Browse: Departments   Dates   Agencies  

ENVIRONMENTAL PROTECTION AGENCY

Western Area Power Administration

CFR Citation: 40 CFR Part 63

EPA ID: [EPA-HQ-OAR-2003-0146; FRL-8461-3]

RIN ID: RIN 2060-AO55

NOTICE: Part II

DOCUMENT ACTION: Proposed rule.

SUBJECT CATEGORY: National Emission Standards for Hazardous Air Pollutants From Petroleum Refineries

DATES: Comments must be received on or before November 5, 2007.

DOCUMENT SUMMARY: This action proposes amendments to the national emission standards for petroleum refineries to address the risk remaining after application of the 1995 standards. This action also provides the results of EPA's 8year review of developments in practices, processes, and control technologies that have occurred since the time EPA adopted the emissions standards. Based on the results of the residual risk and technology review, this action proposes two options for both wastewater treatment systems and storage vessels. For wastewater treatment systems, the first option would not require any additional controls as necessary to address residual risk or under the technology review. The second option would require refineries to apply new or additional requirements for wastewater treatment systems. For storage vessels, the first option would also not require any additional controls as necessary to address residual risk or under the technology review and the second option would require refineries to apply new or additional requirements for storage vessels. Finally, we are also proposing two options for amendments to add emissions standards for cooling towers.

SUMMARY: Environmental Protection Agency,

SUPPLEMENTAL INFORMATION

I. General Information

A. Does this action apply to me?

The regulated category and entities affected by this proposed action include:
NAICS \1\ Examples of regulated Category code entities Industry.......................... 32411 Petroleum refineries located at a major source that are subject to 40 CFR part 63, subpart CC. \1\ North American Industrial Classification System.

This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely to be regulated by the proposed rule. To determine whether your facility would be regulated by the proposed amendments, you should carefully examine the applicability criteria in 40 CFR 63.100 of subpart CC (National
[[Page 50717]]
Emission Standards for Hazardous Air Pollutants From Petroleum Refineries). If you have any questions regarding the applicability of this action to a particular entity, contact either the air permit authority for the entity or your EPA regional representative as listed in 40 CFR 63.13 of subpart A (General Provisions).
B. What should I consider as I prepare my comments for EPA?

Do not submit information containing CBI to EPA through http://www.regulations.gov or email. Send or deliver information as CBI only to the following address: Roberto Morales, OAQPS Document Control Officer (C40402), Office of Air Quality Planning and Standards, Environmental Protection Agency, Research Triangle Park, NC 27711, Attention Docket ID EPAHQOAR20030146 (for petroleum refineries). Clearly mark the part or all of the information that you claim to be CBI. For CBI information in a disk or CD ROM that you mail to EPA, mark the outside of the disk or CD ROM as CBI and then identify electronically within the disk or CD ROM the specific information that is claimed as CBI. In addition to one complete version of the comment that includes information claimed as CBI, a copy of the comment that does not contain the information claimed as CBI must be submitted for inclusion in the public docket. Information so marked will not be disclosed except in accordance with procedures set forth in 40 CFR part 2.

C. Where can I get a copy of this document?

In addition to being available in the docket, an electronic copy of this proposed action will also be available on the Worldwide Web through the Technology Transfer Network (TTN). Following signature, a copy of this proposed action will be posted on the TTN(s policy and guidance page for newly proposed or promulgated rules at the following address: http://www.epa.gov/ttn/oarpg/. The TTN provides information and technology exchange in various areas of air pollution control. D. When would a public hearing occur?

If anyone contacts EPA requesting to speak at a public hearing concerning the proposed amendments by September 17, 2007, we will hold a public hearing on October 1, 2007. If you are interested in attending the public hearing, contact Bob Lucas at (919) 5410884 to verify that a hearing will be held. If a public hearing is held, it will be held at 10 a.m. at the EPA's Environmental Research Center Auditorium, Research Triangle Park, NC, or an alternate site nearby.
E. How is this document organized?
I. General Information

A. Does this action apply to me?

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

C. Where can I get a copy of this document?

D. When would a public hearing occur?

E. How is this document organized?

II. Background Information

A. What is the statutory authority for regulating hazardous air pollutants?

B. What source category is affected by this action?

C. What are the emissions sources at petroleum refineries?

D. What hazardous air pollutants are emitted from petroleum refineries?

E. What does the NESHAP require?
III. Summary of Proposed Amendments to NESHAP for Petroleum Refineries

A. What options are we proposing?

B. What are the proposed requirements to meet CAA sections 112(f)(2) and (d)(6) for storage vessels?

C. What are the proposed requirements to meet CAA sections 112 (f)(2) and (d)(6) for EBU used to treat Group 1 wastewater streams?

D. What are the proposed requirements for cooling towers under CAA sections 112(d)(2) and (f)(2)?

E. What other revisions are we proposing?

F. What is the compliance schedule for the proposed amendments? IV. Rationale for Proposed Amendments

A. What actions are we proposing under CAA section 112(d)(2)?

B. How did we estimate residual risk?

C. What are the residual risks from petroleum refineries?

D. What are the uncertainties in risk assessments?

E. What is our proposed decision under CAA section 112(f)?

F. What is EPA proposing pursuant to CAA section 112(d)(6)? V. Request for Comments

VI. 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 Risks and Safety Risks

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

I. National Technology Transfer and Advancement Act

J. Executive Order 12898: Federal Actions to Address Environmental Justice in Minority Populations and LowIncome Populations
II. Background Information
A. What is the statutory authority for regulating hazardous air pollutants?

Section 112 of the Clean Air Act (CAA) establishes a twostage regulatory process to address emissions of hazardous air pollutants (HAP) from stationary sources. In the first stage, after EPA has identified categories of sources emitting one or more of the HAP listed in section 112(b) of the CAA, section 112(d) calls for us to promulgate national emission standards for hazardous air pollutants (NESHAP) for those sources. For ``major sources'' that emit or have the potential to emit any single HAP at a rate of 10 tons or more per year or any combination of HAP at a rate of 25 tons or more per year, these technologybased standards must reflect the maximum reductions of HAP achievable (after considering cost, energy requirements, and nonair quality health and environmental impacts) and are commonly referred to as maximum achievable control technology (MACT) standards.

The MACT floor is the minimum control level allowed for NESHAP and is defined under section 112(d)(3) of the CAA. For new sources, the MACT floor cannot be less stringent than the emission control that is achieved in practice by the bestcontrolled 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 bestperforming 12 percent of existing sources in the category or subcategory (or the bestperforming five sources for categories or subcategories with fewer than 30 sources). In developing MACT, we must also consider control options that are more stringent than the floor. We may establish standards more stringent than the floor based on the consideration of the cost of achieving the emissions reductions, any nonair quality health and environmental impacts, and energy requirements. We published the final MACT standards for petroleum refineries (40 CFR part 63, subpart CC) on August 18, 1995 (60 FR 43620).

The EPA is then required to review these technologybased standards and to revise them ``as necessary (taking into account developments in practices, processes, and control technologies)'' no less frequently than every 8 years, under CAA section 112(d)(6). In this proposal, we are publishing the results of our 8year review for the petroleum refineries source category. We are required by a consent decree to propose the results of our CAA section 112(d)(6) review by
[[Page 50718]]
August 21, 2007. The consent decree also requires EPA to consider and address the application of the NESHAP general provisions in 40 CFR part 63, subpart A to the existing rule.

The second stage in standardsetting focuses on reducing any remaining ``residual'' risk according to CAA section 112(f). This provision requires, first, that EPA prepare a Report to Congress discussing (among other things) methods of calculating risk posed (or potentially posed) by sources after implementation of the MACT standards, the public health significance of those risks, the means and costs of controlling them, actual health effects to persons in proximity of emitting sources, and recommendations as to legislation regarding such remaining risk. EPA prepared and submitted this report (Residual Risk Report to Congress, EPA453/R99001) in March 1999. Congress did not act in response to the report, thereby triggering EPA's obligation under CAA section 112(f)(2) to analyze and address residual risk.

CAA Section 112(f)(2) requires us to determine for source categories subject to certain section 112(d) standards whether the emissions limitations protect public health with an ample margin of safety. If the MACT standards for HAP ``classified as a known, probable, or possible human carcinogen do not reduce lifetime excess cancer risks to the individual most exposed to emissions from a source in the category or subcategory to less than 1in1 million,'' EPA must promulgate residual risk standards for the source category (or subcategory) as necessary to provide an ample margin of safety to protect public health. The EPA must also adopt more stringent standards if necessary to prevent an adverse environmental effect (defined in CAA section 112(a)(7) as any significant and widespread adverse effect * * * to wildlife, aquatic life, or natural resources * * *), but must consider cost, energy, safety, and other relevant factors in doing so. Section 112(f)(2) of the CAA expressly preserves our use of a twostep process for developing standards to address any residual risk and our interpretation of ``ample margin of safety'' developed in the National Emission Standards for Hazardous Air Pollutants: Benzene Emissions from Maleic Anhydride Plants, Ethylbenzene/Styrene Plants, Benzene Storage Vessels, Benzene Equipment Leaks, and Coke ByProduct Recovery Plants (Benzene NESHAP) (54 FR 38044, September 14, 1989).

The first step in this process is the determination of acceptable risk. The second step provides for an ample margin of safety to protect public health, which is the level at which the standards are set (unless a more stringent standard is required to prevent an adverse environmental effect after the consideration of costs, energy, safety, and other relevant factors).

The terms ``individual most exposed,'' ``acceptable level,'' and ``ample margin of safety'' are not specifically defined in the CAA. However, CAA section 112(f)(2)(B) directs us to use the interpretation set out in the Benzene NESHAP. See also, A Legislative History of the Clean Air Act Amendments of 1990, volume 1, p. 877 (Senate debate on Conference Report). We notified Congress in the Residual Risk Report to Congress that we intended to use the Benzene NESHAP approach in making CAA section 112(f) residual risk determinations (EPA453/R99001, p. ES11).

In the Benzene NESHAP, we stated as an overall objective: * * * in protecting public health with an ample margin of safety, we strive to provide maximum feasible protection against risks to health from hazardous air pollutants by (1) Protecting the greatest number of persons possible to an individual lifetime risk level no higher than approximately 1in1 million; and (2) limiting to no higher than approximately 1in10 thousand [i.e., 100in1 million] the estimated risk that a person living near a facility would have if he or she were exposed to the maximum pollutant concentrations for 70 years.

The Agency also stated that, ``The EPA also considers incidence (the number of persons estimated to suffer cancer or other serious health effects as a result of exposure to a pollutant) to be an important measure of the health risk to the exposed population. Incidence measures the extent of health risk to the exposed population as a whole, by providing an estimate of the occurrence of cancer or other serious health effects in the exposed population.'' The Agency went on to conclude that ``estimated incidence would be weighed along with other health risk information in judging acceptability.'' \1\ As explained more fully in our Residual Risk Report to Congress, EPA does not define ``rigid line[s] of acceptability,'' but considers rather broad objectives to be weighed with a series of other health measures and factors (EPA453/R99001, p. ES11).
\1\ In the Benzene NESHAP decision, the Agency considered the same risk measures in the ``acceptability'' analysis as in the ``margin of safety'' analysis, stating: ``In the ample margin decision, the Agency again considers all of the health risk and other health information considered in the first step. Beyond that information, additional factors relating to the appropriate level of control will also be considered, including costs and economic impacts of controls, technological feasibility, uncertainties, and any other relevant factors. Considering all of these factors, the Agency will establish the standard at a level that provides an ample margin of safety to protect the public health, as required by section 112.''

The determination of what represents an ``acceptable'' risk is based on a judgment of ``what risks are acceptable in the world in which we live'' (54 FR 38045, quoting the Vinyl Chloride decision at 824 F.2d 1165) recognizing that our world is not riskfree.

In the Benzene NESHAP, we stated that ``EPA will generally presume that if the risk to [the maximum exposed] individual is no higher than approximately 1in10 thousand, that risk level is considered acceptable.'' We discussed the maximum individual lifetime cancer risk as being ``the estimated risk that a person living near a plant would have if he or she were exposed to the maximum pollutant concentrations for 70 years.'' We explained that this measure of risk ``is an estimate of the upper bound of risk based on conservative assumptions, such as continuous exposure for 24 hours per day for 70 years.'' We acknowledge that maximum individual lifetime cancer risk ``does not necessarily reflect the true risk, but displays a conservative risk level which is an upper bound that is unlikely to be exceeded.''

Understanding that there are both benefits and limitations to using maximum individual lifetime cancer risk as a metric for determining acceptability, we acknowledged in the 1989 Benzene NESHAP that ``consideration of maximum individual risk * * * must take into account the strengths and weaknesses of this measure of risk.'' Consequently, the presumptive risk level of 100in1 million (1in10 thousand) provides a benchmark for judging the acceptability of maximum individual lifetime cancer risk, but does not constitute a rigid line for making that determination.

The Agency also explained in the 1989 Benzene NESHAP the following: ``In establishing a presumption for MIR, rather than rigid line for acceptability, the Agency intends to weigh it with a series of other health measures and factors. These include the overall incidence of cancer or other serious health effects within the exposed population, the numbers of persons exposed within each individual lifetime risk range and associated incidence within, typically, a 50 km exposure radius around facilities, the science policy assumptions and estimation uncertainties associated with the risk measures, weight of the scientific evidence for human health effects, other quantified or unquantified health
[[Page 50719]]
effects, effects due to colocation of facilities, and coemission of pollutants.''

In some cases, these health measures and factors taken together may provide a more realistic description of the magnitude of risk in the exposed population than that provided by maximum individual lifetime cancer risk alone.

B. What source category is affected by this action?

Petroleum refineries are facilities engaged in refining and producing products made from crude oil or unfinished petroleum derivatives. Based on the Energy Information Administration's Refinery Capacity Report 2006, there are 150 operable petroleum refineries in the United States (U.S.) and the U.S. territories. A few of these 150 refineries have integrated operations between two nearby, but non contiguous, locations. Therefore, we have identified and have data on 153 distinct petroleum refinery facilities (according to the definition of facility in the CAA), all of which are major sources of HAP emissions. Petroleum refineries are located in 35 States, as well as Puerto Rico and the U.S. Virgin Islands. Texas, Louisiana, and California are the States with the most petroleum refining capacity. The permitting process has begun for construction of a new refinery in Arizona; this is the only newly constructed refinery anticipated over the next 5 years. However, a few additional refineries have announced significant expansion or modification projects that will essentially double their refining capacity.

EPA listed two separate Petroleum Refinery source categories for regulation under CAA section 112(d), both of which include any facility engaged in producing gasoline, naphtha, kerosene, jet fuels, distillate fuel oils, residual fuel oils, lubricants, or other products from crude oil or unfinished petroleum derivatives. The first and primary source category for which regulations were developed, Petroleum Refineries Other Sources Not Distinctly Listed (Refinery MACT 1), includes all emission sources from petroleum refinery process units except those that were expected to be regulated elsewhere, such as the NESHAP for Boilers and Process Heaters (40 CFR part 63 subpart DDDDD). Refinery process units include, but are not limited to: Crude distillation, vacuum distillation, thermal cracking, catalytic cracking, catalytic reforming, hydrotreating, hydrorefining, isomerization, polymerization, lube oil processing, and hydrogen production. The Refinery MACT 1 rule specifically excludes three types of process vents: Catalytic cracking unit catalyst regeneration vents, catalytic reforming unit catalyst regeneration vents, and sulfur plant vents. These specific vents are regulated by the NESHAP for Petroleum Refineries: Catalytic Cracking Units, Catalytic Reforming Units, and Sulfur Recovery Units (Refinery MACT 2) in 40 CFR part 63, subpart UUU. It is important to note that equipment leaks and wastewater produced from catalytic cracking units, catalytic reforming units, and sulfur recovery units are subject to Refinery MACT 1; only the process vent emissions associated with these units are subject to Refinery MACT 2.

C. What are the emissions sources at petroleum refineries?

The emissions sources subject to the Refinery MACT 1 rule include miscellaneous process vents, storage vessels, wastewater streams, and equipment leaks associated with petroleum refining process units, as well as gasoline loading racks and marine tank vessel loading operations located at a petroleum refinery. Storage vessels and equipment leaks associated with a bulk gasoline terminal or pipeline breakout station located at a petroleum refinery and under common control of the refinery are also subject to Refinery MACT 1. Cooling towers associated with petroleum refining process units are part of the MACT 1 source category although no specific emission limitations were established for cooling towers in the original Refinery MACT 1 rule. Thus, there are seven general types of emission sources under Refinery MACT 1: Miscellaneous process vents, storage vessels, wastewater streams, equipment leaks, gasoline loading racks, marine tank vessel loading operations, and cooling towers. Each of these emission sources are described briefly in sections II.C.1 through II.C.7 of this preamble.

1. Miscellaneous Process Vents

Many unit operations at petroleum refineries generate gaseous streams that contain HAP. These streams may be routed to other unit operations for additional processing (i.e., a gas stream from a reactor that is routed to a distillation unit for separation) or they may be sent to a blowdown system or vented to the atmosphere. Miscellaneous process vents emit gases to the atmosphere, either directly or after passing through recovery and/or control devices.

2. Storage Vessels

Storage vessels contain crude oil, intermediate products, and finished products. Different types of vessels are used to store various types of products. Gases are stored in pressurized vessels that are not vented to the atmosphere during normal operations while liquids are stored in horizontal, fixed roof, or floating roof tanks, depending on properties and volumes to be stored. Liquids with vapor pressures greater than 11 pounds per square inch of air (psia) are typically stored in fixed roof tanks that are vented to a control device. Volatile liquids with vapor pressures up to 11 psia are usually stored in floating roof tanks because such vessels have lower emission rates than fixed roof tanks within this vapor pressure range. Emissions from storage vessels typically occur as working losses. As a storage vessel is filled, HAPladen vapors inside the tank become displaced and can be emitted to the atmosphere. Also, diurnal temperature changes result in breathing losses of organic HAPladen vapors from storage vessels. 3. Wastewater Streams

Many refinery process units generate wastewater streams that contain HAP. Significant wastewater sources include the crude desalting unit, process waters, steam stripper blowdown, and storage tank draws. Organic HAP compounds in the wastewater can volatilize and be emitted to the atmosphere from wastewater collection and treatment units if these units are open or vented to the atmosphere. Potential sources of HAP emissions associated with wastewater collection and treatment systems include drains, manholes, trenches, surface impoundments, oil/ water separators, storage and treatment tanks, junction boxes, sumps, basins, and biological treatment systems.

4. Equipment Leaks

Equipment leaks are releases of process fluid or vapor from processing equipment, including pump and compressor seals, process valves, pressure relief devices, openended lines, flanges and other connectors, agitators, and instrumentation systems. These releases occur primarily at the interface between connected components of equipment or in sealing mechanisms.

5. Gasoline Loading Racks

Loading racks are the collection of equipment, including loading arms, pumps, meters, shutoff valves, relief valves, and other piping and valves used to fill gasoline cargo tanks. Emissions from loading racks may be
[[Page 50720]]
released when gasoline loaded into cargo tanks displaces vapors inside these containers.

6. Marine Vessel Loading Operations

Marine vessel loading operations load and unload liquid commodities in bulk, such as crude oil, gasoline and other fuels, and naphtha. The cargo is pumped from the terminal's large, aboveground storage tanks through a network of pipes and into a storage compartment (tank) on the vessel. The HAP emission result from the displaced vapors during the filling operation.

7. Cooling Towers

Cooling tower systems include closed loop recirculation systems and once through systems that receive noncontact process water from a heat exchanger for the purposes of cooling the process water prior to returning the water to the heat exchanger or discharging the water to another process unit, waste management unit, or to a receiving water body. Cooling towers typically use force draft air ventilation of the process water to cool the process water. Heat exchangers occasionally develop leaks which result in process fluids entering the cooling tower process water. The HAP and other organics in these process fluids are then emitted to the atmosphere due to stripping in the cooling tower. Cooling tower emissions arising from the addition of chemicals to the cooling water to prevent fouling or to decontaminate the water are not covered by this standard, but are instead covered under the Industrial Process Cooling Tower NESHAP.
D. What hazardous air pollutants are emitted from petroleum refineries?

The specific HAP emitted by petroleum refineries varies by facility and process operations but can include a variety of organic and inorganic compounds and metals. Emissions originate from various process vents, storage vessels, wastewater streams, loading racks, marine tank vessel loading operations, and equipment leaks associated with refining facilities. Process vents, wastewater streams, and storage vessels generally emit organic HAP. Organic compounds account for the majority of the total mass of HAP emitted by petroleum refinery sources, with toluene, hexane, mixed and individual isomers of xylenes, benzene, methanol, methyl tertbutyl ether, and ethyl benzene accounting for about 90 percent of the HAP mass emitted. Other HAP emissions may include biphenyl, 1,3butadiene, cumene, carbon disulfide, carbonyl sulfide, cresols, ethylene dibromide, 1,2 dichloroethane, diethanolamine, ethylene glycol, methyl isobutyl ketone, 2,2,4trimethylpentane, naphthalene, and phenol.

The HAP emitted from emissions sources subject to the Refinery MACT 1 rule are associated with a variety of health effects, depending on the specific pollutants involved and the degree and duration of exposure. The range of adverse health effects include cancer and a number of other chronic health disorders (e.g., aplastic anemia, panctopenia, pernicious anemia, lung structural changes) and a number of acute health disorders (difficulty in breathing, upper respiratory tract irritation, conjunctivitis, tremors, delirium, coma, convulsions). More details on the health effects of individual HAP may be found in numerous sources, including http://www.epa.gov/iris.html, http://www.oehha.ca.gov/air/acute_rels/index .html.

E. What does the NESHAP require?

The Refinery MACT 1 rule (40 CFR part 63, subpart CC) applies to petroleum refining process units and their collocated emissions points that are part of a plant site that is a major source and that emit or have equipment containing or contacting one or more of the 28 HAP listed in Table 1 in the appendix to the rule. Section 63.640(c) of the rule specifies that emissions points subject to the rule include an individual miscellaneous process vent, storage vessel, wastewater stream, or equipment leak associated with a petroleum refining process unit; an individual storage vessel or equipment leak associated with a bulk gasoline terminal or pipeline breakout station classified under Standard Industrial Classification (SIC) code 2911 located at a petroleum refinery; a gasoline loading rack classified under SIC code 2911 located at a petroleum refinery and under common control with the refinery; or a marine tank vessel loading operation located at a petroleum refinery. The rule establishes applicability criteria to distinguish between Group 1 emissions points and Group 2 emissions points. Controls are required only for emissions points meeting the Group 1 criteria. Group 2 emissions points are subject to recordkeeping requirements only. We estimate that the 1995 rule reduces HAP emissions by 53,000 tons per year (tpy)a 59percent reduction (60 FR 43248, August 18, 1995).

Section 63.641 of the rule defines Group 1 miscellaneous process vents as those with volatile organic compound (VOC) emissions equal to or greater than 33 kilograms per day (kg/day) (72 pounds per day (lb/ day)) for existing sources and 6.8 kg/day (15 lb/day) for new sources. Under Sec. 63.643, the owner or operator of a Group 1 miscellaneous process vent must reduce organic HAP using a flare that meets the equipment specifications in 40 CFR 63.11 of the general provisions (subpart A) or use a control device to reduce organic HAP emissions by 98 weightpercent or to a concentration of 20 parts per million by volume (ppmv dry basis, corrected to 3 percent oxygen).

Section 63.646(a) of the Refinery MACT 1 rule requires each Group 1 storage vessel to comply with 40 CFR 63.119 through 63.121 of subpart G (National Emission Standards for Organic Hazardous Air Pollutants From the Synthetic Organic Chemical Manufacturing Industry for Process Vents, Storage Vessels, Transfer Operations, and Wastewater). A Group 1 storage vessel at an existing refinery has a design storage capacity and maximum true vapor pressure greater than the values specified in the existing rule. Under 40 CFR 63.119, a Group 1 storage vessel must be equipped with an internal floating roof with proper seals, an external floating roof with proper seals, an external floating roof converted to an internal floating roof with proper seals, or a closed vent system to a control device that reduces HAP emissions by 95 percent or to 20 ppmv. Storage vessels at existing sources are not subject to certain equipment specifications and inspection requirements for automatic bleeder vents, gaskets, slotted membranes, and sleeve seals. See 40 CFR 63.640(c). The requirements for a Group 1 storage vessel at a new refinery apply to tanks with a smaller design capacity and lower vapor pressures and HAP liquid concentration. These tanks also must comply with the storage vessel requirements in 40 CFR part 63, subpart G.

Each Group 1 wastewater stream at a new or existing refinery must comply with 40 CFR 61.340 through 61.355 of the National Emission Standard for Benzene Waste Operations (BWON) in 40 CFR part 61, subpart FF. Group 1 wastewater streams are those wastewater streams (at a petroleum refinery that has a total annual benzene loading of 10 megagrams per year (Mg/yr) or greater) that have a flow rate greater than 0.02 liters per minute, a benzene concentration of 10 parts per million by weight (ppmw) or greater, and are not exempt from control requirements under the BWON. The BWON requires affected waste streams [[Page 50721]]
to comply with one of several options for controlling benzene emissions from waste management units and treating the benzene containing wastes.

The Refinery MACT 1 rule requires the owner or operator of an existing refinery to comply with the equipment leak provisions in 40 CFR part 60, subpart VV (Standards of Performance for Equipment Leaks of VOC in the Synthetic Organic Chemicals Manufacturing Industry) for all equipment in organic HAP service. The term ``in organic HAP service'' means that a piece of equipment either contains or contacts a fluid (liquid or gas) that is at least 5 percent by weight of total organic HAP. The owner or operator of a new facility must comply with a modified version of 40 CFR part 63, subpart H (National Emission Standards for Organic Hazardous Air Pollutants for Equipment Leaks). Both subpart VV of part 60 and modified subpart H of part 63 require inspection and repair of leaking equipment. The leak definition under subpart VV that triggers repair requirements is an instrument reading of 10,000 ppmv. In the modified version of subpart H, the leak definition for pumps and valves begins at 10,000 ppmv but drops to 2,000 ppmv or 1,000 ppmv, respectively, in subsequent years.

Group 1 gasoline loading racks at refineries must comply with the requirements of the National Emission Standards for Gasoline Distribution Facilities (Bulk Gasoline Terminals and Pipeline Breakout Stations) in 40 CFR part 63, subpart R. Marine tank vessel loading operations at refineries must comply with the requirements in 40 CFR part 63, subpart Y (National Emission Standards for Marine Tank Vessel Loading Operations).
III. Summary of Proposed Amendments to NESHAP for Petroleum Refineries A. What options are we proposing?

We are proposing regulatory options for storage vessels with external floating roofs and regulatory options for an enhanced biodegradation unit (EBU) to meet the requirements of CAA sections 112(f)(2) and (d)(6). We are also proposing options to require a leak detection and repair program for cooling towers under section 112(d)(2) and (f)(2).

A detailed summary of the proposed amendments under the requirements of CAA section 112(f)(2) and (d)(6) is provided below. This section also includes our discussion of the proposal to regulate cooling towers under CAA section 112(d)(2) and (f)(2). Our rationale for the proposed amendments is provided in section IV of this preamble. B. What are the proposed requirements to meet CAA sections 112(f)(2) and (d)(6) for storage vessels?

Currently, the Refinery MACT 1 rule requires Group 1 storage vessels at an existing source to comply with the requirements in 40 CFR 63.119 through 63.121 of 40 CFR part 63, subpart G, except where specifically noted. Under 40 CFR 63.640(c) of the rule, storage vessels at existing sources are not subject to the requirements in 40 CFR 63.119(b)(5), (b)(6), (c)(2), and (d)(2) of subpart G. The requirements in 40 CFR 63.119(c)(2) contain equipment specifications for storage tanks with external floating roofs.

EPA is proposing two regulatory options for storage vessels. We believe that either of these options might achieve an ample margin of safety as described in the Benzene NESHAP. The Agency's basis for selecting one of these options in the final rule would reflect our consideration of the relative risk reduction and cost of the options, as well as consideration of other relevant factors as identified in the Benzene NESHAP. For existing storage vessels, Option 1 requires no revisions to the Refinery MACT 1 rule to meet the requirements of CAA section 112(d)(6) and (f)(2). Option 2 would remove the current exemption for the requirements in 40 CFR 63.119(c)(2)(ix) and (x) for slotted guide poles. Removal of this exemption would require the owner or operator of a Group 1 storage vessel at an existing source that is equipped with an external floating roof to equip each slotted guide pole with a gasketed sliding cover or flexible fabric sleeve seal and a gasketed cover or other device which closes off the liquid surface from the atmosphere. The proposed amendments also revise related inspection requirements in 40 CFR 63.646(e) and reporting requirements in 40 CFR 63.654(f)(1)(A)(1), (g)(1), and (g)(3)(iii)(A) to account for the requirements for slotted guide poles.
C. What are the proposed requirements to meet CAA sections 112(f)(2) and (d)(6) for EBU used to treat Group 1 wastewater streams?

EPA is proposing two regulatory options for EBU. We believe that either of these options might achieve an ample margin of safety as described in the Benzene NESHAP. The Agency's basis for selecting one of these options in the final rule would reflect our consideration of the relative risk reduction and cost of the options, as well as consideration of other relevant factors as identified in the Benzene NESHAP.

Option 1 requires no revisions to the Refinery MACT 1 rule to meet the requirements of CAA sections 112(f)(2) and (d)(6). Option 2 for EBU proposes to revise the wastewater provisions in the Refinery MACT 1 rule to add a specific performance standard and monitoring requirement for EBU. The proposed amendments require owners or operators to operate and maintain EBU to achieve a minimum treatment efficiency for benzene of 90 percent. The owner or operator would be required to conduct an initial performance demonstration using the procedures in 40 CFR part 63, appendix C (Determination of the Fraction Biodegraded (Fbio) in a Biological Treatment Unit). Based on the demonstration results, facilities would establish operating limits for the mixed liquor volatile suspended solids (MLVSS) concentration and the foodto microorganism ratio according to the rule requirements. The operating parameters would be monitored at least once a week. Exceedance of an operating limit would be a deviation that must be reported in the periodic (semiannual) report required by 40 CFR 63.654.
D. What are the proposed requirements for cooling towers under CAA sections 112(d)(2) and (f)(2)?

Because the Refinery MACT 1 rule does not address HAP emissions from cooling towers, we are proposing to regulate cooling towers under CAA section 112(d)(2) and (d)(3) in this action. As we are proposing later in the preamble, once cooling towers have been regulated pursuant to CAA section (d)(2) and (d)(3), no additional controls are needed to provide an adequate margin of safety under CAA section (f)(2).

We are proposing work practice standards for cooling towers which would require the owner or operator of a new or existing source to monitor for leaks in the cooling tower return lines from heat exchangers in organic HAP service (i.e., lines that contain or contact fluids with 5 weight percent or greater of total organic HAP listed in Table 1 of the rule) and, where leaks are detected, to repair such leaks within a specified period of time. The two options that are being coproposed differ in the detection methods used to identify leaks for existing sources, and in the frequency of monitoring for new sources. The first option reflects our MACT floor analysis and would reject imposing controls beyond the MACT
[[Page 50722]]
floor. Under this option, the owner or operator of existing source cooling towers receiving cooling water from heat exchangers in organic HAP service would be required to monitor chemical addition rates or other surrogate indicators of leaks. If the surrogate indicators suggest a leak, the owner or operator would conduct sampling and analyses to determine if the indicated leak is an organic HAP leak. For existing sources, an organic HAP leak is defined as an organic HAP concentration in the cooling tower water of 1 ppmw or greater. Owner and operators of new source cooling towers receiving cooling water from heat exchangers in organic HAP service would be required to conduct quarterly sampling and analyses to identify any organic HAP leaks into the cooling tower water and to take appropriate corrective action to fix the leaks.

Under the second option, we would select a control option based on our beyond the floor analysis and would require the owner or operator of new and existing sources to conduct monthly sampling and analyses to identify any organic HAP leaks into the cooling tower water.

Under both options, a leak into the cooling tower water would be defined as either a mass leak rate of 100 pounds of total organic HAP per day or greater or a mass leak rate of 10 pounds of any single organic HAP per day or greater. Under both options, if a leak is detected, the owner or operator would be required to identify the source of the leak as soon as practicable but not later than 30 days after receiving the sampling results. Unless a delay in repair is allowed under the proposed requirements, the owner or operator would be required to repair the leak no later than 30 days after identifying the source of the leak. The proposed rule would allow a delay in repair of the leak if repair of the leak would require the process unit served by the leaking heat exchanger to be shut down, and the shutdown would result in greater emissions than the potential emissions from the cooling tower leak from the time the leaking heat exchanger was first identified and the next planned shutdown. The owner or operator would be required to continue monthly monitoring and repair the heat exchanger within 30 days if sampling results show that the projected emissions from the cooling tower exceed the startup and shutdown emissions estimates. The proposed rule would also allow a delay in repair if the necessary parts are not reasonably available. In this case, the owner or operator would be required to complete the repair as soon as practicable upon receiving the necessary parts, but no later than 120 days after identifying the leaking heat exchanger. All new or existing refineries with a cooling tower system also would be required to prepare and follow a monitoring plan for cooling towers. The plan is necessary to document emissions potential for employing the delay of repair provisions.

E. What other revisions are we proposing?

We are also proposing clarifications to the requirements in the Refinery MACT 1 rule. The proposed amendments clarify that the control requirements for gasoline loading racks apply to Group 1 gasoline loading racks. ``Group 1 gasoline loading rack'' is the term used to define the affected emissions source subject to emissions control requirements. This clarification would amend 40 CFR 63.640 of subpart CC.
F. What is the compliance schedule for the proposed amendments?

The proposed amendments to the Refinery MACT 1 rule would become effective on the date of publication of the final amendments in the Federal Register. Under section 112 (i)(1) of the CAA, any new facility would be required to comply upon startup. For existing sources, CAA section 112(i)(3)(A) requires compliance no later than 3 years after the effective date of the standard. The proposed 3year compliance date is appropriate because it will allow facilities time to perform monitoring and install required controls. For cooling towers, we are allowing 3 years to identify which towers are affected, to identify the ability to repair these cooling towers without a process unit turnaround, to determine the HAP emissions that would occur if a shutdown is required to control a heat exchanger leak, and to establish an appropriate monitoring program that meets the requirements of the proposed rule. For EBU, 3 years is necessary to perform tests of benzene destruction efficiency, to calculate the overall effectiveness of the EBU using the procedures in Appendix C, to establish appropriate monitoring provisions and install and test necessary equipment, and to make modifications to the EBU if necessary to increase the efficiency of the system to meet the proposed requirements. For storage tanks, 3 years are being proposed to allow flexibility in the addition of the guidepole controls for safety and operational concerns. In promulgating similar requirements for storage tanks, we have extended the compliance time until the next scheduled turnaround requiring emptying and degassing of the tank or 10 years, whichever is sooner. This is because the emissions that occur during emptying and degassing exceed the HAP emission reductions that would occur as a result of applying the controls. We are requesting comments on whether it is necessary to empty and degas tanks for retrofitting the proposed controls. IV. Rationale for Proposed Amendments
A. What actions are we proposing under CAA section 112(d)(2)?

We did not establish standards for cooling towers in the Refinery MACT 1 rule. Industry emissions information and data demonstrate that organic HAP emissions from cooling towers at petroleum refineries are significant, and we are proposing to add emissions standards for organic HAP from cooling towers at petroleum refineries under the authority of CAA section 112(d)(2). Because the emissions from cooling towers are not emitted through a stack and are not practically measurable, we have established work practice standards as provided for under CAA section 112(h)(2) to address these emissions.

In evaluating the MACT floor, we must determine the average emissions limitations achieved by the top 12 percent of the affected sources. We have often interpreted the average of the top 12 percent as the performance of the 6th percentile unit. Of the 150 refineries, the 6th percentile is represented by the 9th ranked topperforming unit. Based on available information, we have determined that the top 12 percent of the industry currently implements cooling tower monitoring programs to detect and repair leaks of process fluids into cooling water using chemical usage rates or other surrogate indicators of heat exchanger leaks. Therefore, we have determined that the MACT floor for existing cooling towers is monitoring of surrogate indicators of heat exchanger leaks in cooling water and to repair leaks. The nationwide total annual cost (TAC) to conduct cooling tower monitoring of surrogate indicators and repairs is estimated to be $750,000. This cost includes a product recovery credit of $1.2 million, and includes no costs for repair of heat exchangers under that assumption that refiners would repair leaking heat exchangers when they are made aware of the leak as part of their routine operations. For large leaks, reasons for repairing leaks immediately could be safety concerns or the recovery of large product losses. For smaller leaks, these concerns might not be valid
[[Page 50723]]
and therefore refiners might incur additional costs beyond routine operations. EPA requests comment on the extent to which immediate repairs would be based on these concerns, and on typical costs of repair. The HAP emissions reduction for the MACT floor is estimated to be 373 tpy total HAP and 28.3 tpy of benzene. The HAP baseline for cooling towers was estimated to be 3,024 tpy.

The MACT floor for new sources is represented by the best performing similar unit. Based on all of the information available, the best performance standard currently being implemented is direct organic chemical concentration monitoring of their Refinery MACT 1 cooling towers on a quarterly basis. Based on emissions data for the facility implementing this program, we have determined that the performance of this cooling tower monitoring program would limit leaks into the cooling water to less than 10 lbs/day of a single organic HAP and less than 100 lbs/day of total organic HAP. Therefore, we have determined that the MACT floor for new cooling towers is quarterly organic chemicalspecific monitoring with an action level of 10 lbs/day or greater of a single organic HAP and 100 lbs/day or greater of total organic HAP.

EPA has concluded, based on available data, that existing industry monitoring of surrogate parameters will only detect large leaks, which would miss leaks that would generate significant organic HAP emissions (see memorandum to docket: Cooling towers: Control Options and Impact Estimates). EPA analyzed the amount of HAP that could be emitted from cooling water based on HAP concentration data and flow rates for cooling towers at several petroleum refinery facilities and decided to structure regulatory options to account for variable cooling water flow and minimum detection limit capabilities of 10 parts per billion by weight (ppbw) for the concentrations of individual HAP in water. For example, at a petroleum refinery with total organic HAP concentration of 30 ppbw and a cooling water flow rate of 40,000 gallons per minute (gal/min), the potential organic HAP emissions from the cooling tower are 14 lbs/day or over 2.5 tons if the leak lasted for a year.

As part of our beyond the floor analysis, we considered alternatives more stringent than the MACT floor option for existing and new sources. For existing and new sources, we identified two alternatives that would require monitoring by collecting a cooling water sample and analyzing for speciated HAP. In both alternatives, the cost of the monitoring is likely less than the value of the product that would no longer be lost to the atmosphere. Additionally, we have not included repair costs in any of the options as we considered these costs to be routine operational costs. The costs discussed also apply to new as well as existing sources, since there are no retrofit issues associated with the proposed monitoring program.

One alternative more stringent than the MACT floor includes quarterly monitoring of cooling water by water sampling and a leak definition of greater than or equal to 10 pounds of any single organic HAP or greater than or equal to 100 pounds organic HAP per day and results in a total annualized cost saving of $2.1 million. This savings includes a product recovery credit of $4.4 million. The organic HAP emissions reduction for this alternative regulatory option 1 is 1,330 tpy and the costeffectiveness is $1,600/ton.

Another alternative more stringent than the MACT floor includes monthly monitoring of cooling water by water sampling and a leak definition of greater than or equal to 10 pounds of any single organic HAP or greater than or equal to 100 pounds organic HAP per day. The nationwide TAC is a savings of $1.6 million, including a recovery credit of $5.7 million. The organic HAP emissions reduction for this alternative is 1,720 tpy. The costeffectiveness of this alternative is $920/ton.

EPA is coproposing two options for finalizing MACT standards for new and existing cooling towers. Option 1 represents the MACT floor for new and existing units, as discussed above. Option 2 is more stringent than the MACT floor and is described above as requiring monthly (as opposed to quarterly) monitoring of individual (speciated) organic HAP. Table 1 of this preamble summarizes nationwide impacts of the proposed options.
Table 1.Nationwide Impacts for Cooling Tower Options Product Costeffectiveness ($/ton) Monitoring recovery Total annual HAP emissions Option cost ($1,000) credit cost ($1,000/ (tons/yr HAP) ($1,000/yr) yr) Overall Incremental Baseline................................................ 0 0 0 3,024 0 0 1 (MACT Floor).......................................... 1,990 1,240 750 2,647 1,980 1,980 2 (Beyond the floor).................................... 4,100 5,680 1,590 1,304 920 1,750 Note: The monthly monitoring alternative is projected to result in a positive incremental costeffectiveness of $1,400 per ton (as compared to the quarterly alternative).

This analysis indicates that Option 2 will result in an overall cost savings. Further, the incremental costeffectiveness of Option 2 monitoring compared to Option 1 is a negative $1,750/ton of HAP emissions controlled, which indicates a cost savings above the MACT floor option and is reasonable given these assumptions. However, there are some fundamental assumptions that may affect this analysis, for example, the amount of recovery credit generated by each program is uncertain and we did not consider repair costs or production downtime costs in our analysis. Therefore, we are coproposing Option 1, the MACT floor option, and Option 2 in the event that the costs and feasibility of going beyond the floor are not reasonable. We are requesting comments on this analysis and on these options.

Additionally, under both options, a delay in repair is allowed under the proposed requirements if repair of the leak would require the process unit served by the leaking heat exchanger to be shut down, and the shutdown would result in greater HAP emissions than the projected HAP emissions from the cooling tower leak or if the necessary parts are not reasonably available. We request comments on other possible criteria for delay of repair in addition to these. In addition, we are requesting comments on another option for heat exchanger systems that cannot be repaired without a shutdown that would allow delay of repair until the next unit shutdown. This allowance could be contingent on factors such as the level of HAP emissions from the cooling tower or the duration to the next scheduled shutdown. Finally, we
[[Page 50724]]
request comments on tracking the HAP emissions that occur during the delayed repair and relationship between this monitoring and emission measurement and the reportable quantity requirements under CERCLA. B. How did we estimate residual risk?

EPA modeled available data on the emissions from petroleum refineries to assess the risks associated with petroleum refinery HAP emissions after compliance with the Refinery MACT 1 standard but prior to the proposed MACT amendments for cooling towers. Consistent with previous residual risk assessments, standard air toxics risk assessment practices and principles were used to conduct assessments of potential chronic and acute exposures and risks for both inhalation and non inhalation pathways. In addition, the potential for an adverse environmental effect arising from these sources was also evaluated. Complete documentation for the methods used and results from the risk assessment is available in a report entitled, draft Residual Risk Assessment for MACT 1 Petroleum Refining Sources, which is available in the docket.

Emissions data for 153 petroleum refineries nationwide were developed starting from the EPA's 2002 National Emissions Inventory (NEI), incorporating sitespecific emissions and source information which were provided by the American Petroleum Institute (API) for 22 facilities. The emissions database was published for public comment through an Advanced Notice of Proposed Rulemaking (ANPRM). Comments and corrections to the database received during the public comment period were evaluated by technical reviewers for quality and consistency with engineering data; valid corrections to the database were incorporated for an additional 50 facilities (beyond the 22). No comments or corrections were received on the emissions or source data for 81 facilities.\2\ The 153 refineries included in the database are believed to be all of the sources in the category.
\2\ For an explanation of the corrections we accepted and the corrections we did not accept, see docket.

C. What are the residual risks from petroleum refineries?

Table 2 of this preamble summarizes the results of the inhalation risk assessment. These estimates characterize the lifetime risk of developing cancer or noncancer health effects for individuals living within 50 kilometers (km) of any petroleum refinery.
Table 2.Risk Estimates Due to HAP Exposure Based on 70Year Exposure Duration
Results for refinery MACT Parameter 1 source category Maximum individual lifetime cancer risk (in 1 million).. 70 Maximum hazard index \1\ (chronic respiratory effects).. 0.3 Estimated size of population at risk.................... 90,000,000

greater than 1in1 million......................... 460,000

greater than 10in1 million........................ 6,000

greater than 100in1 million....................... 0 Annual cancer incidence (number of cases per year)...... 0.040.09 \1\ If the hazard index (HI) is calculated to be less than or equal to 1, then no adverse heath effects are expected as a result of the exposure.

We estimate that approximately 90 million people live within 50 km of a refinery. Results from the risk assessment indicate that none of the facilities posed a cancer risk greater than 100in1 million. Approximately 60 percent of the refineries have a maximum individual lifetime cancer risk (MIR) of greater than 1in1 million, and about 14 percent are associated with a MIR greater than 10in1 million. The highest MIR value at any facility is 70in1 million. The cumulative cancer incidence from all MACT 1 refinery emission sources is estimated to be between 0.04 and 0.09 cases per year, or 1 case every 11 to 25 years. Benzene, naphthalene, polycyclic organic matter, and ethylene dibromide emissions are responsible for most of the estimated cancer incidence. Since the benzene cancer unit risk estimate (URE) is reported as a range of values, each end of which is considered to be equally plausible, the range of incidence reflects calculated risks using either end of the range, as well as different methods for extrapolating the risks from subsets of facility emission estimates. Additionally, the maximum noncancer hazard index (HI) associated with emissions from any refinery is estimated to be less than 1. This allows us to conclude that human inhalation exposures to pollution from Refinery MACT 1 sources are without appreciable risk of chronic noncancer health effects, and that direct atmospheric exposures of these pollutants to ecological receptors should not result in any potential environmental impact.

We performed acute screeninglevel assessments of potential acute impacts of concern on each facility and refined those assessments by analyzing aerial photographs of facilities with potential exceedances of acute benchmarks to determine which potential exceedances were truly outside facility boundaries. The results indicated that 12 facilities show a potential to exceed 1hour California acute Reference Exposure Levels (REL) for 3 pollutants (benzene, acrolein, and arsenic). The acute 1hour REL is defined as the concentration level at or below which no adverse health effects are anticipated for a 1hour exposure. Acute REL values are designed to protect the most sensitive individuals in a population by including margins of safety. The highest potential exceedance of any REL was for acrolein, and the REL was exceeded by a factor of 70. Other pollutants showing potential exceedances of the REL value are benzene (exceeded by a factor of 40), and arsenic (exceeded by a factor of 30). In spite of the fact that potential exceedances of these 3 acute REL values are shown by this analysis, none of the facilities investigated showed any potential to exceed available mild 1hour Acute Exposure Guideline Levels (AEGL1) for any of the modeled pollutants. The AEGL1 is the airborne concentration of a substance above which it is predicted that the general population, including susceptible individuals, could experience notable discomfort, irritation, or certain asymptomatic nonsensory effects.

Given the definitions of the acute REL and the AEGL1, it is reasonable to conclude that (1) Health effects in humans could occur as exposures increase above the AEGL1, and (2) exposures below the REL are very unlikely to result in adverse health effects. Potential exposures in between these values (which is what this analysis shows) are more difficult to interpret in terms of health risk. That is, these potential exposures are in the ``gray area'' of uncertainty where the true threshold for adverse effects lies, and thus it is not clear if adverse effects could actually occur at the levels determined by this analysis. Further, we did not refine these results by incorporating actual sitespecific shortterm emission variability into the analysis, so these results are believed to be very conservative and should be interpreted with care.

We also performed a screeninglevel multipathway risk assessment on the
[[Page 50725]]
emissions of mercury, cadmium, lead, and polycyclic aromatic hydrocarbons (PAH), all compounds which are considered to be persistent and bioaccumulative HAP. Based on the results of this screening, noncancer human health risks due to the ingestion of these pollutants were all below levels considered to be without appreciable risk of adverse health effect. One of these pollutants, PAH, showed a potential to cause individual cancer risks as high as 40in1 million, exceeding 1in1 million, but less than 100in1 million. However, because of our inability to accurately speciate and estimate risks for individual compounds within the PAH class, we believe that this result is highly conservative, and that the true risks associated with these PAH are likely to be less than 1in1 million.

For the ecological assessment, two exceedances (cadmium and PAH) of ecological toxicity benchmarks were observed when examining the predicted TRIM.FaTE media concentrations (see Draft Residual Risk Assessment for MACT I Petroleum Refining Sources document). Given the conservative nature of the screening scenario, the results of the comparisons and a review of additional information available on the ecological toxicity of cadmium and PAH, we concluded that it is highly unlikely that these two exceedances are of concern. Overall, the potential for emissions from petroleum refinery sources to result in an adverse environmental impact is likely to be very low for all persistent bioaccumulative HAP emitted.

D. What are the uncertainties in risk assessments?

Uncertainty and the potential for bias are inherent in all risk assessments, including those performed for the petroleum refineries source categories affected by this proposal. A full discussion of uncertainties is found in the Draft Residual Risk Assessment for Petroleum Refining Sources (August 2007), available in the docket.

Although the development of the risk and technology review (RTR) database involved quality assurance/quality control processes, the accuracy of emissions values will vary depending on the source of the data present, incomplete or missing data, errors in estimating emissions values, and other factors. Our review of the data indicates that there may be a low bias in reported emissions for many facilities. It appears that data from several processes and operations are not included in the reported emissions from many facilities. These include exclusion of upset, malfunction, startup, and shutdown events as well as omission of emissions sources that are unexpected, not measured, or not considered in inventories, such as leaks in heat exchanger systems; emissions from process sewers and wastewater systems; fugitive emissions from delayed coking units; and emissions from tank roof landings. Further, the emissions values considered in this analysis are annual totals for a single calendar year (2002) and do not reflect actual fluctuations during the course of the year, as well as variations from year to year. Finally, although we have performed a significant amount of quality control on the data set, for many facilities the physical characteristics (i.e., stack height, physical location) of the reported sources may be inaccurate for detailed risk characterization purposes.

We recently discovered that certain area source location attributes may have been incorrectly incorporated into our atmospheric dispersion simulations, resulting in a positional translation error which may locate certain emission points closer to or farther from potentially exposed populations. While the impact of this error has not been fully evaluated, we believe that it will not dramatically alter the MIR value for the source category, and that it will have very little impact on the total cancer incidence. Nonetheless, we will investigate and correct this error between proposal and promulgation of the final petroleum refineries MACT 1 residual risk decision and

FOR FURTHER INFORMATION CONTACT Mr. Robert Lucas, Office of Air Quality Planning and Standards, Sector Policies and Programs Division, Coatings and Chemicals Group (E14301), Environmental Protection Agency, Research Triangle Park, North Carolina 27711, telephone number (919) 5410884; fax number (919) 5410246; email address:
lucas.bob@epa.gov.