

Emissions. In our initial proposal, our approach required that facilities estimate annual N

O emissions using a simple mass-balance method. This method assumed that all N

O consumed is emitted (

i.e.,

not converted or destroyed). We also requested comment on utilization factors for N

O as well as on data on N

O by-product formation.

In response to our initial proposal, we received comments that clarified that N

O is used primarily in the chemical vapor deposition process. Commenters opposed our proposed method for estimating N

O emissions, which assumed 100 percent N

O used is emitted, and asserted that semiconductor facilities should be permitted to use measured N

O emission factors where these factors were measured using methods consistent with the December 2006 International SEMATECH Manufacturing Initiative's Guideline for Environmental Characterization of Semiconductor Process Equipment (2006 ISMI Guidelines). Commenters also noted that facilities that have not developed N

O emission factors should be allowed to use a default emission factor of 60 percent, reflecting N

O utilization of 40 percent.

Lastly, commenters asserted that those companies that have a measured DRE for N

O abatement be allowed to apply these DREs in the emission estimates.

The 40% utilization rate (60% emission factor) was identified based on a survey of industry conducted by ISMI and provided in comments in response to the initial proposal.

We are now proposing two methods for estimating N

O emissions from electronics manufacturing: one for estimating N

O emissions from chemical vapor deposition and another for estimating N

O emissions from all other manufacturing processes such as chamber cleaning.

Reporting Controlled Emissions From Abatement Systems.

The emissions estimation method originally proposed accounted for destruction by abatement systems only if facilities verified the performance of their systems using one of two methods. In particular, we proposed to require that the DRE be verified by either (1) measurement by the facility using the methods described in EPA's Protocol for Measuring Destruction or Removal Efficiency of Fluorinated Greenhouse Gas Abatement Equipment in Electronics Manufacturing (EPA's DRE Protocol), or (2) purchase by the facility of abatement systems that were tested by a third party using a standard protocol such as EPA's DRE Protocol.

We also proposed to require that facilities use the systems within the manufacturer's specified system lifetime, operate the system within the manufacturer specific limits for the gas mix and exhaust flow rate intended for the fluorinated GHG destruction, and maintain the equipment according to the manufacturer's guidelines.

In response to the initial proposal, commenters were generally opposed to EPA's initial approach for measuring DRE, noting that according to the

Results of the ISMI ESH Technology Center Greenhouse Gas Facility Survey,

less than one percent of installed abatement systems have been properly tested using the draft EPA Protocol and that generally, facilities use the IPCC default factors or manufacturer-supplied measurements. In addition, commenters were also opposed to EPA's proposed requirement that facilities rely on manufacturer-specified system lifetime as properly maintained and serviced abatement systems can last beyond the manufactures' specified lifetime. For purposes of this reporting rule, we are now proposing that facilities that wish to document and report fluorinated GHG and N

O emissions reflecting the use of abatement systems adhere to a method that would require (1) documentation to certify that the abatement device is installed, operated, and maintained according to manufacturers' specifications, (2) accounting for the system's uptime, and (3) either certification that the abatement system is specifically designed for fluorinated GHG and N

O abatement and the use of EPA default DRE value, or directly and properly measured DRE (

i.e.,

in accordance with EPA DRE Protocol) confirming abatement system's performance.

Estimating Emissions from Heat Transfer Fluids.

To estimate the emissions from heat transfer fluids we proposed to require that electronics manufacturers use the 2006 IPCC Tier 2 approach, which is based on a mass-balance method. As we stated in the initial proposal, the 2006 IPCC Tier 2 approach uses company-specific data and accounts for differences among facilities' heat transfer fluids, leak rates, and service practices.

In comments we received on our initial proposal, it was noted that our proposed method for estimating emissions from heat transfer fluids would require companies to compile a detailed inventory of all fluorinated heat transfer equipment and its nameplate capacity. Comments stated that such a mass balance approach would be overly burdensome.

In evaluating these comments, we believe that there was some confusion regarding our intended method. As a result, we are not changing the broad outlines of our initial proposal, but we are clarifying required data elements.

3. Definition of the Source Category

The electronics industry uses multiple long-lived fluorinated GHGs such as PFCs, HFCs, SF

, and NF

, as well as N

O, during manufacturing of semiconductors, LCDs, MEMS, and PV. We understand that there are other electronics manufacturers such as those facilities that manufacture light-emitting diodes (LEDs) and disk readers that use fluorinated GHGs in similar manufacturing processes as semiconductors. As a result, we are seeking information on fluorinated GHG and N

O emissions associated with the manufacture of these products and also comment on whether to include them as part of the electronics manufacturing source category. It is our intent to include these other sources as part of the electronics manufacturing source category in the final rule where their emissions meet or exceed our proposed threshold of 25,000 mtCO

Fluorinated GHGs are used for plasma etching of silicon materials, cleaning deposition tool chambers, and wafer cleaning. N

O is also used in depositing certain films and chamber cleaning. Additionally, electronics manufacturing employs fluorinated GHGs (typically liquids at ambient temperature) as heat transfer fluids. The most common fluorinated GHGs in use for these purposes are CHF

(HFC-23), CF

, C

, NF

, SF

and Fluorinert

and Galden® heat transfer fluids; other compounds such as perfluoropropane (C

) and perfluorocyclobutane (c-C

) are also used in smaller quantities (EPA, 2008a). Table 3 of this preamble presents fluorinated GHGs typically used during manufacture of electronics devices.

Table 3—Examples of Fluorinated GHGs Used by the Electronics Industry

Product type

Fluorinated GHGs used during manufacture

Electronics (

e.g.,

Semiconductor, MEMS, LCD, PV)

, C

, c-C

O, C

, C

, CHF

, CH

, NF

, SF

, and Heat Transfer Fluids (CF3-(O-CF(CF3)-CF2)n-(O-CF2)m-O-CF3, CnF2n+2, CnF2n+1(O)CmF2m+1, CnF2nO, (CnF2n+1)3N)

IPCC Guidelines do not specify the fluorinated GHGs used for MEMS production. Literature reviews revealed that among others CF

, SF

, and the Bosch process (consisting of alternating steps of SF

and c-C

) are used to manufacture MEMS. For further information,

see

the Electronics Manufacturing TSD in the docket for this rulemaking (EPA-HQ-OAR-2009-0927).

Description of Electronics Manufacturing Processes and Activities.

Fluorinated GHG and N

O emissions result from the following electronics processes and activities:

(1) Plasma etching;

(2) Chemical vapor deposition;

(3) Chamber cleaning;

(4) Wafer cleaning; and

(5) Heat transfer fluid use.

Plasma etching, essential to fabricating intricate, nanometer size features in contemporary electronic devices, is the removal of solid material from a substrate surface with gaseous reactants, in plasma, to produce gaseous products, which are then pumped away and disposed. Unless abated, unreacted fluorinated reactants or fluorinated GHG by-products from etching are emitted into the atmosphere.

Typical fluorinated GHG etching reagents, used either individually or in combination, are CF

, CHF

, C

and c-C

for silicon dioxide and nitride films; CF

, NF

and SF

for polysilicon films; and CHF

for aluminum and SF

for tungsten films. A typical fluorinated GHG by-product from etching processes is CF

; in some instances C

may also be formed.

Deposition is a fundamental step in the fabrication of a variety of electronic devices. During deposition, layers of dielectric, barrier, or electrically conductive films are deposited or grown on a wafer or other substrate. Chemical vapor deposition enables the deposition of dielectric or metal films. During the chemical vapor deposition process, gases that contain atoms of the material to be deposited react on the wafer surface to form a thin film of solid material. Films deposited by chemical vapor deposition may be silicon oxide, single-layer crystal epitaxial silicon, amorphous silicon, silicon nitride, dielectric anti-reflective coatings, low k dielectric, aluminum, titanium, titanium nitride, polysilicon, tungsten, refractory metals or silicides. Nitrous oxide may be the oxidizer of choice during deposition of silicon oxide films.

Chambers used for depositing polysilicon, dielectric and metal films are cleaned periodically using fluorinated GHGs, N

O, and other gases. During the cleaning cycle, the gas is converted to fluorine atoms in plasma, which etches away residual silicon-containing material from chamber walls, electrodes, and chamber hardware. Undissociated fluorinated gases and other fluorinated and non-fluorinated products pass from the chamber to waste streams and, unless emissions control systems are employed, into the atmosphere.

Typical fluorinated GHGs used for chamber cleaning are NF

, C

and C

. N

O may also be used to reduce particle formation during chamber cleaning. As with etching films, fluorinated GHG by-products may be formed during chamber cleaning, typically CF

During wafer processing, any residual photoresist material can be removed through an ashing process, which consists of placing partially processed wafers in an oxygen plasma to which CF

may be added

The edges of wafers (the bevel) may require additional cleaning to remove yield-reducing residual material. Bevel cleaning may also use a plasma process with fluorinated gas chemistry. In both of these wafer cleaning processes, unused fluorinated GHGs are emitted unless abated.

Fluorinated GHG liquids (at ambient temperature) such as fully fluorinated linear, branched or cyclic alkanes, ethers, tertiary amines and aminoethers, and mixtures thereof are used as heat transfer fluids at several semiconductor facilities to cool process equipment, control temperature during device testing, and solder semiconductor devices to circuit boards. The fluorinated heat transfer fluid's high vapor pressures can lead to evaporative losses during use.

Electronics Manufacturing TSD (EPA-HQ-OAR-2009-0927); 2006 IPCC Guidelines.

Our understanding is that heat transfer fluids are widely used within semiconductor manufacturing. We are seeking comment on the extent of use and annual replacement quantities of heat transfer fluids in other electronics sectors, such as their use for cooling or cleaning during LCD manufacture.

Total U.S. Emissions From Electronics Manufacturing.

Emissions of fluorinated GHGs from 216 electronics facilities were estimated to be 6.1 million metric tons CO

e in 2006. Below is a breakdown of emissions by electronics product type.

Semiconductors.

Emissions of fluorinated GHGs, including heat transfer fluids, from 175 semiconductor facilities were estimated to be 5.9 million metric tons CO

e in 2006. Of the total estimated semiconductor emissions, 5.4 million metric tons CO

e are from etching/chamber cleaning and 0.5 million metric tons CO

e are from heat transfer fluid usage.

MEMS.

Emissions of fluorinated GHGs from 12 MEMS facilities were estimated to be 0.1 million metric tons CO

e in 2006.

LCDs.

Emissions of fluorinated GHGs from 9 LCD facilities were estimated to be 0.02 million metric tons CO

e in 2006.

PV.

Emissions of fluorinated GHGs from 20 PV facilities were estimated to be 0.07 million metric tons CO

e in 2006. We request comment on the number and capacity of PV facilities that employ thin film technologies (

i.e.,

amorphous silicon) and other PV manufacturing facilities in the United States using fluorinated GHGs.

For additional background information on the electronics industry, refer to the Electronics Manufacturing Technical Support Document (TSD) in the docket for this rulemaking (EPA-HQ-OAR-2009-0927).

4. Threshold for Reporting

For facilities that manufacture semiconductors, LCD, MEMS, and PV, we are proposing an emissions-based threshold of 25,000 mtCO

e. Consistent with other sections of the Final MRR, EPA is proposing that for the purposes of determining whether a facility emits amounts equal to or greater than 25,000 mtCO

e, a facility must include emissions from all source categories for which methods are provided in the rule. For purposes of the threshold determination under subpart I, we are proposing two different methods, depending on whether the facility

manufacturers semiconductors, MEMS, LCDs or PVs (

see

proposed section 98.91). It is important to note that these methods are only for determining whether a facility exceeds the threshold; the proposed methods required for monitoring and reporting emissions data are presented in section 5 below.

To determine whether a manufacturer falls above or below the proposed 25,000 mtCO

e threshold, we are proposing that semiconductor, MEMS, and LCD facilities use gas specific emission factors assuming 100 percent manufacturing capacity to calculate annual metric tons of emissions in CO

equivalents. Because we understand that heat transfer fluids are widely used within semiconductor manufacturing, we are proposing that semiconductor manufacturers add 10 percent of total clean and etch emissions at a facility to their estimate. For applicability purposes, we propose that manufacturing capacity means the facility's full planned design capacity.

The gas specific emission factors we are proposing to use for threshold applicability for semiconductors and LCD facilities are consistent with the 2006 IPCC Tier 1 emission factors. For MEMS, because there are no IPCC factors available, we are assuming that SF

accounts for 100 percent of the sector's total emissions. The emission factor we are proposing for threshold applicability is based on the assumption that the MEMS SF

emission factor is equivalent to the IPCC Tier 1 SF

emission factor for semiconductors, scaled up by a factor of 5.

For a more detailed explanation of MEMS default factor, please refer to the Electronics Manufacturing TSD (EPA-HQ-OAR-2009-0927).

We are proposing that PV facilities multiply annual fluorinated GHG purchases or consumption by the gas-appropriate 100-year GWPs, as defined in Table A-1 of subpart A of part 98, to calculate annual metric tons of emissions in CO

equivalents. None of these calculations would account for emission abatement systems.

We are proposing to require an emissions estimating method that does not account for destruction by abatement systems because actual emissions from facilities employing abatement systems may exceed estimates when based on the manufacturers' rated DREs of the equipment and may therefore exceed the 25,000 mtCO

e threshold without the knowledge of the facility operators. When abatement equipment is used, electronics manufacturers often estimate their emissions using the manufacturer-supplied DRE for the system. However, an abatement system may fail to achieve its rated DRE either because it was not installed properly, is not being properly operated and maintained, or because the DRE value itself was incorrectly measured due to a failure to properly account for the effects of dilution. For example, reported DREs for CF

can be overstated by as much as a factor of 20 to 50, and the corresponding figure for C

can be overstated by a factor of up to 10 because of failure to properly account for dilution (Burton, 2007).

In our analysis of the emissions thresholds, we considered thresholds of 1,000 mtCO

e, 10,000 mtCO

e, 25,000 mtCO

e, and 100,000 mtCO

e per year. To estimate the number of semiconductor facilities that would have to report under each of the various thresholds, we estimated emissions for each facility in the U.S. by using IPCC Tier 1 emission factors. These emissions estimates were then evaluated to determine how many facilities would meet the various thresholds. To estimate the collective emissions from the facilities that would have to report under the various thresholds, we used information from EPA's PFC Reduction/Climate Partnership for Semiconductors and the EPA PFC Emissions Vintaging Model.

To estimate the number of LCD and PV facilities that would have to report under the various thresholds, as well as the collective emissions from these facilities, we used IPCC Tier 1 emission factors. Because IPCC emission factors for MEMS are not available, the number of facilities that would have to report and the collective emissions from these facilities were determined using an emission factor based on a relevant IPCC Tier 1 emission factor for semiconductor production.

All of our analyses assumed no abatement.

For a more detailed explanation of MEMS default emission factor, please refer to the Electronics Manufacturing TSD (EPA-HQ-OAR-2009-0927).

Table 4 of this preamble shows emissions and facilities that would be captured by the respective emissions thresholds.

Table 4—Threshold Analysis for Electronics Industry

Emission threshold level metric tons CO

e/yr

Total

national

emissions

Total number of facilities

Emissions covered

metric tons CO

e/yr

Percent

Facilities covered

Facilities

Percent

1,000

5,984,463

216

5,962,091

99.6

165

10,000

5,984,463

216

5,813,200

114

25,000

5,984,463

216

5,622,570

100,000

5,984,463

216

4,737,622

We selected the 25,000 mtCO

e per year threshold because it maximizes emissions reporting, while excluding small facilities that do not contribute significantly to the overall GHG emissions.

Table 5 of this preamble shows the estimated emissions and number of facilities that would report for each type of source under the proposed emissions-based thresholds.

Table 5—Summary of Rule Applicability Under the Proposed Thresholds

Emissions source

Threshold

Total national facilities

Total emissions of source (metric tons CO

Emissions covered

metric tons CO

e/yr

Percent

Facilities covered

Facilities

Percent

Semi-conductors

25,000 Mt CO

Eq.

175

5,741,676

5,492,066

MEMS

25,000 Mt CO

Eq.

146,115

96,164

LCD

25,000 Mt CO

Eq.

23,632

25,000 Mt CO

Eq.

73,039

34,340

The proposed emissions-based thresholds are estimated to include approximately 50 percent of semiconductor facilities and between approximately 5 percent and 17 percent of the facilities manufacturing PV and MEMS, respectively. At the same time, the thresholds are expected to cover nearly 96 percent of fluorinated GHG emissions from semiconductor facilities, 66 percent of fluorinated GHG emissions from facilities manufacturing MEMS, and 47 percent of fluorinated GHG emissions from facilities manufacturing PV. Combined, these emissions are estimated to account for close to 94 percent of fluorinated GHG emissions from the electronics industry as a whole.

Based on our current analysis, facilities manufacturing LCDs are not expected to meet the proposed threshold. In addition, only 2 MEMS facilities and 1 PV facility are expected to be covered. The data and information that we currently have on MEMS, LCD, and PV manufacturing, however, is limited and incomplete. We are including these sectors because they have similar fluorinated GHG and N

O use and manufacturing processes as those of semiconductor manufacturing and they are high growth sectors. We estimate that emissions from MEMS, LCD, and PV may be higher than our data show currently and we expect them to increase in the future.

For additional background information on the threshold analysis, refer to the Electronics Manufacturing TSD. For specific information on costs, including unamortized first year capital expenditures, please refer to the EIA and the EIA cost appendix.

5. Selection of Proposed Monitoring Methods

We are proposing methods to monitor and estimate fluorinated GHG and N

O emissions from semiconductor, LCD, MEMS, and PV manufacture. The proposed methods discussed below include the following: (a) Estimating emissions from cleaning and etching processes; (b) estimating facility N

O emissions; (c) estimating emissions from heat transfer fluids; and (d) reporting controlled emissions from abatement equipment. The methods described and proposed in this section are for estimating emissions that would be required to be reported under this subpart (

see

proposed sections 98.93 and 98.94). It is important to note that these methods differ from those proposed in the section above which are for determining applicability of the subpart.

a. Methods for Estimating Emissions From Cleaning and Etching Processes

We are proposing different methods for estimating fluorinated GHG emissions from etching and cleaning based on whether the facility is a semiconductor manufacturer or an LCD, MEMS, or PV manufacturer.

Method for Semiconductor Facilities.

Under this proposal, all semiconductor manufacturers that have emissions equal to or greater than 25,000 mtCO

e would be required to estimate and report emissions from etching and cleaning using one of two approaches. First, we are proposing an approach, hereinafter referred to as the “Refined Method,” that is based on:

(1) Gas consumption as calculated using the facility's purchase records, inventory, and gas- and facility-specific heel factors,

(2) Facility-specific methods for apportioning gas consumption by process category

using indicators of GHG-using activity (

e.g.,

wafer passes),

For purposes of electronic manufacturing, “process category” is a set of similar manufacturing steps, performed for the same purpose, associated with substrate (

e.g.,

wafer) processing during device manufacture for which fluorinated GHG and N

O emissions and fluorinated GHG and N

O usages are calculated and reported.

(3) Emission factors for utilization and by-product formation rates based on refined process categories (

e.g.,

categories with more specificity than the simpler cleaning and etching categories listed in the 2006 IPCC Guidelines), and

(4) Methods for reporting controlled emissions (as proposed below).

Alternatively, we are proposing to permit those facilities that have monitoring infrastructure or the necessary data to estimate emissions obtained through recipe-specific measurements to report their emissions using their data by following an approach consistent with the 2006 IPCC Tier 3 method. In addition, for those semiconductor manufacturers that fabricate electronic devices on wafers of measuring greater than 300 mm in diameter, we are proposing to require that they estimate and report their emissions using recipe-specific measurements and follow an approach consistent with the IPCC Tier 3 method. Each of these approaches is discussed below.

Refined Method.

The Refined Method would apply to all covered semiconductor facilities and would not make a distinction between relatively large and other facilities. In the paragraphs below, we discuss in detail each one of the components we are proposing to require under this approach.

Gas consumption as calculated using the facility's purchase records, inventory, and gas- and facility-specific heel factors.

Notwithstanding the definition of “heel” in subpart A of this rule,

we are proposing that for purposes of electronics manufacturing that a heel means, “the amount of gas that remains in a gas cylinder or container after it is discharged or off-loaded (this may vary by cylinder or container type and facility).” We are not planning to use the subpart A definition because it contains a default value of 10 percent. In this action, we are proposing to require facilities to calculate gas- and facility-specific heel factors rather than using a default value.

Pursuant to subpart A of the Final MRR, “heel” means the amount of gas that remains in a shipping container after it is discharged or off-loaded (that is no more than ten percent of the volume of the container).

As part of determining each facility's overall usage of each gas for a reporting period, we are proposing that a facility use their purchase records, inventory, and gas- and facility-specific heel factors. More specifically, for each cylinder/container type for each gas used, we are proposing that semiconductor facilities be required to base their heel factors on the residual

weight or pressure of the gas cylinder or container that a facility uses to change out that cylinder/container. This is common practice in the industry and is typically referred to as the “trigger point for change out.” These points, one for each gas and cylinder/container type, together with the initial container mass or pressure, are used to calculate the unused gas for each container, which when expressed as a fraction of the initial amount in the container is the “heel” (or unused fraction of the container). This gas- and facility-specific heel factor would then be applied to each container for that gas to determine the net amount of that gas used at a facility. In cases where the “trigger point for change out” used at a facility differs by more than one percentage point from that used to calculate the previous gas-specific heel factor, we propose that the gas- and facility-specific heel factor must be recalculated.

Currently most semiconductor facilities rely upon the IPCC default heel factor of 10 percent and apply that value to each cylinder/container. Based on information provided in an industry study of facility-specific, gas-specific heel factors, the heel factor in a given facility for individual cylinders/containers can vary from 3 percent to 25 percent. Given this variation, we conclude that gas- and facility-specific heel factors would provide improved accuracy in emissions estimates over the use of the IPCC default heel factor.

We understand that there are exceptional circumstances when facilities do not always change cylinders/containers exactly when they reach the targeted residual weight or pressure. In those instances, which we expect are infrequent, we are proposing that the cylinder/container must be weighed or the pressure measured using a pressure gauge; as opposed to using the facility-wide gas-specific heel factor as part of determining the net amount of gas used at a facility. We are proposing to define an exceptional circumstance as one which the cylinder/container is changed at a residual mass or pressure that differs by more than 20 percent from the “trigger point for change out.” We request comment on the frequency of these exceptional circumstances and also the percentage difference (

i.e.

20 percent) for which we are proposing to require that the exceptional cylinder/container be weighed or the pressure measured.

When taking an annual inventory, we understand that multiple cylinders/containers are in service. We request comment on the significance of accounting for the quantity of fluorinated GHGs or N

O remaining in cylinders/containers in service at the end of the reporting period. We also request comment and detailed information on other methods and technologies (

i.e.

other than purchase records) that facilities may be using for determining annual gas consumption (

e.g.,

recorded data from an automated gas inventory system).

We are proposing that all flowmeters, weigh scales, pressure gauges, and thermometers used to measure quantities that are monitored or used in calculations in this proposal have an accuracy and precision of 1 percent of full scale or better. We request comment on this requirement including alternative accuracy and precision requirements and detailed information about why particular instruments can not meet the proposed 1 percent standard.

Apportioning gas consumption to process categories.

Estimating facility emissions requires apportioning annual facility-wide gas consumption across a facility's emitting process categories by way of applying facility-specific apportioning factors. A facility's uncontrolled emissions are the product of that apportioned gas consumption and the corresponding emission factor. To determine the share of each gas used by each process category, we are proposing to require that semiconductor facilities use a quantifiable indicator (or metric) of gas usage activity. More specifically, we are proposing facilities track wafer passes as an indicator of activity with which to apportion the facility's gas consumption. Wafer passes is a count of the number of times a silicon wafer is processed for a specific process category. The total number of wafer passes over a reporting year is the number of wafer passes per tool times the number of operational process tools during the reporting year.

To illustrate a case where wafer passes is used as a facility-specific engineering model, consider a facility that uses NF

for chamber cleaning with remote plasma systems and for etching polysilicon and oxide films. With knowledge of the NF

-specfic heel and the number of NF

containers used, the facility knows the amount of NF

consumed. To estimate emissions, the facility must now apportion NF

usage between the chamber cleaning and oxide and polysilicon etching processes. To do this it might use the total number of wafer passes through each and every NF

-cleaning system together with the time and nominal (not measured actual) gas flow rate for each and every NF

-cleaning system and the corresponding figures for oxide and polysilicon etch processes to arrive at the proportion of NF

used for cleaning chambers and etching oxide and polysilicon films. Once developed, these apportioning factors would be used to estimate NF

gas usage for the cleaning and etching process categories proposed in our method. This example is illustrated further in Table 6 of this preamble.

Table 6—Illustrative Calculation for NF

Example at One Facility

Gas type—annual usage, kg.

Process category

Apportioning factor

Process category gas usage, kg.

—56,286 kg

RPS Chamber Cleaning

82%

46,202

Polysilicon Etch

17%

9,561

Oxide Etch

523

Annual gas usage presented is the modeled usage not the nominal usage.

We request comment on using wafer passes as an appropriate quantifiable indicator of activity, and on our description and example of how it would be used.

We recognize that facilities may use other types of quantifiable indicators of gas-usage activity data to develop facility-specific engineering models to estimate gas consumption. We may include additional indicators as options in the final rule if they are quantifiable and if we receive adequate information regarding how they were developed and how they are used, including descriptions, examples, and any additional information that may be necessary to understand how such indicators of activity would be developed and used in a facility-specific engineering model to apportion annual

facility-wide gas usage across a facility's emitting process categories. The use of engineering judgment, for example, is not based on a quantitative metric and would not be considered an acceptable quantifiable indicator of gas usage. We also request comment on the use of a representative sampling method for tracking activity indicators such as wafer passes that may be used in the engineering model.

In many cases, EPA anticipates that the development of apportioning factors will result in a facility-wide consumption estimates that are independent of the estimates calculated using purchase records, inventory, and facility-specific heel factors. In such cases, we propose that facilities report these consumption estimates.

Emission factors for refined process categories.

We are proposing that semiconductor facilities estimate their emissions using a specific set of process categories. Our proposed method would simplify the reporting requirements as compared to the 2006 IPCC Tier 3 method by lowering the number of emitting process categories from up to 455 per facility down to a fixed figure of approximately nine. Our goal in establishing the process categories is to account for most of the variability in emission factors across processes while limiting the total number of process categories whose gas usage must be tracked by semiconductor facilities.

Under this approach, we are proposing to require reporting of fluorinated GHG emissions for the following nine emitting process categories: four subcategories for wafer patterning (etching), three subcategories for chamber cleaning, and two subcategories for wafer cleaning. The nine process categories we are proposing account for distinct and widely-used manufacturing activities during production of discrete, logic and memory devices. We anticipate that these nine categories effectively capture current and projected processes and the differences in emission factors across various semiconductor manufacturing technologies.

Our proposed definitions of these nine emitting categories are:

Wafer patterning subcategories:

Oxide etch

means any process using fluorinated GHG reagents to selectively remove SiO

SiO

-based or fully organic-based thin-film material that has been deposited on a wafer during semiconductor device manufacturing.

Nitride etch

means any process using fluorinated GHG reagents to selectively remove SiN, SiON, Si

SiC, SiCO, SiCN,

etc.

(represented by the general chemical formula, Si

where w,x,y and z are zero or integers and X can be some other element such as carbon) that has been deposited on a wafer during semiconductor manufacturing.

Silicon etch

also often called

polysilicon etch

means any process using fluorinated GHG reagents to selectively remove silicon during semiconductor manufacturing.

Metal etch

means any process using fluorinated GHG reagents associated with removing metal films (such as aluminum or tungsten) that have been deposited on a wafer during semiconductor manufacturing.

Chamber cleaning subcategories:

In situ plasma

means cleaning thin-film production chambers, after processing one or more wafers, with a fluorinated GHG cleaning reagent that is dissociated into its cleaning constituents by a plasma generated inside the chamber where the film was produced.

Remote plasma system

means cleaning thin-film production chambers, after processing one or more wafers, with a fluorinated GHG cleaning reagent dissociated by a remotely located (

e.g.,

upstream) plasma source.

In situ thermal

means cleaning thin-film production chambers, after processing one or more wafers, with a fluorinated GHG cleaning reagent that is thermally dissociated into its cleaning constituents inside the chamber where the thin-film (or thin films) was (were) produced.

Wafer cleaning subcategories:

Bevel cleaning

means any process using fluorinated GHG reagents with plasma to clean the edges of wafers during semiconductor manufacture.

Ashing

means any process using fluorinated GHG reagents with plasma to remove photoresist materials during wafer manufacture.

We request comment on the nine process categories we are proposing, their definitions as specified above, and whether they clearly define a specific process without ambiguity. In addition we request comment on whether the categories should be further refined to better capture the variability in emission rates among fluorinated GHG using manufacturing activities (

e.g.,

whether any additional categories should be added or whether the proposed categories should be combined, and the definition of those categories).

Under this approach of defining a specific set of process categories, we are also considering additional patterning and chamber cleaning subcategories. The alternative patterning subcategories, which may replace or complement the four thin-film based subcategories defined previously, are: contact etch, self-alignment contact etch, gate etch, deep trench etch, isolation trench etch, through silicon vias and regular vias. Each of these subcategories represents a specific feature achieved through etching (instead of subcategories based on the type of thin film etched).

Alternative chamber cleaning categories may distinguish between the types of films being removed from the chamber during cleaning. These might include distinguishing between chambers coated with tungsten and silicon-based films, or distinguishing between thin-film deposition equipment manufacturers. We request comment on these additional process categories and whether or not we should include alternative process categories in addition to the nine process categories that we are proposing. We also request comment on other methods of categorizing processes and detailed information on those categories.

We are proposing nine process categories differentiated by production technology generation (

i.e.,

wafer size). For each of the proposed nine process categories, we are proposing to establish a default emission factor within a range of values presented in Tables I-6, I-7, I-8 of subpart I. Within each process category, factors account for (1) the mass fraction of the input gas that is utilized during (

i.e.,

not emitted from) the process and (2) the mass of each fluorinated GHG by-product formed as a fraction of the mass of the dominant fluorinated GHG input gas used.

EPA is proposing a range of values for each default emission factor because the Agency has not yet received sufficient data to select a specific value within each range.

In the case of mixtures of fluorinated GHGs, the “dominant” fluorinated GHG constitutes the largest mass of gas used for that process.

To develop the proposed ranges for each emission factor, EPA requested from semiconductor device manufacturers and equipment suppliers, information on utilization and by-product formation rates and details on the associated measurement approach (

e.g.,

measured in accordance with the 2006 ISMI Guidelines). EPA evaluated the data received as well as the standard deviations provided in Table 6.9 from Chapter 3 of the 2006 IPCC Guidelines. For additional information on how the ranges were developed, please refer to the Electronics Manufacturing TSD (EPA-HQ-OAR-2009-0927).

In a final rule, EPA intends to publish default emission factors for gas utilization and by-product formation rates for each process category,

differentiating amongst 150 mm, 200 mm and 300 mm wafer technology to the extent feasible. To this end, EPA requests additional utilization and by-product formation rates and supporting information on how they were developed. More specifically, EPA requests emission factors and by-product formation rates and information including but not limited to the specific measurement method used (

e.g.,

measurement using the 2006 ISMI Guidelines), the date of measurement, achievement of fluorine mass balance, associated standard deviations of measured factors, the relevant emissions process types and categories (for the patterning/etching process type noting both film type and etched feature where applicable), substrate size (

i.e.,

150 mm, 200 mm, or 300 mm), the number of wafers used in the measurement study, and the equipment manufacturer name and model number where not considered confidential.

Using additional data received, EPA intends to develop default emission factors for each process category using a method of aggregation similar to the 2006 IPCC factor development methodology.

Where available emission factor data are very limited or produce highly uncertain average factors, EPA may develop emissions factors that are conservative and less likely to underestimate actual emissions. If additional data are received in a timely fashion, EPA may develop draft emission factors prior to issuance of the final rule and will determine an appropriate way to promptly and clearly inform the regulated community. We welcome comments on such draft emission factors, recognizing that depending on when the emission factors are made available, such comments could be submitted after the close of the formal comment period. We will make every effort to consider such comments, including late comments, to the extent practicable in the development of the final rule.

For additional information on the 2006 IPCC factor development methodology,

see Emission Factors for Semiconductor Manufacturing: Sources, Methods, and Results

(February 2006) available in the docket (EPA-HQ-OAR-2009-0927).

In developing emission factors for the final rule, EPA is also considering developing weighted average emission factors, for each wafer technology, with the weights based on the market penetration rates of process recipes used in current device manufacturing practices.

Such weighted emission factors, if possible, may better represent actual emissions from installed manufacturing equipment and operating processes. We request comment on using a weighting scheme and detailed information on how it would be developed and implemented.

Note,

in the creation of the IPCC factors, sufficient information was not available to weigh each general process type (

i.e.,

etch and clean categories for the IPCC Tier 2b method).

The uncertainties associated with the 2006 IPCC Tier 2b method are associated with aggregating, for each gas, all usage into just two process categories (

i.e.,

etching and chamber cleaning) and all wafer technologies (

i.e.,

150 mm, 200 mm, and 300 mm wafer sizes) into one, and giving equal weights to all process recipes. A method based on refined processes categories keeps those processes separate, which reflects actual device manufacturing practices and as a result, produces a more representative and accurate emissions estimate.

As an alternative, we are also considering an approach where each facility would develop for themselves or acquire from process equipment manufacturers emission factors (

i.e.,

gas utilization and by-product formation rates) for the nine process categories. Under this approach, we would require the gas utilization and by-product formation rates to be developed using the 2006 ISMI Guidelines. Facilities would be required to construct and apply averages for each process category. One advantage of this approach is that these facility-specific emission factors would be expected to be more representative of the particular processes at that facility than the default emission factors. On the other hand, we estimate the burden associated with each facility developing its own emission factors would be greater compared to using the factors published by EPA. We request comment on this approach.

We recognize that given the dynamic manufacturing processes by the industry, updates to the process categories and emission factors may be necessary. We request comment on the frequency with which those should be updated.

We estimate that our Refined Method will result in a reduction in burden for the large semiconductor facilities (annual capacities greater than 10,000 m

silicon) and an increase in accuracy as compared to the IPCC Tier 2b method. We estimate the uncertainty from using a set of refined process categories to be roughly one-half the uncertainty of the Tier 2b method, assuming similar methods for apportioning gas usage for each method. For the Tier 2b method the fluorinated GHG consuming processes used during semiconductor production are collapsed into just two categories, resulting in considerable variability for each category. For the Refined Method there are nine fluorinated GHG-using categories, resulting in less variability, on average, per category. Please refer to the Electronics Manufacturing TSD for a more detailed discussion of our uncertainty analysis.

For the relatively smaller semiconductor facilities (annual production of less than 10,500 m

of silicon) we estimate an increase in burden as compared to our initial proposal where we required the use of the 2006 IPCC Tier 2b method; however, we anticipate that these facilities have the necessary data available to comply. The increase in burden for estimating emissions using the Refined Method, as opposed to the IPCC Tier 2b method, can be attributed to the increased level of effort to distinguish between nine refined process categories in comparison to two broad clean and etch categories, respectively.

Recipe-specific measurements.

As an alternative to the Refined Method where EPA default factors would be used, we are also proposing to permit those facilities that have monitoring infrastructure or the necessary data to estimate emissions obtained through recipe-specific measurements to report their emissions using their data (

see

proposed sections in 98.93 98.94(d)). This approach, consistent with the 2006 IPCC Tier 3 method, is based on (1) gas consumption as calculated using the facility's purchase records, inventory, and gas-and facility-specific heel factors (as described above), (2) facility-specific methods for apportioning gas consumption by individual process using indicators of GHG-using activity, (3) recipe-specific gas utilization and by-product formation factors, and also (4) methods for reporting controlled emissions from abatement devices (as proposed below). Under this approach, gas utilization and by-product formation rates would be required to be developed using the 2006 ISMI Guidelines for all fluorinated GHG-using process types at that facility.

According to information provided by one of the commenters in response to our initial proposal, only one company currently estimates their emissions using an approach consistent with the Tier 3 method. Nevertheless, if a facility is using a method that provides more accurate data, then we believe that they should be permitted to use such method. We request comment on the number of companies that are currently

or expecting to in the near future, report their emissions using this method.

We are also proposing to require semiconductor manufacturers that fabricate devices on wafers measuring larger than 300 mm in diameter to estimate their emissions based on an approach consistent with the IPCC Tier 3 method and gas- and facility-specific heel factors for estimating and reporting GHG emissions. Under this approach, gas utilization and by-product formation rates would be required to be developed using the 2006 ISMI Guidelines for all fluorinated GHG using process types at that facility. We understand the industry's conversion to 450 mm is expected to begin in 2011 or shortly thereafter. We are proposing this requirement because we estimate that this method that uses recipe-specific gas utilization and by-product formation factors results in the most accurate facility-specific emission estimate. By including this requirement for only the 450 mm or larger wafers in this proposal, we anticipate a reduction in burden as compared to requiring existing large semiconductor facilities to estimate their emissions using an approach consistent with the IPCC Tier 3 method for the smaller sized wafers as well (

i.e.

300 mm and smaller). We anticipate a reduction in burden because emission factors (

i.e.

gas utilization and by-product formation rates) can be developed over a number of years as semiconductor manufacturers begin to transition to 450 mm tools and develop the estimating and reporting infrastructure. The commissioning process for new tools is an ideal opportunity for emission factor development and/or verification. We request comment on requiring semiconductor manufacturers that fabricate electronic devices on wafers of diameter 450 mm or larger to estimate their emissions based on an approach consistent with the IPCC Tier 3 method.

During the development of this proposal, the 2006 International SEMATECH Manufacturing Initiative's Guideline for Environmental Characterization of Semiconductor Process Equipment was revised and republished (December 2009). We request comment on requiring the use of the revised version of the ISMI Guidelines to measure emission factors as opposed to the 2006 version of the ISMI Guidelines, and also information on emission factors (including utilization by-product formation rates) measured using the revised ISMI Guidelines.

Method for LCD, MEMS, and PV Facilities.

In this action for LCD, MEMS, and PV facilities, we are proposing an approach based on a slightly modified 2006 IPCC Tier 2b method which would include (1) gas consumption calculated using the facility's purchase records, inventory, and gas- and facility-specific heel factors (as described above for semiconductor manufacturing facilities), (2) gas consumption apportioned to 2006 IPCC Tier 2b broad process categories, clean and etch, (3) default emission factors consistent with the 2006 IPCC Tier 2b factors, and (4) methods for reporting controlled emissions from abatement equipment (as proposed below).

The method proposed to develop the gas- and facility-specific heel factors for LCD, MEMS, and PV facilities is the same as proposed for semiconductor facilities including the provisions for exceptional circumstances. Although we don't have complete information on how LCD, MEMS, and PV facilities are currently estimating their emissions from manufacture and how they are currently accounting for heels, their gas use and manufacturing processes are similar to that of semiconductor manufacturing. As a result, we have concluded these facilities have the data required to develop a gas- and facility-specific heel factors and this method can be implemented with minimal burden. Similar to the semiconductor manufacturing case, the use of a gas- and facility-specific heel factor is expected to result in improved accuracy when compared to the IPCC's 10 percent default factor. We request comment on our proposal to require LCD, MEMS, and PV facilities to use gas- and facility-specific heel factors and our understanding that these facilities have the data to develop such a factor with minimal burden.

Under this approach consistent with the 2006 IPCC Tier 2b method, we propose that LCD, MEMS, and PV manufacturing facilities use the calculated mass of gas consumed and apportion this amount to the simplified process categories (

i.e.

etch and chemical vapor deposition chamber cleaning.) The associated emission factors including utilization and by-product formation rates, would then be used to calculate uncontrolled fluorinated GHG emissions. The emission factors being proposed are consistent with the 2006 IPCC default values. For MEMS manufacturing, where an IPCC default value does not exist, we propose the use of factors consistent with the 2006 IPCC Tier 2b factors for semiconductor manufacturing. We selected these factors because we understand MEMS manufacturing is silicon wafer-based and uses processes similar to those found in semiconductor manufacturing.

Additionally, we are proposing that LCD, MEMS, and PV manufacturing facilities abide by the requirements proposed for reporting controlled emissions from abatement systems as proposed below.

We are requesting information on emissions and emission factors from LCD, MEMS, and PV manufacturing. We are requesting such information as a means to verify that the Tier 2b emission factors for each of the manufacturing types are reflective of current fluorinated GHG emitting processes. Based on new information we receive, we may consider updating the emission factors in the final rule.

We expect that LCD, MEMS, and PV manufacturers may also use engineering models and quantifiable indicators (

e.g.,

substrate-area based) of manufacturing activity for apportioning gas consumption by process category similar to the approach described for semiconductors above (

e.g.,

wafer passes). We request detailed information on those indicators, how they were developed, and how they are used in a facility-specific engineering model to apportion annual facility-wide gas usage across a facility's emitting process categories.

We request comment on permitting those LCD, MEMS, and PV manufacturing facilities that have monitoring infrastructure or the necessary data to estimate emissions obtained through recipe-specific measurements to report their emissions using their data by following an approach consistent with the 2006 IPCC Tier 3 method.

Review of Existing Reporting Programs and Methodologies and Consideration of Alternative Methods.

EPA considered various methods for estimating emissions from etching and cleaning processes for electronics manufacturing facilities including the 2006 IPCC Tier 1, 2a, 2b, and Tier 3 method as well as a Tier 2b/3 hybrid which would apply Tier 3 to the most heavily used fluorinated GHGs in all facilities. For a detailed description of our evaluation of these options, please

see

the Electronics Manufacturing section of the initial Mandatory Reporting Rule (74 FR 16499).

For this proposal, to estimate emissions from all semiconductor manufacturing facilities, we are also considering the alternative of a modified Tier 2b method (our preferred option for other electronics manufacturers) which would require the use of the 2006 IPCC Tier 2b default factors and gas- and facility-specific data on heels and gas

use by process category. This approach would be based on a modified version of the 2006 IPCC Tier 2b method for estimating emissions and would require semiconductor facilities to report emissions using (1) gas consumption as calculated using the facility's purchase records, inventory, and gas- and facility-specific heel factors (as described above), (2) facility-specific methods for apportioning gas usage by process category using indicators of activity (as described above,

e.g.,

wafer pass), (3) IPCC Tier 2b emission factors, and (4) methods for reporting controlled emissions using our proposed approach discussed below. We request comment on this approach.

As an alternative to the Refined Method, we are also considering requiring all semiconductor manufacturing facilities to estimate their emissions using an approach consistent with the IPCC Tier 3 method based on (1) gas consumption as calculated using the facility's purchase records, inventory, and gas- and facility-specific heel factors, (2) facility-specific methods for apportioning gas consumption by individual process using indicators of GHG-using activity, (3) recipe-specific gas utilization and by-product formation factors, and also (4) methods for reporting controlled emissions from abatement devices (as proposed below). Under this approach, facilities would be required to develop gas utilization and by-product formation rates using the 2006 ISMI Guidelines for all fluorinated GHG-using process types at that facility. We request comment on this approach.

Another option we are considering is to evaluate emissions from electronics manufacturing using continuous emission monitoring system(s) (CEMS). Under this approach, facilities would be required to install and operate CEMS to measure process emissions. A typical electronics manufacturing facility may have many individual process tools that influence emissions. Process tool exhaust is managed within the facility using stainless steel plumbing and ductwork. Due to the complexity of the manufacturing layout, CEMS would be attached either to every tool or to one or more final exhaust points (

e.g.,

scrubber stacks). One possible option is to use Fourier Transform Infrared Spectrometers (FTIRs) in scrubber stacks to measure facility emissions. FTIR spectroscopy is presently used to conduct short-term fluorinated GHG emission measurements from single tools. EPA requests comment on the use of CEMS at electronics manufacturing facilities. We also request data and other information evaluating the use of CEMS in electronics facilities to determine fluorinated GHG and N

O emissions.

(b) Method for Estimating N

O Emissions

We are proposing that electronics manufacturers estimate N

O emissions from chemical vapor deposition processes and all other electronics manufacturing processes such as chamber cleaning, and that they estimate those emissions using the following proposed methods.

To estimate N

O emissions from chemical vapor deposition we are proposing the use of a facility-specific emission factor based on facility measurements of N

O utilization for chemical vapor deposition, using 2006 ISMI Guidelines. Under this approach, we propose to permit the facility to apply the average N

O utilization emission factor to all N

O using chemical vapor deposition recipes. In cases where a facility has not developed a facility-specific N

O utilization factor for chemical vapor deposition processes, we are proposing a default value in the range of 0 to 40 percent. We are taking comment on this range due to a lack of information for N

O utilization for chemical vapor deposition processes.

In comments received in response to our initial proposal, industry provided information to support a N

O utilization factor of 40 percent, primarily in 300 mm chemical vapor deposition processes. Taking the industry-provided 40 percent utilization into account, we propose to select a N

O utilization factor in the range from 0 to 40 percent. In the industry's survey, the measured utilization factors are largely from newer 300 mm manufacturing equipment. We do not expect these data fairly represent the entire population of all N

O processes and installed equipment, many of which are older tools. In addition, the industry comments did not fully identify the specific processes from which the average N

O utilization factor was calculated. For these reasons, and because we understand that N

O is most commonly used for chemical vapor deposition as opposed to other processes, we are proposing to establish a default value within a range of values with 40 percent as the upper bound and 0 percent as the lower bound to be conservative, reducing potential for underestimating emissions.

To estimate N

O emissions from all other manufacturing processes (

e.g.,

chamber cleaning), we are proposing either a facility-specific utilization factor based on measurements using 2006 ISMI Guidelines, or applying a default utilization factor of 0 percent which assumes N

O is not converted or destroyed during the manufacturing process. We are proposing this method due to a lack of information regarding other processes for which N

O is used and N

O utilization data in those processes.

We request comment on values within the range that we are proposing to estimate N

O emissions from chemical vapor deposition processes and our approach for estimating N

O emissions from all other manufacturing processes. We also request additional information on N

O uses and N

O utilization in electronics manufacturing processes. More specifically, we request N

O emission factors and detailed supporting information including but not limited to the specific measurement method used, date of measurement, standard deviation of measured factors, identification of manufacturing process or process category, substrate size, and equipment manufacturer name and model number where not considered confidential.

In addition, we request comment on using wafer passes or other appropriate quantifiable indictors of activity for apportioning N

O consumption to chemical vapor deposition and other manufacturing processes.

We are proposing that as part of determining annual facility N

O emissions, if a facility employs abatement systems and it wishes to report N

O emission reductions due to these systems it must adhere to the methods for reporting controlled emissions included in this proposal.

To estimate the emissions of heat transfer fluids, we propose that electronics manufacturers use the 2006 IPCC Tier 2b approach, which is a mass-balance approach. We are not changing the broad outlines of our initial proposal; however, we are clarifying required data elements.

In evaluating the comments we received, we understand that there was some confusion regarding our intended method. The proposed method required data on the total nameplate capacity

of equipment that “is installed during the reporting year.” We intended “installed during the reporting year” to mean newly installed during the period,

not in place from the beginning of that period. To eliminate confusion, we are clarifying that facilities are required to provide the total nameplate capacity (charge) of equipment that is “newly installed” during the reporting year. We anticipate that facilities will find it straightforward to track the nameplate capacities of equipment that is newly installed or retired during the reporting year.

Nameplate capacity means the full and proper charge of gas specified by the equipment manufacturer to achieve the equipment's specified performance. The nameplate capacity is typically indicated on the equipment's nameplate; it is not necessarily the actual charge, which may be influenced by leakage and other emissions.

In addition, we are also clarifying that a facility may only subtract the amount of fluorinated heat transfer fluids sent off site if the heat transfer fluids are properly recovered, stored, and sent off site for verifiable recycling or destruction during the reporting year. We are adding this clarification because we understand that facilities may be recovering, storing, and removing from their facility, fluorinated heat transfer fluids in a manner that does not effectively prevent the substance(s) from evaporating to the atmosphere. In such cases, the users of the chemicals would be required to account for these emissions using the mass-balance calculation provided.

As we stated in our initial proposal, in developing our proposal for estimating heat transfer fluid emissions, we reviewed both the IPCC Tier 1 and IPCC Tier 2 approaches. The Tier 1 approach for heat transfer fluid emissions is based on the utilization capacity of the semiconductor facility multiplied by a default emission factor. Although the Tier 1 approach has the advantages of simplicity, it is less accurate than the Tier 2 approach according to the 2006 IPCC Guidelines. The IPCC Tier 2 approach uses company-specific data and accounts for differences among facilities' heat transfer fluids (which vary in their GWPs), leak rates, and service practices. It has an uncertainty on the order of ±20 percent at the 95 percent confidence interval according to the 2006 IPCC Guidelines.

(d) Method for Reporting Controlled Emissions From Abatement Systems

For this proposed rule, we are defining DRE as the efficiency of a control system designed to destroy or remove fluorinated GHGs, N

O, or both. The DRE is equal to one minus the ratio of the mass of all relevant GHGs exiting the emission abatement system to the mass of GHGs entering the emission abatement system. When fluorinated GHGs are formed in an abatement system, DRE is expressed as one minus the ratio of amounts of exiting GHGs to the amounts entering the system in units of CO

-equivalents. In addition, we are clarifying facilities may account for all abatement systems (

e.g.,

multi-chamber POU, central devices) provided that they abide by the requirements below.

We are proposing to use the term destruction or removal efficiency (DRE) as opposed to “destruction efficiency” or “destruction,” terms that are already defined in subpart A of the Final MRR. We are proposing to use DRE because it is the term generally used by the electronics manufacturing industry. Furthermore, in addition to capturing the destruction of materials in the exhaust, the term also captures materials in the exhaust that are recycled or captured for reuse.

For purposes of this reporting rule, we propose that facilities that wish to document and report fluorinated GHG and N

O emissions reflecting the use of abatement systems adhere to a method that would require: (1) Documentation to certify that the abatement system is installed, operated, and maintained in accordance with manufacturers' specifications, (2) accounting for the system's uptime,

and (3) either certification that the abatement system is specifically designed for fluorinated GHG and N

O abatement and the use of an EPA default DRE value, or direct, proper DRE measurement to confirm the performance of the abatement system. Proper DRE measurement means measured in accordance with EPA's Protocol for Measuring Destruction or Removal Efficiency of Fluorinated Greenhouse Gas Abatement Equipment in Electronics Manufacturing (EPA's DRE Protocol). EPA's DRE Protocol is available for review in the docket (EPA-HQ-OAR-2009-0927). Our proposed approach is depicted as a decision tree in Figure 1 of this preamble.

Uptime means the total time during the reporting year when the abatement system for which controlled emissions will be reported was properly installed, operated, and maintained.

The proposed approach requires annual certification to ensure that abatement systems for which controlled emissions are reported are installed, operating, and maintained according to manufacturers' specifications. Our approach would also require that any DRE used in reporting emissions be based on an EPA default DRE value or on recent on-site measurements and actual uptime of the system, accounting for system redundancy. When process tools are equipped with multiple abatement systems designed for fluorinated GHGs and N

O, the facility may account for the combined uptime for the specific calculation of controlled emissions. Each one of these components is discussed in detail in the paragraphs below. We anticipate this method for reporting controlled emissions will ensure that abatement systems have been properly installed, operated and maintained during each reporting period and that best available measured DRE values are used to estimate and report emissions.

BILLING CODE 6560-50-P

EP12AP10.037

BILLING CODE 6560-50-C

Proper Installation, Operation, and Maintenance.

We are proposing that all facilities that use abatement systems and would like to reflect these emissions reductions in their annual emissions estimations be required to document and certify the abatement equipment's proper installation, operation, and maintenance. There are many manufacturers, and for each manufacturer multiple models, that are marketed as fluorinated GHG-destruction capable (Beu, 2005). While some abatement systems may be capable of destroying some fluorinated GHGs, they may not be effective in abating CF

(Beu, 2005), which in some processes can constitute 10 percent—20 percent (by volume) of fluorinated GHG exhaust composition (EPA, 2006). It appears that this variability may be partially attributable to installation as well as operating and maintenance practices although variations in how destruction is measured may also contribute to this variability (Beu, 2005). Evidence indicates abatement devices must be properly installed to ensure achievement of the manufacturer's design goals. For this reason, we propose devices be installed in

accordance with manufacturers' specifications.

In terms of operation and maintenance, we also propose to require that abatement systems be operated and maintained in accordance with the manufacturers' specifications. It is well known across the industry that abatement system performance varies greatly depending on a variety of abatement device and process parameters such as temperature, flow and exhaust composition (Beu, 2005, EPA 2006, 2007)). Our proposed requirement that abatement systems be operated and maintained in accordance with manufacturers' specifications is intended to ensure best performance.

We understand that many times a facility may have an independent quality assurance expert certify the installation, operation, and maintenance of abatement equipment. We are considering the inclusion in the final rule, a requirement for annual, on-site independent inspections of abatement system installation, operation, and maintenance, which could include a review of records and physical inspection of installed equipment. We request comment on whether to require an independent quality assurance audit/inspection for abatement system installation, operation, and maintenance.

Accounting for Abatement System Uptime.

We are proposing that facilities account for abatement systems' uptime to report controlled emissions. Uptime is the total time during the reporting year when the abatement systems for which controlled emissions are being reporting was properly installed, operated, and maintained. Uptime is calculated as the sum of time during the reporting period that an abatement system is in a standby, productive, and engineering state as described in SEMI Standard E10-0304, Specification for Definition and Measurement of Equipment Reliability, Availability, and Maintainability (2004). Abatement system uptime is expressed as the sum of an abatement system's operational productive, standby, and engineering times divided by the total operations time of its associated manufacturing tool. For example, the time during which a system is in by-pass mode, undergoing maintenance, or not operating with O

-flow (in the case of a CF

combustion system) is not included in uptime. An exception to this is time during which exhaust flows are passed through a redundant abatement system that is in the same abatement system class (discussed below) as the primary abatement system. Such time may be included in the uptime of the primary system.

We are proposing this requirement because we anticipate accounting for uptime (

i.e.,

tracking incidents when abatement systems may be “bypassed” or otherwise not in service) will produce a more accurate emissions estimate. We request comment on our proposal to account for and report the uptime of abatement systems. We also request detailed information on how uptime may be monitored and calculated.

EPA Default DRE Value.

In addition to certifying that an abatement system is installed, operated, and maintained according to manufacturers' specifications, and accounting for the system's uptime, the first approach we are proposing includes the following two key elements: (1) Certification that the abatement system is specifically designed for fluorinated GHG and N

O abatement, and (2) an EPA default DRE value. By applying the EPA default DRE value, the facility is not required to measure the DRE of their abatement system(s). We are proposing the use of a default DRE value of 60 percent if the facility certifies that the abatement systems for which this value is applied are specifically designed for fluorinated GHG and N

O abatement.

To develop the default DRE of 60 percent, we reviewed the individual DREs measured under our in-fab DRE measurement program and selected those that constituted discrete values

for systems that had been properly installed, operated and maintained. Of the data from the DRE measurement program, those that met the stated criteria were values for CF

. We calculated the mean and the lower one sided tolerance interval of the (CF

) DRE data set. This yielded an understated, default DRE, reducing the likelihood that the DRE of any particular system will be either overestimated or greatly underestimated. For additional information on how the EPA default DRE was developed, please refer to the Electronics Manufacturing TSD.

Using data available from the in-fab DRE measurement program, we selected discrete numbers rather than the lower bound (

e.g.,

≥ 99%).

While we are now proposing the use of an EPA default DRE value, consistent with our initial proposal we are not planning to permit use of the 2006 IPCC default factors or the manufacturer's DRE values. We are not permitting their use because once installed, abatement equipment may fail to achieve the default or a supplier's claimed DRE. DRE performance claimed by equipment suppliers and upon which the 2006 IPCC default factors were based may have been incorrectly measured due to a failure to account for the effects of dilution (

e.g.,

can be off by as much as a factor of 20 to 50 and C

can be off by a factor of up to 10 [Burton, 2007].) This understanding is supported by industry assessments as presented in Beu, 2005.

We are permitting the use of our default DRE value because we estimate that it strikes an appropriate balance between being conservative and being representative where equipment is properly operated and maintained. Our default DRE value was calculated using data from measurements assured to properly account for the effects of dilution. In addition, the tested systems were properly installed, operated, and maintained.

We request comment on our proposed default DRE value, and additional data and supporting documentation on DREs from studies that have been conducted on properly installed, operated, and maintained abatement systems and consistent with EPA's DRE Protocol.

Proper Measurement of the Abatement DRE.

The second proposed approach for quantifying, documenting, and reporting controlled emissions from abatement systems, described below, would require proper measurement of the abatement system DRE in addition to documentation to certify that the abatement system is installed, operated, and maintained in accordance with manufacturers' specifications, and accounting for uptime.

Consistent with our initial proposal, this second proposed method permits facilities to account for destruction if the abatement system performance is measured and verified using EPA's DRE Protocol. To measure DRE, we propose requiring facilities to conduct annual sampling through a random sampling abatement system testing program (RSASTP), spanning all abatement classes using the methods outlined in EPA's DRE Protocol. “Class” refers to a category of abatement systems grouped by manufacturer model number(s) and by gas for which the system is used to abate, including N

O and CF

direct and by-product formation, and all other fluorinated GHG gas direct and by-product formation.

“Classes” may also include any other abatement systems for which the reporting facility wishes to report controlled emissions provided that class is identified. For each class, the representative or average DRE

factors would then be applied to the yet unmeasured abatement devices of that class.

is a very stable chemical and especially difficult to effectively destroy. It may be used as an input gas and generated as a byproduct of other fluorinated GHG process reactions.

An annual representative sample as part of the RSASTP would consist of three or 20 percent of installed abatement systems, whichever is greater, for each class each year, measuring the DRE for a different three or 20 percent set of systems each year. Where 20 percent of total abatement systems do not equal a whole number, the number of systems to be tested would be rounded up to the nearest integer (

e.g.,

16 abatement devices, 20 percent of which equals 3.2; therefore, four abatement systems would be measured each year). Using the RSASTP and our rounding convention, all systems in each class would be tested within a five-year period. EPA is seeking comment on the required frequency of abatement system performance measurement.

When reporting controlled emissions from manufacturing, we propose that the facility either use the measured DRE or, in those instances where an individual abatement system has not yet undergone proper DRE testing, a simple average of the measured DREs for systems of that class would be used. If redundant abatement systems were used during periods of maintenance or repair, then we propose that the measured or average DRE for that system's class would be used. In any of these cases, the DRE used to report emissions would be adjusted to account for the actual uptime of the system. For example, if the uptime for a device is 98 percent over the reporting period, then the measured DRE (or class average of measured DREs when a system has not yet been measured) would be multiplied by 0.98.

Under the RSASTP, all systems in each class would be tested within a five-year period, after which the process would be repeated as long as controlled emissions were reported. There are two reasons for requiring the DRE to be measured for each abatement device over a time period and by specific class. Some fluorinated GHGs, particularly CF

, are harder to destroy than others; thus, the performance of abatement systems with one fluorinated GHG cannot necessarily be assumed to apply to other fluorinated GHGs.

Second, even if abatement systems rely on the same operating principle (

e.g.,

thermal oxidation) and are used on the same gases, their performance can vary depending on their operation and maintenance.

Moreover, maintenance that is adequate for abatement systems in some applications may not be adequate for abatement systems in others (

e.g.,

those that handle high volumes of etched or cleaned material, which can be deposited inside abatement equipment and clog lines). This argues for gradually testing all of the abatement systems within a class, and for retesting individual abatement systems over time.

There are many manufacturers, and for each manufacturer many models, that are marketed as fluorinated GHGs-destruction capable (Beu, 2005). While some abatement devices may be capable of destroying some fluorinated GHGs, they may not be effective in abating CF4 (Beu, 2005), which in some processes can constitute 10%-20% (by volume) of fluorinated GHGs exhaust composition (EPA, 2006).

Some variability in performance may be partially attributable to installation as well as operating and maintenance practices although variations in how destruction is measured may also contribute to this variability (Beu, 2005).

We request comment on the method proposed for proper measurement of DRE at a facility and the proposed RSASTP for abatement systems by class.

6. Selection of Procedures for Estimating Missing Data

In general, it is not expected that data to estimate emissions from electronics manufacturing would be missing; gas consumption data and indicators of activity data (

e.g.,

wafer passes) is collected as business as usual. For this reason, we are not proposing procedures for estimating missing data from emissions from cleaning, etching or deposition processes. Because our proposal includes an EPA default DRE value for estimating and reporting controlled emissions, we propose that no missing data procedures would apply.

When estimating heat transfer fluid emissions during electronics manufacture, the use of the mass-balance approach requires facilities to correct records for all inputs. Should the facility be missing records for a given input, heat transfer fluid emissions may be estimated using the arithmetic average of the emission rates for the year immediately preceding the period of missing data and the months immediately following the period of missing data. Alternatively it may be possible that the heat transfer fluid supplier has information in their records for the facility.

7. Selection of Data Reporting Requirements

We are proposing that owners and operators be required to report fluorinated GHG and N

O emissions for the facility for each electronics manufacturing process as well as all heat transfer fluid use. In addition, facilities would be required to report the following: method used to calculate emissions; factors used for gas utilization and by-product formation rates and the source for each factor for each fluorinated GHG and N

O; production in terms of substrate surface area (

e.g.,

silicon, PV-cell, LCD); for each fluorinated GHG and N

O, annual gas consumed during the reporting year and gas- and facility-specific heel factors used; the apportioning factors used, a description of the engineering model used for apportioning gas usage, and facility-wide consumption estimates based upon development of the apportioning factors, independent of the consumption value calculated using purchase records; fraction of each gas fed into each process type that is fed into tools with abatement systems; descriptions and information about abatement systems through which fluorinated GHGs and N

O flow; inputs in the mass-balance equation (for heat transfer fluid emissions); and example calculations. Where process categories defined in the Refined Method and/or default gas utilization and by-product formation rates are not used, we propose that facilities provide descriptions of individual processes or processes categories used to estimate emissions consistent with the IPCC Tier 3 method.

For each abatement system for which a facility is reporting controlled emissions, we propose that facilities be required to report the following: certification that the abatement device is installed, operated, and maintained according to manufacturers' specifications; the uptime and the calculations to determine uptime for that reporting year; the DRE used (

i.e.

either the EPA default DRE value or a properly measured DRE); and documentation for the EPA default DRE value or a properly measured DRE.

These data form the basis of the calculations and are needed for us to understand the reported emissions and verify their reasonableness.

8. Selection of Records That Must Be Retained

We propose that facilities keep records of data used to estimate emissions, records supporting values used to estimate emissions, purchase records, and invoices for gas purchases and sales. For those facilities that use facility-specific, recipe-specific gas utilization and by-production formation rates, we are proposing that the following records be maintained: documentation that the rates were measured using the 2006 ISMI Guidelines, documentation that the measurements made are representative of fluorinated GHG and N

O emitting

processes at the facility, and the date and results of the initial and any subsequent tests to determine process tool gas utilization and by-product formation rates.

For those facilities that are reporting controlled emissions, we propose that the following records be kept: documentation to certify that each abatement device used at the facility is installed, maintained, and operated in accordance with manufacturers' specifications; records of the uptime and the calculations to determine uptime; abatement system calibration and maintenance records; documentation for the EPA default DRE value or a properly measured DRE.

These records consist of values that are directly used to calculate the emissions that are reported and are necessary to enable verification that the GHG emissions monitoring and calculations are done correctly.

B. Fluorinated Gas Production

1. Overview of Reporting Requirements

Under this proposal, subpart L would require facilities that produce fluorinated gases to report their fluorinated GHG emissions from fluorinated gas production and transformation and from fluorinated GHG destruction. Fluorinated gases include fluorinated GHGs (HFCs, PFCs, SF

, NF

, HFEs,

etc.

), CFCs, and HCFCs. Certain emissions subject to other subparts or authorities are excluded from this subpart. Specifically, emissions of HFC-23 from HCFC-22 production are addressed under subpart O and are therefore excluded from this subpart. Similarly, as discussed in the Final MRR, emissions of ozone depleting substances (

e.g.,

CFCs and HCFCs) are subject to Title VI of the CAA and are therefore excluded from this subpart.

Under this proposed rule, facilities would be required to estimate their emissions from fluorinated GHG production processes using either a mass-balance approach or an approach based on measured (or in some cases, calculated) emission factors. Facilities would be required to estimate their emissions from CFC and HCFC production processes and from fluorinated gas transformation processes using an emission-factor-based approach. Consistent with the Final MRR, this proposal would establish an annual frequency for reporting and would include provisions to ensure the accuracy of emissions data through monitoring, reporting, and recordkeeping requirements. Reporting would be at the facility level.

2. Summary of Major Changes Since Initial Proposal

In the April 2009 proposed mandatory GHG reporting rule (74 FR 16448; April 10, 2009), the fluorinated GHG production source category was included as proposed subpart L. That initial proposal would have required reporting from facilities emitting more than 25,000 mtCO

e from fluorinated GHG production and other source categories (

e.g.,

stationary combustion). We proposed monitoring based on a daily mass-balance or yield approach that included measurements of the reactants and the fluorinated GHG product and byproducts. Under that approach, facilities would have had to calculate the difference between the expected production of each fluorinated GHG based on the consumption of reactants and the measured production of that fluorinated GHG, accounting for yield losses related to byproducts and wastes and accounting for streams that were recaptured and destroyed. Facilities would have been required to measure the various inputs and outputs daily using scales and flow meters with an accuracy and precision of 0.2 percent of full scale, and to measure concentrations in streams using methods with an accuracy and precision of 5 percent. (For more detailed information on the initial proposal,

see

the fluorinated gas production section of the April 10, 2009 proposed rule.)

We received numerous comments on the proposed approach. Commenters stated that there may be significant uncertainty associated with the mass-balance approach, that EPA's stated accuracy and precision requirement of 0.2 percent for flow meters and weigh equipment was costly and not technically achievable for many streams, that daily calculations were excessive and likely to introduce errors, that it was sometimes impracticable to perform a mass-balance for more than one reactant, and that the mass-balance approach was not appropriate for batch processes.

Commenters also suggested alternatives to the mass-balance approach. Several commenters focused on the use of site-specific or process-specific emission factors. These commenters noted that many facilities in this source category already measure emissions during performance testing to verify compliance with their emission limits under other EPA regulations. Commenters also noted that some fluorinated GHG producers currently estimate their emissions of fluorinated GHG using the emission factor approach and that this approach is both more cost effective and more accurate than the mass-balance approach. One commenter using the emission factor approach stated that the estimated uncertainty of its overall fluorinated GHG emissions estimate was 13 percent (expressed as one standard deviation) and that the uncertainty associated with the estimates that it would develop using the proposed mass-balance approach would be significantly higher. Commenters suggested both emissions testing and chemical engineering calculations as appropriate techniques to develop site-specific emissions factors.

Partly in response to the comments received on the April 2009 proposed MRR (74 FR 16448; April 10, 2009), today's proposed subpart L rule incorporates a number of changes compared to the original proposal, including but not limited to:

• Inclusion of additional emission estimation methodologies, including process-specific, site-specific emission factors, which allow facilities to estimate emissions using methods that may already be in place;

• Revisions to the mass-balance approach, including provisions to allow monthly rather than daily monitoring; greater flexibility in the accuracy and precision of flowmeters, weigh scales, and concentration measurements (as long as the final estimate meets an overall accuracy and precision requirement); and the use of one rather than two reactants in the mass-balance equation;

• Inclusion of fluorinated GHGs emitted as a by-product of the production of CFCs and HCFCs; and

• Inclusion of fluorinated GHGs emitted as a feedstock or by-product of transformation processes that are not intended to produce any fluorinated gases (when those transformation processes are co-located with fluorinated gas production processes).

3. Definition of Source Category

This source category covers emissions of fluorinated GHGs that occur during the production of fluorinated gases, where fluorinated gases include fluorinated GHGs (HFCs, PFCs, SF

, NF

, and fluorinated ethers, among others), CFCs, and HCFCs (except HCFC-22).

It also covers emissions of

fluorinated GHGs from transformation and destruction processes that occur at fluorinated gas production facilities. EPA estimates that total emissions from this source category were 10.6 million metric tons of CO

e in 2006.

In the April 2009 proposal, EPA requested comment on whether emissions of fluorinated GHGs from CFC and HCFC production processes should be subject to the subpart L reporting requirements. While no public comments were received on this topic, EPA has determined that HFCs and PFCs are likely to be generated during the production of several CFCs and HCFCs, and that the quantities generated may be significant. According

to the 2006 IPCC Guidelines and fluorinated gas producers, production of CFCs and HCFCs can generate and emit fluorinated GHGs such as various HFCs and some PFCs. (These HFCs exclude HFC-23 generated during HCFC-22 production, which is already covered under Subpart O). These emissions are by-product emissions that occur due to the chemical similarities between HFCs, PFCs, HCFCs, and CFCs and the common use of halogen replacement chemistry to produce them. HFC-23 generated during HCFC-22 production is already covered under Subpart O.

Emissions from fluorinated gas production facilities can occur from vents, from leaks at flanges and connections in the production line, and from control devices (

e.g.,

thermal oxidizers). Undesired by-products may be deliberately vented, and some product (or reactant) may be vented at the same time due to imperfect separation of by-products, products, and reactants. Emissions can also occur during occasional service work on the production equipment, during blending and recycling of fluorinated GHGs, and during the evacuation and filling of tanks or other containers that are distributed by the producer (

e.g.,

on trucks and railcars).

Fluorinated GHG Emissions from Fluorinated GHG Production.

Emissions that occur during fluorinated GHG production include fluorinated GHG products that are emitted before the production measurement and fluorinated GHG byproducts that are generated and emitted either without or despite recapture or destruction.

These emissions are not counted as “mass produced” under the final requirements for suppliers of industrial GHGs in 40 CFR part 98, subpart OO (74 FR 56260; October 30, 2009).

Byproducts that are emitted or destroyed at the production facility are excluded from the Subpart OO definition of “produce a fluorinated GHG.” Any HFC-23 generated during the production of HCFC-22 is also excluded from this definition, even if the HFC-23 is recaptured. However, other fluorinated GHG byproducts that are recaptured for any reason are considered to be “produced.”

Fluorinated GHG emissions from U.S. facilities producing fluorinated GHGs are estimated to range from 0.8 percent to 2 percent of the amount of fluorinated GHG produced, depending on the facility. In 2006, 12 U.S. facilities produced over 350 million metric tons CO

e of HFCs, PFCs, SF

, and NF

, and an additional 6 facilities produced approximately 1 million metric tons CO

e of fluorinated anesthetics. Based on an emission rate of 1.5 percent, facilities are estimated to have emitted approximately 5.3 million metric tons CO

e of HFCs, PFCs, SF

, and NF

, and approximately 15,000 metric tons CO

e of fluorinated anesthetics.

Fluorinated GHG Emissions from CFC and HCFC Production.

Our proposal to include fluorinated GHG emissions that occur during CFC and HCFC production processes is based on two important considerations. First, while the quantity of by-product emissions is uncertain, we believe that it is significant and could be similar to total estimated emissions from fluorinated GHG production. Second, many CFC and HCFC production processes are co-located with fluorinated GHG production facilities, allowing for efficiencies in the application of estimation methods and monitoring and reporting infrastructures. These issues are discussed in more detail in the Fluorinated Gas Production Technical Support Document in the docket for this rulemaking (EPA-HQ-OAR-2009-0927).

Although we do not have precise estimates of the magnitude of fluorinated GHG emissions from production of CFCs and HCFCs, we estimate that if CFC and HCFC production processes emitted fluorinated GHGs equivalent to one percent of their CFC and HCFC production (excepting HCFC-22 production), U.S. emissions from this source would be 5.3 mtCO

e, the same as from fluorinated GHG production. EPA requests comment on the extent to which fluorinated GHGs are generated and emitted during CFC and HCFC production. EPA also requests comment on the extent to which fluorinated GHGs may be generated and emitted during production of other ozone-depleting substances such as methyl chloroform and carbon tetrachloride and on whether such emissions should be reported under this rule.

CFCs and HCFCs are often produced at the same facilities that produce fluorinated GHGs. In these cases, these facilities would need to quantify their fluorinated GHG emissions from a few processes in addition to those producing fluorinated GHGs. In other cases, CFCs or HCFCs are produced at facilities that do not produce fluorinated GHGs. In these cases, which EPA estimates include 2 facilities, the facilities would not have been covered by the initially proposed subpart L, but would be covered by today's proposal. This coverage is reflected in the threshold analysis discussed below.

Fluorinated GHG Emissions from Other Processes.

Facilities producing fluorinated gases would also be required to report emissions of fluorinated GHG feedstocks that occur during the transformation of these feedstocks into other fluorinated substances such as fluoropolymers, as well as emissions of fluorinated GHGs that occur during destruction of fluorinated GHGs that are removed from the supply of industrial gases.

The reasons for requiring reporting of fluorinated GHG emissions from transformation processes that are co-located with fluorinated gas production processes are similar to those for requiring reporting of fluorinated GHG emissions from CFC and HCFC production. First, although EPA does not have precise estimates of the magnitude of fluorinated GHG emissions from transformation processes, discussions with fluoropolymer producers indicate that these emissions do occur. Second, facilities could apply similar methods and monitoring approaches to estimate emissions from both fluorinated gas production and fluorinated gas transformation. The rationale for requiring reporting of emissions from the destruction of fluorinated GHGs that are removed from the supply of industrial gases is discussed below under

Relationship between emissions covered under subpart L and those covered under subpart OO.

EPA is also considering requiring reporting of fluorinated GHG emissions from two other types of processes. The first type includes processes (other than CFC and HCFC production processes) in which fluorinated GHGs are neither reactants nor products of the process but are nevertheless generated as by-products or intermediates. To the extent that such processes may generate or emit significant amounts of fluorinated GHGs, it may be appropriate to require reporting of those emissions. This would be particularly true if the processes were co-located with fluorinated GHG production processes, permitting effiencies in the application of estimation methods and reporting infrastructures. EPA requests comment on whether, how often, and where such processes occur (

i.e.,

at fluorinated gas production facilities or elsewhere). The second type of process includes fluorinated gas transformation processes that are not co-located with fluorinated gas production facilities. Again, it may be appropriate to require reporting of fluorinated GHG emissions from such processes if these emissions are significant. EPA requests comment on both of these options.

Relationship between emissions covered under subpart L and those covered under subpart OO.

Subpart L would require reporting from many of

the same facilities (fluorinated GHG producers) that are required to report under subpart OO, which contains the industrial gas supply reporting provisions of the final MRR. In general, subpart OO is intended to capture the quantities of fluorinated GHGs that are entering and leaving the U.S. supply of industrial gases,

while subpart L is intended to capture the quantities of fluorinated GHGs emitted at fluorinated gas production facilities.

Specifically, subpart OO tracks the quantities of fluorinated GHGs that are (1) produced, (2) transformed, (3) destroyed, (4) imported, and (5) exported.

There are several areas of possible overlap between the emissions that could be reported under this subpart and those reported under subpart OO. The areas of overlap all concern emissions that occur at the fluorinated GHG production facility after (downstream of) the fluorinated GHG production measurement. These include emissions from:

• Fluorinated GHG transformation processes (including polymerization),

• Destruction of fluorinated GHGs that are removed from the supply of industrial gases,

• Cylinder filling (if this occurs after the production measurement),

• Blending of fluorinated GHGs,

• Recycling or reclamation of fluorinated GHGs, and

• Evacuation of fluorinated GHG heels from returned cylinders.

The MRR is intended to inform a range of possible policies for reducing emissions of GHGs, including both upstream and downstream approaches. Under a policy that focused primarily on supply, the fluorinated GHGs added to and subtracted from the gas supply would be tracked, and only the on-site emissions that occurred before (upstream of) the fluorinated GHG production measurement would need to be covered for completeness. On-site emissions that occurred after the production measurement would be assumed to be captured by the production measurement. Under a policy that focused on actual emissions (

i.e.,

“downstream coverage”) rather than supply, on-site emissions that occurred both before and after the production measurement would need to be tracked.

Maintaining flexibility to adopt either upstream or downstream approaches argues for some counting under L of emissions that are counted upstream (as supply) under OO.

(

See

the October 30, 2009 Final MRR, 74 FR 56260, for more discussion of the rationale for including both upstream and downstream emissions under the rule.) As noted above, EPA is proposing to require reporting of fluorinated GHG emissions from transformation and destruction processes that are located at fluorinated gas production facilities. However, EPA is also considering requiring reporting of fluorinated GHG emissions from the other activities that occur at fluorinated GHG production facilities downstream of the production measurement. EPA requests comment on the magnitude of these other on-site emissions and on whether or not they should be required to be reported under subpart L.

In theory, it might be possible to track emissions from transformation and destruction simply using quantities reported under OO. However, this would require that (1) fluorinated GHGs that are produced only to be transformed or destroyed be tracked separately, (2) production, transformation, and destruction be measured to very good precision and accuracy (

e.g.,

0.2 percent), and (3) that no by-products be formed or emitted during these processes. If all of these conditions were met, emissions could be equated to the differences between production and transformation and production and destruction. In practice, however, it would be difficult to meet all of these conditions.

4. Selection of Reporting Threshold

Under today's proposed rule, owners and operators of fluorinated gas production facilities would be required to estimate and report GHG emissions if those emissions, including both combustion and fluorinated GHG emissions, would exceed 25,000 mtCO

e in the absence of control technology (

e.g.,

thermal oxidation).

Following the precedents set by other Clean Air Act regulations, EPA is using the term “uncontrolled” to describe such emissions. Specifically, EPA is proposing to define “uncontrolled fluorinated GHG emissions” as a gas stream containing fluorinated GHG which has exited the process (or process condenser, where applicable), but which has not yet been introduced into an air pollution control device to reduce the mass of fluorinated GHGs in the stream. The term does not imply that the emissions are never controlled, but is synonymous with “pre-control emissions.”

In developing the threshold, we considered multiple controlled and uncontrolled emissions thresholds, including 1,000, 10,000, 25,000, and 100,000 metric tons CO

e. For fluorinated GHG production processes (including fluorinated anesthetics production processes), uncontrolled (pre-control) emissions were estimated by multiplying a factor of 3 percent by the estimated production at each facility. For CFC and HCFC production processes (except for HCFC-22 production processes), uncontrolled emissions were estimated by multiplying a factor of 2 percent by the estimated production at each facility. Uncontrolled emissions are strongly influenced by by-product generation rates, which are known to vary between zero and several percent for fluorinated gas production processes; thus, these estimates are uncertain. Controlled emissions were assumed to be half of uncontrolled emissions at each facility. Because EPA has little information on combustion-related emissions at fluorinated gas production facilities, these emissions were not included in the analysis. The results of the analysis for production of HFCs, PFCs, SF

, NF

, CFCs, and HCFCs are shown in Tables 7 and 8 of this preamble.

Table 7—Threshold Analysis for Fluorinated GHG Emissions From Production of HFCs, PFCs, SF6, NF3, CFCs, and HCFCs

[Uncontrolled Emissions]

Threshold level

(metric tons CO

e/r)

Total national emissions

(metric tons CO

e )

Number of facilities

Emissions covered

Metric tons CO

Percent

Facilities covered

Number

Percent

1,000

10,600,000

100

10,000

10,600,000

100

25,000

10,600,000

100

100,000

10,600,000

100

Table 8—Threshold Analysis for Fluorinated GHG Emissions From Production of HFCs, PFCs, SF

, NF

, CFCs, and HCFCs

[Controlled Emissions]

Threshold level

(metric tons CO

e/r)

Total national emissions

(metric tons CO

e )

Number of facilities

Emissions covered

Metric tons CO

Percent

Facilities covered

Number

Percent

1,000

10,600,000

100

10,000

10,600,000

100

25,000

10,600,000

100

100,000

10,600,000

10,300,000

As can be seen from the tables, most HFC, PFC, SF

e , NF

, CFC, and HCFC production facilities would be covered by all the thresholds considered. Although we do not have facility-specific production information for producers of fluorinated anesthetics, we believe that few or none of these facilities are likely to have uncontrolled emissions above the proposed threshold.

EPA is proposing to use a threshold based on uncontrolled (pre-control) rather than controlled (post-control) emissions to ensure that facilities that generate significant quantities fluorinated GHGs fully characterize and quantify their emissions, even if they initially believe those emissions to be small. Discussions with fluorinated gas manufacturers indicate that occasionally, fluorinated GHG by-products may be generated and emitted from production processes unexpectedly. If these by-products are relatively difficult to destroy (

e.g.,

), facilities' post-control emissions may be significantly higher than expected.

The initial scoping test described in the next section is intended to identify the full range of fluorinated GHGs in potentially emitted streams. Applying the full methodologies on the basis of the initial scoping study will provide EPA and the facilities with critical information on the extent to which control technologies are actually reducing emissions and therefore on the actual emissions from the facility.

It is important to note that even if a threshold based on controlled emissions were adopted, failure to report as required when a source's actual emissions were above that threshold would be a violation of these regulations and the Clean Air Act. Lack of test data or other errors of omission do not excuse such violations as the Clean Air Act is a strict liability statute.

EPA is requesting comment on an alternative approach in which all fluorinated gas production facilities, regardless of their estimated pre-control emissions, would analyze their emissions using the initial scoping test discussed in the next section. This approach would ensure that facilities understood the identities, and therefore the GWPs, of the fluorinated GHGs potentially emitted. EPA requests comment on this option, as well as on the option of simply eliminating the threshold for fluorinated gas production facilities and making this an “all-in” category.

As is true for the source categories covered by the Final MRR, fluorinated GHG production facilities could cease reporting if their controlled (post-control) emissions were less than 25,000 mtCO

e per year for five consecutive years or less than 15,000 mtCO

e per year for three consecutive years. This approach may be appropriate if control technologies are effective and there is no evidence of unexpected uncontrolled emissions. However, EPA requests comment on an alternative “off-ramp” for this source category. Under this alternative approach, the 25,000 and 15,000 mtCO

e triggers would be based on the level of emissions that is estimated before accounting for the use of any control technology (

e.g.,

thermal oxidation). EPA is requesting comment on this approach because emissions can become quite large if the destruction device malfunctions, is not operated properly, or is not used for some other reason.

As noted above, EPA estimates that under this proposal, all HFC, PFC, SF

, and NF

production facilities would be covered, and few or no anesthetics producing facilities would be covered. However, it is possible that EPA has underestimated total pre-control emissions from anesthetics. In its threshold analysis for fluorinated GHG production, EPA has assumed that emissions have GWPs similar to those of the product produced. However, fluorinated anesthetics are hydrofluoroethers, and other HFE production processes of which EPA is aware generate by-products with higher GWPs than the product. EPA requests comment on this issue.

A full discussion of the threshold selection analysis is available in the revised Fluorinated Gas Production TSD. For specific information on costs, including unamortized first year capital expenditures, please refer to the Economic Impact Analysis (EIA) for this rulemaking.

5. Selection of Proposed Monitoring Methods

a. Summary of Proposed Monitoring Methods

We are proposing to allow facilities to use either a mass-balance approach or a site-specific, process-vent-specific emission factor (PSEF) approach to estimate their fluorinated GHG emissions from fluorinated GHG production. Facilities would be required to use the PSEF approach to estimate their fluorinated GHG emissions from CFC and HCFC production or from fluorinated gas transformation. The mass-balance approach is similar to that proposed in April, 2009, but has been modified in some details in response to comments. Facilities using either approach would be required to perform a one-time scoping test to identify the fluorinated GHGs in certain emitted streams and to verify the destruction efficiency (DE) of any destruction devices every five years. These approaches are discussed in more detail below.

b. Initial Scoping Test of Potentially Emitted Fluorinated GHGs

In today's action, we are proposing that facilities that produce fluorinated gases perform an initial scoping test (proposed 40 CFR part 98.124(a)). The purpose of the scoping test is to ensure that all of the fluorinated GHGs that occur in emitted streams are properly identified. EPA is concerned that without the test, facilities could mischaracterize the set of fluorinated GHGs that was emitted, leading to inaccurate emissions estimates. We are aware that in general, facilities will have already identified most if not all of the fluorinated GHGs occurring in emitted streams during process design and bench and pilot scale testing. However, as noted above, we are also aware of

situations in which producers have analyzed process or emissions streams and found fluorinated GHGs that they were not expecting. Such by-product fluorinated GHGs can have high GWPs, making their CO

-equivalent emissions significant.

Under this requirement, which would be one-time for any given process, facilities would be required to sample the vent(s) or stream(s) that, alone or together, would be expected to contain all the fluorinated GHG by-products of the process. Facilities would be required to use EPA Method 18 (GC/ECD, GC/MS), EPA Method 320 (FTIR), or ASTM D6348-03 (FTIR) to identify fluorinated GHGs that occur in concentrations above 0.1 percent in emitted streams.

For facilities using the mass-balance approach, the scoping test could be used to determine whether some emissions that are assumed to occur in the form of the product are actually occurring as by-products. For facilities using the process-vent-specific emission factor approach (PSEF), the test would identify by-products to measure in subsequent emissions testing to develop emission factors.

To avoid the need to survey a large number of processes with relatively small fluorinated GHG emissions, EPA is proposing to limit the scoping test requirement to processes that would emit more than one metric ton per year of fluorinated GHGs before the imposition of control technologies. We are proposing a limit in tons of fluorinated GHGs rather than in tons of CO

e because the identities, and therefore the GWPs, of some fluorinated GHG constituents of the stream may not be known. Acquiring this information is the purpose of the test. We developed the one-ton limit by starting with a limit of 10,000 mtCO

e for each process and making the reasonably conservative assumption that the unknown fluorinated GHG could have a GWP of 10,000. For purposes of estimating the mass of fluorinated GHG emitted from the process, facilities could use the same types of engineering calculations that they would use to determine whether process vent testing was required under the PSEF approach (described in more detail below). They could assume that the mass of carbon, fluorine, or another relevant element is emitted in the form of fluorinated GHGs that were previously identified in bench- or pilot-scale testing.

We are proposing that the one-metric-ton trigger be applied to emissions before rather than after control because some byproducts, particularly CF

, are very difficult to destroy. If these by-products occurred unexpectedly in a stream and if the trigger were applied to emissions after control, the facility would underestimate controlled emissions. Consequently, the facility could fail to undertake the scoping test when it was actually appropriate and could overlook the occurrence and emissions of the by-products.

We are proposing that facilities test the streams before the control device because emissions streams are often diluted during destruction processes (

e.g.,

due to fuel and air feeds), which would make it more difficult to detect and identify fluorinated GHGs that survived the destruction process. However, we request comment on this requirement as well as on the scoping test requirement as a whole.

For example, suppose that a facility believed that all of the fluorinated GHG by-products from a certain process consisted of HFCs, which its destruction device destroyed with a destruction efficiency of 99.9 percent, but that one of these by-products was actually CF

, which the destruction device destroyed with an efficiency of only 50 percent. In this case, the facility could underestimate its fluorinated GHG emissions by more than an order of magnitude, neither seeking nor finding the CF

that it was actually emitting.

c. Mass-Balance Approach

We are proposing that facilities producing fluorinated GHGs have the option of monitoring emissions using the mass-balance approach. In this approach, facilities would calculate the difference between the expected production of each fluorinated GHG based on the consumption of reactants and the measured production of that fluorinated GHG, accounting for yield losses related to byproducts (including intermediates permanently removed from the process) and wastes. Yield losses that could not be accounted for would be attributed to emissions of the fluorinated GHG product. This calculation could be performed for any fluorine- or carbon-containing reactant (

e.g.,

HF or hydrocarbon) to estimate emissions of the fluorinated GHG product for that reactant (

i.e.,

the mass balance may be based on a carbon balance or a fluorine balance). If fluorinated GHG byproducts were produced and were not completely recaptured or completely destroyed, facilities would also estimate emissions of each fluorinated GHG by-product.

Because the mass-balance approach assumes that losses from the process are emissions of the product, EPA believes that the mass-balance approach would only be appropriate for estimating emissions from fluorinated GHG production, not production of CFCs, HCFCs, or polymers. (In the last three situations, the product is not a fluorinated GHG.) However, EPA requests comment on this issue.

To be eligible to use the mass-balance approach, facilities would have to demonstrate that their planned measurements could meet a statistical error limit required in the rule (described below). If the facility could not demonstrate that it could meet the error limit, it would have to improve the accuracy and/or precision of its monitoring and measurement devices or opt to use another monitoring approach offered in the rule.

To carry out the mass-balance approach, the facility would choose a reactant for yield calculation purposes. The facility would then weigh or meter the mass of that reactant fed into the process, any primary fluorinated GHG produced by the process, the mass of the reactant permanently removed from the process (

i.e.,

sent to the thermal oxidizer or other equipment, not immediately recycled back into the process), any fluorinated GHG byproducts generated, and any streams that contain the product or fluorinated GHG byproducts and that are recaptured or destroyed. These measurements would be tracked monthly or more frequently and consolidated and recorded on a monthly basis. If monitored streams (including relevant process streams, emissions streams, and destroyed streams) included more than one component (product, byproducts, or other materials) in more than trace concentrations,

the facility would be required to monitor concentrations of products and byproducts in these streams. Finally, the facility would be required to perform monthly mass-balance calculations for each product produced.

EPA is proposing to define “trace concentration” as any concentration less than 0.1 percent by mass of the stream.

Statistical Error Estimate.

To estimate the statistical error associated with use of the mass-balance approach, facilities would be required to use error propagation, considering the accuracy and precision of their measurements and the calculation methods of the mass-balance approach. This approach is described in more detail in the TSD for this proposal. Under this approach, EPA would not specify precision and accuracy requirements for individual mass or concentration measurements. Instead, EPA would require that the error associated with the overall estimate of fluorinated GHG emissions fall under 30 percent (relative error) or under 3,000 mtCO

e (absolute error). (Both errors are expressed as halves of 95 percent confidence intervals; for normal distributions, this is quite close

to two standard deviations). Facilities could achieve this level of precision however they chose.

We are proposing to require the error estimate to ensure that the use of the mass-balance approach yields accurate emission estimates. As observed by several groups that commented on the initial proposal, the mass-balance approach can result in large errors if measurements of the flow of fluorinated GHGs in one or more streams have significant errors.

We recognize that the proposed approach requires facilities to calculate the overall error of their own estimates, which adds complication and introduces opportunities for mistakes. We therefore plan to develop a calculation tool that would permit reporters to develop an error estimate, reducing both their burden and the likelihood of errors.

The mass-balance approach works by subtracting the masses of process outputs from those of process inputs. As a result, errors that are a relatively small share of these masses become a large share of the difference between them. Errors are particularly a concern for streams where the fluorinated GHG is only one component of the total flow, and where, therefore, fluorinated GHG concentrations must be measured. In general, the accuracy and precision of concentration measurements is expected to be approximately +/−10 percent, although this can be as low as five percent and as high as 20 percent, depending on the circumstances. If this 10 percent error applies to a stream that constitutes a significant input or (more likely) output of the process, it can lead to an emissions estimate with a high relative error.

We are proposing a maximum relative error of 30 percent because this error is comparable to that cited by the facility that has used an emission factor approach to estimate its fluorinated GHG emissions.

It is also comparable to the error that EPA calculates for a facility with an emission rate of two percent and with good precisions and accuracies for its mass flow measurements (+/−0.2 percent) and for its concentration measurement (+/−10 percent) of a waste stream constituting five percent of the process's fluorinated GHG output flow.

A 13 percent error expressed as a standard deviation translates into a 26 percent error expressed as one half of a 95 percent confidence interval.

For facilities whose emissions constitute a very small share of their inputs and outputs (

e.g.,

one percent or less), a relative error of 30 percent will be very difficult to achieve using a mass-balance approach. At the same time, the absolute error of such a facility's estimate may be smaller than the absolute error of a facility that meets the relative error test but that has a higher emission rate. EPA is therefore proposing a maximum permissible absolute error of 3,000 mtCO

e for facilities whose estimates have relative errors greater than 30 percent. This absolute error is equivalent to 30 percent of the 10,000 mtCO

e threshold that is used elsewhere in the subpart to establish requirements for different sources (

e.g.,

process vents). Under this approach, processes whose emissions were lower than 10,000 mtCO

e could have relative errors higher than 30 percent so long as they met the limit on absolute error. This approach avoids penalizing processes and facilities with low emissions. EPA requests comment on the absolute error limit of 3,000 mtCO

e. EPA is also considering a higher limit,

e.g.,

5,000 mtCO

Another approach that would avoid penalizing facilities with low emission rates would be to express the maximum relative error as a fraction of the total mass of reactants fed into (or consumed by) the process. For a given process, this mass would remain relatively constant regardless of the emission rate. For the model facility described above, with errors of 0.2 percent in its mass flow measurements and of 10 percent in its concentration measurements, the error of the emissions estimate relative to the total mass of reactants is about 0.3 percent. One advantage of this approach compared to the absolute limit is that this approach limits the relative errors for processes with small throughputs, while the absolute limit could permit very large relative errors for processes with small throughputs. EPA requests comment on this approach.

In developing the approach to specifying maximum absolute and/or relative errors for the overall emissions estimate, we considered the alternative of specifying the maximum allowable errors (precisions and accuracies) of the individual measurements that feed into the mass-balance equation. This is the approach that EPA took in the initial proposal. This approach limits error, but it also limits flexibility, a concern raised by several commenters. Even a facility with a relatively large error in one stream may be able to bring the total error of its emissions estimate to a tolerable level by improving the accuracy and precision of other measurements that are used in the mass-balance equation, such as the mass flows of reactants and products. Neve

This text is long and has been trimmed here. Open the source document for the complete record.

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.