Skip to Content

Rule

Petroleum Refinery Sector Risk and Technology Review and New Source Performance Standards

Document Details

Information about this document as published in the Federal Register.

Published Document

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

Start Preamble Start Printed Page 75178

AGENCY:

Environmental Protection Agency (EPA).

ACTION:

Final rule.

SUMMARY:

This action finalizes the residual risk and technology review conducted for the Petroleum Refinery source categories regulated under national emission standards for hazardous air pollutants (NESHAP) Refinery MACT 1 and Refinery MACT 2. It also includes revisions to the Refinery MACT 1 and MACT 2 rules in accordance with provisions regarding establishment of MACT standards. This action also finalizes technical corrections and clarifications for the new source performance standards (NSPS) for petroleum refineries to improve consistency and clarity and address issues related to a 2008 industry petition for reconsideration. Implementation of this final rule will result in projected reductions of 5,200 tons per year (tpy) of hazardous air pollutants (HAP) which will reduce cancer risk and chronic health effects.

DATES:

This final action is effective on February 1, 2016. The incorporation by reference of certain publications for part 63 listed in the rule is approved by the Director of the Federal Register as of February 1, 2016. The incorporation by reference of certain publications for part 60 listed in the rule were approved by the Director of the Federal Register as of June 24, 2008.

ADDRESSES:

The Environmental Protection Agency (EPA) has established a docket for this action under Docket ID No. EPA-HQ-OAR-2010-0682. All documents in the docket are listed on the www.regulations.gov Web site. Although listed in the index, some information is not publicly available, e.g., confidential business information (CBI) or other information whose disclosure is restricted by statute. Certain other material, such as copyrighted material, is not placed on the Internet and will be publicly available only in hard copy form. Publicly available docket materials are available either electronically through http://www.regulations.gov, or in hard copy at the EPA Docket Center, WJC West Building, Room Number 3334, 1301 Constitution Ave. NW., Washington, DC. The Public Reading Room hours of operation are 8:30 a.m. to 4:30 p.m. Eastern Standard Time (EST), Monday through Friday. The telephone number for the Public Reading Room is (202) 566-1744, and the telephone number for the Air and Radiation Docket and Information Center is (202) 566-1742.

Start Further Info

FOR FURTHER INFORMATION CONTACT:

For questions about this final action, contact Ms. Brenda Shine, Sector Policies and Programs Division, Refining and Chemicals Group (E143-01), Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina, 27711; telephone number: (919) 541-3608; fax number: (919) 541-0246; and email address: shine.brenda@epa.gov. For specific information regarding the risk modeling methodology, contact Mr. Ted Palma, Health and Environmental Impacts Division (C539-02), Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711; telephone number: (919) 541-5470; fax number: (919) 541-0840; and email address: palma.ted@epa.gov. For information about the applicability of the NESHAP to a particular entity, contact Ms. Maria Malave, Office of Enforcement and Compliance Assurance, U.S. Environmental Protection Agency, William Jefferson Clinton Building, 1200 Pennsylvania Ave. NW., Washington, DC 20460; telephone number: (202) 564-7027; fax number: (202) 564-0050; and email address: malave.maria@epa.gov.

End Further Info End Preamble Start Supplemental Information

SUPPLEMENTARY INFORMATION:

Preamble Acronyms and Abbreviations. We use multiple acronyms and terms in this preamble. While this list may not be exhaustive, to ease the reading of this preamble and for reference purposes, the EPA defines the following terms and acronyms here:

10/25 tpy emissions equal to or greater than 10 tons per year of a single pollutant or 25 tons per year of cumulative pollutants

AEGL acute exposure guideline levels

APCD air pollution control devices

API American Petroleum Institute

BAAQMD Bay Area Air Quality Management District

BDT best demonstrated technology

BLD bag leak detectors

BSER best system of emission reductions

Btu/ft2 British thermal units per square foot

Btu/scf British thermal units per standard cubic foot

CAA Clean Air Act

CBI confidential business information

CCU catalytic cracking units

CDX Central Data Exchange

CEDRI Compliance and Emissions Data Reporting Interface

CEMS continuous emission monitoring system

CFR Code of Federal Regulations

CO carbon monoxide

CO2 carbon dioxide

CO2 e carbon dioxide equivalents

COMS continuous opacity monitoring system

COS carbonyl sulfide

CPMS continuous parameter monitoring system

CRA Congressional Review Act

CRU catalytic reforming units

CS2 carbon disulfide

DCU delayed coking units

EPA Environmental Protection Agency

ERPG emergency response and planning guidelines

ERT Electronic Reporting Tool

ESP electrostatic precipitator

FCCU fluid catalytic cracking unit

FGCD fuel gas combustion device

FMP flare management plan

FR Federal Register

FTIR Fourier transform infrared spectroscopy

GC gas chromatograph

GHG greenhouse gases

H2 S hydrogen sulfide

HAP hazardous air pollutants

HCl hydrogen chloride

HCN hydrogen cyanide

HF hydrogen fluoride

HFC highest fenceline concentration

HI hazard index

HQ hazard quotient

ICR information collection request

IRIS Integrated Risk Information System

km kilometers

LAER lowest achievable emission rate

lb/day pounds per day

LDAR leak detection and repair

LEL lower explosive limit

LTD long tons per day

MACT maximum achievable control technology

MIR maximum individual risk

mph miles per hour

MPV miscellaneous process vent

NAICS North American Industry Classification System

NESHAP National Emissions Standards for Hazardous Air Pollutants

NFS near-field interfering source

NHVCZ combustion zone net heating value

Ni nickel

NOX nitrogen oxides

NRDC Natural Resources Defense Council

NSPS new source performance standards

NTTAA National Technology Transfer and Advancement Act

OAQPS Office of Air Quality Planning and standards

OECA Office of Enforcement and Compliance Assurance

OEHHA Office of Environmental Health Hazard Assessment

OEL open-ended line

OMB Office of Management and Budget

PM particulate matter

PM2.5 particulate matter 2.5 micrometers in diameter and smaller

ppbv parts per billion by volume

ppm parts per millionStart Printed Page 75179

ppmv parts per million by volume

PRA Paperwork Reduction Act

PRD pressure relief device [1]

psia pounds per square inch absolute

psig pounds per square inch gauge

REL reference exposure level

REM Model Refinery Emissions Model

RFA Regulatory Flexibility Act

RTC response to comment

RTR Risk and Technology Review

SAB Science Advisory Board

SBA Small Business Administration

SCAQMD South Coast Air Quality Management District

SCR selective catalytic reduction

SISNOSE significant economic impact on a substantial number of small entities

SO2 sulfur dioxide

SRP sulfur recovery plant

SRU sulfur recovery unit

SSM startup, shutdown and malfunction

TOSHI target organ-specific hazard index

tpy tons per year

UMRA Unfunded Mandates Reform Act

URE unit risk estimate

UV-DOAS ultraviolet differential optical absorption spectroscopy

VCS voluntary consensus standards

VOC volatile organic compounds

°F degrees Fahrenheit

ΔC the concentration difference between the highest measured concentration and the lowest measured concentration

μg/m3 micrograms per cubic meter

Background Information. On June 30, 2014, the EPA proposed revisions to both of the petroleum refinery NESHAP based on our residual risk and technology review (RTR). In that action, we also proposed to revise the NESHAP pursuant to CAA section 112(d)(2) and (3), to revise the SSM provisions in the NESHAP, and to make technical corrections to the NSPS to address issues related to reconsideration of the final NSPS subpart Ja rule in 2008. In this action, we are finalizing decisions and revisions for these rules. We summarize some of the more significant comments received regarding the proposed rule and provide our responses in this preamble. A summary of all other public comments on the proposal and the EPA's responses to those comments is provided in the “Response to Comment” document, which is available in Docket ID No. EPA-HQ-OAR-2010-0682. The “track changes” version of the regulatory language that incorporates the changes in this final action is also available in the docket for this rulemaking.

Organization of this Document. This preamble is organized as follows:

I. General Information

A. Does this action apply to me?

B. Where can I get a copy of this document and other related information?

C. Judicial Review and Administrative Reconsideration

II. Background

A. What is the statutory authority for this action?

B. How do the NESHAP and NSPS regulate air pollutant emissions from refineries?

C. What changes did we propose for the Petroleum Refinery NESHAP and NSPS in our June 30, 2014 RTR proposal?

III. What is included in this final rule?

A. What are the final NESHAP amendments based on the risk review for the Petroleum Refinery source categories?

B. What are the final NESHAP amendments based on the technology review for the Petroleum Refinery source categories?

C. What are the final NESHAP amendments pursuant to section 112(d)(2) & (3) for the Petroleum Refinery source categories?

D. What are the final NESHAP amendments addressing emissions during periods of SSM?

E. What other revisions to the NESHAP and NSPS are being promulgated?

F. What are the requirements for submission of performance test data to the EPA?

G. What are the effective and compliance dates of the NESHAP and NSPS?

H. What materials are being incorporated by reference?

IV. What is the rationale for our final decisions and amendments to the Petroleum Refinery NESHAP and NSPS?

A. Residual Risk Review for the Petroleum Refinery Source Categories

B. Technology Review for the Petroleum Refinery Source Categories

C. Refinery MACT Amendments Pursuant to CAA section 112(d)(2) and (d)(3)

D. NESHAP Amendments Addressing Emissions During Periods of SSM

E. Technical Amendments to Refinery MACT 1 and 2

F. Technical Amendments to Refinery NSPS Subparts J and Ja

V. Summary of Cost, Environmental, and Economic Impacts and Additional Analyses Conducted

A. What are the affected facilities, the air quality impacts and cost impacts?

B. What are the economic impacts?

C. What are the benefits?

D. Impacts of This Rulemaking on Environmental Justice Populations

E. Impacts of This Rulemaking on Children's Health

VI. Statutory and Executive Order Reviews

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

B. Paperwork Reduction Act (PRA)

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 Concerning Regulations That Significantly Affect Energy Supply, Distribution or Use

I. National Technology Transfer and Advancement Act (NTTAA) and 1 CFR part 51

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

K. Congressional Review Act (CRA)

I. General Information

A. Does this action apply to me?

Regulated Entities. Categories and entities potentially regulated by this action are shown in Table 1 of this preamble.

Table 1—Industrial Source Categories Affected by This Final Action

NESHAP and source categoryNAICS a Code
Petroleum Refining Industry324110
a North American Industry Classification System.

Table 1 of this preamble is not intended to be exhaustive, but rather to provide a guide for readers regarding entities likely to be affected by the final action for the source categories listed. To determine whether your facility is affected, you should examine the applicability criteria in the appropriate NESHAP or NSPS. If you have any questions regarding the applicability of any aspect of these NESHAP or NSPS, please contact the appropriate person listed in the preceding FOR FURTHER INFORMATION CONTACT section of this preamble.

B. Where can I get a copy of this document and other related information?

In addition to being available in the docket, an electronic copy of this final action will also be available on the Internet through the Technology Transfer Network (TTN) Web site, a forum for information and technology exchange in various areas of air pollution control. Following signature by the EPA Administrator, the EPA will post a copy of this final action at: http://www.epa.gov/​ttn/​atw/​petref.html. Following publication in the Federal Register, the EPA will post the Federal Register version and key technical documents at this same Web site.

Additional information is available on the RTR Web site at http://www.epa.gov/​ttn/​atw/​rrisk/​rtrpg.html. This information includes an overview of the RTR program, links to project Web sites Start Printed Page 75180for the RTR source categories, and detailed emissions and other data we used as inputs to the risk assessments.

C. Judicial Review and Administrative Reconsideration

Under CAA section 307(b)(1), judicial review of this final action is available only by filing a petition for review in the United States Court of Appeals for the District of Columbia Circuit by February 1, 2016. Under CAA section 307(b)(2), the requirements established by this final rule may not be challenged separately in any civil or criminal proceedings brought by the EPA to enforce the requirements.

Section 307(d)(7)(B) of the CAA further provides that “[o]nly an objection to a rule or procedure which was raised with reasonable specificity during the period for public comment (including any public hearing) may be raised during judicial review.” This section also provides a mechanism for the EPA to reconsider the rule “[i]f the person raising an objection can demonstrate to the Administrator that it was impracticable to raise such objection within [the period for public comment] or if the grounds for such objection arose after the period for public comment (but within the time specified for judicial review) and if such objection is of central relevance to the outcome of the rule.” Any person seeking to make such a demonstration should submit a Petition for Reconsideration to the Office of the Administrator, U.S. EPA, Room 3000, WJC Building, 1200 Pennsylvania Ave. NW., Washington, DC 20460, with a copy to both the person(s) listed in the preceding FOR FURTHER INFORMATION CONTACT section, and the Associate General Counsel for the Air and Radiation Law Office, Office of General Counsel (Mail Code 2344A), U.S. EPA, 1200 Pennsylvania Ave. NW., Washington, DC 20460.

II. Background

A. What is the statutory authority for this action?

1. NESHAP

Section 112 of the CAA establishes a two-stage regulatory process to address emissions of hazardous air pollutants (HAP) from stationary sources. In the first stage, we must identify categories of sources emitting one or more of the HAP listed in CAA section 112(b) and then promulgate technology-based NESHAP for those sources. “Major sources” are those that emit, or have the potential to emit, any single HAP at a rate of 10 tons per year (tpy) or more, or 25 tpy or more of any combination of HAP. For major sources, these standards are commonly referred to as maximum achievable control technology (MACT) standards and must reflect the maximum degree of emission reductions of HAP achievable (after considering cost, energy requirements, and non-air quality health and environmental impacts). In developing MACT standards, CAA section 112(d)(2) directs the EPA to consider the application of measures, processes, methods, systems or techniques, including but not limited to those that reduce the volume of or eliminate HAP emissions through process changes, substitution of materials, or other modifications; enclose systems or processes to eliminate emissions; collect, capture, or treat HAP when released from a process, stack, storage, or fugitive emissions point; are design, equipment, work practice, or operational standards; or any combination of the above.

For these MACT standards, the statute specifies certain minimum stringency requirements, which are referred to as MACT floor requirements, and which may not be based on cost considerations. See CAA section 112(d)(3). For new sources, the MACT floor cannot be less stringent than the emission control achieved in practice by the best-controlled similar source. The MACT standards for existing sources can be less stringent than floors for new sources, but they cannot be less stringent than the average emission limitation achieved by the best-performing 12-percent of existing sources in the category or subcategory (or the best-performing 5 sources for categories or subcategories with fewer than 30 sources). In developing MACT standards, we must also consider control options that are more stringent than the floor, under CAA section 112(d)(2). We may establish standards more stringent than the floor, based on the consideration of the cost of achieving the emissions reductions, any non-air quality health and environmental impacts, and energy requirements.

In the second stage of the regulatory process, the CAA requires the EPA to undertake 2 different analyses, which we refer to as the technology review and the residual risk review. Under the technology review, we must review the technology-based standards and revise them “as necessary (taking into account developments in practices, processes, and control technologies)” no less frequently than every eight years, pursuant to CAA section 112(d)(6). Under the residual risk review, we must evaluate the risk to public health remaining after application of the technology-based standards and revise the standards, if necessary, to provide an ample margin of safety to protect public health or to prevent, taking into consideration costs, energy, safety and other relevant factors, an adverse environmental effect. The residual risk review is required within eight years after promulgation of the technology-based standards, pursuant to CAA section 112(f). In conducting the residual risk review, if the EPA determines that the current standards provide an ample margin of safety to protect public health, it is not necessary to revise the MACT standards pursuant to CAA section 112(f).[2] For more information on the statutory authority for this rule, see 79 FR 36879.

2. NSPS

Section 111 of the CAA establishes mechanisms for controlling emissions of air pollutants from stationary sources. Section 111(b) of the CAA provides authority for the EPA to promulgate NSPS that apply only to newly constructed, reconstructed and modified sources. Once the EPA has elected to set NSPS for new and modified sources in a given source category, CAA section 111(d) calls for regulation of existing sources, with certain exceptions explained below.

Specifically, section 111(b) of the CAA requires the EPA to establish emission standards for any category of new and modified stationary sources that the Administrator, in his or her judgment, finds “causes, or contributes significantly to, air pollution which may reasonably be anticipated to endanger public health or welfare.” The EPA has previously made endangerment findings under this section of the CAA for more than 60 stationary source categories and subcategories that are now subject to NSPS.

Section 111 of the CAA gives the EPA significant discretion to identify the affected facilities within a source category that should be regulated. To define the affected facilities, the EPA can use size thresholds for regulation and create subcategories based on source type, class or size. Emission limits also may be established either for equipment within a facility or for an entire facility. For listed source categories, the EPA must establish “standards of performance” that apply Start Printed Page 75181to sources that are constructed, modified or reconstructed after the EPA proposes the NSPS for the relevant source category.[3]

The EPA also has significant discretion to determine the appropriate level for the standards. Section 111(a)(1) of the CAA provides that NSPS are to reflect the degree of emission limitation achievable through the application of the best system of emission reduction which (taking into account the cost of achieving such reduction and any non-air quality health and environmental impact and energy requirements) the Administrator determines has been adequately demonstrated. This level of control is commonly referred to as best demonstrated technology (BDT) or the best system of emission reduction (BSER). The standard that the EPA develops, based on the BSER achievable at that source, is commonly a numerical emission limit, expressed as a performance level (i.e., a rate-based standard). Generally, the EPA does not prescribe a particular technological system that must be used to comply with a NSPS. Rather, sources remain free to elect whatever combination of measures will achieve equivalent or greater control of emissions.

Costs are also considered in evaluating the appropriate standard of performance for each category or subcategory. The EPA generally compares control options and estimated costs and emission impacts of multiple, specific emission standard options under consideration. As part of this analysis, the EPA considers numerous factors relating to the potential cost of the regulation, including industry organization and market structure, control options available to reduce emissions of the regulated pollutant(s) and costs of these controls.

B. How do the NESHAP and NSPS regulate air pollutant emissions from refineries?

The EPA promulgated the petroleum refinery NESHAP pursuant to CAA section 112(d)(2) and (3) for refineries located at major sources in two separate rules. On August 18, 1995, the first petroleum refinery MACT standard was promulgated in 40 CFR part 63, subpart CC (60 FR 43620). This rule is known as “Refinery MACT 1” and covers the “Sources Not Distinctly Listed,” meaning it includes all emissions sources from petroleum refinery process units, except those listed separately under the section 112(c) source category list and expected to be regulated by other MACT standards (for example, boilers and process heaters). Some of the emission sources regulated in Refinery MACT 1 include miscellaneous process vents (MPV), storage vessels, wastewater, equipment leaks, gasoline loading racks, marine tank vessel loading and heat exchange systems.

On April 11, 2002 (67 FR 17762), EPA promulgated a second MACT standard regulating certain process vents that were listed as a separate source category under CAA section 112(c) and that were not addressed as part of the Refinery MACT 1. This standard, which is referred to as “Refinery MACT 2”, covers process vents on catalytic cracking units (CCU) (including FCCU), CRU and SRU and is codified as 40 CFR part 63, subpart UUU.

Finally, on October 28, 2009, we revised Refinery MACT 1 by adding MACT standards for heat exchange systems, which the EPA had not addressed in the original 1995 Refinery MACT 1 rule (74 FR 55686). In this same 2009 action, we updated the cross-references to the General Provisions in 40 CFR part 63. On June 20, 2013 (78 FR 37133), we promulgated minor revisions to the heat exchange provisions of Refinery MACT 1.

On September 27, 2012, Air Alliance Houston, California Communities Against Toxics and other environmental and public health groups filed a lawsuit alleging that the EPA missed statutory deadlines to review and revise Refinery MACT 1 and 2. The EPA reached an agreement to settle that litigation and entered into a Consent Decree. The Consent Decree provides for the Administrator to sign a final action no later than September 30, 2015.

Refinery NSPS subparts J and Ja regulated criteria pollutant emissions, including particulate matter (PM), sulfur dioxide (SO2), nitrogen oxides (NOX) and carbon monoxide (CO) from FCCU catalyst regenerators, fuel gas combustion devices (FGCD) and sulfur recovery plants. Refinery NSPS subpart Ja also regulates criteria pollutant emissions from fluid coking units and DCU.

The NSPS for petroleum refineries (40 CFR part 60, subpart J) were promulgated in 1974, amended in 1976 and amended again in 2008, following a review of the standards. As part of the review that led to the 2008 amendments to the Refinery NSPS subpart J, the EPA developed separate standards of performance for new process units (40 CFR part 60, subpart Ja). However, the EPA received multiple petitions for reconsideration on issues related to those standards. The Administrator granted the petitions for reconsideration. The EPA addressed petition issues related to process heaters and flares by promulgating amendments to the Refinery NSPS subparts J and Ja on September 12, 2012 (77 FR 56422). In this action, we are finalizing technical corrections and clarifications to NSPS subparts J and Ja raised by American Petroleum Institute (API) in their 2008 petition for reconsideration that were not addressed by the final NSPS amendments of 2012.

The petroleum refining industry consists of facilities that engage in converting crude oil into refined products, including liquefied petroleum gas, gasoline, kerosene, aviation fuel, diesel fuel, fuel oils, lubricating oils and feedstocks for the petrochemical industry. Currently, 142 facilities have emission sources regulated by either or both Refinery MACT 1 and 2.

Petroleum refinery activities start with the receipt of crude oil for storage at the refinery, include all the petroleum handling and refining operations, and terminate with loading of refined products into pipelines, tank or rail cars, tank trucks, or ships or barges that take products from the refinery to distribution centers. Petroleum-specific process units include FCCU and CRU. Other units and processes found at petroleum refineries (as well as at many other types of manufacturing facilities) include storage vessels and wastewater treatment plants. HAP emitted by this industry include organics (e.g., acetaldehyde, benzene, formaldehyde, hexane, phenol, naphthalene, 2-methylnaphthalene, dioxins, furans, ethyl benzene, toluene and xylene); reduced sulfur compounds (i.e., carbonyl sulfide (COS), carbon disulfide (CS2))); inorganics (e.g., hydrogen chloride (HCl), hydrogen cyanide (HCN), chlorine, hydrogen fluoride (HF)); and metals (e.g., antimony, arsenic, beryllium, cadmium, chromium, cobalt, lead, mercury, manganese and nickel (Ni)). This industry also emits criteria pollutants and other non-HAP, including NOX, PM, SO2, volatile organic compounds (VOC), CO, greenhouse gases (GHG) and total reduced sulfur.Start Printed Page 75182

C. What changes did we propose for the Petroleum Refinery NESHAP and NSPS in our June 30, 2014, RTR proposal?

On June 30, 2014, the EPA published a proposed rule in the Federal Register addressing the RTR for the Petroleum Refinery NESHAP, 40 CFR part 63, subparts CC and UUU. The proposal also included changes pursuant to section 112(d)(2) and (3) and technical revisions to the NSPS. Specifically, we proposed:

(1) Pursuant to CAA sections 112(d)(2) and (3):

a. Refinery MACT 1:

  • Adding MACT Standards for DCU decoking operations.
  • Adding operational requirements for flares used as APCD in Refinery MACT 1 and 2.
  • Adding requirements and clarifications for vent control bypasses in Refinery MACT 1.

b. Refinery MACT 2:

  • Revising the CRU purge vent exemption.

(2) Pursuant to CAA sections 112(d)(6) and 112(f)(2):

  • Revising Refinery MACT 1 to cross-reference the corresponding storage vessel requirements in the Generic MACT (40 CFR part 63, subpart WW, as applicable), and revising the definition of Group 1 storage vessels to include smaller capacity storage vessels and to include storage vessels storing materials with lower vapor pressures.

(3) Pursuant to CAA section 112(d)(6):

a. Refinery MACT 1:

  • Allowing refineries to meet the leak detection and repair (LDAR) requirements in Refinery MACT 1 by monitoring for leaks using optical gas imaging in place of EPA Method 21, once the monitoring protocol set forth in Appendix K is promulgated.
  • Amending the Marine Tank Vessel Loading Operations NESHAP, 40 CFR part 63, subpart Y, to delete the exclusion for marine vessel loading operations at petroleum refineries.
  • Establishing a fenceline monitoring work practice standard to improve the management of fugitive emissions.

b. Refinery MACT 2:

  • Incorporating requirements consistent with those in Refinery NSPS subpart Ja for FCCU including:
  • Requiring the use of 3-hour averages rather than daily averages for parameter operating limits (e.g., depending on the type of control device: Opacity, total power, secondary current, pressure drop, and/or liquid-to-gas ratio).
  • Removing the Refinery NSPS subpart J incremental PM emissions allowance for post combustion devices when burning liquid or solid fuels, and removing the 30 percent opacity limit for units complying with NSPS subpart J.
  • Adding requirements for FCCU controls to include bag leak detectors (BLD) as an option to continuous opacity monitoring system (COMS).
  • Incorporating total power and the secondary current operating limits for electrostatic precipitators (ESP).
  • Requiring daily checks of the air or water pressure to the spray nozzles on jet ejector-type wet scrubber or other type of wet scrubber equipped with atomizing spray nozzles.
  • Requiring FCCU periodic performance testing on a frequency of once every 5 years, as opposed to the current rule, which only requires an initial performance test.
  • Including a correlation equation for the use of oxygen-enriched air for SRU.
  • Allowing SRU subject to Refinery NSPS subpart Ja with a capacity greater than 20 long tons per day (LTD) to comply with Refinery NSPS subpart Ja as a means of complying with Refinery MACT 2.

(4) Other proposed changes include:

  • Removing exemptions from the rule requirements for periods of SSM in order to ensure that the NESHAP are consistent with the court decision in Sierra Club v. EPA, 551 F. 3d 1019 (D.C. Cir. 2008).
  • Clarifying requirements related to open-ended valves or lines.
  • Adding electronic reporting requirements.
  • Updating the General Provisions cross-reference tables.
  • Making technical corrections and clarifications to NSPS subparts J and Ja.

III. What is included in this final rule?

This action finalizes the EPA's determinations pursuant to the RTR provisions of CAA section 112 for the Petroleum Refinery source categories and amends the Petroleum Refinery NESHAP based on those determinations. This action also finalizes other changes to the NESHAP including revising Refinery MACT 1 and 2 pursuant to CAA section 112 (d)(2) and (3), including revising requirements for flares and pressure relief devices (PRD). This action finalizes changes to the SSM provisions to ensure that the subparts are consistent with the court decision in Sierra Club v. EPA, 551 F. 3d 1019 (D.C. Cir. 2008), adds electronic reporting requirements in Refinery MACT 1 and 2; and updates the General Provisions cross-reference tables. Finally, this action finalizes technical corrections and clarifications to Refinery NSPS subparts J and Ja to address issues raised in the reconsideration of these rules.

A. What are the final NESHAP amendments based on the risk review for the Petroleum Refinery source categories?

The EPA is promulgating final amendments to the Petroleum Refinery NESHAP pursuant to CAA section 112(f) that expand the existing Refinery MACT 1 control requirements and extend these requirements to smaller tanks and tanks with lower vapor pressures. Specifically, consistent with the proposal, the EPA is amending Refinery MACT 1 by revising the definition of Group 1 storage vessels to include storage vessels with capacities greater than or equal to 20,000 gallons but less than 40,000 gallons if the maximum true vapor pressure is 1.0 psia or greater and to include storage tanks greater than 40,000 gallons if the maximum true vapor pressure is 0.75 psia or greater. The EPA is also adding a cross-reference to the storage vessel requirements in the Generic MACT (40 CFR part 63, subpart WW and subpart CC), which include requirements for guide pole controls and other fittings as well as inspection requirements. After considering the public comments, the final amendments include minor changes from our proposed requirements to clarify language and correct typographical and referencing errors.

B. What are the final NESHAP amendments based on the technology review for the Petroleum Refinery source categories?

1. Refinery MACT 1

We determined that there are developments in practices, processes and control technologies that warrant revisions to the MACT standards for this source category. Therefore, to satisfy the requirements of CAA section 112(d)(6), we are revising the MACT standards to amend 40 CFR part 63, subpart Y to delete the exclusion for marine vessel loading operations at petroleum refineries. Removing this exclusion will require small marine vessel loading operations (i.e., operations with HAP emissions less than 10/25 tpy) and offshore marine vessel loading operations to use submerged filling based on the cargo filling line requirements in 46 CFR 153.282, as proposed.

We are also finalizing a fenceline monitoring work practice standard to improve the management of fugitive emissions and finalizing EPA Methods 325A and 325B to support the work Start Printed Page 75183practice, with some changes from proposal to address issues raised by commenters. Key revisions include: New provisions for reduced monitoring for facilities with consistently low fenceline concentrations; requirements for alternatives to passive monitoring; revised placement guidance to allow perimeter monitoring within a facility's property boundary provided all sources are encompassed within the monitoring perimeter; reductions in the number of monitors required for subareas and segregated areas; clarifications on monitor placement for internal roadways or other right-of-ways and marine docks; and revised timelines for submitting periodic reports (quarterly rather than semiannually) and implementing the work practice standard (2 years after promulgation rather than 3 years as proposed). We are also revising Refinery MACT 1 storage vessel requirements as described above under the risk review, as proposed.

2. Refinery MACT 2

We determined that there are developments in practices, processes and control technologies that warrant revisions to the MACT standards for this source category. Therefore, to satisfy the requirements of CAA section 112(d)(6), we are revising the Refinery MACT 2 standard for FCCU subject to Refinery NSPS subpart J or those electing to comply with the Refinery NSPS subpart J requirements. As proposed, we are removing the incremental PM limit when burning liquid or solid fuels. We are finalizing a 20-percent opacity operating limit evaluated on a 3-hour average, which differs from the proposal to eliminate the 30-percent opacity limit and instead allow only for a site-specific opacity operating limit or control device parameter monitoring. As proposed, we are finalizing requirements to make Refinery MACT 2 consistent with Refinery NSPS subpart Ja for FCCU by including 3-hour averages rather than daily averages for parameter operating limits, and by including 3-hour averages rather than daily averages for the site-specific opacity operating limit. We are also finalizing requirements, as proposed, for FCCU controls to include adding BLD as an option to COMS, incorporating total power and the secondary current operating limits for ESP and requiring daily checks of the air or water pressure to the spray nozzles on jet ejector-type wet scrubbers or other types of wet scrubbers equipped with atomizing spray nozzles.

Finally, we are finalizing, as proposed, requirements for FCCU periodic performance testing at a frequency of once every 5 years rather than the current requirements for a one-time initial performance test. However, for owners or operators complying with the Refinery NSPS subpart J option (with the 20-percent opacity operating limit discussed above), if the PM emissions are within 80-percent of the PM limit during any periodic performance test (i.e., emissions exceed 0.8 lb PM/1,000 lbs of coke burn-off), the refinery owner or operator must conduct subsequent performance tests on an annual basis. Based on comments received, we are also adding requirements in the final rule for owners or operators of FCCU to conduct a one-time test for HCN emissions from the FCCU concurrent with their first periodic performance test, which must be conducted on or before August 1, 2017 for all FCCU subject to Refinery MACT 2.

For SRU, as proposed, we are finalizing a correlation equation for the use of oxygen-enriched air. Additionally, as proposed, we are finalizing requirements to allow sulfur recovery plants subject to Refinery NSPS subpart Ja with a capacity greater than 20 LTD to comply with Refinery NSPS subpart Ja as a means of complying with Refinery MACT 2.

C. What are the final NESHAP amendments pursuant to section 112(d)(2) & (3) for the Petroleum Refinery source categories?

1. Refinery MACT 1

We are finalizing MACT standards for DCU decoking operations that require that each coke drum be depressured to a closed blowdown system until the coke drum pressure is 2 psig with minor revisions from proposal. Specifically, we are finalizing provisions for existing DCU affected sources to average over a 60-cycle (i.e., 60 batch) basis to comply with the 2 psig limit, rather than the proposed requirement to meet the 2 psig limit on a per venting event basis. In addition, we are finalizing requirements for new DCU affected sources to depressure to 2.0 psig on a per-event, not-to-exceed basis, adding one significant digit to the limit for new DCU affected sources. For both new and existing DCU affected sources, we are finalizing specific provisions for DCU with water overflow design and for double quenching.

We are finalizing operational requirements and the associated monitoring, recordkeeping and reporting requirements for flares used as APCD in Refinery MACT 1 and 2 with revisions to the requirements proposed. Prior to these amendments, Refinery MACT 1 and 2 cross-referenced the General Provisions requirements at 40 CFR 63.11(b). As proposed, this final action replaces the cross reference to the General Provisions and incorporates enhanced flare operational requirements directly into the Refinery MACT regulations. As proposed, the final rule amendments require that refinery flares operate with continuously lit pilot flames at all times. Consistent with our proposal, we are finalizing requirements for flares to operate with no visible emissions and comply with consolidated requirements related to flare tip velocity, but in the final rule these direct emissions limits apply when flare vent gas flow is below the smokeless capacity of the flare rather than at all times. Above the smokeless capacity of the flare, we are establishing a work practice standard related to the visible emissions and velocity limits; these work practice standards are described in more detail in section III.D.1 of this preamble.

We are finalizing new operational requirements related to combustion zone gas properties with revisions from proposal. In response to comments on the proposal, we are finalizing requirements that flares meet a minimum operating limit of 270 BTU/scf NHVcz on a 15-minute average, and are allowing refinery owners or operators to use a corrected heat content of 1,212 BTU/scf for hydrogen to demonstrate compliance with this operating limit. We had proposed two separate sets of limits, one being more stringent if an olefins/hydrogen mixture was present in the waste gas. For each set of limits, we proposed three different alternative combustion zone operating limits: One based on the combustion zone net heat content with no correction for the heat content of hydrogen, one based on the lower flammability limit and one based on the combustibles concentration. We proposed that these limits be determined on a 15-minute “feed-forward” block average approach (i.e., compositional data are collected every 15 minutes, after which adjustments are made). We have included an additional option for refiners to comply where more frequent data are collected (using direct net heating value monitoring) to calculate the combustion limit using net heating value data from the same 15-minute block period. We are simplifying the compliance approach to a single operating limit based only on the combustion zone net heating value (with a hydrogen correction). As proposed, we are requiring refinery owners or operators to characterize the composition of waste gas, assist gas and Start Printed Page 75184fuel to demonstrate compliance with the operational requirements.

As proposed, we are also finalizing in this rule a burden reduction option to use grab sampling every 8 hours rather than continuous vent gas composition or heat content monitors. We are also including, based on public comment, provisions to conduct limited initial sampling and process knowledge to characterize flare gas composition for flares in “dedicated” service as an alternative to collecting grab samples during each specific event. We are finalizing a requirement for daily visible emissions observations as proposed, but, based on public comment, we are allowing owners or operators to use video surveillance cameras to demonstrate compliance with the visible emissions limit as an alternative to the daily visible emissions observations.

For PRD, we are finalizing requirements for monitoring systems that are capable of identifying and recording the time and duration of each pressure release to the atmosphere, as proposed. Certain PRD with low set pressures or low emission potential or in liquid service would not be subject to these monitoring requirements. We are finalizing requirements to minimize or prevent atmospheric releases of HAP through PRD. Instead of the proposed prohibition on such releases, we are finalizing work practice requirements that require both preventive measures as well as root cause analysis and corrective action that will incentivize refinery owners or operators to eliminate the causes of the releases.

We are finalizing requirements for bypass lines with minor revisions from those proposed. Specifically, we are not adopting the proposed requirement to install quantitative flow monitors and thus are leaving in place the requirement to use flow indicators on bypass lines. In addition, we are maintaining the requirements to estimate and report the quantity of organic HAP released. In response to public comment, we are also clarifying changes to remove the proposed reference to air intrusion and specifying that reporting of bypasses is only required when “regulated material” is discharged to the atmosphere as a result of a bypass of a control device.

We are also finalizing revisions to the definition of miscellaneous process vent, as proposed. These revisions include deletion of exclusions associated with episodic releases and vents from in situ sampling systems. As proposed, the final amendments require that these vents must meet the standards applicable to MPV.

2. Refinery MACT 2

For CRU vents, we are finalizing the vessel pressure limit exclusion of 5 psig to apply only to passive depressurization, as proposed.

D. What are the final NESHAP amendments addressing emissions during periods of SSM?

We are finalizing, as proposed, changes to Refinery MACT 1 and 2 to eliminate the SSM exemption. Consistent with Sierra Club v. EPA, 551 F. 3d 1019 (D.C. Cir. 2008), the EPA has established standards in this rule that apply at all times. EPA is revising Table 6 of subpart CC of 40 CFR part 63 and Table 44 to subpart UUU of 40 CFR part 63 (the General Provisions Applicability Tables) to change several references related to requirements that apply during periods of SSM. We also are eliminating or revising certain recordkeeping and reporting requirements related to the eliminated SSM exemptions. We also are removing or modifying inappropriate, unnecessary or redundant language in the absence of the SSM exemption. Further, for certain emission sources in both MACT 1 and 2, we are establishing standards to address emissions during these periods. These are described below.

1. Refinery MACT 1

We are finalizing a work practice standard for PRD that requires refinery owners or operators to establish prevention measures for each PRD in organic HAP service. Under the work practice standard, where a direct release occurs, the refinery is required to perform root cause analysis and implement corrective action. The work practice standard also limits the number of events that a PRD may release to the atmosphere during a 3-year period, as explained further in the section IV.D. of this preamble.

We are also finalizing a work practice standard for emergency flaring events that requires refinery owners or operators to establish prevention measures, including the development of a flare management plan (FMP), and perform root cause analysis and implement corrective action following flaring events during which the velocity of waste gas going to the flare or visible emissions limits (i.e., opacity) at the flare tip are exceeded, and to limit the number of these events allowed in a 3-year period, as explained further in section IV.D. of this preamble. Both of these work practice standards are consistent with the EPA's goal to improve the effectiveness of the rules. These requirements will provide a strong incentive for facilities, over time, to better operate their processes to prevent PRD and flare releases.

We are also finalizing requirements for opening process equipment to the atmosphere during maintenance events after draining and purging to a closed system, provided the hydrocarbon content is less than or equal to 10-percent of the lower explosive limit (LEL). For those situations where 10-percent LEL cannot be demonstrated, the equipment may be opened and vented to the atmosphere if the pressure is less than or equal to 5 psig, provided there is no active purging of the equipment to the atmosphere until the LEL criterion is met. This 5 psig allowance is only available during shutdown. We are also providing additional allowances for situations where it is not technically feasible to depressurize a control system where there is no more than 72 lbs VOC per day vented to the atmosphere, consistent with our Group 1 applicability cutoff for control of process vents, or for catalyst changeout activities where hydrotreater pyrophoric catalyst must be purged. Provisions to demonstrate that process equipment is opened only after the LEL, pressure or mass in the vessel requirement is met includes documenting the procedures for equipment openings and procedures for verifying that the openings meet the specific, above-discussed requirements using site-specific procedures used to de-inventory equipment for safety purposes (i.e., hot work or vessel entry procedures).

2. Refinery MACT 2

The Refinery MACT 2 standards regulate all HAP emissions from the three refinery process vents subject to Refinery MACT 2. For FCCU, the standard specifies a CO limit as a surrogate for organic HAP and specifies a PM limit (or Ni limit) as a surrogate for metal HAP. Compliance with the organic HAP emissions limit is demonstrated using a continuous CO monitor; compliance with the metal HAP emissions limit is demonstrated using either COMS or control device parameter monitoring systems (CPMS). At proposal, with the removal of the exemptions in the Refinery MACT 2 rule for periods of startup and shutdown, we recognized the need for alternative standards during some startup and shutdown situations, and we proposed alternative requirements.

For this final rule, we are including a 1-percent minimum oxygen limit as an alternative to the 500 ppmv hourly CO limit during FCCU startup for partial Start Printed Page 75185burn FCCU with CO boilers, as proposed. We are extending that alternative limit to all FCCU and extending it to apply during shutdown.

We are not finalizing the proposed alternative opacity limit for FCCU during startup. Instead, based on public comments received, we are finalizing an alternative minimum cyclone face velocity limit as a means to demonstrate compliance with the PM limit during both startup and shutdown, regardless of the type of FCCU and its control device. We are finalizing alternative standards for sulfur recovery plant (SRP) incinerator temperature and excess oxygen limits during SRP shutdown, as proposed, and we are extending the proposed alternative standards to startup as well.

E. What other revisions to the NESHAP and NSPS are being promulgated?

We are finalizing technical amendments to NSPS subparts J and Ja with limited changes from what we proposed. First, in response to comments, we are revising the NSPS requirements that a flow sensor have a “measurement sensitivity” of no more than 5-percent of the flow rate to an “accuracy” requirement that the flow sensor have an accuracy of 5-percent of the flow rate. This change will make the requirements more clear and consistent between the flow meter requirements in the NSPS and the MACT standards since it is the same flow meter subject to these requirements. We are also revising flare flow rate accuracy requirements in Refinery NSPS subpart Ja to make them consistent with those we are finalizing in Refinery MACT 1. Finally, we are revising 40 CFR 60.101a(b) to begin as “Except for flares and delayed coking units . . .” to correct an inadvertent error. We proposed revisions to this sentence solely to allow sources subject to Refinery NSPS subpart J to comply with the provisions in Refinery NSPS subpart Ja instead. However, the words “and delayed coking units” were inadvertently omitted from the initial part of the sentence. Thus, as intended, we are finalizing revisions to this sentence to allow sources subject to Refinery NSPS subpart J to comply with the provisions in Refinery NSPS subpart Ja.

F. What are the requirements for submission of performance test data to the EPA?

As proposed, the EPA is taking a step to increase the ease and efficiency of data submittal and data accessibility. Specifically, the EPA is finalizing the requirement for owners or operators of Petroleum Refinery facilities to submit electronic copies of certain required performance test reports through the EPA's Central Data Exchange (CDX) using the Compliance and Emissions Data Reporting Interface (CEDRI). The EPA believes that the electronic submittal of the reports addressed in this rulemaking will increase the usefulness of the data contained in those reports, is in keeping with current trends in data availability, will further assist in the protection of public health and the environment and will ultimately result in less burden on the regulated community. Electronic reporting can also eliminate paper-based, manual processes, thereby saving time and resources, simplifying data entry, eliminating redundancies, minimizing data reporting errors and providing data quickly and accurately to the affected facilities, air agencies, the EPA and the public.

As mentioned in the preamble of the proposal, the EPA Web site that stores the submitted electronic data, WebFIRE, will be easily accessible to everyone and will provide a user-friendly interface that any stakeholder could access. By making the records, data and reports addressed in this rulemaking readily available, the EPA, the regulated community and the public will benefit when the EPA conducts its CAA-required technology and risk-based reviews. As a result of having reports readily accessible, our ability to carry out comprehensive reviews will be increased and achieved within a shorter period of time.

We anticipate fewer or less substantial information collection requests (ICRs) in conjunction with prospective CAA-required technology and risk-based reviews may be needed. We expect this to result in a decrease in time spent by industry to respond to data collection requests. We also expect the ICRs to contain less extensive stack testing provisions, as we will already have stack test data electronically. Reduced testing requirements would be a cost savings to industry. The EPA should also be able to conduct these required reviews more quickly. While the regulated community may benefit from a reduced burden of ICRs, the general public benefits from the agency's ability to provide these required reviews more quickly, resulting in increased public health and environmental protection.

Air agencies could benefit from more streamlined and automated review of the electronically submitted data. Having reports and associated data in electronic format will facilitate review through the use of software “search” options, as well as the downloading and analyzing of data in spreadsheet format. The ability to access and review air emission report information electronically will assist air agencies to more quickly and accurately determine compliance with the applicable regulations, potentially allowing a faster response to violations which could minimize harmful air emissions. This benefits both air agencies and the general public.

For a more thorough discussion of electronic reporting required by this rule, see the discussion in the preamble of the proposal. In summary, in addition to supporting regulation development, control strategy development, and other air pollution control activities, having an electronic database populated with performance test data will save industry, air agencies, and the EPA significant time, money, and effort while improving the quality of emission inventories, air quality regulations, and enhancing the public's access to this important information.

G. What are the effective and compliance dates of the NESHAP and NSPS?

The final amendments to the NESHAP and NSPS in this action are effective on February 1, 2016. As proposed, new sources must comply with these requirements by the effective date of the final rule or upon startup, whichever is later.

As proposed, existing sources are required to comply with the final DCU and CRU requirements no later than 3 years after the effective date of the final rule. Similarly, as proposed, owners or operators are required to comply with the new operating and monitoring requirements for existing flares no later than 3 years after the effective date of the final rule.

We proposed to provide 3 years from the effective date of the final rule for refinery owners or operators to install and begin monitoring (collecting samples) around the fenceline of their existing facility. If refinery owners and operators determined that a site-specific monitoring plan was needed, they would also need to submit and receive approval for such a plan during the 3-year compliance period. Based on information submitted during the comment period, we are finalizing requirements that refinery owners or operators begin collecting samples around the fenceline within 2 years of the effective date of the final rule. Based on information submitted during the comment period, 1 year is sufficient time to identify proper monitoring locations and to install the required monitoring stations around the facility Start Printed Page 75186fenceline. However, owners or operators may need additional monitoring systems to account for near-field interfering sources (NFS), for which the development and approval of a site-specific fenceline monitoring plan is required. We expect that the site-specific fenceline monitoring plans can take an additional year to develop, submit and obtain approval. Consequently, we are providing 2 years from the effective date of the final rule for refinery owners or operators to install and begin collecting samples around the fenceline of their facility.

As proposed, we are requiring that existing sources comply with the submerged filling requirement for marine vessel loading on the effective date of the final rule.

As proposed, we are providing 18 months after the effective date of the final rule to conduct required performance tests and comply with any revised operating limits for FCCU.

We proposed to require refinery owners or operators to comply with the revisions to the SSM provisions of Refinery MACT 1 and 2 on the effective date of the final rule. As proposed, this final rule requires refinery owners or operators to comply with the limits in Refinery MACT 2 or the alternative limits in this final rule during startup and shutdown for FCCU and SRU on the effective date of the final rule.

The flare work practice standards for high-load flaring events (events exceeding the smokeless capacity of the flare) require development of FMP (or revision of an existing plan) to specifically consider emergency shutdown and other high load events. In this FMP, refinery owners or operators must consider measures that can be implemented to reduce the frequency and magnitude of these high-load flaring events. This may include installation of a flare gas recovery system. Additionally, the work practice standards will require refinery owners or operators to identify and implement measures that may involve process changes. Therefore, we are establishing a compliance date of 3 years from the effective date of the final rule for refinery owners or operators to comply with the work practice standards for high load flaring events. We also note that this compliance period is consistent with the compliance time provided for the flare operating limits.

For atmospheric PRD in HAP service we are establishing a work practice standard that requires a process hazard analysis and implementation of a minimum of three redundant measures to prevent atmospheric releases. Alternately, refinery owners or operators may elect to install closed vent systems to route these PRD to a flare, drain (for liquid thermal relief valves) or other control system. We anticipate that sources will need to identify the most appropriate preventive measures or control approach; design, install and test the system; install necessary process instrumentation and safety systems; and may need to time installations with equipment shutdown or maintenance outages. Therefore, we have established a compliance date of 3 years from the effective date of the final rule for refinery owners or operators to comply with the work practice standards for atmospheric PRD.

As proposed, we are requiring compliance with the electronic reporting provisions for performance tests conducted for Refinery MACT 1 and 2 on the effective date of the final rule.

Finally, we are finalizing additional requirements for storage vessels under CAA sections 112(d)(6) and (f)(2) with a compliance date 90 days after the effective date of the final rule, as proposed.

H. What materials are being incorporated by reference?

In this final rule, the EPA is including regulatory text that includes incorporation by reference. In accordance with requirements of 1 CFR 51.5, the EPA is incorporating by reference the following documents described in the amendments to 40 CFR 63.14:

  • ASTM D1945-03 (Reapproved 2010), Standard Test Method for Analysis of Natural Gas by Gas Chromatography, (Approved January 1, 2010).
  • ASTM D1945-14, Standard Test Method for Analysis of Natural Gas by Gas Chromatography.
  • ASTM D6196-03 (Reapproved 2009), Standard Practice for Selection of Sorbents, Sampling, and Thermal Desorption Analysis Procedures for Volatile Organic Compounds in Air, (Approved March 1, 2009).
  • ASTM D6348-03 (Reapproved 2010), Standard Test Method for Determination of Gaseous Compounds by Extractive Direct Interface Fourier Transform Infrared (FTIR) Spectroscopy, including Annexes A1 through A8, (Approved October 1, 2010).
  • ASTM D6348-12e1, Standard Test Method for Determination of Gaseous Compounds by Extractive Direct Interface Fourier Transform Infrared (FTIR) Spectroscopy.
  • ASTM D6420-99 (Reapproved 2010), Standard Test Method for Determination of Gaseous Organic Compounds by Direct Interface Gas Chromatography-Mass Spectrometry.
  • ASTM UOP539-12, Refinery Gas Analysis by GC.
  • BS EN 14662-4:2005, Ambient air quality—Standard method for the measurement of benzene concentrations—Part 4: Diffusive sampling followed by thermal desorption and gas chromatography, June 27, 2005.
  • EPA-454/B-08-002, Quality Assurance Handbook for Air Pollution Measurement Systems, Volume IV: Meteorological Measurements, Version 2.0 (Final), March 2008.
  • EPA-454/R-99-005, Meteorological Monitoring Guidance for Regulatory Modeling Applications, February 2000.
  • ISO 16017-2:2003(E): Indoor, ambient and workplace air—Sampling and analysis of volatile organic compounds by sorbent tube/thermal desorption/capillary gas chromatography—Part 2: Diffusive sampling, May 15, 2003.
  • Air Stripping Method (Modified El Paso Method) for Determination of Volatile Organic Compound Emissions from Water Sources” Revision Number One, dated January 2003, Sampling Procedures Manual, Appendix P: Cooling Tower Monitoring, prepared by Texas Commission on Environmental Quality, January 31, 2003.[4]

The EPA has made, and will continue to make, these documents available electronically through www.regulations.gov and/or in hard copy at the appropriate EPA office (see the ADDRESSES section of this preamble for more information).

IV. What is the rationale for our final decisions and amendments to the Petroleum Refinery NESHAP and NSPS?

A. Residual Risk Review for the Petroleum Refinery Source Categories

1. What did we propose pursuant to CAA section 112(f) for the Petroleum Refinery source categories?

The results of our residual risk review for the Petroleum Refinery source categories were published in the June 30, 2014 proposal at (79 FR 36934 through 36942), and included assessment of chronic and acute inhalation risk, as well as multipathway and environmental risk, to inform our decisions regarding acceptability and ample margin of safety. The results indicated that both the actual and Start Printed Page 75187allowable inhalation cancer risks to the individual most exposed are no greater than approximately 100-in-1 million, which is the presumptive limit of acceptability. In addition, the maximum chronic non-cancer target organ-specific hazard index (TOSHI) due to inhalation exposures was less than 1. The evaluation of acute non-cancer risks, which was conservative, showed acute risks below a level of concern. Based on the results of the refined site-specific multipathway analysis, we also concluded that the ingestion cancer risk to the individual most exposed through ingestion is considerably less than 100-in-1 million. In determining risk acceptability, we also evaluated population impacts because of the large number of people living near facilities in the source category. We estimated that 5-million people are exposed to increased cancer risks of greater than 1-in-1 million and 100,000 people are exposed to increased cancer risks of greater than 10-in-1 million, but, as noted previously, no individual is exposed to increased cancer risks of greater than 100-in-1 million. Considering the above information, we proposed that the risks remaining after implementation of the existing NESHAP for the Refinery MACT 1 and 2 source categories is acceptable. However, we noted that the risks based on allowable emissions are at the presumptive limit of acceptable risk, and that a large number of people are exposed to risks of greater than 1-in-1 million, and we solicited comment on whether EPA should conclude that the risk was unacceptable based on the health information before the Agency. We also proposed that the original Refinery MACT 1 and 2 MACT standards, along with the proposed requirements for storage vessels, provide an ample margin of safety to protect public health. Finally, we proposed that it is not necessary to set a more stringent standard to prevent, taking into consideration costs, energy, safety, and other relevant factors, an adverse environmental effect.

2. How did the risk review change for the Petroleum Refinery source categories?

As part of the final risk assessment, we conducted a screening level analysis of how the information we received during the public comment period, along with the changes we are making to the proposed rule, would change our proposed risk estimates (More details can be found in the “Final Residual Risk Assessment for the Petroleum Refining Source Sector”, Docket ID No. EPA-HQ-OAR-2010-0682).

First, we received approximately 20 emissions inventory updates for specific facilities. These updates included revised emission estimates, revised release latitude/longitude locations and other release characteristic revisions. The updates provided evidence that the quantity of HAP emitted at these specific facilities is lower than considered in the risk modeling for the proposed rule. Our assessment of the effects of these changes suggests that the cancer maximum individual risk (MIR) based on actual emissions may be closer to 40-in-1 million, as opposed to 60-in-1 million, as projected at proposal. We did not quantify the reductions in chronic or acute non-cancer risks from these updates. We calculated allowable emissions using the Refinery Emissions Model (REM), which estimates emissions based on each refinery's capacities and throughputs [See discussion at 79 FR 36888, June 30, 2014.] The allowable emission estimates for point and fugitive sources were not specific to a particular latitude/longitude location so we assumed them to release from the centroid of the facility. Therefore, the predicted cancer MIR of approximately 100-in-1 million based on allowable emissions and reported in the proposal risk characterization does not change based on the submitted emissions revisions. We did not quantify changes to other actual risk metrics as part of the screening level analysis (i.e., incidence, populations in risk bins, multipathway and ecological analyses), but we would expect some minor reductions from those presented in the proposed risk characterization.

Second, we are establishing work practice standards in the final rule for PRD releases and emergency flaring events, which under the proposed rule would not have been allowed. Thus, because we did not consider such non-routine emissions under our risk evaluation for the proposed rule, we performed a screening assessment of risk associated with these non-routine events for the final rule. [We provide further details on the screening approach in “Final Residual Risk Assessment for the Petroleum Refining Source Sector” in Docket ID No. EPA-HQ-OAR-2010-0682.] We extracted information on these events from the 2011 Petroleum Refinery ICR data that included the process unit identification, mass of emissions, duration of release, and description of the incident. We identified the highest HAP mass releases for both PRDs and flares from these non-routine events. We assumed these HAP emission releases could occur at any facility in the source category. Our analysis suggests that these HAP emissions could increase the MIR based on actual emissions by as much as 2-in-1 million. Because the PRD and flaring events were the worst case HAP mass emission release events reported in the 2011 Refinery ICR for the source category, we are assuming that actual and allowable risks are no different for these events (i.e., a MIR of 2-in-1 million). A MIR increase of 2-in-1 million attributable to these events, added to our previous estimate for allowable risk at proposal will not appreciably change our proposed determination that the MIR based on allowable emissions are approximately 100-in-1 million. We note that the MIR estimate attributable to these non-routine PRD and flaring events was estimated using a conservative, screening-level assessment, while the MIR estimate at proposal was based on a refined risk assessment. By adding a screening estimate to a refined risk estimate, we are merely defining an upper limit that we expect the combined risks from both the routine and non-routine emissions to be. Similarly, we estimate chronic non-cancer hazard index (HI) values attributable to the additional exposures resulting from non-routine flaring and PRD HAP emissions to be well below 1 (HIimmune-system of 0.007) such that there is no appreciable change in the maximum chronic non-cancer HI of 0.9 estimated at proposal for routine emissions, which was based on neurological effects.

The screening analysis projects that the maximum predicted acute non-cancer risk from non-routine PRD and flare emissions results in a hazard quotient (HQ) based on a recommended reference exposure level limit (REL) of up to 14 from benzene emissions. While the analysis shows that there is a potential for HQs exceeding 1 for benzene, because of the many uncertainties and conservative nature of this screening analysis, the likelihood of such exposure and risk are low. At proposal, we projected a HQ based on the REL for benzene of up to 2 from routine emissions. If we conservatively combine the routine and non-routine emissions analyses, we would expect the potential for HQs based on the REL for benzene to have the potential to increase above 2. However, as projected at proposal, we estimate that the acute HQs calculated using acute exposure guideline levels (AEGL) and emergency response and planning guidelines (ERPG) values for all pollutants including benzene would still be well Start Printed Page 75188below 1 considering both routine and non-routine emissions.

Considering all of these factors, we do not project risks to be significantly different from what we proposed. Based on the risk analysis, as informed by the screening level analysis based on information obtained during the comment period, we are finalizing our determination that the risk remaining after promulgation of the NESHAP is acceptable.

3. What key comments did we receive on the risk review and what are our responses?

We received numerous comments on the residual risk assessment analyses and results. We summarize the key comments received below, along with our responses. A complete summary of all public comments received and our responses are in the “Response to Comment” Document in the public docket (Docket ID No. EPA-HQ-OAR-2010-0682).

Comment: Several commenters agreed that the EPA has correctly concluded that the proposed rule requirements protect the public with an ample margin of safety from refinery emissions. Other commenters noted that EPA found residual risks remaining after implementation of the MACT standards to be acceptable, and in light of the acceptability determination argued that the proposed changes to the rule are not justified. The commenters noted that the EPA's detailed emissions inventory assessment and risk modeling results demonstrated that, at every U.S. refinery, category-specific risks are below the EPA's presumptive limit of acceptable risk (i.e., cancer risk of less than 100-in-1 million).

Other commenters stated the EPA's risk estimates are understated and that the EPA should reduce the benchmark of what it considers acceptable lifetime cancer risk instead of the upper limit of 100-in-1 million. One commenter provided an extensive critique of the cancer, chronic and acute affects levels used in the risk assessment and recommended that the EPA use California Office of Environmental Health Hazard Assessment's (OEHHA) new toxicity values for several chemicals. The commenter provided some references for the approaches used to derive the California values. The commenter also asserted that risks would be unacceptable had these more protective values been used in the risk assessment. Some commenters stated the risks from petroleum refinery emissions are underestimated because the EPA did not but should have included interaction of multiple pollutants, accounted for exposure to multiple sources, and assessed the cumulative risks from facility-wide emissions and multiple nearby sources impacting an area.

Response: The approximately 100-in-1 million benchmark was established in the Benzene NESHAP (54 FR 38044, September 14, 1989), which Congress specifically referenced in CAA section 112(f)(2)(B). While this presumptive level provides a benchmark for judging the acceptability of MIR, it is important to recognize that it does not constitute a rigid line for making that determination. The EPA considers the specific uncertainties of the emissions, health effects and risk information for the source category in question when deciding whether the risk posed by that source category is acceptable. In addition, the source category-specific decision of what constitutes an acceptable level of risk is a holistic one; that is, the EPA considers all potential health impacts—chronic and acute, cancer and non-cancer, and multipathway—along with their uncertainties, when determining whether the source category presents an unacceptable risk.

Regarding the comment that in light of the acceptability determination the proposed changes to the rule are not justified, we note that we also are required to ensure that the standards provide an ample margin of safety to protect public health. That analysis is separate from the acceptability analysis, and the determination of acceptability does not automatically lead us to conclude that the standards provide an ample margin of safety to protect public health.

Regarding the comments that the EPA should use the new California OEHHA values, we disagree. The EPA's chemical-specific toxicity values are derived using risk assessment guidelines and approaches that are well established and vetted through the scientific community, and follow rigorous peer review processes.[5] The RTR program gives preference to the EPA values for use in risk assessments and uses other values, as appropriate, when those values are derived with methods and peer review processes consistent with those followed by the EPA. The approach for selecting appropriate toxicity values for use in the RTR Program has been endorsed by the Science Advisory Board (SAB).6

The EPA scientists reviewed the information provided by the commenter regarding the California values and concluded that further information is needed to evaluate the scientific basis and rationale for the recent changes in California OEHHA risk assessment methods. The EPA will work on gathering the necessary information to conduct an evaluation of the scientific merit and the appropriateness of the use of California OEHHA's new toxicity values in the agency decisions. Until the EPA has completed its evaluation, it is premature to determine what role these values might play in the RTR process. Therefore, the EPA did not use the new California OEHHA toxicity values as part of this current action. For more detailed responses regarding appropriate reference values for specific pollutants, see the “Response to Comment” document in the public docket (Docket ID No. EPA-HQ-OAR-2010-0682).

Concerning comments that we should consider aggregate risks from multiple pollutants and sources, we note that we have done this to the extent it is appropriate to do so. We modeled whole-facility risks for both chronic cancer and non-cancer impacts to understand the risk contribution of the sources within the Petroleum Refinery source categories. The individual cancer risks for the source categories were aggregated for all carcinogens. In assessing non-cancer hazard from chronic exposures to pollutants that have similar modes of action or (where this information is absent) that affect the same target organ, we summed the HQs. This process creates, for each target organ, a TOSHI, defined as the sum of HQs for individual HAP that affect the same organ or organ system. Whole-facility risks were estimated based on the 2011 ICR emissions data obtained from facilities, which included emissions from all sources at the refinery, not just Refinery MACT 1 and 2 emission sources (e.g., emissions were included for combustion units and units subject to the Hazardous Organic NESHAP, if present at the refinery). We disagree with the commenter's assertion that additional quantitative assessment of risks from sources outside the source category is required under the statute. The statute requires the EPA to provide the quantitative risk information necessary to inform RTR regulatory decisions, and to this end, the EPA conducted a comprehensive assessment of the risks associated with exposure to the HAP emitted by the source category and supplemented that with additional Start Printed Page 75189information available about other possible concurrent and relevant risks.

Further, the risk assessment modeling accounts for the effects of multiple facilities that may be in close proximity when estimating concentration and risk impacts at each block centroid. When evaluating the risks associated with a particular source category, we combined the impacts of all facilities within the same source category and assessed chronic exposure and risk for all census blocks with at least one resident (i.e., locations where people may reasonably be assumed to reside). The MIR considers the combined impacts of all sources in the category that may be in close proximity (i.e., cumulative impact of all refineries).

Comment: Several commenters stated that the EPA underestimated exposure because emissions are underreported and underestimated. The commenters noted that for the risk assessment for the refineries rule, the EPA evaluated (1) the emissions reported to the agency pursuant to the 2011 Petroleum Refinery ICR as sources' “actual” emissions, and (2) the emissions the EPA estimates that the existing standards currently allow sources to emit using the REM, which it describes as “allowable” emissions. According to the commenters, both the EPA's “actual” and “allowable” emissions data sets are incomplete and undercount emissions, causing the EPA to significantly underestimate the resulting risk in its risk analysis. For example, the commenters noted that the EPA assumed the flare destruction efficiency to be 98 percent, while the EPA's own estimates suggest flare efficiency is 93.9 percent. The commenters also noted that the EPA has further understated risks by ignoring emissions during unplanned SSM events and by ignoring HAP for which no reference values are established. One commenter cited the TCEQ Emissions Event Database as evidence that SSM emissions are a severe public health problem because data show that nearly 1 million pounds of HAP are reported from Texas refineries between 2009 and 2013. According to these commenters, the EPA needs to adopt standards that provide greater protection, including protection from the risks of accidents.

Response: We used the best and most robust facility-specific HAP emissions inventory available to us, which was the 2011 ICR, in performing the analysis for the proposed rule. We conducted a thorough and exhaustive review of the data submitted through the ICR and we followed up on source-specific information on a facility-by-facility basis, as documented in the “Emissions Data Quality Memorandum and Development of the Risk Model Input File” (see Docket ID No. EPA-HQ-OAR-2010-0682-0076). In addition, we took steps ahead of issuing the 2011 ICR to make sure that facilities could, as accurately as practicable, estimate their HAP emissions for purposes of responding to the inventory portion of that ICR. We prepared a Refinery Protocol to provide guidance to refinery owners or operators to use the best available, site-specific data when developing their emissions inventory, to ensure all emission sources are included in the inventory, and to have a consistent set of emission factors that all respondents use if no site-specific emissions data were available. If site-specific emissions data were available, sites were to use these data preferentially over the default factors. We developed the default factors provided in the protocol from the best data available at the time.

The ICR-submitted information for allowable emissions did not include emission estimates for all HAP and all emission sources. Consequently, we used the REM to estimate allowable emissions. The REM relies on model plants that vary based on throughput capacity. Each model plant contains process-specific default emission factors, adjusted for compliance with the Refinery MACT 1 and 2 emission standards.

We agree with the commenters that studies have shown that many refinery flares are operating less efficiently than 98 percent. Prior to proposing this rule, we conducted a flare ad hoc peer review to advise the EPA on factors affecting flare performance (see discussion in the June 30, 2014, proposal at 79 FR 36905). However, we disagree with the commenters that the risk analysis should consider this level of performance since the existing MACT standard does not allow it. For purposes of the risk analysis, we evaluate whether it is necessary to tighten the existing MACT standard in order to provide an ample margin of safety. Thus, in reviewing whether the existing standards provide an ample margin of safety, we review the level of emissions the MACT standards allow. In the present case, we considered the level of performance assumed in establishing the MACT standard for purposes of determining whether the MACT standard provides an ample margin of safety. However, we did recognize that facilities were experiencing performance issues with flares and that many flares were not meeting the assumed performance level at the time we promulgated the MACT standard. Thus, we proposed, and are finalizing, revisions to the flare operating requirements to ensure that the flares meet the required performance level. These provisions are consistent with the EPA's goals to improve the effectiveness of our rules.

Similarly, we do not include startup, shutdown (including maintenance events) and malfunction emissions that are not allowed under the standard as part of our evaluation of whether the standards provide an ample margin of safety. Regarding the HAP emissions from SSM events that the commenter is concerned with, we note that our review of the TCEQ incident database indicates that many of the large reported release events were of SO2 emissions and only a few had significant HAP emissions.

Because in the final rule we are establishing work practice standards for PRD and emergency flaring events, we performed a screening-level risk analysis to address changes in facility HAP emission releases due to these events. Details on this analysis are presented in the final risk report for the source category (For more details see Appendix 13 of the “Final Residual Risk Assessment for the Petroleum Refining Source Sector,” Docket ID No. EPA-HQ-OAR-2010-0682).

As for HAP with no reference value, the SAB addressed this issue in its May 7, 2010, response to the EPA Administrator. In that response, the SAB Panel recommended that, for HAP that do not have dose-response values from the EPA's list, the EPA should consider and use, as appropriate, additional sources for such values that have undergone adequate and rigorous scientific peer review. The SAB panel further recommended that the inclusion of additional sources of dose-response values into the EPA's list should be adequately documented in a transparent manner in any residual risk assessment case study. We agree with this approach and have considered other sources of dose-response data when conducting our risk determinations under RTR. However, in some instances no sources of information beyond the EPA's list are available. Compounds without health benchmarks are typically those without significant health effects compared to compounds with health benchmarks, and in such cases we assume these compounds will have a negligible contribution to the overall health risks from the source category. A tabular summary of HAPs that have dose response values for which an exposure assessment was conducted is presented in Table 3.1-1 of the “Final Residual Risk Assessment for the Petroleum Refining Source Sector”, Docket ID No. EPA-HQ-OAR-2010-0682.Start Printed Page 75190

Comment: A few commenters asserted that the EPA should decide that it is unjust and inconsistent with the CAA's health protection purpose to allow the high health risks caused by refineries to fall disproportionately on communities of color and lower income communities who are least equipped to deal with the resulting health effects. Because of that disparity, the commenter stated that the EPA should recognize that the risks found are unacceptable and set stronger national standards for all exposed Americans.

Response: For this rulemaking, the EPA conducted both pre- and post-control risk-based assessments with analysis of various socio-economic factors for populations living near petroleum refineries (see Docket ID Nos. EPA-HQ-OAR-2010-0682-0226 and -0227) and determined that there are more African-Americans, Other and multiracial groups, Hispanics, low-income individuals, and individuals with less than a high school diploma compared to national averages. In determining the need for tighter residual risk standards, the EPA strives to limit to no higher than 100-in-1 million the estimated cancer risk for persons living near a plant if exposed to the maximum pollutant concentration for 70 years and to protect the greatest number of persons to an individual lifetime risk of no higher than 1-in-1 million. Although we consider the risk for all people regardless of racial or socioeconomic status, communities near petroleum refineries will particularly benefit from the risk reductions associated with this rule. In particular, as discussed later, the fenceline monitoring work practice standard will be a further improvement in the way fugitive emissions are managed and will provide an extra measure of protection for surrounding communities.

4. What is the rationale for our final decisions for the risk review?

As described in section IV.A.2 of this preamble, we performed a screening-level analysis to assess the risks associated with inventory updates we received for specific facilities and with emissions events that were previously not included in the risk assessment because the proposed rule did not allow them. Because we are finalizing work practice standards to regulate emission events associated with PRD releases and emergency flaring, we considered the effect these work practice standards would have on risks. As discussed in section IV.A.2 of this preamble, we project that accounting for these emergency events in the baseline risks after implementation of the MACT standards does not appreciably change the risks, and at most, could increase the proposed rule estimate of MIR by approximately 2-in-1 million. Therefore, we would project that any controls applied to these emergency events, including the work practice standards for PRDs and emergency flaring in this final rule, would not appreciably change the proposed post-control risks. Although we would anticipate minimal additional risk reductions, we reviewed more stringent alternatives to the work practice standards for PRD releases and emergency flaring events included in this final rule, and we found that the costs of increasing flare capacity to control all PRD releases and to eliminate all visible emissions during emergency flaring were too high. We estimate the capital costs of applying the velocity and visible emissions limit at all times would be approximately $3 billion, and we estimate that the costs of controlling all PRD releases with flares would be approximately $300 million. [See the discussion in the “Flare Control Option Impacts for Final Refinery Sector Rule”, Docket ID No. EPA-HQ-OAR-2010-0682 and the PRD work practice standard discussion in section IV.C of this preamble.] Further, we did not receive comments on additional control technologies that we should have considered for other emission sources (e.g., tanks, DCUs) beyond those considered and described at proposal. Consequently, as discussed in section IV.A.2, we conclude that the risks from the Petroleum Refinery source categories are acceptable and that, with the additional requirements for storage vessels that we are finalizing, as proposed, the Refinery MACT 1 and 2 rules provide an ample margin of safety to protect public health. We also maintain, based on the rationale presented in the preamble to the proposed rule, that the current standards prevent, taking into consideration costs, energy, safety and other relevant factors, an adverse environmental effect.

B. Technology Review for the Petroleum Refinery Source Categories

1. What did we propose pursuant to CAA section 112(d)(6) for the Refinery MACT 1 (40 CFR part 63, subpart CC) source category?

The results of our technology review for the Petroleum Refinery source categories were published in the June 30, 2014, proposal at (79 FR 36913 through 36928). The technology review was conducted for both MACT source categories as described below.

a. Refinery MACT 1

Refinery MACT 1 sources include MPV, storage vessels, equipment leaks, gasoline loading racks, marine vessel loading operations, cooling towers/heat exchange systems and wastewater. Based on technology reviews for the sources described above, we proposed that it was not necessary to revise Refinery MACT 1 requirements for MPV, gasoline loading racks, cooling towers/heat exchange systems, and wastewater. For storage vessels, we proposed revisions pursuant to the technology review. Specifically, we proposed to cross-reference the storage vessel requirements in the Generic MACT (40 CFR part 63, subpart WW) to require controls on floating roof fittings (e.g., guidepoles, ladder wells and access hatches) and to revise the definition of Group 1 storage vessels to include smaller tanks with lower vapor pressures. For equipment leaks, we proposed to allow refineries to meet LDAR requirements in Refinery MACT 1 by monitoring for leaks via optical gas imaging in place of the EPA Method 21, using monitoring requirements to be specified in a not-yet-proposed appendix K to 40 CFR part 60. For marine vessel loading, we proposed to amend the Marine Tank Vessel Loading Operations MACT standards (40 CFR part 63, subpart Y) to require small marine vessel loading operations (i.e., operations with HAP emissions less than 10/25 tpy) and offshore marine vessel loading operations at petroleum refineries to use submerged filling based on the cargo filling line requirements in 46 CFR 153.282.

We also proposed an additional work practice standard under the technology review to manage fugitive emissions from the entire petroleum refinery through a fenceline monitoring and corrective action standard. As part of the work practice standard, we specified the monitoring technology and approach that must be used, and we developed a fenceline benzene concentration action level above which refinery owners or operators would be required to implement corrective action to reduce their fenceline concentration to below this action level. The action level we proposed was consistent with the emissions projected from fugitive sources compliant with the provisions of the refinery MACT standards as modified by the additional controls proposed for storage vessels.

b. Refinery MACT 2

The Refinery MACT 2 source category regulates HAP emissions from FCCU, CRU and SRU process vents. We Start Printed Page 75191proposed to revise Refinery MACT 2 to incorporate the developments in monitoring practices and control technologies reflected in Refinery NSPS subpart Ja (73 FR 35838). This included proposing to incorporate the Refinery NSPS subpart Ja PM limit for new FCCU sources and to revise the monitoring provisions in Refinery MACT 2 to require all FCCU sources to meet operating limits consistent with the requirements in Refinery NSPS subpart Ja. The existing MACT standard provided that a refiner could demonstrate compliance with the PM limit in the MACT by meeting the 30-percent opacity limit requirement of Refinery NSPS subpart J; we proposed to eliminate that provision and instead establish control device operating limits or site-specific opacity limits similar to those required in Refinery NSPS subpart Ja. We also proposed to incorporate the use of 3-hour averages rather than daily averages for monitoring data to demonstrate compliance with the FCCU site-specific opacity and Ni operating limits. We proposed additional control device-specific monitoring alternatives for various control devices on FCCU, including BLD monitoring as an option to COMs for owners or operators of FCCU using fabric filter-type control systems, and total power and secondary current operating limits for owners or operators of ESPs. We also proposed to add a requirement to perform daily checks of the air or water pressure to atomizing spray nozzles for owners or operators of FCC wet gas scrubbers. Finally, we proposed to require a performance test once every 5 years for all FCCU in place of the one-time performance test required by the current Refinery MACT 2.

At proposal, we did not identify any developments in practices, processes and control technologies for CRU process vents based on our technology review. For SRU, we proposed to include the Refinery NSPS subpart Ja allowance for oxygen-enriched air as a development in practice and to allow SRU to comply with Refinery NSPS subpart Ja as a means of complying with Refinery MACT 2.

2. How did the technology review change for the Petroleum Refinery source categories?

a. Refinery MACT 1

We are finalizing most of our technology review decisions for Refinery MACT 1 emissions sources as proposed; however, as described briefly below, we are revising certain proposed requirements.

We are not taking final action adopting the use of appendix K to 40 CFR part 60 for optical gas imaging for refinery equipment subject to the LDAR requirements in Refinery MACT 1 because we have not yet proposed appendix K.

After considering the public comments, we are finalizing the proposed fenceline monitoring requirements, with a few revisions. First, we have made numerous clarifications in this final rule to the language for the fenceline monitoring siting method and analytical method (i.e., Methods 325 A and B, respectively). Specific comments on these methods, along with our responses and explanations of the revisions to the regulatory text are discussed in the “Response to Comment” document. Second, we are finalizing a revised compliance schedule for fenceline monitoring, which will require refinery owners or operators to have the fenceline monitors in place and collecting benzene concentration data no later than 2 years from the effective date of the final rule, as opposed to 3 years in the proposed rule. Third, we have removed the requirement for refinery owners or operators to obtain the EPA approval for the corrective action plan. Fourth, we are requiring the submittal of the fenceline monitoring data on a quarterly basis, as opposed to on a semiannual basis as proposed. Fifth, we are providing guidelines for operators to use in requesting use of an alternative fenceline monitoring technology to the passive sorbent samplers set forth in Method 325B. Finally, to reduce the burden of monitoring, we are finalizing provisions that would allow refinery owners or operators to reduce the frequency of fenceline monitoring for areas that consistently stay well below the fenceline benzene concentration action level. Specifically, we are allowing refinery owners or operators to monitor every other two weeks (i.e., skip period monitoring) if over a two-year period, each sample collected at a specific monitoring location is at or below 0.9 μg/m3. If every sample collected from that sampling location during the subsequent 2-years is at or below 0.9 μg/m3, the monitoring frequency may be reduced from every other two weeks to quarterly. After an additional two years, the monitoring can be reduced to semiannually and finally to annually, provided the samples continue to be at or below 0.9 μg/m3 during all sampling events at that location. If at any time a sample for a monitoring location that is monitored at a reduced frequency returns a concentration greater than 0.9 μg/m3, the owner or operator must return to the original sampling requirements for one quarter (monitor every two weeks for the next six monitoring periods for that location); if every sample collected from this quarter is at or below 0.9 ug/m3, then the sampling frequency reverts back to the reduced monitoring frequency for that monitoring location; if not then the sampling frequency reverts back to the original biweekly monitoring frequency.

b. Refinery MACT 2

We are finalizing, as proposed, our determination that it is not necessary to revise the requirements for CRU pursuant to the technology review and we are finalizing our determination that it is necessary to revise the MACT for SRU and FCCU. For SRU, we are finalizing the revisions as proposed. For FCCU, we are making modifications to the proposed requirements in light of public comment.

As discussed previously, we proposed to remove the alternative in Refinery MACT 2 for owners or operators to demonstrate compliance with the PM limits on FCCU by meeting a 30-percent opacity standard as provided in Refinery NSPS subpart J and instead make the FCCU operating limits in Refinery MACT 2 consistent with Refinery NSPS subpart Ja. Based on the Refinery NSPS subpart J review in 2008, we determined that a 30-percent opacity limit does not adequately assure compliance with the PM emissions limit (see discussion in the proposed rule at 79 FR 36929, June 30, 2014). Thus, we included other monitoring approaches in Refinery NSPS subpart Ja.

Comments received on this proposal, along with data available to the Agency, confirmed that the 30-percent opacity standard is not adequate on its own to demonstrate compliance with the PM (or metal HAP) emissions limit in Refinery MACT 2. We also received comments that the site-specific opacity alternative, which is the only compliance option proposed for FCCU with tertiary cyclones, would essentially require owners or operators with these FCCU configurations to meet an opacity limit of 10-percent. According to commenters, opacity increases with decreasing particle size, so that it is common to exceed 10-percent opacity during soot blowing or other similar events that produce very fine particulates even though mass emissions have not changed appreciably.

Based on the available data, we have determined that a 20-percent opacity operating limit is well correlated with Start Printed Page 75192facilities meeting a limit of 1.0 lb PM/1,000 lbs coke burn-off. Therefore, we are retaining the option in Refinery MACT 2 to comply with Refinery NSPS subpart J except we are adding a 20-percent opacity operating limit in Refinery MACT 2, evaluated on a 3-hour basis. To ensure that FCCU owners or operators complying with the Refinery NSPS subpart J option can meet the 1.0 lb PM/1,000 lbs emissions limit at all times, we are finalizing requirements that owners or operators conduct the performance test during higher PM periods, such as soot blowing. Where the PM emissions are within 80-percent of the PM limit during any periodic performance test, we are requiring the refinery owner or operator to conduct subsequent performance tests on an annual basis instead of on a 5-year basis.

We are finalizing our proposed requirement that compliance with the control device operating limits in the other compliance alternatives be demonstrated on a 3-hour basis, instead of the 24-hour basis currently allowed in Refinery MACT 2.

3. What key comments did we receive on the technology review, and what are our responses?

a. Refinery MACT 1

The majority of comments received regarding the proposed amendments to Refinery MACT 1 pursuant to our technology review dealt with the proposed fenceline monitoring requirements. The primary comments on the fenceline monitoring requirements are in this section along with our responses. Comment summaries and the EPA's responses for additional issues raised regarding the proposed requirements resulting from our technology review are in the “Response to Comment” document in the public docket (Docket ID No. EPA-HQ-OAR-2010-0682).

i. Legal Authority and Need for Fenceline Monitoring

Comment: Numerous commenters claimed that the proposed fenceline monitoring program would unlawfully impose what is effectively an ambient air quality standard for benzene, which is not authorized by CAA section 112, which only authorizes the control of emission sources. The commenters argued it is an ambient standard because sources are required to meet the benzene level set or “perform injunctive relief which may or may not address the source of the benzene.” The commenter quoted language from the proposal as support that EPA has described the benzene level as an ambient standard: “We are proposing a HAP concentration to be measured in the ambient air around a refinery, that if exceeded, would trigger corrective action to minimize fugitive emissions.” 79 FR at 36920 (June 30, 2014). The commenter further noted that this requirement is not just “monitoring” because it establishes a “not-to-be exceeded” level. Therefore, the commenters stated, the EPA should not finalize this portion of the proposal.

Response: We disagree with the comment that the fenceline proposal is an ambient air standard. First, the owner or operator must place the monitors on the facility fenceline to measure emissions from the facility, i.e., on the property of the refiner. While we recognize that we used the term “ambient air” in the preamble to the proposal, we note that the placement requirements for the monitors make clear that the monitors are not monitoring ambient air, which EPA has defined at 40 CFR 50.1(e) as “that portion of the atmosphere, external to buildings, to which the general public has access.” Second, the proposed EPA Method 325A sets out procedures to subtract background concentrations and contributions to the fenceline benzene concentrations from non-refinery emission sources, so that the benzene concentrations measured are attributable to the refinery. In other words, the fenceline monitoring work practice standard uses a benzene concentration difference, referred to as the ΔC (essentially an upwind and downwind concentration difference) to isolate the refinery's emissions contribution.

Furthermore, we disagree that the fact that refiners are required to perform corrective action if the fenceline benzene concentration action level is exceeded makes the benzene action level an ambient standard. As an initial matter sources are not directly responsible for demonstrating that an area is meeting an ambient standard; rather that burden falls on states. See e.g., CAA section 110(a)(2). Moreover, the “corrective action” is simply that sources must ensure that fugitive emission sources on the property are not emitting HAP at levels that will result in exceedances of the fenceline benzene concentration action level. In other words, the purpose of the fenceline monitoring work practice is to ensure that sources are limiting HAP emissions at the fenceline, which are solely attributable to emissions from sources within the facility. In fact, the fenceline benzene concentration action level was established using emissions inventories reported by the facilities, assuming compliance with the MACT standards. Finally, monitoring is conducted as part of the work practice standard to identify sources that will require additional controls to reduce their impact on the fenceline benzene concentration. In that sense, the fenceline monitoring work practice standard is not different than, for example, our MACT standard for refinery heat exchangers. If a facility is exceeding the relevant cooling water pollutant concentration “level” when it performs a periodic test, it must undertake corrective action to bring the concentration down below the action level.

Comment: Several commenters noted that EPA's authority under section 112(d) is to set “emissions standards” and quoted the CAA definition of that term: “A requirement . . . which limits the quantity, rate, or concentration of emissions of air pollutants on a continuous basis, including any requirement relating to the operation or maintenance of a source to assure continuous emission reduction, and any design, equipment, work practice or operational standard promulgated under this Act.” 42 U.S.C. 7602(k). The commenters argued that the proposed fenceline monitoring standard does not meet this definition because it would not “limit the quantity, rate, or concentration of emissions” from any given emissions point. Also, the commenters claimed that the EPA did not designate fenceline monitoring as a work practice under CAA section 112(h) since the EPA did not even mention CAA section 112(h), nor did it conduct any analysis to show that fenceline monitoring meets the CAA section 112(h) factors.

Response: We disagree with the commenters' assertion that the proposed fenceline monitoring work practice standard is not authorized under CAA section 112(d)(6). Contrary to the commenter's claims, we specifically proposed the fenceline monitoring standard under CAA section 112(d)(6) to be a work practice standard that is applied broadly to fugitive emissions sources located at petroleum refineries. As discussed above, the proposed standard does more than impose monitoring as some commenters suggested; it also will limit emissions from refineries because it requires the owner or operator to identify and reduce HAP emissions through a monitoring and repair program, as do many work practice standards authorized under CAA Section 112(h) and 112(d).

We note that the sources addressed by the fenceline monitoring standard—refinery fugitive emissions sources such as wastewater collection and treatment Start Printed Page 75193operations, equipment leaks, heat exchange systems and storage vessels in the Refinery MACT 1 rule—are already subject to work practice standards. Our review of these requirements indicates that this fenceline monitoring work practice standard would be a further improvement in the way fugitive emissions are managed and would provide an extra measure of protection for surrounding communities. The commenter claims EPA did not analyze how the fenceline monitoring requirement meets the criteria in section 112(h). However, that is a misinterpretation of how the criteria apply. The criteria are assessed with regard to whether it is feasible to “prescribe or enforce an emission standard for a source”, and do not apply to the work practice standard. Consistent with the criteria in section 112(h)(2), we determined and established that work practice standards are appropriate for these Refinery MACT fugitive emissions at the time we established the initial MACT standard. In the proposal, (79 FR at 36919, June 30, 2014), we reaffirmed that it is impracticable to directly measure fugitive emission sources at refineries but did not consider it necessary to reiterate these findings as part of this proposal to revise the existing MACT for these sources under CAA section 112(d)(6). We note that the commenters do not provide any grounds to support a reevaluation of whether these fugitive emission sources are appropriately regulated by a work practice standard.

Comment: Several commenters questioned the EPA's authority under the CAA to promulgate a rule that amounts to an ongoing information gathering and reporting obligation. The commenters stated that the EPA has not demonstrated that the proposed fenceline monitoring program represents an actual emission reduction technology improvement. A commenter stated that compliance assurance methods, including monitoring, for fugitive emissions and other emission standards are established as part of the emission standard and EPA's authority to gather information that is not directly required for compliance with a specific standard but is related to air emissions is found in CAA section 114. Under CAA section 114, the requirement must be related to one of the stated purposes and must be reasonable. The commenter did not believe that the EPA has demonstrated that the costs of fenceline monitoring are reasonable in light of the information already available to the EPA and in light of many other means by which the EPA could obtain such information.

Response: We disagree with the commenters' assertion that the authority for the fenceline monitoring requirement falls under CAA section 114 and not CAA section 112(d) because it is an “ongoing information gathering and reporting obligation.” The issue here is not whether EPA could have required the fenceline monitoring requirement under CAA section 114, but rather did EPA support that it was a development in processes practices or controls technology under section 112(d)(6).

As an initial matter, we disagree with the commenters' characterization of the fenceline monitoring standard as “an information gathering and reporting obligation.” We have repeatedly stated that we consider the fenceline monitoring requirement to be a work practice standard that will ensure sources take corrective action if monitored benzene levels (as a surrogate for HAP emissions from fugitive emissions sources) exceed the fenceline benzene concentration action level. The standard requires refinery owners or operators to monitor the benzene concentration at the refinery perimeter, to evaluate the refinery's contribution as estimated by taking the concentration difference between the highest and lowest concentrations (ΔC) in each period, and to conduct root cause analysis and take corrective action to minimize emissions if the concentration difference is higher (on an annual average) than the benzene concentration action level. Thus, the fenceline monitoring requirement goes well beyond “information gathering and reporting.”

In addition, the commenters again read section 112(d)(6) too narrowly by suggesting that a program considered as a development must be a “technology” improvement. Section 112(d)(6) of the CAA requires the EPA to review and revise the MACT standards, as necessary, taking into account developments in “practices, processes and control technologies.” Consistent with our long-standing practice for the technology review of MACT standards, in section III.C of the proposal (see 79 FR 36900, June 30, 2014), we list five types of “developments” we consider. Fenceline monitoring fits squarely within two of those five types of developments (emphasis added):

  • Any add-on control technology or other equipment that was not identified and considered during development of the original MACT standards.
  • Any work practice or operational procedure that was not identified or considered during development of the original MACT standards.

As used here, “other equipment” is clearly separate from and in addition to “add-on control” technology and is broad enough to include monitoring equipment. In this case, fenceline monitoring is a type of equipment that we did not identify and consider during development of the original MACT standards. Additionally, the fenceline standard is a work practice standard, involving monitoring, root cause analysis and corrective action not identified at the time of the original MACT standards. Therefore, the fenceline requirements are a development in practices that will improve how facilities manage fugitive emissions and EPA appropriately relied on section 112(d)(6) in requiring this standard.

Comment: Some commenters contended that because the fenceline monitoring standard is in essence an ambient standard, the only justification that can be used to support it would be under CAA section 112(f)(2). The commenters stated that EPA determined that the MACT standards pose an acceptable level of risk and protect the public health with an ample margin of safety and thus, section 112(f) does not support imposition of the fenceline monitoring requirement. Several commenters stated that the Agency expressly acknowledges that imposition of additional emission standards for fugitive emissions from refinery sources are not warranted under CAA section 112(f). Some commenters suggested that because the existing MACT standards protect public health with an ample margin of safety, the fenceline monitoring requirement imposes an unnecessary burden on industry because it is not necessary to achieve acceptable risk or provide an ample margin of safety.

Response: EPA is not relying on section 112(f)(2) as the basis for the fenceline monitoring requirement. As provided in a previous response to comment, we disagree with the commenters that the fenceline monitoring requirement is an ambient standard and therefore, we do not need to consider what authority would be appropriate for establishing an ambient standard that would apply to fugitive sources of emissions at refineries. We also disagree with the commenters who suggest that EPA may not require fenceline monitoring pursuant to section 112(d)(6) because EPA has not determined that fenceline monitoring is necessary to ensure an acceptable level of risk or the provide an ample margin of safety. Section 112(d)(6) does not Start Printed Page 75194require EPA to factor in the health considerations provided in section 112(f)(2) when making a determination whether it is “necessary” to revise the MACT.

Comment: Commenters stated that the pilot studies undertaken by the EPA and pilot studies undertaken by the refining industry (see the API Fenceline Study in the docket for this rulemaking) demonstrate either that there is no underestimation of emissions and thus, no need for the fenceline monitoring work practice standard, or that fenceline benzene data cannot be used to validate emission estimates. Commenters stated that none of the refineries in the API study of the proposed refinery fenceline standard had study-averaged ΔC concentrations that exceeded the proposed action level of 9 µg/m3 and thus the study provides some evidence that U.S. refineries are not underestimating emissions. Furthermore, the commenter stated that there is significant ambient air monitoring performed that further supports low benzene concentrations in the vicinities of refineries and cited ambient monitoring data collected by the Southeast Texas Regional Planning Commission Air Quality Group and the Texas Commission on Environmental Quality (TCEQ).

Response: We disagree that the API fenceline study demonstrates that there is no underestimation of emissions. The API report referred to by the commenter actually shows higher ΔC concentrations than what we expected, when we compare the distribution of ΔC's presented in the API fenceline study to the distribution of benzene concentrations at the 142 refineries we modeled (see memorandum “Fenceline Ambient Benzene Concentrations Surrounding Petroleum Refineries”, EPA-HQ-OAR-2010-0682-0208). [Note that API did not identify the facilities in their study, so we were not able to perform a one-to-one comparison of the measured ΔC concentrations with the modeled fenceline concentrations.] Furthermore, the API conducted the study primarily during the fall and winter months (October to March) when the ambient temperatures are lower than the annual averages. While this may not impact equipment leak emissions, temperature can have a significant impact on emissions from storage vessels and wastewater treatment systems, so it is likely that the annual average ΔC for the facilities tested could be higher than the “winter” averages measured in the API study. Based on our review of the API study data, we interpret the results to indicate that there may be higher concentrations of benzene on the fenceline attributable to fugitive emissions than anticipated at some facilities. These studies are an indication that the standard we are finalizing will achieve the goal of ensuring that the owners or operators manage fugitive emissions within the refinery.

This regulatory approach also fits with the EPA's goals to improve the effectiveness of rules. Specifically, in this case, we are improving the effectiveness of the rule in two ways. First, we are establishing a fenceline benzene trigger to manage overall fugitive HAP emissions, rather than establishing further requirements on many individual emission points. Secondly, the rule incentivizes facilities to reduce fugitive HAP emissions below the fenceline benzene trigger by providing regulatory options for reduced monitoring.

Regarding ambient monitoring data, we note that existing ambient monitors are not located at the fenceline; they are located away from sources, and concentrations typically decrease exponentially with distance from the emissions source. We are encouraged that data referenced by the commenter indicate that ambient levels of benzene are within levels that are protective of human health in communities, but note that analysis of benzene concentrations in communities does not necessarily indicate that refineries located near these communities are adequately managing their fugitive HAP emissions.

Comment: Several commenters reiterated that they do not believe the proposed fenceline monitoring is a technology development for equipment leaks, storage vessels or wastewater sources. However, if the EPA finalizes the fenceline monitoring requirements, the commenters suggested that there is no longer a need or regulatory basis for imposing both the fenceline monitoring requirements and the existing MACT standards for fugitive HAP emission sources. Thus, the EPA should remove the current MACT requirements for LDAR, storage vessels and wastewater handling and treatment from Refinery MACT 1 if the EPA promulgates fenceline monitoring. Addition of fenceline monitoring on top of the existing MACT requirements, they argue, would violate the Executive Order 12866 mandate to avoid redundant, costly regulatory requirements that provide no emission reductions.

Response: We disagree that the fenceline monitoring standards we are finalizing in this rule are redundant to MACT emissions standards for fugitive HAP emissions sources. The MACT standards impose requirements on fugitive HAP emissions sources consistent with the requirements in CAA section 112(d)(2) & (3), and the fenceline monitoring requirement is not a replacement for those requirements. Rather, based on our review of these standards, we concluded that fenceline monitoring is a development in practices, processes or control technologies that would improve management of fugitive emissions in a cost-effective manner. In selecting this development as an across-the-board means of improving management of fugitive emissions, we rejected other more costly developments that would have applied independently to each fugitive emissions source. Requiring refineries to establish a fenceline monitoring program that identifies HAP emission sources that cause elevated benzene concentrations at the fenceline and correcting high emissions through a more focused effort augments but does not replace the existing requirements. We found that, through early identification of significant fugitive HAP releases through fenceline monitoring, compliance with the existing MACT standards for these emissions sources could be improved and that it was necessary to revise the existing standards because fenceline monitoring is a cost-effective development in processes, practices, and control technologies.

We note that the existing MACT requirements are based on the MACT floor (the best performers), and as such, provide a significant degree of emission reductions from the baseline. The action level for the fenceline work practice standard, by contrast, is not based on the best performers but rather on the highest value expected on the fenceline from any refinery, based on the modeling of refinery emission inventories. As such it is not representative of the best performers and could not be justified as meeting the requirements of section 112(d)(2)and (3). If we were to remove the existing standards for fugitive emission sources at the refinery, we would not be able to justify that sources are meeting the level of control we identified as the MACT floor when we first promulgated the MACT. Nor could we justify the fenceline monitoring program we are promulgating as representing the MACT floor because we considered cost (and not the best performers as previously noted) in identifying the components of the program. Although the fenceline monitoring standard on its own cannot be justified as meeting the MACT floor requirement for each of the separate Start Printed Page 75195types of fugitive emission sources, that does not mean that it is not an effective enhancement of those MACT requirements. To the contrary, it works in tandem with the existing MACT requirements to provide improved management of fugitive emissions and, in that sense, it is precisely the type of program that we believe Congress had in mind when enacting section 112(d)(6).

ii. Rule Should Require Real-Time Monitoring Technology for Fenceline Monitoring.

Comment: Numerous commenters stated that the proposed fenceline standards, which require monitoring using 2-week integrated passive samplers, are flawed and weak for a number of reasons, including that the monitoring method does not provide real-time data, does not provide adequate spatial coverage of the fenceline, and does not provide a mechanism to identify the specific emission source impacting the fenceline to manage fugitive emissions. Several commenters suggested that this monitoring technology is not state of the art. They claimed that there are superior systems in place at refineries that are technically and economically feasible, including at Shell Deer Park, Texas; BP Whiting, Indiana; and Chevron Richmond, California. Further, they claimed that these systems more effectively achieve the objective of reducing fugitive emissions. They claimed several systems are superior to the proposed system, including open-path systems such as ultraviolet differential optical absorption (UV DOAS) and Fourier transform infrared spectroscopy (FTIR), as well as point monitors such as gas chromatographs. A number of commenters suggested that open-path monitors should be required, stating that this technology is capable of providing real-time analysis and data on air pollution, is able to analyze multiple pollutants simultaneously at low, near-ambient concentrations, and is capable of providing more complete geographic coverage.

The commenters also stated that the benefits of real-time monitors are particularly important in communities close to refineries, where they believe refinery emissions are a major source of toxic pollutants and short-term upset events that can have significant public health impacts. In particular, the commenters stated that open-path monitors promote an individual's right-to-know, in real-time, about harmful pollution events affecting their communities, and will allow refinery owners or operators to immediately identify fugitive emissions and undertake swift corrective action to reduce these emissions. Some commenters suggested that, if the EPA rejects these open-path real-time monitors, then at a minimum the EPA should require the use of active daily monitoring, such as auto-gas chromatograph (GC) systems.

Finally, a number of commenters recommended that the EPA provide sufficient flexibility in its regulations to allow state and local jurisdictions to develop, demonstrate, and subsequently require the use of alternative monitoring programs, provided these monitoring programs are at least equivalent to those in the final rule.

Response: We understand that many commenters believe real-time monitoring would not only help refinery owners or operators in identifying emission sources, but also would warn the community of releases in real time.

Both open-path systems and active sampling systems (such as auto-GCs) mentioned by the commenters, are monitoring systems capable of yielding monitoring data quickly—ranging from a few minutes to about a day. However, these “real-time” systems have not been demonstrated to be able to achieve all of the goals stated by the commenters—specifically, able to provide real-time analysis and data on multiple pollutants simultaneously at low-, near-ambient concentrations, with more complete geographic (or spatial) coverage of the fenceline.

The real-time open-path systems suggested by the commenters are all limited in that they are not sensitive enough to detect benzene at the levels needed to ensure that fenceline monitoring achieves its intended goal. The fenceline monitoring system needs to be capable of measuring at sub-ppbv levels—well below the 9 μg/m3 fenceline benzene concentration action level in the final rule, in order to determine the ΔC. In the proposal, we discussed two open-path monitoring technologies, FTIR and UV-DOAS. For the proposed rule, we analyzed the feasibility of employing UV-DOAS over FTIR because the UV-DOAS is more sensitive to detection of benzene than FTIR, as we described in the proposal. We reviewed performance data on several UV-DOAS systems in support of the proposed rule, and for this final rule, we considered information submitted during the comment period. We found that the lowest detection limit reported for any commercially-available UV-DOAS system is on the order of 3 ppbv over a 200-meter path length, whereas the fenceline benzene concentration action level is 2.8 ppbv (equivalent concentration to 9 μg/m3). This system is being installed at the Shell Deer Park refinery but has not been field validated yet. Thus, we do not yet know the detection capabilities of the system, as installed. Based on the lowest reported detection limit, it cannot achieve the detection levels needed to demonstrate compliance with the fenceline standard in this final rule. This system also will only cover approximately 5 percent of the fenceline at Shell Deer Park, instead of the full fenceline coverage of the passive diffusive tube monitoring system we proposed. Facilities would have to deploy a monitoring system consisting of many open-path monitors to achieve the same spatial coverage as the passive diffusive tube monitoring system.

For the final rule, we also reviewed other UV-DOAS systems in operation at refineries that commenters identified. However, reported detection limits for these systems are even higher than for the type of system being installed at Shell Deer Park. For example, we reviewed the open-path UV-DOAS system information from BP Whiting and found that they were able to verify a detection limit of 8 ppbv path average concentration for benzene over a 1,500-meter optical path. This is well above the 2.8 ppbv fenceline benzene concentration action level, let alone the sub-ppbv levels necessary to determine the ΔC. Moreover, this system, though commercially available, was optimized by developing alternative software to improve the detection limit (see memorandum “Meeting Minutes for April 21, 2015, Meeting Between the U.S. EPA and BP Whiting” in Docket ID No. EPA-HQ-OAR-2010-0682). Thus, the system, as installed, would not be readily available to other refineries. We reviewed data for the UV-DOAS system at the Chevron Richmond refinery and found that this system, with optical path lengths ranging from 500 to 1,000 meters, has a reported benzene detection limit of 5 ppbv averaged over the path length. Again, this is above the fenceline benzene concentration action level at the fenceline established in this final rule. In addition, we could not find any information to support the reported detection limit. We note that the public Web site operated by the City of Richmond, California indicates that information provided by the system is informational only, not quality assured, and not to be used for emergency response or health purposes.

We also disagree with the commenter's claim that if the EPA does not finalize requirements for real-time open-path monitors then, at a minimum, the EPA should require active daily monitoring. There are two methods of Start Printed Page 75196active monitoring. One method, which we will refer to as the “auto-GC method,” uses a dedicated gas chromatograph at each monitoring location and can return ambient air concentration results multiple times a day or even hourly. The other method, which we refer to as “method 2,” uses an active pump to collect gas in a sorbent tube or in an evacuated canister over a 1-day period, for later analysis at a central location. While active sampling monitoring networks are capable of measuring multiple pollutants and would likely be able to detect benzene at sub-ppbv levels as necessary to demonstrate compliance with the fenceline requirements in this final rule, they consist of discreet monitors and would not provide any better spatial coverage of the refinery fenceline than a passive diffusive tube monitoring network. Further, as shown in Table 9 of the proposed rule (see 79 FR 36923, June 30, 2014), like open-path systems, an active sampling monitoring network would cost many times that of a passive diffusive tube monitoring network. At proposal, we estimated the costs of active daily sampling based on “method 2” to be approximately 10 times higher than for the proposed passive monitoring (see memorandum “Fenceline Monitoring Technical Support Document”, Docket ID No. EPA-HQ-OAR-2010-0682-0210). We note that this type of active daily sampling based on method 2 does not necessarily yield results within 24 hours as the sample analysis would be conducted separately. We did not specifically estimate the costs of an auto-GC alternative, but the capital costs would be at least 20 to 30 times that for the passive diffusive tube system, would require shelters and power supplies at all monitoring locations and would have operating costs similar to the “method 2” active monitoring option we considered.

To date, there are no commercially-available, real-time open-path monitors capable of detecting benzene at the sub-ppbv levels necessary to demonstrate compliance with the fenceline requirements in this final rule. Only a system that can detect such levels will result in effective action by facilities to identify and control fugitive emissions in excess of those contemplated by the MACT standards. Further, active monitoring systems, while potentially capable of detecting benzene at sub-ppbv levels, like open-path systems, become very costly when enough monitors are located around the facility to approach the spatial coverage of the passive diffusive tubes. However, we believe that the state of technology is advancing and that the capabilities of these systems will continue to improve and that the costs will likely decrease. If a refinery owner or operator can demonstrate that a particular technology would be able to comply with the fenceline standards, the owner or operator can request the use of an alternative test method under the provisions of 40 CFR 63.7(f). A discussion of the specific requirements for these requests can be found in the first comment and response summary of Chapter 8.3 of the “Response to Comment” document.

Comment: One commenter stated that the required monitoring should include real-time monitoring of all chemicals released by refineries that pose risks to human health. The commenter stated that the limited scope of monitoring required by the proposed rule appears to be guided by the EPA's judgment that fugitive, or “unintended” emissions pose the greatest threat to public health. On the contrary, communities may well suffer from the effects of chemicals released into the air under normal, permitted emissions. A more expansive monitoring strategy would account for both routine and fugitive emissions.

Several commenters noted that monitoring is limited to benzene as opposed to multiple HAP. One commenter noted that ill health experienced by refinery neighbors is due in large part to the synergistic effects of multiple chemicals. Therefore, the commenter stated that it is essential that the rule require monitoring of the full range of chemicals with health implications. Other commenters recommended that the fenceline monitoring requirement be amended to include additional contaminants, such as VOC, that may negatively impact human health and the environment. Conversely, other commenters stated that the EPA has appropriately selected benzene as a target analyte and surrogate for HAP emissions from petroleum refineries, as benzene is a common constituent in refinery feedstocks and numerous refinery streams, and is present in most HAP-containing streams in a refinery.

Response: As part of the CAA section 112(d)(6) technology review, the EPA identified the fenceline monitoring standard as a development in practices, processes or control technologies that could improve management of fugitive HAP emissions. Thus, to the extent the commenter is suggesting that the EPA require the fenceline monitoring system to monitor for emissions of non-HAP pollutants, such request goes beyond the scope of our action. Furthermore, to the extent that the commenter is raising health concerns, although we address residual risk remaining after implementation of the MACT standards under CAA section 112(f)(2), we note that the MACT standards themselves, including this requirement, are aimed at protecting public health, especially in surrounding communities. As we explained in the proposal, and as we determine for this final rule, the MACT standards as modified by additional requirements for storage vessels, provide an ample margin of safety to protect public health. We did not propose and are not finalizing a fenceline monitoring requirement as necessary to provide an ample margin of safety under CAA section 112(f)(2).

Petroleum refining emissions can contain hundreds of different compounds, including many different HAP, and no single method can detect every HAP potentially emitted from refineries. While several HAP are amenable to quantification via passive diffusive tube monitoring using the same adsorbent tubes used for benzene (e.g., toluene, xylenes and ethyl benzene, which have uptake rates in Table 12.1 in Method 325B), we selected benzene as a surrogate because it is present in nearly all refinery fugitive emissions. By selecting a single HAP as a surrogate for all fugitive HAP, we are able to establish a clear action level, which simplifies the determination of compliance for refinery owners or operators and simplifies the ability of regulators and the public to determine whether sources are complying with the work practice standard. As described in the proposal preamble, benzene is ubiquitous at refineries and present in nearly all refinery process streams, including crude oil, gasoline and wastewater. Additionally, benzene is primarily emitted from ground level, fugitive sources that are the focus of the work practice standard. Thus, we conclude that monitoring of benzene is appropriate and sufficient to identify emission events for which the monitoring program is targeting. Consequently, we are not requiring quantification of other pollutants although refinery owners or operators could choose to analyze the diffusive tube samples for additional HAP in conducting root cause analysis and corrective action.

iii. Fenceline Monitoring Action Level

Comment: Several commenters stated that the action level for fenceline monitoring (i.e., 9 μg/m3 or 2.8 ppbv), was set too high. Some of these commenters noted that the EPA selected 9 μg/m3 as the highest modeled benzene Start Printed Page 75197concentration at any refinery fenceline. One commenter stated that this was arbitrary and capricious and stated the action threshold level makes little sense because only 2 of the 142 modeled facilities are expected to have fenceline concentrations above 4 μg/m3. Several commenters noted that the average modeled benzene concentration is 0.8 μg/m3, which is more than an order of magnitude less than the proposed fenceline benzene concentration action level.

Two commenters argued for a lower action level threshold, citing the proposed California OEHHA rule, which finalized new and revised benzene reference exposure levels (REL) that are more stringent than the ones the EPA used in the residual risk assessment supporting the proposed rule.

Two commenters stated that while the fenceline benzene concentration action level of 9 μg/m3 is relatively protective compared to standards adopted by many states, including Louisiana and Texas, it is still 80-percent higher than the European Union's standard of 5 μg/m3. The commenter urged the agency to consider adopting a stricter standard comparable to what other industrialized nations use.

Several commenters stated that the EPA's 9 μg/m3 action level is inconsistent with the statutory text and objectives of CAA sections 112(d) and (f), which direct the EPA to focus on the best-performing, lowest-emitting sources, in order to require the “maximum achievable” emission reductions. The commenters stated that the EPA promulgated the 9 μg/m3 limit without properly following the statutory requirements for establishing MACT floor limits, pointing out that the EPA made no determination of whether or not these general models were representative of the emissions levels actually achieved by the submitting refinery, and no connection was drawn between the best performing sources and the eventual 9 μg/m3 limit.

On the other hand, several commenters opposed the 9 μg/m3 action level suggesting that it was not achievable and that it is arbitrary. Some commenters noted that emission/dispersion models are always very site-specific and do not necessarily yield a result that is reliable or reproducible. Several commenters stated that additional studies are necessary to allow the agency to account for these variables and set a more appropriate concentration corrective action level. Commenters suggested a 2-year data gathering effort at all refineries and data evaluation before determining a specific threshold to use.

Several commenters recommended action levels ranging from 15 μg/m3 to 20 μg/m3 of benzene to account for the variability expected in monitoring data. The commenters stated that modeling biases have underestimated the necessary action level to achieve the stated goals of the program.

Response: First, it is important to note that the purpose of the standard has not changed between proposal and promulgation, namely that it is a technology-based standard that is an advancement in practices to manage fugitive emissions. It is not intended to be a separate or new MACT standard promulgated pursuant to CAA sections 112(d)(2) and (3) for which a “floor” analysis would be required.[7] Nor is it a standard that we are promulgating pursuant to CAA section 112(f)(2) as necessary to provide an ample margin of safety to protect public health or prevent an adverse environmental effect.[8] Thus, claims that a standard should reflect European Union health-based standards or the California OEHHA rule are misplaced. We also disagree with the suggestion that the proposed monitoring requirement will allow for higher emissions. As noted elsewhere, we are retaining all of the source-specific requirements for fugitive emissions sources that exist in Refinery MACT 1.

We disagree with the commenters that suggest that the proposed action level of 9 μg/m3 is too low and may not be achievable even for well-performing facilities. As discussed in the preamble for the proposed rule, we selected the 9 μg/m3 benzene action level because it is the highest value on the fenceline predicted by the dispersion modeling and, thus, is a level that we estimate that no refinery should exceed when in full compliance with the MACT standards, as amended by this final rule. All of the results of our pilot study, the API study, and the other ambient monitoring data near refineries clearly indicate that this level is achievable. Furthermore, we expect the fenceline concentration difference measured following the procedures in the final rule to be indicative of refinery source contributions and we have provided procedures to isolate these concentrations from outside sources, as well as background.

We expect that the fenceline monitoring standard will result in improved fugitive HAP emissions management as it will alert the refinery owners or operators of fugitive sources releasing high levels of HAPs, such as large leaks, faulty tank seals, etc.

iv. Fenceline Monitoring Root Cause Analysis and Corrective Action Provisions

Comment: A number of commenters objected to the proposal's “open-ended” provisions allowing the EPA to direct refinery owners or operators to change their operations in order to achieve the fenceline limit, with no regulatory limits on costs and without consideration of the impact to safe operations or operability of the plant. Another commenter stated that the EPA must properly assess the costs associated with the root cause analysis/corrective action requirements and should establish a cost effectiveness threshold for any required root cause analysis/corrective action to ensure that limited resources are effectively and efficiently applied for the control of emissions.

One commenter stated the proposed fenceline benzene concentration action level is effectively an ambient air standard, because corrective action to achieve that level is required and that if a facility's initial corrective action is unsuccessful, the rule provides that further action is required and the EPA must approve that further corrective action plan. Thus, the commenter argued, the EPA would essentially be able to dictate corrective actions, with no bounds on what could be required and no consideration of whether any cost-effective actions are available to assure the action level is met. The commenter continued that such a requirement converts a work practice program to an emission limitation and such ambient air limits are not authorized by CAA section 112. Several commenters noted that LDAR and current work practice programs have no similar requirement for the EPA approval, and the commenters suggested that the requirement for the EPA approval of any second corrective action should not be included in 40 CFR 63.658(h).

Another commenter recommended that, if after corrective action, a facility still has an exceedance for the next sampling episode, then the facility should be required to do more than it Start Printed Page 75198did after the first root cause analysis, as the prior corrective action clearly did not correct the problem. The commenter stated that one corrective action measure the EPA should include in all such instances is higher-quality monitoring such as UV-DOAS for at least 1 year to monitor, identify, correct and assure ongoing compliance after the exceedance problem is fixed.

Response: The “on-going” requirement to achieve the fenceline benzene concentration action level is no different in concept from the LDAR requirements for equipment or heat exchange systems in the Refinery MACT 1 rule, which requires the refinery owner or operator to repair the source of the emissions regardless of what it takes until compliance with the standard is achieved.

We disagree with the claim that the EPA must assess the costs associated with the root cause analysis/corrective action requirements and establish a cost effectiveness threshold for any required root cause analysis/corrective action to ensure that limited resources are effectively and efficiently applied for the control of emissions. We did not attempt to project the costs of the root cause analysis/corrective action for at least two reasons. First, based on the dispersion modeling of the benzene emissions reported in response to the inventory section of the 2011 ICR, we project that no refinery should exceed that fenceline benzene concentration action level if in full compliance with the MACT standards, as amended by this action. Thus, assuming compliance with the MACT standards, we would expect that there are no costs for root cause analysis/corrective action. To the extent that there are exceedances of the action level, the premise of the fenceline monitoring is to provide the refinery owners or operators with the flexibility to identify the most efficient approaches to reduce the emissions that are impacting the fenceline level. Since the choice of control is a very site-specific decision, we would have no way to know how to estimate the costs. Thus, the source is in the best position to ensure that resources are effectively and efficiently spent to address any exceedance.

We intended the proposed requirement for refinery owners or operators to submit a corrective action plan for the EPA approval to provide the Administrator with information that they were making a good-faith effort to reduce emissions below the fenceline benzene concentration action level, as expeditiously as practicable. However, we understand the importance for refinery owners or operators to begin corrective action as soon as possible, without having to wait for the EPA approval. Therefore, we are finalizing the requirement for refinery owners or operators to submit such plans but we are not finalizing the requirement that the EPA must approve the plan prior to the corrective action being taken.

We previously responded to comments regarding UV-DOAS or other open-path monitoring systems in this section, explaining that the current detection limits for these systems exceeds the action level threshold and, thus, these systems would not provide usable data to inform corrective action. Thus, we disagree that the EPA should require these systems for all facilities whose first attempt at corrective action is ineffective.

v. Fenceline Monitor Siting Requirements

Comment: Numerous commenters provided suggestions on, or requested clarification of, the monitor siting requirements. Several commenters stated that proposed Method 325A uses the terms “fenceline or property boundary,” while it should consistently use the term “property boundary” or even “property line” as the fenceline location. Several commenters stated that Sections 8.2.2.1.4 and 8.2.2.3 of Draft Method 325A specify that samplers be placed just beyond the intersection where the measured angle intersects the property boundary and this could require placing monitors on other people's property, in a road, in a water body or in a railroad right-of-way. The commenters suggested that facilities should be allowed to place monitors at any vector location that meets other requirements between the property boundary and the source nearest the property boundary. They stated that facilities need this clarification to avoid obstructions (e.g., buildings or trees) that may be at the property line.

Numerous commenters requested that the rule clarify where monitors need to be placed in special circumstance, such as refineries bisected by a road, railroad or other public right-of-way or a boundary next to a navigable waterway. Several commenters stated that refiners should not need to place monitors on these property boundaries or other property boundaries where there are no residences within 500 feet of the property line. Commenters also asked if areas that had non-refinery operations, but are still inside the property boundary, would be included for purposes of determining where to site monitors.

A few commenters expressed concern about the approach for determining the number of required monitors at a site based on the acreage, noting that it is unfair to small facilities and will leave gaps in monitoring coverage for very large facilities. Some commenters recommended amending the proposed rule to require the placement of fenceline monitors at fixed distances along facilities' perimeters with no maximum number of monitors. Some commenters stated that the rule should specify an acceptable range on the 2,000-foot spacing requirement or the radial placement requirement as it may be necessary to address accessibility or safety concerns. Several commenters suggested that a lower minimum number of sampling monitors should be required for very small refineries or small “subareas.” These commenters noted that refineries often include disconnected parcels that can be very small (e.g., 10 acres or less). If each disconnected parcel must be treated as a separate subarea, then both sampler siting options in Draft Method 325A would result in unnecessarily large numbers of samplers extremely close together. Some commenters recommended that Method 325A specify that samplers need not be placed closer than 500 feet (versus the normal 2,000-foot interval specified in Option 2) along the fenceline from an adjoining sampler, regardless of whether the radial or linear approach is used and should waive the minimum number of samplers specified in Sections 8.2.2.1.1, 8.2.2.2.1, and 8.2.3.1. Another commenter added that the rule should waive the requirement for additional samplers in Sections 8.2.2.1.5 and 8.2.3.5 if the 500-foot minimum spacing criterion is compromised.

Response: We agree that the Method 325A should provide clear and consistent language. We have revised the language to be consistent in referring to the “property boundary”. We have also revised the Method to allow placement of monitors at any radial distance along either a vector location or linear location (that meets the other placement requirements) between the property boundary and the source nearest the property boundary. That is, the monitors do not need to be placed exactly on the property boundary or outside of the property boundary. They may be placed within the property closer to the center of the plant as long as the monitor is still external to all potential emission sources. We do note that if the monitors are placed farther in from the property boundary, the owner or operator should take care to ensure, if possible, that the radial distance from the sources to the monitors is at least 50 Start Printed Page 75199meters. If the perimeter line of the actual placement of the fenceline monitors is closer than 50 meters to one or more sources, then the additional monitor citing requirements will apply. We have revised subparagraphs of Section 8.2.2 to provide this allowance. This clarification should address issues related to obstructions such as tall walls located at the facility boundary.

We intended that the fenceline monitoring would create a monitoring perimeter capable of detecting emissions from all fugitive emission sources at the refinery facility. We have long established that a road or other right of way that bisects a plant site does not make the plant site two separate facilities, and, thus, would not be considered part of the property boundary. As we agree that monitors need only be placed around the property boundary of the facility, it would not be necessary to place monitors along a road or other right-of-way that bisects a facility. We have clarified this in the final rule and Method 325A.

If the facility is bounded by a waterway on one or more sides, then the shoreline is the facility boundary and monitors should be placed along this boundary. If the waterway bisects the facility, the waterway would be considered internal to the facility and monitors would only be needed at the facility perimeter.

Regarding the comment that monitors should not be required where there is no residence within 500 feet of the property line, we disagree. We proposed and are finalizing the fenceline monitoring standards under CAA section 112(d)(6) as a means to improve fugitive HAP emissions management, regardless of whether there are people living near a given boundary of the facility.

Regarding the clarification requested about monitor placement considering non-refinery operations, the property boundary monitors should be placed outside of all sources at the refinery. This is because moving the monitoring line inward to exclude the non-refinery source could lead to an underestimation of the ΔC compared to the monitoring external of the entire site. If the non-refinery source is suspected of contributing significantly to the maximum concentration measured at the fenceline, a site-specific monitoring plan and monitoring location specific near-field interfering source (NFS) corrections will be needed to address this situation.

Section 8.2.3 of Method 325A includes language to provide some flexibility when using the linear placement (±10% or ±250 feet). We consider it reasonable to provide similar placement allowance criteria for the radial placement option (±1 degree). We are not providing requirements that would allow small area refineries to use fewer than 12 monitoring sites. We do not consider that any refinery would be so small as to warrant fewer than 12 monitors; however, we did not necessarily consider very small subareas for irregularly shaped facilities or segregated operations. When considering these subareas, we agree that fewer than 12 monitoring sites should be appropriate. Therefore, we have provided that monitors do not need to be placed closer than 152 meters (500 feet) (or 76 meters (250 feet) if known sources are within 50 meters (162 feet) of the monitoring perimeter, which is likely for these subareas or segregated areas) with a stipulation that a minimum of 3 monitoring locations be used per subarea or segregated area. We note, however, that this distance provision does not obviate the near source extra monitoring siting requirements or the requirement to have a minimum of three monitors per subarea or segregated area.

If facility owners or operators have questions regarding the required locations of monitors for a specific application, they should contact the EPA (or designated authority) to resolve questions about acceptable monitoring placement.

vi. Compliance Time for Fenceline Monitoring Requirements

Comment: Some commenters supported EPA's proposal to provide 3 years to put a fenceline monitoring program in place, but the commenters believe that timing is unclear in the proposed regulatory language, which appears in Table 11 to subpart CC, and requested that the EPA add the initial compliance date to 40 CFR 63.658(a). One commenter stated that instituting this program for all 142 major source U.S. refineries would require considerable time. Based on their experience with their pilot study, one commenter noted that commercially available weather guards meeting the specifications of proposed Method 325A are not available and would need to be fabricated. Additionally, a commenter stated that only a limited number of laboratories in the U.S. are able to perform the necessary analyses. According to the commenter, considerable time and effort will be needed to qualify additional laboratories and to expand the capacity of existing laboratories to handle the samples from 142 refineries.

Other commenters disagreed with the EPA's proposed compliance time and suggested that the EPA shorten the timeline for implementation at refineries so that possible corrective action occurs much sooner than proposed. The commenters suggested that deployment of passive samplers can proceed more promptly than proposed, especially since the EPA has simultaneously proposed specific “monitor siting and sample collection requirements as EPA method 325A of 40 CFR part 63, Appendix A, and specific methods analyzing the sorbent tube samples as EPA Method 325B of 40 CFR part 63, Appendix A.” Moreover, the commenter noted, a principal reason that the EPA selected passive monitors over active monitors was due to the relative “ease of deployment.” The commenter claimed this ease of deployment rationale is undermined by a 3-year grace period to deploy passive monitors when the EPA is providing very specific criteria for their use. The commenter suggested that the EPA require full compliance with the passive monitoring requirement within 1 year of the effective date of the rule.

Response: While we realize that it will take some time for the refinery owners or operators to understand the final rule and develop a compliant monitoring program, we agree that in requiring the passive sampler monitoring system, we recognized the ease of implementation and deployment. Although industry commenters identified issues they faced in the API pilot study while trying to implement the monitoring method, we note that the 12 facilities that participated in the API pilot study installed the fenceline monitors and began sampling in late 2013 with relative ease and within months of obtaining the draft methods. Thus, we disagree with the suggestion that 3 years is insufficient and agree with other commenters that 3 years is in fact too long. However, we also are aware that the API pilot facilities used the direct ΔC approach proposed and did not attempt to develop site-specific monitoring programs to correct for interfering near-field sources. Although we expect that facilities could complete direct implementation of the proposed fenceline monitoring requirement within 1 year after the effective date of the rule, as suggested by some commenters, facilities that choose to develop a site-specific monitoring plan would need a longer period of time. Therefore, we are finalizing requirements that specify that facilities must begin monitoring for the official Start Printed Page 75200determination of ΔC values no later than 2 years after the effective date of the rule.

vii. Fenceline Monitoring Recordkeeping and Reporting Requirements

Comment: Some commenters suggested that facilities should be required to submit the monitoring data via the ERT only if they exceed the fenceline benzene concentration action level and that all remaining data should be kept on-site and available for inspection or upon request of the EPA, citing that this is consistent with EPA's semiannual NESHAP reporting of only exceptions (i.e., deviations). Other commenters requested that the EPA only post the rolling annual average concentration values and not the 2-week monitoring data. These commenters indicated concern that if errors are present in the raw data that are submitted semiannually to the EPA, the data, errors and all, will be released to the public and correcting them will not take place or will not take place in a timely manner. One commenter added that there is very little useful information that can be gleaned from the raw data and posting it simply invites misunderstandings.

Commenters also stated that the EPA should adopt reporting requirements to ensure that facilities report the monitoring data appropriately. Specifically, commenters recommended that 40 CFR 63.655(h)(8)(i) should be clarified to only require reporting of valid data and cautioned that data should be processed to allow accurate calculations of annual averages to be used for reporting and evaluation. To accomplish this, commenters recommended that the rule provide 75 days from the end of a 6-month sampling period to report to the EPA, rather than the proposed 45-day period, in order to provide adequate time to obtain quality-assured results for all 2-week sampling periods.

One commenter applauded the proposal's requirements for electronic reporting of the fenceline concentration data and making the resulting information publicly available. However, the commenter recommended that the EPA consider a more truncated data reporting period that is more consistent with the associated milestones of collecting a 14-day sampling episode. As is, the commenter claimed, the proposed rule would have a lag time of up to 7.5 months between data collection and posting. The commenter indicated that data reporting on a more frequent schedule will not only provide transparency, but will provide states and local agencies with information about air quality at refineries at a frequency that could allow informed activities to address leaks much more quickly and protect public health.

Response: We disagree with the commenters who suggest that facilities only report the rolling annual average or only exceedances of the fenceline benzene concentration action level because the commenters believe there is little information to be gleaned from the raw data. Monitoring data are useful in understanding emissions, testing programs, and in determining and ensuring compliance. We generally require reporting of all test data, not just values calculated from test data and/or where a facility exceeds an emissions or operating limit. For example, when we conduct risk and technology reviews for source categories, we are adding requirements for facilities to submit performance test data into the ERT, not just performance test data that indicates an exceedance of an applicable requirement. In the Mercury and Air Toxics Rule, we require facilities to report direct measurements made with CEMS, such as gas concentrations, and we require hourly reporting of all measured and calculated emissions values (see discussion at 77 FR 9374, February 16, 2012). In particular, for the fenceline monitoring requirements in this final rule, we offer facilities options for delineating background benzene emissions and benzene emissions not attributable to the refinery, and we offer options for reduced monitoring, making it even more necessary that we have all of the data to review to ensure that testing and analyses are being done correctly and in compliance with the requirements set out in the regulations, and that root cause analyses and corrective actions are being performed where necessary. Therefore, as proposed, we are finalizing the requirements that facilities report the individual 2-week sampling period results for each monitor, in addition to the calculated ΔC values in their quarterly reporting.

Regarding commenters' concerns that facilities post accurate data and have sufficient time to perform quality assurance on the data, in the final rule, we have established provisions for how sources are to address outliers and data corrections. Additionally, as proposed, we do not require an initial report until facilities have collected 1 year of data so that facilities do not report the data until a rolling annual average value can be determined. This will allow refinery staff and analytical laboratories to iron out any issues that might arise as they implement these methods for the first time. Once this initial data collection period is complete, we anticipate that data quality issues should be infrequent. Therefore, we are providing a 45-day period following each quarterly period before facilities must submit the monitoring results, which should provide facilities adequate time to correct any data errors prior to reporting the data.

Regarding comments that suggest reporting each 2-week sample result soon after its collection, we disagree. This frequency would put undue burden on the refinery owners or operators in trying to collect, review and quality assure the data prior to reporting. However, we agree with commenters that more frequent reporting of the fenceline monitoring data would be useful. Therefore, we have revised the reporting frequency for the fenceline monitoring data to be quarterly in the final rule rather than semiannually as proposed. Additionally, we understand that there is a lot of interest in how these data will be presented to the public, and we plan to reach out to all stakeholders on appropriate approaches for presenting this information in ways that are helpful and informative.

b. Refinery MACT 2

This section provides comment and responses for the key comments received regarding the technology review amendments proposed for Refinery MACT 2. Comment summaries and the EPA's responses for additional issues raised regarding the proposed requirements resulting from our technology review are in the “Response to Comment” document in the public docket (Docket ID No. EPA-HQ-OAR-2010-0682).

i. FCCU

We received comments on the consideration of developments in pollution controls, the averaging time for FCCU PM limits, and the FCCU opacity limit, as discussed below.

Comment: One commenter stated that the EPA failed to consider developments in pollution controls for HAP from FCCUs for two reasons. First, the commenter contended that cost is not a valid consideration to evaluate if a “development” in pollution control is necessary pursuant to section 7412(d)(2), (3), (6), unless the EPA is setting a “beyond-the-floor” requirement.

Second, the commenter claimed that the EPA's review of developments is nearly 10 years old and misses some important pollution control Start Printed Page 75201improvements in the industry. For example, the commenter noted that Valero Benicia installed a combination of controls in 2012 including a scrubber, SCR and CO Boiler that combine exhaust streams from the FCCU and coking and reportedly eliminate HAP emissions entirely from these sources.

The commenter also asserted that EPA consent decrees impose lower effective limits on PM than the EPA considered under the technology review. The commenter identified the BP Whiting facility as subject to 0.7 lb PM/1,000 lbs coke burn-off at one FCCU and 0.9 lb PM/1,000 lbs coke burn-off at another and claimed these limits are lower than the 1.0 lb PM/1,000 lbs coke burn-off limit currently mandated by Refinery MACT 2.

Response: We disagree that we cannot consider costs when determining if it is necessary to revise an existing MACT standard based on developments in practices, processes and control technologies. The commenter suggests that we cannot consider costs because of the requirements in CAA section 112(d)(2) and (3) for establishing initial MACT standards and which do not allow for consideration of costs until the second, “beyond the floor” phase of the analysis. As discussed previously in this preamble where we respond to comments on the fenceline monitoring requirements, to the extent that the commenters are suggesting that EPA must re-perform the MACT floor analysis for purposes of setting a standard pursuant to section 112(d)(6), we note that the D.C. Circuit has rejected this argument numerous times, most recently in National Association for Surface Finishing et al. v. EPA No. 12-1459 in the U.S. Court of Appeals for the District of Columbia.

Regarding the claim that the EPA did not consider the types of controls at the Valero and BP facilities, we disagree. The control measures for both of those facilities are controls that existed at the time of the development of the MACT standard. Thus, we did not identify these technologies as developments in control technologies during the technology review. However, we did identify developments in processes or practices that reflect better control by the existing technology and we reviewed modified emission limits that reflect that better level of control. The commenter suggested that we failed to consider a level of zero when the Valero facility was able to achieve zero emissions through a combined SCR, boiler and scrubber. However, the commenter provided no information to support such a claim and we are skeptical that such a result could be achieved. We note that the SCR is designed specifically to reduce NOX emissions, and would not be capable of reducing significantly, much less eliminating completely, HAP emissions. Similarly, based on our long-standing understanding of the processes, neither a boiler nor a scrubber could achieve such a result. Regarding the level of emissions achieved at the BP Whiting facility, we note that we evaluated control systems that can meet 0.5 lb PM/1,000 lb coke burn-off, which is a lower limit than that at BP Whiting. We determined that these were cost-effective to require for new units that are installing a new control system. However, we determined that retrofitting controls designed to meet a PM limit of 1.0 lb PM/1,000 lbs coke burn-off to now meet a limit of 0.5 lb PM/1,000 lbs coke burn-off was not cost-effective when considering PM and PM2.5 emissions reductions. We projected the cost of the 0.5 lb PM/1,000 lbs coke burn-off limit in retrofit cases to be $23,000 per ton PM emissions reduced. To meet a limit of 0.7 lb PM/1,000 lbs coke burn-off or 0.9 lb PM/1,000 lbs coke burn-off, as is the case for BP Whiting, the retrofit costs would be similar to this 0.5 lb PM/1,000 lb coke burn-off option, but the reductions would be even less, resulting in costs over $23,000 per ton. As metal HAP content of FCCU PM is approximately 0.1 to 0.2-percent of the total PM, the cost of requiring this lower limit for existing FCCU is over $10 million per ton of metal HAP reduced. Therefore, we determined that it is not necessary to revise the PM standard for existing FCCU sources.

Comment: Refinery MACT 2 requires the owner or operator to demonstrate compliance with the PM FCCU limits by complying with the operating limits established during the performance test on a daily (i.e., 24-hour) average basis. Several commenters objected to the EPA's proposal to revise this requirement to a 3-hour averaging time. Commenters restated EPA's arguments for 3-hour averaging time as: (1) Daily average could allow FCCUs to exceed limits for short periods while still complying with the daily average, (2) consistency with NSPS subpart Ja and (3) consistency with duration of testing. The commenters stated that the EPA had not provided any data that show that the daily average could allow FCCUs to exceed limits for short periods and, therefore, the EPA is using a hypothetical compliance assurance argument to change emission limits. The commenters stated that a change in emission limits is not authorized by CAA section 112 because the emission limitations in Refinery MACT 2 for FCCUs were established as daily averages following the floor and ample margin of safety requirements in section 112(d)(2) of the CAA.

The commenters also state that the EPA's additional arguments for the change to a 3-hour average are irrelevant and legally deficient. The commenters stated that the combination of a numerical emission limit and an averaging period frames the stringency of a limitation and that a reduction in either of those factors results in a significant lowering of the operating limit. The commenters conclude that the EPA has proposed to change the stringency of the requirements without justification, and the CAA requires that such a change in stringency be justified pursuant to CAA section 112(d)(6) or (f)(2). The commenters stated that increasing stringency for consistency with NSPS rules is not a criterion for a CAA section 112(d)(6) action. Rather that section requires a change to be due to “developments.” The only change in technology since the 2002 promulgation of Refinery MACT 2 is the availability of PM continuous emission monitoring system (CEMS), which is unproven.

One commenter noted that changing the averaging time is a very significant modification considering that the compliance limits would apply for periods of SSM. This commenter stated that it is unlikely that existing operations can consistently be in compliance with a new 3-hour average since the current daily averaging was put in place to recognize that there will be periods of operating variability that do not represent the longer term performance of an FCCU. The commenters recommended that the EPA retain the daily averaging requirement.

Response: We disagree with the commenters' statement that reducing the averaging time from a 24-hour basis to a 3-hour basis for demonstrating compliance with the FCCU PM emission limit, using operating limits established during the performance test, is a change to the MACT floor. The emission limit of 1.0 lb PM/1,000 lbs coke burn-off is the MACT floor, and we are not changing the PM emissions limit (or alternate Ni limits) in Table 1 to subpart UUU (except to remove the incremental PM limit that did not comport with the MACT floor emissions limitation).

However, whether or not it is a change from the MACT floor is not relevant. Pursuant to CAA section 112(d)(6), the EPA must revise MACT standards “as necessary” considering developments in practices, processes and control technologies. For this Start Printed Page 75202exercise, we considered any of the following to be a “development”:

  • Any add-on control technology or other equipment that was not identified and considered during development of the original MACT standards.
  • Any improvements in add-on control technology or other equipment (that were identified and considered during development of the original MACT standards) that could result in additional emissions reduction.
  • Any work practice or operational procedure that was not identified or considered during development of the original MACT standards.
  • Any process change or pollution prevention alternative that could be broadly applied to the industry and that was not identified or considered during development of the original MACT standards.
  • Any significant changes in the cost (including cost effectiveness) of applying controls (including controls the EPA considered during the development of the original MACT standards).

In determining whether there are “developments,” we review, among other things, EPA regulations promulgated after adoption of the MACT, such as the NSPS we identified in this instance. We identified the enhanced monitoring requirements for these operating limits as a development in practices that will help ensure FCCU owners or operators are properly operating control devices and, thus, are meeting the PM emission limit at all times. We further determined that this enhanced monitoring was cost effective and proposed that it was necessary to revise the existing standard pursuant to CAA section 112(d)(6).

While we do not have continuous PM emissions data that show actual deviations of the PM limit, we do not need such data in order to conclude that such deviations could occur when daily averages are used. The Refinery MACT 2 (i.e., subpart UUU) rule requires owners or operators to establish operating limits based on three 1-hour runs during the performance test. As a matter of simple mathematics, a source could demonstrate that it is meeting the operating limit based on a 24-hour average but could be exceeding the 1.0 lb PM/1,000 lbs coke burn-off emission limit based on a 24-hour average or for one or more individual 3-hour periods during that 24-hour average. For example, an owner or operator could operate with a power input 5-percent higher than the operating limit for 23 hours, have the ESP off (zero power) for one hour, and still comply with a 24-hour average operating limit. However, it would be difficult for this same unit to meet the 1.0 lb PM/1,000 lbs coke burn-off emissions limit over a 24-hour period, and it certainly would not meet the limit for every 3-hour period during that day. As the operating limit can be established to correspond with 1.0 lb PM/1,000 lbs coke burn-off, the 5-percent higher power input would likely correspond with a 0.95 lb PM/1,000 lbs coke burn-off emissions rate (5-percent lower). Uncontrolled emissions are typically 6 to 8 lbs/1,000 lbs coke burn-off. Thus, this unit would have emissions averaging approximately 1.2 lbs PM/1,000 lbs coke burn-off during this 24-hour period [i.e., (0.95*23+7)/24], but would be in compliance with the 24-hour average operating limit. The unit would obviously also be out of compliance with the 3-hour average over the period when the power was turned off. We also have concerns that the operating limits are not always linear with the emissions, so that the longer averaging times do not effectively ensure compliance with the PM emissions limit. Therefore, as proposed, we are finalizing the requirement for owners or operators to comply with the operating limits on a 3-hour basis, rather than the 24-hour basis currently in the rule.

Comment: The technology review for FCCUs resulted in the EPA proposing to remove the 30-percent opacity alternative limit for demonstrating compliance with the PM emissions limit that is available for refineries complying with the Refinery NSPS 40 CFR part 63, subpart J. Two commenters supported the EPA's proposed removal of the 30-percent opacity limit for FCCUs. Other commenters stated that current technology is good enough for a 10- or 20-percent opacity limit. On the other hand, several commenters stated that the proposed removal of the 30-percent opacity limit must meet the criteria specified in CAA section 112(d)(6) and (f)(2), which requires analysis of the statutory basis, environmental impacts, costs, operational and compliance feasibility and impacts, that the EPA has not conducted. The commenters claimed that had the EPA conducted a proper analysis, the EPA would have determined that the proposed change to remove the 30-percent opacity limit is not necessary or supportable. Additionally, these commenters stated that since the underlying PM emissions limit is unchanged, there is no emission reduction justification for this proposed change, and the change would not meet the CAA section 112(d)(6) requirement of being cost effective. The commenters also noted that processes or practices for existing FCCUs have not changed, as required for a CAA section 112(d)(6) revision.

Several commenters urged the EPA to maintain the 30-percent opacity limit for these FCCUs. As a practicable and cost-effective alternative to address the EPA's concern as to whether compliance with a 30-percent opacity limit ensures compliance with the PM emissions limit, commenters suggested annual performance tests to confirm that the FCCU is meeting the PM emissions limit, rather than performance tests every 5 years, as proposed.

One commenter stated that the EPA never intended for the opacity limit in Refinery NSPS subpart J to be used to demonstrate compliance with the PM emissions limit, but instead to assure the PM controls operate properly. The commenter stated that the EPA's conclusion that the 30-percent opacity limit may not be sufficiently stringent to ensure compliance with the underlying PM emissions limit is based on a false premise as to the purpose of the opacity standard because as the EPA states, “Opacity of emissions is indicative of whether control equipment is properly maintained and operated.”

Several commenters stated that the proposed elimination of the 30-percent opacity limit currently in Refinery MACT 2 leaves existing FCCUs that use cyclones with no viable alternative approach to demonstrate compliance with the PM emissions limit without adding or replacing controls. They stated the other approaches for demonstrating compliance with the PM emissions limit in Refinery MACT 2 (such as development of a site-specific opacity limit) do not work for them. The commenters stated that although they believe that more frequent performance tests would show that the FCCUs are in fact meeting the PM emissions limit, the absence of the 30-percent opacity limit would force FCCUs using cyclones for PM control to install additional, costly PM controls (e.g., ESPs or wet gas scrubbers). They projected that these additional controls would cost tens of millions of dollars per FCCU and would require at least 3 years of compliance time. Additionally, one commenter stated that even FCCUs with additional downstream PM controls would not be able to achieve a site-specific limit at all times and needed the availability of the alternative 30-percent opacity limit. One commenter estimated that installing an ESP to meet the proposed 10-percent opacity limit would cost approximately $121,000/ton, assuming a 32 tpy PM emission reduction. The commenter noted that the ESP would also increase GHG emissions and require more energy Start Printed Page 75203resources from the facility. The commenter concluded that installing an ESP is neither cost effective nor appropriate considering non-air quality environmental and health impacts and energy requirements, and recommended that the EPA maintain the current NSPS subpart J alternative limits and add additional alternative limits into Refinery MACT 2 only as optional limits for demonstrating compliance with the PM emissions limit.

Response: In promulgating Refinery MACT 2, the EPA identified the 1.0 lb PM/1,000 lbs coke burn-off limit as the MACT floor but allowed a compliance option for FCCUs subject to Refinery NSPS subpart J to comply with an opacity limit up to 30 percent with one 6-minute allowance to exceed the 30-percent opacity in any 1-hour period. As stated in the proposal, compliance studies have shown that the 30-percent opacity limit does not correlate well with the 1.0 lb PM/1,000 lbs coke burn-off limit, and that an FCCU can comply with the 30-percent opacity limit while its emissions exceed the PM emissions limit.[9] Regardless of whether the 30-percent opacity limit in Refinery NSPS subpart J was designed to “ensure that the control device was operated properly,” Refinery MACT 2 allows sources subject to NSPS subpart J to use the 30-percent opacity limit to demonstrate continuous compliance with the PM emissions limit. We have determined that the 30-percent opacity limit is inadequate for the purpose of demonstrating continuous compliance with the PM emissions limits in Refinery MACT 2. As such, we proposed to remove this opacity limit and require the owner or operator to either demonstrate compliance with the PM emissions limit by continuously monitoring the control device parameters established during the performance test or establish and monitor a site-specific opacity limit. For clarity, we note that we proposed to allow a site-specific opacity limit, not a 10-percent opacity limit as some commenters suggest. The site-specific opacity limit can be significantly higher than 10 percent, but it cannot be lower than 10 percent.

While the compliance study indicates that a 30-percent opacity limit does not correlate well with a 1.0 lb PM/1,000 lbs coke burn-off emissions limit, further review of this same study indicates that a 20-percent opacity limit provides a reasonable correlation with units meeting the 1.0 lb PM/1,000 lbs coke burn-off emissions limit. We also reviewed the data submitted by the commenters regarding PM emissions and opacity correlation. While the data suggest that there is variability and uncertainty in the PM/opacity correlation, the data do not support that a 30-percent opacity limit would ensure compliance even when considering the uncertainty associated with the PM/opacity correlation. Based on the variability of the 3-run average opacity limits, we determined that, if the 3-hour average opacity exceeded 20-percent, then it was highly likely (98 to 99-percent confidence) that the FCCU emissions from the unit tested would exceed the PM emissions limit.

After considering the public comments, reviewing the data submitted with those comments, and further review of the compliance study, in this final rule we are adding a 20-percent opacity limit, evaluated on a 3-hour average basis for units subject to NSPS subpart J. As we noted above, a 20-percent opacity limit provides a reasonable correlation with the PM emissions limit, and an exceedance of this 20-percent opacity limit will provide evidence that the PM emissions limit is exceeded. However, it is possible that units could still exceed the PM emissions limit while complying with the 20-percent opacity limit, if those units operate close to the 1 lb PM/1,000 lbs coke burn-off emissions limit. To address this concern, we considered the commenters' suggestion to require a performance test annually rather than once every 5 years. Some commenters suggested that this option specifically apply to FCCUs with cyclones, but this option is applicable to any control system operating very near the PM emissions limit and using an opacity limit to demonstrate continuous compliance. We have determined that the Refinery NSPS subpart J compliance procedures in Refinery MACT 2, in combination with a 20-percent opacity limit demonstrated on a 3-hour average basis and with annual performance tests when a test indicates PM emissions are greater than 80-percent of the limit (i.e., 0.80 lb PM/1,000 lbs coke burn-off), will ensure continuous compliance with the PM emissions limit. FCCUs with measured PM emissions during the performance test at or below 0.80 lb PM/1000 lbs of coke burn-off will remain subject to the requirement to conduct performance tests once every 5 years, consistent with the requirements we proposed.

We do not agree with commenters that the proposed opacity revision would add significant cost or compliance burden. The control device-specific monitoring parameters that were proposed rely on parameters commonly used to control the operation of the control device, so the monitoring systems should be already available. Further, since we are merely changing the opacity limit, we expect these units will already have opacity monitoring systems needed to demonstrate compliance with the PM emissions limit and would not incur costs for new equipment.

Comment: Several commenters stated that they agree with the EPA's determination in the proposal that the current CO limits provide adequate control of HCN. Two commenters stated that there are limited HCN emissions data and that more data are needed before the Agency can appropriately determine whether an HCN standard is necessary and justified. One commenter noted that the process undertaken by the EPA to estimate HCN emissions was flawed, and likely overestimates HCN emissions significantly. Another commenter stated that they performed HCN stack testing at three refineries and subsequent modeling at two refineries and concluded that the ambient HCN emissions were well below the applicable health limits.

In contrast, some commenters expressed concerns about high HCN levels. One commenter stated that the EPA should consider re-evaluating the benefit of low NOX emissions from the FCCU, if that is indeed the cause of higher HCN emissions, because exposing people to HCN is not acceptable. The commenter also noted that the community now also has the increased dangers of storing and transporting aqueous ammonia, which is used in some cases to achieve low NOX emissions from the FCCU.

One commenter stated that the EPA must set stronger HCN standards on FCCU emissions because of the high release amounts reported, the fact that non-cancer risk is driven by emissions of HCN from FCCU, and the fact that the EPA has never set standards for HCN emissions. The commenter provided a report that they believe shows that the EPA has not shown that CO is a reasonable or lawful surrogate to control HCN and has not shown that the conditions necessary for a surrogate are met with regard to CO and HCN, which is an inorganic nonmetallic HAP. Further, the report indicates that SCR is a reasonable and cost effective method for controlling HCN and that the EPA failed to review and consider other viable methods to control HCN and must do so to satisfy its legal obligations in this rulemaking.Start Printed Page 75204

Response: At the time we promulgated the MACT, we determined that the control strategy used by the best performing facilities to reduce organic HAP emissions was the use of complete combustion, which occurs when the CO concentration is reduced to 500 ppmv (see the proposal for Refinery MACT 2 at 63 FR 48899, September 11, 1998). We rejected arguments that some facilities operate at CO levels well below 500 ppmv and, thus, the MACT floor should be set at a lower CO concentration because once CO concentrations reached 500 ppmv, there was no longer a correlation between reduced CO concentrations and reduced HAP concentrations. And, in fact, emissions of certain HAP, such as formaldehyde, tended to increase as CO concentrations were reduced below 500 ppmv.[10]

In the current rulemaking action, we determined at the time of the proposed rule that this also holds true for HCN emissions. That is, once CO emissions are reduced to below 500 ppmv (i.e., complete combustion is achieved), we no longer see a direct correlation between CO concentrations and HCN emissions.

All of the HCN emissions data we have were reported from units operating at or below the 500 ppmv CO limit (i.e., in the complete combustion range), so it is not surprising that there is not a strong correlation between CO and HCN from the FCCU ICR source test data. However, catalyst vendor data and combustion kinetic theory support the fact that, in the partial burn mode (with CO concentrations of 2 to 6-percent, which is 20,000 to 60,000 ppmv), HCN concentrations exiting the FCCU regenerator are much greater than for units using complete combustion FCCU regenerators or the concentration exiting a post-combustion device used in conjunction with a partial burn FCCU regenerator. Therefore, we maintain that complete combustion is the primary control needed to achieve controlled levels of HCN emissions.

We initially thought the higher levels of HCN emissions that were reported by sources achieving complete combustion might be due to a switch away from platinum-based combustion promoters to palladium-based combustion promoters. However, many of the units that were tested and that had some of the lowest HCN emissions used palladium-based oxygen promoters. Therefore, it appears unlikely that palladium-based catalyst promoters are linked to the higher HCN emissions. We also evaluated one commenter's argument that CO is not a good surrogate for HCN emissions, but that SCR are a reasonable and cost-effective control strategy. We are not aware of any data that suggest that an SCR removes HCN and the commenter did not provide any support for that premise. At proposal, we evaluated HCN control on units using extra oxygen or converting back to platinum-based promoters to oxidize any HCN formed. This would cause more NOX formation, which would then require post-combustion NOX control, such as an SCR. However, if HCN emissions are not a function of CO concentration beyond that required to achieve complete combustion (as noted by the commenter), then more aggressive combustion conditions and the use of an SCR (to remove the NOX formed) may not be a viable control strategy. Therefore, considering all of the data currently available and the comments received regarding HCN emissions and controls, we maintain that the only proven control technique is the use of complete combustion as defined by a CO level of 500 ppmv or less. We are not establishing a more stringent CO level because, once complete combustion is achieved, (i.e., CO concentrations drop below 500 ppmv), no further reduction in HCN emissions are achieved.

For the purposes of Refinery MACT 2, we consider the emission limits and operating requirements for organic HAP in Tables 8 through 14 to subpart UUU of part 63 adequate to also limit HCN emissions.

Finally, we understand concerns about the reported HCN emissions being higher than anticipated and the need for more data to better determine HCN emissions levels. To address these concerns, we are finalizing a requirement that facility owners or operators conduct a performance test for HCN from all FCCU at the same time they conduct the first PM performance test on the FCCU following promulgation of this rule. Facility owners or operators that conducted a performance test for HCN from a FCCU in response to the refinery ICR or subsequent to the 2011 Petroleum Refinery ICR following appropriate methods are not required to retest that FCCU.

4. What is the rationale for our final approach for the technology review?

a. Refinery MACT 1

We did not receive substantive comments concerning our proposal that it was not necessary to revise Refinery MACT 1 requirements for MPV, gasoline loading racks and cooling towers/heat exchange systems. Based on the rationale provided in the preamble to the proposed rule, we are taking final action concluding that it is not necessary pursuant to CAA section 112(d)(6) to revise the MACT requirements for MPV, gasoline loading racks and cooling towers/heat exchange systems emission sources at refineries.

We proposed that the options for additional wastewater controls are not cost effective and thus it was not necessary to revise the MACT for these emission sources. We received public comments suggesting that emissions from wastewater systems are higher than modeled and that we should develop additional technology standards for wastewater treatment systems regardless of cost. As we discussed in the proposal, emissions from wastewater are difficult to measure and emission estimates rely on process data and empirical correlations, which introduces uncertainty into the estimates. Although we do not have evidence, based on the process data we collected, that emissions are higher than modeled at proposal, we note that the fenceline monitoring program effectively ensures that wastewater emissions are not significantly greater than those included in the emissions inventory and modeled in the risk assessment. Furthermore, we believe that cost is a valid consideration in determining whether it is necessary within the meaning of section 112(d)(6) to revise requirements and that we are not required to establish additional controls regardless of cost. Consequently, we conclude that it is not necessary to revise the Refinery MACT 1 requirements for wastewater systems pursuant to CAA section 112(d)(6).

For storage vessels, we identified a number of options, including requiring tank fitting controls for external and internal floating roof tanks, controlling smaller tanks with lower vapor pressures and requiring additional monitoring to prevent roof landings, liquid level overfills and to identify leaking vents as developments in practices, processes and control technology. We proposed to cross-reference the storage vessel requirements in the Generic MACT (effectively requiring additional control for tank roof fittings) and to revise the Start Printed Page 75205definition of Group 1 storage vessels to include smaller tanks with lower vapor pressures. We received comments that we could have required additional controls on tanks and monitoring for landings, overfills and leaking vents described above. We also received comments related to clarifications of specific rule references and overlap provisions. We addressed these comments in the “Response to Comments” document, and we maintain that the additional control options described by the commenters (tank roof landing/degassing requirements or use of geodesic domes to retrofit external floating roofs) are not cost-effective. Consequently, based on the rationale provided in the preamble to the proposed rule and our consideration of public comments, we are finalizing the requirements as proposed with minor clarifications of the rule references. However, as with wastewater systems, we note that the fenceline monitoring program will ensure that the owner or operator is effectively managing fugitive emissions sources and should detect landings, overfills, and leaking vents.

For equipment leaks, we identified specific developments in practices, processes and control technologies that included requiring repair of leaking components at lower leak definitions, requiring monitoring of connectors, and allowing the use of the optical imaging camera as an alternative method of monitoring for leaks. We proposed to establish an alternative method for refineries to meet LDAR requirements in Refinery MACT 1. This alternative would allow refineries to monitor for leaks via optical gas imaging in place of EPA Method 21, using monitoring requirements to be specified in a not yet proposed appendix K to 40 CFR part 60. However, the development of appendix K is taking longer than anticipated. Therefore, we are not finalizing this alternative monitoring method in Refinery MACT 1.

We received comments suggesting that additional requirements be imposed to further reduce emissions from leaking equipment components, such as requiring “leakless” equipment, reducing the leak threshold, and eliminating delay of repair provisions. As provided in the “Response to Comments” document, we do not agree that these additional requirements are cost-effective. Based on the rationale provided in the preamble to the proposed rule and our consideration of public comments, we conclude that it is not necessary to revise the Refinery MACT 1 requirements for equipment leaks. Again, however, the fenceline monitoring program is intended to ensure that large leaks from fugitive emissions sources, including equipment leaks, are more quickly identified and repaired, thereby helping to reduce emissions from leaking equipment components.

For marine vessel loading, we identified control of marine vessel loading operations with HAP emissions of less than 10/25 tpy and the use of lean oil absorption systems as developments that we considered in the technology review. We proposed to amend 40 CFR part 63, subpart Y to require small marine vessel loading operations (i.e., operations with HAP emissions less than 10/25 tpy) and offshore marine vessel loading operations to use submerged filling based on the cargo filling line requirements in 46 CFR 153.282. We received comments that other options considered during the technology review of the standard were cost-effective for small marine vessel loading operations and should be required. As provided in the “Response to Comments,” we continue to believe those other controls are not cost-effective because of the high costs of controls for limited additional organic HAP emission reduction. Therefore, we are finalizing these amendments as proposed.

Finally, we proposed that it was necessary to revise the MACT to require fenceline monitoring as a means to manage fugitive emissions from the entire petroleum refinery, which includes sources such as wastewater collection and treatment operations, equipment leaks and storage vessels. We received numerous comments regarding the proposed requirement to conduct fenceline monitoring, many of which we address above and the remainder of which we respond to in the “Response to Comments” document. After considering comments, we maintain that the proposed work practice standard is authorized under section 112 of the CAA and will improve fugitive management at the refinery. Therefore, we are finalizing the key components of fenceline monitoring work practice as proposed. These requirements include the use of passive diffusive tube samplers (although we are providing a mechanism to request approval for alternative monitoring systems provided certain criteria are met), the 9 μg/m3 on a rolling annual average basis action level, and the need to perform corrective action to comply with the action level.

Based on public comments received, we are making numerous revisions to clarify the fenceline monitor siting requirements. This includes provisions to allow siting of monitors within the property boundary as long as all emissions sources at the refinery are included within the monitoring perimeter. We are also clarifying that we do not consider public roads or public waterways that bisect a refinery to be property boundaries, and owners or operators do not need to place monitors along the internal public right-of-ways. We are also providing provisions to allow fixed placement of monitors at 500 feet intervals (with a minimum of 3 monitors) for subareas or segregated areas. If an emissions source is near the monitoring perimeter, an additional monitor siting requirement would still apply. The 500 feet provision is provided to reduce burden for facilities with irregular shapes or noncontiguous property areas that we did not fully consider at proposal.

We also received comments on the compliance time and reporting requirements associated with the fenceline monitoring provisions. Upon consideration of public comments, we have revised the compliance period to 2 years after the effective date of the final rule. Thus, beginning no later than 2 years after the effective date of the rule, the source must have a fenceline monitoring system that is collecting samples such that the first rolling annual average ΔC value would be completed no later than 3 years after the effective date of the final rule. Facilities will have 45 days after the completion of the first year of sampling, as proposed, to submit the initial data set. We are reducing the proposed compliance period from 3 years to 2 years because the passive diffusive tube monitors are easy to deploy and pilot study demonstrations indicate that significant time is not needed to deploy the monitors. However, the reduced compliance period still provides time to resolve site-specific monitor placement issues and to provide time to develop and implement a site-specific monitoring plan, if needed. We are increasing the fenceline monitoring reporting frequency (after the first year of data collection) from semiannually to quarterly to provide more timely dissemination of the data collected via this monitoring program.

b. Refinery MACT 2

We proposed to revise Refinery MACT 2 to incorporate the developments in monitoring practices and control technologies reflected in the Refinery NSPS subpart Ja limits and monitoring provisions (73 FR 35838, June 24, 2008). We are finalizing most of these provisions as proposed. Specifically, we are incorporating the Start Printed Page 75206Refinery NSPS subpart Ja PM limit for new FCCU sources. We are also finalizing compliance options for FCCU that are not subject to Refinery NSPS subpart J or Ja. These options would allow such sources to elect to comply with the Refinery NSPS subpart Ja monitoring provisions to demonstrate compliance with the emissions PM limit. We are revising the averaging period for the control device operating limits or site-specific opacity limits to be on a 3-hour average basis in order to more directly link the operating limit to the duration of the performance test runs, on which they are based, as proposed. We are incorporating additional control device-specific monitoring alternatives for various control devices on FCCU, including BLD monitoring as an option to COMS for owners or operators of FCCU using fabric filter-type control systems and total power and secondary current operating limits for owners or operators of ESPs. We are adding an additional requirement to perform daily checks of the air or water pressure to atomizing spray nozzles for owners or operators of FCCU wet gas scrubbers not subject to the pressure drop operating limit, as proposed. Finally, we finalizing requirements to conduct a performance test at least once every 5 years for all FCCU, as proposed. These requirements are being finalized to ensure that control devices are continuously operated in a manner similar to the operating conditions of the performance test and to ensure that the emissions limits, which are assessed based on the results of three 1-hour test runs, are achieved at all times.

We also proposed to eliminate the Refinery NSPS subpart J compliance option that allows refineries to meet the 30-percent opacity emissions limit requirement and revise the MACT to include control device operating limits or site-specific opacity limits identical to those required in Refinery NSPS subpart Ja. We received numerous comments, particularly from owners or operators of FCCU that employ tertiary cyclones to control FCCU PM emissions. According to the commenters, opacity is not a direct indicator of PM emissions because finer particles will increase opacity readings without a corresponding mass increase in PM emissions. Additionally, the commenters stated that the site-specific opacity limit generally leads to a site-specific operating limit of 10-percent opacity, which is too stringent and does not adequately account for variability between PM emissions and opacity readings. According to the commenters, FCCU with tertiary cyclones would need to be retrofitted with expensive and costly controls in order to meet the 10-percent opacity limit, even though they are meeting the 1 lb/1000 lbs coke burn PM emissions limit. It was not our intent to require units to retrofit their controls simply to meet the site-specific opacity limit. However, the existing 30-percent opacity limit in the subpart J compliance option is not adequate to ensure compliance with the PM emissions limit at all times. After reviewing the public comments and available data, we determined that, rather than removing the subpart J compliance option altogether, it is sufficient to add an opacity operating limit of 20-percent opacity determined on a 3-hour average basis to the existing subpart J compliance option and to require units complying with this operating limit to conduct annual performance tests (rather than one every 5 years) when the PM emissions measured during the source test are greater than 0.80 lb PM/1,000 lbs coke burn-off. These provisions improve assurance that these units are, in fact, achieving the required PM emissions limitation without requiring units to retrofit controls due to variability in the correlation of PM emissions and opacity.

We did not propose to revise the organic HAP emissions limits for FCCU to further address HCN emissions. We received numerous comments on this issue. We continue to believe that complete combustion is the appropriate control needed to control HCN emissions. Consequently, for the purposes of Refinery MACT 2, we are not changing the MACT standards to further reduce emissions of HCN. However, we understand that there are uncertainties and high variability in HCN emissions measured from FCCU. In order to address the need for more data to better characterize HCN emissions levels, we are finalizing a requirement for refinery owners or operators to conduct a performance test for HCN from all FCCU (except those units that were tested previously using acceptable methods as outlined in the 2011 Refinery ICR) during the first PM test required as part of the on-going compliance requirements for FCCU metal HAP emissions. These data will be useful to the EPA in understanding HCN emissions from FCU and may help to inform future regulatory reviews for this source category.

We proposed that there have been no developments in practices, processes, and control technologies for CRU based on our technology review and that therefore it is not necessary to revise these standards. Based on the rationale provided in the preamble to the proposed rule and our consideration of public comments, we are finalizing our conclusion.

For SRU, we identified the Refinery NSPS subpart Ja allowance for oxygen-enriched air as a development in practice and we proposed that it was necessary to revise the MACT to allow SRU to comply with Refinery subpart Ja as a means of complying with Refinery MACT 2. The key issue identified by commenters was that Refinery NSPS subpart Ja includes a flow monitoring alternative for determining the average oxygen concentration in the enriched air stream and that this was not included in the proposed amendments to Refinery MACT 2. This was an oversight on our part. We are, based on the rationale provided in the preamble to the proposed rule and our consideration of public comments, finalizing the SRU revisions as proposed but with inclusion of the flow monitoring alternative provisions that are in Refinery NSPS subpart Ja for this source.

C. Refinery MACT Amendments Pursuant to CAA Section 112(d)(2) and (d)(3)

1. What did we propose pursuant to CAA section 112(d)(2) and (d)(3) for the Petroleum Refinery source categories?

We proposed the following revisions to the Refinery MACT 1 and 2 standards pursuant to CAA section 112(d)(2) and (3) [11] : (1) Adding MACT standards for DCU decoking operations; (2) revising the CRU purge vent pressure exemption; (3) adding operational requirements for flares used as APCD in Refinery MACT 1 and 2; and (4) adding requirements and clarifications for vent control bypasses in Refinery MACT 1.

For DCU, we proposed to require that prior to venting or draining, each coke drum must be depressured to a closed blowdown system until the coke drum vessel pressure is 2 psig or less. As proposed, the 2 psig limit would apply to each vessel opening/venting/draining event at new or existing affected DCU facilities.

For the CRU, we proposed to require that any emissions during the active Start Printed Page 75207purging or depressuring of CRU vessels meet the applicable organic HAP emission limitations in Tables 15 and 16 to subpart UUU regardless of the vessel pressure.

For flares, we proposed to remove cross references to the General Provisions requirements for flares used as control devices at 40 CFR 63.11(b) and to incorporate enhanced flare operational requirements directly into the Refinery MACT rules. The proposed rule amendments included:

  • A ban on flaring of halogenated vent streams.
  • A requirement to operate with continuously lit pilot flames at all times and to equip the pilot system with an automated device to relight the pilot if it is extinguished.
  • A requirement to operate with no visible emissions except for periods not to exceed a total of 5 minutes during any 2 consecutive hours and to monitor for visible emissions daily.
  • A requirement to operate with the flare tip velocity less than 60-feet-per-second or the velocity limit calculated by an equation provided in the proposed rule.
  • A requirement to meet one of three combustion zone gas properties operating limits based on the net heating value, lower flammability limit, or combustion concentration. Owners or operators could elect to comply with any one of the three limits at any time. Two separate sets of operating limits were proposed: One for gas streams not meeting all three “hydrogen-olefin interaction criteria” specified in the rule and a more stringent set of limits for gas streams meeting all three hydrogen-olefin interaction criteria. The combustion zone net heating value considered steam assist rates but not “perimeter air” assist rates.
  • For air-assisted flares, a requirement to meet an additional “dilution parameter” operating limit determined based on the combustion zone net heating values above, the diameter of the flare and the perimeter air assist rates.

The proposed amendments for flares also included detailed monitoring requirements to determine these operating parameters either through continuous parameter monitoring systems or grab sampling, detailed calculation instructions for determining these parameters on a 15-minute block average, and detailed recordkeeping and reporting requirements. We also proposed provisions to allow owners or operators to request alternative emissions limitations that would apply in place of the proposed operating limits.

We proposed to revise the definition of MPV to remove the current exclusion for in situ sampling systems (onstream analyzers). We also proposed to limit the exclusion for gaseous streams routed to a fuel gas system to apply only to those systems for which any flares receiving gas from the fuel gas system are in compliance with the proposed flare monitoring and operating limits. We note that we also proposed revisions related to monitoring of bypass lines, but these revisions were proposed to address concerns related to SSM releases and are described in further detail in section IV.D. of this preamble.

We proposed that emissions of HAP may not be discharged to the atmosphere from PRD in organic HAP service to address concerns related to SSM releases. To ensure compliance with this proposed amendment, we proposed to require that sources monitor PRD using a system that is capable of identifying and recording the time and duration of each pressure release and of notifying operators that a pressure release has occurred. This proposed requirement was addressed in section IV.A.4. of the preamble for the proposal.

2. How did the revisions pursuant to CAA section 112(d)(2) and (3) change since proposal?

We proposed identical standards for existing and new DCU decoking operations, but we are finalizing standards for new and existing sources that are not identical. We are finalizing provisions that will require owners or operators of existing DCU sources to comply with a 2 psig limit averaged over 60 cycles (i.e., 60 venting events), rather than meet the 2 psig limit on a per venting event basis, as proposed. We are finalizing provisions that will require owners or operators of new DCU sources to comply with a 2.0 psig limit on a per event, not-to-exceed basis. We are adding one significant digit to the limit for new DCU affected sources because our re-review of permit requirements conducted in response to comments identified that the best performing DCU source is required to comply with a 2.0 psig limit on a per event basis. In response to comments regarding the proposed prohibition on draining prior to achieving the pressure limit, we are finalizing specific provisions for DCU with water overflow design and for double quenching.

For flares, we are not finalizing the ban that we proposed on halogenated vent streams and we are not finalizing the proposed requirement to equip the flare pilot system with an automated device to relight an extinguished pilot.

We are revising the MACT to include the proposed no visible emissions limit and the flare tip velocity limit as direct emissions limits only when the flare vent gas flow rate is below the smokeless capacity of the flare. Under the revised standard, when the flare is operating above the smokeless capacity, an exceedance of the no visible emission limit and/or flare tip velocity limit is not a violation of the standard but instead triggers a work practice standard. Flares operate above the smokeless capacity only when there is an emergency release event and thus the work practice standard is intended to address emissions during such emergency release events. (See section IV.D. of this preamble for more details regarding this work practice standard). We are also adding provisions that would allow sources to use video surveillance of the flare as an alternative to daily Method 22 visible emissions observations.

For flares, we are also simplifying the combustion zone gas property operating limits by finalizing a requirement only for the net heating value of the combustion zone gas. We are finalizing requirements that flares meet a minimum operating limit of 270 BTU/scf NHVcz on a 15-minute average, as proposed, and we are allowing refinery owners or operators to use a corrected heat content of 1212 BTU/scf for hydrogen to demonstrate compliance with this operating limit. We are not finalizing separate combustion zone operating limits for gases meeting the hydrogen-olefin interaction criteria that were proposed. We are also not finalizing the alternative combustion zone operating limits based on lower flammability limit or combustibles concentration.

We are finalizing “dilution parameter” requirements for air-assisted flares, but we are providing a limit only for the net heating value dilution parameter. Similar to the requirements we are finalizing for the combustion zone parameters, we are finalizing requirements that flares meet a minimum operating limit of 22 BTU/ft2 NHVdil on a 15-minute average, as proposed, and we are allowing refinery owners or operators to use a corrected heat content of 1,212 BTU/scf for hydrogen to demonstrate compliance with this operating limit. We are not finalizing separate dilution parameter operating limits for gases meeting the hydrogen-olefin interaction criteria that were proposed. We are also not finalizing the alternative dilution parameter operating limits based on lower flammability limit or combustibles concentration.Start Printed Page 75208

We are providing an alternative to use initial sampling period and process knowledge for flares in dedicated service as an alternative to continuous or on-going grab sample requirements for determining waste gas net heat content.

We are finalizing revisions to the definition of MPV, as proposed.

We are establishing work practice standards that apply to PRD releases in place of the proposed prohibition on PRD releases to the atmosphere. The work practice standards that we are finalizing for PRD require refiners to establish proactive, preventative measures for each PRD to identify and correct direct releases of HAP to the atmosphere as a result of pressure release events. Over time, these proactive measures will reduce the occurrence of releases and the magnitude of releases when they occur, while avoiding the environmental disbenefits of having additional flare capacity on standby to control these unpredictable and infrequent events. Refinery owners or operators will be required to perform a root cause analysis/corrective action following such pressure release events. In addition, a second release event in a 3-year period from the same PRD with the same root cause on the same equipment is a deviation of the work practice standard. A third release event in a3-year period from the same PRD is a deviation of the work practice standard regardless of the root cause. PRD release events related to force majeure events are not considered in these hard limits.

3. What key comments did we receive on the proposed revisions pursuant to CAA section 112(d)(2) and (3) and what are our responses?

i. DCU

Comment: Several commenters argued that the EPA incorrectly set the MACT floor emission limitation for DCU. Commenters noted that CAA section 112(d)(3)(A) states that the MACT limit for existing sources “shall not be less stringent, and may be more stringent than the average emission limitation achieved by the best performing 12-percent of the existing sources” excluding those first achieving that level within 18 months prior to proposal or 30 months prior to promulgation, whichever is later. According to the commenters, the EPA failed to follow this procedure in setting the 2 psig vent limit as a MACT floor because the EPA incorrectly considered permit limits and other non-performance based criteria instead of basing the MACT floor on the actual performance of sources. Commenters stated that the EPA improperly considered permit limits that should have been excluded from consideration, as well as considering permit limits for closed facilities instead of using more accurate data from operating DCUs at sources that submitted actual emissions data. Specifically, commenters stated that the DCU at the non-operational plant (Hovensa) should not be included. One commenter noted that they operate one of the South Coast DCU listed as subject to a 2 psig limit and asserted that it does not currently meet that emission limitation. The commenter claimed that significant capital investment would be required in order for the DCU to comply with the 2 psig limit. According to one commenter, data for six of the eight DCU they claim the EPA considered for the MACT floor should not be counted in determining the limit that represents the average emission limitation actually achieved 18 months prior to the proposal.

Response: CAA section 112(d)(3)(A) states that the existing source standard shall not be less stringent than the average emission limitation achieved by the best performing 12-percent of the existing sources (for which the Administrator has emissions information), excluding those sources that have, within 18 months before the emission standard is proposed or within 30 months before such standard is promulgated, whichever is later, first achieved a level of emission rate or emission reduction which complies, or would comply if the source is not subject to such standard, with the lowest achievable emission rate (as defined by section 171) applicable to the source category and prevailing at the time, in the category or subcategory for categories and subcategories with 30 or more sources. We consider a 2 psig emissions limitation to be equivalent to the lowest achievable emission rate (LAER) emission limits. Thus, we agree with the commenter that sources that first meet the 2 psig limit on or after December 30, 2012, should be excluded from the MACT floor analysis. We also agree that under CAA section 112(d)(3)(A), the MACT floor analysis focuses on those sources that are achieving the emission limit (i.e., the emission limitation “achieved by. . . ”). The EPA has previously determined that the 6th-percentile unit is a reasonable estimate of the average emission limitation achieved by the best performing 12-percent of sources especially when averaging across units with and without control requirements. As noted in our DCU MACT floor analysis memorandum (Docket ID No. EPA-HQ-OAR-2010-0682-0203), the 6th-percentile is represented by the fifth-best performing DCU. If we exclude the two South Coast refineries and the two Marathon Garyville DCU because these sources were not implementing the 2 psig permit limit prior to December 30, 2012, the fifth-best performing DCU would be represented by the Bay Area refineries (4.6 psig). However, based on the 2011 Petroleum Refinery ICR responses, 25 out of 75 (33-percent) DCU have a “typical coke drum pressure when first vented to the atmosphere” of 2 psig or less and 10 out of 75 (13-percent) DCU have a “typical coke drum pressure when first vented to the atmosphere” of 1 psig or less. While we acknowledge that these data represent “typical” operations and not necessarily a never-to-be-exceeded emissions limitation, we conclude that this information is sufficient for us to conclude that the average emission limitation achieved by the best performing 12-percent of sources is consistent with a 2 psig emissions limitation. This is because facility owners or operators commonly target to operate at approximately half the allowable emissions limit to ensure that they can comply with the emissions limit at all times. Therefore, we maintain that an average venting pressure of 2 psig is the MACT floor level for decoking operation at existing sources based on the ICR responses and considering the average performance expected.

Comment: Four commenters suggested that the 2 psig limit, if finalized, should be based on a rolling 30-day average per DCU rather than a never to be exceeded “instantaneous” standard. According to the commenters, an instantaneous standard is unnecessary to address HAPs with chronic health impacts and adds cost and compliance challenges. According to the commenters, chronic health impacts are not materially affected by short-term variability, but instead depend on the average concentration of exposure over a 70-year lifetime; therefore, there is no health based or environmental reason for requiring an instantaneous limit. The commenters noted that there would be additional capital costs to comply with a 2 psig not-to-be-exceeded limit compared to a 30-day average 2 psig limit vent pressure. One commenter specifically requested that the EPA also confirm that a pressure of 2.4 psig is compliant with the 2 psig limit vent pressure. Another commenter also requested clarification that the vent pressure can be rounded to Start Printed Page 75209one significant figure when determining compliance.

Response: For new sources, the MACT floor emission limit for DCU is based on the best-performing source. Based on this and other comments received, we again reviewed existing permit conditions. Based on this review, we found that one of the permit requirements specified the pressure limit as 2.0 psig for each coke drum venting event. Therefore, we are finalizing the new source MACT floor as 2.0 psig on a per coke drum venting event basis.

As discussed in response to the previous comment, we are basing the MACT floor for existing source DCU on responses we received from the 2011 Petroleum Refinery ICR. Because the ICR requested the “typical coke drum pressure when first vented to the atmosphere,” we do not consider the information provided in ICR responses to reflect a “never-to-be-exceeded” limit. Therefore, we evaluated whether it is reasonable to allow averaging, and if so, what averaging period should be provided.

Health risks are not considered in establishing MACT requirements, so we do not consider the argument that chronic effects are evaluated over a 70-year period to be relevant to a determination of the MACT floor. However, a primary consideration regarding averaging periods is how the averaging period was considered in setting the floor and whether the intended reductions will occur under a different averaging period. According to the heat balance method for estimating DCU emissions, DCU decoking operations emissions are directly proportional to the average bed temperature. While the relationship is not exactly linear, the average bed temperature is expected to be a function of the venting pressure. Moreover, the shape of the pressure-temperature correlation curve is such that the emissions at 6 psig are almost exactly but not quite three times the emissions at 2 psig. Given the expected linearity of the emissions with venting pressures, we are not concerned with an occasional venting event above 2 psig because the average emissions from a facility meeting an average 2 psig pressure limit would be identical to the emissions achieved by a facility that vented each time at 2 psig. That is, given the expected linearity in the projected DCU emissions to the venting pressure, we conclude that it is reasonable to allow averaging across events and that the precise averaging period is not a critical concern.

Most industry commenters requested a 30-day average. However, different facilities have different numbers of DCU, different numbers of drums per DCU and different cycle times. Consequently, basing the averaging period across a given time period would result in significantly different number of venting events included in a 30-day average for different facilities and generally provide more flexibility to larger refineries and less flexibility to smaller refineries. Based on the ICR responses, almost half of all DCU operate with two drums and about 90-percent of DCU have two to four coke drums; however, a few DCU have six or even eight drums. Also, based on the ICR responses, the average complete coke drum cycle time is 32 hours, but can be as short as 18 hours and as long as 48 hours. Reviewing the ICR responses, we found that a 30-day average would include 30 events for some facilities and more than 250 events at other facilities.

Since the existing source MACT standards apply “in combination” to “all releases associated with decoking operations” at a given facility, we determined that it was reasonable to consider an averaging period that applies to the number of venting events from all coke drums at the facility rather than to all coke drums for a specific DCU for a specified period of time. This provides a more consistent basis for the averaging period and allows the same operational flexibility for small refineries as large refineries. Based on the ICR responses, the median (typical) DCU has 60 venting events in a 30-day period. Providing an averaging period of 60 venting events provides a more consistent averaging basis for all facilities, regardless of the number of DCU at the facility and the number of drums and cycle times for different DCU. Additionally, it eliminates issues with respect to how to handle operating days versus non-operating days, e.g., in the event of a turn-around resulting in a limited number of venting events in a 30-calendar day period. Therefore, we are establishing a 2 psig limit based on a 60-event average considering all coke drum venting events at an existing source and we are finalizing a 2.0 psig limit on a per coke drum venting event for DCU at new sources.

We have consistently maintained our policy to round to the last digit provided in the emission limit, a pressure of 2.4 psig would round to 2 psig and would be compliant with a requirement to depressure each coke drum to a closed blowdown system until the coke drum vessel pressure is 2 psig or less, but it would not be compliant with the revised new source provision to depressure until the coke drum vessel pressure is 2.0 psig or less. A coke drum pressure of 2.04, however, would be compliant with the revised new source requirement pressure limit of 2.0 psig.

ii. Refinery Flares

Comment: Several commenters suggested that the proposed flare operating limits were too complex. The commenters recommended that the EPA eliminate the dual flare combustion zone heat content limits related to the proposed hydrogen-olefin interaction criteria and instead finalize a single combustion zone net heating value of approximately 200 BTU/scf, which would minimize the unnecessary burning of supplemental gas but still ensure good combustion efficiency.

A few commenters suggested that the EPA based the proposed combustion zone limits on an invalid data analysis, that the 1 minute PFTIR data should not be used to establish combustion efficiency correlations, and that the emission limits should be set so as to provide an equal chance of false positives and negatives. A few commenters suggested that the EPA should assign hydrogen a heating value of 1,212 BTU/scf to more accurately reflect its flammability in a NHV basis and that doing so is consistent with some recent flare consent decrees and would help reduce natural gas supplementation for facilities complying only with the NHVcz metric.

Several commenters suggested that neither scientific literature nor the available flare test data support the EPA's claim of an adverse hydrogen-olefin interaction on combustion efficiency and that the EPA should not finalize the more restrictive combustion zone operating limits for all flare types. These commenters suggested that the EPA did not provide any evidence the assumed hydrogen-olefin effect actually exists; that statistical analysis demonstrates the EPA developed their limit based on random differences in data; that the PFTIR data analysis method of using the individual minute-by-minute data instead of the test average data is flawed and leads to invalid conclusions; and that proper analysis of the data demonstrates the more stringent operating limits for hydrogen-olefin conditions cannot be supported.

Some commenters suggested that there is evidence to support more stringent flare combustion zone limits for a narrowly defined high concentration propylene-only condition as outlined in some of the recent flare Start Printed Page 75210consent decrees but that the flare test data do not support more stringent operating limits for the proposed hydrogen-olefins criteria by the EPA. Additionally, one commenter suggested that if the EPA decides to proceed with the more restrictive combustion zone limits for the hydrogen-olefins interaction cases then the final rule should not expand beyond an interaction between hydrogen and propylene.

Several commenters suggested that the proposed 15-minute feed forward averaging time for flares (e.g., combustion zone parameters, air-assist dilution parameters and associated flow rates) is arbitrary, unrealistic and unworkable and that the feed forward compliance determination should not be finalized and, if it is finalized, the averaging time should be extended to 1-hour, 3-hour, or 24-hour. To support these suggested averaging periods, commenters claimed that typical standards for combustion devices are averaged over these suggested timeframes, noting as an example, recent refinery flare consent decrees that contain a 3-hour average. The commenters also asserted that both a GC and calorimeter will be needed to obtain data rapidly enough to try and maintain a 15-minute average; that the feed forward approach requires calculation artifices to attempt to correct for the fact that compliance cannot be determined until the averaging period is over; and that a longer averaging time is needed for instrument and control response time.

Response: In addressing these comments, we further analyzed the flare emissions test data. First, to address concerns that the minute-by-minute analysis produced flawed results, we re-compiled the data into approximate “15-minute averages” to the extent practical based on the duration of a given test run (e.g., a 10-minute run was used as 1 run and a 32-minute run was divided into 2 runs of 16 minutes each). We do not find significant differences in the data or that different conclusions would be drawn from the data based on this approach as compared with the minute-by-minute analysis used for the proposed rule.

Next, we evaluated the 15-minute run data using the normal net heating value for hydrogen of 274 Btu/scf, which is the value we used in the analysis for the proposed rule and also evaluated the data using the 1,212 Btu/scf, the value recommended by some commenters. The 1,212 Btu/scf value is based on a comparison between the lower flammability limit and net heating value of hydrogen compared to light organic compounds and has been used in several consent decrees to which the EPA is a party. Based on our analysis, we determined that using a 1,212 Btu/scf value for hydrogen greatly improves the correlation between combustion efficiency and the combustion zone net heating value over the entire array of data. Using the net heating value of 1,212 Btu/scf for hydrogen also greatly reduced the number of “type 2 failures” (instances when the combustion efficiency is high, but the gas does not meet the NHVcz limit). One of the primary motivations for the proposed approach to provide alternative limits based on lower flammability limits and combustibles concentrations was to reduce these type 2 failures. Therefore, we proposed all three of these parameters (i.e., NHVcz, LFL and total combustibles) and allowed flare owners or operators to comply with any of the parameter limits at any time. When using the net heating value of 1,212 Btu/scf for hydrogen, the other two alternatives no longer provide any improvement in the ability to predict good flare performance. Consequently, we are simplifying the operating limits to use only NHVcz.

Next, we re-evaluated whether to finalize the proposed dual combustion zone operating limits for refinery flares that met certain hydrogen-olefins interactions or to finalize a single combustion zone net heating value limit. The newly re-compiled PFTIR run average flare dataset suggests that higher operating limits may be appropriate for some olefin-hydrogen mixtures. However, the dataset using 15-minute test average runs is much smaller than the set using 1-minute runs and thus creates a greater level of uncertainty. In addition, we cannot definitively conclude that a dual combustion zone limit for refinery flares meeting certain hydrogen-olefins interactions is appropriate given these uncertainties. Thus, in order to minimize these uncertainties and streamline the compliance requirements, we used all of the 15-minute test run average data together as a single dataset in an effort to determine an appropriate, singular combustion zone net heating value operational limit.

Finally, we conducted a Monte Carlo analysis to help assess the impacts of extending the averaging time on the test average flare dataset of 15-minute runs to 1-hour or 3-hour averaging time alternatives. While we consider it reasonable to provide a longer averaging time for logistical reasons, the Monte Carlo analysis demonstrated, consistent with concerns described in our proposal, that short periods of poor performance can dramatically limit the ability of a flare to achieve the desired control efficiency. Consequently, we find it necessary to finalize the proposed 15-minute averaging period to ensure that the 98-percent control efficiency for flares is achieved at all times. However, we understand that flare vent gas flow and composition are variable. While a short averaging time is needed to ensure adequate control given this variability, we also understand the complications that this variability places on flare process control in efforts to meet the NHVcz limit. Therefore, we are clarifying that the 270 Btu/scf NHVcz value is an operational limit that must be calculated according to the requirements in this rule. We also clarify that compliance with this operational limit must be evaluated using the equations and calculation methods provided in the rule. We proposed a feed forward calculation method to allow refinery owners or operators a means by which to adjust steam (or air) and, if necessary, supplemental natural gas flow, in order to meet the limit. In other words, “feed forward” refers to the fact that the rule requires the refinery owners or operators to use the net heating value of the vent gas (NHVvg) going into the flare in one 15-minute period to adjust the assist media (i.e., steam or air) and/or the supplemental gas in the next 15-minute period, as necessary for the equation in the rule to calculate an NHVcz limit of 270 BTU/scf or greater. We recognize that when a subsequent measurement value is determined, the instantaneous NHVcz based on that compositional analysis and the flow rates that exist at the time may not be above 270 Btu/scf. We clarify that this is not a deviation of the operating limit. Rather, the owner or operator is only required to make operational adjustments based on that information to achieve, at a minimum, the net heating value limit for the subsequent 15-minute block average. Failure to make adjustments to assist media or supplemental natural gas using the equation provided for calculating an NHVcz limit of 270 BTU/scf, using the NHVvg from the previous period, would be a deviation of the operating limit.

Alternatively, if the owner or operator is able to directly measure the NHVvg on a more frequent basis, such as with a calorimeter (and optional hydrogen analyzer), the process control system is able to adjust more quickly, and the owner or operator can make adjustments to assist media or supplemental natural gas more quickly. In this manner, the owner or operator is not limited by Start Printed Page 75211relying on NHVvg data that may not represent the current conditions. Therefore, the owner or operator may opt to use the NHVvg from the same period to comply with the operating limit.

Based on the results of all of our analyses, the EPA is finalizing a single minimum NHVcz operating limit for flares subject to the Petroleum Refinery MACT standards of 270 BTU/scf during any 15-minute period. The agency believes, given the results from the various data analyses conducted, that this operating limit is appropriate, reasonable and will ensure that refinery flares meet 98-percent destruction efficiency at all times when operated in concert with the other suite of requirements refinery flares need to achieve (e.g., flare tip velocity requirements, visible emissions requirements, and continuously lit pilot flame requirements). For more detail regarding our data re-analysis, please see the memorandum titled “Flare Control Option Impacts for Final Refinery Sector Rule” in Docket ID No. EPA-HQ-OAR-2010-0682.

Comment: Numerous commenters objected to the proposed requirements to have the velocity and visible emissions limits apply at all times for flares. Commenters suggested that flares are not designed to meet the visible emissions and flare tip velocity requirements when being operated beyond their smokeless capacity and suggested several alternative approaches: remove the visible emissions and flare tip velocity requirements from the rule altogether; exempt flares from these requirements during emergencies; or add a requirement to maintain a visible flame present at all times or include a work practice standard in the rule when flares are operated beyond their smokeless capacity at full hydraulic load. The commenters identified full hydraulic load as the maximum flow the flare can receive based on the piping diameter of the flare header and operating pressure of processes connected to the flare header system. They also specified that full hydraulic load would only occur if all sources connected to the flare header vented at the same time, which might result from an emergency shutdown due to a plant-wide power failure. According to commenters, flares are typically designed to operate in a smokeless manner at 20 to 30-percent of full hydraulic load. Thus, they claimed, flares have two different design capacities: A “smokeless capacity” to handle normal operations and typical process variations and a “hydraulic load capacity” to handle very large volumes of gases discharged to the flare as a result of an emergency shutdown. According to commenters, this is inherent in all flare designs and it has not previously been an issue because the flare operating limits did not apply during malfunction events. However, if flares are required to operate in a smokeless capacity during emergency releases, the commenters claimed that refineries would have to quadruple the number of flares at each refinery to control an event that may occur once every 2 to 5 years.

To support their suggestions, commenters pointed out that flaring during emergencies is the optimum way of handling very large releases and that the flare test data clearly demonstrate that visible emissions and/or high flare tip velocity do not suggest poor destruction efficiency during such events. The commenters also argued that operators should not have conflicting safety and environmental considerations to deal with during these times. The commenters stated that refiners are still subject to a civil suit even if the EPA uses its enforcement discretion where such a release would violate the limit and in order to avoid such liability, many new flares would have to be built. Commenters estimated that 500 new large flare systems at a capital cost in excess of $10-20 billion would need to be built because of the amount of smokeless design capacity that would be needed and that this significant investment would take the industry at least a decade to install.

Response: At the time of the proposed rule, we did not have any information indicating that flares were commonly operated during emergency releases at exit velocities greater than 400 ft/sec (which is 270 miles per hour (mph)). Similarly, we did not have information to indicate that flares were commonly designed to have a smokeless capacity that is only 20 to 30-percent of their “hydraulic load capacity.” While we are uncertain that refineries actually would install additional flares to the degree the commenters claim, based on the possibility that there may be an event every 2 to 5 years that would result in a deviation of the smokeless limit, we also recognize that it would be environmentally detrimental to operate hundreds of flares on hot standby in an effort to never have any releases to a flare that exceed the smokeless capacity of that flare. This is because operating hundreds of new flares to prevent smoking during these rare events will generate more ongoing emissions from idling flares than the no visible emissions limit might prevent during one of these events. Therefore, we considered alternative operating limits or alternative standards that could apply during these emergency release events.

As an alternative to the proposed requirement that flares meet the visible emissions and velocity limits at all times, we considered a work practice standard for the limited times when the flow to the flare exceeds the smokeless capacity of the flare. Owners or operators of flares would establish the smokeless capacity of the flare based on design specification of the flare. Below this smokeless capacity, the velocity and visible emissions standards would apply as proposed. Above the smokeless capacity, flares would be required to perform root cause analysis and take corrective action to prevent the recurrence of a similarly caused event. Multiple events from the same flare in a given time period would be a deviation of the work practice standard. Force majeure events would not be included in the event count for this requirement.

Based on industry claims that there is a hydraulic load flaring event, on average, every 4.4 years, we assumed the best performers would have no more than one event every 6 years, or a probability of 16.7-percent of having an event in any given year. We found that, over a long period of time such as 20 years, half of these best performers would have 2 events in a 3 year period, which would still result in over half the “best performing” flares having a deviation of the work practice standard if it was limited to 2 events in 3 years. Conversely, only 6 percent would have 3 events in 3 years over this same time horizon. Based on this analysis, 3 events in 3 years would appear to be “achievable” for the average of the best performing flares.

Pursuant to CAA section 112(d)(2) and (3), we are finalizing a work practice standard for flares that is based on the best practices of the industry, and considers the rare hydraulic load events that inevitably occur at even the best performing facilities.

The best performing facilities have flare management plans that include measures to minimize flaring during events that may cause a significant release of material to a flare. Therefore, we are requiring owners or operators of affected flares to develop a flare management plan specifically to identify procedures that will be followed to limit discharges to the flare as a result of process upsets or malfunctions that cause the flare to exceed its smokeless capacity. We are specifically requiring refinery owners or operators to implement appropriate prevention measures applicable to these Start Printed Page 75212emergency flaring events (similar to the prevention measures we are requiring in this final rule to minimize the likelihood of a PRD release). Refiners will be required to develop a flare minimization plan that describes these proactive measures and reports smokeless capacity. Refiners will need to conduct a specific root cause analysis and take corrective action for any flare event above smokeless design capacity that also exceeds the velocity and/or visible emissions limit. If the root cause analysis indicates that the exceedance is caused by operator error or poor maintenance, the exceedance is a deviation from the work practice standard. A second event within a rolling 3-year period from the same root cause on the same equipment is a deviation from the standard. Events caused by force majeure, which is defined in this subpart, would be excluded from a determination of whether there has been a second event. Finally, and again excluding force majeure events, a third opacity or velocity limit exceedance occurring from the same flare in a rolling 3-year period is a deviation of the work practice standard, regardless of the cause.

Comment: Several commenters suggested that the EPA should revise the combustion efficiency requirements to apply only to steam-assisted flares used as Refinery MACT control devices during periods of time that the flares are controlling Refinery MACT regulated streams. One commenter suggested that the EPA misused the TCEQ data in proposing the NHVcz metric and that the proposed limits are overly conservative. The commenter requested that the EPA work with stakeholders to conduct additional testing to determine what, if any, operating parameters are appropriate and necessary to achieve an adequate destruction efficiency for non-steam-assisted flares.

Response: We disagree with the commenters that the combustion efficiency requirements should apply only to steam-assisted flares. The available data (for runs where steam assist is turned off) as well as the available combustion theories suggest that the combustion zone net heating value minimum limit, which is the vent gas net heating value for unassisted or perimeter air-assisted flares, is necessary to ensure proper flare performance. While we agree that additional data on air-assisted flares would allow for a more robust analysis, the data we do have strongly indicate that air-assisted flares can be over-assisted and that the combustion efficiency of air-assisted flares that are over-assisted is below 98-percent control efficiency.

Comment: A few commenters suggested that the proposed flare regulations should not apply to part 63, subpart R (gasoline loading) and subpart Y (marine vessel loading) facilities, and to part 61, subpart FF (benzene waste) facilities. The commenters recommended that flares associated with gasoline loading, marine vessel loading and wastewater treatment emissions need to comply only with the General Provisions for flares. Some of these commenters argued that these sources are more consistent in flow and composition than other refinery sources, so the new requirements are not necessary to ensure good combustion for these “dedicated” flares. Some commenters suggested that operators of flares with consistent flow and composition be allowed to use process knowledge or engineering judgment rather than be required to install continuous monitors or be subject to ongoing grab sampling requirements.

Some commenters noted that the required control efficiency for some refinery emissions sources subject to subpart CC sources is 95-percent. One commenter also requested that the EPA provide overlap provisions so flares used to control sources from different MACT sources would not have duplicative requirements.

Response: The regulatory revisions that we are finalizing apply to petroleum refinery sources subject to part 63, subparts CC and UUU. Gasoline loading, marine vessel loading and wastewater treatment operations that are part of the refinery affected source as defined at 40 CFR 63.640 are subject to subpart CC. Gasoline loading, marine vessel loading and wastewater treatment operations located at non-refinery source categories are not subject to part 63, subpart CC and, thus, would not be subject to the revisions to subpart CC being finalized in this action. To the extent that the commenters are requesting that the EPA establish flare requirements that would apply to flares that are not part of the refinery affected source, that request is beyond the scope of this rulemaking, which only addresses revisions to Refinery MACT 1 and 2. When we issue rules addressing requirements for other sources with flares, we will consider issues similar to those we considered in this action and determine at that time whether revisions to those other flare requirements are necessary.

The commenters note that some subpart CC emissions sources have only a control efficiency requirement of 95-percent. While this may be true, where the owner or operator chooses to control these sources through the use of a flare, operation of that flare was subject to operational requirements in the General Provisions at 40 CFR 63.11 and the best performing flares were achieving 98-percent control at the time the General Provisions were promulgated. At the time the General Provisions were promulgated, we received no comments that the EPA should set different operational limits for flares that are controlling emissions from sources where the standard may vary by level of control efficiency and we see no basis to do so now. The purpose of the revisions to the flare operating requirements is to ensure that flares are operating consistent with the MACT floor requirements for any and all sources that may use flares as a control device (79 FR 36905, June 30, 2014). As the MACT floor control requirements of certain refinery sources that allow the use of a flare as a control device is 98-percent, we established operational limits to ensure flares used as control devices meet this MACT requirement.

To the extent that the commenters are requesting that the EPA establish an alternative monitoring approach for flares in dedicated service that have consistent composition and flow, we agree that these types of flares, which have limited flare vent gas streams, do not need to have the same type of on-going monitoring requirements as those with more variable waste streams. Thus, we are establishing an option that refinery owners or operators can use to demonstrate compliance with the operating requirements for flares that are in dedicated service to a specific emission source, such as a wastewater treatment operation. Refinery owners or operators will need to submit an application for the use of this alternative. The application must include a description of the system, characterization of the vent gases that could be routed to the flare based on a minimum of 7 grab samples (14 daily grab samples for continuously operated flares) and specification of the net heating value that will be used for all flaring events (based on the minimum net heating value of the grab samples). We are also allowing engineering estimates to characterize the amount of gas flared and the amount of assist gas introduced into the system. For example, the use of fan curves to estimate air assist rates is acceptable. Flare owners or operators would use the net heating value determined from the initial sampling phase and measured or estimated flare vent gas and assist gas Start Printed Page 75213flow rates, if applicable, to demonstrate compliance with the standards.

Comment: A few commenters suggested that the EPA's proposed work practice and monitoring standards for flares are CAA section 112(d) “developments” required by law and supported by the evidence, and reflect best practices at many refineries today. One commenter suggested that the EPA must allow companies with consent decrees to meet their consent decree requirements as an alternative compliance approach and in lieu of the proposed requirements.

Response: We proposed the enhanced monitoring requirements and operating limits under authority of CAA sections 112(d)(2) and (d)(3) to ensure that flares used to control regulated Refinery MACT 1 or 2 gas streams are meeting the prescribed control efficiencies established at the time the MACT standard was promulgated. And, we continue to believe that these revisions are appropriate under CAA sections 112(d)(2) and (d)(3). The commenter has not suggested, and we do not believe, that the revisions promulgated would differ in substance if they were instead promulgated under CAA section 112(d)(6).

In general, we expect that the NHVcz monitoring requirements that we are finalizing for flares will be consistent with the requirements in various consent decrees. However, we have not conducted a rigorous evaluation of equivalency between various requirements and therefore we are not at this time providing an allowance for flare owners or operators to comply with the NHVcz operating limits and any provisions for necessary monitoring needed in the consent decree in lieu of the NHVcz limits and monitoring requirements established in this rule. In the event that an owner or operator wishes to continue complying only with the requirements of a consent decree, the rule contains provisions by which owner or operator can seek approval for alternative limits that are at least equivalent to the performance achieved from complying with the operating limits included in the final rule.

iii. Pressure Relief Devices

Comment: Several commenters suggested that the EPA develop a work practice approach for atmospheric PRD rather than a prohibition on releases. One commenter recommended that the EPA establish a work practice standard for atmospheric PRDs that requires refiners to implement a base level of preventative measures including: Basic process controls, instrumented alarms, documented and verified routine inspection and maintenance programs, safety-instrumented systems, disposal systems, provide redundant equipment, increase vessel design pressure and systems that reduce fire exposure on equipment. Additionally, the commenter recommended that the EPA require refiners to perform root cause analysis and implement corrective action in the event of a release. The commenter stated these requirements would be similar to the root cause analysis/corrective action requirements recently promulgated for flares under NSPS subpart Ja and provided specific regulatory language for a proposed work practice approach. (See section 2.4.1.8 in Docket item EPA-HQ-OAR-2010-0682-0583.) One commenter requested that the EPA allow a process for companies to submit an application for case-by-case limits to be approved by the agency, either the EPA or a delegated state similar to the alternate NOX limits for process heaters provided in NSPS subpart Ja. This commenter recommended that the EPA establish reasonable work practice standards, specifically suggesting that the EPA develop work practice standards consistent with API 521. The commenter stated that the EPA should provide an implementation period for compliance that goes beyond the timeframe provided under CAA section 112(d). The commenter added that the EPA should adopt specified changes to the definition of an atmospheric pressure relief safety valve and provided suggested regulatory language for a proposed work practice standard for PRDs in EPA-HQ-OAR-2010-0682-0549.

Another commenter stated that the EPA should require, as the Bay Area Air Quality Management District (BAAQMD) does, that any refinery that has a reportable PRD event must take certain steps to prevent such releases in the future (BAAQMD Rule 8-28-304). In particular, such a refinery must create a Process Hazard Analysis, meet the Prevention Measures Procedures specified in section 8-28-405, and conduct a failure analysis of the incident, to prevent recurrence of similar incidents (Id. Reg. section 8-28-304.1). If a second release occurs, then, within one year, the facility must vent its PRDs to a vapor recovery or disposal system that meets certain requirements (Id. Reg. section 8-28-304.2). The commenter asserted that the EPA's prohibition on releases to the atmosphere from PRD will ensure that refineries take the necessary steps to prevent such releases, or install control devices so that any releases from PRDs that must occur are vented through a control device to reduce the amount of toxic air pollution they emit. At a minimum, the commenter stated, the EPA must prohibit these uncontrolled emissions and require monitoring and reporting to assure compliance and ensure that the emission standards apply at all times, as required by the Act. The commenter argued that the EPA must also, however, consider requiring the additional developments that have been put into place in the BAAQMD and also require control devices to be used for all PRD, as some local air districts require. In addition, the commenter supported the EPA's monitoring and reporting requirements for PRD releases and the proposed electronic reporting requirements, which the EPA recognized are needed to assure compliance and assist with future rulemakings and as that provision requires, the EPA also must make all information reported publicly available online promptly and in an accessible and understandable format.

Response: We agree that, under the proposal, refineries would consider installing add-on controls to comply with the prohibition on atmospheric releases from PRDs. In addition, they would consider venting these control devices to existing control devices, including flares. However, it may not be feasible to vent some or all of the PRDs to existing flares if the flares are near their hydraulic load capacity based on the processes already connected to the flares. Flares have negative secondary impacts when operated at idle conditions for the vast majority of time, which could be the case if they were installed solely to address PRD releases. These secondary impacts result from GHG, CO and NOX emissions. Some PRDs may vent materials that are not compatible with flare control and would need to be vented to other controls.

To estimate the impact of the proposed prohibition on venting PRDs to the atmosphere, we estimated that at least one new flare per facility would be required to handle releases from PRDs, based on the number of atmospheric PRDs reported at refineries; that 60-percent of the PRDs could be piped to existing controls at minimal costs and the other 40-percent would have to be piped to new flares; and that, on average, each new flare would service 40 PRDs. Based on these assumptions, 151 new flares would be needed or approximately one new flare per refinery. At a capital cost of $2 million for each new flare, which would not include long pipe runs, if needed, to PRD that are dispersed across the plant, we estimate that the capital cost of the Start Printed Page 75214prohibition on venting to the atmosphere would exceed $300 million. Considering the fuel needed (approximately 50,000 scf/day per flare) and a natural gas price of $4.50 per 1,000 scf, we estimate the annual operating cost for these new flares to be $12 million.

PRDs are unique in that they are designed for the purpose of releasing or “popping” as a safety measure to address pressure build-up in various systems—pipes, tanks, reactors—at a facility. These pressure build-ups are typically a sign of a malfunction of the underlying equipment. While it would be difficult to regulate most malfunction events because they are unpredictable and can vary widely, in the case of PRDs, they are equipment installed specifically to release during malfunctions and as such, we have information on PRDs in our 2011 Refinery ICR and through the SCAAMD and BAAQ rules to establish standards for them. After reviewing these comments, we thus examined whether it would be feasible to regulate these devices under CAA section 112(d)(2) and (3).

After reviewing the comments, we agree with the commenters who suggest that the BAAQMD rule, as well as a similar South Coast Air Quality Management District (SCAQMD) rule that address PRD releases (SCAQMD Rule 1173), provide work practice standards that reflect the level of control that applies to the best performers. Consequently, we developed a work practice standard for PRD based on a detailed MACT analysis considering the requirements in these rules. Our rationale for the selected MACT requirements is provided in section IV.C.4 of this preamble. The work practice standards that we are finalizing for PRDs require refiners to establish proactive measures for each affected PRD to prevent direct release of HAP to the atmosphere as a result of pressure release events. In the event of an atmospheric release, we are requiring refinery owners or operators to conduct root cause analysis to determine the cause of a PRD release event. If the root cause was due to operator error or negligence, then the release would be a deviation of the standard. For any other release (not including those caused by force majeure events), the owner or operator would have to implement corrective action. A second release due to the same root cause for the same equipment in a 3-year period would be a deviation of the work practice standard. Finally, a third release in a 3-year period would be a deviation of the work practice standard, regardless of the root cause. Force majeure events would not count in determining whether there has been a second or third event.

With respect to defining “atmospheric pressure relief safety valve” as suggested by the commenter, we note that the June 30, 2014, proposed amendments in 40 CFR 63.648(j) used the term “relief valve” because this was a defined term in Refinery MACT 1. However, the proposed amendments included clauses such as “if the relief valve does not consist of or include a rupture disk.” Thus, we specifically intended to apply the pressure relief management requirements broadly to “pressure relief devices” and not just “valves.” To clarify this, we have revised the regulatory language to use the term “pressure relief device” rather than “relief valve” to clearly include rupture disks or similar types of equipment that may be used for pressure relief.

4. What is the rationale for our final approach and final decisions for the revisions pursuant to CAA section 112(d)(2) and (3)?

We revised the MACT floor determination for DCU sources. CAA section 112(d)(3)(A) requires the MACT floor for existing sources to exclude “. . . those sources that have, within 18 months before the emission standard is proposed or within 30 months before such standard is promulgated, whichever is later, first achieved a level of emission rate or emission reduction which complies, or would comply if the source is not subject to such standard, with the lowest achievable emission rate (as defined by section 171) applicable to the source category and prevailing at the time, in the category or subcategory for categories and subcategories with 30 or more sources.” Because we have determined that a 2 psig emissions limitation is equivalent with a LAER emission limit for DCU, we revised the MACT floor analysis in order to exclude sources that first met the 2 psig limit on or after December 30, 2012. For existing sources, based on the revised MACT analysis, we concluded that the MACT floor is still 2 psig. However, because the information on which we relied was submitted in response to the 2011 Petroleum Refinery ICR which requested “typical” venting pressures and because providing an allowance to average across venting periods does not reduce the emissions reductions achieved, we are providing a 60-event averaging period for existing sources in response to public comments received.

For new DCU sources, our revised analysis identified one DCU subject to permit emission limitations of 2.0 psig pressure limit prior to venting on a per event basis. Under CAA section 112(d)(3), the MACT standard for new sources cannot be less stringent than the emission control achieved in practice by the best-controlled similar source. Thus, we are finalizing a limit of 2.0 for new DCU sources. We note that as 2.0 psig limit is more stringent than a 2 psig limit because of the rounding convention of rounding to the number of significant digits for which the standard is expressed. For example, a 2.4 psig venting pressure is compliant with a 2 psig limit, while it is not compliant with a 2.0 psig limit.

We evaluated the costs of requiring existing sources to meet a 2.0 psig limit as a beyond-the-MACT-floor option. We determined the incremental cost of going from a 2 psig limit with an allowance to average over 60 events to a 2.0 psig limit on a per event basis was approximately $70,000 per ton of HAP reduced considering VOC credits. Based on this high incremental cost-effectiveness, we concluded that the MACT floor requirement for existing DCU sources was MACT. As discussed in detail in the proposal, we do not consider it technically feasible to meet a 1 psig pressure limit (effectively a 1.4 psig limit) on a not-to-be-exceeded basis. Thus, we rejected this beyond the floor control option for both existing and new DCU sources. See the memorandum titled “Reanalysis of MACT for Delayed Coking Unit Decoking Operations” in Docket ID No. EPA-HQ-OAR-2010-0682 for additional details regarding our re-analysis of MACT for DCU decoking operations.

In response to comments received on the prohibition of draining prior to achieving the proposed pressure limit (see Section 7.2.1 in the “National Emission Standards for Hazardous Air Pollutants from Petroleum Refineries—Background Information for Final Amendments: Summary of Public Comments and Responses” in Docket ID No. EPA-HQ-OAR-2010-0682), we are providing specific provisions to allow for draining under special conditions. The specific provision and our rationale for providing them are provided below.

First, we learned that certain DCU are designed to completely fill the drum with water and allow the water to overflow in the overhead line and drain to a receiving tank in order to more effectively cool the coke bed. Owners or operators of this DCU design were concerned that the water overflow may be considered a drain and also stated that overhead temperature rather than pressure would be a better indicator of effective bed cooling. In reviewing this Start Printed Page 75215type of DCU design, we find that this design has some unique advantages to traditional DCU to effect better cooling of the coke drum, and therefore we do not want to preclude its use. Based on saturated steam properties, we determined that an overhead temperature of 220 °F would achieve equivalent or greater emissions reductions than a 2 psig pressure limitation and an overhead temperature of 218 °F would achieve equivalent or greater emissions reductions than a 2.0 psig pressure limitation. Therefore, we are including these temperature limits as alternatives to the 2 or 2.0 psig pressure limitations for existing and new DCU affected sources, respectively. With respect to the overflow “drain,” we remain concerned with emissions from draining superheated water. However, if submerged fill is used in the atmospheric tank receiving the overflow water, the superheated water will be cooled by the water within the tank and emissions that occur during the conventional draining of water (from the flashing of superheated water into steam) can be prevented. Therefore, we are allowing the use of water overflow provided the overflow “drain” water is hard-piped to the receiving tank via a submerged fill pipe (pipe below the existing liquid level) whenever the overflow water exceeds 220 °F.

Second, we received comments that, for conventional DCU (those not designed to allow water overflow), there is a limit to the maximum water level in the drum, which limits to some extent how much cooling water can be added to the coke drum. In rare cases, the coke drum does not cool sufficiently using the typical cooling steps. In this case, the common industry practice is to partially drain the coke drum and refill it with additional cooling water. This “double-quench” process is needed for safety reasons to sufficiently cool the coke drum contents prior to the decoking operations. Therefore, commenters requested provisions to allow double-quenching of the coke drum. We recognize the safety issues associated with coke blow-out during coke cutting if there is a portion of the coke bed that is not sufficiently cooled and we agree that double-quenching is an effective means to cool the coke drum in those rare instances that the typical cooling cycle does not sufficiently cool the coke drum contents, so we considered granting the commenters' request. As noted previously, the primary concern with early draining of the coke drum is the emissions that are expected to occur as a result of draining superheated water. We recognize, however, that the water temperature near the bottom of the coke drum is typically much lower than at the top of the coke drum. If the temperature of the water drained from the bottom of the coke drum remains below 210 °F, this would minimize steam flashing and associated HAP emissions since the water drained would not be superheated. We conclude that the use of double quenching is appropriate for cases when the coke drum is not sufficiently cooled using the normal cooling procedures provided the temperature of the water drained remains below 210 °F, and it is consistent with the practices of the best performing sources. Consequently, we are finalizing provisions to allow the use of double-quenching for DCU provided the temperature of the water drained remains below 210 °F.

For the CRU, we are finalizing the proposed revisions to require CRU that employ active purging to meet the MACT emissions limitations in Tables 15 and 16 in subpart UUU at all times regardless of vessel pressure. We received limited comments regarding our proposal; these comments generally concerned the costs associated with the proposed emissions limitations. As discussed in our proposal, and based on data submitted in response to the ICR, emissions using active purging are much higher than those not using active purging. In the original rule, we based the MACT floor on the best performing facilities that used sequential pressurizations and depressurizations rather than active purging. Thus, in the proposal, we concluded that allowing owners or operators to actively purge while at low pressures was inconsistent with the MACT floor emissions limitations achieved by the best performing 12-percent of sources when the MACT floor was originally established. As we are simply requiring these facilities to meet the same emission levels determined to be MACT, we do not consider costs of potential additional controls to be a viable rationale to allow these units to emit several times more HAP than the units upon which the MACT requirements were based and the emissions levels achieved in practice by the vast majority of other CRU sources.

For flares, we are finalizing proposed revisions to include detailed flare monitoring and operating requirements. We are including the flaring provisions for refineries in the Refinery MACT rules and removing the cross-references to the flaring requirements in the General Provisions. The final regulatory requirements differ from the proposed requirements in several respects. First, we are not finalizing the ban on halogenated vent streams because we did not include sufficient justification or include cost estimates for this proposed provision and we did not include any monitoring requirements to ensure compliance with this ban on halogenated vent streams.

We are finalizing the proposed no visible emissions limit and the flare tip velocity limit but they will apply only when the flare vent gas flow rate is below the smokeless capacity of the flare. We received a number of comments stating that the no visible emissions limit and the flare tip velocity limit cannot be met during large malfunctions and emergency shutdown events. In response to comments, we are finalizing work practice standards for emergency flaring events using the proposed no visible emission limit and flare tip velocity limit as thresholds in the final rule to trigger root cause analysis when the flare vent gas flow rate is above the smokeless capacity of the flare. The final work practice standard includes requirements to develop a flare management plan, to implement prevention measures, and to perform root cause analysis and implement corrective action following each flaring event that exceeds the smokeless capacity of the flare. There is also a limit on the number of these flaring events that a given flare may have in the 3-year period. We are establishing these provisions because we now recognize that flares have two different design capacities: A smokeless design capacity and a hydraulic load capacity. We determined that the proposed visible emissions limit and the flare tip velocity limit for very large flow events are not the MACT floor for such events. The final work practice standards for flaring events are based on the best performing facilities and will result in emission reductions in a technically feasible manner without any negative secondary impacts.

We consider it appropriate to establish a work practice standard for flares as provided in CAA section 112(h). While it is possible to monitor gaseous streams going into the flare (as we have required for the flare operating requirements) it is not possible to design and construct a conveyance to capture the emissions from a flare. While knowledge of the composition and flow of gases entering the flare provides a reasonable basis for establishing operating requirements for normal operations, we have no data on flare performance under conditions in the hydraulic load range. While smoke in the flare exhaust is an indication of incomplete combustion, it is uncertain Start Printed Page 75216how much deterioration of HAP destruction efficiency occurs during a smoking event. We also consider that the application of a measurement methodology for flare exhaust is not practicable due to technological and economic limitations. Passive FTIR has been used to determine combustion efficiency in flare exhaust, but these are essentially manual tests, and the measurement accuracy is dependent on how well the monitor is aligned with the flare exhaust plume. Changes in wind direction require manual movement of the monitoring system. It is also unclear if these systems can accurately measure combustion efficiency during high smoking events. These systems also require very specialized expertise, and we consider that it is both technologically and economically infeasible to measure flare exhaust emissions, particularly during high load events. Consequently, for emergency flare releases, we conclude that it is appropriate to establish a work practice standard as provided in CAA section 112(h).

We also received comments that the daily visible emissions observations were burdensome and unnecessary and some commenters suggested that facilities be allowed to use video surveillance cameras. We concluded that video surveillance cameras would be at least as effective as the proposed daily 5-minute visible emissions observations using Method 22. We are finalizing the proposed visible emissions monitoring requirements Method 22 and the alternative of using video surveillance cameras.

We are simplifying the combustion zone gas property operating limits in response to public comments received. Specifically, we are finalizing requirements that all flares meet a minimum operating limit of 270 BTU/scf NHVcz on a 15-minute average, and we are providing that refiners use a corrected heat content of 1,212 BTU/scf for hydrogen to demonstrate compliance with this operating limit. We determined that a corrected heat content of 1212 BTU/scf for hydrogen provided a better indication of flare performance than without the correction. We also determined that the other combustion zone parameters, which were primarily proposed to provide suitable methods for flares that had high hydrogen concentrations, were no longer necessary when a 1,212 Btu/scf net heating value is used for hydrogen. Therefore, we are not finalizing the alternative combustion zone operating limits based on lower flammability limit or combustibles concentration. We are also not finalizing separate combustion zone operating limits for gases meeting the proposed hydrogen-olefin interaction criteria. In our revised analysis of the data, we analyzed all of the data together and determined the 270 Btu/scf NHVcz operating limit provided in the final rule would adequately ensure that flares achieve the desired 98-percent control efficiency regardless of the composition of gas sent to the flare.

For air-assisted flares, we are finalizing the additional “dilution parameter” operating limit only for the net heating value dilution parameter, NHVdil. Similar to the requirements we are finalizing for the combustion zone parameters, we are finalizing requirements that flares meet a minimum operating limit of 22 BTU/ft2 NHVdil on a 15-minute average, and we are providing that refiners use a corrected heat content of 1,212 BTU/scf for hydrogen to demonstrate compliance with this operating limit. For the reasons explained above, we are not finalizing the proposed alternative dilution parameter operating limits based on lower flammability limit or combustibles concentration, and we are not finalizing separate dilution parameter operating limits for gases meeting the proposed hydrogen-olefin interaction criteria.

For flares in dedicated service, we are establishing an alternative to continuous or on-going grab sample requirements for determining waste gas net heating content to reduce the burden of sampling for flare waste gases that have consistent compositions. Flares in dedicated service can use initial sampling period and process knowledge to determine a fixed net heating value of the flare vent gas to be used in the calculations of NHVcz and, if applicable, NHVdil.

We are revising the definition of MPV to remove the exemption for in situ sampling systems for the reasons provided in the proposed rule.

We received comments recommending that a work practice standard be adopted for PRD rather than the proposed prohibition of atmospheric PRD releases. Commenters stated that the prohibition was infeasible due to the proposed immediate timing of the requirement and impractical due to cost considerations. After reviewing these comments as well as the BAAQMD rule (Regulation 8, Rule 8-28-304) and the SCAQMD rule (Rule 1173), we have determined that the work practice standards in these rules reflect the level of control that applies to the best performers. Therefore, we proceeded to evaluate appropriate MACT requirements based on the provisions in these rules.

The BAAQMD rule requires sources to implement a minimum of three prevention measures to limit the possibility of a release. The BAAQMD uses a “release event” threshold of 10 lbs/day of organic or inorganic pollutants; the SCAQMD rule effectively uses a release event threshold of 500 lbs VOC/day. When a release event occurs, both rules require that the refiner perform a root cause analysis and take corrective action (including additional prevention measures). In addition, both rules require piping the PRD to a flare if there are more than two release events (releases above a certain release size threshold) in a 5-year period. Both rules include a number of exemptions for certain types of PRD that are not expected to release significant amounts of pollutants to the air or that are not feasible to control because of pressure considerations. These include PRD associated with storage tanks, vacuum systems and equipment in heavy liquid service as well as liquid thermal relief valves that are vented to process drains.

There are five refineries subject to the BAAQMD rule and seven refineries subject to the SCAQMD rule, accounting for 8-percent of refineries nationwide and representing the industry's best performers. We consider the BAAQMD rule to be the more stringent of the two because this rule requires sources to implement a minimum of three prevention measures to limit the possibility of a release (the SCAQMD rule has no similar requirement) and uses a lower mass threshold for what is considered a “release event” (10 lbs/day of organic or inorganic pollutants versus the 500 lbs VOC release threshold in the SCAQMD rule). Therefore, the BAAQMD rule is considered to be the MACT floor requirement for PRDs associated with new affected sources and the SCAQMD rule is considered to be the MACT floor for PRDs associated with existing affected sources.

In general, an open PRD is essentially the same as a miscellaneous process vent that is vented directly to the atmosphere. Consistent with our treatment of miscellaneous process vents and consistent with the two California rules, we believe that it is appropriate to exclude certain types of PRD that have very low potential to emit based on their type of service, size and/or pressure. For example, PRD that have a potential to emit less than 72 pounds per day of VOC, considering the size of the valve opening, design release pressure, and equipment contents, would be considered in a similar manner as Group 2 miscellaneous Start Printed Page 75217process vents and would not require additional control. The two California rule requirements do not apply to PRD on storage tanks and vacuum systems. Most of these PRD have a design release pressure of 2.5 psig and thus have a very limited potential to emit. It is technically infeasible to pipe these sources to a flare (or other similar control system) because the back pressure in the flare header system generally exceeds 2.5 psig. We note that some storage tanks can operate at elevated pressure (i.e., pressure tanks). Therefore, rather than follow exactly the requirements in the California rules, we determined it more practical to exclude PRD with design release pressure of less than 2.5 psig.

Any release from a PRD in heavy liquid service would have a visual indication of a leak and any repairs to the valve would have to be further inspected and, if necessary, repaired under the existing equipment leak provisions. Therefore, consistent with the BAAQMD rule, we are exempting PRD in heavy liquid service from the work practice standards we are establishing in this final rule.

Both the BAAQMD and SCAQMD rules exempt thermal expansion valves that are “vented to process drains or back to the pipeline.” We are unclear what is meant by “vented to process drains”; however, if a liquid is released from a PRD via hard-piping to a drain system that meets the control requirements specified in Refinery MACT 1, we consider that these PRD are controlled and they would not be subject to the work practice standard established in this final rule. Similarly, all PRD in light liquid service that are hard-piped to a controlled drain system (or back to the process or pipeline) are otherwise subject to a MACT requirement and would not be subject to the work practice standard.

In considering thermal relief valves not vented to process drains or back to the pipeline, we expect that releases from these thermal relief valves will be small and generally under the release event thresholds specified in the California rules. Therefore, the work practice standards do not apply to PRD that are designed solely to release due to liquid thermal expansion.

The primary goal of the PRD work practice standard is to reduce the size and frequency of releases. The SCAQMD rule is targeted towards fairly large releases (compared to the direct PRD releases reported in response to the Refinery ICR), so it will reduce the frequency of large releases, but it does little to reduce the frequency of smaller releases. To more effectively reduce the size and frequency of all releases, we consider it important to require the implementation of prevention measures (as required in the BAAQMD rule) and require root cause analysis and corrective action for PRD releases from all PRD subject to the work practice standard. While we recognize that if a PRD opens for a short period of time, the release might be below the release thresholds in the SCAQMD rules, we believe the release may be indicative of an important issue or design flaw. Because the potential for large emissions exist from the PRD subject to the work practice standard, we think it is reasonable to require a root cause analysis be conducted and appropriate corrective action implemented to potentially identify this issue and prevent a second release which, if the issue remains uncorrected, could be significant.

Requiring that prevention measures be implemented on all PRD subject to the work practice standard and not establishing a release threshold for release events is a variation from the SCAQMD rule. However, we also considered the allowable release frequency. We believe that our adoption of this approach is balanced by our not adopting the SCAQMD provisions requiring that PRD be vented to a flare or other control system or that refiners pay a fee if there are multiple releases of a certain size within a specified timeframe.[12] In place of this system, we are limiting the number of events from each PRD that can occur in a 3 year time period (2, if root causes are different), and in place of a fine, or routing to control, stating that the 3rd release in 3 years for any root cause is a deviation of the standard.

Because we are not including a size threshold for release events as in the SCAQMD rule, it is natural to assume release events would occur more frequently than release events subject to the SCAQMD rules. Also, based on our Monte Carlo analysis of random rare events, we note that it is quite likely to have two or three events in a 5-year period when a long time horizon (e.g., 20 years) is considered. Therefore, considering our analysis of emergency flaring events and the lack of a 500 lb/day release threshold, we considered it reasonable to use a 3-year period rather than a 5-year period as the basis of a deviation of the work practice standard.

The SCAQMD work practice standards do not apply to releases that are demonstrated to “result from natural disasters, acts of war or terrorism, or external power curtailment beyond the refinery's control, excluding power curtailment due to an interruptible service agreement.” These types of events, which we are referring to as “force majeure” events, are beyond the control of the refinery owner or operator. We are providing that these events should not be included in the event count, but that they would be subject to the root cause analysis in order to confirm whether the release was caused by a force majeure event.

Consistent with the requirements in the SCAQMD rule, we are requiring refinery owners or operators to conduct a root cause analysis for a PRD release event. If the root cause was due to operator error or negligence, then the release would be a deviation of the standard. For any other release (not including those caused by force majeure events), the owner or operator would have to implement corrective action. We consider that a second release due to the same root cause for the same equipment in a 3-year period would be a deviation of the work practice standard. This provision will help ensure that root cause/corrective action are conducted effectively. Finally, a third release in a 3-year period (not including those caused by force majeure events) would be a deviation of the work practice standard, regardless of the root cause. While we are using a 3-year interval rather than the 5-year interval provided in the SCAQMD, we consider that the requirements as included in this final rule (i.e., the inclusion of prevention measure requirements and no thresholds for release events) will achieve equivalent if not greater emissions reductions than the SCAQMD rule. We also consider that, given the prevention measure requirements and a 3-year period, there is less likelihood of unusual random events that happen over a short period of time that may cause refinery owners or operators to feel compelled to vent the PRD to a flare to eliminate concerns regarding potential non-compliance. Thus, we project that the requirements that we have included in the final rule will achieve emissions reductions commensurate to or exceeding the requirements in the SCAQMD rule (that serves as the MACT floor for existing sources) but will achieve those Start Printed Page 75218reductions in a more cost-effective manner.

We also considered requiring all PRD to be vented through a closed vent system to a control device as an alternative beyond-the-MACT floor requirement. While this requirement would provide additional emission reductions beyond those we are establishing as the MACT floor, these reduction come at significant costs. Capital costs for requiring control of all atmospheric PRD is estimated to be approximately $300 million compared to $11 million for the requirements described above. The total annualized cost for requiring control of all atmospheric PRD is estimated to be approximately $41 million/year compared to $3.3 million/year for the requirements described above. We estimate that the incremental cost-effectiveness of requiring control of all atmospheric PRD compared to the requirements described above exceeds $1 million per ton of HAP reduced. Consequently, we conclude that this is not a cost-effective option for existing sources.

The final requirements that we have developed for PRD achieve equal or greater emission reductions than those achieved by the SCAQMD rule (MACT floor). To the extent those requirements are more stringent that the SCAQMD, they are cost-effective. We could not identify an alternative requirement that provided further emission reductions in a cost-effective manner. Thus, we conclude that the work practice standards described above represent MACT for existing sources.

The BAAQMD rule, which represents the requirements applicable to the best performing sources, is the basis for new source MACT for PRD. Based on the specific provisions for PRD in the BAAQMD rule, we conclude that the MACT floor requirement is to have all PRD in HAP service associated with a new affected source vented through a closed vent system to a control device. As with existing sources, the PRD WPS would also contain the same exclusions (e.g., heavy liquid service PRDs, thermal expansion valves, liquid PRDs that are hard-piped to controlled drains, PRD with release pressures of less than 2.5 psig, PRD with emission potential of less than 72 lbs/day, and PRD on mobile equipment). These provisions are similar to the applicability provisions of the BAAQMD rule. Thus, we retain the same applicability of the work practice standard for PRDs on new or existing equipment, but all affected PRD on a new source would be required to be controlled. This is essentially equivalent to the proposed requirement of no atmospheric releases. We could not identify a control option more stringent than the BAAQMD rule as applied to new sources. Therefore, we conclude that venting all PRD in HAP service through a closed vent system to a flare or similar control system is MACT for PRD associated with new affected sources.

We consider it appropriate to establish a work practice standard for PRD as provided in CAA section 112(h). While it may be possible to design and construct a conveyance for PRD releases, we consider that the application of a measurement methodology for PRDs is not practicable due to technological and economic limitations. First, it is not practicable to use a measurement methodology for PRD releases. The venting time can be very short and may vary widely in composition and flow rate. The often-short duration of an event makes it infeasible to collect a grab sample of the gases when a release occurs, and a single grab sample would not account for potential variation in vent gas composition. It would be economically prohibitive to construct an appropriate conveyance and install and operate continuous monitoring systems for each individual PRD in order to attempt to quantitatively measure a release event that may occur only a few times in a 3-year period. Additionally, we have not identified an available, technically feasible continuous emission monitoring systems that can determine a mass VOC or HAP release quantity accurately given the flow, composition and composition variability of potential PRD releases from refineries. Consequently, we conclude that it is appropriate to establish a work practice standard for PRD releases as provided in CAA section 112(h).

D. NESHAP Amendments Addressing Emissions During Periods of SSM

1. What amendments did we propose to address emissions during periods of SSM?

We proposed to eliminate the SSM exemption in 40 CFR part 63, subparts CC and UUU. Consistent with Sierra Club v. EPA, we proposed standards in these rules that apply at all times. We also proposed several revisions to Table 6 of subpart CC of 40 CFR part 63 and to Table 44 to subpart UUU of 40 CFR part 63 (the General Provisions Applicability tables for each subpart), including eliminating the incorporation of the General Provisions' requirement that the source develop an SSM plan, and eliminating and revising certain recordkeeping and reporting requirements related to the SSM exemption.

For Refinery MACT 1, we proposed that the use of a bypass at any time to divert a Group 1 miscellaneous process vent to the atmosphere is a deviation of the emission standard, and specified that refiners install, maintain and operate a continuous parameter monitoring system (CPMS) for flow that is capable of recording the volume of gas that bypasses the APCD.

We also proposed to revise the definition of MPV to remove the exclusion for “Episodic or non-routine releases such as those associated with startup, shutdown, malfunction, maintenance, depressuring and catalyst transfer operations.” We also proposed that the control requirements for Group 1 MPV apply at all times, including startup and shutdowns.

For Refinery MACT 2, we proposed alternate standards for three emission sources for periods of startup or shutdown. We proposed PM standards for startup of FCCU controlled with an ESP under Refinery MACT 2 because of safety concerns associated with operating an ESP during an FCCU startup. For FCCU controlled by an ESP, we proposed a 30-percent opacity limit (on a 6-minute rolling average basis) during the period that torch oil is used during FCCU startup. For startup of FCCU without a post-combustion device under Refinery MACT 2, we proposed a CO standard based on an excess oxygen concentration of 1 volume percent (dry basis) based on a 1-hour average. For periods of SRU shutdown, we proposed to allow diverting the SRU purge gases to a flare meeting the design and operating requirements in 40 CFR 63.670 (or, for a limited transitional time period, 40 CFR 63.11) or to a thermal oxidizer operated at a minimum temperature of 1,200 °F and a minimum outlet oxygen concentration of 2 volume percent (dry basis). For other emission sources in Refinery MACT 2, we proposed that the requirements that apply during normal operations should apply during startup and shutdown.

2. How did the SSM provisions change since proposal?

a. Refinery MACT 1

We proposed that when process equipment is opened to the atmosphere (e.g., for maintenance), the existing MPV emissions limits apply. In this final rule, we are instead finalizing startup and shutdown provisions that apply to these venting events. These startup and shutdown provisions are work practice standards that allow refinery owners or operators to open process equipment Start Printed Page 75219during startup and shutdown provided that the equipment is drained and purged to a closed system until the hydrocarbon content is less than or equal to 10-percent of the LEL. For those situations where 10-percent LEL cannot be demonstrated (no direct measurement location), the equipment may be opened and vented to the atmosphere if the pressure is less than or equal to 5 psig. Active purging of the equipment is only allowed after the 10-percent LEL level is achieved, regardless of the pressure of the equipment/vessel. We are establishing a separate requirement for very small process equipment, defined as equipment where it is physically impossible to release more than 72 lbs VOC per equipment opening based on the size and contents of the equipment. This definition is consistent with the Group 1 applicability cutoff for control of miscellaneous process vents. We also developed requirements specific to catalyst changeout activities where pyrophoric catalyst (e.g., hydrotreater or hydrocracker catalysts) must be purged using recovered hydrogen. These provisions include: Documenting the procedures for equipment openings and procedures for verifying that events meet the specific conditions above using site procedures used to de-inventory equipment for safety purposes (i.e., hot work or vessel entry procedures) and documenting any deviations from the work practice standard requirements.

b. Refinery MACT 2

We are expanding the proposed 1-percent minimum oxygen operating limit alternative for organic HAP to apply for all FCCU startup and shutdown events (rather than only partial burn FCCU with CO boilers during startup). We are replacing the proposed opacity limit alternative to the metal HAP standard with a minimum cyclone face velocity limit and we are extending that alternative limit to all FCCU (regardless of control device) for both startup and shutdown in this final rule.

We are extending the proposed alternative for SRU to monitor incinerator temperature and excess oxygen limits during SRU shutdowns to also apply during periods of startup.

3. What key comments did we receive on the SSM revisions and what are our responses?

a. Refinery MACT 1

Comment: Many commenters stated that the proposed extension of the MPV definition to episodic maintenance startup and shutdown vents and elimination of the SSM exception for storage tanks would create hundreds or thousands of new vents per refinery per year and generate massive on-going burdens. The commenters argued that the EPA has not included in the record any analysis of the potential environmental benefits, costs or operational and compliance feasibility and impacts associated with this requirement and that many of these requirements will result in delayed and extended equipment and process outages. One commenter asserted that the EPA has articulated no justification for applying emission standards to these events, nor any analysis consistent with CAA section 112 for a determination that MACT standards are appropriately applied to these emission events under the criteria in CAA section 112(d).

Many commenters stated that every time a vessel is opened for inspection or maintenance each vent point will have to be evaluated as a potential MPV or storage tank vent. If a particular vent point (e.g., bleeder) used for maintenance, startup or shutdown handles material that is initially greater than 20 ppm HAP, then it is a MPV. If there is a potential to emit greater than or equal 72 lbs/day of VOC, then it is a Group 1 MPV and must be controlled. If there is a potential of less than 72 lb/day VOC release, then it is a Group 2 MPV and subject to recordkeeping requirements. Commenters stated that in a refinery there would be tens or more such activities per day associated with normal maintenance and inspection; during turnarounds, there could be hundreds of such MPVs. Commenters added that these MPVs may then need to be individually accounted for and permitted creating an unnecessary permitting and recordkeeping burden for these periodic emissions.

Commenters recommended a general set of work practice requirements for maintenance, startup and shutdown of vents, based on state requirements, that do not impose the permitting, notice and evaluation requirements associated with identifying these vents individually. Commenters explained that states have dealt with these episodic vents by establishing them as a special class of process vent with limited recordkeeping requirements and subject to a work practice standard, rather than the normal MPV requirements. A key element of these work practices is clear identification of the criteria for releasing these vents to the atmosphere and for routing these vents to control after hydrocarbon is reintroduced, which the commenters asserted the current rule does not provide. Commenters proposed that a work practice standard could include removing process liquids to the extent practical and depressuring smaller volume equipment until a pressure of <5 psig is achieved and/or purging and depressuring to a control device until the vent has a hydrocarbon concentration of less than 10-percent of the LEL. The commenters suggested that these standards should provide clear easily monitored criteria for when this equipment can be vented to the atmosphere, and should not impose the permitting, notice and evaluation requirements associated with identifying these vents as individual MPVs. One commenter provided draft regulatory language for a work practice requirement.

Response: We proposed to eliminate the episodic and non-routine emission exclusion in order to ensure that the MACT includes emission limits that apply at all times consistent with the holding in Sierra Club. At the time of the proposal, we expected that essentially all SSM event emissions would be routed to flares that are subject to the MACT standards and, thus, would serve to control these emissions. However, we recognize that maintenance activities that require equipment openings are a separate class of startup/shutdown emissions because there must be a point in time when the vessel can be opened and any emissions vented to the atmosphere. We acknowledge that it would require a significant effort to identify and characterize each of these potential release points for permitting purposes.

In considering these comments and whether we should establish a separate limit that would apply to these equipment openings, we reviewed state permit requirements and the practices employed by the best performing sources. We found that some state or local agencies required depressuring to 5 psig prior to atmospheric releases while others required the gases to have organic concentrations at or below 10-percent of LEL prior to atmospheric venting. In the final rule, we are establishing a requirement that prior to opening process equipment to the atmosphere, the equipment must first be drained and purged to a closed system so that the hydrocarbon content is less than or equal to 10-percent of the LEL. For those situations where 10-percent LEL cannot be demonstrated, the equipment may be opened and vented to the atmosphere if the pressure is less than or equal to 5 psig, provided there is no active purging of the equipment to the atmosphere until the LEL criterion is met. For equipment where it is not technically possible to depressurize to a Start Printed Page 75220control system, we allow venting to the atmosphere where there is no more than 72 lbs VOC per day potential, consistent with our Group 1 applicability cutoff for control of process vents. For catalyst changeout activities where hydrotreater pyrophoric catalyst must be purged we have provided limited allowances for direct venting. Provisions to demonstrate compliance with this work practice include documenting the procedures for equipment openings and procedures for verifying that events meet the specific conditions above using site procedures used to de-inventory equipment for safety purposes (i.e., hot work or vessel entry procedures).

b. Refinery MACT 2

Comment: Several commenters noted that there was a proposed specific alternative metal HAP/PM standard for startup of an FCCU controlled with an ESP, but took issue with the fact that no alternative PM limits were proposed for startup of FCCU equipped with other types of PM controls, or for any FCCU during periods of shutdown or hot standby. Regarding the proposed alternative for startup, which would provide an alternative in the form of an opacity limit when torch oil is in use, commenters stated that there are serious process safety concerns which prevent most FCCU ESPs from being operated when torch oil is in the regenerator, that is, during periods of startup, shutdown and hot standby. To avoid the possibility of a fire and explosion, the commenters claimed ESPs are usually de-energized and bypassed during these periods and, consequently, these FCCUs are generally unable to meet the proposed 30-percent opacity limit.

Several commenters stated that the EPA's limits on FCCU opacity during SSM are unreasonable and ignore the technical requirements for transitional operations of those units. The commenters indicated that they have ESPs located downstream of the CO boiler and claimed that for safety reasons the CO boiler cannot operate during startup, shutdown or hot standby. Further, a commenter indicated that the ESP cannot operate if the CO boiler is not operating and thus both the CO boiler and the ESP must be bypassed during startup, shutdown, and hot standby operations.

Another commenter stated that the EPA offers no data to support the achievability of this requirement in practice and discusses information for 26 startup/shutdown events that found that none complied with a 30-percent opacity requirement. Several commenters also noted that experience has shown that the 30-percent opacity limit is unachievable during these periods for FCCUs controlled with tertiary cyclones, when regenerator gas flow is below cyclone minimum design flow.

Several commenters suggested that the EPA establish a standard based on the operation of FCCU catalyst regenerators' internal cyclones that function to retain the catalyst in the regenerators and thereby minimize catalyst and metal HAP emissions from the regenerators. Additional control to meet the Refinery MACT 2 emission limit of not more than 1.0 lb PM/1,000 lbs coke burn-off is provided by a bag house, wet gas scrubber (WGS), ESP or tertiary (external) cyclone. The efficiency of a cyclone is a function of the inlet gas velocity. Assuring adequate velocity to the internal cyclones ensures that the catalyst sent to these additional controls is minimized and ensures that they are operating as effectively as possible. Similarly, even if the FCCU cannot meet the normal opacity limits during startup, shutdown or hot standby (e.g. due to the ESP being off-line for safety reasons or the tertiary cyclones or WGS operating at non-routine conditions), assuring adequate velocity to the internal regenerator cyclones will control and minimize particulate emissions. Several commenters stated support for another commenter's position that all FCCUs should be allowed the option of complying with a 20 feet/second minimum inlet velocity to the primary regenerator cyclones during periods of startup and shutdown, including hot standby, and these commenters provided additional technical explanations in their comments.

On the other hand, some commenters seemed to support the proposed opacity limits, but suggested minor revisions. One commenter noted that the SCAQMD has granted Valero's request for variances from visible emission standards during startup of the FCCU of up to 65-percent opacity for up to five minutes, in aggregate, during any 1-hour period, and 30-percent as an hourly average for the remaining period, during startup events. The application of this variance reflects the unavailability and/or ineffectiveness of the ESP during the startup condition. Another commenter recommended that either the opacity standard should be raised or the time period for averaging should be extended so FCCUs can be operated safely during SSM events and still remain in compliance.

Response: We have reviewed the data submitted by the commenters to support their assertion that the 30-percent opacity limit (determined on a 6-minute average basis) is not achievable during startup and shutdown events. While the data are limited, and it is unclear if the data provided are indicative of the performance achieved by the best performing sources, we do not have adequate data to refute the assertion that the 30-percent opacity limit (determined on a 6-minute average basis) is not achievable during startup and shutdown events. We considered the two options suggested by the commenters, the minimum velocity for the internal FCCU regenerator cyclones and the 30-percent hourly average opacity limit excluding 5 minutes not exceeding 65-percent opacity. Again, due to the limited data available during startup and shutdown events, we are not able to determine which requirement would provide greater HAP emissions reduction. However, we note that some facilities may not be required to have an opacity monitoring system in place and opacity monitoring is not applicable for FCCU controlled with wet scrubbers. Therefore, we find that the minimum internal cyclone inlet velocity requirement is more broadly applicable than the opacity limit. Also, based on the data provided by the commenters, the minimum internal cyclone inlet velocity requirement will provide PM (and therefore metal HAP) emissions reductions during startup and shutdown periods. Therefore, considering the available data, we conclude that MACT for FCCU startup and shutdown events is maintaining the minimum internal cyclone inlet velocity of 20 feet/second.

Comment: Several commenters stated that the EPA should provide alternate standards for startups of FCCU equipped with CO boilers and for any FCCU during periods of shutdown and hot standby. The commenters stated that the EPA incorrectly assumes that refiners are able to safely and reliably start up their FCCU with flue gas boilers in service and meet the normal operating limit of 500 ppm CO. They claimed that most refiners are unable to reliably start up their FCCU with flue gas boilers in service due to the design of the boiler and the fact that many boilers are not able to safely and reliably handle the transient FCCU operations that can occur during startup, shutdown, and hot standby. One commenter stated that FCCU built with CO boilers experience issues with flame stability due to fluctuating flue gas compositions and rates when starting up and shutting down. Accordingly, the commenter stated, startup and shutdown activities at FCCU using a boiler as an APCD are not currently meeting the Refinery MACT 2 standard Start Printed Page 75221of 500 ppm CO on a 1-hour basis, and this level of control does not qualify as the MACT floor. The commenter gave examples of facilities where FCCU, including those equipped with post-combustion control systems, do not consistently demonstrate compliance with a 500 ppm CO concentration standard during all startup and shutdown events.

Commenters stated that reliable boiler operation is critical to the overall refinery steam system and refineries must avoid jeopardizing boiler operation to prevent major upsets of process operations. A major upset or site-wide shutdown could result in flaring and emissions of HAP far in excess of that emitted while bypassing the CO boiler.

Commenters stated that combustion of torch oil in the FCCU regenerator during startup is one of the primary reasons the CO limit cannot be met during these operations. Torch oil is also used during shutdown to control the cooling rate (and potential equipment damage) and during hot standby and, thus, the normal CO standard cannot be met at these times either. Hot standby is used to hold an FCCU regenerator at operating temperature for outages where a regenerator shutdown is not needed and to avoid full FCCU shutdowns. Full cold shutdown also increases personnel exposures associated with removing catalyst and securing equipment. Additionally, this can produce additional emissions over maintaining the unit in hot standby. Commenters claimed that because of the variability of CO during torch oil operations, it is not possible for the EPA to establish a CAA section 112(d) standard for startup and shutdown activities at FCCU because refineries cannot measure a constant level of emissions reductions.

The commenters recommended expansion of the proposed standard of greater than 1-percent hourly average excess regenerator oxygen to all FCCU, including units with fired boilers. These commenters suggested that maintaining an adequate level of excess oxygen for the combustion of fuel in the regenerator is the best way to minimize CO and organic HAP emissions from FCCU during these periods.

Response: After reviewing the comments and discussing CO boiler operations with facility operators, we agree that the 1-percent minimum oxygen limit should be more broadly applicable to FCCU startup and shutdown regardless of the control device configuration and have revised the final rule accordingly.

Comment: Several commenters stated that the proposed alternative standards for SRP shutdowns should be extended to startups as well since the normal SRP emission limitation cannot always be achieved during SRP startups. Several commenters gave examples of startup activities where this relief is needed, and noted there may be other startup activities that also need this relief.

Response: For the control of sulfur HAP, we determined that incineration effectively controls these HAP. We were not aware that there would be unusual sulfur loads in the SRU tail gas during startup. We agree that the alternative standard we proposed for periods of shutdown is also the MACT floor for periods of startup because incineration meeting the limits proposed will achieve the MACT control requirements for sulfur HAP during periods of either startup or shutdown even though sulfur loadings during these periods may be elevated. For many SRU configurations, compliance during normal operations is demonstrated by monitoring SO2 emissions. However, during startup and shutdown, high sulfur loadings in the SRU tail gas entering the incinerator will cause high SO2 emissions even though sulfur HAP emissions are well controlled. Consequently, the proposed incinerator operating limits provide a better indication of sulfur HAP control during startup and shutdown than SO2 emissions. Owners or operators that use incinerators or thermal oxidizers during normal operations may meet the site-specific temperature and excess oxygen operating limits that were determined based on their performance test during periods of startup and shutdown.

4. What is the rationale for our final approach and final decisions to address emissions during periods of SSM?

a. Refinery MACT 1

We did not receive comments regarding the proposed amendments to Table 6 of subpart CC of 40 CFR part 63; therefore, for the reasons provided in the preamble to the proposed rule, we finalizing these amendments as proposed.

We determined that it was overly burdensome and in most cases technically infeasible to consider every potential equipment or vessel opening and classify these “openings” (newly classified as MPV in the proposal) as either Group 1 or Group 2 MPV. We also determined that it is not always technically feasible, depending on the opening, to demonstrate compliance with the MPV emissions limitations. After considering the public comments, we determined it was appropriate to establish separate startup and shutdown provisions for MPV associated with process equipment openings. We reviewed state and local requirements and based the final rule requirements on the emissions limitations required to be followed by the best performing sources. Therefore, we are finalizing requirements for refinery owners or operators to open process equipment during these startup and shutdown events without directly permitting these “vents” as Group 1 or Group 2 MPV provided that the equipment is drained and purged to a closed system until the hydrocarbon content is less than or equal to 10-percent of the LEL. As described in further detail previously in this section, we have provided provisions for special cases where the 10-percent LEL limit cannot be demonstrated and provisions for less significant equipment openings, consistent with the practices used by the best performing facilities.

b. Refinery MACT 2

We did not receive significant comments regarding the proposed amendments to Table 44 to subpart UUU of 40 CFR part 63; therefore, we finalizing these amendments as proposed.

In response to comments, we determined that the limited provisions that were provided for startup only or for shutdown only were too limited and we have expanded the proposed provisions to both startup and shutdown regardless of control device used. For the FCCU organic HAP emissions limit, we are finalizing an alternative limit for periods of startup of no less than 1-percent oxygen in the exhaust gas as proposed, but we are extending that alternative limit to shutdown and to all FCCU in this final rule.

For the FCCU metal HAP emissions limit, we proposed a specific startup limit for FCCU controlled be an ESP of 30-percent opacity. We received comments along with limited data suggesting that this limit was not achievable. Commenters suggested that the best performing units maintain a minimum face velocity of at least 20 feet/second to minimize catalyst PM losses during startup and shutdowns. Operators of wet scrubbers also noted that they cannot maintain pressure drops and that one cannot meet the PM emissions limit normalized by coke burn-off rate when the coke burn-off rate approaches zero. Consequently, commenters stated that the alternative limits should be provided for startup and shutdown regardless of control device. Upon consideration of the comments, we determined that it was necessary to revise the proposed Start Printed Page 75222alternative to be based on minimum inlet face velocity to the FCCU regenerator internal cyclones and provide the alternative for both startup and shutdown. We also expanded this limit to all FCCU; however, we also required FCCU with wet scrubbers to meet only the liquid to gas ratio operating limit during periods of startup and shutdown to allow wet scrubbers to use a consistent compliance method at all times.

For SRU, we are finalizing an alternative standard during periods of startup and shutdown to use a flare that meets the operating limits included in the final rule or a thermal oxidizer or incinerator operated at a minimum hourly average temperature of 1,200 °F and a minimum hourly average outlet oxygen concentration of 2 volume percent (dry basis). We proposed these alternatives for periods of shutdown only, but based on comments received regarding startup issues, we determined that high sulfur loadings can occur during periods of startup and that the alternative limit proposed was appropriate for both startup and shutdown.

E. Technical Amendments to Refinery MACT 1 and 2

1. What other amendments did we propose for Refinery MACT 1 and 2?

We proposed a number of amendments to Refinery MACT 1 and 2 to address technical issues such as rule language clarifications and reference corrections. First, we proposed to amend Refinery MACT 1 to clarify what is meant by “seal” for open-ended valves and lines that are “sealed” by the cap, blind flange, plug, or second valve by stating that sealed means when there are no detectable emissions from the open-ended valve or line at or above an instrument reading of 500 ppm. Second, we also proposed electronic reporting requirements where owners or operators of petroleum refineries must submit electronic copies of required performance test and performance evaluation reports for compliance with Refinery MACT 1 and 2 by direct computer-to-computer electronic transfer using EPA-provided software. Third, we proposed to update the General Provisions Tables 6 (for Refinery MACT 1) and 44 (for Refinery MACT 2) to correct cross references and to incorporate additional sections of the General Provisions that are necessary to implement these rules.

2. How did the other amendments for Refinery MACT 1 and 2 change since proposal?

We are not finalizing the definition of “seal” for open-ended lines as proposed. We are finalizing changes to update the General Provisions cross-reference tables as proposed, with one minor change to provide an option for the administrator to issue guidance on performance test reporting timeframes in order to address issues relating to submittal of data to the ERT.

3. What key comments did we receive on the other amendments for Refinery MACT 1 and 2 and what are our responses?

Comment: Numerous commenters objected to the proposal to clarify the meaning of “seal” as it relates to open-ended line (OEL) standards. Commenters contend that there is no basis for the EPA to assert that the proposed definition merely “clarifies” an established interpretation of the term “seal” and stated that the proposed revision constitutes an illegal change in the requirements for OELs, and the clarification should not be finalized.

One commenter stated that none of the MACT standards in place before this proposal have stated or suggested that a “sealed” OEL is one with detectable emissions below 500 ppm. This commenter added this unique interpretation of the requirement to “seal” an OEL with a cap or plug is incompatible with the historical interpretation of this requirement by affected facilities and by the EPA, and the EPA has not issued any sort of definitive guidance or interpretation setting out this position. The commenter detailed numerous references to considerations the EPA has made relative to OEL requirements in LDAR programs. In addition to the examples cited, the commenter noted that in 2006, the EPA proposed to add a “no detectible emissions” limit and monitoring requirement for OELs to NSPS VV (71 FR 65317, November 7, 2006). Two commenters noted that the proposed monitoring was not finalized in either NSPS VV or VVa (72 FR 64860, November 16, 2007) because it was not considered BDT due to the low emission reductions and the cost effectiveness of the requirement. Another commenter agreed that there is no explanation provided for why this information could now support the need for a new OEL seal standard that requires monitoring to ensure compliance when it was deemed to be unjustified previously.

In addition, the commenter collected OEL monitoring data and submitted it to the EPA (see Docket Item No. EPA-HQ-OAR-2010-0869-0058). Based on these data, the commenter asserted that the existence of leaks from OELs that are not properly sealed is extremely low.

The commenter noted that the EPA is claiming this change is only a clarification of current requirements, allowing the EPA to bypass the need to cite a CAA authorization for this change to the existing CAA section 112(d)(2) standard or meet the process requirements associated with such a change, including providing emission reduction, cost and burden estimates in the record and the associated PRA Information Collection Request (ICR).

Several commenters claimed that this clarification would result in retroactive impact and also addressed the implication of the proposed change on other fugitive emissions standards. One commenter stated that the EPA cannot retroactively reinterpret the OEL requirements or define the word “seal” and added that the EPA should account for the thousands of additional monitoring events per year per refinery that this new requirement would add to LDAR programs and provide proper cost justification under CAA sections 112(d)(6) or 112(f)(2).

Several commenters also stated that the proposed definition will effectively change all equipment leak rules in parts 40 CFR parts 60, 61 and 63 and the change should not be finalized. One commenter added that by claiming this change is only a clarification of current requirements, the EPA would set a precedent applicable to all OELs in all industries subject to any similar OEL equipment leak requirement.

Response: We have decided not to finalize the proposed clarification of the term “seal” for OELs at this time. The fenceline monitoring requirements we are finalizing will detect any significant leaks from a cap, blind flange, plug or second valve that does not properly seal an OEL, as well as significant leaks from numerous other types of fugitive emission sources.

Comment: A few commenters stated that the proposed use of the ERT is not appropriate because the costs and burdens imposed are additive to the costs of producing and submitting the written report, and there is no benefit that justifies the additional cost. One commenter also stated that the EPA has not developed or articulated a reasonable approach to using information that would be uploaded to the ERT. The commenters recommended that the EPA remove this portion of the proposal until the ERT is demonstrated to handle all the information from refinery performance Start Printed Page 75223tests (rather than only portions), thereby eliminating the need for both written and electronic reporting and until the Agency demonstrates that it is using the electronic data to develop improved air quality emission factors.

Other commenters stated that the ERT requirement does not supersede or replace any state reporting requirements and thus the regulated industry will be subject to dual reporting requirements. These commenters disagreed with the preamble claim that eliminating the recordkeeping requirements for performance test reports is a burden savings, and stated that it may duplicate burdens already borne by the regulated community.

The commenters expressed further concern that duplicative reporting requirements will strain the regulated industry to comply with deadlines established by rule for report submittals. One commenter stated that there is no mechanism for obtaining extensions for special circumstances. Under proposed 40 CFR 63.655(h)(9)(i), all reports are due in 60 days. The commenter claimed that by not referencing reporting requirements to the General Provisions in 40 CFR 63.10(d)(2), there is no allowance for obtaining additional time due to unforeseen circumstances or due to the difficulties involved with completing particularly complex reports.

One commenter stated that the primary performance test method (Method 18) required for determining compliance is not currently included in the list of methods supported by the ERT. The commenter stated that the regulated community's experience with Method 18 is that it is a very broad methodology and can be exceptionally complex to execute and to report. The commenter stated that the EPA is aware that Method 18 reporting is complex, that it may be difficult to incorporate into the ERT, and that no time schedule has been defined for development or implementation for this method.

The commenter also stated that without formal notice of changes to the ERT, the regulated community is at risk of non-compliance. The only way for the regulated community to know that changes have occurred in the ERT is to monitor the Web site directly because the EPA does not formally announce changes to the ERT in the Federal Register. As such, it would be possible for a regulated entity to be unaware of changes made such as the incorporation of Method 18. The commenter expressed concern that the proposal language is an open-ended commitment subject to change without notice. The commenter stated that the EPA should clearly indicate when facilities would be required to use the ERT when new test methods are included in the ERT.

Response: We disagree that use of the ERT for completing stack test reports is an added cost and burden. While the requirement to report the results of stack tests with the ERT does not supersede state reporting requirements, we are aware of several states that already require the use of the ERT, and we are aware of more states that are considering requiring its use. We note that where states will not accept an electronic ERT submittal, the ERT provides an option to print the report, and the printed report can be mailed to the state agency. We have no reason to believe that the time savings in the ability to reuse data elements within reports does not, at a minimum, offset the cost incurred by printing out and mailing a copy of the report and the commenters have provided no support for their cost claims.

Furthermore, based on the analysis performed for the Electronic Reporting and Recordkeeping Requirements for the New Source Performance Standards Rulemaking (ERRRNSPS) (80 FR 15100), electronic reporting results in an overall cost savings to industry when annualized over a 20-year period. The cost savings is achieved through means such as standardization of data, embedded quality assurance checks, automatic calculation routines and reduced data entry through the ability to reuse data in files instead of starting from scratch with each test. As outlined in the ERRRNSPS, there are many benefits to electronic reporting. These benefits span all users of the data—the EPA, state and local regulators, the regulated entities and the public. We note that in the preamble to this proposed rule we provided a number of reasons why the use of the ERT will provide benefit going forward and that most of the benefits we outlined were longer-term benefits (e.g., reducing burden of future information collection requests). Additionally, we note that in 2011, in response to Executive Order 13563, the EPA developed a plan [13] to periodically review its regulations to determine if they should be modified, streamlined, expanded or repealed in an effort to make regulations more effective and less burdensome. The plan includes replacing outdated paper reporting with electronic reporting. In keeping with this plan and the White House's Digital Government Strategy, [14] in 2013 the EPA issued an agency-wide policy specifying that new regulations will require reports to be electronic to the maximum extent possible. By requiring electronic submission of stack test reports in this rule, we are taking steps to implement this policy. We also disagree that we have not developed or articulated a reasonable approach to using information that would be uploaded to the ERT. To the contrary, we have discussed at length our plans for the use of stack test data collected via the ERT. In 2009, we published an advanced notice of proposed rulemaking (74 FR 52723) for the Emissions Factors Program Improvements. In that notice, we first outlined our intended approach for revising our emissions factors development procedures. This approach included using stack test data collected with the ERT. We reiterated this position in our “Recommended Procedures for the Development of Emissions Factors and Use of the WebFIRE Database” (http://www.epa.gov/​ttn/​chief/​efpac/​procedures/​procedures81213.pdf), which was subject to public notice and comment before being finalized in 2013. Finally, we discussed uses of these data in the preamble to the proposed rule and at length in the preamble to the ERRRNSPS.

We think that it is a circular argument to say that the agency should eliminate the use of the ERT until it demonstrates that it is using the electronic data. It would be impossible for the agency to use data that it does not have. We can only use electronic data once we have electronic data. We do note that we are nearing completion of programming the WebFIRE database with our new emissions factor development procedures and anticipate running the routines on existing data sets in the near future.

We continue to improve and upgrade the ERT on an ongoing basis. The current version of the ERT supports 41 methods, including EPA Methods 1-4, 5, 5B, 5F, 25A 26, and 26A. We note that the ERT does not currently support EPA Method 18, and for performance tests using Method 18, the source will still have to produce a paper report. However, we are aware of the need to add Method 18 to the ERT, and we are currently looking at developing this capability. As noted in the ERRRNSPS, when new methods are added to the Start Printed Page 75224ERT, we will not only post them to the Web site; we will also send out a listserv notice to the Clearinghouse for Inventories and Emissions Factors (CHIEF) listserv. Information on joining the CHIEF listserv can be found at http://www.epa.gov/​ttn/​chief/​listserv.html#chief. We are requiring the use of the ERT if the method is supported by the ERT, as listed on the ERT Web site (http://www.epa.gov/​ttn/​chief/​ert/​ert_​info.html) at the time of the test. We do not agree that it is overly burdensome to check a Web site for updates prior to conducting a performance test.

We did revise the MACT 1 and 2 tables referencing reporting requirements to the general provisions (Table 6 for Refinery MACT 1 and Table 44 for Refinery MACT 2) to provide flexibility in the 60-day reporting timeline to accommodate unforeseen circumstances or difficulties involved with completing particularly complex reports.

4. What is the rationale for our final approach and final decisions for the other amendments for Refinery MACT 1 and 2?

We are not finalizing the definition of seal, as proposed. The fenceline monitoring work practice standard will detect any significant leaks from a cap, blind flange, plug or second valve that does not properly seal an OEL, as well as significant leaks from numerous other types of fugitive emission sources.

We are finalizing requirements for electronic reporting, as proposed, with a minor clarification. Specifically, we are revising Tables 6 in subpart CC and 44 in subpart UUU, which cross-reference the applicable provisions in the General Provisions to provide flexibility in the ERT 60-day reporting timeline. Refiners can seek approval from the EPA or a delegated state additional time for submittal of data due to unforeseen circumstances or due to the difficulties involved with completing particularly complex reports.

F. Technical Amendments to Refinery NSPS Subparts J and Ja

1. What amendments did we propose for Refinery NSPS Subparts J and Ja?

We proposed a number of amendments to Refinery NSPS subparts J and Ja to address reconsideration issues and minor technical clarifications. First, we proposed revisions to 40 CFR 60.100a(b) to include a provision that sources subject to Refinery NSPS subpart J could elect to comply instead with the provisions of Refinery NSPS subpart Ja.

Second, we proposed a series of amendments to the requirements for SRP in 40 CFR 60.102a, to clarify the applicable emission limits for different types of SRP based on whether oxygen enrichment is used. The amendments proposed also clarified that emissions averaging across a group of emission points within a given SRP is allowed for each of the different types of SRP, and that emissions averaging is specific to the SO2 or reduced sulfur standards (and not to the 10 ppmv hydrogen sulfide (H2 S) limit). We also proposed a series of corresponding amendments in 40 CFR 60.106a to clarify the monitoring requirements, particularly when oxygen enrichment or emissions averaging is used. We also proposed clarifications in 40 CFR 60.106a to consistently use the term “reduced sulfur compounds” when referring to the emission limits and monitoring devices needed to comply with the reduced sulfur compound emission limits for sulfur recovery plants with reduction control systems not followed by incineration.

Third, we proposed amendments to 40 CFR 60.102a(g)(1) to clarify that CO boilers, while part of the FCCU affected facility, can also be FGCD.

Fourth, we proposed several revisions to 40 CFR 60.104a to clarify the performance testing requirements. We proposed revision to 40 CFR 60.104a(a) to clarify that an initial compliance demonstration is needed for the H2 S concentration limit in 40 CFR 60.103a(h). We proposed revisions to the annual PM testing requirement in 40 CFR 60.104a(b) to clarify that annually means once per calendar year, with an interval of at least 8 months but no more than 16 months between annual tests. We also proposed to amend 40 CFR 60.104a(f) to clarify that the provisions of that paragraph are specific to owners or operators of an FCCU or FCU that use a cyclone to comply with the PM emissions limit in 40 CFR 60.102a(b)(1) and not to facilities electing to comply with the PM emissions limit using a PM CEMS. We also proposed to amend 40 CFR 60.104a(j) to delete the requirements to measure flow for the H2 S concentration limit for fuel gas.

Fifth, we proposed several amendments to clarify the requirements for control device operating parameters in 40 CFR 60.105a. Specifically, we proposed amendments to 40 CFR 60.105a(b)(1)(ii)(A) to require corrective action be completed to repair faulty (leaking or plugged) air or water lines within 12 hours of identification of an abnormal pressure reading during the daily checks. We also proposed revisions to 40 CFR 60.105a(i) to specify that periods when abnormal pressure readings for a jet ejector-type wet scrubber (or other type of wet scrubber equipped with atomizing spray nozzles) are not corrected within 12 hours of identification and periods when a bag leak detection system alarm (for a fabric filter) is not alleviated within the time period specified in the rule are considered to be periods of excess emissions.

We also proposed amendments to 40 CFR 60.105(b)(1)(iv) and 60.107a(b)(1)(iv) to provide flexibility in span range to accommodate different manufacturers of the length-of-stain tubes. We also proposed to delete the last sentence in 40 CFR 60.105(b)(3)(iii).

Finally, we proposed clarification to the performance test requirements for the H2 S concentration limit for affected flares in 40 CFR 60.107a(e)(1)(ii) and (e)(2)(ii) to remove the distinction between flares with or without routine flow.

2. How did the amendments to Refinery NSPS Subparts J and Ja change since proposal?

We are making very few changes to the amendments proposed for Refinery NSPS subparts J and Ja. In response to comments, we are revising the NSPS requirements to replace the “measurement sensitivity” requirements with accuracy requirements consistent with those used in Refinery MACT 1 and 2. Specifically, we are revising 40 CFR 60.106a(a)(6)(i)(B) and (7)(i)(B) to require use of a flow sensor meeting an accuracy requirement of ±5-percent over the normal range of flow measured or 10-cubic-feet-per-minute, whichever is greater. We are also revising the flare accuracy requirements in 40 CFR 60.107a(f)(1)(ii) to require use of a flow sensor meeting an accuracy requirement of ±20-percent of the flow rate at velocities ranging from 0.1 to 1 feet per second and an accuracy of ±5-percent of the flow rate for velocities greater than 1-feet-per-second.

Finally, we are revising 40 CFR 60.101a(b) to correct an inadvertent error where the phrase “and delayed coking units” was not included in the proposed sentence revision.

3. What key comments did we receive on the amendments to Refinery NSPS Subparts J and Ja and what are our responses?

Comment: Two commenters noted concern with the term “measurement sensitivity” in proposed 40 CFR 60.106a(a)(6)(i)(B) and (a)(7)(i)(B) for sulfur recovery unit monitoring alternatives and in existing regulations 40 CFR 60.107a(f)(1)(ii) for flares because “sensitivity” is not a term Start Printed Page 75225found on typical monitoring system data sheets. Typical flow meter characteristics include terms such as accuracy and resolution and the commenters requested that the EPA revise the terminology to match the wording found in 40 CFR part 63, subpart CC, Table 13 for flow meters (i.e., accuracy requirements). Additionally, several commenters suggested that the EPA flow monitor accuracy specifications are inconsistent with those in the SCAQMD Flare Rule and many refinery consent decrees. The commenters recommended revising both the flare flow meter sensitivity specification and accuracy specification in Refinery MACT 1 Table 13 and in Refinery NSPS subpart Ja to be consistent with the accuracy specification from the Shell Deer Park Consent Decree, Appendix 1.10, which specifies the required flare flow meter accuracy as “±20% of reading over the velocity range of 0.1-1 feet per second (ft/s) and ±5% of reading over the velocity range of 1-250 ft/s.”

Response: We proposed the term “measurement sensitivity” in proposed 40 CFR 60.106a(a)(6)(i)(B) and (a)(7)(i)(B) to be internally consistent within Refinery NSPS subpart Ja [i.e., consistent with the existing language in § 60.107a(f)(1)(ii)]. However, we agree with the commenters that this term may be unclear. This term is not defined in Refinery NSPS subpart Ja and it is not commonly used in the flow monitoring system's technical specification sheets. Therefore, to be consistent with the terminology used by instrument vendors and used in Refinery MACT 1 and 2, we are revising these sections to replace the term “measurement sensitivity” with “accuracy.” We are also revising the flow rate accuracy provisions specific for flares to provide an accuracy requirement of ±20-percent over the velocity range of 0.1-1 ft/s and ±5% for velocities exceeding 1 ft/s in 40 CFR 60.107a(f)(1)(ii) and in Table 13 of subpart CC. We are providing this provision specifically for flares because they commonly operate at high turndown ratios. For other flow measurements, we are retaining the 10-cubic-foot-per-minute accuracy requirement. We are also clarifying that the ±5-percent accuracy requirement for the SRU alternatives apply to the “the normal range of flow measured” consistent with the requirements in Refinery MACT 1 and 2.

Comment: One commenter stated that in the proposed revisions to 40 CFR 60.100a, (79 FR 36956), the EPA proposes to remove the phrase “and delayed coker units” from 40 CFR 60.100a(b). However, we state the compliance date for both flares and delayed coker units separately in the same paragraph. The commenter believes the EPA should explain the reason for and implications of the removal of this phrase.

Response: The removal of the phrase “and delayed coking units” from the first sentence in 40 CFR 60.100a(b) was an inadvertent error. The only revision that we intended to make in 40 CFR 60.100a was to allow owners or operators subject to subpart J to elect to comply with the requirements in subpart Ja. In the final amendments, we have included the phrase “and delayed coking units” in the first sentence in 40 CFR 60.100a(b).

4. What is the rationale for our final approach and final decisions for the amendments to Refinery NSPS Subparts J and Ja?

We are finalizing amendments for Refinery NSPS subparts J and Ja as proposed with minor revisions. In response to comments, we are revising the “measurement sensitivity” requirements to be an “accuracy” requirement. This change will make the requirements more clear and consistent between the flow meter requirements in the NSPS and the MACT standards since the same flow meter will be subject to each of these requirements. We are also providing a dual accuracy requirement for flare flow meters. This accuracy requirement is necessary because flares, which can have large diameters to accommodate high flows, are commonly operated at low flow rates. Together, this makes it technically infeasible for many flares to meet the lower flow 10 cfm accuracy requirement. Therefore, we are providing specific accuracy requirements for flares of ±20-percent over the velocity range of 0.1-1 ft/s and ±5-percent for velocities exceeding 1 ft/s, consistent with recent consent decrees and equipment vendor specifications.

Finally, we are revising the introductory phrase in the first sentence in 40 CFR 60.101a(b) to read “Except for flares and delayed coking units . . .” to correct an inadvertent error. We intended to revise this sentence only to include the proposed provision to allow sources subject to Refinery NSPS subpart J to comply with Refinery NSPS subpart Ja. The redline text posted on our Web site showed no revisions to this introductory phrase, but the amendatory text did not include the words “and delayed coking units” in this phrase. This was an inadvertent error, which we are correcting in the final rule.

V. Summary of Cost, Environmental and Economic Impacts and Additional Analyses Conducted

A. What are the affected facilities, the air quality impacts and cost impacts?

The sources affected by significant amendments to the petroleum refinery standards include flares, storage vessels, pressure relief devices, fugitive emissions and DCU. The amendments for other sources subject to one or more of the petroleum refinery standards are expected to have minimal air quality and cost impacts.

The total capital investment cost of the final amendments and standards is estimated at $283 million, $112 million from the final amendments for storage vessels, DCU and fenceline monitoring and $171 million from standards to ensure compliance. We estimate annualized costs of the final amendments for storage vessels, DCU and fenceline monitoring to be approximately $13.0 million, which includes an estimated $11.0 million for recovery of lost product and the annualized cost of capital. We also estimated annualized costs of the final standards to ensure compliance to be approximately $50.2 million. The final amendments for storage vessels, DCU and fenceline monitoring would achieve a nationwide HAP emission reduction of 1,323 tpy, with a concurrent reduction in VOC emissions of 16,660 tpy and a reduction in methane emissions of 8,700 metric tonnes per year. Table 2 of this preamble summarizes the cost and emission reduction impacts of the final amendments, and Table 3 of this preamble summarizes the costs of the final standards to ensure compliance.Start Printed Page 75226

Table 2—Nationwide Impacts of Final Amendments (2010$)

Affected sourceTotal capital investment (million $)Total annualized cost without credit (million $/yr)Product recovery credit (million $/yr)Total annualized costs (million $/yr)Methane emission reductions (metric tpy)VOC emission reductions (tpy)Cost effectiveness ($/ton VOC)HAP emission reductions (tpy)Cost effectiveness ($/ton HAP)
Storage Vessels18.53.13(8.16)(5.03)14,600(345)910(5,530)
Delayed Coking Units8114.5(2.80)11.78,7002,0605,68041328,330
Fugitive Emissions (Fenceline Monitoring)12.56.366.36
Total11224.0(11.0)13.08,70016,6607801,3239,830

Table 3—Nationwide Costs of Final Amendments To Ensure Compliance (2010$)

Affected SourceTotal capital investment (million $)Total annualized cost without credit (million $/yr)Product recovery credit (million $/yr)Total annualized costs (million $/yr)
Relief Device Monitoring11.13.33.3
Flare Monitoring16046.546.5
FCCU Testing0.40.4
Total17150.250.2

The impacts shown in Table 2 do not include costs, product recovery credits, or emissions reductions associated with any root cause analysis or corrective action taken in response to the final amendments for fenceline monitoring. The impacts shown in Table 3 do not include (i) the costs or emissions reductions associated with any root cause analysis and corrective action taken in response to the final source performance testing at the FCCUs, or (ii) emissions reductions associated with corrective action taken in response to pressure relief device or (iii) emissions reductions associated with the flare operating and monitoring provisions. The operational and monitoring requirements for flares at refineries have the potential to reduce excess emissions from flares by up to approximately 3,900 tpy of HAP and 33,000 tpy of VOC. The operational and monitoring requirements for flares also have the potential to reduce methane emissions by 25,800 metric tonnes per year while increasing emissions of carbon dioxide (CO2) and nitrous oxide by 267,000 metric tonnes per year and 2 metric tonnes per year, respectively, yielding a net reduction in GHG emissions of 377,000 metric tonnes per year of CO2 equivalents (CO2 e).

B. What are the economic impacts?

We performed a national economic impact analysis for petroleum product producers. All petroleum product refiners will incur annual compliance costs of less than 1-percent of their sales. For all firms, the minimum cost-to-sales ratio is <0.01-percent; the maximum cost-to-sales ratio is 0.87-percent; and the mean cost-to-sales ratio is 0.03-percent. Therefore, the overall economic impact of this proposed rule should be minimal for the refining industry and its consumers.

In addition, the EPA performed a screening analysis for impacts on small businesses by comparing estimated annualized engineering compliance costs at the firm-level to firm sales. The screening analysis found that the ratio of compliance cost to firm revenue falls below 1-percent for the 28 small companies likely to be affected by the proposal. For small firms, the minimum cost-to-sales ratio is <0.01-percent; the maximum cost-to-sales ratio is 0.62-percent; and the mean cost-to-sales ratio is 0.07-percent.

More information and details of this analysis is provided in the technical document “Economic Impact Analysis for Petroleum Refineries Proposed Amendments to the National Emissions Standards for Hazardous Air Pollutants”, which is available in the docket for this rule (Docket ID No. EPA-HQ-OAR-2010-0682).

C. What are the benefits?

The final rule is anticipated to result in a reduction of 1,323 tpy of HAP (based on allowable emissions under the MACT standards) and 16,660 tpy of VOC, not including potential emission reductions that may occur as a result of the operating and monitoring requirements for flares and fugitive emission sources via fenceline monitoring. These avoided emissions will result in improvements in air quality and reduced negative health effects associated with exposure to air pollution of these emissions; however, we have not quantified or monetized the benefits of reducing these emissions for this rulemaking.

D. Impacts of This Rulemaking on Environmental Justice Populations

To examine the potential impacts on vulnerable populations (minority, low-income and indigenous communities) that might be associated with the Petroleum Refinery source categories addressed in this final rule, we evaluated the percentages of various social, demographic and economic groups in the at-risk populations living near the facilities where these sources are located and compared them to national averages. Our analysis of the demographics of the population with estimated risks greater than 1-in-1 million indicates potential disparities in risks between demographic groups including the African American, Other and Multiracial, Hispanic, Below the Poverty Level, and Over 25 without a High School Diploma when compared to the nationwide percentages of those groups. These groups will benefit the most from the emission reductions achieved by this final rulemaking, which is projected to result in 1 million fewer people exposed to risks greater than 1-in-1 million.

Additionally, these communities will benefit from this rulemaking, as this rulemaking for the first time ever requires fenceline monitoring, and reporting of fenceline data. The agency during the pre-proposal period and Start Printed Page 75227during the comment period received feedback from communities on the importance of having fenceline monitoring in their communities and the importance of communities having access to this data. The EPA believes that vulnerable communities will benefit from this data and the requirements that EPA has put in place in this rulemaking to manage fugitive emissions.

E. Impacts of This Rulemaking on Children's Health

Under Executive Order 13045 the EPA must evaluate the effects of the planned regulation on children's health and safety. This action's health and risk assessments are contained in section IV.A of this preamble. We believe we have adequately estimated risk for children, and we do not believe that the environmental health risks addressed by this action present a disproportionate risk to children. When the EPA derives exposure reference concentrations and unit risk estimates (URE) for HAP, it also considers the most sensitive populations identified (i.e., children) in the available literature, and importantly, these are the values used in our risk assessments. With regard to children's potentially greater susceptibility to non-cancer toxicants, the assessments rely on the EPA (or comparable) hazard identification and dose-response values which have been developed to be protective for all subgroups of the general population, including children. With respect to cancer, the EPA uses the age-dependent adjustment factor approach, and applies these factors to carcinogenic pollutants that are known to act via mutagenic mode of action. Further details are provided in the “Final Residual Risk Assessment for the Petroleum Refining Source Sector”, Docket ID No. EPA-HQ-OAR-2010-0682.

VI. Statutory and Executive Order Reviews

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

This action is an economically significant regulatory action that was submitted to the Office of Management and Budget (OMB) for review. Any changes made in response to OMB recommendations have been documented in the docket. The EPA prepared an analysis of the potential costs and benefits associated with this action. This analysis, “Economic Impact Analysis: Petroleum Refineries—Final Amendments to the National Emissions Standards for Hazardous Air Pollutants and New Source Performance Standards” is available in Docket ID Number EPA-HQ-OAR-2010-0682.

B. Paperwork Reduction Act (PRA)

The information collection requirements in this rule have been submitted for approval to the Office of Management and Budget (OMB) under the Paperwork Reduction Act, 44 U.S.C. 3501 et se. The information collection requirements are not enforceable until OMB approves them.

Adequate recordkeeping and reporting are necessary to ensure compliance with these standards as required by the CAA. The ICR information collected from recordkeeping and reporting requirements is also used for prioritizing inspections and is of sufficient quality to be used as evidence in court.

The ICR document prepared by the EPA for the amendments to the Petroleum Refinery MACT standards for 40 CFR part 63, subpart CC has been assigned the EPA ICR number 1692.08. Burden changes associated with these amendments would result from new monitoring, recordkeeping and reporting requirements. The estimated annual increase in recordkeeping and reporting burden hours is 99,722 hours; the frequency of response is quarterly and semiannual for reports for all respondents that must comply with the rule's reporting requirements; and the estimated average number of likely respondents per year is 95 (this is the average in the second year). The cost burden to respondents resulting from the collection of information includes the total capital cost annualized over the equipment's expected useful life (about $18 million, which includes monitoring equipment for fenceline monitoring, pressure relief devices, and flares), a total operation and maintenance component (about $21 million per year for fenceline and flare monitoring), and a labor cost component (about $8.3 million per year, the cost of the additional 99,722 labor hours). Burden is defined at 5 CFR 1320.3(b).

The ICR document prepared by the EPA for the amendments to the Petroleum Refinery MACT standards for 40 CFR part 63, subpart UUU has been assigned the EPA ICR number 1844.06. Burden changes associated with these amendments would result from new testing, recordkeeping and reporting requirements being finalized with this action. The estimated average burden per response is 25 hours; the frequency of response ranges from annually up to every 5 years for respondents that have FCCU, and the estimated average number of likely respondents per year is 67. The cost burden to respondents resulting from the collection of information includes the performance testing costs (approximately $778,000 per year over the first 3 years for the initial PM and one-time HCN performance tests and $235,000 per year starting in the fourth year), and a labor cost component (approximately $410,000 per year for 4,940 additional labor hours). Burden is defined at 5 CFR 1320.3(b).

An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB control number. The OMB control numbers for EPA's regulations in 40 CFR are listed in 40 CFR part 9. When this ICR is approved by OMB, the Agency will publish a technical amendment to 40 CFR part 9 in the Federal Register to display the OMB control number for the approved information collection requirements contained in this final rule.

C. Regulatory Flexibility Act (RFA)

I certify that this action will not have a significant economic impact on a substantial number of small entities (SISNOSE) under the RFA. The small entities subject to the requirements of this action are small businesses, small organizations and small governmental jurisdictions. For purposes of assessing the impacts of this rule on small entities, a small entity is defined as: (1) A small business in the petroleum refining industry having 1,500 or fewer employees (Small Business Administration (SBA), 2011); (2) a small governmental jurisdiction that is a government of a city, county, town, school district or special district with a population of less than 50,000; and (3) a small organization that is any not-for-profit enterprise which is independently owned and operated and is not dominant in its field. Details of this analysis are presented in the economic impact analysis which can be found in the docket for this rule (Docket ID No. EPA-HQ-OAR-2010-0682).

D. Unfunded Mandates Reform Act (UMRA)

This action does not contain an unfunded mandate of $100 million or more as described in UMRA, 2 U.S.C. 1531-1538, and does not significantly or uniquely affect small governments. As discussed earlier in this preamble, these amendments result in nationwide costs of $63.2 million per year for the private sector. Additionally, the rule contains no requirements that apply to small Start Printed Page 75228governments and does not impose obligations upon them.

E. Executive Order 13132: Federalism

This action does not have federalism implications. It will not have substantial direct effects on the states, on the relationship between the national government and the states, or on the distribution of power and responsibilities among the various levels of government.

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

This action does not have tribal implications, as specified in Executive Order 13175. The final amendments impose no requirements on tribal governments. Thus, Executive Order 13175 does not apply to this action. Consistent with the EPA Policy on Consultation and Coordination with Indian Tribes, the EPA consulted with tribal officials during the development of the proposed rule and specifically solicited comment on the proposed amendments from tribal officials.

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

This action is not subject to Executive Order 13045 because the EPA does not believe the environmental health or safety risks addressed by this action present a disproportionate risk to children. This action's health and risk assessments are contained in section IV.A of this preamble.

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

This action is not a “significant energy action” because it is not likely to have a significant adverse effect on the supply, distribution or use of energy. The overall economic impact of this final rule should be minimal for the refining industry and its consumers.

I. National Technology Transfer and Advancement Act (NTTAA) and 1 CFR Part 51

This rulemaking involves technical standards. Therefore, the EPA conducted searches for the Petroleum Refinery Sector Risk and Technology Review and New Source Performance Standards through the Enhanced National Standards Systems Network (NSSN) Database managed by the American National Standards Institute (ANSI). We also contacted voluntary consensus standards (VCS) organizations and accessed and searched their databases. We conducted searches for EPA Methods 18, 22, 320, 325A, and 325B of 40 CFR parts 60 and 63, appendix A. No applicable VCS were identified for EPA Method 22.

The following voluntary consensus standards were identified as acceptable alternatives to the EPA test methods for the purpose of this rule.

The voluntary consensus standard ISO 16017-2:2003(E) “Air quality—Sampling and analysis of volatile organic compounds in ambient air, indoor air and workplace air by sorbent tube/thermal desorption/capillary gas chromatography. Part 2: Diffusive sampling” is an acceptable alternative to Method 325A, Sections 1.2, 6.1 and 6.5 and Method 325B Sections 1.3, 7.1.2, 7.1.3, 7.1.4, 12.2.4, 13.0, A.1.1, and A.2. This voluntary consensus standard gives general guidance for the sampling and analysis of volatile organic compounds in air. It is applicable to indoor, ambient and workplace air. This standard is available at International Organization for Standardization, ISO Central Secretariat, Chemin de Blandonnet 8, CP 401, 1214 Vernier, Geneva, Switzerland. See https://www.iso.org.

The voluntary consensus standard BS EN 14662-4:2005 “Ambient Air Quality: Standard Method for the Measurement of Benzene Concentrations—Part 4: Diffusive Sampling Followed By Thermal Desorption and Gas Chromatography” is an acceptable alternative to Method 325A, Section 1.2 and Method 325B, Sections 1.3, 7.1.3, 7.1.4, 12.2.4, 13.0, A.1.1, and A.2. This voluntary consensus standard gives general guidance for the sampling and analysis of benzene in air by diffusive sampling, thermal desorption and capillary gas chromatography. This standard is available the European Committee for Standardization, Avenue Marnix 17—B-1000 Brussels. See https://www.cen.eu.

The voluntary consensus standard ASTM D6420-99 (2010) “Test Method for Determination of Gaseous Organic Compounds by Direct Interface Gas Chromatography/Mass Spectrometry” is an acceptable alternative to EPA Method 18. This voluntary consensus standard employs a direct interface gas chromatography/mass spectrometer (GCMS) to identify and quantify a list of 36 volatile organic compounds (the compounds are listed in the method).

The voluntary consensus standard ASTM D6196-03 (Reapproved 2009) “Standard Practice for Selection of Sorbents, Sampling, and Thermal Desorption Analysis Procedures for Volatile Organic Compounds in Air” is an acceptable alternative to Method 325A, Sections 1.2 and 6.1, and Method 325B, Sections 1.3, 7.1.2, 7.1.3, 7.1.4, 13.0, A.1.1, and A.2. This voluntary consensus standard is intended to assist in the selection of sorbents and procedures for the sampling and analysis of ambient, indoor, and workplace atmospheres for a variety of common volatile organic compounds.

The voluntary consensus standards ASTM D1945-03 and later revision ASTM D1945-14 “Standard Test Method for Analysis of Natural Gas by Gas Chromatography” are acceptable for natural gas analysis. This voluntary consensus standard covers the determination of the chemical composition of natural gases and similar gaseous mixtures. This test method may be abbreviated for the analysis of lean natural gases containing negligible amounts of hexanes and higher hydrocarbons, or for the determination of one or more components, as required.

The voluntary consensus standard ASTM UOP539-12 “Refinery Gas Analysis by GC” is acceptable for refinery gas analysis. This voluntary consensus standard is for determining the composition of refinery gas streams or vaporized liquefied petroleum gas using a preconfigured, commercially available gas chromatograph.

The voluntary consensus standard ASTM D6348-03 (Reapproved 2010) including Annexes A1 through A8, “Determination of Gaseous Compounds by Extractive Direct Interface Fourier Transform (FTIR) Spectroscopy” is an acceptable alternative to EPA Method 320. This voluntary consensus standard is a field test method that employs an extractive sampling system to direct stationary source effluent to an FTIR spectrometer for the identification and quantification of gaseous compounds. This field test method provides near real time analysis of extracted gas samples from stationary sources.

The voluntary consensus standard ASTM D6348-12e1 “Determination of Gaseous Compounds by Extractive Direct Interface Fourier Transform (FTIR) Spectroscopy” is an acceptable alternative to EPA Method 320 with the following two caveats: (1) The test plan preparation and implementation in the Annexes to ASTM D 6348-03 (Reapproved 2010), Sections A1 through A8 are mandatory; and (2) In ASTM D6348-03 (Reapproved 2010) Annex A5 (Analyte Spiking Technique), the percent (%) R must be determined for each target analyte (Equation A5.5). In order for the test data to be acceptable for a compound, %R must be 70% ≥ R ≤ 130%. If the %R value does not meet this criterion for a target compound, the test data is not acceptable for that compound and the test must be repeated Start Printed Page 75229for that analyte (i.e., the sampling and/or analytical procedure should be adjusted before a retest). The %R value for each compound must be reported in the test report, and all field measurements must be corrected with the calculated %R value for that compound by using the following equation:

Reported Result = (Measured Concentration in the Stack × 100)/% R.

This voluntary consensus standard is a field test method that employs an extractive sampling system to direct stationary source effluent to an FTIR spectrometer for the identification and quantification of gaseous compounds. This field test method provides near real time analysis of extracted gas samples from stationary sources.

The EPA solicited comments on VCS and invited the public to identify potentially-applicable VCS; however, we did not receive comments regarding this aspect of 40 CFR part 60, subparts J and Ja, and part 63, subparts CC, UUU, and Y. Under 40 CFR 63.7(f) and 63.8(f), a source may apply to the EPA for permission to use alternative test methods or alternative monitoring requirements in place of any required testing methods, performance specifications, or procedures in this final rule.

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

Executive Order 12898 (59 FR 7629; February 16, 1994) establishes federal executive policy on environmental justice. Its main provision directs federal agencies, to the greatest extent practicable and permitted by law, to make environmental justice part of their mission by identifying and addressing, as appropriate, disproportionately high and adverse human health or environmental effects of their programs, policies and activities on minority populations and low-income populations in the U.S. The EPA defines environmental justice as the fair treatment and meaningful involvement of all people regardless of race, color, national origin or income with respect to the development, implementation and enforcement of environmental laws, regulations and policies. The EPA has this goal for all communities and persons by working to ensure that everyone enjoys the same degree of protection from environmental and health hazards and equal access to the decision-making process to have a healthy environment in which to live, learn and work.

The EPA believes the human health or environmental risk addressed by this action will not have potential disproportionately high and adverse human health or environmental effects on minority, low-income or indigenous populations. As discussed in section V.D. of this preamble, the EPA conducted an analysis of the characteristics of the population with greater than 1-in-1 million risk living within 50 km of the 142 refineries affected by this rulemaking and determined that there are more African-Americans, Other and multiracial groups, Hispanics, low-income individuals, individuals with less than a high school diploma compared to national averages. Therefore, these populations are expected to experience the benefits of the risk reductions associated with this rule. The results of this evaluation are contained in two technical reports, “Risk and Technology Review—Analysis of Socio-Economic Factors for Populations Living Near Petroleum Refineries”, available in the docket for this action (See Docket ID Nos. EPA-HQ-OAR-2010-0682-0226 and -0227). Additionally, a discussion of the final risk analysis is included in Sections IV.A and V.D of this preamble.

The EPA has determined that this final rule will not have disproportionately high and adverse human health or environmental effects on minority, low-income or indigenous populations because it maintains or increases the level of environmental protection for all affected populations without having any disproportionately high and adverse human health or environmental effects on any population, including any minority, low-income or indigenous populations. Further, the EPA believes that implementation of this rule will provide an ample margin of safety to protect public health of all demographic groups.

K. Congressional Review Act (CRA)

This action is subject to the CRA, and the EPA will submit a rule report to each House of the Congress and to the Comptroller General of the United States. This action is a “major rule” as defined by 5 U.S.C. 804(2).

Start List of Subjects

List of Subjects

40 CFR Part 60

  • Environmental protection
  • Administrative practice and procedures
  • Air pollution control
  • Hazardous substances
  • Incorporation by reference
  • Intergovernmental relations
  • Reporting and recordkeeping requirements

40 CFR Part 63

  • Environmental protection
  • Administrative practice and procedures
  • Air pollution control
  • Hazardous substances
  • Incorporation by reference
  • Intergovernmental relations
  • Reporting and recordkeeping requirements
End List of Subjects Start Signature

Dated: September 29, 2015.

Gina McCarthy,

Administrator.

End Signature

For the reasons stated in the preamble, title 40, chapter I, of the Code of Federal Regulations is amended as follows:

Start Part

PART 60—STANDARDS OF PERFORMANCE FOR NEW STATIONARY SOURCES

End Part Start Amendment Part

1. The authority citation for part 60 continues to read as follows:

End Amendment Part Start Authority

Authority: 42 U.S.C. 7401 et seq.

End Authority

Subpart J—Standards of Performance for Petroleum Refineries

Start Amendment Part

2. Section 60.105 is amended by revising paragraphs (b)(1)(iv) and (b)(3)(iii) to read as follows:

End Amendment Part
Monitoring of emissions and operations.
* * * * *

(b) * * *

(1) * * *

(iv) The supporting test results from sampling the requested fuel gas stream/system demonstrating that the sulfur content is less than 5 ppmv. Sampling data must include, at minimum, 2 weeks of daily monitoring (14 grab samples) for frequently operated fuel gas streams/systems; for infrequently operated fuel gas streams/systems, seven grab samples must be collected unless other additional information would support reduced sampling. The owner or operator shall use detector tubes (“length-of-stain tube” type measurement) following the “Gas Processors Association Standard 2377-86 (incorporated by reference—see § 60.17), using tubes with a maximum span between 10 and 40 ppmv inclusive when 1≤N≤10, where N = number of pump strokes, to test the applicant fuel gas stream for H2 S; and

* * * * *

(3) * * *

(iii) If the operation change results in a sulfur content that is outside the range of concentrations included in the original application and the owner or operator chooses not to submit new information to support an exemption, the owner or operator must begin H2 S monitoring using daily stain sampling to demonstrate compliance using length-of Start Printed Page 75230stain tubes with a maximum span between 200 and 400 ppmv inclusive when 1≤N≤5, where N = number of pump strokes. The owner or operator must begin monitoring according to the requirements in paragraph (a)(1) or (2) of this section as soon as practicable but in no case later than 180 days after the operation change. During daily stain tube sampling, a daily sample exceeding 162 ppmv is an exceedance of the 3-hour H2 S concentration limit.

* * * * *

Subpart Ja—Standards of Performance for Petroleum Refineries for Which Construction, Reconstruction, or Modification Commenced After May 14, 2007

Start Amendment Part

3. Section 60.100a is amended by revising the first sentence of paragraph (b) to read as follows:

End Amendment Part
Applicability, designation of affected facility, and reconstruction.
* * * * *

(b) Except for flares and delayed coking units, the provisions of this subpart apply only to affected facilities under paragraph (a) of this section which either commence construction, modification or reconstruction after May 14, 2007, or elect to comply with the provisions of this subpart in lieu of complying with the provisions in subpart J of this part. * * *

* * * * *
Start Amendment Part

4. Section 60.101a is amended by:

End Amendment Part Start Amendment Part

a. Revising the definition of “Corrective action”; and

End Amendment Part Start Amendment Part

b. Adding, in alphabetical order, a definition for “Sour water”.

End Amendment Part

The revision and addition read as follows:

Definitions.
* * * * *

Corrective action means the design, operation and maintenance changes that one takes consistent with good engineering practice to reduce or eliminate the likelihood of the recurrence of the primary cause and any other contributing cause(s) of an event identified by a root cause analysis as having resulted in a discharge of gases from an affected facility in excess of specified thresholds.

* * * * *

Sour water means water that contains sulfur compounds (usually H2 S) at concentrations of 10 parts per million by weight or more.

* * * * *
Start Amendment Part

5. Section 60.102a is amended by revising paragraphs (b)(1)(i) and (iii), (f), and (g)(1) introductory text to read as follows:

End Amendment Part
Emissions limitations.
* * * * *

(b) * * *

(1) * * *

(i) 1.0 gram per kilogram (g/kg) (1 pound (lb) per 1,000 lb) coke burn-off or, if a PM continuous emission monitoring system (CEMS) is used, 0.040 grain per dry standard cubic feet (gr/dscf) corrected to 0 percent excess air for each modified or reconstructed FCCU.

* * * * *

(iii) 1.0 g/kg (1 lb/1,000 lb) coke burn-off or, if a PM CEMS is used, 0.040 grain per dry standard cubic feet (gr/dscf) corrected to 0 percent excess air for each affected FCU.

* * * * *

(f) Except as provided in paragraph (f)(3) of this section, each owner or operator of an affected sulfur recovery plant shall comply with the applicable emission limits in paragraph (f)(1) or (2) of this section.

(1) For a sulfur recovery plant with a design production capacity greater than 20 long tons per day (LTD), the owner or operator shall comply with the applicable emission limit in paragraph (f)(1)(i) or (ii) of this section. If the sulfur recovery plant consists of multiple process trains or release points, the owner or operator shall comply with the applicable emission limit for each process train or release point individually or comply with the applicable emission limit in paragraph (f)(1)(i) or (ii) as a flow rate weighted average for a group of release points from the sulfur recovery plant provided that flow is monitored as specified in § 60.106a(a)(7); if flow is not monitored as specified in § 60.106a(a)(7), the owner or operator shall comply with the applicable emission limit in paragraph (f)(1)(i) or (ii) for each process train or release point individually. For a sulfur recovery plant with a design production capacity greater than 20 long LTD and a reduction control system not followed by incineration, the owner or operator shall also comply with the H2 S emission limit in paragraph (f)(1)(iii) of this section for each individual release point.

(i) For a sulfur recovery plant with an oxidation control system or a reduction control system followed by incineration, the owner or operator shall not discharge or cause the discharge of any gases into the atmosphere (SO2) in excess of the emission limit calculated using Equation 1 of this section. For Claus units that use only ambient air in the Claus burner or that elect not to monitor O2 concentration of the air/oxygen mixture used in the Claus burner or for non-Claus sulfur recovery plants, this SO2 emissions limit is 250 ppmv (dry basis) at zero percent excess air.

Where:

ELS = Emission limit for large sulfur recovery plant, ppmv (as SO2, dry basis at zero percent excess air);

k1 = Constant factor for emission limit conversion: k1 = 1 for converting to the SO2 limit for a sulfur recovery plant with an oxidation control system or a reduction control system followed by incineration and k1 = 1.2 for converting to the reduced sulfur compounds limit for a sulfur recovery plant with a reduction control system not followed by incineration; and

%O2 = O2 concentration of the air/oxygen mixture supplied to the Claus burner, percent by volume (dry basis). If only ambient air is used for the Claus burner or if the owner or operator elects not to monitor O2 concentration of the air/oxygen mixture used in the Claus burner or for non-Claus sulfur recovery plants, use 20.9% for %O2.

(ii) For a sulfur recovery plant with a reduction control system not followed by incineration, the owner or operator shall not discharge or cause the discharge of any gases into the atmosphere containing reduced sulfur compounds in excess of the emission limit calculated using Equation 1 of this section. For Claus units that use only ambient air in the Claus burner or for non-Claus sulfur recovery plants, this reduced sulfur compounds emission limit is 300 ppmv calculated as ppmv SO2 (dry basis) at 0-percent excess air.

(iii) For a sulfur recovery plant with a reduction control system not followed by incineration, the owner or operator shall not discharge or cause the discharge of any gases into the atmosphere containing hydrogen sulfide (H2 S) in excess of 10 ppmv calculated as ppmv SO2 (dry basis) at zero percent excess air.Start Printed Page 75231

(2) For a sulfur recovery plant with a design production capacity of 20 LTD or less, the owner or operator shall comply with the applicable emission limit in paragraph (f)(2)(i) or (ii) of this section. If the sulfur recovery plant consists of multiple process trains or release points, the owner or operator may comply with the applicable emission limit for each process train or release point individually or comply with the applicable emission limit in paragraph (f)(2)(i) or (ii) as a flow rate weighted average for a group of release points from the sulfur recovery plant provided that flow is monitored as specified in § 60.106a(a)(7); if flow is not monitored as specified in § 60.106a(a)(7), the owner or operator shall comply with the applicable emission limit in paragraph (f)(2)(i) or (ii) for each process train or release point individually. For a sulfur recovery plant with a design production capacity of 20 LTD or less and a reduction control system not followed by incineration, the owner or operator shall also comply with the H2 S emission limit in paragraph (f)(2)(iii) of this section for each individual release point.

(i) For a sulfur recovery plant with an oxidation control system or a reduction control system followed by incineration, the owner or operator shall not discharge or cause the discharge of any gases into the atmosphere containing SO2 in excess of the emission limit calculated using Equation 2 of this section. For Claus units that use only ambient air in the Claus burner or that elect not to monitor O2 concentration of the air/oxygen mixture used in the Claus burner or for non-Claus sulfur recovery plants, this SO2 emission limit is 2,500 ppmv (dry basis) at zero percent excess air.

Where:

ESS = Emission limit for small sulfur recovery plant, ppmv (as SO2, dry basis at zero percent excess air);

k1 = Constant factor for emission limit conversion: k1 = 1 for converting to the SO2 limit for a sulfur recovery plant with an oxidation control system or a reduction control system followed by incineration and k1 = 1.2 for converting to the reduced sulfur compounds limit for a sulfur recovery plant with a reduction control system not followed by incineration; and

%O2 = O2 concentration of the air/oxygen mixture supplied to the Claus burner, percent by volume (dry basis). If only ambient air is used in the Claus burner or if the owner or operator elects not to monitor O2 concentration of the air/oxygen mixture used in the Claus burner or for non-Claus sulfur recovery plants, use 20.9% for %O2.

(ii) For a sulfur recovery plant with a reduction control system not followed by incineration, the owner or operator shall not discharge or cause the discharge of any gases into the atmosphere containing reduced sulfur compounds in excess of the emission limit calculated using Equation 2 of this section. For Claus units that use only ambient air in the Claus burner or for non-Claus sulfur recovery plants, this reduced sulfur compounds emission limit is 3,000 ppmv calculated as ppmv SO2 (dry basis) at zero percent excess air.

(iii) For a sulfur recovery plant with a reduction control system not followed by incineration, the owner or operator shall not discharge or cause the discharge of any gases into the atmosphere containing H2 S in excess of 100 ppmv calculated as ppmv SO2 (dry basis) at zero percent excess air.

(3) The emission limits in paragraphs (f)(1) and (2) of this section shall not apply during periods of maintenance of the sulfur pit, which shall not exceed 240 hours per year. The owner or operator must document the time periods during which the sulfur pit vents were not controlled and measures taken to minimize emissions during these periods. Examples of these measures include not adding fresh sulfur or shutting off vent fans.

(g) * * *

(1) Except as provided in (g)(1)(iii) of this section, for each fuel gas combustion device, the owner or operator shall comply with either the emission limit in paragraph (g)(1)(i) of this section or the fuel gas concentration limit in paragraph (g)(1)(ii) of this section. For CO boilers or furnaces that are part of a fluid catalytic cracking unit or fluid coking unit affected facility, the owner or operator shall comply with the fuel gas concentration limit in paragraph (g)(1)(ii) for all fuel gas streams combusted in these units.

* * * * *
Start Amendment Part

6. Section 60.104a is amended by:

End Amendment Part Start Amendment Part

a. Revising the first sentence of paragraph (a) and paragraphs (b), (f) introductory text, and (h) introductory text;

End Amendment Part Start Amendment Part

b. Adding paragraph (h)(6); and

End Amendment Part Start Amendment Part

c. Removing and reserving paragraphs (j)(1) through (3).

End Amendment Part

The revisions and additions read as follows:

Performance tests.

(a) The owner or operator shall conduct a performance test for each FCCU, FCU, sulfur recovery plant and fuel gas combustion device to demonstrate initial compliance with each applicable emissions limit in § 60.102a and conduct a performance test for each flare to demonstrate initial compliance with the H2 S concentration requirement in § 60.103a(h) according to the requirements of § 60.8. * * *

(b) The owner or operator of a FCCU or FCU that elects to monitor control device operating parameters according to the requirements in § 60.105a(b), to use bag leak detectors according to the requirements in § 60.105a(c), or to use COMS according to the requirements in § 60.105a(e) shall conduct a PM performance test at least annually (i.e., once per calendar year, with an interval of at least 8 months but no more than 16 months between annual tests) and furnish the Administrator a written report of the results of each test.

* * * * *

(f) The owner or operator of an FCCU or FCU that uses cyclones to comply with the PM per coke burn-off emissions limit in § 60.102a(b)(1) shall establish a site-specific opacity operating limit according to the procedures in paragraphs (f)(1) through (3) of this section.

* * * * *

(h) The owner or operator shall determine compliance with the SO2 emissions limits for sulfur recovery plants in § 60.102a(f)(1)(i) and (f)(2)(i) and the reduced sulfur compounds and H2 S emissions limits for sulfur recovery plants in § 60.102a(f)(1)(ii), (f)(1)(iii), (f)(2)(ii), and (f)(2)(iii) using the following methods and procedures:

* * * * *

(6) If oxygen or oxygen-enriched air is used in the Claus burner and either Equation 1 or 2 of this subpart is used to determine the applicable emissions limit, determine the average O2 concentration of the air/oxygen mixture supplied to the Claus burner, in percent by volume (dry basis), for the performance test using all hourly average O2 concentrations determined Start Printed Page 75232during the test runs using the procedures in § 60.106a(a)(5) or (6).

* * * * *
Start Amendment Part

7. Section 60.105a is amended by:

End Amendment Part Start Amendment Part

a. Revising paragraphs (b)(1)(i), (b)(1)(ii)(A), (b)(2), (h)(1), (h)(3)(i), and (i)(1);

End Amendment Part Start Amendment Part

b. Redesignating paragraphs (i)(2) through (6) as (i)(3) through (7);

End Amendment Part Start Amendment Part

c. Adding paragraph (i)(2); and

End Amendment Part Start Amendment Part

d. Revising newly redesignated paragraph (i)(7).

End Amendment Part

The revisions and additions read as follows:

Monitoring of emissions and operations for fluid catalytic cracking units (FCCU) and fluid coking units (FCU).
* * * * *

(b) * * *

(1) * * *

(i) For units controlled using an electrostatic precipitator, the owner or operator shall use CPMS to measure and record the hourly average total power input and secondary current to the entire system.

(ii) * * *

(A) As an alternative to pressure drop, the owner or operator of a jet ejector type wet scrubber or other type of wet scrubber equipped with atomizing spray nozzles must conduct a daily check of the air or water pressure to the spray nozzles and record the results of each check. Faulty (e.g., leaking or plugged) air or water lines must be repaired within 12 hours of identification of an abnormal pressure reading.

* * * * *

(2) For use in determining the coke burn-off rate for an FCCU or FCU, the owner or operator shall install, operate, calibrate, and maintain an instrument for continuously monitoring the concentrations of CO2, O2 (dry basis), and if needed, CO in the exhaust gases prior to any control or energy recovery system that burns auxiliary fuels. A CO monitor is not required for determining coke burn-off rate when no auxiliary fuel is burned and a continuous CO monitor is not required in accordance with paragraph (h)(3) of this section.

(i) The owner or operator shall install, operate, and maintain each CO2 and O2 monitor according to Performance Specification 3 of appendix B to this part.

(ii) The owner or operator shall conduct performance evaluations of each CO2 and O2 monitor according to the requirements in § 60.13(c) and Performance Specification 3 of appendix B to this part. The owner or operator shall use Method 3 of appendix A-3 to this part for conducting the relative accuracy evaluations.

(iii) If a CO monitor is required, the owner or operator shall install, operate, and maintain each CO monitor according to Performance Specification 4 or 4A of appendix B to this part. If this CO monitor also serves to demonstrate compliance with the CO emissions limit in § 60.102a(b)(4), the span value for this instrument is 1,000 ppm; otherwise, the span value for this instrument should be set at approximately 2 times the typical CO concentration expected in the FCCU of FCU flue gas prior to any emission control or energy recovery system that burns auxiliary fuels.

(iv) If a CO monitor is required, the owner or operator shall conduct performance evaluations of each CO monitor according to the requirements in § 60.13(c) and Performance Specification 4 of appendix B to this part. The owner or operator shall use Method 10, 10A, or 10B of appendix A-3 to this part for conducting the relative accuracy evaluations.

(v) The owner or operator shall comply with the quality assurance requirements of procedure 1 of appendix F to this part, including quarterly accuracy determinations for CO2 and CO monitors, annual accuracy determinations for O2 monitors, and daily calibration drift tests.

* * * * *

(h) * * *

(1) The owner or operator shall install, operate, and maintain each CO monitor according to Performance Specification 4 or 4A of appendix B to this part. The span value for this instrument is 1,000 ppmv CO.

* * * * *

(3) * * *

(i) The demonstration shall consist of continuously monitoring CO emissions for 30 days using an instrument that meets the requirements of Performance Specification 4 or 4A of appendix B to this part. The span value shall be 100 ppmv CO instead of 1,000 ppmv, and the relative accuracy limit shall be 10 percent of the average CO emissions or 5 ppmv CO, whichever is greater. For instruments that are identical to Method 10 of appendix A-4 to this part and employ the sample conditioning system of Method 10A of appendix A-4 to this part, the alternative relative accuracy test procedure in section 10.1 of Performance Specification 2 of appendix B to this part may be used in place of the relative accuracy test.

* * * * *

(i) * * *

(1) If a CPMS is used according to paragraph (b)(1) of this section, all 3-hour periods during which the average PM control device operating characteristics, as measured by the continuous monitoring systems under paragraph (b)(1), fall below the levels established during the performance test. If the alternative to pressure drop CPMS is used for the owner or operator of a jet ejector type wet scrubber or other type of wet scrubber equipped with atomizing spray nozzles, each day in which abnormal pressure readings are not corrected within 12 hours of identification.

(2) If a bag leak detection system is used according to paragraph (c) of this section, each day in which the cause of an alarm is not alleviated within the time period specified in paragraph (c)(3) of this section.

* * * * *

(7) All 1-hour periods during which the average CO concentration as measured by the CO continuous monitoring system under paragraph (h) of this section exceeds 500 ppmv or, if applicable, all 1-hour periods during which the average temperature and O2 concentration as measured by the continuous monitoring systems under paragraph (h)(4) of this section fall below the operating limits established during the performance test.

Start Amendment Part

8. Section 60.106a is amended by:

End Amendment Part Start Amendment Part

a. Revising paragraph (a)(1)(i);

End Amendment Part Start Amendment Part

b. Adding paragraphs (a)(1)(iv) through (vii);

End Amendment Part Start Amendment Part

c. Revising paragraphs (a)(2) introductory text, (a)(2)(i) and (ii), and the first sentence of paragraph (a)(2)(iii);

End Amendment Part Start Amendment Part

d. Removing paragraphs (a)(2)(iv) and (v);

End Amendment Part Start Amendment Part

e. Redesignating paragraphs (a)(2)(vi) through (ix) as (a)(2)(iv) through (vii);

End Amendment Part Start Amendment Part

f. Revising the first sentence of paragraph (a)(3) introductory text and paragraph (a)(3)(i);

End Amendment Part Start Amendment Part

g. Adding paragraphs (a)(4) through (7); and

End Amendment Part Start Amendment Part

h. Revising paragraphs (b)(2) and (3).

End Amendment Part

The revisions and additions read as follows:

Monitoring of emissions and operations for sulfur recovery plants.

(a) * * *

(1) * * *

(i) The span value for the SO2 monitor is two times the applicable SO2 emission limit at the highest O2 concentration in the air/oxygen stream used in the Claus burner, if applicable.

* * * * *

(iv) The owner or operator shall install, operate, and maintain each O2 monitor according to Performance Specification 3 of appendix B to this part.

(v) The span value for the O2 monitor must be selected between 10 and 25 percent, inclusive.Start Printed Page 75233

(vi) The owner or operator shall conduct performance evaluations for the O2 monitor according to the requirements of § 60.13(c) and Performance Specification 3 of appendix B to this part. The owner or operator shall use Methods 3, 3A, or 3B of appendix A-2 to this part for conducting the relative accuracy evaluations. The method ANSI/ASME PTC 19.10-1981 (incorporated by reference—see § 60.17) is an acceptable alternative to EPA Method 3B of appendix A-2 to this part.

(vii) The owner or operator shall comply with the applicable quality assurance procedures of appendix F to this part for each monitor, including annual accuracy determinations for each O2 monitor, and daily calibration drift determinations.

(2) For sulfur recovery plants that are subject to the reduced sulfur compounds emission limit in § 60.102a(f)(1)(ii) or (f)(2)(ii), the owner or operator shall install, operate, calibrate, and maintain an instrument for continuously monitoring and recording the concentration of reduced sulfur compounds and O2 emissions into the atmosphere. The reduced sulfur compounds emissions shall be calculated as SO2 (dry basis, zero percent excess air).

(i) The span value for the reduced sulfur compounds monitor is two times the applicable reduced sulfur compounds emission limit as SO2 at the highest O2 concentration in the air/oxygen stream used in the Claus burner, if applicable.

(ii) The owner or operator shall install, operate, and maintain each reduced sulfur compounds CEMS according to Performance Specification 5 of appendix B to this part.

(iii) The owner or operator shall conduct performance evaluations of each reduced sulfur compounds monitor according to the requirements in § 60.13(c) and Performance Specification 5 of appendix B to this part. * * *

* * * * *

(3) In place of the reduced sulfur compounds monitor required in paragraph (a)(2) of this section, the owner or operator may install, calibrate, operate, and maintain an instrument using an air or O2 dilution and oxidation system to convert any reduced sulfur to SO2 for continuously monitoring and recording the concentration (dry basis, 0 percent excess air) of the total resultant SO2. * * *

(i) The span value for this monitor is two times the applicable reduced sulfur compounds emission limit as SO2 at the highest O2 concentration in the air/oxygen stream used in the Claus burner, if applicable.

* * * * *

(4) For sulfur recovery plants that are subject to the H2 S emission limit in § 60.102a(f)(1)(iii) or (f)(2)(iii), the owner or operator shall install, operate, calibrate, and maintain an instrument for continuously monitoring and recording the concentration of H2 S, and O2 emissions into the atmosphere. The H2 S emissions shall be calculated as SO2 (dry basis, zero percent excess air).

(i) The span value for this monitor is two times the applicable H2 S emission limit.

(ii) The owner or operator shall install, operate, and maintain each H2 S CEMS according to Performance Specification 7 of appendix B to this part.

(iii) The owner or operator shall conduct performance evaluations for each H2 S monitor according to the requirements of § 60.13(c) and Performance Specification 7 of appendix B to this part. The owner or operator shall use Methods 11 or 15 of appendix A-5 to this part or Method 16 of appendix A-6 to this part for conducting the relative accuracy evaluations. The method ANSI/ASME PTC 19.10-1981 (incorporated by reference—see § 60.17) is an acceptable alternative to EPA Method 15A of appendix A-5 to this part.

(iv) The owner or operator shall install, operate, and maintain each O2 monitor according to Performance Specification 3 of appendix B to this part.

(v) The span value for the O2 monitor must be selected between 10 and 25 percent, inclusive.

(vi) The owner or operator shall conduct performance evaluations for the O2 monitor according to the requirements of § 60.13(c) and Performance Specification 3 of appendix B to this part. The owner or operator shall use Methods 3, 3A, or 3B of appendix A-2 to this part for conducting the relative accuracy evaluations. The method ANSI/ASME PTC 19.10-1981 (incorporated by reference—see § 60.17) is an acceptable alternative to EPA Method 3B of appendix A-2 to this part.

(vii) The owner or operator shall comply with the applicable quality assurance procedures of appendix F to this part for each monitor, including annual accuracy determinations for each O2 monitor, and daily calibration drift determinations.

(5) For sulfur recovery plants that use oxygen or oxygen enriched air in the Claus burner and that elects to monitor O2 concentration of the air/oxygen mixture supplied to the Claus burner, the owner or operator shall install, operate, calibrate, and maintain an instrument for continuously monitoring and recording the O2 concentration of the air/oxygen mixture supplied to the Claus burner in order to determine the allowable emissions limit.

(i) The owner or operator shall install, operate, and maintain each O2 monitor according to Performance Specification 3 of appendix B to this part.

(ii) The span value for the O2 monitor shall be 100 percent.

(iii) The owner or operator shall conduct performance evaluations for the O2 monitor according to the requirements of § 60.13(c) and Performance Specification 3 of appendix B to this part. The owner or operator shall use Methods 3, 3A, or 3B of appendix A-2 to this part for conducting the relative accuracy evaluations. The method ANSI/ASME PTC 19.10-1981 (incorporated by reference—see § 60.17) is an acceptable alternative to EPA Method 3B of appendix A-2 to this part.

(iv) The owner or operator shall comply with the applicable quality assurance procedures of appendix F to this part for each monitor, including annual accuracy determinations for each O2 monitor, and daily calibration drift determinations.

(v) The owner or operator shall use the hourly average O2 concentration from this monitor for use in Equation 1 or 2 of § 60.102a(f), as applicable, for each hour and determine the allowable emission limit as the arithmetic average of 12 contiguous 1-hour averages (i.e., the rolling 12-hour average).

(6) As an alternative to the O2 monitor required in paragraph (a)(5) of this section, the owner or operator may install, calibrate, operate, and maintain a CPMS to measure and record the volumetric gas flow rate of ambient air and oxygen-enriched gas supplied to the Claus burner and calculate the hourly average O2 concentration of the air/oxygen mixture used in the Claus burner as specified in paragraphs (a)(6)(i) through (iv) of this section in order to determine the allowable emissions limit as specified in paragraphs (a)(6)(v) of this section.

(i) The owner or operator shall install, calibrate, operate and maintain each flow monitor according to the manufacturer's procedures and specifications and the following requirements.

(A) Locate the monitor in a position that provides a representative measurement of the total gas flow rate.Start Printed Page 75234

(B) Use a flow sensor meeting an accuracy requirement of ±5 percent over the normal range of flow measured or 10 cubic feet per minute, whichever is greater.

(C) Use a flow monitor that is maintainable online, is able to continuously correct for temperature, pressure and, for ambient air flow monitor, moisture content, and is able to record dry flow in standard conditions (as defined in § 60.2) over one-minute averages.

(D) At least quarterly, perform a visual inspection of all components of the monitor for physical and operational integrity and all electrical connections for oxidation and galvanic corrosion if the flow monitor is not equipped with a redundant flow sensor.

(E) Recalibrate the flow monitor in accordance with the manufacturer's procedures and specifications biennially (every two years) or at the frequency specified by the manufacturer.

(ii) The owner or operator shall use 20.9 percent as the oxygen content of the ambient air.

(iii) The owner or operator shall use product specifications (e.g., as reported in material safety data sheets) for percent oxygen for purchased oxygen. For oxygen produced onsite, the percent oxygen shall be determined by periodic measurements or process knowledge.

(iv) The owner or operator shall calculate the hourly average O2 concentration of the air/oxygen mixture used in the Claus burner using Equation 10 of this section:

Where:

%O2 = O2 concentration of the air/oxygen mixture used in the Claus burner, percent by volume (dry basis);

20.9 = O2 concentration in air, percent dry basis;

Qair = Volumetric flow rate of ambient air used in the Claus burner, dscfm;

%O2,oxy = O2 concentration in the enriched oxygen stream, percent dry basis; and

Qoxy = Volumetric flow rate of enriched oxygen stream used in the Claus burner, dscfm.

(v) The owner or operator shall use the hourly average O2 concentration determined using Equation 8 of § 60.104a(d)(8) for use in Equation 1 or 2 of § 60.102a(f), as applicable, for each hour and determine the allowable emission limit as the arithmetic average of 12 contiguous 1-hour averages (i.e., the rolling 12-hour average).

(7) Owners or operators of a sulfur recovery plant that elects to comply with the SO2 emission limit in § 60.102a(f)(1)(i) or (f)(2)(i) or the reduced sulfur compounds emission limit in § 60.102a(f)(1)(ii) or (f)(2)(ii) as a flow rate weighted average for a group of release points from the sulfur recovery plant rather than for each process train or release point individually shall install, calibrate, operate, and maintain a CPMS to measure and record the volumetric gas flow rate of each release point within the group of release points from the sulfur recovery plant as specified in paragraphs (a)(7)(i) through (iv) of this section.

(i) The owner or operator shall install, calibrate, operate and maintain each flow monitor according to the manufacturer's procedures and specifications and the following requirements.

(A) Locate the monitor in a position that provides a representative measurement of the total gas flow rate.

(B) Use a flow sensor meeting an accuracy requirement of ±5 percent over the normal range of flow measured or 10 cubic feet per minute, whichever is greater.

(C) Use a flow monitor that is maintainable online, is able to continuously correct for temperature, pressure, and moisture content, and is able to record dry flow in standard conditions (as defined in § 60.2) over one-minute averages.

(D) At least quarterly, perform a visual inspection of all components of the monitor for physical and operational integrity and all electrical connections for oxidation and galvanic corrosion if the flow monitor is not equipped with a redundant flow sensor.

(E) Recalibrate the flow monitor in accordance with the manufacturer's procedures and specifications biennially (every two years) or at the frequency specified by the manufacturer.

(ii) The owner or operator shall correct the flow to 0 percent excess air using Equation 11 of this section:

Where:

Qadj = Volumetric flow rate adjusted to 0 percent excess air, dry standard cubic feet per minute (dscfm);

Cmeas = Volumetric flow rate measured by the flow meter corrected to dry standard conditions, dscfm;

20.9c = 20.9 percent O2−0.0 percent O2 (defined O2 correction basis), percent;

20.9 = O2 concentration in air, percent; and

%O2 = O2 concentration measured on a dry basis, percent.

(iii) The owner or operator shall calculate the flow weighted average SO2 or reduced sulfur compounds concentration for each hour using Equation 12 of this section:

Start Printed Page 75235

Where:

Cave = Flow weighted average concentration of the pollutant, ppmv (dry basis, zero percent excess air). The pollutant is either SO2 (if complying with the SO2 emission limit in § 60.102a(f)(1)(i) or (f)(2)(i)) or reduced sulfur compounds (if complying with the reduced sulfur compounds emission limit in § 60.102a(f)(1)(ii) or (f)(2)(ii));

N = Number of release points within the group of release points from the sulfur recovery plant for which emissions averaging is elected;

Cn = Pollutant concentration in the nth release point within the group of release points from the sulfur recovery plant for which emissions averaging is elected, ppmv (dry basis, zero percent excess air);

Qadj,n = Volumetric flow rate of the nth release point within the group of release points from the sulfur recovery plant for which emissions averaging is elected, dry standard cubic feet per minute (dscfm, adjusted to 0 percent excess air).

(iv) For sulfur recovery plants that use oxygen or oxygen enriched air in the Claus burner, the owner or operator shall use Equation 10 of this section and the hourly emission limits determined in paragraph (a)(5)(v) or (a)(6)(v) of this section in-place of the pollutant concentration to determine the flow weighted average hourly emission limit for each hour. The allowable emission limit shall be calculated as the arithmetic average of 12 contiguous 1-hour averages (i.e., the rolling 12-hour average).

(b) * * *

(2) All 12-hour periods during which the average concentration of reduced sulfur compounds (as SO2) as measured by the reduced sulfur compounds continuous monitoring system required under paragraph (a)(2) or (3) of this section exceeds the applicable emission limit; or

(3) All 12-hour periods during which the average concentration of H2 S as measured by the H2 S continuous monitoring system required under paragraph (a)(4) of this section exceeds the applicable emission limit (dry basis, 0 percent excess air).

Start Amendment Part

9. Section 60.107a is amended by revising paragraphs (a)(1)(i) and (ii), (b)(1)(iv), the first sentence of paragraph (b)(3)(iii), (d)(3), (e)(1) introductory text, (e)(1)(ii), (e)(2) introductory text, (e)(2)(ii), (e)(2)(vi)(C), (e)(3), (f)(1)(ii), and (h)(5) to read as follows:

End Amendment Part
Monitoring of emissions and operations for fuel gas combustion devices and flares.

(a) * * *

(1) * * *

(i) The owner or operator shall install, operate, and maintain each SO2 monitor according to Performance Specification 2 of appendix B to this part. The span value for the SO2 monitor is 50 ppmv SO2.

(ii) The owner or operator shall conduct performance evaluations for the SO2 monitor according to the requirements of § 60.13(c) and Performance Specification 2 of appendix B to this part. The owner or operator shall use Methods 6, 6A, or 6C of appendix A-4 to this part for conducting the relative accuracy evaluations. The method ANSI/ASME PTC 19.10-1981 (incorporated by reference—see § 60.17) is an acceptable alternative to EPA Method 6 or 6A of appendix A-4 to this part. Samples taken by Method 6 of appendix A-4 to this part shall be taken at a flow rate of approximately 2 liters/min for at least 30 minutes. The relative accuracy limit shall be 20 percent or 4 ppmv, whichever is greater, and the calibration drift limit shall be 5 percent of the established span value.

* * * * *

(b) * * *

(1) * * *

(iv) The supporting test results from sampling the requested fuel gas stream/system demonstrating that the sulfur content is less than 5 ppmv H2 S. Sampling data must include, at minimum, 2 weeks of daily monitoring (14 grab samples) for frequently operated fuel gas streams/systems; for infrequently operated fuel gas streams/systems, seven grab samples must be collected unless other additional information would support reduced sampling. The owner or operator shall use detector tubes (“length-of-stain tube” type measurement) following the “Gas Processors Association Standard 2377-86 (incorporated by reference—see § 60.17), using tubes with a maximum span between 10 and 40 ppmv inclusive when 1≤N≤10, where N = number of pump strokes, to test the applicant fuel gas stream for H2 S; and

* * * * *

(3) * * *

(iii) If the operation change results in a sulfur content that is outside the range of concentrations included in the original application and the owner or operator chooses not to submit new information to support an exemption, the owner or operator must begin H2 S monitoring using daily stain sampling to demonstrate compliance using length-of-stain tubes with a maximum span between 200 and 400 ppmv inclusive when 1≤N≤5, where N = number of pump strokes. * * *

* * * * *

(d) * * *

(3) As an alternative to the requirements in paragraph (d)(2) of this section, the owner or operator of a gas-fired process heater shall install, operate and maintain a gas composition analyzer and determine the average F factor of the fuel gas using the factors in Table 1 of this subpart and Equation 13 of this section. If a single fuel gas system provides fuel gas to several process heaters, the F factor may be determined at a single location in the fuel gas system provided it is representative of the fuel gas fed to the affected process heater(s).

Where:

Fd = F factor on dry basis at 0% excess air, dscf/MMBtu.

Xi = mole or volume fraction of each component in the fuel gas.

MEVi = molar exhaust volume, dry standard cubic feet per mole (dscf/mol).

MHCi = molar heat content, Btu per mole (Btu/mol).

1,000,000 = unit conversion, Btu per MMBtu.

* * * * *

(e) * * *

(1) Total reduced sulfur monitoring requirements. The owner or operator shall install, operate, calibrate and maintain an instrument or instruments for continuously monitoring and recording the concentration of total reduced sulfur in gas discharged to the flare.

* * * * *

(ii) The owner or operator shall conduct performance evaluations of each total reduced sulfur monitor according to the requirements in § 60.13(c) and Performance Specification 5 of appendix B to this part. The owner or operator of each total Start Printed Page 75236reduced sulfur monitor shall use EPA Method 15A of appendix A-5 to this part for conducting the relative accuracy evaluations. The method ANSI/ASME PTC 19.10-1981 (incorporated by reference-see § 60.17) is an acceptable alternative to EPA Method 15A of appendix A-5 to this part. The alternative relative accuracy procedures described in section 16.0 of Performance Specification 2 of appendix B to this part (cylinder gas audits) may be used for conducting the relative accuracy evaluations, except that it is not necessary to include as much of the sampling probe or sampling line as practical.

* * * * *

(2) H2S monitoring requirements. The owner or operator shall install, operate, calibrate, and maintain an instrument or instruments for continuously monitoring and recording the concentration of H2 S in gas discharged to the flare according to the requirements in paragraphs (e)(2)(i) through (iii) of this section and shall collect and analyze samples of the gas and calculate total sulfur concentrations as specified in paragraphs (e)(2)(iv) through (ix) of this section.

* * * * *

(ii) The owner or operator shall conduct performance evaluations of each H2 S monitor according to the requirements in § 60.13(c) and Performance Specification 7 of appendix B to this part. The owner or operator shall use EPA Method 11, 15 or 15A of appendix A-5 to this part for conducting the relative accuracy evaluations. The method ANSI/ASME PTC 19.10-1981 (incorporated by reference—see § 60.17) is an acceptable alternative to EPA Method 15A of appendix A-5 to this part. The alternative relative accuracy procedures described in section 16.0 of Performance Specification 2 of appendix B to this part (cylinder gas audits) may be used for conducting the relative accuracy evaluations, except that it is not necessary to include as much of the sampling probe or sampling line as practical.

* * * * *

(vi) * * *

(C) Determine the acceptable range for subsequent weekly samples based on the 95-percent confidence interval for the distribution of daily ratios based on the 10 individual daily ratios using Equation 14 of this section.

Where:

AR = Acceptable range of subsequent ratio determinations, unitless.

RatioAvg = 10-day average total sulfur-to-H2 S concentration ratio, unitless.

2.262 = t-distribution statistic for 95-percent 2-sided confidence interval for 10 samples (9 degrees of freedom).

SDev = Standard deviation of the 10 daily average total sulfur-to-H2 S concentration ratios used to develop the 10-day average total sulfur-to-H2 S concentration ratio, unitless.

* * * * *

(3) SO2monitoring requirements. The owner or operator shall install, operate, calibrate, and maintain an instrument for continuously monitoring and recording the concentration of SO2 from a process heater or other fuel gas combustion device that is combusting gas representative of the fuel gas in the flare gas line according to the requirements in paragraph (a)(1) of this section, determine the F factor of the fuel gas at least daily according to the requirements in paragraphs (d)(2) through (4) of this section, determine the higher heating value of the fuel gas at least daily according to the requirements in paragraph (d)(7) of this section, and calculate the total sulfur content (as SO2) in the fuel gas using Equation 15 of this section.

Where:

TSFG = Total sulfur concentration, as SO2, in the fuel gas, ppmv.

CSO2 = Concentration of SO2 in the exhaust gas, ppmv (dry basis at 0-percent excess air).

Fd = F factor gas on dry basis at 0-percent excess air, dscf/MMBtu.

HHVFG = Higher heating value of the fuel gas, MMBtu/scf.

* * * * *

(f) * * *

(1) * * *

(ii) Use a flow sensor meeting an accuracy requirement of ±20 percent of the flow rate at velocities ranging from 0.1 to 1 feet per second and an accuracy of ±5 percent of the flow rate for velocities greater than 1 feet per second.

* * * * *

(h) * * *

(5) Daily O2limits for fuel gas combustion devices. Each day during which the concentration of O2 as measured by the O2 continuous monitoring system required under paragraph (c)(6) or (d)(8) of this section exceeds the O2 operating limit or operating curve determined during the most recent biennial performance test.

Start Part

PART 63—NATIONAL EMISSION STANDARDS FOR HAZARDOUS AIR POLLUTANTS FOR SOURCE CATEGORIES

End Part Start Amendment Part

10. The authority citation for part 63 continues to read as follows:

End Amendment Part Start Authority

Authority: 42 U.S.C. 7401 et se.

End Authority

Subpart A—General Provisions

Start Amendment Part

11. Section 63.14 is amended by:

End Amendment Part Start Amendment Part

a. Revising paragraph (h)(14);

End Amendment Part Start Amendment Part

b. Redesignating paragraphs (h)(82) through (99) as (h)(86) through (103), paragraphs (h)(77) through (81) as (h)(80) through (84), paragraphs (h)(73) through (76) as paragraphs (h)(75) through (78), and paragraphs (h)(15) through (72) as (16) through (73), respectively;

End Amendment Part Start Amendment Part

c. Revising newly redesignated paragraph (h)(78);

End Amendment Part Start Amendment Part

d. Adding paragraphs (h)(15), (74), (79), (85), (104) and (j)(2);

End Amendment Part Start Amendment Part

e. Redesignating paragraph (m)(3) through (21) as (m)(5) through (23), respectively, and paragraph (m)(2) as (m)(3).

End Amendment Part Start Amendment Part

f. Adding paragraphs (m)(2) and (4) and (n)(3); and

End Amendment Part Start Amendment Part

g. Revising paragraph (s)(1).

End Amendment Part

The revisions and additions read as follows:

Incorporation by reference.
* * * * *

(h) * * *

(14) ASTM D1945-03 (Reapproved 2010), Standard Test Method for Analysis of Natural Gas by Gas Chromatography, Approved January 1, 2010, IBR approved for §§ 63.670(j), 63.772(h), and 63.1282(g).

(15) ASTM D1945-14, Standard Test Method for Analysis of Natural Gas by Gas Chromatography, Approved Start Printed Page 75237November 1, 2014, IBR approved for § 63.670(j).

* * * * *

(74) ASTM D6196-03 (Reapproved 2009), Standard Practice for Selection of Sorbents, Sampling, and Thermal Desorption Analysis Procedures for Volatile Organic Compounds in Air, Approved March 1, 2009, IBR approved for appendix A to this part: Method 325A and Method 325B.

* * * * *

(78) ASTM D6348-03 (Reapproved 2010), Standard Test Method for Determination of Gaseous Compounds by Extractive Direct Interface Fourier Transform Infrared (FTIR) Spectroscopy, including Annexes A1 through A8, Approved October 1, 2010, IBR approved for § 63.1571(a), tables 4 and 5 to subpart JJJJJ, tables 4 and 6 to subpart KKKKK, tables 1, 2, and 5 to subpart UUUUU and appendix B to subpart UUUUU.

* * * * *

(79) ASTM D6348-12e1, Standard Test Method for Determination of Gaseous Compounds by Extractive Direct Interface Fourier Transform Infrared (FTIR) Spectroscopy, Approved February 1, 2012, IBR approved for § 63.1571(a).

* * * * *

(85) ASTM D6420-99 (Reapproved 2010), Standard Test Method for Determination of Gaseous Organic Compounds by Direct Interface Gas Chromatography-Mass Spectrometry, Approved October 1, 2010, IBR approved for § 63.670(j) and appendix A to this part: Method 325B.

* * * * *

(104) ASTM UOP539-12, Refinery Gas Analysis by GC, Copyright 2012 (to UOP), IBR approved for § 63.670(j).

* * * * *

(j) * * *

(2) BS EN 14662-4:2005, Ambient air quality standard method for the measurement of benzene concentrations—Part 4: Diffusive sampling followed by thermal desorption and gas chromatography, Published June 27, 2005, IBR approved for appendix A to this part: Method 325A and Method 325B.

* * * * *

(m) * * *

(2) EPA-454/B-08-002, Office of Air Quality Planning and Standards (OAQPS), Quality Assurance Handbook for Air Pollution Measurement Systems, Volume IV: Meteorological Measurements, Version 2.0 (Final), March 24, 2008, IBR approved for § 63.658(d) and appendix A to this part: Method 325A.

* * * * *

(4) EPA-454/R-99-005, Office of Air Quality Planning and Standards (OAQPS), Meteorological Monitoring Guidance for Regulatory Modeling Applications, February 2000, IBR approved for appendix A to this part: Method 325A.

* * * * *

(n) * * *

(3) ISO 16017-2:2003(E): Indoor, ambient and workplace air—sampling and analysis of volatile organic compounds by sorbent tube/thermal desorption/capillary gas chromatography—Part 2: Diffusive sampling, May 15, 2003, IBR approved for appendix A to this part: Method 325A and Method 325B.

* * * * *

(s) * * *

(1) “Air Stripping Method (Modified El Paso Method) for Determination of Volatile Organic Compound Emissions from Water Sources,” Revision Number One, dated January 2003, Sampling Procedures Manual, Appendix P: Cooling Tower Monitoring, January 31, 2003, IBR approved for §§ 63.654(c) and (g), 63.655(i), and 63.11920.

* * * * *

Subpart Y—National Emission Standards for Marine Tank Vessel Loading Operations

Start Amendment Part

12. Section 63.560 is amended by revising paragraph (a)(4) to read as follows:

End Amendment Part
Applicability and designation of affected source.

(a) * * *

(4) Existing sources with emissions less than 10 and 25 tons must meet the submerged fill standards of 46 CFR 153.282.

* * * * *

Subpart CC—National Emission Standards for Hazardous Air Pollutants From Petroleum Refineries

Start Amendment Part

13. Section 63.640 is amended by:

End Amendment Part Start Amendment Part

a. Revising paragraphs (a) introductory text and (c) introductory text;

End Amendment Part Start Amendment Part

c. Adding paragraph (c)(9);

End Amendment Part Start Amendment Part

d. Revising paragraphs (d)(5), (h), (k)(1), (l) introductory text, (l)(2) introductory text, (l)(2)(i), (l)(3) introductory text, (m) introductory text, (n) introductory text, (n)(1) through (5), (n)(8) introductory text, and (n)(8)(ii);

End Amendment Part Start Amendment Part

e. Adding paragraphs (n)(8)(vii) and (viii);

End Amendment Part Start Amendment Part

f. Revising paragraph (n)(9)(i);

End Amendment Part Start Amendment Part

g. Adding paragraph (n)(10);

End Amendment Part Start Amendment Part

h. Revising paragraph (o)(2)(i) introductory text;

End Amendment Part Start Amendment Part

i. Adding paragraph (o)(2)(i)(D);

End Amendment Part Start Amendment Part

j. Revising paragraph (o)(2)(ii) introductory text; and

End Amendment Part Start Amendment Part

k. Adding paragraphs (o)(2)(ii)(C) and (s).

End Amendment Part

The revisions and additions read as follows:

Applicability and designation of affected source.

(a) This subpart applies to petroleum refining process units and to related emissions points that are specified in paragraphs (c)(1) through (9) of this section that are located at a plant site and that meet the criteria in paragraphs (a)(1) and (2) of this section:

* * * * *

(c) For the purposes of this subpart, the affected source shall comprise all emissions points, in combination, listed in paragraphs (c)(1) through (9) of this section that are located at a single refinery plant site.

* * * * *

(9) All releases associated with the decoking operations of a delayed coking unit, as defined in this subpart.

(d) * * *

(5) Emission points routed to a fuel gas system, as defined in § 63.641, provided that on and after January 30, 2019, any flares receiving gas from that fuel gas system are subject to § 63.670. No other testing, monitoring, recordkeeping, or reporting is required for refinery fuel gas systems or emission points routed to refinery fuel gas systems.

* * * * *

(h) Sources subject to this subpart are required to achieve compliance on or before the dates specified in table 11 of this subpart, except as provided in paragraphs (h)(1) through (3) of this section.

(1) Marine tank vessels at existing sources shall be in compliance with this subpart, except for §§ 63.657 through 63.660, no later than August 18, 1999, unless the vessels are included in an emissions average to generate emission credits. Marine tank vessels used to generate credits in an emissions average shall be in compliance with this subpart no later than August 18, 1998, unless an extension has been granted by the Administrator as provided in § 63.6(i).

(2) Existing Group 1 floating roof storage vessels meeting the applicability criteria in item 1 of the definition of Group 1 storage vessel shall be in compliance with § 63.646 at the first degassing and cleaning activity after August 18, 1998, or August 18, 2005, whichever is first.Start Printed Page 75238

(3) An owner or operator may elect to comply with the provisions of § 63.648(c) through (i) as an alternative to the provisions of § 63.648(a) and (b). In such cases, the owner or operator shall comply no later than the dates specified in paragraphs (h)(3)(i) through (iii) of this section.

(i) Phase I (see table 2 of this subpart), beginning on August 18, 1998;

(ii) Phase II (see table 2 of this subpart), beginning no later than August 18, 1999; and

(iii) Phase III (see table 2 of this subpart), beginning no later than February 18, 2001.

* * * * *

(k) * * *

(1) The reconstructed source, addition, or change shall be in compliance with the new source requirements in item (1), (2), or (3) of table 11 of this subpart, as applicable, upon initial startup of the reconstructed source or by August 18, 1995, whichever is later; and

* * * * *

(l) If an additional petroleum refining process unit is added to a plant site or if a miscellaneous process vent, storage vessel, gasoline loading rack, marine tank vessel loading operation, heat exchange system, or decoking operation that meets the criteria in paragraphs (c)(1) through (9) of this section is added to an existing petroleum refinery or if another deliberate operational process change creating an additional Group 1 emissions point(s) (as defined in § 63.641) is made to an existing petroleum refining process unit, and if the addition or process change is not subject to the new source requirements as determined according to paragraph (i) or (j) of this section, the requirements in paragraphs (l)(1) through (4) of this section shall apply. Examples of process changes include, but are not limited to, changes in production capacity, or feed or raw material where the change requires construction or physical alteration of the existing equipment or catalyst type, or whenever there is replacement, removal, or addition of recovery equipment. For purposes of this paragraph (l) and paragraph (m) of this section, process changes do not include: Process upsets, unintentional temporary process changes, and changes that are within the equipment configuration and operating conditions documented in the Notification of Compliance Status report required by § 63.655(f).

* * * * *

(2) The added emission point(s) and any emission point(s) within the added or changed petroleum refining process unit shall be in compliance with the applicable requirements in item (4) of table 11 of this subpart by the dates specified in paragraph (l)(2)(i) or (ii) of this section.

(i) If a petroleum refining process unit is added to a plant site or an emission point(s) is added to any existing petroleum refining process unit, the added emission point(s) shall be in compliance upon initial startup of any added petroleum refining process unit or emission point(s) or by the applicable compliance date in item (4) of table 11 of this subpart, whichever is later.

* * * * *

(3) The owner or operator of a petroleum refining process unit or of a storage vessel, miscellaneous process vent, wastewater stream, gasoline loading rack, marine tank vessel loading operation, heat exchange system, or decoking operation meeting the criteria in paragraphs (c)(1) through (9) of this section that is added to a plant site and is subject to the requirements for existing sources shall comply with the reporting and recordkeeping requirements that are applicable to existing sources including, but not limited to, the reports listed in paragraphs (l)(3)(i) through (vii) of this section. A process change to an existing petroleum refining process unit shall be subject to the reporting requirements for existing sources including, but not limited to, the reports listed in paragraphs (l)(3)(i) through (vii) of this section. The applicable reports include, but are not limited to:

* * * * *

(m) If a change that does not meet the criteria in paragraph (l) of this section is made to a petroleum refining process unit subject to this subpart, and the change causes a Group 2 emission point to become a Group 1 emission point (as defined in § 63.641), then the owner or operator shall comply with the applicable requirements of this subpart for existing sources, as specified in item (4) of table 11 of this subpart, for the Group 1 emission point as expeditiously as practicable, but in no event later than 3 years after the emission point becomes Group 1.

* * * * *

(n) Overlap of this subpart with other regulations for storage vessels. As applicable, paragraphs (n)(1), (3), (4), (6), and (7) of this section apply for Group 2 storage vessels and paragraphs (n)(2) and (5) of this section apply for Group 1 storage vessels.

(1) After the compliance dates specified in paragraph (h) of this section, a Group 2 storage vessel that is subject to the provisions of 40 CFR part 60, subpart Kb, is required to comply only with the requirements of 40 CFR part 60, subpart Kb, except as provided in paragraph (n)(8) of this section. After the compliance dates specified in paragraph (h) of this section, a Group 2 storage vessel that is subject to the provisions of 40 CFR part 61, subpart Y, is required to comply only with the requirements of 40 CFR part 61, subpart Y, except as provided in paragraph (n)(10) of this section.

(2) After the compliance dates specified in paragraph (h) of this section, a Group 1 storage vessel that is also subject to 40 CFR part 60, subpart Kb, is required to comply only with either 40 CFR part 60, subpart Kb, except as provided in paragraph (n)(8) of this section or this subpart. After the compliance dates specified in paragraph (h) of this section, a Group 1 storage vessel that is also subject to 40 CFR part 61, subpart Y, is required to comply only with either 40 CFR part 61, subpart Y, except as provided in paragraph (n)(10) of this section or this subpart.

(3) After the compliance dates specified in paragraph (h) of this section, a Group 2 storage vessel that is part of a new source and is subject to 40 CFR 60.110b, but is not required to apply controls by 40 CFR 60.110b or 60.112b, is required to comply only with this subpart.

(4) After the compliance dates specified in paragraph (h) of this section, a Group 2 storage vessel that is part of a new source and is subject to 40 CFR 61.270, but is not required to apply controls by 40 CFR 61.271, is required to comply only with this subpart.

(5) After the compliance dates specified in paragraph (h) of this section, a Group 1 storage vessel that is also subject to the provisions of 40 CFR part 60, subpart K or Ka, is required to only comply with the provisions of this subpart.

* * * * *

(8) Storage vessels described by paragraph (n)(1) of this section are to comply with 40 CFR part 60, subpart Kb, except as provided in paragraphs (n)(8)(i) through (vi) of this section. Storage vessels described by paragraph (n)(2) electing to comply with part 60, subpart Kb of this chapter shall comply with subpart Kb except as provided in paragraphs (n)(8)(i) through (viii) of this section.

* * * * *

(ii) If the owner or operator determines that it is unsafe to perform the seal gap measurements required in § 60.113b(b) of this chapter or to inspect the vessel to determine compliance with Start Printed Page 75239§ 60.113b(a) of this chapter because the roof appears to be structurally unsound and poses an imminent danger to inspecting personnel, the owner or operator shall comply with the requirements in either § 63.120(b)(7)(i) or (ii) of subpart G (only up to the compliance date specified in paragraph (h) of this section for compliance with § 63.660, as applicable) or either § 63.1063(c)(2)(iv)(A) or (B) of subpart WW.

* * * * *

(vii) To be in compliance with § 60.112b(a)(1)(iv) or (a)(2)(ii) of this chapter, guidepoles in floating roof storage vessels must be equipped with covers and/or controls (e.g., pole float system, pole sleeve system, internal sleeve system or flexible enclosure system) as appropriate to comply with the “no visible gap” requirement.

(viii) If a flare is used as a control device for a storage vessel, on and after January 30, 2019, the owner or operator must meet the requirements of § 63.670 instead of the requirements referenced from part 60, subpart Kb of this chapter for that flare.

(9) * * *

(i) If the owner or operator determines that it is unsafe to perform the seal gap measurements required in § 60.113a(a)(1) of this chapter because the floating roof appears to be structurally unsound and poses an imminent danger to inspecting personnel, the owner or operator shall comply with the requirements in either § 63.120(b)(7)(i) or (ii) of subpart G (only up to the compliance date specified in paragraph (h) of this section for compliance with § 63.660, as applicable) or either § 63.1063(c)(2)(iv)(A) or (B) of subpart WW.

* * * * *

(10) Storage vessels described by paragraph (n)(1) of this section are to comply with 40 CFR part 61, subpart Y, except as provided in paragraphs (n)(10)(i) through (vi) of this section. Storage vessels described by paragraph (n)(2) electing to comply with 40 CFR part 61, subpart Y, shall comply with subpart Y except as provided for in paragraphs (n)(10)(i) through (viii) of this section.

(i) Storage vessels that are to comply with § 61.271(b) of this chapter are exempt from the secondary seal requirements of § 61.271(b)(2)(ii) of this chapter during the gap measurements for the primary seal required by § 61.272(b) of this chapter.

(ii) If the owner or operator determines that it is unsafe to perform the seal gap measurements required in § 61.272(b) of this chapter or to inspect the vessel to determine compliance with § 61.272(a) of this chapter because the roof appears to be structurally unsound and poses an imminent danger to inspecting personnel, the owner or operator shall comply with the requirements in either § 63.120(b)(7)(i) or (ii) of subpart G (only up to the compliance date specified in paragraph (h) of this section for compliance with § 63.660, as applicable) or either § 63.1063(c)(2)(iv)(A) or (B) of subpart WW.

(iii) If a failure is detected during the inspections required by § 61.272(a)(2) of this chapter or during the seal gap measurements required by § 61.272(b)(1) of this chapter, and the vessel cannot be repaired within 45 days and the vessel cannot be emptied within 45 days, the owner or operator may utilize up to two extensions of up to 30 additional calendar days each. The owner or operator is not required to provide a request for the extension to the Administrator.

(iv) If an extension is utilized in accordance with paragraph (n)(10)(iii) of this section, the owner or operator shall, in the next periodic report, identify the vessel, provide the information listed in § 61.272(a)(2) or (b)(4)(iii) of this chapter, and describe the nature and date of the repair made or provide the date the storage vessel was emptied.

(v) Owners and operators of storage vessels complying with 40 CFR part 61, subpart Y, may submit the inspection reports required by § 61.275(a), (b)(1), and (d) of this chapter as part of the periodic reports required by this subpart, rather than within the 60-day period specified in § 61.275(a), (b)(1), and (d) of this chapter.

(vi) The reports of rim seal inspections specified in § 61.275(d) of this chapter are not required if none of the measured gaps or calculated gap areas exceed the limitations specified in § 61.272(b)(4) of this chapter. Documentation of the inspections shall be recorded as specified in § 61.276(a) of this chapter.

(vii) To be in compliance with § 61.271(a)(6) or (b)(3) of this chapter, guidepoles in floating roof storage vessels must be equipped with covers and/or controls (e.g., pole float system, pole sleeve system, internal sleeve system or flexible enclosure system) as appropriate to comply with the “no visible gap” requirement.

(viii) If a flare is used as a control device for a storage vessel, on and after January 30, 2019, the owner or operator must meet the requirements of § 63.670 instead of the requirements referenced from part 61, subpart Y of this chapter for that flare.

(o) * * *

(2) * * *

(i) Comply with paragraphs (o)(2)(i)(A) through (D) of this section.

* * * * *

(D) If a flare is used as a control device, on and after January 30, 2019, the flare shall meet the requirements of § 63.670. Prior to January 30, 2019, the flare shall meet the applicable requirements of 40 CFR part 61, subpart FF, and subpart G of this part, or the requirements of § 63.670.

(ii) Comply with paragraphs (o)(2)(ii)(A) through (C) of this section.

* * * * *

(C) If a flare is used as a control device, on and after January 30, 2019, the flare shall meet the requirements of § 63.670. Prior to January 30, 2019, the flare shall meet the applicable requirements of 40 CFR part 61, subpart FF, and subpart G of this part, or the requirements of § 63.670.

* * * * *

(s) Overlap of this subpart with other regulation for flares. On January 30, 2019, flares that are subject to the provisions of 40 CFR 60.18 or 63.11 and subject to this subpart are required to comply only with the provisions specified in this subpart. Prior to January 30, 2019, flares that are subject to the provisions of 40 CFR 60.18 or 63.11 and elect to comply with the requirements in §§ 63.670 and 63.671 are required to comply only with the provisions specified in this subpart.

Start Amendment Part

14. Section 63.641 is amended by:

End Amendment Part Start Amendment Part

a. Adding, in alphabetical order, definitions of “Assist air,” “Assist steam,” “Center steam,” “Closed blowdown system,” “Combustion zone,” “Combustion zone gas,” “Decoking operations,” “Delayed coking unit,” “Flare,” “Flare purge gas,” “Flare supplemental gas,” “Flare sweep gas,” “Flare vent gas,” “Flexible enclosure device,” “Force majeure event,” “Lower steam,” “Net heating value,” “Perimeter assist air,” “Pilot gas,” “Premix assist air,” “Regulated material,” “Thermal expansion relief valve,” “Total steam,” and “Upper steam”; and

End Amendment Part Start Amendment Part

b. Revising the definitions of “Delayed coker vent,” “Emission point,” “Group 1 storage vessel,” “Miscellaneous process vent,” “Periodically discharged,” and “Reference control technology for storage vessels.”

End Amendment Part

The revisions and additions read as follows:

Definitions.
* * * * *

Assist air means all air that intentionally is introduced prior to or at Start Printed Page 75240a flare tip through nozzles or other hardware conveyance for the purposes including, but not limited to, protecting the design of the flare tip, promoting turbulence for mixing or inducing air into the flame. Assist air includes premix assist air and perimeter assist air. Assist air does not include the surrounding ambient air.

Assist steam means all steam that intentionally is introduced prior to or at a flare tip through nozzles or other hardware conveyance for the purposes including, but not limited to, protecting the design of the flare tip, promoting turbulence for mixing or inducing air into the flame. Assist steam includes, but is not necessarily limited to, center steam, lower steam and upper steam.

* * * * *

Center steam means the portion of assist steam introduced into the stack of a flare to reduce burnback.

Closed blowdown system means a system used for depressuring process vessels that is not open to the atmosphere and is configured of piping, ductwork, connections, accumulators/knockout drums, and, if necessary, flow inducing devices that transport gas or vapor from process vessel to a control device or back into the process.

* * * * *

Combustion zone means the area of the flare flame where the combustion zone gas combines for combustion.

Combustion zone gas means all gases and vapors found just after a flare tip. This gas includes all flare vent gas, total steam, and premix air.

* * * * *

Decoking operations means the sequence of steps conducted at the end of the delayed coking unit's cooling cycle to open the coke drum to the atmosphere in order to remove coke from the coke drum. Decoking operations begin at the end of the cooling cycle when steam released from the coke drum is no longer discharged via the unit's blowdown system but instead is vented directly to the atmosphere. Decoking operations include atmospheric depressuring (venting), deheading, draining, and decoking (coke cutting).

Delayed coker vent means a miscellaneous process vent that contains uncondensed vapors from the delayed coking unit's blowdown system. Venting from the delayed coker vent is typically intermittent in nature, and occurs primarily during the cooling cycle of a delayed coking unit coke drum when vapor from the coke drums cannot be sent to the fractionator column for product recovery. The emissions from the decoking operations, which include direct atmospheric venting, deheading, draining, or decoking (coke cutting), are not considered to be delayed coker vents.

Delayed coking unit means a refinery process unit in which high molecular weight petroleum derivatives are thermally cracked and petroleum coke is produced in a series of closed, batch system reactors. A delayed coking unit includes, but is not limited to, all of the coke drums associated with a single fractionator; the fractionator, including the bottoms receiver and the overhead condenser; the coke drum cutting water and quench system, including the jet pump and coker quench water tank; and the coke drum blowdown recovery compressor system.

* * * * *

Emission point means an individual miscellaneous process vent, storage vessel, wastewater stream, equipment leak, decoking operation or heat exchange system associated with a petroleum refining process unit; an individual storage vessel or equipment leak associated with a bulk gasoline terminal or pipeline breakout station classified under Standard Industrial Classification code 2911; a gasoline loading rack classified under Standard Industrial Classification code 2911; or a marine tank vessel loading operation located at a petroleum refinery.

* * * * *

Flare means a combustion device lacking an enclosed combustion chamber that uses an uncontrolled volume of ambient air to burn gases. For the purposes of this rule, the definition of flare includes, but is not necessarily limited to, air-assisted flares, steam-assisted flares and non-assisted flares.

Flare purge gas means gas introduced between a flare header's water seal and the flare tip to prevent oxygen infiltration (backflow) into the flare tip. For a flare with no water seal, the function of flare purge gas is performed by flare sweep gas and, therefore, by definition, such a flare has no flare purge gas.

Flare supplemental gas means all gas introduced to the flare in order to improve the combustible characteristics of combustion zone gas.

Flare sweep gas means, for a flare with a flare gas recovery system, the gas intentionally introduced into the flare header system to maintain a constant flow of gas through the flare header in order to prevent oxygen buildup in the flare header; flare sweep gas in these flares is introduced prior to and recovered by the flare gas recovery system. For a flare without a flare gas recovery system, flare sweep gas means the gas intentionally introduced into the flare header system to maintain a constant flow of gas through the flare header and out the flare tip in order to prevent oxygen buildup in the flare header and to prevent oxygen infiltration (backflow) into the flare tip.

Flare vent gas means all gas found just prior to the flare tip. This gas includes all flare waste gas (i.e., gas from facility operations that is directed to a flare for the purpose of disposing of the gas), that portion of flare sweep gas that is not recovered, flare purge gas and flare supplemental gas, but does not include pilot gas, total steam or assist air.

Flexible enclosure device means a seal made of an elastomeric fabric (or other material) which completely encloses a slotted guidepole or ladder and eliminates the vapor emission pathway from inside the storage vessel through the guidepole slots or ladder slots to the outside air.

* * * * *

Force majeure event means a release of HAP, either directly to the atmosphere from a relief valve or discharged via a flare, that is demonstrated to the satisfaction of the Administrator to result from an event beyond the refinery owner or operator's control, such as natural disasters; acts of war or terrorism; loss of a utility external to the refinery (e.g., external power curtailment), excluding power curtailment due to an interruptible service agreement; and fire or explosion originating at a near or adjoining facility outside of the refinery owner or operator's control that impacts the refinery's ability to operate.

* * * * *

Group 1 storage vessel means:

(1) Prior to February 1, 2016:

(i) A storage vessel at an existing source that has a design capacity greater than or equal to 177 cubic meters and stored-liquid maximum true vapor pressure greater than or equal to 10.4 kilopascals and stored-liquid annual average true vapor pressure greater than or equal to 8.3 kilopascals and annual average HAP liquid concentration greater than 4 percent by weight total organic HAP;

(ii) A storage vessel at a new source that has a design storage capacity greater than or equal to 151 cubic meters and stored-liquid maximum true vapor pressure greater than or equal to 3.4 kilopascals and annual average HAP liquid concentration greater than 2 percent by weight total organic HAP; or

(iii) A storage vessel at a new source that has a design storage capacity greater than or equal to 76 cubic meters and less than 151 cubic meters and stored-Start Printed Page 75241liquid maximum true vapor pressure greater than or equal to 77 kilopascals and annual average HAP liquid concentration greater than 2 percent by weight total organic HAP.

(2) On and after February 1, 2016:

(i) A storage vessel at an existing source that has a design capacity greater than or equal to 151 cubic meters (40,000 gallons) and stored-liquid maximum true vapor pressure greater than or equal to 5.2 kilopascals (0.75 pounds per square inch) and annual average HAP liquid concentration greater than 4 percent by weight total organic HAP;

(ii) A storage vessel at an existing source that has a design storage capacity greater than or equal to 76 cubic meters (20,000 gallons) and less than 151 cubic meters (40,000 gallons) and stored-liquid maximum true vapor pressure greater than or equal to 13.1 kilopascals (1.9 pounds per square inch) and annual average HAP liquid concentration greater than 4 percent by weight total organic HAP;

(iii) A storage vessel at a new source that has a design storage capacity greater than or equal to 151 cubic meters (40,000 gallons) and stored-liquid maximum true vapor pressure greater than or equal to 3.4 kilopascals (0.5 pounds per square inch) and annual average HAP liquid concentration greater than 2 percent by weight total organic HAP; or

(iv) A storage vessel at a new source that has a design storage capacity greater than or equal to 76 cubic meters (20,000 gallons) and less than 151 cubic meters (40,000 gallons) and stored-liquid maximum true vapor pressure greater than or equal to 13.1 kilopascals (1.9 pounds per square inch) and annual average HAP liquid concentration greater than 2 percent by weight total organic HAP.

* * * * *

Lower steam means the portion of assist steam piped to an exterior annular ring near the lower part of a flare tip, which then flows through tubes to the flare tip, and ultimately exits the tubes at the flare tip.

* * * * *

Miscellaneous process vent means a gas stream containing greater than 20 parts per million by volume organic HAP that is continuously or periodically discharged from a petroleum refining process unit meeting the criteria specified in § 63.640(a). Miscellaneous process vents include gas streams that are discharged directly to the atmosphere, gas streams that are routed to a control device prior to discharge to the atmosphere, or gas streams that are diverted through a product recovery device prior to control or discharge to the atmosphere. Miscellaneous process vents include vent streams from: Caustic wash accumulators, distillation tower condensers/accumulators, flash/knockout drums, reactor vessels, scrubber overheads, stripper overheads, vacuum pumps, steam ejectors, hot wells, high point bleeds, wash tower overheads, water wash accumulators, blowdown condensers/accumulators, and delayed coker vents. Miscellaneous process vents do not include:

(1) Gaseous streams routed to a fuel gas system, provided that on and after January 30, 2019, any flares receiving gas from the fuel gas system are in compliance with § 63.670;

(2) Pressure relief device discharges;

(3) Leaks from equipment regulated under § 63.648;

(4) [Reserved]

(5) In situ sampling systems (onstream analyzers) until January 30, 2019. After this date, these sampling systems will be included in the definition of miscellaneous process vents;

(6) Catalytic cracking unit catalyst regeneration vents;

(7) Catalytic reformer regeneration vents;

(8) Sulfur plant vents;

(9) Vents from control devices such as scrubbers, boilers, incinerators, and electrostatic precipitators applied to catalytic cracking unit catalyst regeneration vents, catalytic reformer regeneration vents, and sulfur plant vents;

(10) Vents from any stripping operations applied to comply with the wastewater provisions of this subpart, subpart G of this part, or 40 CFR part 61, subpart FF;

(11) Emissions associated with delayed coking unit decoking operations;

(12) Vents from storage vessels;

(13) Emissions from wastewater collection and conveyance systems including, but not limited to, wastewater drains, sewer vents, and sump drains; and

(14) Hydrogen production plant vents through which carbon dioxide is removed from process streams or through which steam condensate produced or treated within the hydrogen plant is degassed or deaerated.

Net heating value means the energy released as heat when a compound undergoes complete combustion with oxygen to form gaseous carbon dioxide and gaseous water (also referred to as lower heating value).

* * * * *

Perimeter assist air means the portion of assist air introduced at the perimeter of the flare tip or above the flare tip. Perimeter assist air includes air intentionally entrained in lower and upper steam. Perimeter assist air includes all assist air except premix assist air.

Periodically discharged means discharges that are intermittent and associated with routine operations, maintenance activities, startups, shutdowns, malfunctions, or process upsets.

* * * * *

Pilot gas means gas introduced into a flare tip that provides a flame to ignite the flare vent gas.

* * * * *

Premix assist air means the portion of assist air that is introduced to the flare vent gas, whether injected or induced, prior to the flare tip. Premix assist air also includes any air intentionally entrained in center steam.

* * * * *

Reference control technology for storage vessels means either:

(1) For Group 1 storage vessels complying with § 63.660:

(i) An internal floating roof, including an external floating roof converted to an internal floating roof, meeting the specifications of § 63.1063(a)(1)(i) and (b);

(ii) An external floating roof meeting the specifications of § 63.1063(a)(1)(ii), (a)(2), and (b); or

(iii) [Reserved]

(iv) A closed-vent system to a control device that reduces organic HAP emissions by 95 percent, or to an outlet concentration of 20 parts per million by volume (ppmv).

(v) For purposes of emissions averaging, these four technologies are considered equivalent.

(2) For all other storage vessels:

(i) An internal floating roof meeting the specifications of § 63.119(b) of subpart G except for § 63.119(b)(5) and (6);

(ii) An external floating roof meeting the specifications of § 63.119(c) of subpart G except for § 63.119(c)(2);

(iii) An external floating roof converted to an internal floating roof meeting the specifications of § 63.119(d) of subpart G except for § 63.119(d)(2); or

(iv) A closed-vent system to a control device that reduces organic HAP emissions by 95 percent, or to an outlet concentration of 20 parts per million by volume.

(v) For purposes of emissions averaging, these four technologies are considered equivalent.

* * * * *

Regulated material means any stream associated with emission sources listed Start Printed Page 75242in § 63.640(c) required to meet control requirements under this subpart as well as any stream for which this subpart or a cross-referencing subpart specifies that the requirements for flare control devices in § 63.670 must be met.

* * * * *

Thermal expansion relief valve means a pressure relief valve designed to protect equipment from excess pressure due to thermal expansion of blocked liquid-filled equipment or piping due to ambient heating or heat from a heat tracing system. Pressure relief valves designed to protect equipment from excess pressure due to blockage against a pump or compressor or due to fire contingency are not thermal expansion relief valves.

* * * * *

Total steam means the total of all steam that is supplied to a flare and includes, but is not limited to, lower steam, center steam and upper steam.

Upper steam means the portion of assist steam introduced via nozzles located on the exterior perimeter of the upper end of the flare tip.

* * * * *
Start Amendment Part

15. Section 63.642 is amended by:

End Amendment Part Start Amendment Part

a. Adding paragraph (b);

End Amendment Part Start Amendment Part

b. Revising paragraphs (d)(3), (e), (i), (k) introductory text, (k)(1), (l) introductory text, and (l)(2); and

End Amendment Part Start Amendment Part

c. Adding paragraph (n).

End Amendment Part

The revisions and additions read as follows:

General standards.
* * * * *

(b) The emission standards set forth in this subpart shall apply at all times.

* * * * *

(d) * * *

(3) Performance tests shall be conducted according to the provisions of § 63.7(e) except that performance tests shall be conducted at maximum representative operating capacity for the process. During the performance test, an owner or operator shall operate the control device at either maximum or minimum representative operating conditions for monitored control device parameters, whichever results in lower emission reduction. An owner or operator shall not conduct a performance test during startup, shutdown, periods when the control device is bypassed or periods when the process, monitoring equipment or control device is not operating properly. The owner/operator may not conduct performance tests during periods of malfunction. The owner or operator must record the process information that is necessary to document operating conditions during the test and include in such record an explanation to support that the test was conducted at maximum representative operating capacity. Upon request, the owner or operator shall make available to the Administrator such records as may be necessary to determine the conditions of performance tests.

* * * * *

(e) All applicable records shall be maintained as specified in § 63.655(i).

* * * * *

(i) The owner or operator of an existing source shall demonstrate compliance with the emission standard in paragraph (g) of this section by following the procedures specified in paragraph (k) of this section for all emission points, or by following the emissions averaging compliance approach specified in paragraph (l) of this section for specified emission points and the procedures specified in paragraph (k)(1) of this section.

* * * * *

(k) The owner or operator of an existing source may comply, and the owner or operator of a new source shall comply, with the applicable provisions in §§ 63.643 through 63.645, 63.646 or 63.660, 63.647, 63.650, and 63.651, as specified in § 63.640(h).

(1) The owner or operator using this compliance approach shall also comply with the requirements of §§ 63.648 and/or 63.649, 63.654, 63.655, 63.657, 63.658, 63.670 and 63.671, as applicable.

* * * * *

(l) The owner or operator of an existing source may elect to control some of the emission points within the source to different levels than specified under §§ 63.643 through 63.645, 63.646 or 63.660, 63.647, 63.650, and 63.651, as applicable according to § 63.640(h), by using an emissions averaging compliance approach as long as the overall emissions for the source do not exceed the emission level specified in paragraph (g) of this section. The owner or operator using emissions averaging shall meet the requirements in paragraphs (l)(1) and (2) of this section.

* * * * *

(2) Comply with the requirements of §§ 63.648 and/or 63.649, 63.654, 63.652, 63.653, 63.655, 63.657, 63.658, 63.670 and 63.671, as applicable.

* * * * *

(n) At all times, the owner or operator must operate and maintain any affected source, including associated air pollution control equipment and monitoring equipment, in a manner consistent with safety and good air pollution control practices for minimizing emissions. The general duty to minimize emissions does not require the owner operator to make any further efforts to reduce emissions if levels required by the applicable standard have been achieved. Determination of whether a source is operating in compliance with operation and maintenance requirements will be based on information available to the Administrator which may include, but is not limited to, monitoring results, review of operation and maintenance procedures, review of operation and maintenance records, and inspection of the source.

Start Amendment Part

16. Section 63.643 is amended by revising paragraphs (a) introductory text and (a)(1) and adding paragraph (c) to read as follows:

End Amendment Part
Miscellaneous process vent provisions.

(a) The owner or operator of a Group 1 miscellaneous process vent as defined in § 63.641 shall comply with the requirements of either paragraph (a)(1) or (2) of this section or, if applicable, paragraph (c) of this section. The owner or operator of a miscellaneous process vent that meets the conditions in paragraph (c) of this section is only required to comply with the requirements of paragraph (c) of this section and § 63.655(g)(13) and (i)(12) for that vent.

(1) Reduce emissions of organic HAP's using a flare. On and after January 30, 2019, the flare shall meet the requirements of § 63.670. Prior to January 30, 2019, the flare shall meet the requirements of § 63.11(b) of subpart A or the requirements of § 63.670.

* * * * *

(c) An owner or operator may designate a process vent as a maintenance vent if the vent is only used as a result of startup, shutdown, maintenance, or inspection of equipment where equipment is emptied, depressurized, degassed or placed into service. The owner of operator does not need to designate a maintenance vent as a Group 1 or Group 2 miscellaneous process vent. The owner or operator must comply with the applicable requirements in paragraphs (c)(1) through (3) of this section for each maintenance vent.

(1) Prior to venting to the atmosphere, process liquids are removed from the equipment as much as practical and the equipment is depressured to a control device, fuel gas system, or back to the process until one of the following conditions, as applicable, is met.

(i) The vapor in the equipment served by the maintenance vent has a lower Start Printed Page 75243explosive limit (LEL) of less than 10 percent.

(ii) If there is no ability to measure the LEL of the vapor in the equipment based on the design of the equipment, the pressure in the equipment served by the maintenance vent is reduced to 5 psig or less. Upon opening the maintenance vent, active purging of the equipment cannot be used until the LEL of the vapors in the maintenance vent (or inside the equipment if the maintenance is a hatch or similar type of opening) equipment is less than 10 percent.

(iii) The equipment served by the maintenance vent contains less than 72 pounds of VOC.

(iv) If the maintenance vent is associated with equipment containing pyrophoric catalyst (e.g., hydrotreaters and hydrocrackers) at refineries that do not have a pure hydrogen supply, the LEL of the vapor in the equipment must be less than 20 percent, except for one event per year not to exceed 35 percent.

(2) Except for maintenance vents complying with the alternative in paragraph (c)(1)(iii) of this section, the owner or operator must determine the LEL or, if applicable, equipment pressure using process instrumentation or portable measurement devices and follow procedures for calibration and maintenance according to manufacturer's specifications.

(3) For maintenance vents complying with the alternative in paragraph (c)(1)(iii) of this section, the owner or operator shall determine mass of VOC in the equipment served by the maintenance vent based on the equipment size and contents after considering any contents drained or purged from the equipment. Equipment size may be determined from equipment design specifications. Equipment contents may be determined using process knowledge.

Start Amendment Part

17. Section 63.644 is amended by revising paragraphs (a) introductory text, (a)(2), and (c) to read as follows:

End Amendment Part
Monitoring provisions for miscellaneous process vents.

(a) Except as provided in paragraph (b) of this section, each owner or operator of a Group 1 miscellaneous process vent that uses a combustion device to comply with the requirements in § 63.643(a) shall install the monitoring equipment specified in paragraph (a)(1), (2), (3), or (4) of this section, depending on the type of combustion device used. All monitoring equipment shall be installed, calibrated, maintained, and operated according to manufacturer's specifications or other written procedures that provide adequate assurance that the equipment will monitor accurately and, except for CPMS installed for pilot flame monitoring, must meet the applicable minimum accuracy, calibration and quality control requirements specified in table 13 of this subpart.

* * * * *

(2) Where a flare is used prior to January 30, 2019, a device (including but not limited to a thermocouple, an ultraviolet beam sensor, or an infrared sensor) capable of continuously detecting the presence of a pilot flame is required, or the requirements of § 63.670 shall be met. Where a flare is used on and after January 30, 2019, the requirements of § 63.670 shall be met.

* * * * *

(c) The owner or operator of a Group 1 miscellaneous process vent using a vent system that contains bypass lines that could divert a vent stream away from the control device used to comply with paragraph (a) of this section either directly to the atmosphere or to a control device that does not comply with the requirements in § 63.643(a) shall comply with either paragraph (c)(1) or (2) of this section. Use of the bypass at any time to divert a Group 1 miscellaneous process vent stream to the atmosphere or to a control device that does not comply with the requirements in § 63.643(a) is an emissions standards violation. Equipment such as low leg drains and equipment subject to § 63.648 are not subject to this paragraph (c).

(1) Install, calibrate and maintain a flow indicator that determines whether a vent stream flow is present at least once every hour. A manual block valve equipped with a valve position indicator may be used in lieu of a flow indicator, as long as the valve position indicator is monitored continuously. Records shall be generated as specified in § 63.655(h) and (i). The flow indicator shall be installed at the entrance to any bypass line that could divert the vent stream away from the control device to the atmosphere; or

(2) Secure the bypass line valve in the non-diverting position with a car-seal or a lock-and-key type configuration. A visual inspection of the seal or closure mechanism shall be performed at least once every month to ensure that the valve is maintained in the non-diverting position and that the vent stream is not diverted through the bypass line.

* * * * *
Start Amendment Part

18. Section 63.645 is amended by revising paragraphs (e)(1) and (f)(2) to read as follows:

End Amendment Part
Test methods and procedures for miscellaneous process vents.
* * * * *

(e) * * *

(1) Methods 1 or 1A of 40 CFR part 60, appendix A-1, as appropriate, shall be used for selection of the sampling site. For vents smaller than 0.10 meter in diameter, sample at the center of the vent.

* * * * *

(f) * * *

(2) The gas volumetric flow rate shall be determined using Methods 2, 2A, 2C, 2D, or 2F of 40 CFR part 60, appendix A-1 or Method 2G of 40 CFR part 60, appendix A-2, as appropriate.

* * * * *
Start Amendment Part

19. Section 63.646 is amended by adding introductory text and revising paragraph (b)(2) to read as follows:

End Amendment Part
Storage vessel provisions.

Upon a demonstration of compliance with the standards in § 63.660 by the compliance dates specified in § 63.640(h), the standards in this section shall no longer apply.

* * * * *

(b) * * *

(2) When an owner or operator and the Administrator do not agree on whether the annual average weight percent organic HAP in the stored liquid is above or below 4 percent for a storage vessel at an existing source or above or below 2 percent for a storage vessel at a new source, an appropriate method (based on the type of liquid stored) as published by EPA or a consensus-based standards organization shall be used. Consensus-based standards organizations include, but are not limited to, the following: ASTM International (100 Barr Harbor Drive, P.O. Box CB700, West Conshohocken, Pennsylvania 19428-B2959, (800) 262-1373, http://www.astm.org), the American National Standards Institute (ANSI, 1819 L Street NW., 6th floor, Washington, DC 20036, (202) 293-8020, http://www.ansi.org), the American Gas Association (AGA, 400 North Capitol Street NW., 4th Floor, Washington, DC 20001, (202) 824-7000, http://www.aga.org), the American Society of Mechanical Engineers (ASME, Three Park Avenue, New York, NY 10016-5990, (800) 843-2763, http://www.asme.org), the American Petroleum Institute (API, 1220 L Street NW., Washington, DC 20005-4070, (202) 682-8000, http://www.api.org), and the North American Energy Standards Board (NAESB, 801 Travis Street, Suite 1675, Houston, TX 77002, (713) 356-0060, http://www.naesb.org).

* * * * *
Start Printed Page 75244 Start Amendment Part

20. Section 63.647 is amended by:

End Amendment Part Start Amendment Part

a. Revising paragraph (a);

End Amendment Part Start Amendment Part

b. Redesignating paragraph (c) as paragraph (d); and

End Amendment Part Start Amendment Part

c. Adding paragraph (c).

End Amendment Part

The revisions and additions read as follows:

Wastewater provisions.

(a) Except as provided in paragraphs (b) and (c) of this section, each owner or operator of a Group 1 wastewater stream shall comply with the requirements of §§ 61.340 through 61.355 of this chapter for each process wastewater stream that meets the definition in § 63.641.

* * * * *

(c) If a flare is used as a control device, on and after January 30, 2019, the flare shall meet the requirements of § 63.670. Prior to January 30, 2019, the flare shall meet the applicable requirements of part 61, subpart FF of this chapter, or the requirements of § 63.670.

* * * * *
Start Amendment Part

21. Section 63.648 is amended by:

End Amendment Part Start Amendment Part

a. Adding paragraph (a)(3);

End Amendment Part Start Amendment Part

b. Revising paragraph (c) introductory text; and

End Amendment Part Start Amendment Part

c. Adding paragraphs (c)(11) and (12) and (j).

End Amendment Part

The revisions and additions read as follows:

Equipment leak standards.

(a) * * *

(3) If a flare is used as a control device, on and after January 30, 2019, the flare shall meet the requirements of § 63.670. Prior to January 30, 2019, the flare shall meet the applicable requirements of part 60, subpart VV of this chapter, or the requirements of § 63.670.

* * * * *

(c) In lieu of complying with the existing source provisions of paragraph (a) in this section, an owner or operator may elect to comply with the requirements of §§ 63.161 through 63.169, 63.171, 63.172, 63.175, 63.176, 63.177, 63.179, and 63.180 of subpart H except as provided in paragraphs (c)(1) through (12) and (e) through (i) of this section.

* * * * *

(11) [Reserved]

(12) If a flare is used as a control device, on and after January 30, 2019, the flare shall meet the requirements of § 63.670. Prior to January 30, 2019, the flare shall meet the applicable requirements of §§ 63.172 and 63.180, or the requirements of § 63.670.

* * * * *

(j) Except as specified in paragraph (j)(4) of this section, the owner or operator must comply with the requirements specified in paragraphs (j)(1) and (2) of this section for pressure relief devices, such as relief valves or rupture disks, in organic HAP gas or vapor service instead of the pressure relief device requirements of § 60.482-4 or § 63.165, as applicable. Except as specified in paragraphs (j)(4) and (5) of this section, the owner or operator must also comply with the requirements specified in paragraph (j)(3) of this section for all pressure relief devices.

(1) Operating requirements. Except during a pressure release, operate each pressure relief device in organic HAP gas or vapor service with an instrument reading of less than 500 ppm above background as detected by Method 21 of 40 CFR part 60, appendix A-7.

(2) Pressure release requirements. For pressure relief devices in organic HAP gas or vapor service, the owner or operator must comply with the applicable requirements in paragraphs (j)(2)(i) through (iii) of this section following a pressure release.

(i) If the pressure relief device does not consist of or include a rupture disk, conduct instrument monitoring, as specified in § 60.485(b) or § 63.180(c), as applicable, no later than 5 calendar days after the pressure relief device returns to organic HAP gas or vapor service following a pressure release to verify that the pressure relief device is operating with an instrument reading of less than 500 ppm.

(ii) If the pressure relief device includes a rupture disk, either comply with the requirements in paragraph (j)(2)(i) of this section (not replacing the rupture disk) or install a replacement disk as soon as practicable after a pressure release, but no later than 5 calendar days after the pressure release. The owner or operator must conduct instrument monitoring, as specified in § 60.485(b) or § 63.180(c), as applicable, no later than 5 calendar days after the pressure relief device returns to organic HAP gas or vapor service following a pressure release to verify that the pressure relief device is operating with an instrument reading of less than 500 ppm.

(iii) If the pressure relief device consists only of a rupture disk, install a replacement disk as soon as practicable after a pressure release, but no later than 5 calendar days after the pressure release. The owner or operator may not initiate startup of the equipment served by the rupture disk until the rupture disc is replaced. The owner or operator must conduct instrument monitoring, as specified in § 60.485(b) or § 63.180(c), as applicable, no later than 5 calendar days after the pressure relief device returns to organic HAP gas or vapor service following a pressure release to verify that the pressure relief device is operating with an instrument reading of less than 500 ppm.

(3) Pressure release management. Except as specified in paragraphs (j)(4) and (5) of this section, the owner or operator shall comply with the requirements specified in paragraphs (j)(3)(i) through (v) of this section for all pressure relief devices in organic HAP service no later than January 30, 2019.

(i) The owner or operator must equip each affected pressure relief device with a device(s) or use a monitoring system that is capable of:

(A) Identifying the pressure release;

(B) Recording the time and duration of each pressure release; and

(C) Notifying operators immediately that a pressure release is occurring. The device or monitoring system may be either specific to the pressure relief device itself or may be associated with the process system or piping, sufficient to indicate a pressure release to the atmosphere. Examples of these types of devices and systems include, but are not limited to, a rupture disk indicator, magnetic sensor, motion detector on the pressure relief valve stem, flow monitor, or pressure monitor.

(ii) The owner or operator must apply at least three redundant prevention measures to each affected pressure relief device and document these measures. Examples of prevention measures include:

(A) Flow, temperature, level and pressure indicators with deadman switches, monitors, or automatic actuators.

(B) Documented routine inspection and maintenance programs and/or operator training (maintenance programs and operator training may count as only one redundant prevention measure).

(C) Inherently safer designs or safety instrumentation systems.

(D) Deluge systems.

(E) Staged relief system where initial pressure relief valve (with lower set release pressure) discharges to a flare or other closed vent system and control device.

(iii) If any affected pressure relief device releases to atmosphere as a result of a pressure release event, the owner or operator must perform root cause analysis and corrective action analysis according to the requirement in paragraph (j)(6) of this section and implement corrective actions according to the requirements in paragraph (j)(7) of this section. The owner or operator must also calculate the quantity of organic HAP released during each pressure Start Printed Page 75245release event and report this quantity as required in § 63.655(g)(10)(iii). Calculations may be based on data from the pressure relief device monitoring alone or in combination with process parameter monitoring data and process knowledge.

(iv) The owner or operator shall determine the total number of release events occurred during the calendar year for each affected pressure relief device separately. The owner or operator shall also determine the total number of release events for each pressure relief device for which the root cause analysis concluded that the root cause was a force majeure event, as defined in this subpart.

(v) Except for pressure relief devices described in paragraphs (j)(4) and (5) of this section, the following release events are a violation of the pressure release management work practice standards.

(A) Any release event for which the root cause of the event was determined to be operator error or poor maintenance.

(B) A second release event not including force majeure events from a single pressure relief device in a 3 calendar year period for the same root cause for the same equipment.

(C) A third release event not including force majeure events from a single pressure relief device in a 3 calendar year period for any reason.

(4) Pressure relief devices routed to a control device. If all releases and potential leaks from a pressure relief device are routed through a closed vent system to a control device, back into the process or to the fuel gas system, the owner or operator is not required to comply with paragraph (j)(1), (2), or (3) (if applicable) of this section. Both the closed vent system and control device (if applicable) must meet the requirements of § 63.644. When complying with this paragraph (j)(4), all references to “Group 1 miscellaneous process vent” in § 63.644 mean “pressure relief device.” If a pressure relief device complying with this paragraph (j)(4) is routed to the fuel gas system, then on and after January 30, 2019, any flares receiving gas from that fuel gas system must be in compliance with § 63.670.

(5) Pressure relief devices exempted from pressure release management requirements. The following types of pressure relief devices are not subject to the pressure release management requirements in paragraph (j)(3) of this section.

(i) Pressure relief devices in heavy liquid service, as defined in § 63.641.

(ii) Pressure relief devices that only release material that is liquid at standard conditions (1 atmosphere and 68 degrees Fahrenheit) and that are hard-piped to a controlled drain system (i.e., a drain system meeting the requirements for Group 1 wastewater streams in § 63.647(a)) or piped back to the process or pipeline.

(iii) Thermal expansion relief valves.

(iv) Pressure relief devices designed with a set relief pressure of less than 2.5 psig.

(v) Pressure relief devices that do not have the potential to emit 72 lbs/day or more of VOC based on the valve diameter, the set release pressure, and the equipment contents.

(vi) Pressure relief devices on mobile equipment.

(6) Root cause analysis and corrective action analysis. A root cause analysis and corrective action analysis must be completed as soon as possible, but no later than 45 days after a release event. Special circumstances affecting the number of root cause analyses and/or corrective action analyses are provided in paragraphs (j)(6)(i) through (iv) of this section.

(i) You may conduct a single root cause analysis and corrective action analysis for a single emergency event that causes two or more pressure relief devices installed on the same equipment to release.

(ii) You may conduct a single root cause analysis and corrective action analysis for a single emergency event that causes two or more pressure relief devices to release, regardless of the equipment served, if the root cause is reasonably expected to be a force majeure event, as defined in this subpart.

(iii) Except as provided in paragraphs (j)(6)(i) and (ii) of this section, if more than one pressure relief device has a release during the same time period, an initial root cause analysis shall be conducted separately for each pressure relief device that had a release. If the initial root cause analysis indicates that the release events have the same root cause(s), the initially separate root cause analyses may be recorded as a single root cause analysis and a single corrective action analysis may be conducted.

(7) Corrective action implementation. Each owner or operator required to conduct a root cause analysis and corrective action analysis as specified in paragraphs (j)(3)(iii) and (j)(6) of this section shall implement the corrective action(s) identified in the corrective action analysis in accordance with the applicable requirements in paragraphs (j)(7)(i) through (iii) of this section.

(i) All corrective action(s) must be implemented within 45 days of the event for which the root cause and corrective action analyses were required or as soon thereafter as practicable. If an owner or operator concludes that no corrective action should be implemented, the owner or operator shall record and explain the basis for that conclusion no later than 45 days following the event.

(ii) For corrective actions that cannot be fully implemented within 45 days following the event for which the root cause and corrective action analyses were required, the owner or operator shall develop an implementation schedule to complete the corrective action(s) as soon as practicable.

(iii) No later than 45 days following the event for which a root cause and corrective action analyses were required, the owner or operator shall record the corrective action(s) completed to date, and, for action(s) not already completed, a schedule for implementation, including proposed commencement and completion dates.

Start Amendment Part

22. Section 63.649 is amended by revising definition of C c term in the equation in paragraph (c)(6)(i) to read as follows:

End Amendment Part
Alternative means of emission limitation: Connectors in gas/vapor service and light liquid service.
* * * * *

(c) * * *

(6) * * *

(i) * * *

Cc = Optional credit for removed connectors = 0.67 × net number (i.e., the total number of connectors removed minus the total added) of connectors in organic HAP service removed from the process unit after the applicability date set forth in § 63.640(h)(3)(iii) for existing process units, and after the date of start-up for new process units. If credits are not taken, then Cc = 0.

* * * * *
Start Amendment Part

23. Section 63.650 is amended by revising paragraph (a) and adding paragraph (d) to read as follows:

End Amendment Part
Gasoline loading rack provisions.

(a) Except as provided in paragraphs (b) through (d) of this section, each owner or operator of a Group 1 gasoline loading rack classified under Standard Industrial Classification code 2911 located within a contiguous area and under common control with a petroleum refinery shall comply with subpart R of this part, §§ 63.421, 63.422(a) through (c) and (e), 63.425(a) through (c) and (e) through (i), 63.427(a) and (b), and 63.428(b), (c), (g)(1), (h)(1) through (3), and (k).

* * * * *
Start Printed Page 75246

(d) If a flare is used as a control device, on and after January 30, 2019, the flare shall meet the requirements of § 63.670. Prior to January 30, 2019, the flare shall meet the applicable requirements of subpart R of this part, or the requirements of § 63.670.

Start Amendment Part

24. Section 63.651 is amended by revising paragraphs (a) and (d) and adding paragraph (e) to read as follows:

End Amendment Part
Marine tank vessel loading operation provisions.

(a) Except as provided in paragraphs (b) through (e) of this section, each owner or operator of a marine tank vessel loading operation located at a petroleum refinery shall comply with the requirements of §§ 63.560 through 63.568.

* * * * *

(d) The compliance time of 4 years after promulgation of 40 CFR part 63, subpart Y, does not apply. The compliance time is specified in § 63.640(h)(1).

(e) If a flare is used as a control device, on and after January 30, 2019, the flare shall meet the requirements of § 63.670. Prior to January 30, 2019, the flare shall meet the applicable requirements of subpart Y of this part, or the requirements of § 63.670.

Start Amendment Part

25. Section 63.652 is amended by:

End Amendment Part Start Amendment Part

a. Revising paragraph (a);

End Amendment Part Start Amendment Part

b. Removing and reserving paragraph (f)(2); and

End Amendment Part Start Amendment Part

c. Revising paragraphs (g)(2)(iii)(B)( 1), (h)(3), (k) introductory text, and (k)(3).

End Amendment Part

The revisions and additions read as follows:

Emissions averaging provisions.

(a) This section applies to owners or operators of existing sources who seek to comply with the emission standard in § 63.642(g) by using emissions averaging according to § 63.642(l) rather than following the provisions of §§ 63.643 through 63.645, 63.646 or 63.660, 63.647, 63.650, and 63.651. Existing marine tank vessel loading operations located at the Valdez Marine Terminal source may not comply with the standard by using emissions averaging.

* * * * *

(g) * * *

(2) * * *

(iii) * * *

(B) * * *

(1) The percent reduction shall be measured according to the procedures in § 63.116 of subpart G if a combustion control device is used. For a flare meeting the criteria in § 63.116(a) of subpart G or § 63.670, as applicable, or a boiler or process heater meeting the criteria in § 63.645(d) or § 63.116(b) of subpart G, the percentage of reduction shall be 98 percent. If a noncombustion control device is used, percentage of reduction shall be demonstrated by a performance test at the inlet and outlet of the device, or, if testing is not feasible, by a control design evaluation and documented engineering calculations.

* * * * *

(h) * * *

(3) Emissions from storage vessels shall be determined as specified in § 63.150(h)(3) of subpart G, except as follows:

(i) For storage vessels complying with § 63.646:

(A) All references to § 63.119(b) in § 63.150(h)(3) of subpart G shall be replaced with: § 63.119(b) or § 63.119(b) except for § 63.119(b)(5) and (6).

(B) All references to § 63.119(c) in § 63.150(h)(3) of subpart G shall be replaced with: § 63.119(c) or § 63.119(c) except for § 63.119(c)(2).

(C) All references to § 63.119(d) in § 63.150(h)(3) of subpart G shall be replaced with: § 63.119(d) or § 63.119(d) except for § 63.119(d)(2).

(ii) For storage vessels complying with § 63.660:

(A) Section 63.1063(a)(1)(i), (a)(2), and (b) or § 63.1063(a)(1)(i) and (b) shall apply instead of § 63.119(b) in § 63.150(h)(3) of subpart G.

(B) Section 63.1063(a)(1)(ii), (a)(2), and (b) shall apply instead of § 63.119(c) in § 63.150(h)(3) of subpart G.

(C) Section 63.1063(a)(1)(i), (a)(2), and (b) or § 63.1063(a)(1)(i) and (b) shall apply instead of § 63.119(d) in § 63.150(h)(3) of subpart G.

* * * * *

(k) The owner or operator shall demonstrate that the emissions from the emission points proposed to be included in the average will not result in greater hazard or, at the option of the State or local permitting authority, greater risk to human health or the environment than if the emission points were controlled according to the provisions in §§ 63.643 through 63.645, 63.646 or 63.660, 63.647, 63.650, and 63.651, as applicable.

* * * * *

(3) An emissions averaging plan that does not demonstrate an equivalent or lower hazard or risk to the satisfaction of the State or local permitting authority shall not be approved. The State or local permitting authority may require such adjustments to the emissions averaging plan as are necessary in order to ensure that the average will not result in greater hazard or risk to human health or the environment than would result if the emission points were controlled according to §§ 63.643 through 63.645, 63.646 or 63.660, 63.647, 63.650, and 63.651, as applicable.

* * * * *
Start Amendment Part

26. Section 63.653 is amended by revising paragraphs (a) introductory text, (a)(3)(i) and (ii), and (a)(7) to read as follows:

End Amendment Part
Monitoring, recordkeeping, and implementation plan for emissions averaging.

(a) For each emission point included in an emissions average, the owner or operator shall perform testing, monitoring, recordkeeping, and reporting equivalent to that required for Group 1 emission points complying with §§ 63.643 through 63.645, 63.646 or 63.660, 63.647, 63.650, and 63.651, as applicable. The specific requirements for miscellaneous process vents, storage vessels, wastewater, gasoline loading racks, and marine tank vessels are identified in paragraphs (a)(1) through (7) of this section.

* * * * *

(3) * * *

(i) Perform the monitoring or inspection procedures in § 63.646 and either § 63.120 of subpart G or § 63.1063 of subpart WW, as applicable; and

(ii) For closed vent systems with control devices, conduct an initial design evaluation as specified in § 63.646 and either § 63.120(d) of subpart G or § 63.985(b) of subpart SS, as applicable.

* * * * *

(7) If an emission point in an emissions average is controlled using a pollution prevention measure or a device or technique for which no monitoring parameters or inspection procedures are specified in §§ 63.643 through 63.645, 63.646 or 63.660, 63.647, 63.650, and 63.651, as applicable, the owner or operator shall establish a site-specific monitoring parameter and shall submit the information specified in § 63.655(h)(4) in the Implementation Plan.

* * * * *
Start Amendment Part

27. Section 63.655 is amended by:

End Amendment Part Start Amendment Part

a. Revising paragraphs (f) introductory text, (f)(1) introductory text, (f)(1)(i)(A) introductory text, (f)(1)(i)(A)(2) and (3), (f)(1)(i)(B) introductory text, (f)(1)(i)(B)(2), (f)(1)(i)(D)(2), (f)(1)(iv) introductory text, and (f)(1)(iv)(A);

End Amendment Part Start Amendment Part

b. Adding paragraphs (f)(1)(vii) and (viii);

End Amendment Part Start Amendment Part

c. Revising paragraphs (f)(2) introductory text, (f)(3) introductory text, the first sentence of (f)(6), (g) introductory text, (g)(1) through (5), (g)(6)(i)(D), (g)(6)(iii), and (g)(7)(i); Start Printed Page 75247

End Amendment Part Start Amendment Part

d. Adding paragraphs (g)(10) through (14);

End Amendment Part Start Amendment Part

e. Removing and reserving paragraph (h)(1);

End Amendment Part Start Amendment Part

f. Revising paragraphs (h)(2) introductory text, (h)(2)(i)(B), (h)(2)(ii), and (h)(5)(iii);

End Amendment Part Start Amendment Part

g. Adding paragraphs (h)(8) and (9) and (i) introductory text;

End Amendment Part Start Amendment Part

h. Revising paragraph (i)(1) introductory text and paragraph (i)(1)(ii);

End Amendment Part Start Amendment Part

i. Adding paragraphs (i)(1)(v) and (vi);

End Amendment Part Start Amendment Part

j. Redesignating paragraphs (i)(4) and (5) as paragraphs (i)(5) and (6), respectively;

End Amendment Part Start Amendment Part

k. Adding paragraph (i)(4);

End Amendment Part Start Amendment Part

l. Revising newly redesignated paragraph (i)(5) introductory text; and

End Amendment Part Start Amendment Part

m. Adding paragraphs (i)(7) through (12).

End Amendment Part

The revisions and additions read as follows:

Reporting and recordkeeping requirements.
* * * * *

(f) Each owner or operator of a source subject to this subpart shall submit a Notification of Compliance Status report within 150 days after the compliance dates specified in § 63.640(h) with the exception of Notification of Compliance Status reports submitted to comply with § 63.640(l)(3) and for storage vessels subject to the compliance schedule specified in § 63.640(h)(2). Notification of Compliance Status reports required by § 63.640(l)(3) and for storage vessels subject to the compliance dates specified in § 63.640(h)(2) shall be submitted according to paragraph (f)(6) of this section. This information may be submitted in an operating permit application, in an amendment to an operating permit application, in a separate submittal, or in any combination of the three. If the required information has been submitted before the date 150 days after the compliance date specified in § 63.640(h), a separate Notification of Compliance Status report is not required within 150 days after the compliance dates specified in § 63.640(h). If an owner or operator submits the information specified in paragraphs (f)(1) through (5) of this section at different times, and/or in different submittals, later submittals may refer to earlier submittals instead of duplicating and resubmitting the previously submitted information. Each owner or operator of a gasoline loading rack classified under Standard Industrial Classification Code 2911 located within a contiguous area and under common control with a petroleum refinery subject to the standards of this subpart shall submit the Notification of Compliance Status report required by subpart R of this part within 150 days after the compliance dates specified in § 63.640(h).

(1) The Notification of Compliance Status report shall include the information specified in paragraphs (f)(1)(i) through (viii) of this section.

(i) * * *

(A) Identification of each storage vessel subject to this subpart, and for each Group 1 storage vessel subject to this subpart, the information specified in paragraphs (f)(1)(i)(A)(1) through (3) of this section. This information is to be revised each time a Notification of Compliance Status report is submitted for a storage vessel subject to the compliance schedule specified in § 63.640(h)(2) or to comply with § 63.640(l)(3).

* * * * *

(2) For storage vessels subject to the compliance schedule specified in § 63.640(h)(2) that are not complying with § 63.646, the anticipated compliance date.

(3) For storage vessels subject to the compliance schedule specified in § 63.640(h)(2) that are complying with § 63.646 and the Group 1 storage vessels described in § 63.640(l), the actual compliance date.

(B) If a closed vent system and a control device other than a flare is used to comply with § 63.646 or § 63.660, the owner or operator shall submit:

* * * * *

(2) The design evaluation documentation specified in § 63.120(d)(1)(i) of subpart G or § 63.985(b)(1)(i) of subpart SS (as applicable), if the owner or operator elects to prepare a design evaluation; or

* * * * *

(D) * * *

(2) All visible emission readings, heat content determinations, flow rate measurements, and exit velocity determinations made during the compliance determination required by § 63.120(e) of subpart G or § 63.987(b) of subpart SS or § 63.670(h), as applicable; and

* * * * *

(iv) For miscellaneous process vents controlled by flares, initial compliance test results including the information in paragraphs (f)(1)(iv)(A) and (B) of this section.

(A) All visible emission readings, heat content determinations, flow rate measurements, and exit velocity determinations made during the compliance determination required by §§ 63.645 and 63.116(a) of subpart G or § 63.670(h), as applicable; and

* * * * *

(vii) For pressure relief devices in organic HAP service subject to the requirements in § 63.648(j)(3)(i) and (ii), this report shall include the information specified in paragraphs (f)(1)(vii)(A) and (B) of this section.

(A) A description of the monitoring system to be implemented, including the relief devices and process parameters to be monitored, and a description of the alarms or other methods by which operators will be notified of a pressure release.

(B) A description of the prevention measures to be implemented for each affected pressure relief device.

(viii) For each delayed coking unit, identification of whether the unit is an existing affected source or a new affected source and whether monitoring will be conducted as specified in § 63.657(b) or (c).

(2) If initial performance tests are required by §§ 63.643 through 63.653, the Notification of Compliance Status report shall include one complete test report for each test method used for a particular source. On and after February 1, 2016, performance tests shall be submitted according to paragraph (h)(9) of this section.

* * * * *

(3) For each monitored parameter for which a range is required to be established under § 63.120(d) of subpart G or § 63.985(b) of subpart SS for storage vessels or § 63.644 for miscellaneous process vents, the Notification of Compliance Status report shall include the information in paragraphs (f)(3)(i) through (iii) of this section.

* * * * *

(6) Notification of Compliance Status reports required by § 63.640(l)(3) and for storage vessels subject to the compliance dates specified in § 63.640(h)(2) shall be submitted no later than 60 days after the end of the 6-month period during which the change or addition was made that resulted in the Group 1 emission point or the existing Group 1 storage vessel was brought into compliance, and may be combined with the periodic report. * * *

(g) The owner or operator of a source subject to this subpart shall submit Periodic Reports no later than 60 days after the end of each 6-month period when any of the information specified in paragraphs (g)(1) through (7) of this section or paragraphs (g)(9) through (14) of this section is collected. The first 6-month period shall begin on the date the Notification of Compliance Status report is required to be submitted. A Periodic Report is not required if none of the events identified in paragraphs (g)(1) Start Printed Page 75248through (7) of this section or paragraphs (g)(9) through (14) of this section occurred during the 6-month period unless emissions averaging is utilized. Quarterly reports must be submitted for emission points included in emission averages, as provided in paragraph (g)(8) of this section. An owner or operator may submit reports required by other regulations in place of or as part of the Periodic Report required by this paragraph (g) if the reports contain the information required by paragraphs (g)(1) through (14) of this section.

(1) For storage vessels, Periodic Reports shall include the information specified for Periodic Reports in paragraphs (g)(2) through (5) of this section. Information related to gaskets, slotted membranes, and sleeve seals is not required for storage vessels that are part of an existing source complying with § 63.646.

(2) Internal floating roofs. (i) An owner or operator who elects to comply with § 63.646 by using a fixed roof and an internal floating roof or by using an external floating roof converted to an internal floating roof shall submit the results of each inspection conducted in accordance with § 63.120(a) of subpart G in which a failure is detected in the control equipment.

(A) For vessels for which annual inspections are required under § 63.120(a)(2)(i) or (a)(3)(ii) of subpart G, the specifications and requirements listed in paragraphs (g)(2)(i)(A)(1) through (3) of this section apply.

(1) A failure is defined as any time in which the internal floating roof is not resting on the surface of the liquid inside the storage vessel and is not resting on the leg supports; or there is liquid on the floating roof; or the seal is detached from the internal floating roof; or there are holes, tears, or other openings in the seal or seal fabric; or there are visible gaps between the seal and the wall of the storage vessel.

(2) Except as provided in paragraph (g)(2)(i)(A)(3) of this section, each Periodic Report shall include the date of the inspection, identification of each storage vessel in which a failure was detected, and a description of the failure. The Periodic Report shall also describe the nature of and date the repair was made or the date the storage vessel was emptied.

(3) If an extension is utilized in accordance with § 63.120(a)(4) of subpart G, the owner or operator shall, in the next Periodic Report, identify the vessel; include the documentation specified in § 63.120(a)(4) of subpart G; and describe the date the storage vessel was emptied and the nature of and date the repair was made.

(B) For vessels for which inspections are required under § 63.120(a)(2)(ii), (a)(3)(i), or (a)(3)(iii) of subpart G (i.e., internal inspections), the specifications and requirements listed in paragraphs (g)(2)(i)(B)(1) and (2) of this section apply.

(1) A failure is defined as any time in which the internal floating roof has defects; or the primary seal has holes, tears, or other openings in the seal or the seal fabric; or the secondary seal (if one has been installed) has holes, tears, or other openings in the seal or the seal fabric; or, for a storage vessel that is part of a new source, the gaskets no longer close off the liquid surface from the atmosphere; or, for a storage vessel that is part of a new source, the slotted membrane has more than a 10 percent open.

(2) Each Periodic Report shall include the date of the inspection, identification of each storage vessel in which a failure was detected, and a description of the failure. The Periodic Report shall also describe the nature of and date the repair was made.

(ii) An owner or operator who elects to comply with § 63.660 by using a fixed roof and an internal floating roof shall submit the results of each inspection conducted in accordance with § 63.1063(c)(1), (d)(1), and (d)(2) of subpart WW in which a failure is detected in the control equipment. For vessels for which inspections are required under § 63.1063(c) and (d), the specifications and requirements listed in paragraphs (g)(2)(ii)(A) through (C) of this section apply.

(A) A failure is defined in § 63.1063(d)(1) of subpart WW.

(B) Each Periodic Report shall include a copy of the inspection record required by § 63.1065(b) of subpart WW when a failure occurs.

(C) An owner or operator who elects to use an extension in accordance with § 63.1063(e)(2) of subpart WW shall, in the next Periodic Report, submit the documentation required by § 63.1063(e)(2).

(3) External floating roofs. (i) An owner or operator who elects to comply with § 63.646 by using an external floating roof shall meet the periodic reporting requirements specified in paragraphs (g)(3)(i)(A) through (C) of this section.

(A) The owner or operator shall submit, as part of the Periodic Report, documentation of the results of each seal gap measurement made in accordance with § 63.120(b) of subpart G in which the seal and seal gap requirements of § 63.120(b)(3), (4), (5), or (6) of subpart G are not met. This documentation shall include the information specified in paragraphs (g)(3)(i)(A)(1) through (4) of this section.

(1) The date of the seal gap measurement.

(2) The raw data obtained in the seal gap measurement and the calculations described in § 63.120(b)(3) and (4) of subpart G.

(3) A description of any seal condition specified in § 63.120(b)(5) or (6) of subpart G that is not met.

(4) A description of the nature of and date the repair was made, or the date the storage vessel was emptied.

(B) If an extension is utilized in accordance with § 63.120(b)(7)(ii) or (b)(8) of subpart G, the owner or operator shall, in the next Periodic Report, identify the vessel; include the documentation specified in § 63.120(b)(7)(ii) or (b)(8) of subpart G, as applicable; and describe the date the vessel was emptied and the nature of and date the repair was made.

(C) The owner or operator shall submit, as part of the Periodic Report, documentation of any failures that are identified during visual inspections required by § 63.120(b)(10) of subpart G. This documentation shall meet the specifications and requirements in paragraphs (g)(3)(i)(C)(1) and (2) of this section.

(1) A failure is defined as any time in which the external floating roof has defects; or the primary seal has holes or other openings in the seal or the seal fabric; or the secondary seal has holes, tears, or other openings in the seal or the seal fabric; or, for a storage vessel that is part of a new source, the gaskets no longer close off the liquid surface from the atmosphere; or, for a storage vessel that is part of a new source, the slotted membrane has more than 10 percent open area.

(2) Each Periodic Report shall include the date of the inspection, identification of each storage vessel in which a failure was detected, and a description of the failure. The Periodic Report shall also describe the nature of and date the repair was made.

(ii) An owner or operator who elects to comply with § 63.660 by using an external floating roof shall meet the periodic reporting requirements specified in paragraphs (g)(3)(ii)(A) and (B) of this section.

(A) For vessels for which inspections are required under § 63.1063(c)(2), (d)(1), and (d)(3) of subpart WW, the owner or operator shall submit, as part of the Periodic Report, a copy of the inspection record required by § 63.1065(b) of subpart WW when a failure occurs. A failure is defined in § 63.1063(d)(1).Start Printed Page 75249

(B) An owner or operator who elects to use an extension in accordance with § 63.1063(e)(2) or (c)(2)(iv)(B) of subpart WW shall, in the next Periodic Report, submit the documentation required by those paragraphs.

(4) [Reserved]

(5) An owner or operator who elects to comply with § 63.646 or § 63.660 by installing a closed vent system and control device shall submit, as part of the next Periodic Report, the information specified in paragraphs (g)(5)(i) through (v) of this section, as applicable.

(i) The Periodic Report shall include the information specified in paragraphs (g)(5)(i)(A) and (B) of this section for those planned routine maintenance operations that would require the control device not to meet the requirements of either § 63.119(e)(1) or (2) of subpart G, § 63.985(a) and (b) of subpart SS, or § 63.670, as applicable.

(A) A description of the planned routine maintenance that is anticipated to be performed for the control device during the next 6 months. This description shall include the type of maintenance necessary, planned frequency of maintenance, and lengths of maintenance periods.

(B) A description of the planned routine maintenance that was performed for the control device during the previous 6 months. This description shall include the type of maintenance performed and the total number of hours during those 6 months that the control device did not meet the requirements of either § 63.119(e)(1) or (2) of subpart G, § 63.985(a) and (b) of subpart SS, or § 63.670, as applicable, due to planned routine maintenance.

(ii) If a control device other than a flare is used, the Periodic Report shall describe each occurrence when the monitored parameters were outside of the parameter ranges documented in the Notification of Compliance Status report. The description shall include: Identification of the control device for which the measured parameters were outside of the established ranges, and causes for the measured parameters to be outside of the established ranges.

(iii) If a flare is used prior to January 30, 2019 and prior to electing to comply with the requirements in § 63.670, the Periodic Report shall describe each occurrence when the flare does not meet the general control device requirements specified in § 63.11(b) of subpart A and shall include: Identification of the flare that does not meet the general requirements specified in § 63.11(b) of subpart A, and reasons the flare did not meet the general requirements specified in § 63.11(b) of subpart A.

(iv) If a flare is used on or after the date for which compliance with the requirements in § 63.670 is elected, which can be no later than January 30, 2019, the Periodic Report shall include the items specified in paragraph (g)(11) of this section.

(v) An owner or operator who elects to comply with § 63.660 by installing an alternate control device as described in § 63.1064 of subpart WW shall submit, as part of the next Periodic Report, a written application as described in § 63.1066(b)(3) of subpart WW.

(6) * * *

(i) * * *

(D) For data compression systems under paragraph (h)(5)(iii) of this section, an operating day when the monitor operated for less than 75 percent of the operating hours or a day when less than 18 monitoring values were recorded.

* * * * *

(iii) For periods in closed vent systems when a Group 1 miscellaneous process vent stream was detected in the bypass line or diverted from the control device and either directly to the atmosphere or to a control device that does not comply with the requirements in § 63.643(a), report the date, time, duration, estimate of the volume of gas, the concentration of organic HAP in the gas and the resulting mass emissions of organic HAP that bypassed the control device. For periods when the flow indicator is not operating, report the date, time, and duration.

(7) * * *

(i) Results of the performance test shall include the identification of the source tested, the date of the test, the percentage of emissions reduction or outlet pollutant concentration reduction (whichever is needed to determine compliance) for each run and for the average of all runs, and the values of the monitored operating parameters.

* * * * *

(10) For pressure relief devices subject to the requirements § 63.648(j), Periodic Reports must include the information specified in paragraphs (g)(10)(i) through (iii) of this section.

(i) For pressure relief devices in organic HAP gas or vapor service, pursuant to § 63.648(j)(1), report any instrument reading of 500 ppm or greater.

(ii) For pressure relief devices in organic HAP gas or vapor service subject to § 63.648(j)(2), report confirmation that any monitoring required to be done during the reporting period to show compliance was conducted.

(iii) For pressure relief devices in organic HAP service subject to § 63.648(j)(3), report each pressure release to the atmosphere, including duration of the pressure release and estimate of the mass quantity of each organic HAP released, and the results of any root cause analysis and corrective action analysis completed during the reporting period, including the corrective actions implemented during the reporting period and, if applicable, the implementation schedule for planned corrective actions to be implemented subsequent to the reporting period.

(11) For flares subject to § 63.670, Periodic Reports must include the information specified in paragraphs (g)(11)(i) through (iv) of this section.

(i) Records as specified in paragraph (i)(9)(i) of this section for each 15-minute block during which there was at least one minute when regulated material is routed to a flare and no pilot flame is present.

(ii) Visible emission records as specified in paragraph (i)(9)(ii)(C) of this section for each period of 2 consecutive hours during which visible emissions exceeded a total of 5 minutes.

(iii) The 15-minute block periods for which the applicable operating limits specified in § 63.670(d) through (f) are not met. Indicate the date and time for the period, the net heating value operating parameter(s) determined following the methods in § 63.670(k) through (n) as applicable.

(iv) For flaring events meeting the criteria in § 63.670(o)(3):

(A) The start and stop time and date of the flaring event.

(B) The length of time for which emissions were visible from the flare during the event.

(C) The periods of time that the flare tip velocity exceeds the maximum flare tip velocity determined using the methods in § 63.670(d)(2) and the maximum 15-minute block average flare tip velocity recorded during the event.

(D) Results of the root cause and corrective actions analysis completed during the reporting period, including the corrective actions implemented during the reporting period and, if applicable, the implementation schedule for planned corrective actions to be implemented subsequent to the reporting period.

(12) For delayed coking units, the Periodic Report must include the information specified in paragraphs (g)(12)(i) through (iv) of this section.

(i) For existing source delayed coking units, any 60-cycle average exceeding the applicable limit in § 63.657(a)(1).

(ii) For new source delayed coking units, any direct venting event Start Printed Page 75250exceeding the applicable limit in § 63.657(a)(2).

(iii) The total number of double quenching events performed during the reporting period.

(iv) For each double quenching draining event when the drain water temperature exceeded 210 °F, report the drum, date, time, the coke drum vessel pressure or temperature, as applicable, when pre-vent draining was initiated, and the maximum drain water temperature during the pre-vent draining period.

(13) For maintenance vents subject to the requirements in § 63.643(c), Periodic Reports must include the information specified in paragraphs (g)(13)(i) through (iv) of this section for any release exceeding the applicable limits in § 63.643(c)(1). For the purposes of this reporting requirement, owners or operators complying with § 63.643(c)(1)(iv) must report each venting event for which the lower explosive limit is 20 percent or greater.

(i) Identification of the maintenance vent and the equipment served by the maintenance vent.

(ii) The date and time the maintenance vent was opened to the atmosphere.

(iii) The lower explosive limit, vessel pressure, or mass of VOC in the equipment, as applicable, at the start of atmospheric venting. If the 5 psig vessel pressure option in § 63.643(c)(1)(ii) was used and active purging was initiated while the lower explosive limit was 10 percent or greater, also include the lower explosive limit of the vapors at the time active purging was initiated.

(iv) An estimate of the mass of organic HAP released during the entire atmospheric venting event.

(14) Any changes in the information provided in a previous Notification of Compliance Status report.

(h) * * *

(2) For storage vessels, notifications of inspections as specified in paragraphs (h)(2)(i) and (ii) of this section.

(i) * * *

(B) Except as provided in paragraph (h)(2)(i)(C) of this section, if the internal inspection required by § 63.120(a)(2), (a)(3), or (b)(10) of subpart G or § 63.1063(d)(1) of subpart WW is not planned and the owner or operator could not have known about the inspection 30 calendar days in advance of refilling the vessel with organic HAP, the owner or operator shall notify the Administrator at least 7 calendar days prior to refilling of the storage vessel. Notification may be made by telephone and immediately followed by written documentation demonstrating why the inspection was unplanned. This notification, including the written documentation, may also be made in writing and sent so that it is received by the Administrator at least 7 calendar days prior to the refilling.

* * * * *

(ii) In order to afford the Administrator the opportunity to have an observer present, the owner or operator of a storage vessel equipped with an external floating roof shall notify the Administrator of any seal gap measurements. The notification shall be made in writing at least 30 calendar days in advance of any gap measurements required by § 63.120(b)(1) or (2) of subpart G or § 63.1062(d)(3) of subpart WW. The State or local permitting authority can waive this notification requirement for all or some storage vessels subject to the rule or can allow less than 30 calendar days' notice.

* * * * *

(5) * * *

(iii) An owner or operator may use an automated data compression recording system that does not record monitored operating parameter values at a set frequency (for example, once every hour) but records all values that meet set criteria for variation from previously recorded values.

(A) The system shall be designed to:

(1) Measure the operating parameter value at least once every hour.

(2) Record at least 24 values each day during periods of operation.

(3) Record the date and time when monitors are turned off or on.

(4) Recognize unchanging data that may indicate the monitor is not functioning properly, alert the operator, and record the incident.

(5) Compute daily average values of the monitored operating parameter based on recorded data.

(B) You must maintain a record of the description of the monitoring system and data compression recording system including the criteria used to determine which monitored values are recorded and retained, the method for calculating daily averages, and a demonstrations that they system meets all criteria of paragraph (h)(5)(iii)(A) of this section.

* * * * *

(8) For fenceline monitoring systems subject to § 63.658, within 45 calendar days after the end of each quarterly reporting period covered by the periodic report, each owner or operator shall submit the following information to the EPA's Compliance and Emissions Data Reporting Interface (CEDRI). (CEDRI can be accessed through the EPA's Central Data Exchange (CDX) (https://cdx.epa.gov/​). The owner or operator need not transmit this data prior to obtaining 12 months of data.

(i) Individual sample results for each monitor for each sampling period during the quarterly reporting period. For the first reporting period and for any period in which a passive monitor is added or moved, the owner or operator shall report the coordinates of all of the passive monitor locations. The owner or operator shall determine the coordinates using an instrument with an accuracy of at least 3 meters. Coordinates shall be in decimal degrees with at least five decimal places.

(ii) The biweekly annual average concentration difference (Δc) values for benzene for the quarterly reporting period.

(iii) Notation for each biweekly value that indicates whether background correction was used, all measurements in the sampling period were below detection, or whether an outlier was removed from the sampling period data set.

(9) On and after February 1, 2016, if required to submit the results of a performance test or CEMS performance evaluation, the owner or operator shall submit the results according to the procedures in paragraphs (h)(9)(i) and (ii) of this section.

(i) Within 60 days after the date of completing each performance test as required by this subpart, the owner or operator shall submit the results of the performance tests following the procedure specified in either paragraph (h)(9)(i)(A) or (B) of this section.

(A) For data collected using test methods supported by the EPA's Electronic Reporting Tool (ERT) as listed on the EPA's ERT Web site (http://www.epa.gov/​ttn/​chief/​ert/​index.html) at the time of the test, the owner or operator must submit the results of the performance test to the EPA via the CEDRI. (CEDRI can be accessed through the EPA's CDX.) Performance test data must be submitted in a file format generated through the use of the EPA's ERT or an alternate electronic file format consistent with the extensible markup language (XML) schema listed on the EPA's ERT Web site. If an owner or operator claims that some of the performance test information being submitted is confidential business information (CBI), the owner or operator must submit a complete file generated through the use of the EPA's ERT or an alternate electronic file consistent with the XML schema listed on the EPA's ERT Web site, including information claimed to be CBI, on a compact disc, flash drive or other commonly used electronic storage media to the EPA. The electronic storage media must be clearly marked as CBI and mailed to U.S. EPA/OAQPS/Start Printed Page 75251CORE CBI Office, Attention: Group Leader, Measurement Policy Group, MD C404-02, 4930 Old Page Rd., Durham, NC 27703. The same ERT or alternate file with the CBI omitted must be submitted to the EPA via the EPA's CDX as described earlier in this paragraph (h)(9)(i)(A).

(B) For data collected using test methods that are not supported by the EPA's ERT as listed on the EPA's ERT Web site at the time of the test, the owner or operator must submit the results of the performance test to the Administrator at the appropriate address listed in § 63.13.

(ii) Within 60 days after the date of completing each CEMS performance evaluation as required by this subpart, the owner or operator must submit the results of the performance evaluation following the procedure specified in either paragraph (h)(9)(ii)(A) or (B) of this section.

(A) For performance evaluations of continuous monitoring systems measuring relative accuracy test audit (RATA) pollutants that are supported by the EPA's ERT as listed on the EPA's ERT Web site at the time of the evaluation, the owner or operator must submit the results of the performance evaluation to the EPA via the CEDRI. (CEDRI can be accessed through the EPA's CDX.) Performance evaluation data must be submitted in a file format generated through the use of the EPA's ERT or an alternate file format consistent with the XML schema listed on the EPA's ERT Web site. If an owner or operator claims that some of the performance evaluation information being submitted is CBI, the owner or operator must submit a complete file generated through the use of the EPA's ERT or an alternate electronic file consistent with the XML schema listed on the EPA's ERT Web site, including information claimed to be CBI, on a compact disc, flash drive or other commonly used electronic storage media to the EPA. The electronic storage media must be clearly marked as CBI and mailed to U.S. EPA/OAQPS/CORE CBI Office, Attention: Group Leader, Measurement Policy Group, MD C404-02, 4930 Old Page Rd., Durham, NC 27703. The same ERT or alternate file with the CBI omitted must be submitted to the EPA via the EPA's CDX as described earlier in this paragraph (h)(9)(ii)(A).

(B) For any performance evaluations of continuous monitoring systems measuring RATA pollutants that are not supported by the EPA's ERT as listed on the EPA's ERT Web site at the time of the evaluation, the owner or operator must submit the results of the performance evaluation to the Administrator at the appropriate address listed in § 63.13.

(i) Recordkeeping. Each owner or operator of a source subject to this subpart shall keep copies of all applicable reports and records required by this subpart for at least 5 years except as otherwise specified in paragraphs (i)(1) through (12) of this section. All applicable records shall be maintained in such a manner that they can be readily accessed within 24 hours. Records may be maintained in hard copy or computer-readable form including, but not limited to, on paper, microfilm, computer, flash drive, floppy disk, magnetic tape, or microfiche.

(1) Each owner or operator subject to the storage vessel provisions in § 63.646 shall keep the records specified in § 63.123 of subpart G except as specified in paragraphs (i)(1)(i) through (iv) of this section. Each owner or operator subject to the storage vessel provisions in § 63.660 shall keep records as specified in paragraphs (i)(1)(v) and (vi) of this section.

* * * * *

(ii) All references to § 63.122 in § 63.123 of subpart G shall be replaced with § 63.655(e).

* * * * *

(v) Each owner or operator of a Group 1 storage vessel subject to the provisions in § 63.660 shall keep records as specified in § 63.1065 or § 63.998, as applicable.

(vi) Each owner or operator of a Group 2 storage vessel shall keep the records specified in § 63.1065(a) of subpart WW. If a storage vessel is determined to be Group 2 because the weight percent total organic HAP of the stored liquid is less than or equal to 4 percent for existing sources or 2 percent for new sources, a record of any data, assumptions, and procedures used to make this determination shall be retained.

* * * * *

(4) For each closed vent system that contains bypass lines that could divert a vent stream away from the control device and either directly to the atmosphere or to a control device that does not comply with the requirements in § 63.643(a), the owner or operator shall keep a record of the information specified in either paragraph (i)(4)(i) or (ii) of this section, as applicable.

(i) The owner or operator shall maintain records of periods when flow was detected in the bypass line, including the date and time and the duration of the flow in the bypass line. For each flow event, the owner or operator shall maintain records sufficient to determine whether or not the detected flow included flow of a Group 1 miscellaneous process vent stream requiring control. For periods when the Group 1 miscellaneous process vent stream requiring control is diverted from the control device and released either directly to the atmosphere or to a control device that does not comply with the requirements in § 63.643(a), the owner or operator shall include an estimate of the volume of gas, the concentration of organic HAP in the gas and the resulting emissions of organic HAP that bypassed the control device using process knowledge and engineering estimates.

(ii) Where a seal mechanism is used to comply with § 63.644(c)(2), hourly records of flow are not required. In such cases, the owner or operator shall record the date that the monthly visual inspection of the seals or closure mechanisms is completed. The owner or operator shall also record the occurrence of all periods when the seal or closure mechanism is broken, the bypass line valve position has changed or the key for a lock-and-key type lock has been checked out. The owner or operator shall include an estimate of the volume of gas, the concentration of organic HAP in the gas and the resulting mass emissions of organic HAP from the Group 1 miscellaneous process vent stream requiring control that bypassed the control device or records sufficient to demonstrate that there was no flow of a Group 1 miscellaneous process vent stream requiring control during the period.

(5) The owner or operator of a heat exchange system subject to this subpart shall comply with the recordkeeping requirements in paragraphs (i)(5)(i) through (v) of this section and retain these records for 5 years.

* * * * *

(7) Each owner or operator subject to the delayed coking unit decoking operations provisions in § 63.657 must maintain records specified in paragraphs (i)(7)(i) through (iii) of this section.

(i) The average pressure or temperature, as applicable, for the 5-minute period prior to venting to the atmosphere, draining, or deheading the coke drum for each cooling cycle for each coke drum.

(ii) If complying with the 60-cycle rolling average, each 60-cycle rolling average pressure or temperature, as applicable, considering all coke drum venting events in the existing affected source.

(iii) For double-quench cooling cycles:Start Printed Page 75252

(A) The date, time and duration of each pre-vent draining event.

(B) The pressure or temperature of the coke drum vessel, as applicable, for the 15 minute period prior to the pre-vent draining.

(C) The drain water temperature at 1-minute intervals from the start of pre-vent draining to the complete closure of the drain valve.

(8) For fenceline monitoring systems subject to § 63.658, each owner or operator shall keep the records specified in paragraphs (i)(8)(i) through (x) of this section on an ongoing basis.

(i) Coordinates of all passive monitors, including replicate samplers and field blanks, and if applicable, the meteorological station. The owner or operator shall determine the coordinates using an instrument with an accuracy of at least 3 meters. The coordinates shall be in decimal degrees with at least five decimal places.

(ii) The start and stop times and dates for each sample, as well as the tube identifying information.

(iii) Sampling period average temperature and barometric pressure measurements.

(iv) For each outlier determined in accordance with Section 9.2 of Method 325A of appendix A of this part, the sampler location of and the concentration of the outlier and the evidence used to conclude that the result is an outlier.

(v) For samples that will be adjusted for a background, the location of and the concentration measured simultaneously by the background sampler, and the perimeter samplers to which it applies.

(vi) Individual sample results, the calculated Δc for benzene for each sampling period and the two samples used to determine it, whether background correction was used, and the annual average Δc calculated after each sampling period.

(vii) Method detection limit for each sample, including co-located samples and blanks.

(viii) Documentation of corrective action taken each time the action level was exceeded.

(ix) Other records as required by Methods 325A and 325B of appendix A of this part.

(x) If a near-field source correction is used as provided in § 63.658(i), records of hourly meteorological data, including temperature, barometric pressure, wind speed and wind direction, calculated daily unit vector wind direction and daily sigma theta, and other records specified in the site-specific monitoring plan.

(9) For each flare subject to § 63.670, each owner or operator shall keep the records specified in paragraphs (i)(9)(i) through (xii) of this section up-to-date and readily accessible, as applicable.

(i) Retain records of the output of the monitoring device used to detect the presence of a pilot flame as required in § 63.670(b) for a minimum of 2 years. Retain records of each 15-minute block during which there was at least one minute that no pilot flame is present when regulated material is routed to a flare for a minimum of 5 years.

(ii) Retain records of daily visible emissions observations or video surveillance images required in § 63.670(h) as specified in the paragraphs (i)(9)(ii)(A) through (C), as applicable, for a minimum of 3 years.

(A) If visible emissions observations are performed using Method 22 at 40 CFR part 60, appendix A-7, the record must identify whether the visible emissions observation was performed, the results of each observation, total duration of observed visible emissions, and whether it was a 5-minute or 2-hour observation. If the owner or operator performs visible emissions observations more than one time during a day, the record must also identify the date and time of day each visible emissions observation was performed.

(B) If video surveillance camera is used, the record must include all video surveillance images recorded, with time and date stamps.

(C) For each 2 hour period for which visible emissions are observed for more than 5 minutes in 2 consecutive hours, the record must include the date and time of the 2 hour period and an estimate of the cumulative number of minutes in the 2 hour period for which emissions were visible.

(iii) The 15-minute block average cumulative flows for flare vent gas and, if applicable, total steam, perimeter assist air, and premix assist air specified to be monitored under § 63.670(i), along with the date and time interval for the 15-minute block. If multiple monitoring locations are used to determine cumulative vent gas flow, total steam, perimeter assist air, and premix assist air, retain records of the 15-minute block average flows for each monitoring location for a minimum of 2 years, and retain the 15-minute block average cumulative flows that are used in subsequent calculations for a minimum of 5 years. If pressure and temperature monitoring is used, retain records of the 15-minute block average temperature, pressure and molecular weight of the flare vent gas or assist gas stream for each measurement location used to determine the 15-minute block average cumulative flows for a minimum of 2 years, and retain the 15-minute block average cumulative flows that are used in subsequent calculations for a minimum of 5 years.

(iv) The flare vent gas compositions specified to be monitored under § 63.670(j). Retain records of individual component concentrations from each compositional analyses for a minimum of 2 years. If NHVvg analyzer is used, retain records of the 15-minute block average values for a minimum of 5 years.

(v) Each 15-minute block average operating parameter calculated following the methods specified in § 63.670(k) through (n), as applicable.

(vi) [Reserved]

(vii) All periods during which operating values are outside of the applicable operating limits specified in § 63.670(d) through (f) when regulated material is being routed to the flare.

(viii) All periods during which the owner or operator does not perform flare monitoring according to the procedures in § 63.670(g) through (j).

(ix) Records of periods when there is flow of vent gas to the flare, but when there is no flow of regulated material to the flare, including the start and stop time and dates of periods of no regulated material flow.

(x) Records when the flow of vent gas exceeds the smokeless capacity of the flare, including start and stop time and dates of the flaring event.

(xi) Records of the root cause analysis and corrective action analysis conducted as required in § 63.670(o)(3), including an identification of the affected facility, the date and duration of the event, a statement noting whether the event resulted from the same root cause(s) identified in a previous analysis and either a description of the recommended corrective action(s) or an explanation of why corrective action is not necessary under § 63.670(o)(5)(i).

(xii) For any corrective action analysis for which implementation of corrective actions are required in § 63.670(o)(5), a description of the corrective action(s) completed within the first 45 days following the discharge and, for action(s) not already completed, a schedule for implementation, including proposed commencement and completion dates.

(10) [Reserved]

(11) For each pressure relief device subject to the pressure release management work practice standards in § 63.648(j)(3), the owner or operator shall keep the records specified in paragraphs (i)(11)(i) through (iii) of this section.

(i) Records of the prevention measures implemented as required in § 63.648(j)(3)(ii), if applicable.Start Printed Page 75253

(ii) Records of the number of releases during each calendar year and the number of those releases for which the root cause was determined to be a force majeure event. Keep these records for the current calendar year and the past five calendar years.

(iii) For each release to the atmosphere, the owner or operator shall keep the records specified in paragraphs (i)(11)(iii)(A) through (D) of this section.

(A) The start and end time and date of each pressure release to the atmosphere.

(B) Records of any data, assumptions, and calculations used to estimate of the mass quantity of each organic HAP released during the event.

(C) Records of the root cause analysis and corrective action analysis conducted as required in § 63.648(j)(3)(iii), including an identification of the affected facility, the date and duration of the event, a statement noting whether the event resulted from the same root cause(s) identified in a previous analysis and either a description of the recommended corrective action(s) or an explanation of why corrective action is not necessary under § 63.648(j)(7)(i).

(D) For any corrective action analysis for which implementation of corrective actions are required in § 63.648(j)(7), a description of the corrective action(s) completed within the first 45 days following the discharge and, for action(s) not already completed, a schedule for implementation, including proposed commencement and completion dates.

(12) For each maintenance vent opening subject to the requirements in § 63.643(c), the owner or operator shall keep the applicable records specified in (i)(12)(i) through (v) of this section.

(i) The owner or operator shall maintain standard site procedures used to deinventory equipment for safety purposes (e.g., hot work or vessel entry procedures) to document the procedures used to meet the requirements in § 63.643(c). The current copy of the procedures shall be retained and available on-site at all times. Previous versions of the standard site procedures, is applicable, shall be retained for five years.

(ii) If complying with the requirements of § 63.643(c)(1)(i) and the lower explosive limit at the time of the vessel opening exceeds 10 percent, identification of the maintenance vent, the process units or equipment associated with the maintenance vent, the date of maintenance vent opening, and the lower explosive limit at the time of the vessel opening.

(iii) If complying with the requirements of § 63.643(c)(1)(ii) and either the vessel pressure at the time of the vessel opening exceeds 5 psig or the lower explosive limit at the time of the active purging was initiated exceeds 10 percent, identification of the maintenance vent, the process units or equipment associated with the maintenance vent, the date of maintenance vent opening, the pressure of the vessel or equipment at the time of discharge to the atmosphere and, if applicable, the lower explosive limit of the vapors in the equipment when active purging was initiated.

(iv) If complying with the requirements of § 63.643(c)(1)(iii), identification of the maintenance vent, the process units or equipment associated with the maintenance vent, the date of maintenance vent opening, and records used to estimate the total quantity of VOC in the equipment at the time the maintenance vent was opened to the atmosphere for each applicable maintenance vent opening.

(v) If complying with the requirements of § 63.643(c)(1)(iv), identification of the maintenance vent, the process units or equipment associated with the maintenance vent, records documenting the lack of a pure hydrogen supply, the date of maintenance vent opening, and the lower explosive limit of the vapors in the equipment at the time of discharge to the atmosphere for each applicable maintenance vent opening.

Start Amendment Part

28. Section 63.656 is amended by revising paragraph (c)(1) to read as follows:

End Amendment Part
Implementation and enforcement.
* * * * *

(c) * * *

(1) Approval of alternatives to the requirements in §§ 63.640, 63.642(g) through (l), 63.643, 63.646 through 63.652, 63.654, 63.657 through 63.660, and 63.670. Where these standards reference another subpart, the cited provisions will be delegated according to the delegation provisions of the referenced subpart. Where these standards reference another subpart and modify the requirements, the requirements shall be modified as described in this subpart. Delegation of the modified requirements will also occur according to the delegation provisions of the referenced subpart.

* * * * *
Start Amendment Part

29. Section 63.657 is added to read as follows:

End Amendment Part
Delayed coking unit decoking operation standards.

(a) Except as provided in paragraphs (e) and (f) of this section, each owner or operator of a delayed coking unit shall depressure each coke drum to a closed blowdown system until the coke drum vessel pressure or temperature measured at the top of the coke drum or in the overhead line of the coke drum as near as practical to the coke drum meets the applicable limits specified in paragraph (a)(1) or (2) of this section prior to venting to the atmosphere, draining or deheading the coke drum at the end of the cooling cycle.

(1) For delayed coking units at an existing affected source, meet either:

(i) An average vessel pressure of 2 psig determined on a rolling 60-event average; or

(ii) An average vessel temperature of 220 degrees Fahrenheit determined on a rolling 60-event average.

(2) For delayed coking units at a new affected source, meet either:

(i) A vessel pressure of 2.0 psig for each decoking event; or

(ii) A vessel temperature of 218 degrees Fahrenheit for each decoking event.

(b) Each owner or operator of a delayed coking unit complying with the pressure limits in paragraph (a)(1)(i) or (a)(2)(i) of this section shall install, operate, calibrate, and maintain a monitoring system, as specified in paragraphs (b)(1) through (5) of this section, to determine the coke drum vessel pressure.

(1) The pressure monitoring system must be in a representative location (at the top of the coke drum or in the overhead line as near as practical to the coke drum) that minimizes or eliminates pulsating pressure, vibration, and, to the extent practical, internal and external corrosion.

(2) The pressure monitoring system must be capable of measuring a pressure of 2.0 psig within ±0.5 psig.

(3) The pressure monitoring system must be verified annually or at the frequency recommended by the instrument manufacturer. The pressure monitoring system must be verified following any period of more than 24 hours throughout which the pressure exceeded the maximum rated pressure of the sensor, or the data recorder was off scale.

(4) All components of the pressure monitoring system must be visually inspected for integrity, oxidation and galvanic corrosion every 3 months, unless the system has a redundant pressure sensor.

(5) The output of the pressure monitoring system must be reviewed Start Printed Page 75254daily to ensure that the pressure readings fluctuate as expected between operating and cooling/decoking cycles to verify the pressure taps are not plugged. Plugged pressure taps must be unplugged or otherwise repaired prior to the next operating cycle.

(c) Each owner or operator of a delayed coking unit complying with the temperature limits in paragraph (a)(1)(ii) or (a)(2)(ii) of this section shall install, operate, calibrate, and maintain a continuous parameter monitoring system to measure the coke drum vessel temperature (at the top of the coke drum or in the overhead line as near as practical to the coke drum) according to the requirements specified in table 13 of this subpart.

(d) The owner or operator of a delayed coking unit shall determine the coke drum vessel pressure or temperature, as applicable, on a 5-minute rolling average basis while the coke drum is vented to the closed blowdown system and shall use the last complete 5-minute rolling average pressure or temperature just prior to initiating steps to isolate the coke drum prior to venting, draining or deheading to demonstrate compliance with the requirements in paragraph (a) of this section. Pressure or temperature readings after initiating steps to isolate the coke drum from the closed blowdown system just prior to atmospheric venting, draining, or deheading the coke drum shall not be used in determining the average coke drum vessel pressure or temperature for the purpose of compliance with the requirements in paragraph (a) of this section.

(e) The owner or operator of a delayed coking unit using the “water overflow” method of coke cooling must hardpipe the overflow water or otherwise prevent exposure of the overflow water to the atmosphere when transferring the overflow water to the overflow water storage tank whenever the coke drum vessel temperature exceeds 220 degrees Fahrenheit. The overflow water storage tank may be an open or fixed-roof tank provided that a submerged fill pipe (pipe outlet below existing liquid level in the tank) is used to transfer overflow water to the tank. The owner or operator of a delayed coking unit using the “water overflow” method of coke cooling shall determine the coke drum vessel temperature as specified in paragraphs (c) and (d) of this section regardless of the compliance method used to demonstrate compliance with the requirements in paragraph (a) of this section.

(f) The owner or operator of a delayed coking unit may partially drain a coke drum prior to achieving the applicable limits in paragraph (a) of this section in order to double-quench a coke drum that did not cool adequately using the normal cooling process steps provided that the owner or operator meets the conditions in paragraphs (f)(1) and (2) of this section.

(1) The owner or operator shall install, operate, calibrate, and maintain a continuous parameter monitoring system to measure the drain water temperature at the bottom of the coke drum or in the drain line as near as practical to the coke drum according to the requirements specified in table 13 of this subpart.

(2) The owner or operator must maintain the drain water temperature below 210 degrees Fahrenheit during the partial drain associated with the double-quench event.

Start Amendment Part

30. Section 63.658 is added to read as follows:

End Amendment Part
Fenceline monitoring provisions.

(a) The owner or operator shall conduct sampling along the facility property boundary and analyze the samples in accordance with Methods 325A and 325B of appendix A of this part and paragraphs (b) through (k) of this section.

(b) The target analyte is benzene.

(c) The owner or operator shall determine passive monitor locations in accordance with Section 8.2 of Method 325A of appendix A of this part.

(1) As it pertains to this subpart, known sources of VOCs, as used in Section 8.2.1.3 in Method 325A of appendix A of this part for siting passive monitors means a wastewater treatment unit, process unit, or any emission source requiring control according to the requirements of this subpart, including marine vessel loading operations. For marine loading operations that are located offshore, one passive monitor should be sited on the shoreline adjacent to the dock.

(2) The owner or operator may collect one or more background samples if the owner or operator believes that an offsite upwind source or an onsite source excluded under § 63.640(g) may influence the sampler measurements. If the owner or operator elects to collect one or more background samples, the owner of operator must develop and submit a site-specific monitoring plan for approval according to the requirements in paragraph (i) of this section. Upon approval of the site-specific monitoring plan, the background sampler(s) should be operated co-currently with the routine samplers.

(3) The owner or operator shall collect at least one co-located duplicate sample for every 10 field samples per sampling period and at least two field blanks per sampling period, as described in Section 9.3 in Method 325A of appendix A of this part. The co-located duplicates may be collected at any one of the perimeter sampling locations.

(4) The owner or operator shall follow the procedure in Section 9.6 of Method 325B of appendix A of this part to determine the detection limit of benzene for each sampler used to collect samples, background samples (if the owner or operator elects to do so), co-located samples and blanks.

(d) The owner or operator shall collect and record meteorological data according to the applicable requirements in paragraphs (d)(1) through (3) of this section.

(1) If a near-field source correction is used as provided in paragraph (i)(1) of this section or if an alternative test method is used that provides time-resolved measurements, the owner or operator shall:

(i) Use an on-site meteorological station in accordance with Section 8.3 of Method 325A of appendix A of this part.

(ii) Collect and record hourly average meteorological data, including temperature, barometric pressure, wind speed and wind direction and calculate daily unit vector wind direction and daily sigma theta.

(2) For cases other than those specified in paragraph (d)(1) of this section, the owner or operator shall collect and record sampling period average temperature and barometric pressure using either an on-site meteorological station in accordance with Section 8.3 of Method 325A of appendix A of this part or, alternatively, using data from a United States Weather Service (USWS) meteorological station provided the USWS meteorological station is within 40 kilometers (25 miles) of the refinery.

(3) If an on-site meteorological station is used, the owner or operator shall follow the calibration and standardization procedures for meteorological measurements in EPA-454/B-08-002 (incorporated by reference—see § 63.14).

(e) The owner of operator shall use a sampling period and sampling frequency as specified in paragraphs (e)(1) through (3) of this section.

(1) Sampling period. A 14-day sampling period shall be used, unless a shorter sampling period is determined to be necessary under paragraph (g) or (i) of this section. A sampling period is defined as the period during which sampling tube is deployed at a specific sampling location with the diffusive Start Printed Page 75255sampling end cap in-place and does not include the time required to analyze the sample. For the purpose of this subpart, a 14-day sampling period may be no shorter than 13 calendar days and no longer than 15 calendar days, but the routine sampling period shall be 14 calendar days.

(2) Base sampling frequency. Except as provided in paragraph (e)(3) of this section, the frequency of sample collection shall be once each contiguous 14-day sampling period, such that the beginning of the next 14-day sampling period begins immediately upon the completion of the previous 14-day sampling period.

(3) Alternative sampling frequency for burden reduction. When an individual monitor consistently achieves results at or below 0.9 μg/m3, the owner or operator may elect to use the applicable minimum sampling frequency specified in paragraphs (e)(3)(i) through (v) of this section for that monitoring site. When calculating Δc for the monitoring period when using this alternative for burden reduction, zero shall be substituted for the sample result for the monitoring site for any period where a sample is not taken.

(i) If every sample at a monitoring site is at or below 0.9 μg/m3 for 2 years (52 consecutive samples), every other sampling period can be skipped for that monitoring site, i.e., sampling will occur approximately once per month.

(ii) If every sample at a monitoring site that is monitored at the frequency specified in paragraph (e)(3)(i) of this section is at or below 0.9 μg/m3 for 2 years (i.e., 26 consecutive “monthly” samples), five 14-day sampling periods can be skipped for that monitoring site following each period of sampling, i.e., sampling will occur approximately once per quarter.

(iii) If every sample at a monitoring site that is monitored at the frequency specified in paragraph (e)(3)(ii) of this section is at or below 0.9 μg/m3 for 2 years (i.e., 8 consecutive quarterly samples), twelve 14-day sampling periods can be skipped for that monitoring site following each period of sampling, i.e., sampling will occur twice a year.

(iv) If every sample at a monitoring site that is monitored at the frequency specified in paragraph (e)(3)(iii) of this section is at or below 0.9 μg/m3 for an 2 years (i.e., 4 consecutive semi-annual samples), only one sample per year is required for that monitoring site. For yearly sampling, samples shall occur at least 10 months but no more than 14 months apart.

(v) If at any time a sample for a monitoring site that is monitored at the frequency specified in paragraphs (e)(3)(i) through (iv) of this section returns a result that is above 0.9 μg/m3, the sampling site must return to the original sampling requirements of contiguous 14-day sampling periods with no skip periods for one quarter (six 14-day sampling periods). If every sample collected during this quarter is at or below 0.9 μg/m3, the owner or operator may revert back to the reduced monitoring schedule applicable for that monitoring site prior to the sample reading exceeding 0.9 μg/m3 If any sample collected during this quarter is above 0.9 μg/m3, that monitoring site must return to the original sampling requirements of contiguous 14-day sampling periods with no skip periods for a minimum of two years. The burden reduction requirements can be used again for that monitoring site once the requirements of paragraph (e)(3)(i) of this section are met again, i.e., after 52 contiguous 14-day samples with no results above 0.9 μg/m3 .

(f) Within 45 days of completion of each sampling period, the owner