Greenhouse Gas Reporting Program: Proposed Amendments and Confidentiality Determinations for Subpart I

Federal RegisterOct 16, 2012

Ask Donna

What actually matters in this document.

Text

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Part 98

[EPA-HQ-OAR-2011-0028; FRL-9726-7]

RIN 2060-AR61

Greenhouse Gas Reporting Program: Proposed Amendments and Confidentiality Determinations for Subpart I

AGENCY:

Environmental Protection Agency (EPA).

ACTION:

Proposed rule; Grant of Reconsideration.

SUMMARY:

This action proposes amending the calculation and monitoring methodologies for the Electronics Manufacturing, of the Greenhouse Gas Reporting Rule. Proposed changes include revising certain calculation methods and adding a new method, amending data reporting requirements, and clarifying terms and definitions. This action also proposes confidentiality determinations for the reporting of the new and revised data elements. Many of these proposed actions are in response to a petition to reconsider specific aspects of our regulations. This document also proposes amendments to the General Provisions of the Greenhouse Gas Reporting Rule to reflect proposed changes to the reporting requirements for the Electronics Manufacturing sector.

DATES:

Comments.

Comments must be received on or before December 17, 2012.

Public Hearing.

The EPA does not plan to conduct a public hearing unless requested. To request a hearing, please contact the person listed in the

FOR FURTHER INFORMATION CONTACT

section by October 23, 2012. Upon such request, the EPA will hold the hearing on October 31, 2012 in the Washington, DC area starting at 9 a.m., local time. The EPA will provide further information about the hearing on its Web page if a hearing is requested.

ADDRESSES:

Submit your comments, identified by Docket ID No. EPA-HQ-OAR-2011-0028, by one of the following methods:

•

Federal eRulemaking Portal:

http://www.regulations.gov

. Follow the online instructions for submitting comments.

•

Email:

GHGReportingCBI@epa.gov

•

Fax:

(202) 566-1741.

•

Mail:

Environmental Protection Agency, EPA Docket Center (EPA/DC), Mailcode 6102T, Attention Docket ID No. EPA-HQ-OAR-2011-0028, 1200 Pennsylvania Avenue NW., Washington, DC 20460.

•

Hand Delivery:

EPA Docket Center, Public Reading Room, EPA West Building, Room 3334, 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-2011-0028. The 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 confidential business information (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 email. Send or deliver information identified as CBI to only the mail or hand/courier delivery address listed above, attention: Docket ID No. EPA-HQ-OAR-2011-0028. The

http://www.regulations.gov

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

http://www.regulations.gov,

your email 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, the 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 the EPA cannot read your comment due to technical difficulties and cannot contact you for clarification, the 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 the Air Docket, EPA/DC, EPA West, Room B102, 1301 Constitution Ave. NW., Washington, DC. 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 GENERAL 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; email address:

GHGReportingRule@epa.gov

. For technical information, contact the Greenhouse Gas Reporting Rule Hotline at:

http://www.epa.gov/climatechange/emissions/ghgrule_contactus.htm

Alternatively, contact Carole Cook at (202) 343-9263.

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, email address:

GHGReportingRule@epa.gov.

Worldwide Web (WWW).

In addition to being available in the docket, an electronic copy of this proposal, memoranda to the docket, and all other related information will also be available through the WWW on the EPA's Greenhouse Gas Reporting Rule Web site at

http://www.epa.gov/climatechange/

Acronyms and Abbreviations.

The following acronyms and abbreviations are used in this document.

BAMM best available monitoring methods

CAA Clean Air Act

e carbon dioxide equivalent

CBI confidential business information

CFR Code of Federal Regulations

CVD chemical vapor deposition

DRE destruction or removal efficiency

EIA Economic Impact Analysis

EPA U.S. Environmental Protection Agency

F-GHG fluorinated greenhouse gas

FDL field detection limit

FTIR Fourier transform infrared

GHG greenhouse gas

GWP global warming potential

HTF heat transfer fluid

ICR Information Collection Request

IPCC Intergovernmental Panel on Climate Change

ISBN International Standard Book Number

ISMI International SEMATECH Manufacturing Initiative

LCD liquid crystal display

MEMS micro-electro-mechanical systems

mtCO

e metric ton carbon dioxide equivalent

NAICS North American Industrial Classification System

O nitrous oxide

NTTAA National Technology Transfer and Advancement Act of 1995

OMB Office of Management & Budget

PFC perfluorocarbon

POU point of use

ppbv parts per billion by volume

QMS quadrupole mass spectroscopy

RFA Regulatory Flexibility Act

RSASTP random sampling abatement system testing program

RSD relative standard deviation

SEMATECH SEmiconductor MAnufacturing TECHnology

SIA Semiconductor Industry Association

UMRA Unfunded Mandates Reform Act of 1995

U.S. United States

VCS voluntary consensus standard

WWW Worldwide Web

Organization of This Document.

The following outline is provided to aid in locating information in this preamble.

I. General Information

A. What is the purpose of this action?

B. Does this action apply to me?

C. Legal Authority

D. What should I consider as I prepare my comments to the EPA?

II. Background for Proposed Amendments to GHG Monitoring and Calculation Methodologies and Other Technical Revisions

A. Background for Proposed Amendments

B. How would these amendments apply to 2012 and 2013 reports?

III. Summary and Rationale for Proposed Amendments to GHG Monitoring and Calculation Methodologies and Other Revisions

A. Summary of Proposed Rule Amendments in Response to Petition for Reconsideration

B. Rationale for Proposed Amendments

C. Proposed Rule Changes to Reporting and Recordkeeping Requirements

D. Proposed Changes to Remove BAMM Provisions and Language Specific to Reporting Years 2011, 2012, and 2013.

IV. Background for Confidentiality Determinations for Subpart I of Part 98

A. Overview and Background

B. Approach to Proposed CBI Determinations for New or Revised Subpart I Data Elements

C. Proposed Confidentiality Determinations for Individual Data Elements in Two Direct Emitter Data Categories

D. Request for Comments on Proposed Confidentiality Determinations

V. Statutory and Executive Order Reviews

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

B. Paperwork Reduction Act

C. Regulatory Flexibility Act (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. General Information

A. What is the purpose of this action?

The EPA is proposing amendments to the calculation and monitoring methodologies for Subpart I, Electronics Manufacturing, of the Greenhouse Gas Reporting Rule (“subpart I”). In addition, the EPA is proposing conforming changes to the reporting and recordkeeping requirements of subpart I. Changes include revising certain calculation methods and adding a new method, amending data reporting requirements, and clarifying terms and definitions. The EPA is proposing these amendments to (1) Modify calculation methods and data requirements to better reflect new industry data and current practice; (2) provide additional calculation methods to allow individual facilities to choose the method best suited for their operations; (3) reduce the burden associated with existing requirements; and (4) address sensitive business information concerns raised by members of the Semiconductor Industry Association (SIA). Amendments being proposed today affect all facilities that manufacture electronics including those that manufacture semiconductors (including light emitting diodes), micro-electro-mechanical systems (MEMS), liquid crystal displays (LCDs), or photovoltaic (PV) cells. Because we are planning an effective date of January 1, 2014 for the final amendments, we are also proposing to remove the rule language for certain provisions that will not apply after 2013. Sections II and III of this preamble contain more detailed information on the background and rationale for these proposed amendments. Many of the proposed changes are in response to a petition to reconsider specific aspects of subpart I.

The EPA is also proposing confidentiality determinations for the new and revised data elements under the proposed amendments to subpart I. Section IV of this preamble provides the background and rationale for these proposed confidentiality determinations. Finally, Section V of this preamble describes the statutory and executive order requirements applicable to this action.

B. Does this action apply to me?

This proposal affects entities that are required to submit annual greenhouse gas (GHG) reports under subpart I of 40 CFR part 98 (“Part 98”). 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 CAA section 307(d) apply to “such other actions as the Administrator may determine”). Part 98 and this action affect owners and operators of electronics manufacturing facilities. Affected categories and entities 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

Liquid crystal display unit screens manufacturing facilities.

334419

Micro-electro-mechanical systems manufacturing facilities.

Table 1 of this preamble lists the types of entities that potentially could be affected by the reporting requirements under the subpart covered by this proposal. However, this list is not intended to be exhaustive, but rather provides a guide for readers regarding facilities likely to be affected by this action. Other types of facilities 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 as well as 40 CFR part 98, subpart I. If you have questions regarding the applicability of this action to a particular facility, consult the person listed in the

FOR FURTHER INFORMATION CONTACT

section of this preamble.

C. Legal Authority

The EPA is proposing rule amendments to Part 98 under its existing CAA authority, specifically authorities provided in CAA section 114. As stated in the preamble to the 2009 final rule (74 FR 56260, October 30, 2009) and the Response to Comments on the Proposed Rule, Volume 9, Legal Issues, CAA section 114 provides the EPA broad authority to obtain the information in Part 98, including subpart I, because such data would inform and are relevant to the EPA's carrying out a wide variety of CAA provisions. As discussed in the preamble to the initial Part 98 proposal (74 FR 16448, April 10, 2009), CAA section 114(a)(1) authorizes the Administrator to require emissions sources, persons subject to the CAA, manufacturers of control or process equipment, 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.

In addition, the EPA is proposing confidentiality determinations for proposed data elements in subpart I, under its authorities provided in sections 114, 301, and 307 of the CAA. As mentioned, CAA section 114 provides the EPA authority to obtain the information in Part 98, including those in subpart I. Section 114(c) requires that the EPA make publicly available information obtained under section 114 except for information (excluding emission data) that qualify for confidential treatment.

The Administrator has determined that this action (proposed amendments and confidentiality determinations) is subject to the provisions of section 307(d) of the CAA.

D. What should I consider as I prepare my comments to the EPA?

1. Submitting Comments That Contain CBI

Clearly mark the part or all of the information that you claim to be CBI. For CBI information in a disk or CD-ROM that you mail to the EPA, mark the outside of the disk or CD-ROM as CBI and then identify electronically within the disk or CD-ROM the specific information that is claimed as CBI. In addition to one complete version of the comment that includes information claimed as CBI, a copy of the comment that does not contain the information claimed as CBI must be submitted for inclusion in the public docket. Information marked as CBI will not be disclosed except in accordance with procedures set forth in 40 CFR part 2.

Do not submit information that you consider to be CBI or otherwise protected through

http://www.regulations.gov

or email. Send or deliver information identified as CBI to only the mail or hand/courier delivery address listed above, attention: Docket ID No. EPA-HQ-OAR-2011-0028.

If you have any questions about CBI or the procedures for claiming CBI, please consult the person identified in the

FOR FURTHER INFORMATION CONTACT

section.

2. Tips for Preparing Your Comments

When submitting comments, remember to:

Identify the rulemaking by docket number and other identifying information (e.g., subject heading,

Federal Register

date and page number).

Follow directions. The EPA may ask you to respond to specific questions or organize comments by referencing a CFR part or section number.

Explain why you agree or disagree, and suggest alternatives and substitute language for your requested changes.

Describe any assumptions and provide any technical information and/or data that you used.

If you estimate potential costs or burdens, explain how you arrived at your estimate in sufficient detail to allow us to reproduce your estimate.

Provide specific examples to illustrate your concerns and suggest alternatives.

Explain your views as clearly as possible, avoiding the use of profanity or personal threats.

Make sure to submit your information and comments by the comment period deadline identified in the preceding section titled

DATES

. To ensure proper receipt by the EPA, be sure to identify the docket ID number assigned to this action in the subject line on the first page of your response. You may also provide the name, date, and

Federal Register

citation.

GHGReportingCBI@epa.gov

. You are also encouraged to send a separate copy of your CBI information to Carole Cook at the provided mailing address in the

FOR FURTHER INFORMATION CONTACT

section. Please do not send CBI to the electronic docket or by email.

II. Background for Proposed Amendments to GHG Monitoring and Calculation Methodologies and Other Technical Revisions

A. Background for Proposed Amendments

The GHG reporting requirements for subpart I were finalized on December 1, 2010 (75 FR 74774, hereafter referred to as “final subpart I rule”). Following the publication of the final subpart I rule in the

Federal Register

, the SIA (hereafter referred to as “the Petitioner”) submitted on January 31, 2011 an administrative petition titled “Petition for Reconsideration and Request for Stay Pending Reconsideration of Subpart I of the Final Rule for Mandatory Reporting of Greenhouse Gases” (hereafter referred to as the “Petition for Reconsideration”, available in docket EPA-HQ-OAR-2009-0927), requesting reconsideration of numerous provisions in the final subpart I rule. Since that petition was filed, the EPA has published five actions related to subpart I.

• Additional Sources of Fluorinated GHGs: Extension of Best Available Monitoring Provisions for Electronics Manufacturing (76 FR 36339, published June 22, 2011). Granted the Petition for Reconsideration with respect to the provisions for the use of Best Available Monitoring Methods (BAMM). Extended three of the deadlines in subpart I related to using the BAMM provisions from June 30, 2011 to September 30, 2011.

• Changes to Provisions for Electronics Manufacturing to Provide Flexibility (76 FR 59542, published September 27, 2011). Amended the calculation and monitoring provisions for the largest semiconductor manufacturing facilities to provide flexibility through the end of 2013 and extended two deadlines in the BAMM provisions.

• Proposed Confidentiality Determinations for Subpart I and Proposed Amendments to Subpart I Best Available Monitoring Methods Provisions (77 FR 10434, published February 22, 2012). Re-proposed confidentiality determinations for data elements in subpart I and proposed amendments to the provisions regarding

the calculation and reporting of emissions from facilities that use BAMM.

• Revisions to Heat Transfer Fluid Provisions (77 FR 10373, published February 22, 2012). Amended the definition of fluorinated heat transfer fluids (fluorinated HTFs) and the provisions to estimate and report emissions from fluorinated HTFs.

• Final Confidentiality Determinations for Nine Subparts and Amendments to Subpart A and I under the Mandatory Reporting of Greenhouse Gases Rule; Final Rule (77 FR 48072, published August 13, 2012). Final confidentiality determinations for data elements in subpart I and final amendments to the provisions regarding the calculation and reporting of emissions from facilities that use BAMM.

B. How would these amendments apply to 2012 and 2013 reports?

The EPA intends to address the comments on these proposed amendments and publish any final amendments in 2013. Facilities would be required to follow one of the new or revised methods to estimate emissions beginning in 2014. The first reports of emissions estimated using the new methods would be submitted in 2015. For the reports for reporting years 2012 and 2013, reporters would be expected to calculate emissions and other relevant data using the existing requirements under Part 98. These existing requirements include the flexibility for the largest semiconductor manufacturing facilities added in the September 27, 2011 rule titled “Changes to Provisions for Electronics Manufacturing to Provide Flexibility.”

Given the timing and extent of the proposed changes, and the likelihood that the final rule will not be published until the second half of 2013, we have determined that it is not feasible for sources to implement these changes for reporting year 2013. The proposed revisions would change and replace existing calculation methods and regulatory requirements, and would greatly affect how emissions are calculated and the data that would be reported. For example, we are proposing to add a new stack testing option to measure and calculate fab-level fluorinated greenhouse gas (F-GHG) emissions, revise process categories and associated gas utilization rates and by-product formation rates, and eliminate existing methods that require using recipe-specific gas utilization rates and by-product formation rates to calculate emissions. Because of the different data collection requirements compared to the current subpart I requirements, we do not anticipate that facilities would have enough time after the final rule is published to schedule stack tests, revise their current tracking and monitoring methods, or revise the data collection methods for reporting year 2013.

Thus, reporters using the current methods in subpart I would continue to use these methods for collecting data and calculating emissions for 2013 that are reported in 2014. Reporters would be required to select calculation methods based on any final revisions to the rule to calculate the emissions for 2014 that are reported in 2015.

III. Summary and Rationale for Proposed Amendments to GHG Monitoring and Calculation Methodologies and Other Revisions

A. Summary of Proposed Rule Amendments in Response to Petition for Reconsideration

In this action, we are granting reconsideration on all issues in the Petition for Reconsideration not already addressed in the final rules published June 22, 2011 (Additional Sources of Fluorinated GHGs: Extension of Best Available Monitoring Provisions for Electronics Manufacturing); September 27, 2011 (Changes to Provisions for Electronics Manufacturing to Provide Flexibility); and August 13, 2012 (Confidentiality Determinations for Subpart I and Amendments to Subpart I Best Available Monitoring Methods Provisions). Those final rules are described in Section II.A of this preamble. Section III.B of this preamble discusses the specific issues raised in the Petition for Reconsideration that are addressed in this action and the changes the EPA is proposing in response to the petition. The EPA intends to complete its response to the Petition for Reconsideration through this rulemaking.

Following consideration of the issues raised in the Petition for Reconsideration and data presented by the Petitioner, the EPA is proposing certain amendments to subpart I. Table 2 of this preamble presents a summary of the outstanding issues raised by the Petitioner and the corresponding proposed changes to the rule. Section III.B of this preamble provides further detail including the EPA's rationale for each proposed change

Table 2—Proposed Changes to the Rule Based on Petition for Reconsideration and the Petitioner's May 26, 2011 Letter Supporting the Development of the Rule Changes To Provide Flexibility That Were Finalized September 27, 2011

Technical issue

Proposed changes to rule

Rows 2 and 12 apply to semiconductor facilities only. All other rows apply to all electronics manufacturing facilities.

1. Addition of an emission estimation method as an alternative to recipe-specific emission factors. (See Section III.B.1)

Revising 40 CFR 98.93 to provide an option for using stack testing as an alternative method for determining fab-level emission factors for determining fab-level F-GHG emissions for all electronics manufacturing facilities.

Revising 40 CFR 98.94 to 98.98 to include the monitoring methods, QA/QC, missing data, reporting, recordkeeping, and definition requirements for the stack testing alternative.

2. Revision of default gas utilization rates and by-product formation rates for the plasma etch process type for semiconductor manufacturing. (See Section III.B.2)

Revise 40 CFR 98.92(a) and 40 CFR 98.93(a)(2) and (a)(4) to combine wafer cleaning and plasma etch emission processes and associated gas utilization rates and by-product formation rates. Revise Tables I-3 and I-4 for semiconductor manufacturing with new gas utilization rates and by-product formation rates based on gas type and process type or sub-type using additional data submitted by the Petitioner.

3. Removing recipe-specific emission factors: Requirements for (1) Largest semiconductor manufacturing facilities (defined as those facilities with annual manufacturing capacity of greater than 10,500 m

of substrate) to use recipe-specific gas utilization rates and by-product formation rates to estimate emissions from plasma etch processes; and (2) semiconductor facilities using wafers greater than 300 mm diameter to estimate all of their emissions from processes that use fluorinated GHGs using recipe-specific gas utilization rates and by-product formation rates. (See Section III.B.3)

Revising 40 CFR 98.93, 98.94, 98.96, and 98.97 to remove provisions to use recipe-specific gas utilization rates and by-product formation rates and to combine the wafer cleaning process type with the plasma etch process type. Under this proposal, all semiconductor manufacturing facilities, regardless of manufacturing capacity, would have the option to use default gas utilization rates and by-product formation rates to estimate emissions from the plasma etching/wafer cleaning process type and from the following three subtypes of the chamber cleaning process type: in-situ plasma chamber cleaning, remote plasma chamber cleaning, and in-situ thermal chamber cleaning.

4. Calculation for determining manufacturing capacity. (See Section III.B.4)

Revising the terminology and definition of

maximum designed substrate starts

in 40 CFR 98.98 to be

maximum substrate starts,

meaning for the purposes of Equation I-5 in subpart I, the maximum quantity of substrates, expressed as surface area, that could be started each month in a reporting year based on the equipment installed in that facility and assuming that the equipment were fully utilized. Manufacturing equipment would be considered installed when it is on the manufacturing floor and connected to the required utilities.

5. Reporting provisions for facilities that have integrated production and research and development (R&D) activities. (See Section III.B.5)

Facilities would be allowed to report integrated production and R&D emissions and, if doing so, would be required to provide an estimate of the fraction of total emissions from their R&D activities under 40 CFR 98.96.

6. Requirements for the accuracy and precision of the equipment measuring gas consumption. (See Section III.B.6)

Removing the requirement for one percent of full-scale accuracy for “all flow meters, weigh scales, pressure gauges and thermometers* * *” in 40 CFR 98.93(i) and referencing the calibration accuracy requirements in 40 CFR 98.3(i) for all measurement devices used to measure quantities that are monitored in subpart I.

7. Provisions for re-calculating the facility-wide gas specific heel factor and handling exceptional circumstances. (See Section III.B.7)

Revising the criteria for an “exceptional circumstance” in 40 CFR 98.94(b)(4) from 20 percent of the original trigger point for change out to 50 percent for small cylinders (containing less than 9.08 kilograms (20 pounds) of gas). For large containers, the “exceptional circumstance” would remain as a change out point that differs by 20 percent of the trigger point used to calculate the gas specific heel factor. Clarifying the requirements for recalculating the facility-wide heel factor.

8. Requirements for verifying the model used to apportion gas consumption. (See Section III.B.8)

Revising 40 CFR 98.94(c) to allow for development of apportioning factors by using direct measurements using gas flow meters or weigh scales, to measure process sub-type, process type, stack system, or fab-specific input gas consumption.

Revising 40 CFR 98.94(c)(2)(i) to allow reporters to select a period of the reporting year and its duration that is representative of normal operations for the model verification. The representative period would be at least 30 days in duration, and may be as long as one year. The model would be verified using the F-GHG used in the greatest quantity, and would be corrected if it does not meet the verification requirements. A facility would be able to use two F-GHG for model verification if they both meet the criteria and if at least one of them is used in the greatest quantity.

Increasing the maximum allowed difference between the modeled and actual gas consumption in the verification process from 5 percent to 20 percent.

9. Provisions for calculating N

O emissions. (See Section III.B.9)

Revising 40 CFR 98.93(b), 40 CFR 98.96(c)(3) and 40 CFR 98.96(k) to clarify that facilities must calculate annual fab-level N

O emissions from the chemical vapor deposition (CVD) process type and from the aggregate of other electronics manufacturing production processes using default emission factors (facilities are not required to report emissions from each CVD process and from each other N

O using process).

10. Provisions for reporting controlled emissions from abatement systems. (See Section III.B.10)

Revising 40 CFR 98.94(f) to allow facilities to use either revised default destruction or removal efficiency (DRE) values or to establish a site-specific DRE value for each combination of input gas or by-product gas and process type or sub-type using directly measured DREs. Providing alternative methods for a facility to directly measure DRE.

11. Provisions for determining and calculating abatement system uptime. (See Section III.B.11)

Revising Equation I-15 to allow reporters to calculate the average uptime for the group of systems for each combination of input gas or by-product gas and process type or sub-type, using the same process categories in which F-GHG use and emissions are calculated. Abatement system uptime monitoring and calculation would be simplified by assuming that connected process tools operate with F-GHGs or N

O flowing continuously once they are installed; this would apply for all methods (both default emission factors and stack testing).

12. Absence of a method for updating gas utilization rates and by-product formation rates and DRE values for semiconductor manufacturing. (See Section III.B.12)

Revising the data reporting requirements in 40 CFR 98.96 to require certain semiconductor manufacturing facilities to provide a report to the EPA every 3 years covering technology changes at the facility that may affect gas utilization rates and by-product formation rates or DRE values.

The EPA is not staying subpart I pending reconsideration as requested in the Petition for Reconsideration because the EPA believes that the concerns prompting the stay request have been addressed through the BAMM process and through the September 27, 2011 final rule (Changes to Provisions for Electronics Manufacturing to Provide Flexibility), which amended the calculation and monitoring provisions for the largest semiconductor manufacturing facilities to provide flexibility through the end of 2013. As stated in the preamble to the September 27, 2011 final rule, the EPA intends to finalize revisions to subpart I in 2013 so that semiconductor manufacturing facilities can implement the revised subpart I beginning in 2014. The EPA is not reopening the entirety of subpart I for comment but is taking comment only on the remaining issues raised by the Petitioner, as listed in Table 2 of this preamble, and the proposed amendments described in Section III.B of this preamble, with the exception that we request comment on whether new data are available to update the default gas utilization rates and by-product formation rates for the facilities that manufacture MEMS, LCDs, or PV cells (see Section III.B.2 of this preamble), and whether new data are available on measured DRE values for abatement systems used at MEMS, LCD, or PV cell manufacturing facilities (see Section III.B.10 of this preamble).

In summary, the major changes we are proposing are to revise the calculation methods to provide all electronics manufacturing facilities the choice of two methods to calculate annual emissions and to remove the option for electronics manufacturing facilities to determine and use recipe-specific gas utilization rates and by-product formation rates. The proposed rule would provide the option for reporters to use either default gas utilization rates and by-product formation rates, which the EPA is proposing to revise for semiconductor manufacturing facilities to reflect new industry data provided to the EPA, or to conduct stack testing to establish site-specific emission factors for F-GHGs that would be used to calculate F-GHG emissions. The proposed amendments would ensure that the EPA receives accurate and current facility-specific data. The proposed amendments also include provisions for the periodic review of industry advances and changes that may impact the default gas utilization rates and by-product formation rates and default DRE values used to estimate emissions, to encourage the continued collection of data that represent current industry practices. Additionally, the proposed stack testing approach allows for estimation of emissions based on periodic direct measurements of stack emissions from facilities. These proposed amendments would allow the EPA to accurately characterize and analyze GHG emissions from facilities in the electronics manufacturing industry while reducing burden to the industry.

B. Rationale for Proposed Amendments

1. Stack Testing as an Alternative Emission Monitoring Method for Facilities that Manufacture Electronics

After subpart I was promulgated, the Petitioner expressed interest in developing a method to use stack testing to quantify F-GHG emissions from electronics manufacturing facilities as an alternative to the recipe-specific method in the final subpart I rule. Specifically, the Petitioner proposed an approach in which they would (1) develop emission factors by measuring emissions from their stacks over a certain period and dividing them by an activity metric (e.g., gas consumption) measured over the same period; and (2) estimate annual emissions by multiplying the emission factors by the appropriate annual activity. They noted that stack testing is already widely accepted in the industry and commonly used to quantify non-F-GHG emissions for compliance with other state and federal air programs. They also noted that in most facilities, a large number of tools using F-GHGs are exhausted through a relatively small number of stacks, and stack testing in such a situation could be at least as accurate as the other methods in the final subpart I rule, and could be more cost-effective for the facility depending on how often testing is conducted (see “Technical Support for the Stack Test Option for Estimating Fluorinated Greenhouse Gas Emissions from Electronics Manufacturing Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028).

The EPA recognizes that stack testing is an important tool that has historically been required for specified non-F-GHG pollutants to determine a facility's compliance with emission limits, capture or control efficiencies, or monitoring parameters established pursuant to certain provisions of the CAA. Stack testing performed and verified according to the procedures in validated EPA methods is considered a reliable method to quantify facility emissions as long as a robust and predictable relationship is found between emissions and the selected activity metric. Because stack testing is a direct measurement of facility emissions, it has the potential to provide a high-quality characterization

of the emissions from the electronics manufacturing industry. Electronics manufacturers are already using stack testing to comply with other air rules and operating permit requirements. For example, semiconductor manufacturers subject to 40 CFR part 63, subpart BBBBB, National Emission Standards for Hazardous Air Pollutants for Semiconductor Manufacturing, are already required to perform stack testing using EPA Method 320 at 40 CFR part 63, appendix A (hereafter “EPA Method 320”), among others, to comply with subpart BBBBB, although they are not required to use EPA Method 320 to quantify F-GHG emissions.

To determine whether stack testing might be appropriate to quantify F-GHG emissions from electronics manufacturing, EPA evaluated whether it demonstrates (1) The ability of a method and technology to accurately measure F-GHG emissions from electronics manufacturing facilities during the test; (2) the ability to accurately measure a corresponding activity metric during the test; and (3) the existence of a reasonably constant and predictable relationship between F-GHG emissions and the chosen activity metric. The first and third factors were particularly important given the relatively low concentrations of F-GHGs in exhaust streams at electronics facilities and the potential variability of emission factors over time at those facilities as the mix of products and processes changed over time.

The Petitioner provided data from stack testing and supporting data on F-GHG consumption and production to demonstrate that that stack testing can be used to estimate annual emissions. These data were provided to the EPA in support of the Petitioner's request in the petition for reconsideration to add a stack testing option to subpart I for semiconductor manufacturing. The data were collected using EPA Method 320, “Measurement Of Vapor Phase Organic And Inorganic Emissions By Extractive Fourier Transform Infrared (FTIR) Spectroscopy” (40 CFR part 63, appendix A), at three companies manufacturing a variety of semiconductor products on different sized wafers. The data provided to the EPA demonstrated that F-GHG emissions are a direct and reasonably constant function of F-GHG consumption over the test period. Moreover, data from multiple tests at two facilities showed that emission factors (kg gas emitted/kg gas consumed) did not vary widely in the absence of significant technology and abatement level changes, even though the mix of products at one of the facilities appeared likely to have changed during the months since the previous test. This indicates that emissions from one period at a facility, when converted to emission factors based on F-GHG consumption, can be used to determine emissions at the same facility over an extended period of time (i.e., one year, and longer under certain circumstances), and can be scaled to estimate annual F-GHG emissions.

The data provided by the Petitioner (see “Technical Support for the Stack Test Option for Estimating Fluorinated Greenhouse Gas Emissions from Electronics Manufacturing Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028) demonstrated that current FTIR methods, such as EPA Method 320, have sufficient sensitivity, when used in conjunction with detectors optimized to detect F-GHGs, to provide accurate measurements of F-GHG emissions. EPA Method 320 can be used to measure concentrations of the commonly emitted F-GHGs down to a few parts per billion by volume (ppbv), and the field detection limits for the same F-GHGs can be as low as 1 or 2 ppbv.

The same data provided by the Petitioner provided evidence that F-GHG consumption can be accurately measured or estimated over the proposed test period of 8 hours as long as varying temperatures, non-ideal gas behavior, and low drawdown rates are appropriately accounted for. (Methods for accounting for these are discussed in “

Stack testing requirements

” in Section III.B.1 of this preamble.) This ensures that gas consumption can be accurately determined, either directly for the test period or by interpolating from longer-term consumption data. Accurate gas consumption measurements ensure that gas consumption can be used with the stack emission measurements as the basis for emission factors to calculate annual emissions.

Finally, the data provided by the Petitioner demonstrated that emissions estimated from stack testing were in agreement with emissions for the same facilities estimated using other methods, such as the default gas utilization rates and by-product formation rate method in subpart I (see “Technical Support for the Stack Test Option for Estimating Fluorinated Greenhouse Gas Emissions from Electronics Manufacturing Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028).

The EPA is proposing to revise subpart I to include a stack testing option for estimating annual F-GHG emissions at 40 CFR 98.93(i). This option would apply to all electronic manufacturing facilities, including those making semiconductors, MEMS, LCDs, and PV cells. We are not proposing this option for estimating N

O emissions; a review of the stack test data provided to the EPA revealed inconsistent results for stack measurements of N

O emissions for which the cause could not be determined (see “Technical Support for the Stack Test Option for Estimating Fluorinated Greenhouse Gas Emissions from Electronics Manufacturing Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028). Therefore, we do not have sufficient data to show that stack testing is appropriate for development of N

O emission estimates. However, the rule already includes an option based on default emission factors for estimating N

O emissions (see 40 CFR 98.93(b))

(Proposed amendments to the provisions and emission factors for estimating N

O emissions are discussed in Section III.B.9 of this preamble.)

In this action, we are also proposing to allow all electronics manufacturing facilities to use separate methods (i.e., stack testing or default utilization and by-product formation rates) to estimate emissions from each fab within a single facility. Facilities would report GHG emissions on a fab basis. Many electronics manufacturing facilities are divided into separate fabs, which generally consist of separate buildings constructed at different times in which the processing tools are located. Most facilities have only one fab, but some facilities have two or more fabs. Each fab may be dedicated to a different product type, or may represent different generations of manufacturing technology because they were built at different times. In the semiconductor manufacturing industry, separate fabs may use different size wafers.

Because of differences among fabs (e.g., differences in the number of stacks), a reporter may wish to use different methods to estimate emissions from each fab. We are proposing to allow reporters to use different methods for separate fabs, but would also require that emissions be reported at the fab level. We are proposing to define a “fab” in 40 CFR 98.98 as “the portion of an electronics manufacturing facility located in a separate physical structure that began manufacturing on a certain date.”

Selection of Stack Systems for Testing.

The EPA recognizes that given the diversity of facility designs among electronics manufacturers, some facilities may have some stacks that account for only a small percent of total facility emissions. In order to avoid the burden of testing a large number of stacks, the proposed amendments

would not require that all stacks be tested. Instead, the reporter would develop a preliminary estimate of the annual emissions from each “stack system” in a fab and would not be required to test those stack systems that account for relatively small emissions. A stack system would be considered to be one or more stacks that are connected by a common header or manifold, through which a fluorinated GHG-containing gas stream originating from one or more fab processes is, or has the potential to be, released to the atmosphere. For purposes of subpart I, stack systems would not include emergency vents or bypass stacks through which emissions are not usually vented under typical operating conditions.

Under the proposed rule, the reporter would develop a preliminary estimate of F-GHG emissions from each stack system on a metric ton carbon dioxide equivalent (mtCO

e) basis using the gas consumption in the tools associated with the stack system and gas utilization rates and by-product formation rates in proposed Tables I-11 through I-15, and accounting for the DRE of the “point of use” (POU) abatement systems and the uptime (the fraction of time the system is operating within manufacturer's specifications) of the POU systems. The gas utilization rates and by-product formation rates in proposed Tables I-11 through I-15 are based on the 2006 Intergovernmental Panel on Climate Change (IPCC) Tier 2a factors.

The factors in proposed Tables I-11 and I-12 for semiconductor manufacturing facilities were updated from the 2006 IPCC factors based on additional data collected by the Petitioner (see “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028).

2006 IPCC Guidelines for National Greenhouse Gas Inventories, Prepared by the National Greenhouse Gas Inventories Programme, Eggleston H.S., Buendia L., Miwa K., Ngara T. and Tanabe K. (eds). Hayama, Kanagawa, Japan.

In the preliminary estimate, reporters would be required to use data from the previous reporting year for the DRE of abatement and the total uptime of all abatement systems in each stack system. The consumption of each F-GHG in each stack system would be estimated as the total gas consumption of that F-GHG times the ratio of the number of tools using that F-GHG that are feeding to that stack system to the total number of tools in the fab using that F-GHG. The reporter would convert the F-GHG emissions to CO

e using the global warming potential (GWP) values for F-GHG in Table A-1 of subpart A of Part 98. For F-GHG in Tables I-11 through I-15 for which Table A-1 of subpart A of Part 98 does not list a GWP value, reporters would use a default value of 2,000 for the GWP. Based on this preliminary estimate, the reporter would rank the F-GHG emitting stack systems at the facility from the lowest to highest emitting. The reporter would not have to test emissions from low-emitting stack systems, defined as those F-GHG emitting stack systems meeting all of the following three criteria:

(1) The sum of the F-GHG emissions from all combined stack systems in the fab that are not tested is less than 10,000 mtCO

e per year;

(2) Each of the stack systems that are not tested are within the fab's lowest F-GHG emitting stack systems that together emit 15 percent or less of total CO

e F-GHG emissions from the fab; and

(3) The F-GHG emissions from each of the stack systems that are not tested can be attributed to only one particular collection of process tools during the test (i.e., the stack cannot be used as a bypass from other tools that are normally vented through a stack system that does not meet these criteria).

For those low-emitting stack systems that are not tested, the reported F-GHG emissions would be the preliminary estimate made using the gas consumption and the gas utilization rates and by-product formation rates in proposed Tables I-11 through I-15 in subpart I, accounting for the DRE and uptime of the POU abatement systems. The default emission factors in proposed Tables I-11 through I-15 are simplified default emission factors based on just F-GHG species, and do not account for different rates by process type or sub-type. This approach minimizes reporting burden to industry because it does not require allocation of gas consumption between process types or sub-types (e.g., etch and chamber clean), as is required for the default emission factor based method. However, we recognize that there may be a need for facilities to reconfigure low-emitting stack systems following testing for production reasons. As a result, we are specifically requesting comment on how often such stack flow configuration changes occur. In addition, we are specifically requesting comment on whether reporters should be allowed to calculate emissions for low-emitting stack systems that are not tested using average fab-specific emission factors developed for the stack systems that are tested. We are specifically requesting comment on how such a provision would affect emission calculations from differences in gas and process types, and in DRE abatement system uptime between stack systems that are tested and stack systems that are not tested.

Stack testing requirements.

For those higher-emitting stack systems in each fab that are not exempt from measurement, the reporter would measure each F-GHG concentration (parts per million by volume, ppmv) and the total stack flow to determine the hourly mass flow rate (kg/hr) of each F-GHG emitted from each applicable stack system. If a stack system has more than one stack from a common header, the reporter would be required to measure F-GHG concentration and flow in each stack from that header because it is known from prior testing that F-GHG concentrations and flow rates are not consistent in such systems because of incomplete mixing. The reporter would use EPA Method 320 or another validated method to measure F-GHG concentration, and EPA Methods 1 through 4 at 40 CFR part 60, appendices A-1, A-2, and A-3 to measure other stack gas parameters needed to convert F-GHG concentration to mass emissions for the test period. Reporters would also be required to measure the fab-specific consumption of each F-GHG for the test period.

Reporters would be required to determine the F-GHGs expected to be emitted from the stack system, including by-product F-GHG, based on a facility analysis of all F-GHGs consumed or emitted in the previous reporting year, and all F-GHGs expected to be consumed or emitted in the current reporting year by process tools vented to the stack system. Documented results of the analysis would be kept as a record by the facility. The facility would not be required to test for all F-GHG consumed in the previous year if they are no longer being used, but only to consider the use of those F-GHG in the analysis of the F-GHG previously consumed or emitted and expected to be consumed or emitted. The reporter would also need to consider in the analysis the by-product gases that are included in Tables I-3 to I-7 that are applicable to the reporter's industry segment (semiconductors, PV, MEMS, or LCD). Based on this analysis, reporters would be required to measure emissions for all F-GHG used as input gases and any expected by-product F-GHG, except for any intermittent low-use F-GHG. Intermittent low-use F-GHGs would be defined as F-GHG that meet all of the following:

(1) The F-GHG is used by the fab but was not used on the day of the actual stack testing;

(2) The emissions of that F-GHG do not constitute more than 5 percent of the total annual F-GHG emissions from the fab on a CO

e basis; and

(3) The sum of all F-GHG that are considered intermittent low-use F-GHGs does not exceed 10,000 mtCO

e for that year.

We are proposing that reporters would specifically test for CF

and C

as by-product F-GHG from all stack systems that are subject to testing. These two F-GHG are commonly formed by-product gases in the electronics manufacturing industry from the plasma etch and chamber cleaning process types, and some may also be formed in the abatement systems.

We are also considering an option that would require testing for all F-GHGs that have been identified as by-products of any input gas in previous testing throughout the electronics industry. This set would include C

, C

, and CHF

in addition to CF

and C

. We are considering this option because the identities and quantities of by-products generated at a particular facility at a particular time can be difficult to predict, and the costs of testing for additional by-products are expected to be modest. In the one set of semiconductor facility stack tests that tested for the full range of potential by-products listed above, a perfluorocarbon (PFC) by-product was found, C

, which accounted for up to 40 percent of the GWP-weighted by-product emissions of the fab (and up to two percent of the total GWP-weighted emissions). If unexpected by-products occur in similar proportions at other facilities, failing to measure for them could lead to routine underestimates of emissions at those facilities. This option is discussed further in the memorandum “Technical Support for the Stack Test Option for Estimating Fluorinated Greenhouse Gas Emissions from Electronics Manufacturing Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028. We are specifically requesting comment on the option of requiring facilities to test for the six by-products listed above.

Reporters would calculate annual emissions of intermittent low-use F-GHGs using the gas consumption and the gas utilization rates and by-product formation rates in proposed Tables I-11 through I-15 in the rule, accounting for the DRE and uptime of the POU systems during the year for which emissions are being estimated.

The testing period would be 8 hours for each stack, with the option for a longer duration. The EPA understands that a 24-hour testing duration may be burdensome and may increase testing costs; however, reporters could elect to conduct longer testing to improve the accuracy of gas consumption and F-GHG concentration measurements for gases used in smaller quantities.

Reporters would not be required to measure all stacks simultaneously, but reporters would be required to certify there are no changes between tests in the stack flow configuration (i.e., the relationship between sets of process tools and any connected POU systems and their corresponding waste streams that are ultimately vented through the stack). Reporters would also be required to certify there are no changes in the centralized abatement systems; if any are present. The tests would have to be conducted during a period in which the fab is operating at a representative operating level and with the POU abatement systems connected to the stack being tested operating with at least 90 percent uptime during the 8-hour (or longer) period, or at no less than 90 percent of the average uptime measured during the previous reporting year. The representative operating level would be considered to be operating the fab, in terms of substrate starts for the period of testing, at no less than 50 percent of installed production capacity or no less than 70 percent of the average production rate for the reporting year, where production rate for the reporting year is represented in average monthly substrate starts. For the purposes of stack testing, the period for determining the representative operating level must be the 30-day period ending on the same date on which testing is concluded.

To convert the measured F-GHG emission rates into fab-specific emission factors, the reporter would measure the consumption of each F-GHG used in the tools associated with the stack systems being tested, excluding gas consumption allocated to tools venting to low-emitting stack systems that are not tested. Consumption could be measured using gas flow meters, weigh scales, or pressure measurements (corrected for temperature and non-ideal gas behavior). For gases with low volume consumption for which it is infeasible to measure consumption accurately over the 8-hour testing duration, short-term consumption could be estimated by using one or more of the following:

(1) Drawing from single gas containers in cases where gas is normally drawn from a series of containers supplying a manifold;

(2) Increasing the length of the test period to greater than 8 hours; or

(3) Calculating consumption from long-term consumption (e.g., monthly) that is pro-rated to the test duration.

F-GHGs not detected by Method 320.

The EPA is proposing that the concentrations of F-GHG in stacks systems be measured using EPA Method 320. This has been shown to be a valid method for measuring these target compounds, but it is expected that some F-GHG may occur in concentrations that are below the field detection limit (FDL), as defined in EPA Method 320. Therefore, we are proposing that the following procedures be followed to account for different scenarios in which a F-GHG is used, but not detected by Method 320 measurements:

• If a F-GHG is consumed during testing, but emissions are not detected, the reporter would use one-half of the FDL for the concentration of that F-GHG in calculations.

• If a F-GHG is consumed during testing and detected intermittently during the test run, the reporter would use the detected concentration for the value of that F-GHG when available and use one-half of the FDL for the value when the F-GHG is not detected.

• If a F-GHG is not consumed during testing but is detected intermittently as a by-product gas, the reporter would use the measured concentration when available and use one-half of the FDL for the value when the F-GHG is not detected.

• If a F-GHG is an expected by-product gas (e.g., CF

, C

, and other gases listed as by-products in Tables I-3, I-4, I-5, I-6, I-7, and proposed Tables I-11 to I-15) of the stack system tested and is not detected during the test run, use one-half of the FDL for the value of that F-GHG.

• If a F-GHG is not used, and is not an expected by-product of the stack system and is not detected, then assume zero emissions for that F-GHG for the tested stack system.

We are specifically requesting comment on the option of listing specific by-product gases as “expected” to be emitted even when they are not detected. Based on a review of the default emission factor tables listed above, CF

and C

are almost always generated as by-products (that is, they are generated by a wide range of process types and input gases), and CHF

is frequently generated. Other by-products appear to be generated less frequently. Thus, it may be appropriate to specify CF

and C

, and possibly also CHF

, as the set of by-products for which a value of one half of the FDL should be assumed in calculating emissions during the test. This approach would simplify the rule, provide certainty for purposes of implementation, and relieve facilities of the burden of determining

which by-products are “expected” to be emitted.

EPA Method 320 requires the specification of maximum FDLs because the FDLs achieved by a method and detector can have a significant impact on the quality of the measurements. For example, if the FDL for a F-GHG were so high that large emissions of that GHG were never detected, the uncertainty of the resulting emissions estimate (i.e., one-half the FDL), would be correspondingly high. The EPA is proposing maximum FDLs based on (1) review of the FDLs that have been achieved at three different semiconductor facilities, and (2) analysis of the magnitude of the emissions that would occur (in CO

e) at various possible maximum FDLs. The latter provides an indication of the uncertainty of emissions measurements using methods and detectors with those FDLs. The proposed maximum FDLs can be found in proposed Table I-10 of the regulatory text.

The EPA expects that the proposed treatment of these non-detect values using one-half of the FDL will avoid any potential under-counting of any F-GHGs that are expected to be in the emissions from a given process and F-GHG input gas combination. At the same time, the proposed treatment will provide a reasonable estimate of emissions of F-GHGs that occur in concentrations that are below the FDL. The EPA's analysis of testing data provided by the Petitioner has shown that emission measurements of gases known to be used and for which the concentration was below the FDL accounted for about 0.1 percent of F-GHG consumption and would account for about 0.1 percent of emissions on a CO

e basis if the concentration was assumed to be one-half of the FDL as outlined in this section (see “Technical Support for the Stack Test Option for Estimating Fluorinated Greenhouse Gas Emissions from Electronics Manufacturing Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028).

Alternative stack test methods.

To provide flexibility for facilities utilizing the stack test option, we are proposing that reporters may use an alternative stack test method to measure the concentration of F-GHG in each stack provided that the method is validated using EPA Method 301 of 40 CFR part 63, appendix A (hereafter “EPA Method 301”), and the EPA approves its use.

Under the proposed approval process in 40 CFR 98.94(k), the reporter would be required to notify the Administrator of the intent to use an alternative test method. The notification would need to include a test plan describing the alternative method and procedures, the range of test conditions over which the validation is intended to be applicable, and also an alternative means of calculating the fab-level F-GHG emissions if the Administrator denies the use of the results of the alternative method. The reporter would be required to validate the alternative method using EPA Method 301 and submit the results of the Method 301 validation process along with the notification of intention and a rationale for not using the specified method.

The Administrator would review and determine whether the validation of the proposed alternative method is adequate and issue an approval or disapproval of the alternative test plan within 120 days of the reporter submitting the notification and test plan. The reporter would be required to respond to any of the Administrator's questions on the test plan before obtaining approval and take into account the Administrator's comments on the test plan in conducting the test using the alternative method. The reporter would be required to respond to the Administrator's questions or request for additional information on the plan during the 120-day review period and the Administrator's questions or request for additional information would not extend that review period. Therefore, it would be the reporter's obligation to respond in a timely manner. If an alternative test plan were not approved, a reporter would need to begin the process to have an alternative test method approved starting with the notification of intent to use an alternative test method.

The reporter would report the results of stack testing using the alternative method and procedure specified in the approved test plan. The report would include all methods, calculations and data used to determine F-GHG emissions. The Administrator would review the results of the test using the alternative methods and procedure and then approve or deny the use of the results of the alternative test method and procedure no later than 120 days after they are submitted to the EPA. During this 120-day period, the reporter would be required to respond to any of the Administrator's questions on the test report before obtaining approval of the final test results using the alternative method. If the Administrator were to find reasonable grounds to dispute the results obtained by the alternative method, the Administrator could require the use of the method specified in subpart I instead of the alternative method.

Once the Administrator approved the use of the alternative method, that method could be used by any other facility for the same F-GHGs and types of stack systems, if the approved conditions apply to that facility. In granting approval, the Administrator would limit the range of test conditions and emission characteristics for which that approval is granted and under which the alternative method could be used without seeking further approval. The Administrator would specify those limitations, if any, in the approval of the alternative method.

Accounting for Abatement System Downtime.

To account for the effect of POU abatement system downtime in estimating emissions using the stack testing method, reporters would record the abatement system downtime in each fab during testing and for the entire reporting year. Using the downtime measured during testing, the reporters would correct the measured emission factors to assume no abatement system downtime (i.e., 100 percent abatement system uptime). The downtime measured over the entire reporting year would be used to calculate the excess F-GHG emissions that occur as a result of abatement system downtime events.

The reporter would measure the amount of POU abatement system downtime (in minutes) during the emission tests for any tools that are vented to the stacks being tested. For example, if five POU abatement systems are down for times of 10, 15, 25, 30, and 40 minutes during an 8-hour test, the total POU system downtime would be 120 minutes, or 5.0 percent of the total possible abatement system and tool operating time for the five tools (2,400 minutes). Using these data and the average DRE for the POU abatement systems, the emission factor measured during the testing would be adjusted to an emission factor representing POU abatement systems with 100 percent uptime (zero percent downtime).

The downtime measured over the year would be used to determine an uptime factor that would be an aggregate for all abatement systems in the fab, and calculated using proposed Equation I-23 in subpart I. Abatement system downtime would be considered any time during which the abatement system was not operating according to the manufacturer's specifications. The reporter would determine the sum of the downtime for all abatement systems during the year, and divide this sum by the sum of the possible annual operating time for each of the tools connected to those abatement systems in the fab to determine the downtime fraction. The downtime fraction would be the

decimal fraction of operating time that the abatement systems were not operating according to the manufacturer's specifications. The uptime fraction used in the emissions calculations would be equal to 1 minus the downtime fraction.

The total possible annual tool operating time would be calculated by assuming that tools that were installed for the whole of the year were operated for the entire year. The total possible tool operating time would be prorated to account for the days in which a tool was not installed; any partial day that a tool was installed would be treated as a full day of tool operation. For an abatement system with more than one connected tool, the tool operating time would be equivalent to a full year if at least one tool was installed at all times throughout the year. The reporter would also be able to account for time that tools are idle and no gas is flowing through the tools to the abatement system.

It is important to note that the proposed calculation of the uptime factor is different when a reporter would be using the proposed stack testing method than when the reporter would be using the default gas utilization rate and by-product formation rate method. In the proposed stack testing method, the uptime would not be determined for each gas and process type combination, as it would be under the proposed revisions to the default emission factor method. Instead, the uptime factor would be based on an aggregate for all tools in the fab for which the stack testing method is being used. This aggregate method is possible because the emissions measured at the stack already account for the fact that the emissions have been abated, and the uptime factor is only needed to account for the relatively small percent of time that the abatement systems are not operating and excess emissions need to be calculated. In contrast, the default gas utilization rates and by-product formation rates in the current rule and in the proposed amendments are for “unabated emissions” and the uptime factor needs to be determined for each gas and process type combination to determine the relatively large percent of emissions that have been abated.

To calculate an unabated emission factor during periods of downtime in the stack testing method, the reporter would divide the abated emission factor by (1-d

), where

is the average weighted fraction of F-GHG

destroyed or removed in the POU abatement system(s) in the fab. The factor

would be calculated using proposed Equation I-24 in subpart I, based on the gas consumption and destruction and removal efficiency (DRE) for the abatement system(s) for each gas and process type combination.

When calculating annual emissions, the reporter would continue to collect abatement system downtime data and calculate the fraction of abatement system uptime for the fab. Excess emissions from abatement system downtime events would be determined based on the actual amount of downtime as a percent of the total annual abatement system operating time for the reporting year. If a fab had 2.0 percent downtime for the year, then the unabated emission factor would be applied to 2.0 percent of the gas consumption for the year to calculate the excess emissions. The abated emission factor would be applied to the other 98 percent of gas consumption for the fab. The excess emissions and the abated emissions would be added together to determine the total annual emission from the fab.

Calculating an average fab-specific emission factor.

The reporter would calculate an average fab-specific emission factor using proposed Equation I-19 in subpart I for each input F-GHG and proposed Equation I-20 for each by-product F-GHG, based on the testing results (average kg/hr) and the F-GHG gas consumption (average kg/hr). The fab-specific emission factor for each input F-GHG and each F-GHG formed as a by-product would take into account the mass emission rate, the gas consumption, the abatement system uptime, and the F-GHG destroyed or removed from the abatement systems. The fab-specific emission factor for input gases would be in units of kg gas emitted per kg of the same gas consumed (kg/kg).

For gases generated as by-products, we are proposing that the fab-specific emission factor would be the mass of the by-product emitted divided by the summed masses of all the F-GHGs consumed, as presented in proposed Equation I-20. This equation would apply to those F-GHGs that are emitted only as by-products and not consumed as input gases.

The reporter would calculate annual emissions for each F-GHG by-product gas as the product of the fab-specific emission factor and the total annual amount of F-GHG consumed, corrected for any POU abatement system downtime as described in this section of the preamble.

In some cases, emissions of a particular F-GHG input gas may exceed consumption of that gas because the F-GHG is generated as a by-product of the other input gases. This is often the case for CF

. In these cases, we are proposing that the reporter use 1.0 as the input F-GHG emission factor and treat the remainder of that F-GHG's emissions as a by-product of the other input gases. The reporter would use Equation I-20 to calculate the emission factor for the by-product emissions. For example, if during the testing, the fab consumed 100 kg of an F-GHG, but the stack testing measured 300 kg of that gas, the reporter would assign 100 kg of that F-GHG as an input gas used in proposed Equation I-19, and 200 kg of that gas as a by-product gas used in proposed Equation I-20. In this instance, we are also proposing that the denominator in Equation I-20 would include the consumption of all other F-GHGs, with the exception of the F-GHG being included in the numerator. This treatment of the denominator reflects the fact that we are assuming that the F-GHG in the numerator is formed as a by-product from all other F-GHGs, while the emissions from the actual consumption of that F-GHG as an input are being accounted by proposed Equation I-19. For calculating emissions from an F-GHG with an input emission factor equal to 1.0 and with a by-product emission factor, the input F-GHG emissions would be assumed to equal consumption of that F-GHG, and the by-product emissions would be determined by multiplying the by-product emission factor by the sum of the consumption of all F-GHGs excluding the by-product F-GHG.

The advantage of this approach is that it reflects the physical mechanism through which emissions of an input gas exceed consumption of that gas. Because mass is conserved, the emissions of an input gas that are in excess of consumption of that gas must be attributable to the other input gases. These “excess” emissions are expected to vary with the facility's consumption of the other input gases rather than with the facility's consumption of the “excessively” emitted gas. Reflecting this in the by-product emission factor will lead to more accurate emission estimates and will help to prevent large swings in emission factors that could result when consumption of the “excessively” emitted gas varies from test to test. For example, this could help a facility to avoid a 20 percent or greater relative standard deviation in its CF

emission factor, which would otherwise prevent the facility from qualifying to skip testing for five years (see “

Testing frequency”

in Section III.B.1 of this preamble).

Note that the proposed approach includes a simplification that would in some cases affect the “extra” emissions

that are reassigned as by-products of other input gases. This simplification, and its potential impacts are discussed in more detail in the document entitled “Technical Support for the Stack Test Option for Estimating Fluorinated Greenhouse Gas Emissions from Electronics Manufacturing Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028. Although we expect that the effect of this simplification will generally be small, we are specifically requesting comment on the simplification.

We are also specifically seeking comment on the proposed treatment of F-GHGs whose emissions exceed consumption, and comment on which F-GHG should be included in the denominator of proposed Equation I-20 for calculating the emission factor for by-product F-GHG. The currently proposed equation includes all F-GHG used in the fab in the denominator for the calculation of all by-product F-GHGs, except when the emission factor for an input F-GHG exceeds 1.0. If the emission factor for a F-GHG exceeds 1.0, the emissions greater than 1.0 would be assumed to be by-product F-GHG instead of un-utilized input F-GHG. This proposed approach is based on the assumption that all F-GHG used as inputs could be contributors of fluoride (F) atoms that could be involved in the formation of F-GHG by-product gases, which are primarily carbon containing F-GHG, even if those input F-GHG do not contain carbon, such as SF

or NF

. An alternative approach on which the EPA is seeking comment is not to include in the denominator SF

, NF

, and other F-GHG that do not contain carbon (C) atoms, assuming that they are less involved in the formation of carbon containing by-product F-GHG than the F-GHG used as inputs that contain carbon.

Testing frequency.

Based on the potential for multiple process changes and numerous R&D activities that may affect emissions at an individual facility, as discussed in the Petition for Reconsideration, the EPA is proposing in 40 CFR 98.94(j)(5)(i) to require annual testing of each stack system and annual calculation of emission factors, excluding those low-emitting stack systems that are exempt from testing. However, to offer flexibility, the EPA is also proposing in 40 CFR 98.94(j)(5)(ii) to allow reduced testing frequency based on variability in measured emission factors. If the reporter meets criteria for low measured variability in emission factors calculated from the test results, then testing frequency could be reduced to every 5 years instead of annually. Under this option, a reporter would conduct a minimum of three emission tests for each non-exempt stack, with at least 2 months between the tests on a single stack system. All tests could be done in one year, or the reporter could use three annual tests for this analysis. If the relative standard deviation (RSD) of the emission factors calculated from each of the three tests, expressed as CO

e for all F-GHG combined, was less than or equal to 15 percent, and the RSD of the emission factors for each single F-GHG that individually accounts for 5 percent or more of CO

e emissions was less than 20 percent, the facility could use the averages of the three emission factors for each F-GHG for annual reporting for that year and the next 4 years without testing, unless conditions change that affect the emission factors and trigger retesting, as specified in proposed 40 CFR 98.94(j)(8) and described in this section of the preamble. If the variability between the three tests did not meet these criteria, then the facility would use the emission factors from the most recent testing for reporting for that year and continue the annual testing. Facilities could repeat the RSD analysis each year using the previous three sets of data. We anticipate that this provision will provide additional incentive for careful measurements of emissions and gas consumption during each stack test to maximize the repeatability of the results in subsequent tests.

In addition, previously completed tests that were performed and verified according to EPA Method 320 or an alternative method validated using EPA Method 301 could be applied towards the three tests required under this option, as long as all three tests were completed no earlier than the date 3 years before the date of publication of the final rule amendments and they meet the final rule requirements for stack testing, which are being proposed under 40 CFR 98.94(j). Allowing facilities to use prior completed tests would allow them to use data that were collected in support of developing this proposed stack testing option, and in support of developing the revised default gas utilization rates and by-product formation rates that are also being proposed in this action. The reporter would be required to conduct testing of each stack system, regardless of the results of the most recent stack tests, if certain changes take place in the reporter's annual consumption of F-GHGs or in the equipment and processes at the fab. Testing would need to be repeated to develop a new fab-specific emission factor if consumption of a specific input gas used during the emissions test changes by more than 10 percent of total annual gas consumption in CO

e, relative to gas consumption in CO

e for that gas during the year in which the most recent emissions test was conducted. For example, if use of a single gas goes from 25 percent of CO

e to more than 35 percent of CO

e, that would trigger the need for a new test. If there is a change in the reporter's use of an intermittent low-use F-GHG that was not used during the emissions test and not reflected in the fab-specific emission factor, such that it no longer meets the proposed definition of intermittent low-use F-GHG (see “

Stack testing requirements”

in Section III.B.1 of this preamble), the reporter would also be required to re-test using that gas. Additionally, if there is: (1) A decrease by more than 10 percent in the fraction of tools with abatement systems, compared to the fraction of tools with abatement systems during the most recent emissions test; (2) a change in the wafer or substrate size used by the fab since the most recent emissions test; or (3) a change in a stack system that formerly met the criteria for not being subject to testing such that it no longer meets those criteria, then the reporter would also be required to re-test.

Finally, if a reporter is using a F-GHG that was not used during the emissions test, the reporter would be required to conduct additional stack tests in that year during a period when that gas is being used to determine an emission factor for that gas. If a F-GHG is no longer used or is an intermittent low-use gas, re-testing would not be required, and F-GHG emissions would be calculated according to the process for intermittent low-use gases.

The EPA is specifically soliciting comment on other changes that may occur at a fab, including the adoption of specific new process technologies that should be included in the list of activities that would be expected to affect emissions to the point that those changes should require a fab to retest the stacks to develop new emission factors.

As stacks are re-tested, reporters would update the fab-specific emission factors with the new data from those stacks, replacing the data from the earlier testing of the same stack. The reporters would also be required to annually review the current data for determining which stacks were exempt from testing to ensure that the low-emitting stacks still qualify for exemption. If a stack no longer meets the criteria for exemption from testing as a low-emitting stack, it would need

to be tested and the fab-specific emission factor would need to be recalculated including those data. This provision would ensure that the fab-specific emission factors determined through testing are based on approximately 85 percent of the F-GHG consumed in the fab on a CO

e basis. Finally, if a requirement to re-test stacks were triggered, facilities would also be required to re-evaluate the RSD of the emission factors including the most recent test results and the previous two test results to see if they still complied with the provisions that allow them to skip testing. If they did not meet those provisions, they would have to resume annual testing for at least the next 3 years to complete a new RSD analysis. Even if they met those requirements, they still would be required to resume annual testing no later than the fifth year after the original RSD analysis that was performed before the retesting requirement was triggered.

We specifically request comment on the proposed option to allow less frequent emission testing (i.e., the 5-year testing exemption). Commenters are encouraged to supply rationale and any available data in support of submitted comments.

2. Revise the Default Gas Utilization Rates and By-Product Formation Rates for the Plasma Etch Process Category for Facilities That Manufacture Semiconductors

The EPA is proposing to amend the default plasma etch and chamber cleaning gas utilization rates and by-product formation rates and the requirements in 40 CFR 98.93(a)(2) for estimating F-GHG emissions from plasma etch processes at semiconductor manufacturing facilities. The EPA is not proposing to amend the default emission factors for other types of electronics manufacturing facilities. As discussed in this section of this preamble, the current provisions allow certain facilities the option to use default plasma etch and chamber cleaning rates based on wafer size, gas input, and process type/sub-type. The default emission factors are based on two different wafer size classes (one set of default emission factors for both 150 mm and 200 mm wafers combined, and a second set of default emission factors for 300 mm wafers) and five process types/sub-types (plasma etching; chamber cleaning including in situ plasma cleaning, remote plasma cleaning, in situ thermal cleaning; and wafer cleaning).

As discussed in this section of this preamble, following the promulgation of the final subpart I rule, the Petitioner submitted additional utilization and by-product formation data for various size wafers (200 mm and 300 mm) from semiconductor manufacturing facilities. The Petitioner requested that the EPA consider revising the default gas utilization rates and by-product formation rates based on gas input, process type, and wafer size. They also requested that the rule be revised to allow all semiconductor manufacturing facilities to use the revised default emission factors in lieu of requiring certain manufacturers to develop recipe-specific utilization rates and by-product formation rates (see “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID. No EPA-HQ-OAR-2011-0028).

The Petitioner, in documents submitted to the EPA after the Petition for Reconsideration, also questioned the EPA's establishment of separate default gas utilization rates and by-product formation rates for the wafer cleaning process type in the final subpart I rule. The Petitioner stated that the wafer cleaning process represents a very small fraction of overall semiconductor manufacturing GHG consumption and emissions. At 12 facilities analyzed by the Petitioner, wafer cleaning represented 1 percent or less of the gas used at each facility. The Petitioner also noted that wafer cleaning is basically the same process as the wafer plasma etch process (see “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028). Plasma etching is defined in 40 CFR 98.98 as “a process type that consists of any production process using fluorinated GHG reagents to selectively remove materials from a substrate during electronics manufacturing.” Wafer cleaning is defined in 40 CFR 98.98 as “a process type that consists of any production process using fluorinated GHG reagents to clean wafers at any step during production.” The Petitioner stated in documents submitted to the EPA that the tools specifically designated for wafer cleaning are using the same gases in plasma to remove materials as used in the tools designated for plasma etching. The Petitioner also noted that the gas utilization rates for wafer cleaning and plasma etching in subpart I are similar for the four gases most commonly used in both plasma etch and wafer cleaning (CF

, CH

, NF

, and SF

), especially for SF

and CF

. The Petitioner also provided additional data to support their recommendation to combine the wafer cleaning process type with the plasma etch process type (see “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028).

In response to the concerns raised in the Petition for Reconsideration about the recipe-specific measurements, the EPA is proposing to amend the default utilization and by-product formation rates for the semiconductor manufacturing industry. Based on the amendments in the September 27, 2011 final rule titled “Changes to Provisions for Electronics Manufacturing to Provide Flexibility,” the larger semiconductor facilities that manufacture wafers measuring 300 mm or less may use the default utilization and by-product formation rates currently in subpart I to estimate emissions, instead of the recipe-specific method that would have otherwise been required, only through December 31, 2013.

First, the EPA is proposing that all semiconductor manufacturing facilities, regardless of manufacturing capacity, would have the option to calculate F-GHG emissions from the plasma etching process type using the appropriate default gas utilization rates and by-product formation rates provided in Tables I-3 and I-4 of subpart I. We would no longer distinguish between “large” and “other” semiconductor manufacturing facilities based on the calculated annual manufacturing capacity. That distinction exists in the current subpart I because the EPA chose not to require the recipe-specific method for the “other” semiconductor manufacturing facilities. However, the calculation methods we are proposing in today's action would apply to all semiconductor manufacturing facilities. Under this proposal, no electronics manufacturing facility would have the option to determine and use recipe-specific gas utilization rates and by-product formation rates for the plasma etch process type, as described in Section III.B.3 of this preamble. The EPA is proposing to remove the distinction between large and other semiconductor facilities, such that all semiconductor manufacturing facilities could use the default gas utilization rates and by-product formation rates, independent of facility size. The EPA had required only the largest semiconductor manufacturing facilities to use the recipe-specific plasma etch method to ensure that smaller facilities

had a lower burden consistent with their lower expected F-GHG emissions. However, in proposing to remove the recipe-specific plasma etch method, the burden on the largest facilities would be reduced significantly and would eliminate the need to distinguish between “large” and “other” semiconductor manufacturing facilities.

Second, we are proposing to revise the default emission factors for the plasma etch process type in Tables I-3 and I-4 of subpart I. The proposed revised default emission factors are based on an expanded data set provided to the EPA by semiconductor manufacturing facilities after subpart I was originally promulgated in December 2010. The data were provided to the EPA in support of the Petitioner's request to develop alternatives to the recipe-specific method. The proposed revised plasma etch default emission factors are based on 976 data records (representing additional data submitted after December 1, 2010; see the EPA's analysis in “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028), whereas the plasma etch default emission factors in the final subpart I are based on 93 records. As in the final rule, the proposed plasma etch default emission factors were developed using data characterizing un-abated emissions for specific process equipment that follows a version of the International SEMATECH Manufacturing Initiative (ISMI) measurement guidelines. Because the set of tool manufacturers and processes included in the 976 data records is larger than that included in the 93 records, the proposed revised plasma etch default emission factors are expected to be more representative of the F-GHG emitting processes and tools than the default emission factors in the final subpart I rule promulgated in December 2010. However, please see the “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028, for more discussion of this issue and of the estimated uncertainty associated with the use of the default emission factor approach.

In developing the proposed revised default emission factors for the plasma etch process type in semiconductor manufacturing, the EPA considered alternatives that would reduce the burden compared to the recipe-specific approach in the current rule, while still providing F-GHG emission estimates with generally acceptable uncertainty.

The EPA considered including film type as a variable in the tables of default emission factors for the plasma etch process type, in addition to the input gas type and wafer size. However, based on the EPA and the Petitioner's analysis of the available data, the EPA determined that including film type would provide only a marginal improvement (about 4 percent) in the uncertainty of the emission estimates, but it would also introduce a potential for error because F-GHG consumption would need to be apportioned to plasma etch processes based on the film type being etched. The potential error introduced by apportioning F-GHG consumption by film type would offset the reduction in uncertainty by including the film type. In addition, including film type would also increase the burden associated with this approach because facilities would need to apportion gas consumption by film type. The EPA also considered establishing default emission factors for different sub-types of the plasma etch process type. However, based on an analysis of the available data, no difference in default emission factors could be accurately determined for any identifiable sub-type of the plasma etch process type. Based on these findings, the EPA concluded that including only input F-GHG type and wafer size in the default emission factors for the plasma etch process type would achieve the best balance between the burden and uncertainty in estimating F-GHG emissions from the plasma etch process type. (See “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028.)

The EPA performed an uncertainty analysis that found that, depending on the wafer size and gas usage patterns of the fab, the default emission factor approach would result in estimates with uncertainties between approximately 10 and 40 percent; see “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028.

The EPA also considered two averaging conventions in developing the proposed revised default by-product emission factors for etch process input F-GHG for multi-gas processes. The first convention used the simple arithmetic mean of all available by-product emission factor data where a non-zero measurement was recorded. This method averaged all available non-zero by-product emission factor data (by by-product) for each gas, wafer size, process type or sub-type combination. This approach is appropriate if zeros indicate that a by-product was not looked for during the test.

The second convention used the simple arithmetic mean of all available by-product emission factor data, but included the use of zeros when by-product emissions were not recorded. This method averaged all available by-product emissions factor data (by by-product) including records that did not indicate by-product emissions (zeros) for each gas, wafer size, process type or sub-type combination. This approach is appropriate if zeros indicate that a by-product was looked for during the test, but was not detected.

The EPA compared the resulting by-product emission factors from using both averaging conventions. The comparison showed that including versus not including the zeros for cases where no detected by-product was reported resulted, on average, in a 38 to 45 percent difference in the by-product emission factors (see “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities Under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028).

Because the EPA was not certain whether zeros indicate that particular by-products were not looked for or whether they were looked for but not detected, we are conservatively proposing by-product emission factors that do not include zeros. We specifically request comment on whether and to what extent zeros in the emission factor data indicate that a by-product was looked for, but not detected. We also specifically request comment on what the detection limits were for such by-products. To the extent that zeros represent instances where a by-product was looked for, but not detected, we recognize that not including zeros in the by-product emission factor development may result in overstating by-product emissions. Therefore, we are specifically requesting comment on the method for averaging the available by-product emission factor data to determine the default by-product emission factors.

Third, the EPA is proposing to revise the default by-product formation rates for the chamber cleaning process type/sub-types in Tables I-3 and I-4 of subpart I. In developing the proposed default utilization and by-product emission factors for etch processes, the EPA also reviewed emissions from chamber cleaning processes for completeness. The EPA did not receive new data to support revised default

utilization rates for the chamber cleaning process type/sub-types established in the final subpart I rule. However, the EPA evaluated the averaging conventions used to develop the proposed revised default by-product emission factors for etch processes for use in developing default by-product emission factors for the chamber cleaning process type/sub-types. Using data from the final subpart I rule, the EPA analyzed the emission estimates from chamber cleaning process type/sub-types using the two averaging conventions described in this section of this preamble. Again, for simplicity, we are proposing to not include zeros for the development of by-product emission factors. As with the proposed revised default etch emission factors, the averaging comparison showed that including versus not including the zeros for cases where no detected by-product was reported could result in overstating by-product emissions. Therefore, we are proposing to follow the same averaging convention for chamber cleaning process type/sub-types. The revised default by-product formation rates for the chamber cleaning process type/sub-types in Tables I-3 and I-4 of subpart I reflect the simple arithmetic mean of the available by-product emission factor data, without the use of zeros. As for the revised default etch emission factors, we are specifically seeking comment on the method for averaging the available by-product emission factor data to determine the default by-product emission factors for chamber cleaning process type/sub-types.

Finally, the EPA is proposing to combine the semiconductor wafer cleaning process type with the plasma etch process type; the amended rule would not have separate default emission factors for semiconductor wafer cleaning in the revised Table I-3 and I-4 of subpart I. The EPA has reviewed the available data (see “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028), and believes that it is appropriate to combine these process types. The same gases are used for plasma etch and wafer clean, with similar gas utilization rates and by-product formation rates, and the wafer clean process represents 1 percent or less of gas consumption at a typical facility. Furthermore, the burden associated with apportioning gas consumption to the various process types is expected to be reduced by combining the wafer cleaning and the plasma etch process types because some gases used for wafer cleaning are also used in etching processes.

For the chamber clean process type, we are not proposing any changes to the three chamber clean sub-types. Under the revised default emission factors, semiconductor manufacturing facilities would estimate emissions from chamber clean and plasma etch processes using the following four process types/sub-types: (1) Plasma etch/wafer cleaning process type; and (2) chamber cleaning process type, including (2a) in situ plasma chamber cleaning; (2b) remote plasma chamber cleaning; and (2c) in situ thermal chamber cleaning.

If gas utilization rates and by-product formation rates are not available for a gas/process combination in Tables I-3 or I-4 of subpart I, we are proposing that reporters would assume that the utilization and by-product formation rates are zero (i.e., assume that emissions of a gas equal consumption of that gas). This approach is consistent with the methodology in the current subpart I rule, except that we are proposing to remove the option for facilities to develop recipe-specific factors.

All other provisions related to the method using default gas utilization rates and by-product formation rates, such as the wafer size classes used for the default emission factors in Tables I-3 and I-4, would remain the same. The only exception would be that the default emission factors in Table I-4 that apply to 300 mm wafers would also apply to wafers greater than 300 mm (e.g., 450 mm wafers). As more data (i.e., utilization and by-product formation rates) become available for the semiconductor manufacturing industry in the future, the EPA would consider adding new default emission factors to Tables I-3 and I-4 for new gas and process type/sub-type combinations, including adding any new default emission factors specifically for semiconductor manufacturing facilities using wafers greater than 300 mm diameter (e.g., 450 mm wafers). However, for this proposal, facilities using wafers greater than 300 mm diameter would use the same default emission factors as those using 300 mm wafers. Section III.B.12 of this preamble describes the proposed process for updating default emission factors as more information is collected from the electronics manufacturing industry.

We request comment on whether new data are available for gas utilization and by-product formation rates for any of the process types or sub-types in the semiconductor manufacturing industry that could be used to further update the default emission factors for semiconductor manufacturing. Commenters are encouraged to submit available data with their comments using the “Electronics Manufacturing Data Request Sheet” (see Docket ID No. EPA-HQ-OAR-2011-0028). Commenters can fill out the “Electronics Manufacturing Data Request Sheet” and submit the data to Docket ID No. EPA-HQ-OAR-2011-0028 for consideration by the EPA on whether to update the proposed default emission factors for semiconductor manufacturing. If the EPA does update the proposed default emission factors using such new data, if approved by the EPA, for the final rule, it will do so using the same methodologies as described in the “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028). The EPA will use the same criteria for accepting new data that were used in accepting data as specified in that document.

The EPA has not developed any specific changes to the default gas utilization rates and by-product formation rates for MEMS, LCD, and PV in Tables I-5 (MEMS), I-6 (LCD), and I-7 (PV) of subpart I because we have not received any new utilization and by-product formation rate data. However, we request comment on whether new data are available to update the default emission factors for the facilities that manufacture MEMS, LCD, or PV cells; commenters are encouraged to submit available data and supporting information with their comments using the “Electronics Manufacturing Data Request Sheet” (see Docket ID No. EPA-HQ-OAR-2011-0028). Commenters can fill out the “Electronics Manufacturing Data Request Sheet” and submit the data to Docket ID No. EPA-HQ-OAR-2011-0028 for consideration by the EPA on whether to update the default emission factors for MEMS, LCD, or PV manufacturing. If the EPA does update the default emission factors using such new data, if approved by the EPA, it will do so using the same methodologies as described in the “Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028). The EPA will use the same criteria for accepting new data that were used in accepting data as specified in that document.

3. Removing the Provisions for Using Recipe-Specific Gas Utilization Rates and By-Product Formation Rates for Facilities That Manufacture Electronics

The EPA is proposing to remove the provisions to use recipe-specific gas utilization rates and by-product formation rates in 40 CFR 98.93(a)(2)(ii)(A), (a)(3), and (a)(4). Under 40 CFR 98.93(a)(2)(ii)(A) of the final subpart I rule, semiconductor manufacturing facilities with an annual manufacturing capacity greater than 10,500 square meters of substrate per year manufacturing wafers with a diameter of 300 mm or less were required to use recipe-specific gas utilization rates and by-product formation rates to estimate emissions for the plasma etch process. However, the September 27, 2011 final rule titled “Changes to Provisions for Electronics Manufacturing to Provide Flexibility” provided these facilities the option to use the default emission factors in lieu of recipe-specific rates for emissions estimated for the 2011, 2012, and 2013 reporting years. Under the current provisions (40 CFR 98.93(a)(3)), all electronics manufacturing facilities (including PV, MEMS, LCD, and semiconductor manufacturers) are given the option to estimate their F-GHG emissions using recipe-specific rates. Under 40 CFR 98.93(a)(4), semiconductor manufacturers are required to use recipe-specific rates for all F-GHG processes if manufacturing on wafers that are greater than 300 mm in diameter.

After subpart I was promulgated on December 1, 2010 (75 FR 75774), the Petitioner requested the EPA to reconsider and remove the requirement to develop and use recipe-specific gas utilization rates and by-product formation rates for certain semiconductor manufacturing processes and facilities. The Petitioner cited three primary concerns with using recipe-specific rates in place of other methods:

• The technical burden of determining rates for numerous recipes used at a facility, which could number in the hundreds.

• The technical and logistical burden of tracking gas consumption and other facility parameters on a recipe-specific basis to accurately implement recipe-specific rates.

• Recipe-specific information could be used to reverse engineer individual recipes and otherwise compromise trade secrets.

The Petitioner noted that the recipes used at a facility could number in the hundreds. In the Petition for Reconsideration, the Petitioner provided industry survey results for 19 facilities each having over 200 recipes, in which three facilities had over 500 recipes, and two facilities had greater than 800 recipes. For facilities with R&D activities, the Petitioner noted that the number of unique recipes could run “into the thousands.” The Petitioner explained in the petition that the EPA defined individual recipes in a way that presumed that each recipe has a “specific combination of gases” “used repeatedly” and “under specific conditions of reactor temperature, pressures, flow, radio frequency (RF) power and duration.” The Petitioner stated that a manufacturer may have many complex recipes that are comprised of upwards of 20 or more individual steps that could each meet the rule definition of “individual recipe,” and that manufacturing facilities may run hundreds to thousands of such recipes per year. Because of the nature of the fabrication process, for each step, a recipe could specify a varying “combination of gases” or a variety of distinct “specific conditions.” The petition stated that the EPA's definition of individual recipes could be interpreted to render each step in a complex recipe as a separate “individual recipe” that would need to be tracked and measured to determine recipe-specific utilization and by-product formation rates.

The Petitioner also stated that the EPA's definition of “similar recipes” could result in each step of a complex recipe to be considered an “individual recipe” under subpart I, due to changes in the chemicals used and the specific conditions for each step. Furthermore, as discussed in Section III.B.5 of this preamble, the Petitioner asserted that many facilities integrate research and development activities into their production lines, and research requires an iterative process and introduces hundreds of recipe variations that would need to be accounted for. The Petitioner stated in the Petition for Reconsideration that the equipment and personnel do not currently exist in most facilities to perform the measurements, testing, and data collection that would be required under subpart I to develop gas utilization rates and by-product formation rates for every recipe or each recipe step. Specifically, the Petitioner provided an industry analysis with the Petition for Reconsideration that stated that only 5 of 24 surveyed facilities had the available equipment, and only one facility had personnel with the expertise to perform the testing to quantify emissions from individual recipes.

The Petitioner further stated in the Petition for Reconsideration that tracking gas consumption and other facility parameters on a recipe-specific basis would present technical and logistical challenges to manufacturers. The Petitioner said that the infrastructure does not currently exist to perform the data collection and testing that would be required on a recipe-specific basis. The Petitioner stated in the petition that many facilities would need to make significant equipment expenditures in order to have the capability to measure and collect the gas consumption data at the recipe-specific level.

In the Petition for Reconsideration, the Petitioner also stated that it is difficult to estimate the quantities of gas used in individual production processes and steps, and it is currently not possible to measure actual consumption because the points at which gases are used (the individual tools) are widely distributed throughout a facility. Although each individual process chamber has a mass flow controller to control the actual flow of each gas introduced in the chamber, collecting this information would require software modifications and the implementation of data gathering capability on the level of each tool at the facility, and then managing the data collected for all tools across the facility. In subsequent information provided to the EPA, the Petitioner stated that apportioning gas consumption to these points on a recipe-specific basis would introduce significant degrees of error that could affect the uncertainty of estimated emissions.

In discussions with the EPA, the Petitioner also suggested that as an alternative to the recipe-specific approach, facilities may be able to estimate emissions using the allocation of F-GHG to specific process types, and an estimate of the overall DRE for those process types. However, because the Petitioner and EPA developed the other F-GHG estimation approaches being proposed today, this alternative method was not developed beyond an initial concept.

In 2010, the EPA's goal was to publish default utilization rates and by-product formation rates for the electronics manufacturing industry that would provide accurate facility-level F-GHG emissions data. This would avoid the need for facilities to determine these rates on a recipe-specific basis. At that time, however, the emission data available to the agency was very limited, particularly with regard to F-GHG emissions from the plasma etch process for the semiconductor industry. At the final rule stage, we decided that we still had insufficient data for estimating

plasma etch process emissions using default emission factors for the largest facilities. For that reason, we required the largest facilities to report their facility-specific plasma etch data using a recipe-specific approach. We intended to use these data to develop emission factors for incorporation into the rule at a later date. Subsequent to the publication of the final rule, the Petitioner provided a substantial amount of plasma etch data as described in this section of the preamble. We have used these data to develop improved emission factors for plasma etch processes. Thus, the recipe-specific approach is no longer a critical part of the rule. As described in Section III.B.12 of this preamble, we are also proposing a mechanism for gathering data from facilities on changes to their processes that may necessitate updates to the default emission factors. We anticipate this addition will ensure that the default emission factors continue to reflect facility emissions going forward.

It is the EPA's position that the recipe-specific requirements in 40 CFR 98.93(a)(2)(ii)(A), (a)(3), and (a)(4) are no longer necessary given the substantial amount of data submitted by the Petitioner following promulgation of subpart I, together with today's proposal to revise the default utilization and by-product formation rate method and introduce a stack testing method. Furthermore, the EPA believes the revised and alternative methods proposed today would provide reliable facility-specific data while avoiding in large part the potential concerns raised regarding the recipe-specific requirements with respect to technical difficulty, burden, and the protection of trade secret information. The EPA is proposing to remove the recipe-specific requirements and revise corresponding requirements in 40 CFR 98.94, 98.96, and 98.97 to remove recipe-specific provisions.

As described in Section III.B.2 of this preamble, after subpart I was promulgated, the EPA received additional data characterizing emissions from the semiconductor manufacturing industry and supporting revised default gas utilization and by-product formation rates for the plasma etch process. As discussed in Section III.B.2 of this preamble, we are proposing revised default utilization rate and by-product formation rates for the plasma etch and chamber cleaning process types. The EPA believes that the revised default emission factors (based on process type, gas, and wafer size) would provide reliable facility-specific GHG data. Like other semiconductor manufacturing facilities, new facilities manufacturing semiconductors on wafers greater than 300 mm diameter would not be required to develop recipe-specific gas utilization rates and by-product formation rates and would use either the default factors for 300 mm wafers or stack testing. In the future, the EPA will likely develop default gas utilization rates and by-product formation rates specifically for facilities using wafers greater than 300 mm as that technology is implemented and emissions data are available and collected by the EPA (see Section III.B.12 of this preamble).

As described in Section III.B.1 of this preamble, the EPA is also proposing to include a method using stack testing to develop fab-specific F-GHG emission factors for all electronics manufacturing facilities. The EPA believes that the addition of the stack testing method would also provide representative facility-specific GHG data for all types of electronics manufacturing facilities, including new facilities manufacturing semiconductors on wafers greater than 300 mm diameter. Allowing a stack test approach in addition to the revised default emission factor approach would give reporters flexibility to choose from alternative methods if the recipe-specific approach is removed as the EPA is proposing. For example, facilities with a large number of stacks may prefer the default emission factor approach, whereas a facility with a small number of stacks may desire the stack test method. Compared to the recipe-specific approach, the default emission factor and stack test options would reduce or eliminate the burden, technical, and logistical feasibility concerns raised by the Petitioner.

Finally, the proposed default gas utilization rates and by-product formation rate and stack test alternatives are more compatible with the existing infrastructure, equipment, data management, and recordkeeping systems currently used by the industry than the recipe-specific approach. The proposed approaches would ensure that the EPA would continue to receive representative data for characterizing the F-GHG emissions from the industry while reducing burden on reporting facilities.

Although the EPA has deferred the mandatory use of recipe-specific gas utilization rates and by-product formation rates through the end of 2013 (76 FR 59542, September 27, 2011), we are proposing that the requirements to use recipe-specific rates in 40 CFR 98.93(a)(2)(ii)(A), (a)(3), and (a)(4) would be removed and therefore no longer be effective beginning January 1, 2014. Under the proposed amendments, no semiconductor manufacturing facility would have the option to use the recipe-specific method or report those data elements after the end of 2013. In addition, the recipe-specific method would be removed as an option for other electronics manufacturing facilities for the same reasons related to burden and technical feasibility that it would be removed for semiconductor manufacturing facilities.

As described in Section II.B of this preamble, the proposed rule may not be finalized until the second half of 2013. Therefore, reporters currently using the recipe-specific methods of 40 CFR 98.93(a)(2)(ii)(A), (a)(3), and (a)(4), if any, would be allowed to continue to use these methods for estimating 2013 emissions reported in 2014. Following the January 1, 2014 effective date, reporters would be required to select new calculation methods to estimate emissions for 2014 reported in 2015, and thereafter, based on the options in the final amendments to subpart I.

Finally, we are also proposing to revise 40 CFR 98.93(a)(6) to remove the option to develop recipe-specific gas utilization rates and by-product formation rates for F-GHG and process combinations for which no default emission factors are available, and to revise 40 CFR 98.93(b)(1)(i) and (b)(2)(i) to remove the option to develop facility-specific N

O emission factors. These options would present essentially the same technical problems as the provisions for developing recipe-specific F-GHG rates elsewhere in the rule, including for the facility-specific N

O factors.

Under 40 CFR 98.93(a)(6), facilities would assume that F-GHG emissions equal F-GHG consumption, which is equivalent to treating the utilization and by-product formation rates for gas and process combinations without default factors as both zero. However, the number of default gas utilization rates and by-product formation rates for different gas and process combination is sufficiently broad that the fraction of total emissions represented by emissions estimated under 40 CFR 98.93(a)(6) would be minimal. Under the proposed revisions to 40 CFR 98.93(b), facilities would use default N

O emission factors for both CVD processes and for the aggregate of all other manufacturing production processes, and would not have the option to develop facility-specific N

O emission factors.

We specifically request comment on whether facilities are currently using or plan to use the recipe-specific approach from the final subpart I rule in 40 CFR 98.93(a)(6), or the facility-specific approach for N

O emissions in 40 CFR

98.93(b), for the 2013 reporting year or beyond and whether removal of these methods would significantly impact facilities.

4. Applicability and Calculating Annual Manufacturing Capacity for Facilities That Manufacture Electronics

The EPA is proposing to revise the calculation to determine annual capacity for electronics manufacturing facilities, which is used in the calculation to determine whether a facility meets the reporting threshold. The current subpart I applicability threshold for semiconductor, MEMS, and LCD manufacturing relies on 2006 IPCC Tier 1 emission factors

and the annual manufacturing capacity of the facility. (For PV manufacturing, emissions for applicability determinations are determined by multiplying annual F-GHG purchases or consumption by the gas-appropriate GWPs.) Electronics manufacturing facilities with total facility emissions equal to or greater than 25,000 mtCO

e must report under subpart I. For the applicability determination, emissions from the electronics manufacturing operations at the facility are calculated using the methods in 40 CFR 98.91 instead of the methods in 40 CFR 98.93. The current methods under 40 CFR 98.91 calculate emissions based on the maximum designed capacity of the facility (measured in surface area of substrate produced) and do not account for the effect of GHG abatement systems. Facilities whose total reported emissions, including the emissions from electronics manufacturing calculated according to 40 CFR 98.93, are below the 25,000 mtCO

e threshold can stop reporting if they meet the criteria in 40 CFR 98.2(i).

http://www.ipcc-nggip.iges.or.jp/public/2006gl/index.html

The current subpart I also requires different methods for semiconductor facilities to calculate and report their F-GHG emissions based on the annual manufacturing capacity of the semiconductor facility and the size of wafers the semiconductor facility is manufacturing.

The facility's manufacturing capacity is calculated using Equation I-5, which specifies the manufacturing capacity as 100 percent of the annual manufacturing capacity of a facility, as determined by summing the area of maximum designed substrate starts of a facility per month over the reporting period. “Maximum designed substrate starts” is currently defined in 40 CFR 98.98 as “the maximum quantity of substrates, expressed as surface area, that could be started each month during a reporting year if the facility were fully equipped as defined in the facility design specifications and if the equipment were fully utilized. It denotes 100 percent of annual manufacturing capacity of a facility.”

Facilities manufacturing MEMS, PVs, and LCDs use the same method regardless of facility manufacturing capacity. Facility manufacturing capacity is still used to determine applicability according to 40 CFR 98.91.

Following the publication of the final subpart I rule, the Petitioner stated in the Petition for Reconsideration that the maximum capacity calculation methods assume that a facility has both a full complement of equipment that corresponds to its design, and that the full complement of equipment is utilized to a maximum degree. The Petitioner stated that the reliance on a “fully equipped” facility and “fully utilized” equipment does not reflect the majority of semiconductor facilities, which may increase or reduce production to meet market demands or update their process to create new products. In the Petition for Reconsideration, the Petitioner noted that many facilities are built to reach a certain maximum capacity but are only equipped in stages (for example, one production line at a time), and that older facilities may have been built for a certain capacity but may only be used partially as part of the original equipment is sold or moved to a newer facility. The Petitioner requested that the method for calculating manufacturing capacity, including the definition of “maximum designed substrate starts,” correlate to a facility's actual current equipped capacity.

The EPA agrees that a facility's annual capacity may not be reflected by the designed capacity of a “fully equipped” and “fully utilized” facility, because some equipment that is part of the original design configuration may not yet be installed, or some equipment may be removed and not replaced. Therefore, the EPA is proposing to replace the phrase “maximum designed substrate starts” in Equation I-5 with the phrase “maximum substrate starts.” Likewise, we are proposing to replace the definition in 40 CFR 98.98 of “maximum designed substrate starts” with that for “maximum substrate starts,” which would mean “the maximum quantity of substrates, expressed as surface area, that could be started each month during a reporting year based on the equipment installed in that facility and assuming that the installed equipment were fully utilized. Manufacturing equipment is considered installed when it is on the manufacturing floor and connected to required utilities.”

A facility would continue to use Equation I-5, with this revision, to determine the annual manufacturing capacity of the facility to determine if they meet the threshold for reporting under subpart I.

The proposed changes retain the requirement to calculate and report the maximum annual capacity of the facility (see 40 CFR 98.96(a)), but clarify that the maximum capacity is based on the equipment on-site in the reporting year, assuming it is fully utilized, rather than the design capacity.

The proposed changes would not affect the applicability of subpart I to any facility that is already reporting GHG emissions under subpart I. If the proposed changes become final, facilities that are already reporting would not be able to re-calculate emissions using the procedures under 40 CFR 98.91 and cease reporting if they do not meet the revised applicability criteria. Facilities may cease reporting only if they meet the criteria in 40 CFR 98.2(i).

We are also proposing to remove the requirement that semiconductor manufacturing facilities calculate and report their F-GHG emissions based on the annual manufacturing capacity of the facility and the size of wafers that the facility is manufacturing. Subpart I currently distinguishes between “large” and “other” semiconductor facilities based on the calculated annual manufacturing capacity. Except as provided in the September 27, 2011 final rule titled “Changes to Provisions for Electronics Manufacturing to Provide Flexibility in 2011 to 2013,” subpart I requires “large” semiconductor facilities (facilities with an annual manufacturing capacity of greater than 10,500 m

of substrate) and those facilities that manufacture wafers greater than 300 mm in diameter to calculate emissions using recipe-specific utilization and by-product formation rates. As discussed in Sections III.B.1 through III.B.3 of this preamble, we are proposing to revise the calculation methodologies for semiconductor manufacturers. The proposed calculation methods would apply to all semiconductor manufacturers and there is no longer a need to distinguish “large” facilities based on manufacturing capacity.

5. Integrated Production and R&D Activities for Facilities That Manufacture Electronics

The October 30, 2009 final GHG reporting rule (74 FR 56260) defined research and development (R&D) activities as “those activities conducted in process units or at laboratory bench-scale settings whose purpose is to conduct research and development for new processes, technologies, or products and whose purpose is not for the manufacture of products for commercial sale, except in a

de minimis

manner.” (See 40 CFR 98.6.) At that time, emissions from R&D were expected to be small, and these activities were not expected to significantly contribute to the total emissions from a reporting facility. The final subpart I rule (75 FR 74774, December 1, 2010) did not change the provisions for R&D activities, but deferred to the requirements found in 40 CFR part 98, subpart A.

Following the publication of the final subpart I rule, the Petitioner stated in the Petition for Reconsideration that the final subpart I rule does not account for semiconductor manufacturing facilities that are unable to segregate their R&D activities from production manufacturing. The Petitioner stated in the petition that in order to remain globally competitive, semiconductor companies must engage in robust R&D efforts aimed at innovating new manufacturing processes and new recipes. The petition further stated that many semiconductor facilities integrate their R&D processes into their manufacturing facilities to better consider process manufacturability. The Petitioner stated that many facilities that have integrated R&D cannot segregate gas consumption and emissions from regular production activities.

To date, no facilities covered by other source categories have requested a change to the R&D exemption. However, based on the additional information provided by facilities subject to subpart I, the EPA believes that certain facilities in the electronics manufacturing industry may have unique R&D activities that are integrated into production. In some cases, facilities with integrated R&D may use the same gases from the same containers for both R&D activities and normal production. The EPA agrees that for these electronics manufacturing facilities, it is not feasible to accurately segregate gas consumption for R&D activities from production activities without measuring consumption at the level of the individual tool, or by the individual wafer. (See “Technical Support for Other Technical Issues Addressed in Revisions to Subpart I,” Docket ID No. EPA-HQ-OAR-2011-0028.) Because gas consumption is the basis for estimating emissions from the electronics industry, segregating gas consumption for R&D and production would be essential to segregating the emissions from the respective processes, and this is not currently feasible at many facilities. Therefore, the EPA is proposing to allow all electronics manufacturing facilities covered by subpart I who cannot segregate R&D emissions to report R&D emissions with their total facility emissions and to identify that emissions associated with R&D activities are included in their overall emissions estimates. We are also proposing that facilities reporting integrated R&D emissions must report an estimate of the range of the percentage of total emissions from their R&D activities as part of their annual report (see proposed 40 CFR 98.96(x) and 40 CFR 98.97(j)).

6. Accuracy and Precision of Monitoring Instrumentation for Facilities That Manufacture Electronics

Subpart I currently requires all flow meters, weigh scales, pressure gauges, and thermometers used for measurements to have an accuracy and precision of one percent of full scale or better (40 CR 98.94(i)). In comments to the April 12, 2010 proposed subpart I rule (75 FR 18652), the Petitioner stated that many older facilities in the electronics manufacturing industry do not have the ability or the available instrumentation to measure all quantities, primarily F-GHG and N

O gas consumption, used to calculate GHG emissions to an accuracy and precision of 1 percent of full scale or better (see “Response to Public Comments, Subpart I—Electronics Manufacturing,” Docket ID. No EPA-HQ-OAR-2009-0927-0228). Therefore, these facilities would have difficulty achieving compliance with the accuracy and precision requirements of the subpart without purchasing and installing new measurement equipment. The Petitioner provided additional data in these comments and in the Petition for Reconsideration that these older facilities typically have accuracies of 2 to 4 percent, and requested that the accuracy requirements for subpart I account for the technical capabilities of older facilities, who may find installing new measurement equipment problematic based on existing equipment configurations.

The EPA recognizes that some of the older facilities required to report under subpart I may have difficulty achieving compliance with the current accuracy and precision requirements. Additionally, the EPA evaluated the existing accuracy and precision requirements in 40 CFR part 98, subpart A, which require flow meters to have a calibration error of not more than 5 percent of the reference value (not full scale) (see 40 CFR 98.3(i)). The 5 percent calibration error requirements of 40 CFR 98.3(i) apply only to gas and liquid flow meters used to measure fuel, process streams, or feedstocks; they do not apply to weigh scales, pressure gauges, and thermometers. Under 40 CFR 98.3(i), these latter measurement devices must be calibrated to meet the accuracy requirement specified for the device in the applicable source category subpart, or, in the absence of an accuracy requirement, the device must be calibrated based on other available standards, such as manufacturer's specifications and industry standards.

The EPA is proposing to remove the 1 percent accuracy and precision requirements in subpart I (40 CFR 98.94(i)). Instead, we are proposing that electronics manufacturing facilities subject to subpart I would be required to meet the existing General Provision calibration accuracy requirements in subpart A (40 CFR 98.3(i)). This would provide a balance between the technical issues raised by the Petitioner and the need to gather data for F-GHGs and N

O with a reasonable degree of accuracy. The EPA believes that the subpart A requirements would be appropriate for electronics manufacturing facilities and would address the concerns of the older facilities. Under this proposal, the calibration accuracy requirements for gas flow measurement devices would be 5 percent, as specified in 40 CFR 98.3(i). Further, other measuring devices (e.g., weigh scales and thermometers) would be required to be calibrated to an accuracy based on an applicable operating standard, including, but not limited to, device manufacturer's specifications and industry standards (see 40 CFR 98.3(i)(1)(i)).

The EPA does not expect that this change will impact the accuracy of facility F-GHG and N

O emission estimates at facilities that are using measurement equipment that meets the one percent of full scale standard. It may affect the accuracy of F-GHG and N

O emission estimates at older facilities that have less accurate measurement equipment. However, the subpart A requirements, which appear in 40 CFR 98.3(i), still require an appropriate amount of accuracy in measurement equipment used for compliance. The accuracy requirements in subpart A that we propose to apply

to subpart I are a minimum requirement. Facilities that are currently meeting the higher accuracy standard in subpart I would be expected to continue to use the same monitoring equipment and achieve the same level of accuracy, and would not be expected to “fall back” to the minimum accuracy requirement in subpart A by, for example, replacing current equipment with less accurate monitoring equipment.

7. Facility-Wide Gas Specific Heel Factor for Facilities That Manufacture Electronics

The 2010 final subpart I rule requires electronics manufacturing facilities to calculate emissions from gas consumption and account for the residual amount of gas left in containers that are returned to the gas supplier. This residual amount of gas is referred to as a “heel.” Facilities establish a trigger point based on cylinder weight or gas pressure for each gas and type or size of container used by the facility to indicate that the cylinder should be changed for a full one.

Specifically, the final subpart I rule requires electronics manufacturing facilities to calculate a facility-wide heel factor for each gas to account for the amount of gas represented by the heel in the emissions calculations. Subpart I also requires facilities to “re-calculate a facility-wide gas-specific heel factor if you use a trigger point for change out for a gas and container type that differs by more than 5 percent from the previously used trigger point for change out for that gas and container type.” Additionally, the final subpart I rule requires measuring the pressure or weight of the container when an exceptional circumstance occurs; an “exceptional circumstance” is a change out point that differs by more than 20 percent from the trigger point for change out used to calculate the facility-wide gas-specific heel factor for that gas and container type. See 40 CFR 98.94(b).

The requirement to re-calculate the facility-wide gas-specific heel factor if the trigger point for change out differs by more than 5 percent is one of the issues identified in the Petition for Reconsideration. In the Petition for Reconsideration, the Petitioner stated that the requirement is technically infeasible for certain facilities using small containers, because the level of accuracy associated with these measurements may not be achievable. Specifically, the Petitioner provided the example of a facility using a 20-pound cylinder with a trigger point of 2 pounds. The Petitioner stated that any change in this trigger point of more than 0.1 pounds would require a facility to “recalculate a facility-wide gas specific heel factor,” and any deviation in the actual change out point of more than 0.4 pounds would require handling as an “exceptional circumstance.” The Petitioner stated that, in the context of using hundreds of cylinders, the re-calculation requirement presents a significant amount of management in terms of tracking and administrative tasks, for a minimal difference in the accuracy of the emission estimates reported.

The EPA did not intend to require facilities to recalculate the facility-wide heel factor whenever the actual heel in a container deviated from trigger point by more than 5 percent. The EPA is proposing to amend the requirements to clarify that recalculating the heel factor is only needed when the trigger point for a specific gas and cylinder type is changed, and not as a result of variation in the actual heel remaining in a cylinder. The trigger point is changed by the facility operators to account for changes in the type or size of containers, or to reflect changes in the process operating requirements that would allow for a lower heel factor to be used to utilize a greater fraction of the gas in a container, or that may require a larger heel factor as a more conservative margin before a container is empty. Subpart I has separate provisions at 40 CFR 98.94(b)(4) to address exceptional circumstances in which the amount of heel in a cylinder deviates substantially from the usual trigger point. We are proposing to amend 40 CFR 98.94(b)(5) to clarify that a gas-specific heel factor must be recalculated when the facility executes a process change to modify the trigger point for a gas and container type that differs by more than 5 percent from the previously used trigger point for that gas and container type. The proposed amendments would clarify the EPA's intent that facilities recalculate the heel factor when there are process changes that would substantially alter the trigger point, and that facilities do not need to recalculate the heel factor to reflect variation in the actual heel quantities in cylinders.

The EPA is also proposing to revise the “exceptional circumstance” criteria at 40 CFR 98.94(b)(4) with respect to small containers because while the current criteria are appropriate for large cylinders, treating small containers in the same manner may be burdensome. Specifically, we are proposing to revise the criteria for an “exceptional circumstance” in 40 CFR 98.94(b)(4) from 20 percent of the original trigger point for change out to 50 percent for small cylinders. We are proposing to define a small cylinder as a container that contains less than 9.08 kg (20 pounds) of gas. For large containers, the “exceptional circumstance” would remain as a change out point that differs by 20 percent of the trigger point used to calculate the gas-specific heel factor. We are proposing to revise the criteria for small containers to 50 percent to reduce the burden for facilities using small containers and still maintain the accuracy needed for accounting for the heel in both small and large containers. These proposed changes take into account the fact that a small amount of F-GHGs can account for a large fraction of the heel factor in a small container, and that normal variation in day-to-day container management could be more likely to trigger an “exceptional circumstance.” At the same time, the proposed revisions would still require facilities to directly measure the heel in cases where the cylinder change out deviated from the established trigger point. For example, a small 15-pound cylinder with a 2-pound trigger point would still need to be measured, in lieu of using the established heel factor, if the difference in the change out point was greater than 1 pound. In this example, this 1-pound difference (based on the proposed 50-percent criteria for an exceptional circumstance) represents less than 8 percent of the usable gas in the cylinder. Under the current 20-percent criteria, a difference from the actual trigger point of 0.4 pounds (20 percent of the 2-pound trigger point), would represent about 3 percent of the usable gas in the cylinder. These small cylinders for which we are proposing to change the exceptional circumstance criteria generally represent a small percentage of overall gas consumption. The EPA understands that cylinder size is generally chosen to reflect overall consumption, with larger cylinder sizes chosen by the facility for those gases used in larger quantities.

8. Apportioning Model Verification for Facilities That Manufacture Electronics

Subpart I requires electronics manufacturing facilities to estimate emissions from gas consumption and report the input gas consumed for each individual process sub-type or process type using Equation I-13. Equation I-13 requires the use of an apportioning factor, which is developed for F-GHG and N

O input gases using a facility-specific engineering model, and is expressed as a fraction of the input gas used for each process sub-type or process type. Reporters have the flexibility to develop the model based on any quantifiable metric selected by the facility (such as wafer passes or wafer starts), but must verify the model

by comparing the modeled and actual gas use for the largest gas used for plasma etch and the largest gas used for chamber cleaning. Additionally, the difference between actual and modeled plasma etch gas consumption must not exceed 5 percent. The provisions of 40 CFR 98.94(c)(2)(i) also require that for verifying the model, facilities analyze a 30-day period of operation during which the utilized capacity of the facility equals or exceeds 60 percent of its design capacity, or if the utilized capacity is less than 60 percent during the reporting year, a period during which the facility experiences its highest 30-day average utilization. This approach allows reporters to select the most appropriate quantifiable metric for their facility while providing consistent verification methods.

The Petition for Reconsideration raised concerns that the verification requirements for the apportioning engineering model were overly burdensome. The Petitioner stated that the hardware and infrastructure for apportioning gas consumption by process type or sub-type to meet this requirement are not in place at most facilities, and would require installation of additional equipment to measure and record gas consumption at the individual tool level for developing and confirming the model at the 5 percent accuracy level.

However, the Petitioner also noted that some facilities may be configured such that they are able to apportion gas consumption to one or more process types or process sub-types based on gas connections and measured flow rates (see “Technical Support for Other Technical Issues Addressed in Revisions to Subpart I,” Docket ID no. EPA-HQ-OAR-2011-0028). They requested that the rule accommodate both a modeling and a measurement approach.

The Petitioner also stated that the verification period criteria in 40 CFR 98.94(c)(2)(i) are not practicable. Specifically, the Petitioner pointed out that the data needed to assess the period with the highest 30-day average utilization may not be available until the end of the reporting year. As a result, facilities may not have enough time to identify and select the assessment period, complete and compare the modeling and measurement analysis, or make corrections prior to the applicable reporting deadline in the following year (see “SIA Revised Proposal to Amend the Apportionment Model Validation Criteria in 40 CFR 98.94(c),” Docket ID no. EPA-HQ-OAR-2011-0028). Based on these concerns, the Petitioner requested that the rule be revised to allow facilities to select a period of operation for model verification that is representative of normal operation, up to and including the full calendar year of operation.

Additionally, in the Petition for Reconsideration the Petitioner questioned the requirement to demonstrate that the model provides a measurement of gas consumption that is accurate to within 5 percent of the actual measurement. The petition stated that data provided from one manufacturer showed that, for a single tool running two recipes, the difference between modeled gas consumption and actual gas consumption was greater than 5 percent (see “Verification Tests to Demonstrate Difficulty of Achieving 5 percent Limit,” Docket ID. No EPA-HQ-OAR-2011-0028). The Petitioner explained that facilities running a number of tools with a larger number of recipes would have greater uncertainties and would be unable to meet the verification requirements of the final rule. Furthermore, they stated that some facilities would require monitoring, collecting, and analyzing data from the mass flow meters for all tools to accurately model, verify, and achieve the 5 percent verification requirement.

The EPA received comments with similar concerns in response to the June 22, 2011 proposed rule titled “Changes to Provisions for Electronics Manufacturing (Subpart I) To Provide Flexibility” (76 FR 36472). In the preamble to the corresponding final rule (76 FR 59542, September 27, 2011), the EPA responded that apportioning is a particularly important component in estimating emissions of F-GHGs from electronics manufacturing because the consumption of gas by process type or sub-type is one of the major sources of error in estimating GHG emissions. The EPA also noted in that response that facilities that could not meet the apportioning model verification requirements in subpart I had the option to apply for, and if approved by the Administrator, use BAMM in 2011, 2012, and 2013. The EPA reported in that preamble that we had received only a small number of requests to use BAMM, relative to the number of facilities expected to report under subpart I. The EPA concluded that while some facilities were unable to meet the model verification requirements, the problem was limited.

Despite the problem being limited to particular facilities, the EPA wants to ensure that all facilities can comply with subpart I. The EPA recognizes that some facilities may still not be able to meet the present apportioning model verification requirements in 40 CFR 98.94(c)(2), even though other changes being proposed today would reduce the need to apportion gas consumption. For example, the proposed stack test alternative and the revised default utilization and by-product formation rates would reduce the need to apportion gas among tools or process types. According to the Petitioner, the situation would be most complicated for semiconductor facilities using 150 or 200 mm wafers because they would typically need to apportion three to five different gases between plasma etch and chamber cleaning process types. At 300 mm fabs, NF

appears to be the only gas that needs to be apportioned between plasma etch and chamber cleaning process types, based on information provided by the Petitioner.

Even though facilities would have a reduced need to apportion gas consumption between the plasma etch and chamber clean process types, the EPA recognizes that many would still need to apportion gas consumption between abated and unabated tools and, if they were to use the proposed stack testing option, they may also need to apportion gas consumption between stack systems that are tested and those that are not. As a result, certain facilities would still face issues of technical feasibility in meeting the apportioning model verification requirement requiring a 5 percent maximum difference between modeled and actual F-GHG consumption.

In light of these concerns, the EPA is proposing to amend the verification requirements. First, the proposed amendments would allow reporters the option to use direct measurements of gas consumption to avoid the need to develop an apportioning model, and to develop an apportioning factor for each process type, sub-type, stack system, or fab using gas flow meters or weigh scales because direct measurements would provide the most accurate data for analysis. However, the proposed rule would retain the option to use an apportioning model to allow for greater flexibility for electronics manufacturers and reduce the burden for facilities with a larger number of tools, gases, or process types and sub-types. The model verification requirements would be retained to ensure that reporters across the industry are providing data of consistent quality. Reporters opting to use the apportioning model would be required to verify the model by comparing actual gas consumption to modeled gas consumption. The reporter would select for comparison the F-GHG that corresponds to the largest quantity, on a mass basis, of F-GHG used at the

fab that has to be apportioned. Reporters would have the flexibility to verify the model for two F-GHGs on an aggregate use basis if one of the gases selected is used in the largest quantity at each fab. In this option, the predicted total volume consumed of the two gases combined would be required to match the actual total volume consumed within the verification percent difference requirements for the apportioning model. Reporters would use this latter option to account for the fact that they may not be able to predict which gas will be used in the largest quantity as of the end of the year, but they want to verify the model at some point early in the year. For example, a facility may predict that one of two gases, CF

and C

, would be used in the largest quantity as of the end of the year, but they do not know which one. However, they believe that the two-month period from March to April is the most representative period of operations, and they may select that period because that is when they will be performing stack testing. The facility could verify the model for both gases based on data from March and April. At the end of the year, the facility would confirm that at least one of those two gases was used in the highest quantity and both gases met the verification criteria on an aggregate basis. Reporters would be required to correct the model if it did not meet the verification requirements.

Second, where a facility opts to develop and use an apportioning model, we are also proposing to revise the verification standard to increase the allowable difference between the actual and modeled gas consumption from a maximum 5 percent difference to a maximum of 20 percent difference. The data provided in an industry analysis submitted with the Petition for Reconsideration have shown that the 5 percent difference criterion would be difficult to achieve under most operating scenarios and would require installation of additional equipment. Increasing the allowable difference between the actual and modeled gas consumption from a maximum 5 percent difference to a maximum 20 percent difference would also reduce the burden on facilities by providing greater flexibility in the methods they use for modeling gas consumption. This will reduce the potential that they will need to purchase and install new equipment to measure, record, and analyze data for gas consumption at the level of the individual tool, process type, or process sub-type.

As a result of other rule changes being proposed today, including the combining of the wafer clean and plasma etch process categories for semiconductor manufacturing and the elimination of the use of recipe-specific gas utilization rates and by-product formation rates for semiconductor manufacturing, the number of gases that would need to be apportioned among process types and sub-types would be reduced for semiconductor manufacturing facilities, especially for semiconductor manufacturing facilities using 300 mm wafers. For facilities that are using 300 mm, only NF

is commonly used in both the plasma etch and chamber clean process types. For facilities that are using 150 mm or 200 mm wafers, several F-GHG are used in both the plasma etch and chamber clean process types. Therefore, the potential effect of the proposed increase in the allowable difference between modeled and actual gas consumption on overall uncertainty of the GHG emission estimates has been minimized for semiconductor manufacturing facilities using 300 mm wafers that need to apportion gas usage among process types or sub-types compared to the standards promulgated in December 2010. However, it is not clear what effect this change will have on facilities using 150 mm and 200 mm wafers because of the number of gases that are used in both plasma etching and chamber cleaning process types.

The proposed change in the apportioning model criteria would also apply to LCD, MEMS, and PV manufacturing facilities. For LCD manufacturing, only SF

is commonly used in both the plasma etching and chamber cleaning process types and would need to be apportioned between those process types. For both MEMS and for PV, several F-GHGs are typically used in both the plasma etching and chamber cleaning process types and would need to be apportioned between the two process types.

It is also important to note that facilities would be required to apportion gas consumption between tools and processes for which they are claiming emission reductions as a result of abatement systems, and some facilities do not have abatement systems on all of their tools. For these reasons, we are specifically seeking comment on the need to change the verification model criterion from 5 percent maximum allowed difference to 20 percent, and the effect that this proposed change may have on the error or uncertainty associated with the F-GHG emission estimates at facilities that need to apportion several gases between process types, or between tools that do or do not have abatement systems.

We also agree with the Petitioner that facilities should be able to select a longer period of operation as the basis for verifying their apportioning models. We agree that they should be able to compare modeled to actual gas consumption for the whole year to verify the model, because it may be difficult to identify in advance a shorter period that meets the production criteria in 40 CFR 98.94(c)(2)(i). The current rule specifies that facilities analyze a period of at least 30-days operation to verify the model, but does not specify a maximum allowed period; it specifies a minimum of 30 days to ensure that data are representative of normal operation.

We are also proposing to allow the facility to select a period of the reporting year when the fab is at a “representative operating level” for the model verification, instead of at a minimum percent of design capacity, or instead of at the highest 30-day average utilization. The concept of a representative operating level would replace the current requirement in 40 CFR 98.94(c)(2)(i) that the facility be operating at 60 percent or more of its design capacity during the model verification, or that the verification occur during the period with the highest 30-day average for facility utilization if the facility operates below 60 percent of design capacity. The Petitioner pointed out that, under the current rule, it is difficult for a facility operating below 60 percent capacity to determine which 30-day period would have the highest average facility utilization. Furthermore, a facility that performs a validation early in the year while operating at less than 60 percent capacity may need to repeat the verification if production dramatically increased later in the year such that the facility was operating above 60 percent of design capacity. (The proposed amendment to adopt the definition of a “representative operating level” is described in detail in Section III.B.1 of this preamble.)

Under this proposal, the representative period would still be at least 30 days, but we are proposing to clarify that it can be up to the whole calendar reporting year in duration. Because the proposed requirements would allow the use of a representative operating level, facilities would be able to determine the assessment period with less chance of having to repeat the verification, complete and compare the modeling and measurement analysis, and make corrections to the model, if needed, prior to the March report submittal deadline for a given reporting year.

9. Calculating N

O Emissions for Facilities That Manufacture Electronics

The EPA is proposing to revise the language for calculating N

O emissions in 40 CFR 98.93(b) to clarify that reporting is at the fab level. In the Petition for Reconsideration, the Petitioner requested clarification of the requirements to calculate annual facility-level N

O emissions for CVD processes for electronics manufacturing facilities. The current subpart I states in 40 CFR 98.93(b) that facilities “must calculate annual facility

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.