National Emission Standards for Hazardous Air Pollutants: Miscellaneous Organic Chemical Manufacturing

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Part 63

[Docket ID No. OAR-2003-0121; FRL-7551-3]

RIN 2060-AE82

National Emission Standards for Hazardous Air Pollutants: Miscellaneous Organic Chemical Manufacturing

AGENCY:

Environmental Protection Agency (EPA).

ACTION:

Final rule.

SUMMARY:

This action promulgates national emission standards for hazardous air pollutants (NESHAP) for miscellaneous organic chemical manufacturing facilities. The final rule establishes emission limits and work practice standards for new and existing miscellaneous organic chemical manufacturing process units, wastewater treatment and conveyance systems, transfer operations, and associated ancillary equipment and implements section 112(d) of the Clean Air Act (CAA) by requiring all major sources to meet hazardous air pollutants (HAP) emission standards reflecting application of the maximum achievable control technology (MACT). The HAP emitted from miscellaneous organic chemical manufacturing facilities include toluene, methanol, xylene, hydrogen chloride, and methylene chloride. Exposure to these substances has been demonstrated to cause adverse health effects such as irritation of the lung, eye, and mucous membranes, effects on the central nervous system, and cancer. We do not have the type of current detailed data on each of the facilities and the people living around the facilities covered by the final rule for this source category that would be necessary to conduct an analysis to determine the actual population exposures to the HAP emitted from these facilities and the potential for resultant health effects. Therefore, we do not know the extent to which the adverse health effects described above occur in the populations surrounding these facilities. However, to the extent the adverse effects do occur, and the final rule reduces emissions, subsequent exposures will be reduced. The final rule will reduce HAP emissions by 16,800 tons per year for existing facilities that manufacture miscellaneous organic chemicals.

DATES:

This rule is effective November 10, 2003.

ADDRESSES:

Docket No. OAR-2003-0121 and A-96-04 are located at the Environmental Protection Agency, Office of Air & Radiation Docket & Information Center (6102T), 1301 Constitution Avenue, NW., Room B108, Washington, DC 20460.

FOR FURTHER INFORMATION CONTACT:

Mr. Randy McDonald, Organic Chemicals Group (C504-04), Emission Standards Division, U.S. EPA, Research Triangle Park, NC 27711; telephone number (919) 541-5402; electronic mail (e-mail) address

mcdonald.randy@epa.gov.

SUPPLEMENTARY INFORMATION:

Regulated Entities.

Categories and entities potentially regulated by this action include:

Category

NAICS

*

Examples of regulated entities

Industry

3251, 3252, 3253, 3254, 3255, 3256, and 3259, with several exceptions.

Producers of specialty organic chemicals, explosives, certain polymers and resins, and certain pesticide intermediates.

*

North American Industry Classification System.

This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely to be regulated by this action. To determine whether your facility is regulated by this action, you should examine the applicability criteria in § 63.2435 of the final rule. If you have any questions regarding the applicability of this action to a particular entity, consult the person listed in the preceding

FOR FURTHER INFORMATION CONTACT

section.

Docket.

We have established official public dockets for this action under Docket ID No. OAR-2003-0121 and A-96-04. The official public docket consists of the documents specifically referenced in this action, any public comments received, and other information related to this action. All items may not be listed under both docket numbers, so interested parties should inspect both docket numbers to ensure that they have received all materials relevant to the final rule. Although a part of the official docket, the public docket does not include confidential business information (CBI) or other information whose disclosure is restricted by statute. The official public docket is the collection of materials that is available for public viewing at the Air and Radiation Docket in the EPA Docket Center, (EPA/DC) EPA West, Room B102, 1301 Constitution Avenue, NW., Washington, DC. The EPA Docket Center Public Reading Room is open from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal holidays. The telephone number for the Reading Room is (202) 566-1744, and the telephone number for the Air Docket Center is (202) 566-1742. A reasonable fee may be charged for copying docket materials.

Electronic Access.

You may access this

Federal Register

document electronically through the EPA Internet under the

Federal Register

listings at

http://www.epa.gov/fedrgstr/.

An electronic version of the public docket also is available through EPA's electronic public docket and comment system, EPA Dockets. You may use EPA Dockets at

http://www.epa.gov/edocket/

to view public comments, access the index listing of the contents of the official public docket, and to access those documents in the public docket that are available electronically. Portions of the docket materials are available electronically through Docket ID No. OAR-2003-0121. Once in the system, select “search,” then key in the appropriate docket identification number. You may still access publicly available docket materials through the Docket ID No. A-96-04.

Worldwide Web (WWW).

In addition to being available in the docket, an electronic copy of the final rule will also be available on the WWW through the Technology Transfer Network (TTN). Following signature, a copy of the rule will be placed on the TTN's policy and guidance page for newly proposed or promulgated rules at

http://www.epa.gov/ttn/oarpg.

The TTN provides information and technology exchange in various areas of air pollution control. If more information regarding the TTN is needed, call the TTN HELP line at (919) 541-5384.

Judicial Review.

Under CAA section 307(b)(1) of the CAA, judicial review of the final NESHAP is available only by filing a petition for review in the U.S. Court of Appeals for the District of Columbia Circuit January 9, 2004. Under section 307(d)(7)(B) of the CAA, only an objection to a rule or procedure raised with reasonable specificity during the period for public comment can be raised during judicial review. Moreover, under CAA section 307(b)(2) of the CAA, the requirements

established by the final rule may not be challenged separately in civil or criminal proceedings brought to enforce these requirements.

Background Information Document.

The EPA proposed the NESHAP for miscellaneous organic chemical manufacturing on April 4, 2002 (67 FR 16154), and received 53 comment letters on the proposal. A background information document (BID) (“National Emission Standards for Hazardous Air Pollutants (NESHAP) for the Miscellaneous Organic Chemical Manufacturing Industry, Summary of Public Comments and Responses,”) containing EPA's responses to each public comment is available in Docket ID No. OAR-2003-0121.

Outline.

The information presented in this preamble is organized as follows:

I. Background

A. What is the source of authority for development of NESHAP?

B. What criteria are used in the development of NESHAP?

C. What is the history of the source categories?

D. What are the health effects associated with the pollutants emitted from miscellaneous organic chemical manufacturing?

E. How did we develop the final rule?

II. Summary of the Final Rule

A. What are the affected sources and emission points?

B. What are the emission limitations and work practice standards?

C. What are the testing and initial compliance requirements?

D. What are the continuous compliance requirements?

E. What are the notification, recordkeeping, and reporting requirements?

III. Summary of Environmental, Energy, and Economic Impacts

A. What are the air emission reduction impacts?

B. What are the cost impacts?

C. What are the economic impacts?

D. What are the non-air health, environmental, and energy impacts?

IV. Summary of Responses to Major Comments

A. What changes to applicability did the commenters suggest?

B. How did we change the compliance dates?

C. How did we develop the standards?

D. Standards for Process Vents

E. Storage Tank Standards

F. Standards for Wastewater Systems

G. Standards for Equipment Leaks

H. Standards for Transfer Racks

I. Pollution Prevention

J. Initial Compliance

K. Ongoing Compliance

L. Recordkeeping and Reporting

M. Startup, Shutdown, and Malfunction

N. Change Management

O. Overlapping Requirements

V. Statutory and Executive Order Reviews

A. Executive Order 12866: Regulatory Planning and Review

B. Paperwork Reduction Act

C. Regulatory Flexibility Act

D. Unfunded Mandates Reform Act

E. Executive Order 13132: Federalism

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

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

H. Executive Order 13211: Actions that Significantly Affect Energy Supply, Distribution, or Use

I. National Technology Transfer Advancement Act

J. Congressional Review Act

I. Background

A. What Is the Source of Authority for Development of NESHAP?

Section 112 of the CAA requires us to list categories and subcategories of major sources and some area sources of HAP and to establish NESHAP for the listed source categories and subcategories. A major source of HAP is a stationary source or group of stationary sources located within a contiguous area under common control that has the potential to emit greater than 9.1 megagrams per year (Mg/yr) (10 tons per year (tpy)) of any one HAP or 22.7 Mg/yr (25 tpy) of any combination of HAP.

B. What Criteria Are Used in the Development of NESHAP?

Section 112 of the CAA requires that we establish NESHAP for the control of HAP from both new and existing major sources. The CAA requires the NESHAP to reflect the maximum degree of reduction in emissions of HAP that is achievable, taking into consideration the cost of achieving the emissions reductions, any non-air quality health and environmental impacts, and energy requirements. This level of control is commonly referred to as MACT.

The MACT floor is the minimum control level allowed for NESHAP and is defined under section 112(d)(3) of the CAA. In essence, the MACT floor ensures that all major sources achieve the level of control already achieved by the better-controlled and lower-emitting sources in each source category or subcategory. For new sources, the MACT floor cannot be less stringent than the emission control that is achieved in practice by the best-controlled similar source. The MACT standards for existing sources can be less stringent than standards for new sources, but they cannot be less stringent than the average emission limitation achieved by the best-performing 12 percent of existing sources (or the best-performing five sources for categories or subcategories with fewer than 30 sources).

In developing MACT, we also consider control options that are more stringent than the floor. In considering whether to establish standards more stringent than the floor, we must consider cost, non-air quality health and environmental impacts, and energy requirements.

C. What Is the History of the Source Categories?

Section 112 of the CAA requires us to establish rules for categories of emission sources that emit HAP. On July 16, 1992, we published an initial list of 174 source categories to be regulated (57 FR 31576). The listing was our best attempt to identify major sources of HAP by manufacturing category. Following the publication of that listing, we published a schedule for the promulgation of emission standards for each of the 174 listed source categories. At the time the initial list was published, we recognized that we might have to revise the list from time to time as better information became available.

Based on information we collected in 1995, we realized that several of the original source categories on the list had similar process equipment, emission characteristics and applicable control technologies. Additionally, many of these source categories were on the same schedule for promulgation, by November 15, 2000. Therefore, we decided to combine a number of source categories from the original listing into one broad set of emission standards. Today's final rule reflects the subsumption of the following source categories into a new source category called Miscellaneous Organic Chemical Manufacturing: benzyltrimethylammonium chloride production, carbonyl sulfide production, chelating agents production, chlorinated paraffins production, ethylidene norbornene production, explosives production, hydrazine production, photographic chemicals production, phthalate plasticizers production, rubber chemicals production, symmetrical tetrachloropyridine production, OBPA/1,3-diisocyanate production, alkyd resins production, polyester resins production, polyvinyl alcohol production, polyvinyl acetate emulsions production, polyvinylbutyral production, polymerized vinylidene chloride production, polymethylmethacrylate production, maleic anhydride copolymers production, ammonium sulfate production—caprolactam by-product plants, and quaternary ammonium

compounds production. Along with these 22 source categories, the Miscellaneous Organic Chemical Manufacturing source category is also defined to include other organic chemical manufacturing processes which are not being covered by any other MACT standards.

Today's action establishes final standards for miscellaneous organic chemical manufacturing (40 CFR part 63, subpart FFFF).

D. What Are the Health Effects Associated With the Pollutants Emitted From Miscellaneous Organic Chemical Manufacturing?

The CAA was created, in part, “to protect and enhance the quality of the Nation's air resources so as to promote the public health and welfare and the productive capacity of the population” (see section 101(b) of the CAA). These NESHAP will protect public health by reducing emissions of HAP from miscellaneous organic chemical manufacturing facilities.

Miscellaneous organic chemical manufacturing facilities emit an estimated 21,900 Mg/yr (24,100 tpy) of organic and inorganic HAP. Organic HAP include toluene, methanol, xylene, methyl ethyl ketone, ethyl benzene, methyl isobutyl ketone, and vinyl acetate. Inorganic HAP emitted by this industry include hydrogen chloride (HCl) and some HAP metals in the form of particulate matter (PM). The final rule reduces HAP emissions from miscellaneous organic chemical manufacturing facilities by 68 percent. As a result of controlling these HAP, the final NESHAP will also reduce emissions of volatile organic compounds (VOC). A summary of the potential health effects caused by exposure to these pollutants is presented in the preamble to the proposed rule (67 FR 16154).

E. How Did We Develop the Final Rule?

We proposed the NESHAP for the miscellaneous organic chemical manufacturing source category on April 4, 2002 (67 FR 16154) and provided an 85-day comment period. We received a total of 55 comment letters. A copy of each of the comment letters is available in Docket No. OAR-2003-0121 or A-96-04.

The final rule reflects full consideration of all the comments we received on the proposed rule, as well as our reassessment of certain data in the rulemaking record. Major public comments on the proposed subpart FFFF, along with our responses to the comments, are summarized in section IV of this preamble. A detailed response to all comments is included in the Background Information Document for the promulgated standards (Docket No. OAR-2003-0121). Comments on the proposed miscellaneous coating manufacturing NESHAP will be summarized and discussed in the subpart HHHHH promulgation package.

II. Summary of the Final Rule

A. What Are the Affected Sources and Emission Points?

Emission points identified from miscellaneous organic chemical manufacturing production include process vents, storage tanks, equipment leaks, transfer operations, and wastewater collection and treatment systems. The affected source subject to this subpart is the facilitywide collection of miscellaneous organic chemical manufacturing process units (MCPU), wastewater treatment and conveyance systems, transfer operations, and associated ancillary equipment such as heat exchange systems that are located at a major source of HAP as defined in section 112(a) of the CAA. An MCPU includes a miscellaneous organic chemical manufacturing process, as defined in 40 CFR 63.2550, and must meet the following criteria: (1) It manufactures any material or family of materials described in 40 CFR 63.2435(b)(1); it processes, uses, or produces HAP described in 40 CFR 63.2435(b)(2); and, except for certain process vents that are part of a chemical manufacturing process unit, as identified in 40 CFR 63.100(j)(4), the MCPU is not part of an affected source under another subpart of 40 CFR part 63. The MCPU is defined according to the equipment used to make the subject material, and it includes storage tanks that are associated with the process.

New sources are created by reconstructing existing sources, constructing new “greenfield” facilities, or constructing an addition to an existing source that is a dedicated MCPU and has the potential to exceed 10 tpy of an individual HAP or 25 tpy of combined HAP. Reconfiguration of existing equipment does not constitute “construction.”

B. What Are the Emission Limits and Work Practice Standards?

The final rule regulates HAP emissions from miscellaneous organic chemical manufacturing facilities that are determined to be major sources. The standards apply to existing sources as well as new sources.

Process Vents

The final standards for existing batch and continuous process vents are set at a floor level of control and include requirements for organic and inorganic HAP. For batch process vents, the final standards require you to reduce uncontrolled organic HAP emissions from the sum of all batch process vents within the process by 98 percent if uncontrolled emissions exceed 4,540 kilograms per year (kg/yr) (10,000 pounds per year (lb/yr)). No control of vents is required for processes that are limited to uncontrolled emissions of 4,540 kg/yr (10,000 lb/yr) or less, as calculated on a rolling 365-day basis. A second control option for batch vents is to reduce the sum of all batch process vents within the process by 95 percent using recovery devices.

For continuous process vents, the final standards require control of vents determined to have a total resource effectiveness (TRE) index equal to or less than 1.9. The standards require you to reduce HAP emissions by at least 98 percent by weight if the TRE of the outlet gaseous stream after the last recovery device is less than 1.9, or to reduce the outlet total organic compound (TOC) concentration to 20 parts per million by volume (ppmv) or less. For continuous process vents, we reference the process vent standards contained in 40 CFR part 63, subpart SS.

For inorganic HAP, we set the standards based on the floor and made no distinction between batch and continuous streams. The standards for hydrogen halide and halogen HAP (

i.e.

, HCl, hydrogen fluoride (HF), and chlorine (C1

2

)) were determined to be 99 percent control of hydrogen halide and halogen HAP from the sum of all process vents in processes with uncontrolled hydrogen halide and halogen HAP emissions equal to or greater than 1,000 lb/yr. The final rule also requires control of hydrogen halide and halogen HAP emissions generated by the combustion control of halogenated streams, which are defined by a mass emission rate of halogen atoms contained in organic compounds of 0.45 kilograms per hour (kg/hr) or more. Specifically, hydrogen halide and halogen HAP emissions must be reduced after the combustion device by 99 percent, to no more than 0.45 kg/hr, or to no more than 20 ppmv. Alternatively, the halogen atom mass rate before the combustion device may be reduced to no more than 0.45 kg/hr or to no more than 20 ppmv. The MACT floor for PM HAP emissions from process vents at existing sources is no

emissions reduction, and we did not set a standard above the floor.

We defined the term “process” to include all equipment that collectively function to produce a material or family of materials that are covered by the source category. For batch process vents, we also established an equivalent mass cutoff of 200 lb/yr in the final rule that corresponds to the 50 ppmv concentration.

The new source standards for batch and continuous process vents follow the same formats as described above. However, some of the applicability triggers are more stringent. All batch process vents within a process for which the uncontrolled organic HAP emissions from batch process vents exceed 1,360 kg/yr (3,000 lb/yr) must be reduced by either 98 percent using a control device or 95 percent using a recovery device. All continuous process vents with a TRE of less than or equal to 5.0 must be controlled by 98 percent. For inorganic HAP, the standards for new sources are identical to the standards for existing sources. The new source standard for PM HAP emissions from process vents is 97 percent control for each process with uncontrolled PM HAP emissions greater than or equal to 400 lb/yr. Control requirements for halogenated streams are also the same as for existing sources.

Storage Tanks

The final rule requires existing sources to control emissions from storage tanks having capacities greater than or equal to 38 cubic meters (m

3

) (10,000 gallons (gal)) and storing material with a HAP partial pressure of greater than 6.9 kilopascals (kPa) (1.0 pound per square inch absolute (psia)). For new sources, the standards require control of storage tanks having capacities greater than or equal to 38 m

3

(10,000 gal) and storing material with a HAP partial pressure of greater than 0.7 kPa (0.1 psia). For both existing and new sources, the required control is to use a floating roof or to reduce the organic HAP emissions by 95 percent by weight or more. We also concluded in a revised analysis that for small storage tanks (capacities <10,000 gal), that there is a “no emission reduction” MACT floor, and we did not specify a standard because the total impacts of a more stringent regulatory alternative were found to be unreasonable. Additionally, we concluded that the new source MACT floor as proposed is appropriate (95 percent control of all tanks with capacities of 10,000 gal and storing material with a HAP partial pressure of 0.1 psia) for all tanks.

Wastewater

The final rule requires management and treatment of Group 1 wastewater streams and residuals removed from Group 1 wastewater streams to be consistent with the requirements contained in 40 CFR part 63, subpart G. For the purposes of 40 CFR part 63, subpart FFFF, the characteristics of Group 1 wastewater streams are defined with the following characteristics at the point of determination (POD):

• Process wastewater containing partially soluble HAP at an annual average concentration greater than 50 parts per million by weight (ppmw) and a combined total annual average concentration of soluble and partially soluble HAP of 10,000 ppmw or greater at any flowrate.

• Process wastewater containing partially soluble HAP at an annual average concentration greater than 50 ppmw and a combined total annual average concentration of soluble and partially soluble HAP of 1,000 ppmw or greater at an annual average flowrate of 1 liter per minute (lpm) or greater.

• Process wastewater containing partially soluble HAP at an annual average concentration of 50 ppmw or less and soluble HAP at an annual average concentration of 30,000 ppmw or greater and a total annual load of soluble HAP of 1 tpy or greater.

At new sources, the requirements are identical to those for existing sources, but the applicability triggers on individual streams are more stringent. In addition to controlling streams that meet the thresholds for existing sources, control is also required for the following streams at their POD:

• Process wastewater containing an annual average HAP concentration exceeding 10 ppmw of compounds listed in Table 8 of 40 CFR part 63, subpart G, with annual average flowrate greater than 0.02 lpm.

• Process wastewater containing partially soluble HAP at an annual average concentration of 50 ppmw or less and soluble HAP at an annual average concentration of 4,500 ppmw or greater and a total annual load of soluble HAP of 1 tpy or greater.

The final rule also requires compliance with the requirements of 40 CFR 63.105 for maintenance wastewater streams, and compliance with the requirements in 40 CFR 63.149 for liquid streams in open systems within an MCPU.

Transfer Racks and Ancillary Sources

The final standards for transfer racks, maintenance wastewater, and heat exchange systems are unchanged from the proposal, and they are identical to the requirements in the hazardous organic NESHAP (HON). For transfer operations, we are requiring the HON level of control for transfer racks that load greater than 0.65 million liters per year (l/yr) (0.17 million gallons per year (gal/yr)) of liquid products that contain organic HAP with a partial pressure of 10.3 kPa (1.5 psia). For each transfer rack that meets these thresholds, total organic HAP emissions must be reduced by 98 percent by weight or more, or the displaced vapors must be returned to the process or originating container. For maintenance wastewater, you must prepare a plan for minimizing emissions. For heat exchange systems, you must implement a monitoring program to detect leaks into the cooling water.

Equipment Leaks

For equipment leaks, the final rule requires implementation of a leak detection and repair (LDAR) program. For processes with no continuous process vents, you must implement the program in 40 CFR part 63, subpart TT. For processes with at least one continuous process vent, you must implement the program in 40 CFR part 63, subpart UU. Alternatively, you may elect to comply with the requirements in 40 CFR part 65, subpart F (

i.e.

, the Consolidated Federal Air Rule).

Pollution Prevention

The final rule also includes a pollution prevention alternative for existing sources that meets the control level of the MACT floor and may be implemented in lieu of the emission limitations and work practice standards described above. The pollution prevention alternative provides a way for facilities to comply with MACT by reducing overall consumption of HAP in their processes; therefore, it is not applicable for HAP that are generated in the process or for new sources. Specifically, you must demonstrate that the production-indexed consumption of HAP has decreased by at least 65 percent from a 3-year average baseline set no earlier than the 1994 through 1996 calendar years. The production-indexed consumption factor is expressed as the mass of HAP consumed, divided by the mass of product produced. The numerator in the factor is the total consumption of the HAP, which describes all the different areas where it can be consumed, either through losses to the environment, consumption in the process as a reactant, or otherwise destroyed.

Emissions Averaging Provisions

The final rule incorporates the emissions averaging provisions in 40 CFR part 63, subpart G (the HON), with some changes to accommodate batch process vents. For example, the final rule specifies that uncontrolled emissions from batch process vents are to be calculated using the procedures in 40 CFR part 63, subpart GGG, and performance testing must be conducted under worst case conditions, as defined in subpart GGG.

Alternative Standard

The final rule contains an alternative standard for process vents and storage tanks. When emissions are controlled using combustion control devices, the alternative standard requires control to an undiluted TOC concentration of 20 ppmv or less and an undiluted hydrogen halide and halogen HAP concentration of 20 ppmv or less. For noncombustion control devices, the TOC concentration and total hydrogen halide and halogen HAP concentration both must be reduced to 50 ppmv or less. Continuous monitoring of outlet TOC and total hydrogen halide and halogen HAP is required for compliance with this alternative standard.

C. What Are the Testing and Initial Compliance Requirements?

Process Vents

The final rule requires calculation of uncontrolled emissions as a first step in demonstrating compliance with the 98 percent or 95 percent reduction requirement for batch process vents. This initial calculation of uncontrolled emissions is not required if you choose to control process vents using the alternative standard or using specified combustion devices. For continuous process vents, the final rule requires calculation of the TRE index values using the procedures contained in the HON for continuous process vents.

To verify that the required reductions have been achieved, you must either test or use calculation methodologies, depending on the emission stream characteristics, control device, and the type of process vent. For each continuous process vent with a TRE less than or equal to 1.9, compliance with the percent reduction emission limitation must be verified through performance testing. For batch process vents, initial compliance demonstrations must be conducted in accordance with the requirements in the Pharmaceuticals Production NESHAP (40 CFR part 63, subpart GGG). Specifically, performance tests are required for control devices handling greater than 9.1 Mg/yr (10 tpy) of HAP, while either engineering assessments or performance tests are allowed for control devices with lower loads and for condensers. Performance tests must be conducted under worst-case conditions if the control device is used to control emissions from batch process vents.

Storage Tanks, Transfer Racks, and Wastewater

To demonstrate initial compliance with emission limits and work practice standards for storage tanks, transfer racks, and wastewater systems, the final rule allows you to either conduct performance tests or document compliance using engineering calculations. The initial compliance procedures are specified in 40 CFR part 63, subpart SS (National Emission Standards for Closed Vent Systems, Control Devices, Recovery Devices and Routing to a Fuel Gas System or a Process), subpart WW (National Emission Standards for Storage Vessels (Tanks—Control Level 2)), and subpart G (the HON), for control devices used to reduce emissions from storage tanks and transfer racks, storage tanks controlled with floating roofs, and wastewater sources, respectively.

D. What Are the Continuous Compliance Requirements?

The final rule requires monitoring, inspections, and calculations to demonstrate ongoing compliance. Typically, continuous monitoring (

i.e.

, every 15 minutes) of emissions or operating parameters is required when using a control device or wastewater treatment device. If operating parameters are monitored, operating limits must be established during the initial compliance demonstration. Periodic inspections are required for emission suppression equipment on waste management units and floating roofs on storage tanks and wastewater tanks. For processes that have Group 2 batch process vents (

i.e.

, total organic HAP emissions less than 10,000 lb/yr), you must track the number of batches produced to show that emissions remain below the Group 1 threshold.

Continuous monitoring requirements for control devices are specified in 40 CFR part 63, subpart SS, with some exceptions specified in the final rule. For example, the final rule requires that monitoring data during periods of startup, shutdown, and malfunction (SSM) be used in daily averages, whereas subpart SS excludes such data from averages. For batch process vents, you may request approval to set operating limits for individual or groups of emission episodes using the results of the performance test and applicable supplementary information. To use this approach, you must provide rationale for your selected operating limits in your precompliance report. As an alternative to daily averaging, the final rule also allows averaging over a batch or segment of a batch for control devices used to reduce emissions from batch process vents. For control devices that do not control more than 1 tpy of HAP emissions, only a daily verification that the control device is operating as designed is required.

Inspections for floating roofs must be conducted in accordance with 40 CFR part 63, subpart WW. All monitoring and inspection requirements for wastewater systems must be conducted in accordance with 40 CFR part 63, subpart G.

E. What Are the Notification, Recordkeeping, and Reporting Requirements?

Recordkeeping and reporting requirements are outlined in the General Provisions to part 63 (40 CFR part 63, subpart A), as well as the requirements in referenced subpart G (the HON), subpart SS (National Emission Standards for Closed Vent Systems, Control Devices, Recovery Devices and Routing to a Fuel Gas System or a Process), subpart TT (National Emission Standards for Equipment Leaks—Control Level 1), subpart UU (National Emission Standards for Equipment Leaks—Control Level 2 Standards), and subpart WW (National Emission Standards for Storage Vessels—Control Level 2). The sections of subpart A that apply to the final rule are designated in Table 12 to subpart FFFF of 40 CFR part 63. Additional recordkeeping and reporting requirements are specific to the final rule. For example, you are required to submit a precompliance report if you choose to comply using an alternative monitoring approach, use an engineering assessment to demonstrate compliance, or comply using a control device handling less than 1 tpy of HAP emissions. The final rule also references the SSM recordkeeping and reporting requirements contained in 40 CFR part 63, subpart SS. Under these provisions, SSM records are required only for events during which excess emissions occur or events when the startup, shutdown, and malfunction plan (SSMP) was not followed.

Consistent with the General Provisions, you must submit an initial notification, a notification of compliance status (NOCS) report, and compliance reports. The initial notification is required within 120 days of the effective date of 40 CFR part 63,

subpart FFFF. That brief notification serves to alert appropriate agencies (State agencies and EPA Regional Offices) of the existence of your affected source and puts them on notice for future compliance actions. The NOCS report, which is due 150 days after the compliance date of the NESHAP, is a comprehensive report that describes the affected source and the strategy being used to comply. The NOCS report is also an important aspect of the title V permitting strategy for sources subject to subpart FFFF. Compliance reports are required every 6 months.

III. Summary of Environmental, Energy, and Economic Impacts

A. What Are the Air Emission Reduction Impacts?

We estimate nationwide baseline HAP emissions from miscellaneous organic chemical manufacturing sources to be 21,900 Mg/yr (24,200 tpy). We project that the final rule will reduce HAP emissions by about 15,200 Mg/yr (16,800 tpy). Because many of the HAP emitted by miscellaneous organic chemical manufacturing facilities are also VOC, the NESHAP will also reduce VOC.

Combustion of fuels in combustion-based control devices and to generate electricity and steam will increase secondary emissions of carbon monoxide (CO), nitrogen oxides (NO

X

), sulfur dioxide (SO

2

), and PM less than 10 microns in diameter (PM

10

) by about 870 Mg/yr (960 tpy). These impacts were estimated assuming electricity is generated in coal-fired power plants, steam is produced in natural gas-fired industrial boilers, and natural gas is used as the auxiliary fuel in incinerators and flares.

B. What Are the Cost Impacts?

The cost impacts include the capital cost to install control devices and monitoring equipment, and include the annual costs involved in operating control devices and monitoring equipment, implementing work practices, and conducting performance tests. The annual cost impacts also include the cost savings generated by reducing the loss of product or solvent in the form of emissions. The total capital cost for existing sources is estimated to be $127 million, and the total annual cost for existing sources is estimated to be $75.1 million per year.

We estimate that in the first 3 years after the effective date of 40 CFR part 63, subpart FFFF, that the annual cost burden will average $3,150/yr per respondent for recordkeeping and reporting requirements. This estimate was based on having 251 sources. Most of these costs are for new and reconstructed sources that must be in compliance upon startup; other costs are for existing sources to prepare initial notifications and plans. In the fourth year after the effective date, existing facilities must begin to monitor and record operating parameters to comply with operating limits and prepare compliance reports, which will significantly increase the annual burden nationwide.

We expect that the actual compliance cost impacts of the NESHAP will be less than described above because of the potential to use common control devices, upgrade existing control devices, implement emissions averaging, or comply with the alternative standard. Because the effect of such practices is highly site-specific and data were unavailable to estimate how often the lower cost compliance practices could be utilized, we could not quantify the amount by which actual compliance costs might be reduced.

C. What Are the Economic Impacts?

The economic impact analysis for 40 CFR part 63, subpart FFFF, shows that the expected price increase for affected output is 0.5 percent, and the expected change in production of affected output is a reduction of 0.3 percent. One plant closure is expected out of the 207 facilities affected by the final rule. It should be noted that the baseline economic conditions of the facility predicted to close affect the closure estimate provided by the economic model, and that the facility predicted to close appears to have low profitability levels currently. Therefore, no adverse impact is expected to occur for those industries that produce miscellaneous organic chemicals affected by the NESHAP, such as soaps and cleaners, industrial organic chemicals, and agricultural chemicals.

D. What Are the Non-air Health, Environmental, and Energy Impacts?

With the assumption that overheads from steam stripping will be recoverable as material or fuel, no solid waste is expected to be generated from steam stripping of wastewater streams. No solid waste is expected to be generated from controls of other emission points. We expect the overall energy demand (

i.e.

, for auxiliary fuel in incinerators, electricity generation, and steam production) to increase by an estimated 6.1 million gigajoules per year (5.8 trillion British thermal units per year).

IV. Summary of Responses to Major Comments

A. What Changes to Applicability Did the Commenters Suggest?

Comment:

Several commenters suggested using only one industrial classification code, preferably the NAICS. The commenters also recommended increasing the specificity of the NAICS codes to six digits. As an alternative, one commenter suggested that the codes be scrapped and applicability be based simply on the manufacture of organic chemicals. Finally, the commenters requested exceptions for all codes that refer to inorganic chemical manufacturing processes.

Response:

We decided to retain both the SIC and NAICS codes in the final rule. Although SIC codes are being phased out, we decided to retain them because many industries still use these codes, and they were the basis for the selecting industries that received the section 114 information request. We rejected the suggestion to use six-digit NAICS codes because the list would be unnecessarily long; listing exclusions is much shorter. For the final rule, we also decided to list only the three-digit NAICS code for the chemical manufacturing subsector (325) rather than the seven four-digit codes for industry groups within this subsector because 40 CFR part 63, subpart FFFF, applies to all of the industry groups. However, there are selected manufacturing processes within both the SIC and NAICS industry groups for which the final rule is not applicable. These processes are exempted in the final rule by listing only the applicable six-digit NAICS code. Thus, a process described by a listed six-digit NAICS code is exempt even if it falls within an otherwise applicable SIC code. The exemptions cover all but three of the processes described by NAICS codes 325131, 325181, 325188, 325314, 325991, and 325992. The three processes within these otherwise exempt categories are hydrazine, reformulating plastics resins from recycled plastics products, and photographic chemicals.

Comment:

Two commenters stated that hydrazine manufacturing should not be subject to 40 CFR part 63, subpart FFFF, and the Hydrazine Manufacturing source category should be delisted because within the next few months, there will no longer be major sources within the source category; emissions from hydrazine manufacturing are too low to trigger controls; and hydrazine is an inorganic compound. If hydrazine is not removed from the miscellaneous

organic chemical manufacturing source category, one of the two commenters suggested that alternative testing methods are needed for hydrazine and that the definition of TOC should be changed to include hydrazine. The other commenter pointed out that the TRE equation is meaningless for hydrazine manufacturing plants because it requires sources to determine the hourly emission rate of organic HAP, and hydrazine and the raw materials used to produce hydrazine (

e.g.

, chlorine, caustic soda, and ammonia) are all inorganic.

Response:

Subpart FFFF covers the manufacture of hydrazine because it was one of the source categories subsumed, and the standards are based on a broad variety of chemical manufacturing processes. We developed separate standards for hydrogen halide and halogen emissions that require 99 percent control when uncontrolled hydrogen halide and halogen emissions exceed 1,000 lb/yr per process. However, hydrazine itself is also a HAP. Therefore, process vents containing hydrogen halide and halogen HAP would be subject to standards for hydrogen halide and halogen emissions. Hydrazine emissions from process vents would be subject to either the continuous process vent standards or the batch process vent standards. For the purposes of calculating the TRE for continuous process vents or mass emissions for comparison with the 10,000 lb/yr applicability threshold for batch process vents, the final rule specifies that hydrazine is to be considered an organic HAP.

Comment:

One commenter requested an exemption for photographic processing chemicals such as fixers, bleaches, and developers because HAP emissions from the processes are minimal, the equipment to manufacture these compounds are mixing vessels, and the processes do not appear to be included in the MACT floor. The commenter suggested that administrative burdens associated with the final rule, including calculating uncontrolled emissions, are not warranted.

Response:

We have not exempted manufacturing processes for photographic processing chemicals. The manufacturing equipment and emission characteristics, such as mixing vessels and their associated emissions from vapor displacement and evaporative losses, are represented by processes contained in the database.

Comment:

Many commenters supported the concept of treating process vents from the production of energetics as a separate class of emission streams subject to alternative requirements or a lesser degree of control for safety reasons. Several commenters provided specifics on the hazards posed by incineration-based controls and made recommendations that included providing definitions for energetics, waiving requirements for energetics or establishing a process where safe control technology can be identified on a case-by-case basis, and considering other control alternatives for compounds such as organic peroxides, powdered metals, metal catalysts, and highly flammable gases such as ethylene oxide and hydrogen. One of the commenters indicated that condensation and carbon adsorption are not effective on some compounds, such as nitroglycerine, which is unstable at low temperatures and cannot be safely controlled by carbon adsorption because it spontaneously combusts. The commenter supported a definition for energetics that includes “propellants, explosives, and pyrotechnics.” A second commenter suggested defining explosives as material included in the U.S. Department of Transportation hazardous materials tables (49 CFR 172.101) and listed as Hazard Class I hazardous material to include all Class I materials, or specifically materials in Divisions 1.1 through 1.6. The commenter indicated that using this approach, explosive manufacturers would know who they are because they are already shipping their materials as explosives; manufacturers who make materials that have some energetic properties, but are not shipped as explosives, would clearly be excluded. A third commenter requested that other compounds also be included in the subclass as explosives, particularly organic peroxides. The commenter cited EPA's rationale in providing a similar exclusion from control according to Resource Conservation and Recovery Act (RCRA), subpart CC for organic peroxide producers. A fourth commenter agreed and requested that EPA incorporate language already included at 40 CFR 264.1080(d) (duplicated at § 265.1080(d)) and 40 CFR 264.1089(i) (duplicated at § 265.1089(i)) in 40 CFR part 63, subpart FFFF. The commenter also suggested that other streams exist in the industry that may also meet this definition. For instance, reactive radioactive mixed waste wastewaters generated under the authority of the Atomic Energy Act and the Nuclear Waste Policy Act are exempted from closed conveyance requirements per 40 CFR 264.1080(b)(6). The U.S. Department of Energy requested this exemption because the radioactive mixed waste (RMW) containers “cannot be tightly sealed due to unacceptable pressure buildup of hydrogen gas to levels which can . . . create a potentially serious explosion hazard.” The commenters requested that EPA include language that allows facilities to document the hazardous nature of their wastewater streams and petition for exemption from the wastewater standards.

Response:

In the proposal, we recognized that the 98 percent control requirement for all process vents within affected processes would force incineration technology, and that this technology might not be appropriate for all process vent streams. Therefore, we also allowed 95 percent reduction of process vents if “recovery” control technology was employed to achieve required reductions. We envisioned at the time that the majority of this technology would be condensation. We solicited comments in the proposal on what commenters would consider achievable reductions from appropriate control technologies and how to define energetics. With the exception of the nitroglycerin example, we did not receive many comments that indicated that 95 percent control could not be achieved in most cases. Regarding organic peroxides, the add-on control requirement of RCRA, subpart CC, is 95 percent; therefore, EPA's earlier decision that indefinitely stayed requirements for producers of organic peroxides is consistent with the assumption that even 95 percent control cannot be achieved in these cases. Similarly, just as some reactive radioactive mixed wastewaters cannot be safely managed in closed systems, as one commenter suggested, there may be other situations that exist where sources may not be able to achieve the control efficiencies required by the final standards because of safety concerns. Based on the specific comments we received, we have concluded that it is appropriate to narrowly define a class of energetics and organic peroxides producers and allow, on a case-specific basis, a procedure to request an alternative compliance option. For these materials, the owner or operator must prepare and submit documentation in the precompliance report similar to the requirements in 40 CFR 264.1089(i) and 265.1089(i), explaining why an undue safety hazard would be created if the air emission controls specified in 40 CFR part 63, subpart FFFF, were installed on process vents, wastewater, and storage tanks containing energetics and organic peroxides, and describing what practices would be implemented to

minimize HAP emissions from energetics and organic peroxides manufacturing.

We did not broadly define energetics to encompass reactive or explosive conditions and the presence of highly flammable gases such as ethylene oxide and hydrogen. Based on past rules, we realize that combustion technology may not be appropriate in these cases, but other control technologies achieving relatively high control efficiencies are available and technically feasible.

Finally, the final rule includes a definition of “energetics” that is based on the definitions suggested by the commenters, and a definition of “organic peroxides” that is taken from 40 CFR 264.1080(d):

Energetics

means propellants, explosives, and pyrotechnics and include materials listed at 49 CFR 172.101 as Hazard Class I Hazardous Materials, Divisions 1.1 through 1.6.

Organic peroxides

means organic compounds containing the bivalent -o-o-structure which may be considered to be a structural derivative of hydrogen peroxide where one or both of the hydrogen atoms has been replaced by an organic radical.

Borrowing from language contained in 40 CFR 264.1080(d), only processes producing “organic peroxides as the predominant products manufactured by the process” and manufacturing “more than one functional family of organic peroxides or multiple organic peroxides within one functional family,” with one or more of these organic peroxides that “could potentially undergo self-accelerating thermal decomposition at or below ambient temperatures” would be eligible for identical treatment as energetics.

Comment:

One commenter asked for clarification that only solvent recovery operations operating at chemical manufacturing facilities are covered under 40 CFR part 63, subpart FFFF. The commenter also suggested adding a paragraph to the final rule to alert wastewater treatment operators that the final rule might apply to them.

Response:

We have not included the suggested language because solvent recovery operations are in fact covered by 40 CFR part 63, subpart FFFF, even if they are not located at a chemical manufacturing facility. However, offsite operations that are part of an affected source under another subpart of 40 CFR part 63, such as the Offsite Waste and Recovery Operations NESHAP (subpart DD), are not subject to subpart FFFF, as specified in § 63.2435(b)(3) of the final rule. Secondly, offsite treatment facilities are not affected sources but they may be required to treat wastewaters according to the provisions in subpart FFFF. Operators will be notified by respective dischargers of their obligation to treat in accordance with § 63.132(g)(1), as referenced in Table 7 to subpart FFFF of part 63.

Comment:

A number of commenters identified concerns with the “family of materials” concept and requested that EPA either eliminate it or make several changes. Several commenters suggested that the term is inconsistent with the floor determination and the information collection request (ICR), which allowed respondents to group materials but did not require it. One commenter suggested that the family of materials concept would discourage innovative or new and changed products due to constantly changing calculations and control requirements and increased administrative burden associated with tracking families. The commenter also stated that the concept is incompatible with flexible batch processes and could lead to division of products and equipment that are emitting to the same vent or groupings of products located in different buildings. The commenter suggested that grouping be conducted on shared process vents rather than families.

Four of the commenters suggested two key concepts to incorporate into the definition: the need to be able to group together processes with essentially identical emission sources and/or stream characteristics; and the recognition that, under some circumstances, functionality (

e.g.

, end use or product characteristics) may be an appropriate option in lieu of chemical composition. One of the commenters also suggested that we revise the list of examples because the proposed examples appear to be much broader categories of products than what other parts of the definition seem to allow and apply the concept only to batch process units in the same operational area.

One commenter stated that if EPA insisted on regulating equipment based on a “family of materials” concept, it should be limited to batch processes, and the emission threshold from the batch database should be recalculated. Finally, one of the commenters suggested that if EPA does not remove the family of materials concept, EPA must allow facilities to exclude from a family of materials grouping all individual products when the manufacture results in uncontrolled HAP emissions of less than 500 lb/yr for nondedicated batch operations or 100 lb/yr for dedicated batch operations.

Response:

The concept of “family of materials” is merely a logical grouping to describe materials that have very similar production and emission stream characteristics such that they can be considered as a single process. The final rule bases its control requirement on the sum of uncontrolled emissions within a process grouping. Only processes with uncontrolled organic HAP emissions of greater than 5 tpy are required to be controlled by 98 percent. Therefore, the definition of process determines what sources are included within a process grouping, which in turn affects applicable requirements and must be clearly specified in the final rule. In the proposed rule, we introduced the term family of materials to describe materials that vary only slightly in molecular structure, functional groups or other characteristics and are produced using procedures that result in essentially identical HAP emission streams from essentially identical emission sources. Our intent in requiring the grouping of these materials is to keep operators from artificially breaking them up into separate “processes” to avoid control requirements. We consider this concept to be important and have retained it in the final rule, with some modifications. Further, from our concept of “standard batch,” we would say that each family of materials has the same “standard batch.”

The standard batch concept was developed to allow owners and operators to identify and characterize emission events associated with a process. Once the emissions from each process are characterized, the owner or operator can merely count the number of batches conducted per year for each process to determine uncontrolled and controlled HAP emissions and compliance requirements. The standard batch concept provides a manageable way to document emissions; processes with the same identical standard batch should be considered the same process.

We agree with the commenters that our proposed definition did not adequately convey the concept of identical emission streams characteristics. We note that as long as groupings are also based on identical HAP emission characteristics, a grouping based on functionality is still compatible with the concept of having only one standard batch per process, which is a cornerstone of our compliance implementation strategy. Therefore, we have incorporated the suggested option so that the final definition requires identical emissions and either similar composition or functionality.

We reject the argument that the database is flawed because we did not require groupings when we surveyed

the industry. Although we did not require groupings, we encouraged respondents to group materials and provided guidance “that products that involve different HAP or different process equipment in case of dedicated processes should not be grouped together.” This language is basically consistent with the family of materials concept, and we note that many processes in our database appear to be material groupings. Therefore, we did not revise the MACT floor or proposed standards for batch vents. We also have not incorporated the suggestion to exempt “individual products for which the manufacture results in less than 500 lb/yr uncontrolled HAP emissions for nondedicated batch operations” because this language is unnecessary and inappropriate. Although the commenter may not have provided information on individual products with less than 500 lb/yr (

e.g.

, the commenter could have grouped families and emissions would be over 500 lb/yr and required to be reported), we expect that some respondents applied the 500 lb/yr reporting test on families of materials, based on the substantive number of groupings reported. Thus, there is no basis for exempting individual products for which the manufacture results in HAP emissions below the suggested 500 lb/yr threshold. Finally, because the final rule makes no distinction between “batch” and “continuous” processes, but rather on batch and continuous emissions, we do not restrict the concept to batch “processes.”

One commenter objected to the grouping of processes that are conducted in separate buildings and areas. Our proposed and final definition of process is not equipment specific. If the same product is manufactured in more than one set of equipment, emissions from all equipment must be considered when comparing to the 5 tpy mass applicability limit. The final rule is written this way because many manufacturers use nondedicated equipment to conduct their processes, and there is the potential that processing can be moved from one area to another easily to avoid regulation. Therefore, we do not restrict the family of materials grouping according to location.

Comment:

Many comments addressed various concepts in the definition of miscellaneous organic chemical manufacturing process. Several commenters considered the definition to be too lengthy and confusing. Some suggested removing statements that do not define the process. Others asked for clarification of various terms used within the definition such as “nondedicated,” “nondedicated solvent recovery,” “equipment,” and “product.” Two commenters stated that “product or isolated intermediate” should be changed to “miscellaneous organic chemical product.”

Several commenters objected to various requirements for nondedicated formulation operations. For example, some commenters opposed the requirement that all nondedicated formulation operations be considered a single process. They noted that the ICR did not request data for aggregated formulation operations and, thus, the MACT floor was based on separate formulation processes. Other commenters requested clarification of the term “contiguous area” as it relates to formulation operations. Several commenters found the exclusion for formulation operations that involve “mixing” to be confusing. They also requested that all formulation operations be exempt, not just those that are nondedicated and involve mixing, because none of these operations result in many emissions. One commenter expressed concern that estimating emissions for “hundreds” of small vents with minimal emissions for all the various formulated products would be burdensome, and control would be very costly. One commenter asked for an explanation of why nondedicated formulation operations (and nondedicated solvent recovery operations) are treated differently than other nondedicated operations.

Several commenters stated that cleaning operations should be part of the process only if they are routine and predictable because these are the only cleaning operations for which emissions can be estimated and included in a standard batch. Other commenters added that cleaning should not be part of the process if it involves opening of process vessels because there are no practical control methods for such events.

Response:

Except for nondedicated solvent recovery and formulation operations, miscellaneous organic chemical manufacturing processes are product based, meaning that all equipment used to manufacture a product is to be included in determining process vent control. We think this product-based approach is necessary because owners and operators may have the flexibility to manufacture the same product in more than one distinct area in a way that would avoid control under an equipment-based standard. However, in the case of solvent recovery operations such as distillation operations, defining a process by product would mean that each separately recovered product would be a separate process, which would result in fewer “processes” triggering the control requirement for the same equipment. The same is true for nondedicated formulation operations, where various finished materials could be formulated for shipment or as final product. Considering these two types of nondedicated operations as single processes also likely reflects the way in which these operations are managed and permitted. Further, we think respondents reported their data following this convention. Often, these operations will vary only in the type of HAP used. If the same HAP solvent is used for a variety of products, the emission stream characteristics per batch will essentially be the same. Therefore, considering a number of these operations as a single process actually simplifies recordkeeping. Note, however, that the final rule contains two key exemptions for batch process vents that may exempt many of the emission sources contributing to “minimal” emissions that the commenter is describing (

i.e.

, 50 ppmv or 200 lb/yr).

Although our proposed definition excluded “mixing,” we meant to exclude “mixing of coatings,” since this operation is to be covered by 40 CFR part 63, subpart HHHHH. When a product is blended or mixed with other materials in equipment that is dedicated to the manufacture of a single product, the mixing is included as part of the miscellaneous organic chemical manufacturing process.

We wanted to limit nondedicated solvent and formulation processes to related operations within the same area, which is the reason for the language regarding “contiguous operations.” However, we agree with one of the commenters that the term contiguous also conveys other meanings and, therefore, have revised the definition to refer to “each nondedicated solvent recovery (or formulation) operation.” The intent is to limit the process to operations located within a distinct operating area.

We agree that nonroutine cleaning operations involving vessel openings should not be considered as part of a process because they are difficult to characterize within a standard batch. These emissions would be attributed to startup and shutdown events, which are addressed separately in the final rule. In some instances, however, cleaning that is conducted within enclosed equipment between batches or between campaigns should be considered part of a process; these operations often consist of conducting solvent rinses through the equipment. Emissions from these

operations are similar to emissions during processing and the final rule's emission estimation procedures are suitable for these events. Therefore, they can be included in a standard batch for a given product and can be practically implemented.

Comment:

Some commenters are confused about how a process ends with the production of an “isolated intermediate” or product. One commenter stated that a process should end with the production of an isolated intermediate. Subsequent manufacturing operations using the intermediate should be considered part of a different process, and emissions from the operation should be managed separately from the emissions for the isolated intermediate process. A second commenter objected to the language in the proposal preamble that qualified the meaning of “stored” to be long-term storage, or that the material must be shipped offsite. The commenters stated that the term “storage” without qualification as to the length of storage or the purpose of storage is sufficient. A third commenter was concerned that the first sentence stated that an isolated intermediate is a “product,” but the second sentence stated that many “isolated intermediates” many be produced in the manufacture of a product; and that to be an isolated intermediate, a material must be stored, but the definition of storage tanks specifically excludes tanks storing isolated intermediates. The commenter stated that the definition needs to define the end of an MCPU where that MCPU produces a material that is not itself a commercial product. Two commenters wanted clarification that the term isolated intermediate refers to an organic material and suggested changing the term to “isolated organic intermediate'; and four commenters suggested that the term be limited to batch processes.

Response:

The concept of isolated intermediate is to identify a repeatable sequence of processing events that yield a material that is stable and subsequently stored before it undergoes further processing. The concept was introduced because many chemical processors have the capability to conduct intermediate processing steps in non-sequential order or even to conduct some processing steps offsite. Requiring an operator to consider all processing steps or campaigns that result in a final product may not yield a repeatable standard batch because of the possibility that not all steps would be conducted every time, or that some processing would depend on the availability of equipment and not be sequential; therefore, we limit the definition of process to the manufacture of an isolated intermediate. The concept that an isolated intermediate must be stored is important in that, if there is no “break” in the processing operations, there is no end of a process. We have, in the final rule, revised the definition of storage tank and process tank. Storage effectively occurs when material is stored and not processed over the course of a batch process. Therefore, we have eliminated the inconsistency regarding storage so that a storage tank can mark the end of a process if it is truly a storage tank and not a process tank, surge control vessel, or bottoms receiver. To limit confusion between listing the various vessel types that could be construed as process tanks, we eliminated the descriptive terms drums, totes, day tanks, and storage tanks.

We have not revised the definition to include the term “organic.” Our proposed and final definition clearly indicates that the material must be described by 40 CFR 63.2435(b). We have not limited the term to batch processes because the revised definitions of storage tank, surge control vessel, and bottoms receiver, make this distinction unnecessary. Additionally, we avoided basing any requirements on the differences between batch and continuous “processes” because processes can often contain both batch and continuous operations. Finally, we agree that the term isolated intermediate also is necessary to clarify that a material that is not itself a commercial product can be considered a product of a process.

B. How Did We Change the Compliance Dates?

Comment:

Several commenters stated that area sources that become major sources should have 3 years to comply. The commenters indicated that the proposed requirement to comply within 1 year deviates from 40 CFR 63.6(c)(5) of the General Provisions and requirements in other rules, and the proposal preamble provides no justification for the shorter time period. One commenter also noted that there is no difference in the level of effort needed to comply relative to that for a major source.

Response:

We agree to reference the General Provisions directly for compliance requirements for an area source that becomes a major source. We consider the 3-year period that the General Provisions allows for areas sources to come into compliance after becoming major sources to be adequate time. The proposed rule was published on April 4, 2002 and the anticipated compliance date is August 2006. Area sources becoming major sources after the effective date will have 4-plus years to become familiar with the applicability of 40 CFR part 63, subpart FFFF. An area source that becomes a major source between the effective date and the compliance date also has 3 years to come into compliance, except if it adds a new affected source (

e.g.

, a dedicated MCPU with the potential to emit 10 tpy of any one HAP or 25 tpy of combined HAP).

Comment:

One commenter operates an offsite treatment facility that could receive wastewater from affected sources under 40 CFR part 63, subpart FFFF. This commenter expressed concern with the requirement that existing sources be in compliance 3 years after the effective date of the final rule because they might not even receive affected wastewater until sometime after the compliance date. Therefore, the commenter suggested adding a new § 63.2445(f) to read as follows: “If you have an offsite treatment operation that receives affected wastewater or residue prior to the effective date of this subpart, then you must comply with the requirements for offsite treatment operations in this subpart no later than the date 3 years after the effective date of the subpart. If you have an offsite treatment operation that receives affected wastewater or residue after the effective date of this subpart, then you must comply with the requirements for offsite treatment operations in this subpart prior to receipt of an affected wastewater or residue.”

Response:

The proposed rule specified that affected wastewater (

i.e.

, “Group 1” wastewater in the final rule) that is sent offsite for treatment would be subject to § 63.132(g) of the HON. Those provisions require the offsite facility to comply with §§ 63.133 through 63.147 for any Group 1 wastewater that they receive. The commenter was concerned that an offsite treatment facility would be considered to be an existing source and might be unable to demonstrate initial compliance (

i.e.

, implement the design and operational requirements for waste management units and determine the performance of control devices and treatment processes) by the compliance date if the facility is not now receiving Group 1 wastewater and the operators are unaware whether the facility may receive such wastewater at some point in the future.

We did not add the suggested language because the proposed language is clear and already satisfies the commenter's concerns. Although an

offsite treatment facility will be required to meet the wastewater standards and associated compliance provisions if it accepts wastewater from an affected source, the offsite treatment facility is not an affected source. Therefore, the compliance date specified in § 63.2445 does not apply to an offsite treatment facility. The burden is also on the affected source operators to inform the offsite treatment facility of their intent, determine if the offsite facility is willing to handle the wastewater, and allow the offsite treatment facility time to achieve initial compliance before the first shipment.

C. How Did We Develop the Standards?

Comment:

One commenter stated that EPA unlawfully failed to set standards for all HAP emitted by the source category. According to the commenter, examples of HAP for which standards were not set include inorganic HAP such as HCl, HF, Cl

2

, potassium compounds; and organic HAP such as maleic and phthalic anhydrides. As support, the commenter referenced

National Lime Association

v.

EPA

, 233 F.3d 625 (D.C. Cir. 2000). Conversely, other commenters noted that the rule as proposed regulates both inorganic and organic HAP, but they suggested it should regulate inorganic HAP only when generated by the combustion of halogenated organic HAP. Some of these commenters stated that focusing on just organic HAP would be consistent with EPA's CAA section 114 data collection, the corresponding MACT floor analysis, and the approach used in other MACT standards. Two commenters noted that EPA recognized the inherent differences in the physical/chemical nature of inorganic HAP and the different technologies required for their control and specifically excluded inorganic HAP from the MACT floor analysis. The two commenters also stated that other standards, such as the HCl Production MACT, already adequately address inorganic HAP reduction requirements. Should EPA decide to regulate inorganic HAP, two commenters indicated that we should conduct additional MACT floor analyses and then propose separate standards for organic and inorganic HAP.

Response:

At proposal, our intent was that all types of gaseous HAP would be subject to the batch and continuous process vent standards. Similarly, the proposed storage tank standards would apply to all gaseous HAP, provided the maximum true vapor pressure for the total HAP in the storage tank exceeded the specified threshold. However, standards for the remaining emission source types are based on the compounds regulated by the HON, which covered organic HAP only. Standards for transfer operations and equipment leaks would also apply to any individual organic HAP or combination of organic HAP that meet a partial pressure threshold. Wastewater standards would apply only to those organic HAP that have the potential to volatilize from water based on modeling analyses conducted during development of the HON.

In response to the comments, we decided to develop a MACT floor and standards for hydrogen halide and halogen HAP (

i.e.,

HCl, HF, and Cl

2

) emissions from process vents that are separate from the analysis for organic HAP emissions. Based on data obtained in responses to the original ICR, this MACT floor was determined to be 99 percent control of hydrogen halide and halogen HAP from the sum of all vents in processes with uncontrolled hydrogen halide and halogen emissions equal to or greater than 1,000 lb/yr. We did not receive any information regarding source reduction techniques for hydrogen halide and halogen HAP. Generally, we would expect that these compounds are emitted as products of reaction, and there may be less opportunity for source reduction from these types of process vent emissions when compared to organic HAP. However, we structured the MACT floor to consider measures of reducing HAP emissions other than add-on control by basing the MACT floor on a percent reduction above some uncontrolled emission value. By default, implementing source reduction measures reduces “uncontrolled emissions.” The performance level of 99 percent is the highest control level achievable across the source category and is achieved by about 50 percent of the processes. The primary control devices used in the industry are packed-bed scrubbers. Control efficiencies for hydrogen halides (acid gases) and halogens depend on the solubility of the HAP in the scrubbing liquid, which in turn will vary with the processes that emit them. Control device vendors estimate that removal efficiencies for inorganic gases range from 95 to 99 percent (EPA-CICA Fact Sheet: Packed-Bed/Packed-Tower Scrubber). Therefore, although the reported control efficiencies for some processes were in excess of 99 percent, levels greater than 99 percent may not be uniformly achievable under all operating conditions. The best performing of these sources are those with the lowest uncontrolled emissions from the sum of all vents within the process. Therefore, we ranked all processes controlling hydrogen halide and halogen emissions to at least 99 percent by their uncontrolled emissions, from lowest to highest. For the best-performing 12 percent of processes, the median uncontrolled emissions rate is 1,000 lb/yr.

In setting the MACT floor for existing sources, we considered whether sources may be using emission reduction techniques other than technological controls for hydrogen halide and halogen HAP to determine whether such techniques might provide the basis for a floor. However, we did not receive any information regarding emission reduction techniques for these HAP in response to our ICR request that sources provide such information. Accordingly, we do not have information indicating that a sufficient percentage of sources are using emission reduction techniques for hydrogen halide and halogen HAP to enable us to set a MACT floor based on such techniques. Generally, we expect that because these HAP are emitted as products of reaction, there may be fewer opportunities to reduce process vent emissions of these HAP than there are opportunities to reduce emissions of organic HAP. (Organic HAP are frequently present in solvents, and solvent use can often be reduced; by contrast, reducing emissions of reaction products is more difficult because fundamental process changes are typically necessary.) Again, however, we do not have any information about the use of emission reduction techniques with which to support a floor determination.

Nevertheless, sources may use the pollution prevention option set out in 40 CFR part 63, subpart FFFF, to meet the 1,000 lb/yr cutoff for process vent emissions of hydrogen halide and halogen HAP and thereby comply with the relevant standards.

For new sources, the MACT floor is the same as for existing sources because reported control efficiencies in excess of 99 percent are not reliable. The final standards for hydrogen halide and halogen HAP emissions from process vents are also based on the MACT floor because the total impacts of a regulatory alternative were determined to be unreasonable.

Based on comments received, we decided to review our available data and develop a MACT floor for HAP metals in the form of PM, which acts as a surrogate for them. Our database shows six facilities emit PM HAP (specifically various metal compounds). One of the six facilities is controlling emissions from three processes with three different control devices, and the lowest control

efficiency is 97 percent. Since there are only six sources, the MACT floor for existing sources is based on the average performance of the top five sources. Since only one of the top five sources is implementing control, we determined the MACT floor is no emissions reduction. The final standard is based on the MACT floor because the total impacts of a regulatory alternative were determined to be unreasonable.

In setting the MACT floor, we considered whether some facilities may implement emission reduction measures to reduce PM HAP emissions, instead of using control technologies. We requested information on emission reduction measures in our section 114 information request. Of the approximately 40 different process changes reported, however, only one facility reported a process change that could be directly associated with PM emissions, which was described as “removing a hopper and vent.” Further, we do not know whether this emission reduction measure was effective in reducing PM HAP emissions. Therefore, because we lack information indicating that a sufficient number of process vents employ such measures to reduce emissions of PM HAP to set a floor, we were unable to set a MACT floor based on emission reduction measures.

The new source MACT floor for PM HAP emissions is based on the control achieved by the best-performing source. As noted above, the best-performing source is routing emission streams from three processes to three different control devices: a baghouse (fabric filter), a spray chamber and a rotoclone. The baghouse (fabric filter) achieves 97 percent control and this level is considered the emission control level that is achieved in practice by the best-controlled similar source, even though the other control devices report higher control efficiencies. Particulate control efficiencies are influenced by factors such as filtration velocity, particle loading, and particle characteristics, which in turn vary depending on the processes that emit them. Variations in stream characteristics make it difficult to conclude that the higher reported control efficiencies for the other control devices could be achieved in practice by all process vents that emit PM HAP. Based on ranking of the sources achieving 97 percent according to each source's lowest uncontrolled PM HAP emission level, the best-performing source is the lowest uncontrolled PM HAP emission level for any of the controlled processes (i.e., 400 lb/yr). Thus, the new source MACT floor for PM HAP emissions from process vents is 97 percent control for each process with uncontrolled PM HAP emissions greater than or equal to 400 lb/yr.

Comment:

One commenter stated that we unlawfully exempted emission points from regulation by establishing applicability cutoffs for both new and existing sources. The commenter stated that the rule must apply to all sources as required under the CAA, and, thus, cutoffs are illegal; and for wastewater, transfer operations, and equipment leaks, EPA illegally borrowed cutoffs and MACT floors from other standards. The commenter stated that standards must reflect the actual performance of the best-performing sources in the miscellaneous organic chemical manufacturing category. The commenter objected to 98 percent control levels for the process vent floors because reported control efficiencies for many process vents exceeded 98 percent. Finally, the commenter objected to the use of a work practice standard for equipment leak controls. Conversely, several other commenters suggested that the rule should specify additional thresholds below which a source would be considered to have “insignificant HAP emissions” and be exempt from control.

Response:

We disagree that every emission point at a major source must be required to reduce emissions. First, section 112(a) of the CAA defines “stationary source” (through reference to section 111(a)) as: “* * * any building, structure, facility, or installation which emits or may emit any air pollutant * * *.” (42 U.S.C. §§ 7412(a)(3) and 7411(a)(3)). The General Provisions for the MACT program define the term “affected source” as “the collection of equipment, activities, or both within a single contiguous area and under common control that is included in a section 112(c) source category or subcategory for which a section 112(d) standard or other relevant standard is established pursuant to section 112.” (40 CFR 63.2). Nothing in the definition of “stationary source” or in the regulatory definition of “affected source” states or implies that each emission point or volume of emissions must be subjected to control requirements in standards promulgated under section 112.

Further, even under the commenter's interpretation of “stationary source,” the Agency would still have discretion in regulating individual emission sources. Section 112(d)(1) allows the Administrator to “distinguish among classes, types, and sizes of sources within a category or subcategory in establishing such standards * * *.” We interpret this provision for the miscellaneous organic chemical manufacturing NESHAP, as we have for previous rules, as allowing emission limitations to be established for subcategories of sources based on size or volume of materials processed at the affected source. Under the discretion allowed by the CAA for the Agency to consider “sizes” of sources, we made the determination that certain small-capacity and low-use operations (e.g., “smaller” storage tanks) can be analyzed separately for purposes of identifying the MACT floor and determining whether beyond-the-floor requirements are reasonable. In addition, our MACT floor determinations for certain categories (e.g., process vents), which are set according to section 112(d)(3) of the CAA, reflect the performance levels and “cutoffs” of the best-performing sources for which we had information.

In general, our MACT floor determinations have focused on the best-performing sources in each source category, and they consider add-on control technologies as well as other practices that reduce emissions. As part of our information collection effort, we requested information on emission source reduction measures. We generally did not receive information indicating that, for the emission points covered by 40 CFR part 63, subpart FFFF, sources are currently reducing emissions by means other than control technologies in sufficient numbers to support a MACT floor based on source reduction measures. Accordingly, our standards include a performance level that represents the level achieved by the best control technology, and a cutoff that represents the lowest emission potential that is controlled by the best 12 percent of sources. Because the miscellaneous organic chemical manufacturing source category is broad in terms of the numbers and types of processing operations that are covered, one challenge was to develop a format by which all sources could be compared to each other to establish the best-performing sources. The performance level generally is of the format that can be applied to different types of control technology and processes and is generally consistent with existing rules. Thus, different types of control technology and emission levels resulting from existing rules are captured in our MACT floor analysis. The cutoff allows owners and operators that have reduced their emissions below a certain level using one or more methods, including process changes to reduce or eliminate pollution at the source, to comply without additional control. Both performance levels and cutoffs have been set to account for variations in emission stream

characteristics so that the standards can be applied consistently across the source category. We believe that this approach is consistent with the language of section 112(d)(3) that requires us to set the MACT floor based on the best-performing 12 percent of existing sources.

Aside from the MACT floor determinations, we also provided a pollution prevention compliance alternative to allow compliance with the standards by demonstrating a reduction in HAP usage per unit of product. This alternative enables owners and operators to comply using emission source reduction measures.

The above discussion notwithstanding, we decided to conduct a MACT floor analysis for storage tanks with capacities less than 10,000 gal. We concluded that the MACT floor for small tanks at existing sources is no emissions reduction because we have information from only eight sources that is not sufficient for setting a floor, and only one of the best-performing five sources is implementing controls. We did not specifically request information for tanks with capacities of less than 10,000 gal. Based on earlier EPA studies on the organic compound manufacturing industry (EPA-450/3-90-025), we estimate the actual number of storage tanks with capacities of less than 10,000 gal in our source category to be 30 percent of the total number of tanks, or approximately 500 tanks. The eight facilities reported information on 19 tanks, which is not enough information to set the floor. We also based the standard for existing storage tanks with capacities less than 10,000 gal on the MACT floor, because a regulatory alternative was determined to be unreasonable.

As for the new source MACT floor for storage tanks with capacities less than 10,000 gal, the best-performing source is controlling emissions from two small tanks by 98 percent with thermal incineration. One tank has a capacity of 9,800 gal and is storing material with a HAP partial pressure of 0.373 psia. The other tank has a capacity of 8,000 gal and is storing material with a HAP partial pressure of 0.574 psia. We consider the first tank to be more stringently controlled because partial pressure is the best indicator of emission potential and controlling a lower partial pressure is an indication of greater stringency. We compared this tank's characteristics to the new source MACT floor for larger tanks, which was set at 95 percent control for tanks with capacities of greater than 10,000 gal and storing materials with HAP partial pressures of 0.1 psia or higher. From an analysis of the tanks in our database, we concluded that the new source MACT floor for larger tanks is more stringent than a floor based on 98 percent reduction for tanks storing material with a HAP partial pressure greater than or equal to 0.4 psia. Therefore, we concluded the new source MACT floor as proposed to be appropriate for all tanks.

Finally, we do not have any information indicating that storage tanks with capacities less than 10,000 gal are reducing emissions through measures other than control technologies. Accordingly, we lacked sufficient information to set a floor based on such measures.

The MACT floors for organic HAP emissions from batch and continuous process vents are 98 percent control because this level has been shown to be uniformly achievable by well-designed and operated combustion devices. During development of the HON, the EPA recognized that thermal incineration may achieve greater than 98 percent reduction in some cases, but test data show that levels greater than 98 percent may not be uniformly achievable under all operating conditions (59 FR 19420, April 22, 1994). Similarities in processes and resulting emission streams in this industry with that of the HON source category processes allow us to draw the same conclusions with regard to achievable combustion control efficiencies. A review of the batch process vent database indicates that most processes with overall control of 98 percent or greater are controlled using thermal incinerators and flares (110 of 132 processes). We found the performance level for the MACT floor to be 98 percent because as much as 15 percent of the 731 processes in the database were controlled by thermal incineration. Similarly, a review of the continuous process vent database indicates that most processes with overall control of 98 percent or greater are controlled using thermal incinerators and flares (31 of 37 processes). We found the performance level for the MACT floor to be 98 percent because as much as 15 percent of the 202 processes in the database were controlled by thermal incineration. We did not use reported control efficiencies for scrubbers used to control organic HAP because we do not know the fate of pollutants captured in the scrubber effluent. If some of these pollutants are re-released to the air, then the reported control efficiencies are not valid.

For equipment leaks, we considered various formats for the standard and determined that a work practice standard based on an LDAR program is the most feasible. Unlike other emission sources, leaking components are not deliberate emission sources but rather result from mechanical limitations associated with process piping and machinery. A well-managed facility follows a preventive maintenance program to minimize leaks, but in all practicality cannot guarantee that no leaks will occur. Therefore, an emission standard for equipment leaks would not be feasible to enforce or prescribe. At the same time, our data indicate that the MACT floor for equipment leaks is an LDAR program. We also developed regulatory alternatives on the use of more effective LDAR programs. Finally, we note that enclosing components and venting to control is allowed, but except in limited cases, we expect the cost to be prohibitive.

Regarding the other commenters' suggestions, we note that the standards for all types of emission points contain cutoff values, consistent with our MACT floors, below which sources are exempt from control. We also concluded that our information did not allow us to develop a relationship between the various emission source types such that we could identify “insignificant” sources merely by the sum of actual emissions from process vents.

Comment:

One commenter stated that we failed to properly evaluate beyond-the-floor options. According to the commenter, in some cases, we stated that the MACT floor option was the most stringent option without identifying or evaluating other options (

e.g.

, LDAR for equipment leaks was assumed to be the most stringent option). In other cases, the commenter noted that the beyond-the-floor option is simply a lowering of the cutoff, and as discussed above for the MACT floor, the commenter stated that cutoffs should not be allowed. Also, where 98 percent control is the MACT floor, the proposed rule did not address why a beyond-the-floor option was not selected where data showed higher reductions are being achieved.

Response:

Our beyond-the-floor options reflect the most stringent performance levels that have been proven and can be applied consistently across our source category. It is true that in many cases, the beyond-the-floor option was based on simply lowering a cutoff, similar to the discussion above for new sources. This is consistent with the intent of section 112(d)(3) because better-performing sources have lower cutoffs.

For example, for batch process vents at existing sources, we evaluated the

feasibility of a regulatory alternative that would require 98 percent control of batch process vents in processes with uncontrolled organic HAP emissions between 5,000 and 10,000 lb/yr. We concluded that the total impacts of this alternative are unreasonable in light of the HAP emission reductions achieved. The incremental HAP reduction achieved by this above-the-floor alternative is 145 Mg/yr, and the incremental cost is about $15,000/Mg of HAP controlled. The incremental electricity consumption to operate exhaust gas fans is 5.1 million kilowatt hours per year (kwh/yr). The incremental steam consumption for steam-assist flares is 6 million lb/yr. The incremental fuel energy consumption to operate incinerators and flares and to generate electricity is 340 billion British thermal units (Btu) per year. Total CO, NO

X

and SO

2

emissions from combustion of the additional fuel is about 66 Mg/yr. There would be no wastewater or solid waste impacts.

We evaluated the feasibility of a regulatory alternative that would require 98 percent control of organic HAP emissions from continuous process vents that have a TRE index value between 1.9 and 5.0 at existing sources. We concluded that the total impacts of this alternative are unreasonable in light of the HAP emission reductions achieved. The incremental HAP reduction achieved by this above-the-floor alternative is about 400 Mg/yr, and the incremental cost is about $29,000/Mg of HAP controlled. The incremental electricity consumption to operate exhaust gas fans is 28 million kwh/yr. The incremental steam consumption for steam-assist flares is 83 million lb/yr. The incremental fuel energy consumption to operate incinerators and flares, generate steam, and generate electricity is 2.4 trillion Btu per year. Total CO, NO

X

, and SO

2

emissions from combustion of the additional fuel is 400 Mg/yr. There would be no wastewater or solid waste impacts.

We evaluated the feasibility of a regulatory alternative that would require 99 percent control of hydrogen halide and halogen emissions from processes with uncontrolled hydrogen halide and halogen emissions between 500 and 1,000 lb/yr at existing sources. We concluded that the total impacts of this alternative are unreasonable in light of the emission reductions achieved. The incremental HAP reduction achieved by this beyond-the-floor alternative is 1.0 Mg/yr, and the incremental cost is about $90,000/Mg of HAP controlled. The incremental electricity consumption to operate exhaust gas fans is 31,000 kwh/yr, and the incremental fuel energy consumption to generate the electricity is 300 million Btu per year. Total CO, NO

X

, and SO

2

emissions from the combustion of the additional fuel is 0.27 Mg/yr. The incremental wastewater generated from scrubber controls is 400,000 gal/yr.

We evaluated the feasibility of a regulatory alternative that would require 97 percent control of PM HAP emissions from process vents at existing sources if the uncontrolled PM HAP emissions exceeded 400 lb/yr. The only facility that meets the threshold for control is already controlled. Thus, we concluded that the total impacts of this alternative are unreasonable in light of the emission reductions achieved for a model facility that was based on the characteristics of the controlled facility. The incremental HAP reduction achieved by the above-the-floor alternative for the model facility is 4.3 Mg/yr, and the incremental cost is $68,000/Mg of HAP controlled. The incremental electricity consumption to operate exhaust gas fans is about 24,000 kwh/yr, and the incremental fuel energy consumption to generate the electricity is 230 million Btu per year. Total CO, NO

X

, and SO

2

emissions from combustion of the additional fuel is 0.2 Mg/yr. The quantity of solid waste generated could be greater if the owner or operator elects to use a dust collector that includes water sprays and discharges the collected dust in a slurry form.

For wastewater, we considered a regulatory alternative that would require HON-equivalent control of wastewater streams at existing sources that contain soluble HAP at concentrations between 15,000 ppmw and 30,000 ppmw or that contain partially soluble or mixed HAP at flowrates between 0.5 and 1.0 lpm. We concluded that the total impacts of this alternative are unreasonable in light of the emission reductions achieved. The incremental HAP reduction achieved by this above-the-floor alternative is 160 Mg/yr, and the incremental cost is about $8,500/Mg of HAP controlled. The incremental electricity consumption to operate pumps is 45,000 kwh/yr. The incremental steam consumption for steam strippers is 8.0 million lb/yr. The incremental fuel energy consumption to generate electricity and steam is 12 billion Btu per year. Total CO, NO

X

, and SO

2

emissions from the combustion of additional fuel to generate the electricity and steam is 1 Mg/yr. There may also be solid waste impacts if condensed steam and pollutants from the steam stripper cannot be reused. Small amounts of wastewater in the form of blowdown from the cooling water system for the condenser may also be generated.

For storage tanks at existing sources, we examined two regulatory alternatives. First, for storage tanks with capacities of at least 10,000 gal, we considered an alternative that would require an internal floating roof, external floating roof, or at least 95 percent reduction if the partial pressure of HAP stored in the tank is between 0.5 and 1.0 psia. We concluded that the total impacts of this alternative are unreasonable in light of the emission reductions achieved. The incremental HAP reduction achieved by this above-the-floor alternative is 30 Mg/yr, and the incremental cost is $19,000/Mg of HAP controlled. The incremental electricity and fuel consumption rates for storage tanks controlled with refrigerated condensers are 16,000 kwh/yr and 155 million Btu per year, respectively. Total CO, NO

x

, and SO

2

emissions from combustion of additional fuel is 0.13 Mg/yr, and there would be no wastewater or solid waste impacts. There also would be no environmental impacts or energy impacts for other storage tanks controlled with floating roofs. The second regulatory alternative that we considered would require 95 percent control for storage tanks with capacities less than 10,000 gal. We concluded that the total impacts of this alternative are unreasonable in light of the emission reductions achieved. On an average tank basis, the incremental HAP reduction achieved by this above-the-floor alternative is less than 0.5 Mg/yr, and the incremental cost would be on the order of $200,000/Mg of HAP controlled. The incremental electricity and fuel energy consumption rates for storage tanks controlled with refrigerated condensers are about 3,100 kwh/yr and 30.0 million Btu per year, respectively. Total CO, NO

x

, and SO

2

emissions from combustion of the additional fuel are about 0.025 Mg/yr. There would be no wastewater or solid waste impacts.

Regarding the specific situation described by the commenter in which we did not propose a more stringent option than the equipment leaks LDAR program, we are not aware of any option that could be applied consistently across the source category that would be effective. For example, enclosing all components and venting to control is allowed for process piping located inside of buildings or enclosures, but except in limited cases, we would expect the costs of such an option to be prohibitive. Furthermore, we have developed a revised MACT floor that consists of an LDAR program consistent with the requirements specified in 40 CFR part 63, subpart TT. We then

evaluated a regulatory alternative based on the more comprehensive LDAR program specified in 40 CFR part 63, subpart UU. We determined that this alternative is reasonable for processes that have at least one continuous process vent, but the costs are unreasonable for other processes. Because the regulatory alternative is implementation of a more stringent LDAR program, there are essentially no energy impacts or non-air quality health and environmental impacts associated with the regulatory alternative.

Finally, we did not evaluate a regulatory alternative for transfer operations because the floor is at the most stringent known requirements.

Comment:

Several commenters recommended referencing the Generic MACT at 40 CFR part 63, subparts SS, UU, and WW, in their entirety to specify all of the initial compliance, monitoring, recordkeeping, and reporting for process vents, transfer operations, storage tanks, closed-vent systems, and equipment leaks. Commenters also recommended referencing §§ 63.132 through 63.149 (and their associated recordkeeping and reporting requirements in §§ 63.151 and 63.152) of the HON for all of the requirements for process wastewater streams and liquid streams in open systems within MCPU, although one commenter recommended referencing the closed-vent system requirements in subpart SS instead of the comparable requirements in the HON. According to the commenters, the piecemeal referencing in the proposed rule was confusing and it expanded some requirements relative to the other subparts and missed some requirements in those subparts, which resulted in inconsistencies. A particular concern was that the proposed approach excluded the use of fuel gas systems and routing emission streams to a process.

Response:

To simplify and streamline the final rule and minimize the compliance burden, we decided to provide more complete references to the other rules with exceptions and additions only where needed. For example, we modified the hierarchy of compliance applicability in § 63.982(f) of the final rule; we overrode some of the initial compliance procedures in 40 CFR part 63, subpart SS, with the procedures in 40 CFR part 63, subpart GGG, for control devices used to control batch process vents; we retained the vapor balancing alternative in subpart GGG for storage tanks; we have specified different thresholds for Group 1 wastewater streams; we referenced 40 CFR part 63, subpart TT, rather than 40 CFR part 63, subpart UU, for equipment leaks in processes with no continuous process vents; we have specified periodic verification procedures rather than continuous monitoring for control devices with inlet HAP load less than 1 tpy; we have allowed averaging periods of operating blocks as well as operating days for batch operations; we retained the recordkeeping concept as proposed based on operating scenarios; we retained the precompliance report; and we have specified recordkeeping and reporting requirements for “deviations.”

Comment:

Two commenters requested that sources be allowed to follow the Synthetic Organic Chemical Manufacturing Industry (SOCMI) Consolidated Federal Air Rule (CAR) for continuous process vents, storage tanks, equipment leaks, and transfer operations so that a facility with HON and miscellaneous organic chemical manufacturing processes can comply with a consistent set of requirements. The commenters stated that the maximum use of standardized programs such as the CAR will provide the maximum flexibility to a facility nominally covered by multiple MACT rule requirements. One commenter stated that the American Chemistry Council, EPA, and many other stakeholders developed the CAR as the lowest burden, clearest, and most consistent set of requirements possible for the chemical industry using the HON model and understood that the CAR rule would be a model for future chemical industry rules.

Response:

The CAR was developed to provide a consolidated set of requirements applicable to storage vessels, process vents, transfer racks, and equipment leaks within the SOCMI. The CAR eliminates the overlapping requirements of numerous new source performance standards (NSPS) and NESHAP for the SOCMI that affect the same processes and equipment. These same requirements have also been codified in the Generic MACT at 40 CFR part 63, subparts SS, UU, and WW. Therefore, a facility with both HON and miscellaneous organic chemical manufacturing processes can essentially comply with the same set of requirements (

i.e.

, the HON processes would use the CAR, and the miscellaneous organic chemical manufacturing processes would follow the Generic MACT). We think that the reference in 40 CFR part 63, subpart FFFF, to the Generic MACT standards already provides the opportunity to consolidate across a facility, and except for equipment leaks, we do not see a benefit to cross-referencing another identical set of standards. We decided to specify in the final rule that you may elect to comply with equipment leak requirements in the CAR because the CAR is equivalent to or more stringent than the requirements in subpart FFFF.

D. Standards for Process Vents

Comment:

Numerous commenters suggested that we adopt the definition of “batch process vent” from the Polymer and Resins IV NESHAP. The commenters noted that this definition includes an applicability cutoff level of 500 lb/yr. Some of the commenters justified using this cutoff, or a similar mass-based limit, for the miscellaneous organic chemical manufacturing source category because 50 percent of batch process vents in the database emit less than 500 lb/yr and account for only 0.2 percent of total emissions, it would be more enforceable, and it would not be affected by dilution. One commenter suggested adding exemptions for vents used less than 300 hours per year (hr/yr) or emitting less than 1,000 lb/yr because batch processes often have hundreds of minor vents that are used only occasionally or have minimal emissions, and it would be prohibitively expensive to control these vents. Other commenters supported the 50 ppmv minimum control threshold but suggested that the concentration should be based on annual average vent HAP concentrations and emissions averaged over numerous emission episodes. They suggested using the existing annual average batch vent flowrate and annual average batch vent concentration equations found in § 63.1323 of subpart JJJ. Many commenters also requested exclusions for opening of process equipment for material addition, inspection, and for health and safety vents. The commenters indicated that the exclusion for opening equipment is supported by the EPA database because those facilities that reported fugitive emissions from batch operations did not control them. Furthermore, the commenters cited the precedent of the Offsite Waste and Recovery Operations MACT, which relieves operators of the requirement to vent emissions through a closed-vent system during sampling of tank contents and removal of sludges.

Response:

In general, we agree with the comments relating to adding a mass cutoff comparable to the 50 ppmv concentration limit. The use of a mass cutoff may be simpler than calculating the concentration in some situations, such as where emissions are known, but not the total volume of air in the system or the duration of an emission event (

e.g.

, emissions data developed from a mass balance). Being allowed to exclude vents based on emissions in addition to concentration may simplify the applicability determination procedure

in some cases. However, we determined that a lower cutoff than the 500 lb/yr is more appropriate. Of the approximately 1,500 process vents with concentrations less than 50 ppmv, the average (mean) mass emission rate is about 235 lb/yr. To establish a mass cutoff in the final rule that corresponds to the 50 ppmv concentration, we rounded this value to 200 lb/yr. If more than one emission episode contributes to a process vent, or if process vents within a process are piped or ducted together, the cutoff applies to the combined total.

The averaging period for determining the concentration was not specified in the proposed process vent definition; however, the proposed rule essentially required emissions to be calculated for each emissions episode. This means the averaging time for a concentration determination is over a single emission episode. The equations found in § 63.1323 of subpart JJJ would divide the total mass per batch by 8,760 hr/yr, which is not our intent. Therefore, we did not revise the definition to be consistent with the definition in subpart JJJ, but we have clarified that the concentration cutoff applies to emission episodes. The mass cutoff discussed above also applies to emission episodes. Thus, if a gas stream from any one episode meets the 50 ppmv cutoff, the process vent is affected.

Streams with less than 50 ppmv were specifically exempted from the vent definition to limit the introduction of dilution gases containing little to no HAP into emission streams as a means of diluting them and exempting them from control. Allowing averaging between streams of less than 50 ppmv with other emission episodes, as the commenters suggested, would effectively allow such dilution.

Therefore, we do not allow averaging across episodes to yield an average concentration for the purposes of determining whether a stream is affected.

We have decided to exempt some emissions releases that result from safety and hygiene practices because it is unlikely that these vents would reach the 50 ppmv concentration level. The exemption also will relieve owners and operators from the burden of demonstrating that they meet the concentration level. Specifically, the definition of “batch process vent” excludes flexible elephant trunk systems that draw ambient air (i.e, systems that are not ducted, piped, or otherwise connected to the unit operations) away from operators that could be exposed to fumes when vessels are opened.

We also note that although equipment openings without the presence of capture hoods and vents were not addressed specifically in the proposed rule, they would be subject to the provisions for certain liquid streams in open systems inside processes. Under these provisions, if the equipment meets the specified design and operating characteristics (

e.g.

, a tank with a capacity greater than 10,000 gal), then routine opening of the equipment would not be allowed. Also, opening events that are not routine and conducted as part of maintenance activities can be addressed in the facility's SSMP. Finally, regarding the commenter's request to exempt emergency vents, the SSMP can be used to address these events as well.

Comment:

One commenter stated that MACT floors must be based on an average of existing regulatory limits, not on actual emissions data. According to the commenter, using actual emissions data violates section 302(k) of the CAA.

Response:

We disagree with this comment. Nothing in section 302(k) of the CAA prohibits the use of actual emission data in setting MACT floors. The MACT floor was developed using all available information. The evaluation included, but was not limited to, information about existing regulatory limits. We also collected information from sources in the industry and States during 1997 that was the source of actual emissions data. A CAA section 114 ICR was sent to 194 facilities in the spring of 1997. The facilities which received the ICR were identified from EPA's 1993 toxic release inventory database which included information on facilities in SIC codes 282, 284, 286, 287, 289, or 386. Information on continuous processes came from emissions and permit databases from the following States: Texas, Louisiana, North Carolina, Illinois, Missouri, California, and New Jersey.

Comment:

One commenter stated that the methodology utilized in determining the MACT floor for batch processes fails to accurately reflect the processes of the adhesives and coatings industry because, to the best of the commenter's knowledge, none of the 11 companies that own all of the sources in the MACT floor analysis makes adhesives. Other commenters noted that air-bearing vents, which cannot safely go to a flare or incinerator, should be considered separately from non-air-bearing vents because it is much harder to obtain high control efficiencies without using a combustion device. One commenter requested that spray dryer operations and post-spray dryer solids handling systems be excluded from the MCPU because the commenter is unaware of any facilities currently controlling such emission streams, it would be very costly to control such streams, and spray dryers are not specifically discussed in the MACT floor documentation.

Response:

In the development of our database, we solicited information from a number of industries thought to be representative of this source category. Processing operations such as the synthesis of resins or polymers that are used as bases for adhesives are expected to result in emission sources with characteristics similar to other specialty chemical processes in this source category. Therefore, we expect the emission stream characteristics of the adhesives industry to have similar characteristics as those of other industries covered by this source category and have, therefore, not developed a separate category for this industry.

We disagree with the suggestion to consider air-bearing vents separately from other vents in the development of the MACT floor. Roughly half of the process vents in our batch process vents database have concentrations of 50 ppmv or less. These streams, which include many air-bearing streams from dryers and other sources, were exempt from the definition of process vent in the proposed rule because we recognize that it is not technically or economically feasible to require control of these streams. For process vents containing greater than 50 ppmv HAP, the final rule also allows compliance by meeting an outlet concentration limit as an alternative to a percent reduction standard. This alternative is provided to assist owners and operators in complying with the standards for low concentration streams.

Our process vent database includes spray dryers at two facilities. It also includes over 25 records for “dryers” at other facilities, some of which may pertain to spray dryers. As noted above, our database also includes air-bearing vents, which have characteristics likely to resemble those of emission streams from spray dryers. Therefore, we determined that these emission sources are represented in our database, and that the MACT floor properly sets the level of control for these vent streams.

Comment:

Various commenters indicated the MACT floor for continuous process vents should be recalculated because of the following perceived problems with the database and analysis: a process vent at the BP Chemicals, Wood River plant (formerly Amoco Petroleum Additives), should be

removed from the database because no such vent ever existed; the database includes errors such as emission points that are not continuous process vents; the analysis was conducted on a facility basis rather than a vent basis; flowrate assumptions are too high; the sample population is too small; and the database is skewed by a disproportionate number of sources in ozone nonattainment areas.

Response:

To develop the MACT floor for continuous process vents, we relied on available information from State permitting databases. To the best of our knowledge, these data reflect the sources that will be subject to requirements for continuous process vents. Although many of these facilities are in ozone nonattainment areas, the commenters have provided no evidence that this is not representative of sources that will be subject to the final standards. We disagree with the commenter's assertion that the analysis should be conducted on a vent basis rather than a facility basis. Our analysis was designed to identify what level of emissions would not be controlled by facilities that would be considered the best-controlled sources in the industry. That level of emissions, characterized by the vent with the highest TRE index value below which all other vents were controlled, became the TRE cutoff value for the facility. We consider the analysis valid and in keeping with the statutory MACT requirements of CAA section 112(d)(3). Regarding our assumption of flowrate in cases where no flowrate data were available, we note that our assumed flowrate is the average of the available flowrates. In response to the objection that the sample population was too small, we note that it is derived from many of the major chemical producing States, and we estimate that it represents about half of the affected sources with continuous process vents. However, we agree that the vent at the BP facility should be excluded because it never existed. Without this vent, the TRE threshold for control of continuous process vents is now 1.9 rather than 2.6.

Comment:

Several commenters requested that control devices installed prior to April 4, 2002, be grandfathered from the 98 percent reduction requirement if they achieve 90 or 95 percent control of organic HAP. The commenters noted that many companies may be faced with abandoning existing control devices and installing new devices to get only an incremental reduction in HAP emissions, and they noted that other MACT rules (

e.g.

, pharmaceuticals and pesticide active ingredients) allow the continued use of existing controls that have a lower efficiency than the standard. One commenter also indicated that regenerative thermal oxidizers (RTO) have difficulty in achieving 98 percent control.

Response:

Since the final rule provides less stringent control requirements for control devices that can recover materials for reuse, we assume that the bulk of the concern related to control devices is for incinerators that will not meet 98 percent. Devices such as RTO are typically installed to control high air flow, low concentration streams. Therefore, while this type of device may not meet the 98 percent control requirement, the final rule also allows sources to demonstrate compliance with an outlet concentration limit, which may be achievable by an RTO when the uncontrolled HAP concentration in the vent stream is low. We note also that the batch vent requirements contain options for monitoring parameters in lieu of correcting outlet concentration for 3 percent oxygen (O

2

). Finally, the final rule includes a provision that may enable some sources to group nondedicated processing equipment together and comply only with the requirements in the rule that apply to the primary product made in the equipment.

E. Storage Tank Standards

Comment:

Several commenters indicated that the proposed definition of “storage tank” is inconsistent with the ICR, MACT floor calculations for both storage tanks and process vents, EPA applicability determination documents, and other MACT rules; likely to lead to compliance confusion; and likely to force replacement of many existing floating roof tank controls at huge costs for negligible benefits. Many of the commenters recommended revising the storage tank definition to match the actual assignment of tanks in the storage tanks database and recalculating the MACT floor.

Response:

The definition of “storage tank” in the proposed rule was based on the treatment of process tanks and storage tanks in the pharmaceuticals industry, a predominantly batch industry. The basis for only considering raw material feedstock tanks as true storage tanks was that the product tanks were seldom of the size at which the storage tank capacity cutoffs were set in many rules, and that a predominant number of tanks were used within processes as feed tanks from one unit operation to another. As such, emission events from these tanks usually would be calculated based on displacement resulting from filling the tank, usually on a per batch basis, and included in the operating scenario for an entire process. Emissions, therefore, were tied to the number of batches produced, as the material was transferred into and out of these tanks during each batch. We consider these tanks to be true process tanks and expect that the batch processors in the miscellaneous organic chemical manufacturing industry would agree with this treatment.

We recognize, however, that this industry contains significant numbers of continuous processors. We also recognize that this industry is more varied than the pharmaceuticals industry and that there are more tanks that are of a size and function that would be treated as storage tanks in other rules. For example, product rundown tanks and product storage tanks are not based on the number of batches, and material remains in the tank or is “stored” on a fairly continuous basis. The tanks are not filled and emptied during batch operations. These tanks are storage tanks and are recognized as such in the final rule.

We agree that the responses to the section 114 ICR would be based on the HON and NSPS definitions, and we have revised the storage tank definition to be consistent with the HON and NSPS. Although defined separately, the HON treats surge control vessels and bottoms receivers, types of tanks found in predominantly continuous processes that function in receiving material between continuous operations, exactly like storage tanks. We kept these terms and requirements in the final rule, but revised the definition of surge control vessel to be consistent with the definition of continuous process vent (

i.e.

, surge control vessels must precede continuous reactors or distillation operations). We also added a definition for “process tank” to clarify which tanks we would consider as part of the batch process vent standards. The changes do not affect MACT floors; they only change applicability under the storage tank standards or under the batch process vent standards.

F. Standards for Wastewater Systems

Comment:

Numerous commenters urged elimination of the requirement to enclose sewers and tankage for conveyance to treatment of wastewater streams with primarily soluble HAP. The commenters stated that soluble HAP do not volatilize significantly from wastewater streams upstream of biological treatment, but the cost to suppress emissions would be significant. Some commenters suggested exempting from control those

wastewater streams that contain soluble HAP unless at least 5 percent of the total soluble HAP is emitted from the waste management units. Commenters were particularly concerned about this issue for the final rule because much more methanol is present in miscellaneous organic chemical manufacturing processes than in processes subject to the HON, particularly because 40 CFR part 63, subpart FFFF, applies to HAP that are used as solvents. Another commenter claimed the available data do not support a floor of HON-equivalent control for streams with HAP concentrations less than 10,000 ppmw.

Response:

We considered the request for separate treatment of wastewater containing soluble HAP. We began by reviewing the miscellaneous organic chemical manufacturing wastewater database, and we determined that wastewater containing soluble HAP compounds are generally managed separately from wastewater containing partially soluble HAP compounds in this industry. This separate treatment by the industry justifies the evaluation of separate floors in accordance with the commenter's requests. For the 60 facilities in the miscellaneous organic chemical manufacturing wastewater database, there are a total of 364 records (streams), excluding streams with HAP that are not listed on Table 9 to subpart G of 40 CFR part 63 (the HON), HAP concentrations less than 1,000 ppmw, and HAP concentrations greater than or equal to 1,000,000 ppmw. Of this total, 192 of the streams contain partially soluble or a mixture of partially soluble and soluble HAP, and 172 of the streams contain only soluble HAP. Only 26 streams contain a mixture of soluble and partially soluble HAP.

When we reevaluated the floors separately, we found that for the partially soluble and mixed streams, data show that considerably more than 12 percent of the streams that meet either of the HON cutoff criteria also received treatment consistent with HON treatment requirements (

i.e.

, the best-performing miscellaneous organic chemical manufacturing sources are those that implement HON-equivalent procedures). Of the 53 streams with flowrates greater than 1 lpm and concentrations of partially soluble or mixed streams less than 10,000 ppmw, nine are managed and treated according to HON levels. Therefore, we revised the flow cutoff in the MACT floor from 10 lpm to 1 lpm for streams with concentrations greater than or equal to 1,000 ppmw and less than 10,000 ppmw; the other cutoffs of greater than or equal to 10,000 ppmw at any flowrate for partially soluble and mixed streams are unchanged. Another 42 streams had flowrates between 0.1 and 1.0, but only one was controlled. Therefore, we concluded that a sufficient number of streams below the cutoffs were not controlled to support a no emissions reduction floor determination.

We also identified a MACT floor for the 172 wastewater streams at 33 facilities that contain only soluble HAP. We ranked the 33 facilities based on the lowest concentration and flowrate of a wastewater stream that was managed and treated according to the HON requirements. The top five facilities were found to manage and treat all their soluble HAP containing wastewater consistent with the requirements in the HON. The median of the lowest concentrations in wastewater streams at these five facilities was found to be 30,000 ppmw. The lowest soluble HAP load for any stream at the five MACT facilities was 1,663 lb/yr (which we rounded to 1 tpy). Therefore, we determined that the MACT floor consists of the management and treatment requirements in the HON for wastewater streams containing at least 30,000 ppmw of soluble HAP and at least 1 tpy of soluble HAP. Wastewater streams with soluble HAP above these concentration and load cutoffs are considered Group 1 wastewater streams in the final rule. We also evaluated a beyond-the-floor alternative based on controlling streams with mixed HAP at flowrates greater than 0.5 lpm and streams that contain soluble HAP at concentrations greater than 15,000 ppmw. The total impacts of this alternative were determined to be unreasonable. Therefore, we set the standard for existing sources at the MACT floor.

For new sources, we determined the MACT floor for wastewater containing soluble HAP to be a concentration of 4,500 ppmw at the 1 tpy load. The 4,500 ppmw corresponds to the lowest concentration of a stream containing only soluble compounds that was managed and treated in accordance with the HON. The 1 tpy load cutoff was not lowered in going from the existing source standard to the new source standard because this level already represents the lowest load cutoff of any stream at the five MACT facilities and, therefore, represents the performance of the best-controlled similar source.

Comment:

Two commenters indicated the proposed rule lacks criteria for evaluating affected wastewater streams from batch process units and specialty chemicals manufacture. One of the commenters suggested revising the rule so that the emission thresholds for wastewater are determined over a representative batch cycle. To accomplish this, the commenter suggested that the following definitions be added to the rule:

• “Annual average” means the average over a designated 12-month period of actual or anticipated operation of the MCPU generating wastewater, except for units that are flexible operations or part of flexible operations. For flexible operation units, “annual average” means the average for a standard batch that is representative of the designated 12-month period of actual or anticipated operation of the MCPU generating wastewater.

• “Standard batch” means a batch process operated within a range of average or typical operating conditions that are documented in an operating scenario. Emissions from a “standard batch” are based on the production activity or product that result in the highest mass of HAP in the wastewaters generated by the process equipment during the batch cycle.

The second commenter noted that the proposed rule refers to § 63.144(c) for establishing the annual average flowrate for wastewater streams (

i.e.

, total wastewater volume divided by 525,600 minutes in a year). The commenter supported this for continuous process units, but recommended that the rule use criteria from 40 CFR part 63, subpart GGG, for batch process units since the wastewater streams from batch operations may only be operational a few months per year.

Response:

The format for applicability is annual average flowrate based on the potential maximum amount of operating hours per year (

i.e.

, 8,760). Although the procedure was developed for continuous processes, it can be applied to batch processes. When multiplied out, the total flow of wastewater equivalent to 1 lpm and 8,760 hr/yr equals 0.14 million gal/yr (530 m 3 /yr). We recognize that the proposed rule did not contain guidance on how to interpret annual average for batch processes although our definition of wastewater stream described a single wastewater stream as being discarded from an MCPU through a single POD. Our intent with this language was to include all wastewater streams from single processes that were discharged through a single POD as one single wastewater stream. In the HON, annual average concentration is the total mass of compounds listed in Table 9 to subpart G of 40 CFR part 63 that are in the wastewater stream during the designated 12-month period divided by the total mass of the wastewater stream during the 12-month period. There is no separate consideration in the HON for

multipurpose batch operations or POD that serve numerous processes because the equipment is part of a flexible operation.

For 40 CFR part 63, subpart FFFF, however, we based the MACT floor on data from wastewater streams that were developed based on our proposed definition of wastewater. Therefore, the definition of annual average is based on wastewater streams from a POD from a single MCPU. For flexible operations (

e.g.

, multipurpose equipment not dedicated to any single process), we have incorporated the concept of a family of materials that considers as a single product the manufacturing processes of multiple materials that are related. Additionally, we consider “nondedicated solvent recovery operations” as a single process. Therefore, in these two circumstances, the definition of wastewater stream should be based on the total mass and flow out of the POD from the sum of all operations considered within the family of materials or within the recovery process. In all other cases, the flow and concentration of HAP should be based on the total flow of wastewater and mass of HAP from all batches of a single process.

The final rule requires a manufacturer of a family of materials in flexible operation units to determine the annual average using a procedure consistent with that described by the commenter. Specifically, the worst-case product would determine the standard batch, and the total flow of wastewater would be based on the total flow of wastewater generated by all batches manufactured in any 12-month period. However, if materials manufactured in the flexible operations fell among more than one product not considered part of a family of materials, we would consider these separate processes, and the annual average concentration and flow would be limited to the characteristics of each process.

Comment:

Consistent with comments on the definition of the miscellaneous organic chemical manufacturing process, one commenter suggested revising the definition of “maintenance wastewater” to clarify that wastewater from routine cleaning operations occurring within a batch process is not considered maintenance wastewater. Another commenter noted that some cleaning operations are performed for equipment preparation and to remove inorganic scale from the equipment on an occasional, though somewhat regular basis. The commenter observed that these operations are performed between batches, though not between every batch or even between batches of different grades. They are performed when maintenance is needed or plugging is evident. The commenter asked for clarification that the types of cleaning operations that do not generate maintenance wastewater are those performed between batches for the purposes of changing grades and not those done to prepare equipment for maintenance or to remove inorganic foulants.

Response:

We agree with the commenters regarding the need to exclude non-routine cleaning operations from other process wastewater streams and have included them in the definition of “maintenance wastewater.” This issue is analogous to the issue of including vents from routine cleaning operations as process vents and covering other types of events under the SSM provisions.

Comment:

One commenter requested an exemption from the offsite certification requirement in 40 CFR 63.132(g)(2), (3) and (4) for any facility electing to discharge wastewater streams to a RCRA-permitted treatment, storage, and disposal facility (TSDF) under 40 CFR parts 264 and 265. The commenter asserted that a RCRA TSDF should be presumed to be acceptable compliance equipment for miscellaneous organic chemical manufacturing facilities, and this presumption should be explicitly stated in the final rule.

Response:

We agree that RCRA TSDF satisfy the compliance requirements in the final rule. The proposed subpart FFFF explicitly stated that performance tests, design evaluations, and related monitoring, recordkeeping, and reporting would not be required when a hazardous waste incinerator is used to meet emission limits. This provision is retained in the final rule through the reference to § 63.988(b)(2), and it applies to offsite treatment facilities as well as affected sources. To simplify and clarify the requirements for offsite treatment facilities, the final rule states that the affected source may indicate in its notification of compliance status (NOCS) report that it is shipping the wastewater to an offsite treatment facility that meets the requirements of 40 CFR 63.138(h), and that the wastewater will be treated as hazardous waste; this documentation may serve as the certification from the offsite treatment facility.

G. Standards for Equipment Leaks

Comment:

Three commenters stated that the docket does not support our conclusion that the HON LDAR program is the MACT floor. Two of the commenters also opposed our approach of assigning a single LDAR program to each facility. They noted that facilities do not always use the same LDAR program for all of their processes. According to one commenter, there also are numerous errors and inconsistencies between various background memoranda, the section 114 ICR responses, and the equipment leaks database that EPA distributed to industry, with no documentation in the docket to explain the differences. After obtaining new information from some of the facilities in the database, the commenter saw no support for a determination that HON-equivalent controls establish the MACT floor (

i.e.

, of the estimated 1,220 processes, only 34, or 2.8 percent, appear to have HON-equivalent programs). The other two commenters indicated that the floor (and standard) should be based on either the LDAR program in the SOCMI NSPS (40 CFR part 60, subpart VV) or subpart TT of 40 CFR part 63 (the Generic MACT).

Response:

After considering the comments and reviewing the available data, we decided to determine the MACT floor on a process basis because some facilities do not implement the same LDAR program for all of their miscellaneous organic chemical manufacturing processes. Therefore, we decided to reevaluate the MACT floor on a process basis. Before revising the analysis, we also reviewed the specific data entries that were disputed by the commenters.

Regrettably, the database that was made available to the industry was not consistent with the final database that we used to develop the MACT floor prior to proposal. As a result, many of the discrepancies identified by commenters are addressed simply by using the correct database.

We also reviewed other changes that the commenter recommended and made corrections to the database under the following two circumstances: when a process is subject to the HON so that only the batch process vent emissions are subject to subpart FFFF, and when a facility representative informed the commenter that a non-HON LDAR program or no program is implemented for a miscellaneous organic chemical manufacturing process. After making the revisions, we found 51 of 1,139 processes are controlled to the HON LDAR (

i.e.

, the best-performing LDAR program in use at miscellaneous organic chemical manufacturing sources), or 4.5 percent controlled. Based on this result, we could not justify a MACT floor at the HON level of control.

Therefore, we reexamined the processes subject to other LDAR programs to develop a revised MACT

floor. A few processes are subject to LDAR programs required by the State of Louisiana, but most other processes subject to LDAR programs are implementing various programs required by the State of Texas or the program in 40 CFR part 60, subpart VV. For this analysis, we considered the Texas programs and the subpart VV program to be essentially equivalent because they all require only sensory monitoring for connectors. These programs also are equivalent to the program in 40 CFR part 63, subpart TT. Only LDAR programs designated as audible/visual/olfactory (

e.g.

, not Method 21 monitoring) were not considered at least equivalent to subpart TT. We found that 236 of the 1,139 processes, or 21 percent, were controlled at least to the subpart TT level. Therefore, we set the floor based on the requirements of 40 CFR part 63, subpart TT.

Based on the revised MACT floor, we conducted an analysis of the cost of going above the floor to the 40 CFR part 63, subpart UU, program. In conducting this analysis, we revised our estimated uncontrolled emissions for our model processes by using the initial leak rates submitted by the industry in their comments. At the leak definitions of 500 ppmv for connectors and valves and 1,000 ppmv for pumps, we calculated leak rates of 0.35 percent for connectors, 6.47 percent for pumps, and 1.66 percent for valves from the data submitted by the industry. We also compared these leak rates and their resulting emission rates to data collected in the development of the Polymers and Resins IV NESHAP and found good agreement. The polymers and resins industry leak rates were 0.61 percent for connectors, 8.71 percent for pumps, and 1.4 to 1.8 percent for valves. To estimate reductions achieved by the LDAR programs, we assumed that the reduction achieved by the subpart UU program would be equal to the emissions estimated at the performance level of the program. We assumed that the subpart TT program would be half as effective as subpart UU for pumps, valves, and connectors, and that the reductions for pressure relief valves, open-ended lines, and sampling connections would be the same under both programs.

We also revised elements in our cost analysis to address commenter concerns. The revised analysis assumes that a facility required to implement an LDAR program will hire a subcontractor based on our understanding that this is the preferred and common alternative over the implementation of an in-house program. The analysis also made use of revised cost data from the project files of the Polymers and Resins IV NESHAP.

The revised cost analysis shows that for processes with continuous process vents, the cost of the subpart TT program (the MACT floor) is $3,200/Mg, the cost of the subpart UU program is $2,800/Mg, and the incremental cost to go beyond the MACT floor to the subpart UU program is $470/Mg. These costs are considered reasonable. Conversely, for batch processes, the costs of the beyond-the-floor option were determined to be unreasonable. Therefore, we decided to set the standard at the MACT floor for processes with only batch process vents, and we selected the beyond-the-floor option of subpart UU for processes with at least one continuous process vent.

Comment:

Several commenters generally supported the pressure testing option in § 63.1036(b) of subpart UU, which requires that new or disturbed equipment be tested for leaks before use. However, the commenters are concerned that § 63.1036(b)(1)(iii) could be interpreted as requiring facilities to conduct leak tests whenever flexible hose connections are changed as part of a reconfiguration to make a different product or intermediate. The commenters stated that these leak tests would be burdensome because (1) changing flexible hoses to make different products may occur as frequently as daily or weekly, which would substantially increase the cost of conducting LDAR programs and take away from operating time, resulting in lost production and sales; (2) more frequent leak tests would also result in more emissions because the equipment must be purged to conduct the tests; and (3) flexible hoses that have been water tested would often have to be flushed with solvent prior to startup, which would add more turn-around time and increase waste generation. According to one commenter, connecting flexible hoses in different configurations is the type of “routine” seal breaks that were not intended to trigger LDAR pressure testing requirements. Thus, the commenters recommended revising § 63.1036(b)(1)(iii) to exempt all routine seal breaks of flexible hoses from LDAR requirements. One commenter also recommended that pressure testing be allowed as an option for sources that comply with the requirements in 40 CFR part 63, subpart TT.

Response:

We agree with the commenters that pressure testing each time process equipment is reconfigured only by changing flexible hose connections at a transfer station is excessively burdensome and likely to lead to more emissions than it prevents. Therefore, the pressure test option in the final rule allows this type of routine disturbance without the requirement to conduct a new pressure test. Since the final rule allows compliance with the requirements of 40 CFR part 63, subpart UU, as an alternative to the requirements of 40 CFR part 63, subpart TT, an owner or operator may comply with the pressure testing option in subpart UU as an alternative to the requirements of subpart TT.

H. Standards for Transfer Racks

Comment:

One commenter indicated the MACT floor for transfer racks was established incorrectly and stated that we have no section 114 ICR data to support the transfer racks MACT floor because this information was not requested for the miscellaneous organic chemical manufacturing source category. The commenter indicated that using transfer rack data from HON sources or Organic Liquid Distribution (OLD) NESHAP sources is not appropriate for the miscellaneous organic chemical manufacturing source category, even if it does streamline the compliance process. The commenter noted that the Group 1 requirements of subpart G of the HON apply to a different source category manufacturing different chemicals in continuous, generally high-volume processes. The commenter claimed we made a “leap of faith” in assuming that the emission and control data for one source category are appropriate to another totally distinct category. The commenter could find no documentation indicating that subpart G continuous process load rates and vapor pressure cutoffs are applicable to batch subpart FFFF facilities. The commenter argued that setting a MACT floor using “existing available data” from a different source category is inconsistent with CAA requirements and requested that an actual transfer rack MACT floor determination be made prior to establishing the subpart FFFF control requirements.

Response:

The MACT floor was based on the HON requirements. We did not have any specific data from our source category, but we relied on information that many of the facilities in this source category are co-located with facilities subject to the HON. The commenter objected to our assumptions because the HON applies to continuous, high volume production processes. Although subpart FFFF applies to many processes, batch specialty chemicals are a major component of the source category, and we agree that individual products are typically manufactured in lesser volumes than typical products in the HON source category. However, we

note that transfer operations, which by definition consist of the loading racks for tank trucks and rail cars, are more specific to the size and type of vessel being loaded than the process that generates the products.

These tank trucks and rail cars are standard in size and configuration so that the same tank trucks and rail cars would be expected to carry material from either source category. Further, pumps, loading arms, and vapor collection and control equipment are not as much dependent on the process that generates the materials as the products themselves which are composed of either pure HAP or solutions containing significant amounts of HAP.

Our data indicate that 60 percent of the facilities that contain miscellaneous organic chemical manufacturing processes also contain processes subject to the HON. Additionally, we would expect that transfer racks located at these facilities would be used to load materials from both HON and miscellaneous organic chemical manufacturing processes. Therefore, we consider it reasonable to assume a MACT floor based on the requirements of the HON.

The HON standards were established based on the lowest yearly loading rates that are controlled in the source category. Because the HON source category manufactures at typically higher volume production than what would typically be expected in the miscellaneous organic chemical manufacturing source category, and control requirements are based on the rack weighted average partial pressure of HAP, it offers a conservative approach to the MACT floor when applied to the batch specialty chemical industry. Therefore, only transfer racks that load miscellaneous organic chemical manufacturing products containing significant amounts of HAP are affected by the control requirements.

I. Pollution Prevention

Comment:

Three commenters stated that the pollution prevention (P2) option should be broadened to allow more nondedicated batch operations or groups of nondedicated batch operations to use P2 for compliance. The commenters maintained that calculating and tracking HAP factors for individual nondedicated processes would not be viable for small operations. One commenter was concerned that only dedicated solvent recovery operations may be included in a P2 demonstration; nondedicated solvent recovery operations may not be considered in conjunction with the processes for which they recover solvents for the P2 alternative standard. Similarly, another commenter stated that the proposal is not viable because waste solvents from numerous nondedicated batch processes are collected and refined at a central recovery unit, and § 63.2495(b)(2) of the proposed rule would preclude the merging of nondedicated solvent recovery with other processes. The commenter suggested including all of the operations in the calculation of a HAP consumption factor (including nondedicated recovery operations that receive and recover solvents for the operations). In addition, the commenter suggested that the production rate should exclude isolated intermediates to appropriately reflect the benefits achieved when measures are taken to eliminate isolation of intermediates. Because the boundaries are well defined, the commenter indicated that such an approach would be clearer to implement and enforce. To incorporate this approach, the commenter suggested adding a statement to the rule that says you may comply with the P2 option for multiple processes and associated recovery operations if the Administrator approves your P2 methodology submitted in the precompliance report.

Response:

After examining the approach suggested by the commenters, we have concerns that it would not be consistent with the goals of P2 and also would not preserve the reductions in HAP consumption that would occur if the P2 alternative were limited to each product. The commenters suggested facilitywide groupings to demonstrate overall reductions in the HAP consumption factors. One of our major concerns stems from the fact that specialty chemical facilities will not manufacture the same products from the baseline years to the contemporaneous period. Under their suggested grouping concept, however, a baseline factor could be developed from a different set of products than those in the contemporaneous period. In this situation, a facility could demonstrate a reduction in the HAP factor by simply not manufacturing products that have high HAP consumption. Although these efforts could result in a net benefit to the environment, they are not considered P2 strategies and, therefore, an owner or operator should not take credit for these changes. Secondly, using the same groupings concept, a manufacturer could effectively reduce the overall usage of HAP in a production process in any given year, but increase the HAP factor for that product and still meet the grouping target reductions, but not the target reductions on individual product lines. This would effectively allow an owner or operator to comply with a P2 alternative that could increase the inefficiency and waste within a process. Therefore, combining processes or groups of processes as suggested by the commenters is not appropriate, and we have not revised the alternative per the commenter's requests.

We also clarified language regarding merging processing steps conducted offsite to onsite for the purposes of redrawing a process boundary and claiming a reduction in consumption. For example, a solvent recovery step conducted offsite or as part of another process cannot later be moved onsite or to another process and used to claim a reduction in consumption. Such a strategy does not result in true emission reductions, but rather is a result of moving process boundaries.

Comment:

Several commenters were concerned that the proposed P2 option would not allow for the generation of HAP other than HAP being used in the process. They noted that based on the definition of “consumption” and § 63.2495(b) of the proposed rule, if the HAP used by the process are not the same as those generated in the process, then the generated HAP must meet the otherwise applicable standards. One commenter suggested revising the definition of consumption to include HAP generated in the process, and the other commenters suggested incorporating generated HAP into the calculation of the HAP factor or the target HAP reduction.

Response:

We do not agree with the suggested changes. The P2 alternative specifies that HAP generated in the process that are not introduced into the process and part of the consumption factor must be controlled per the standard requirements. This restriction is needed to ensure that reductions anticipated from the implementation of the alternative will occur. Consider a situation where the incoming quantity of HAP is considerably less than the amount of HAP generated in the process. Further, suppose the entire quantity of HAP generated in the process is emitted through a process vent (

i.e.

, no waste or wastewater). If the P2 alternative were to allow the quantity of HAP generated to be considered as part of the consumption factor, then the P2 standard could be met by capturing and recovering only 65 percent of the HAP emitted, which may not preserve the reductions we anticipated from the implementation of the standards as written. Therefore, we have not

modified the alternative according to the commenters' requests.

J. Initial Compliance

Comment:

Several commenters indicated that the proposed requirements to complete initial compliance demonstrations and submit the NOCS report by the compliance date are unworkable and unreasonably and unfairly shorten the 3-year compliance period. Based on the commenters' experience, the entire 3-year period is needed to permit, plan, design, procure, install, and shakedown the equipment necessary for MACT compliance. In addition, the 150-day period after the compliance date that other rules allow before the NOCS report is due allows facilities to properly test their control systems, perform necessary shakedown operations, and set the parametric operating limits using actual data. The commenters requested that the final rule defer to the General Provisions regarding the timeline for initial compliance demonstrations and allow the NOCS report to be submitted no later than 150 days after the compliance date. Another commenter requested that area sources that become major sources be allowed up to 3 years to comply with the final rule because the level of effort would be the same as for any existing source when the rule is promulgated.

Response:

We accept the argument that some facilities with numerous processes and controls may need the full 3 years from the promulgation date to the compliance date to bring all of the equipment on-line before completing the initial compliance demonstration. Therefore, we decided to change the due date for the NOCS report. In the final rule, the NOCS report for all sources, including area sources that become major sources, is due no later than 150 days after the compliance date. In addition, the final rule specifies that the compliance date for area sources that become major sources is 3 years after the area source becomes a major source.

Comment:

Several commenters indicated that references to § 63.1257(d)(2)(ii) of the pharmaceutical MACT in the proposed rule inappropriately restrict the use of engineering assessments. The commenters indicated that the rule should not require sources to demonstrate that the calculation methods specified in the rule are not appropriate in order to be allowed to calculate uncontrolled HAP emissions using an engineering assessment. The commenters also objected to language in § 63.1257(d)(2)(ii) that restricts the use of modified equations to those that the source can demonstrate have been used to meet other regulatory obligations. The commenters indicated that they should only be required to show that the selected method for determining uncontrolled HAP emissions is appropriate, and that it has no impact on the applicability assessment or compliance determination.

Response:

We did not revise the restrictions on the use of the modified equations as requested because the suggested changes would not maintain our objective of having a replicable compliance protocol that is applied consistently across the source category. Therefore, the final rule, like the proposed rule, restricts the use of engineering assessments to situations where the equations are not appropriate.

Comment:

Several commenters requested that the procedures for calculating uncontrolled HAP emissions be modified in the final rule so that it represents “post condenser” emissions if the condenser is recovering HAP for reflux, reuse, or use as a fuel. The commenters stated that, for many types of emission events, the proposed equations would require the use of the vessel temperature rather than the temperature of the receiver that receives condensed liquid. The commenters indicated that the procedures ignore the emission reduction realized by the condenser, inflates the uncontrolled emissions, and is inconsistent with the MACT floor database.

Response:

We disagree with the suggested change. Our position is that uncontrolled emissions should be determined at the point the vent stream leaves the process and prior to entering any control device. A condenser that meets the definition of “process condenser” is considered integral to the process, and uncontrolled emissions are calculated at the outlet of the condenser. Process condensers must initiate vapor-to-liquid phase change in an emission stream from equipment that operates above the boiling or bubble point, including condensers located prior to a vacuum source. All other condensers serve primarily to reduce or remove air pollutants, with or without some product recovery benefits; therefore, uncontrolled emissions should be calculated prior to the condenser. This approach does not inflate uncontrolled emissions; it characterizes them properly. Furthermore, if a condenser is determined to be an air pollution control device, the removal efficiency is included as part of the overall control efficiency for the process; it does not ignore the emission reduction realized by the condenser. Finally, we consider the approach to be consistent with our database because we provided clear instructions with

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