National Emission Standards for Hazardous Air Pollutants: Proposed Standards for Hazardous Air Pollutants for Hazardous Waste Combustors (Phase I Final Replacement Standards and Phase II)

[Federal Register: April 20, 2004 (Volume 69, Number 76)]
[Proposed Rules]
[Page 21197-21246]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr20ap04-25]

-----------------------------------------------------------------------

ENVIRONMENTAL PROTECTION AGENCY
40 CFR Parts 63, 264, 265, 266, 270, and 271
[FRL-7644-1]
RIN 2050-AE01

National Emission Standards for Hazardous Air Pollutants:
Proposed Standards for Hazardous Air Pollutants for Hazardous Waste
Combustors (Phase I Final Replacement Standards and Phase II)

AGENCY: Environmental Protection Agency (EPA).
ACTION: Proposed rule.

-----------------------------------------------------------------------

SUMMARY: This action proposes national emission standards for hazardous
air pollutants (NESHAP) for hazardous waste combustors. These
combustors include hazardous waste burning incinerators, cement kilns,
lightweight aggregate kilns, industrial/commercial/institutional
boilers and process heaters, and hydrochloric acid production furnaces,
known collectively as hazardous waste combustors (HWCs). EPA has
identified these HWCs as major sources of hazardous air pollutant (HAP)
emissions. These proposed standards will, when final, implement section
112(d) of the Clean Air Act (CAA) by requiring hazardous waste
combustors to meet HAP emission standards reflecting the application of
the maximum achievable control technology (MACT).
The HAP emitted by facilities in the incinerator, cement kiln,
lightweight aggregate kiln, industrial/commercial/institutional boiler,
process heater, and hydrochloric acid production furnace source
categories include arsenic, beryllium, cadmium, chromium, dioxins and
furans, hydrogen chloride and chlorine gas, lead, manganese, and
mercury. Exposure to these substances has been demonstrated to cause
adverse health effects such as irritation on the lung, skin, and mucus
membranes, effects on the central nervous system, kidney damage, and
cancer. The adverse health effects associated with the exposure to
these specific HAP are further described in the preamble. In general,
these findings have only been shown with concentrations higher than
those typically in the ambient air.
This action also presents our tentative decision regarding the
February 28, 2002, petition for rulemaking submitted by the Cement Kiln
Recycling Coalition to the Administrator, relating to EPA's
implementation of the so-called omnibus permitting authority under
section 3005(c) of the Resource Conservation and Recovery Act (RCRA),
which requires that each permit issued under RCRA contain such terms
and conditions as are determined necessary to protect human health and
the environment. In that petition, the Cement Kiln Recycling Coalition
requests that we repeal the existing site-specific risk assessment
policy and technical guidance for hazardous waste combustors and that
we promulgate the policy and guidance as rules in accordance with the
Administrative Procedure Act if we continue to believe that site-
specific risk assessments may be necessary.

DATES: Submit comments on or before July 6, 2004.

ADDRESSES: Submit your comments, identified by Docket ID No. OAR-2004-
0022 by one of the following methods:
? Federal eRulemaking Portal: http://www.regulations.gov.
Follow the on-line instructions for submitting comments.
? Agency Web site: http://www.epa.gov/edocket.
EDOCKET, EPA's electronic public docket and comment system, is EPA's
preferred method for receiving comments. Follow the on-line
instructions for submitting comments.
? E-mail: http://www.epa.gov/edocket.
? Fax: 202-566-1741.
? Mail: OAR Docket, Environmental Protection
Agency, Mailcode: B102, 1200 Pennsylvania Ave., NW., Washington, DC
20460. Please include a total of 2 copies.
? Hand Delivery: EPA/DC, EPA West, Room B102, 1301
Constitution Ave., NW., Washington, DC. Such deliveries are only
accepted during the Docket's normal hours of operation, and special
arrangements should be made for deliveries of boxed information.
Instructions: Direct your comments to Docket ID No. OAR-2004-0022.
EPA's policy is that all comments received will be included in the
public docket without change and may be made available online at
http://www.epa.gov/edocket, including any personal information provided,
unless the comment includes information claimed to be Confidential
Business Information (CBI) or other information whose disclosure is
restricted by statute. Do not submit information that you consider to
be CBI or otherwise protected through EDOCKET, regulations.gov, or e-
mail. The EPA EDOCKET and the federal regulations.gov Web sites are
``anonymous access'' systems, which means EPA will not know your
identity or contact information unless you provide it in the body of
your comment. If you send an e-mail comment directly to EPA without
going through EDOCKET or regulations.gov, your e-mail address will be
automatically captured and included as part of the comment that is
placed in the public docket and made available on the Internet. If you
submit an electronic comment, EPA recommends that you include your name
and other contact information in the body of your comment and with any
disk or CD-ROM you submit. If EPA cannot read your comment due to
technical difficulties and cannot contact you for clarification, EPA
may not be able to consider your comment. Electronic files should avoid
the use of special characters, any form of encryption, and be free of
any defects or viruses. For additional information about EPA's public
docket visit EDOCKET on-line or see the Federal Register of May 31,
2002 (67 FR 38102).
For additional instructions on submitting comments, go to unit II
of the SUPPLEMENTARY INFORMATION section of this document.
Docket: All documents in the docket are listed in the EDOCKET index
at http://www.epa.gov/edocket. Although listed in the index, some
information is not publicly available, i.e., CBI or other information
whose disclosure is restricted by statute. Certain other material, such
as copyrighted material, is not placed on the Internet and will be
publicly available only in hard copy form. Publicly available docket
materials are available either electronically in EDOCKET or in hard
copy at the OAR Docket, EPA/DC, EPA West, Room B102, 1301 Constitution
Ave., NW., Washington, DC. The 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 Public Reading Room is (202) 566-1744, and the
telephone number for the OAR Docket is (202) 566-1742.

FOR FURTHER INFORMATION CONTACT: For general information, call the RCRA
Call Center at 1-800-424-9346 or TDD 1-800-553-7672 (hearing impaired).
Callers within the Washington Metropolitan Area must dial 703-412-9810
or TDD 703-412-3323 (hearing impaired). The RCRA Call Center is open
Monday-Friday, 9 a.m. to 4 p.m., eastern standard time. For more
information about this proposal, contact Michael Galbraith at 703-605-
0567, or galbraith.michael@epa.gov.

SUPPLEMENTARY INFORMATION:

I. Regulated Entities

The promulgation of the proposed rule would affect the following North

[[Page 21199]]

American Industrial Classification System (NAICS) and Standard
Industrial Classification (SIC) codes:

----------------------------------------------------------------------------------------------------------------
Examples of potentially
Category NAICS code SIC code regulated entities
----------------------------------------------------------------------------------------------------------------
Any industry that combusts 562211............................. 4953 Incinerator, hazardous
hazardous waste as defined in waste.
the proposed rule.
327310............................. 3241 Cement manufacturing,
clinker production.
327992............................. 3295 Ground or treated
mineral and earth
manufacturing.
325................................ 28 Chemical Manufacturers.
324................................ 29 Petroleum Refiners.
331................................ 33 Primary Aluminum.
333................................ 38 Photographic equipment
and supplies.
488, 561, 562...................... 49 Sanitary Services,
N.E.C.
421................................ 50 Scrap and waste
materials.
422................................ 51 Chemical and Allied
Products, N.E.C.
512, 541, 561, 812................. 73 Business Services,
N.E.C.
512, 514, 541, 711................. 89 Services, N.E.C.
924................................ 95 Air, Water and Solid
Waste Management.
----------------------------------------------------------------------------------------------------------------

This table is not intended to be exhaustive, but rather provides a
guide for readers regarding entities likely to be regulated by this
action. This table lists examples of the types of entries EPA is now
aware could potentially be regulated by this action. Other types of
entities not listed could also be affected. To determine whether your
facility, company, business, organization, etc., is regulated by this
action, you should examine the applicability criteria in Part II of
this preamble. 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.

II. What Should I Consider as I Prepare My Comments for EPA?

1. Submitting CBI. Do not submit this information to EPA through
EDOCKET, regulations.gov or e-mail. 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.
2. Tips for Preparing Your Comments. When submitting comments,
remember to:
A. Identify the rulemaking by docket number and other identifying
information (subject heading, Federal Register date and page number).
B. Follow directions--The agency may ask you to respond to specific
questions or organize comments by referencing a Code of Federal
Regulations (CFR) part or section number.
C. Explain why you agree or disagree; suggest alternatives and
substitute language for your requested changes.
D. Describe any assumptions and provide any technical information
and/or data that you used.
E. If you estimate potential costs or burdens, explain how you
arrived at your estimate in sufficient detail to allow for it to be
reproduced.
F. Provide specific examples to illustrate your concerns, and
suggest alternatives.
G. Explain your views as clearly as possible, avoiding the use of
profanity or personal threats.
H. Make sure to submit your comments by the comment period deadline
identified.

Outline

Part One: Background and Summary

I. Background Information
A. What Criteria Are Used in the Development of NESHAP?
B. What Is the Regulatory Development Background of the Source
Categories in the Proposed Rule?
C. What Is the Statutory Authority for this Standard?
D. What Is the Relationship Between the Proposed Rule and Other
MACT Combustion Rules?
E. What Are the Health Effects Associated with Pollutants
Emitted by Hazardous Waste Combustors?
II. Summary of the Proposed Rule
A. What Source Categories Are Affected by the Proposed Rule?
B. What HAP Are Emitted?
C. Does Today's Proposed Rule Apply to My Source?
D. What Emissions Limitations Must I Meet?
E. What Are the Testing and Initial Compliance Requirements?
F. What Are the Continuous Compliance Requirements?
G. What Are the Notification, Recordkeeping, and Reporting Requirements?

Part Two: Rationale for the Proposed Rule

I. How Did EPA Determine Which Hazardous Waste Combustion Sources
Would Be Regulated?
A. How Are Area Sources Regulated?
B. What Hazardous Waste Combustors Are Not Covered by this Proposal?
C. How Would Sulfuric Acid Regeneration Facilities Be Regulated?
II. What Subcategorization Considerations Did EPA Evaluate?
A. What Subcategorization Options Did We Consider for Incinerators?
B. What Subcategorization Options Did We Consider for Cement Kilns?
C. What Subcategorization Options Did We Consider for
Lightweight Aggregate Kilns?
D. What Subcategorization Options Did We Consider for Boilers?
E. What Subcategorization Options Did We Consider for
Hydrochloric Acid Production Furnaces?
III. What Data and Information Did EPA Consider to Establish the
Proposed Standards?
A. Data Base for Phase I Sources
B. Data Base for Phase II Sources
C. Classification of the Emission Data
D. Invitation to Comment on Data Base
IV. How Did EPA Select the Format for the Proposed Rule?
A. What Is the Rationale for Generally Selecting an Emission
Limit Format Rather than a Percent Reduction Format?
B. What Is the Rationale for Selecting a Hazardous Waste Thermal
Emissions Format for Some Standards, and an Emissions Concentration
Format for Others?
C. What Is the Rationale for Selecting Surrogates to Control Multiple HAP?
D. What Is the Rationale for Requiring Compliance with Operating
Parameter Limits to Ensure Compliance with Emission Standards?

[[Page 21200]]

V. How Did EPA Determine the Proposed Emission Limitations for New
and Existing Units?
A. How Did EPA Determine the Proposed Emission Limitations for New Units?
B. How Did EPA Determine the Proposed Emission Limitations for
Existing Units?
VI. How Did EPA Determine the MACT Floor for Existing and New Units?
A. What MACT Methodology Approaches Are Used to Identify the
Best Performers for the Proposed Floors, and When Are They Applied?
B. How Did EPA Select the Data to Represent Each Source When
Determining Floor Levels?
C. How Did We Evaluate Whether It Is Appropriate to Issue
Separate Emissions Standards for Various Subcategories?
D. How Did We Rank Each Source's Performance Levels to Identify
the Best Performing Sources for the Three MACT Methodologies?
E. How Did EPA Calculate Floor Levels That Are Achievable for
the Average of the Best Performing Sources?
F. Why Did EPA Default to the Interim Standards When Establishing Floors?
G. What Other Options Did EPA Consider?
VII. How Did EPA Determine the Proposed Emission Standards for
Hazardous Waste Burning Incinerators?
A. What Are the Proposed Standards for Dioxin and Furan?
B. What Are the Proposed Standards for Mercury?
C. What Are the Proposed Standards for Particulate Matter?
D. What Are the Proposed Standards for Semivolatile Metals?
E. What Are the Proposed Standards for Low Volatile Metals?
F. What Are the Proposed Standards for Hydrogen Chloride and Chlorine Gas?
G. What Are the Standards for Hydrocarbons and Carbon Monoxide?
H. What Are the Standards for Destruction and Removal Efficiency?
VIII. How Did EPA Determine the Proposed Emission Standards for
Hazardous Waste Burning Cement Kilns?
A. What Are the Proposed Standards for Dioxin and Furan?
B. What Are the Proposed Standards for Mercury?
C. What Are the Proposed Standards for Particulate Matter?
D. What Are the Proposed Standards for Semivolatile Metals?
E. What Are the Proposed Standards for Low Volatile Metals?
F. What Are the Proposed Standards for Hydrogen Chloride and Chlorine Gas?
G. What Are the Standards for Hydrocarbons and Carbon Monoxide?
H. What Are the Standards for Destruction and Removal Efficiency?
IX. How Did EPA Determine the Proposed Emission Standards for
Hazardous Waste Burning Lightweight Aggregate Kilns?
A. What Are the Proposed Standards for Dioxin and Furan?
B. What Are the Proposed Standards for Mercury?
C. What Are the Proposed Standards for Particulate Matter?
D. What Are the Proposed Standards for Semivolatile Metals?
E. What Are the Proposed Standards for Low Volatile Metals?
F. What Are the Proposed Standards for Hydrogen Chloride and Chlorine Gas?
G. What Are the Standards for Hydrocarbons and Carbon Monoxide?
H. What Are the Standards for Destruction and Removal Efficiency?
X. How Did EPA Determine the Proposed Emission Standards for
Hazardous Waste Burning Solid Fuel-Fired Boilers?
A. What Is the Rationale for the Proposed Standards for Dioxin and Furan?
B. What Is the Rationale for the Proposed Standards for Mercury?
C. What Is the Rationale for the Proposed Standards for Particulate Matter?
D. What Is the Rationale for the Proposed Standards for Semivolatile Metals?
E. What Is the Rationale for the Proposed Standards for Low Volatile Metals?
F. What Is the Rationale for the Proposed Standards for Total Chlorine?
G. What Is the Rationale for the Proposed Standards for Carbon
Monoxide or Hydrocarbons?
H. What Is the Rationale for the Proposed Standard for
Destruction and Removal Efficiency?
XI. How Did EPA Determine the Proposed Emission Standards for
Hazardous Waste Burning Liquid Fuel-Fired Boilers?
A. What Are the Proposed Standards for Dioxin and Furan?
B. What Is the Rationale for the Proposed Standards for Mercury?
C. What Is the Rationale for the Proposed Standards for Particulate Matter?
D. What Is the Rationale for the Proposed Standards for Semivolatile Metals?
E. What Is the Rationale for the Proposed Standards for Chromium?
F. What Is the Rationale for the Proposed Standards for Total Chlorine?
G. What Is the Rationale for the Proposed Standards for Carbon
Monoxide or Hydrocarbons?
H. What Is the Rationale for the Proposed Standard for
Destruction and Removal Efficiency?
XII. How Did EPA Determine the Proposed Emission Standards for
Hazardous Waste Burning Hydrochloric Acid Production Furnaces?
A. What Is the Rationale for the Proposed Standards for Dioxin
and Furan?
B. What Is the Rationale for the Proposed Standards for Mercury,
Semivolatile Metals, and Low Volatile Metals?
C. What Is the Rationale for the Proposed Standards for Total Chlorine?
D. What Is the Rationale for the Proposed Standards for Carbon
Monoxide or Hydrocarbons?
E. What Is the Rationale for the Proposed Standard for
Destruction and Removal Efficiency?
XIII. What Is the Rationale for Proposing An Alternative Risk-Based
Standard for Total Chlorine in Lieu of the MACT Standard?
A. What Is the Legal Authority to Establish Risk-Based Standards?
B. What Is the Rationale for the National Exposure Standards?
C. How Would You Determine if Your Total Chlorine Emission Rate
Meets the Eligibility Requirements Defined by the National Exposure Standards?
D. What Is the Rationale for Caps on the Risk-Based Emission Limits?
E. What Would Your Risk-Based Eligibility Demonstration Contain?
F. When Would You Complete and Submit Your Eligibility Demonstration?
G. How Would the Risk-Based HCl-Equivalent Emission Rate Limit
Be Implemented?
H. How Would You Ensure that Your Facility Remains Eligible for
the Risk-Based Emission Limit?
I. Request for Comment on an Alternative Approach: Risk-Based
National Emission Standards
XIV. How Did EPA Determine Testing and Monitoring Requirements for
the Proposed Rule?
A. What Is the Rationale for the Proposed Testing Requirements?
B. What Are the Dioxin/Furan Testing Requirements for Boilers
that Would Not Be Subject to a Numerical Dioxin/Furan Emission Standard?
C. What Are the Proposed Test Methods?
D. What Is the Rationale for the Proposed Continuous Monitoring
Requirements?
E. What Are the Averaging Periods for the Operating Parameter
Limits, and How Are Performance Test Data Averaged to Calculate the Limits?
F. How Would Sources Comply with Emissions Standards Based on
Normal Emissions?
G. How Would Sources Comply with Emission Standards Expressed as
Hazardous Waste Thermal Emissions?
H. What Happens if My Thermal Emissions Standard Limits
Emissions to Below the Detection Limit of the Stack Test Methods?
I. Are We Concerned About Possible Negative Biases Associated
With Making Hydrogen Chloride Measurements in High Moisture Conditions?
J. What Are the Other Proposed Compliance Requirements?
XV. How Did EPA Determine Compliance Times for this Proposed Rule?
XVI. How Did EPA Determine the Required Records and Reports for the
Proposed Rule?
A. Summary of Requirements Currently Applicable to Incinerators,
Cement Kilns, and Lightweight Aggregate Kilns and that Would Be
Applicable to Boilers and Hydrochloric Acid Production Furnaces
B. Why Is EPA Proposing Notification of Intent to Comply and
Compliance Progress Report Requirements?
XVII. What Are the Title V and RCRA Permitting Requirements for
Phase I and Phase II Sources?
A. What Is the General Approach to Permitting Hazardous Waste
Combustion Sources?
B. How Will the Replacement Standards Affect Permitting for
Phase I Sources?
C. What Permitting Requirements Is EPA Proposing for Phase II Sources?

[[Page 21201]]

D. How Would this Proposal Affect the RCRA Site-Specific Risk
Assessment Policy?
XVIII. What Alternatives to the Particulate Matter Standard Is EPA
Proposing or Requesting Comment On?
A. What Alternative to the Particulate Matter Standard Is EPA
Proposing for Incinerators, Liquid Fuel-Fired Boilers, and Solid
Fuel-Fired Boilers?
B. What Alternative to the Particulate Matter Standard Is EPA
Requesting Comment On?
XIX. What Are the Proposed RCRA State Authorization and CAA
Delegation Requirements?
A. What Is the Authority for this Rule?
B. Are There Any Changes to the CAA Delegation Requirements for
Phase I Sources?
C. What Are the Proposed CAA Delegation Requirements for Phase
II Sources?

Part Three: Proposed Revisions to Compliance Requirements

I. Why Is EPA Proposing to Allow Phase I Sources to Conduct the
Initial Performance Test to Comply with the Replacement Rules 12
Months After the Compliance Date?
II. Why Is EPA Requesting Comment on Requirements Promulgated as
Interim Standards or as Final Amendments?
A. Interim Standards Amendments to the Startup, Shutdown, and
Malfunction Plan Requirements
B. Interim Standards Amendments to the Compliance Requirements
for Ionizing Wet Scrubbers
C. Why Is EPA Requesting Comment on the Fugitive Emission Requirements?
D. Why Is EPA Requesting Comment on Bag Leak Detector Sensitivity?
E. Final Amendments Waiving Operating Parameter Limits during
Testing without an Approved Test Plan
III. Why Is EPA Requesting Comment on Issues and Amendments that
Were Previously Proposed?
A. Definition of Research, Development, and Demonstration Source
B. Identification of an Organics Residence Time that Is
Independent of, and Shorter than, the Hazardous Waste Residence Time
C. Why Is EPA Not Proposing to Extend APCD Controls after the
Residence Time Has Expired when Sources Operate under Alternative
Section 112 or 129 Standards?
D. Why Is EPA Proposing to Allow Use of Method 23 as an
Alternative to Method 0023A for Dioxin/Furan?
E. Why Is EPA Not Proposing the ``Matching the Profile''
Alternative Approach to Establish Operating Parameter Limits?
F. Why Is EPA Not Proposing to Allow Extrapolation of OPLs?
G. Why Is EPA Proposing to Delete the Limit on Minimum
Combustion Chamber Temperature for Dioxin/Furan for Cement Kilns?
H. Why Is EPA Requesting Additional Comment on Whether to Add a
Maximum pH Limit for Wet Scrubbers to Control Mercury Emissions?
I. How Is EPA Proposing to Ensure Performance of Electrostatic
Precipitators, Ionizing Wet Scrubbers, and Fabric Filters?
IV. Other Proposed Compliance Revisions
A. What Is the Proposed Clarification to the Public Notice
Requirement for Approved Test Plans?
B. What Is the Proposed Clarification to the Public Notice
Requirement for the Petition to Waive a Performance Test?

Part Four: Impacts of the Proposed Rule

I. What Are the Air Impacts?
II. What Are the Water and Solid Waste Impacts?
III. What Are the Energy Impacts?
IV. What are the Control Costs?
V. Can We Achieve the Goals of the Proposed Rule in a Less Costly Manner?
VI. What are the Economic Impacts?
A. Market Exit Estimates
B. Quantity of Waste Reallocated
C. Employment Impacts
VII. What Are the Benefits of Reductions in Particulate Matter Emissions?
VIII. What are the Social Costs and Benefits of the Proposed Rule?
A. Combustion Market Overview
B. Baseline Specification
C. Analytical Methodology and Findings--Social Cost Analysis
D. Analytical Methodology and Findings--Benefits Assessment
IX. How Does the Proposed Rule Meet the RCRA Protectiveness Mandate?
A. Background
B. Assessment of Risks

Part Five: Administrative Requirements

I. Executive Order 12866: Regulatory Planning and Review
II. Paperwork Reduction Act
III. Regulatory Flexibility Act
IV. Unfunded Mandates Reform Act
V. Executive Order 13132: Federalism
VI. Executive Order 13175: Consultation and Coordination with Indian
Tribal Governments
VII. Executive Order 13045: Protection of Children from
Environmental Health and Safety Risks
VIII. Executive Order 13211: Actions that Significantly Affect
Energy Supply, Distribution, or Use
IX. National Technology Transfer and Advancement Act
X. Executive Order 12898: Federal Actions to Address Environmental
Justice in Minority Populations and Low-Income Populations
XI. Congressional Review

Abbreviations and Acronyms Used in This Document

acfm--actual cubic feet per minute
Btu--British thermal units
CAA--Clean Air Act
CFR--Code of Federal Regulations
DRE--destruction and removal efficiency
dscf--dry standard cubic foot
dscm--dry standard cubic meter
EPA--Environmental Protection Agency
FR--Federal Register
gr/dscf--grains per dry standard cubic foot
HAP--hazardous air pollutant(s)
ICR--Information Collection Request
kg/hr--kilograms per hour
kW-hour--kilo Watt hour
MACT--Maximum Achievable Control Technology
mg/dscm--milligrams per dry standard cubic meter
MMBtu--million British thermal unit
ng/dscm--nanograms per dry standard cubic meter
NESHAP--national emission standards for HAP
ng--nanograms
POHC--principal organic hazardous constituent
ppmv--parts per million by volume
ppmw--parts per million by weight
Pub. L.--Public Law
RCRA--Resource Conservation and Recovery Act
SRE--system removal efficiency
TEQ--toxicity equivalence
ug/dscm--micrograms per dry standard cubic meter
U.S.C.--United States Code

Part One: Background and Summary

I. Background Information

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

1. What Information Is Covered in This Preamble and How Is It Organized?
In this preamble, EPA summarizes the important features of these
proposed standards that apply to hazardous waste burning incinerators,
cement kilns, lightweight aggregate kilns, boilers, and hydrochloric
acid production furnaces, known collectively as HWCs. This preamble
describes: (1) The environmental, energy, and economic impacts of these
proposed standards; (2) the basis for each of the decisions made
regarding the proposed standards; (3) requests public comments on
certain issues; and (4) discusses administrative requirements relative
to this action.
2. Where in the Code of Federal Regulations Will These Standards Be Codified?
The Code of Federal Regulations (CFR) is a codification of the
general and permanent rules published in the Federal Register by the
Executive departments and agencies of the Federal Government. The code
is divided into 50 titles that represent broad areas subject to Federal
regulation. These proposed rules would be published in Title 40,
Protection of the Environment, Part 63, Subpart EEE: National Emission
Standards for Hazardous Air Pollutants From Hazardous Waste Combustors.

[[Page 21202]]

3. What Criteria Are Used in the Development of NESHAP?
Section 112 of the Clean Air Act (CAA) requires EPA to promulgate
regulations for the control of HAP emissions from each source category
listed by EPA under section 112(c). The statute requires the
regulations to reflect the maximum degree of reduction in emissions of
HAP that is achievable taking into consideration the cost of achieving
the emission reduction, any nonair quality health and environmental
impacts, and energy requirements. This level of control is commonly
referred to as MACT (i.e., maximum achievable control technology). The
MACT regulation can be based on the emission reductions achievable
through application of measures, processes, methods, systems, or
techniques including, but not limited to: (1) Reducing the volume of,
or eliminating emissions of, such pollutants through process changes,
substitutions of materials, or other modifications; (2) enclosing
systems or processes to eliminate emissions; (3) collecting, capturing,
or treating such pollutants when released from a process, stack,
storage or fugitive emission point; (4) design, equipment, work
practices, or operational standards as provided in subsection 112(h);
or (5) a combination of the above. See section 112(d)(2) of the CAA.
For new sources, MACT standards cannot be less stringent than the
emission control achieved in practice by the best-controlled similar
source. See section 112(d)(3) of the Act. 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 for
categories and subcategories with 30 or more sources, or the best-
performing 5 sources for categories or subcategories with fewer than 30
sources. Id. This level of control is usually referred to as the MACT
``floor'', the term used in the Legislative History.
In essence, MACT standards ensure that all major sources of air
toxic (i.e., HAP) emissions achieve the level of control already being
achieved by the better-controlled and lower-emitting sources in each
category. This approach provides assurance to citizens that each major
source of toxic air pollution will be required to effectively control
its emissions of air toxics. At the same time, this approach provides a
level playing field, ensuring that facilities that employ cleaner
processes and good emission controls are not disadvantaged relative to
competitors with poorer controls.

B. What Is the Regulatory Development Background of the Source
Categories in the Proposed Rule?

Today's notice proposes standards for controlling emissions of HAP
from hazardous waste combustors. Hazardous waste combustors comprise
several categories of sources that burn hazardous waste: incinerators,
cement kilns, lightweight aggregate kilns, boilers and hydrochloric
acid production furnaces. We call incinerators, cement kilns, and
lightweight aggregate kilns Phase I sources because we have already
promulgated standards for those source categories. We call boilers and
hydrochloric acid production furnaces Phase II sources because we
intended to promulgate MACT standards for those source categories after
promulgating MACT standards for Phase I sources. The regulatory
background of Phase I and Phase II source categories is discussed below.
1. Phase I Source Categories
Phase I combustor sources are regulated under the Resource
Conservation and Recovery Act (RCRA), which establishes a ``cradle-to-
grave'' regulatory structure overseeing the safe treatment, storage,
and disposal of hazardous waste. We issued RCRA rules to control air
emissions from incinerators in 1981, 40 CFR parts 264 and 265, subpart
O, and from cement kilns and lightweight aggregate kilns that burn
hazardous waste in 1991, 40 CFR part 266, subpart H. These rules rely
generally on risk-based standards to achieve the RCRA protectiveness mandate.
The Phase I source categories are also subject to standards under
section 112(d) of the Clean Air Act. We promulgated standards for Phase
I sources on September 30, 1999 (64 FR 52828). This final rule is
referred to as the Phase I rule or 1999 final rule. These emission
standards created a technology-based national cap for hazardous air
pollutant emissions from the combustion of hazardous waste in these
devices. The rule regulates emissions of numerous hazardous air
pollutants: dioxin/furans, other toxic organics (through surrogates),
mercury, other toxic metals (both directly and through a surrogate),
and hydrogen chloride and chlorine gas. Where necessary, section
3005(c)(3) of RCRA provides the authority to impose additional
conditions in a RCRA permit to protect human health and the environment.
A number of parties, representing interests of both industrial
sources and of the environmental community, sought judicial review of
the Phase I rule. On July 24, 2001, the United States Court of Appeals
for the District of Columbia Circuit (the Court) granted portions of
the Sierra Club's petition for review and vacated the challenged
portions of the standards. Cement Kiln Recycling Coalition v. EPA, 255
F. 3d 855 (D.C. Cir. 2001). The Court held that EPA had not
demonstrated that its calculation of MACT floors met the statutory
requirement of being no less stringent than (1) the average emission
limitation achieved by the best performing 12 percent of existing
sources and (2) the emission control achieved in practice by the best
controlled similar source for new sources. 255 F.3d at 861, 865-66. As
a remedy, the Court, after declining to rule on most of the issues
presented in the industry petitions for review, vacated the
``challenged regulations,'' stating that: ``[W]e have chosen not to
reach the bulk of industry petitioners' claims, and leaving the
regulations in place during remand would ignore petitioners'
potentially meritorious challenges.'' Id. at 872. Examples of the
specific challenges the Court indicated might have merit were
provisions relating to compliance during start up/shut down and
malfunction events, including emergency safety vent openings, the
dioxin/furan standard for lightweight aggregate kilns, and the
semivolatile metal standard for cement kilns. Id. However, the Court
stated, ``[b]ecause this decision leaves EPA without standards
regulating [hazardous waste combustor]
emissions, EPA (or any of the
parties to this proceeding) may file a motion to delay issuance of the
mandate to request either that the current standards remain in place or
that EPA be allowed reasonable time to develop interim standards.'' Id.
Acting on this invitation, all parties moved the Court jointly to
stay the issuance of its mandate for four months to allow EPA time to
develop interim standards, which would replace the vacated standards
temporarily, until final standards consistent with the Court's mandate
are promulgated. The interim standards were published on February 13,
2002 (67 FR 6792). EPA did not justify or characterize these standards
as conforming to MACT, but rather as an interim measure to prevent the
adverse environmental and other consequences that would result from the
regulatory gap resulting from no standards being in place. Id. at 6795-96.
The motion also indicates that EPA will issue final standards which
comply

[[Page 21203]]

with the Court's opinion by June 14, 2005, and it indicates that EPA
and Petitioner Sierra Club intend to enter into a settlement agreement
requiring us to promulgate final rules by that date, and that date be
judicially enforceable. EPA and Sierra Club entered into that
settlement agreement on March 4, 2002.
The joint motion also details other actions we agreed to take,
including issuing a one-year extension to the September 30, 2002,
compliance date (66 FR 63313, December 6, 2001), and promulgating
several of the compliance and implementation amendments to the rule
which we proposed on July 3, 2001 (66 FR 35126). These final amendments
were published on February 14, 2002 (67 FR 6968).
2. Phase II Source Categories
Phase II combustors--boilers and hydrochloric acid production
furnaces--are also regulated under the Resource Conservation and
Recovery Act (RCRA) pursuant to 40 CFR part 266, subpart H, and (for
reasons discussed below) are also subject to the MACT standard setting
process in section 112(d) of the CAA. We delayed promulgating MACT
standards for these source categories pending reevaluation of the MACT
standard setting methodology following the Court's decision to vacate
the standards for the Phase I source categories. We have also entered
into a judicially enforceable consent decree with Sierra Club which
requires EPA to promulgate MACT standards for the Phase II sources by
June 14, 2005--the same date that (for independent reasons) is required
for the replacement standards for Phase I sources.

C. What Is the Statutory Authority for This Standard?

Section 112 of the Clean Air Act requires that the EPA promulgate
regulations requiring the control of HAP emissions from major and
certain area sources. The control of HAP is achieved through
promulgation of emission standards under sections 112(d) and (in a
second round of standard setting) (f) and, in appropriate
circumstances, work practice standards under section 112(h).
EPA's initial list of categories of major and area sources of HAP
selected for regulation in accordance with section 112(c) of the Act
was published in the Federal Register on July 16, 1992 (57 FR 31576).
Incinerators, cement kilns, lightweight aggregate kilns, industrial/
commercial/institutional boilers and process heaters, and hydrochloric
acid production furnaces are among the listed 174 categories of
sources. The listing was based on the Administrator's determination
that they may reasonably be anticipated to emit several of the 188
listed HAP in quantities sufficient to designate them as major sources.

D. What Is the Relationship Between the Proposed Rule and Other MACT
Combustion Rules?

The proposed amendments to the subpart EEE, part 63, standards for
hazardous waste combustors would apply to the source categories that
are currently subject to that subpart--incinerators, cement kilns, and
lightweight aggregate kilns that burn hazardous waste. Today's proposed
rule, however, would also amend subpart EEE to establish MACT standards
for the Phase II source categories--those boilers and hydrochloric acid
production furnaces that burn hazardous waste.
Generally speaking, you are an affected source pursuant to subpart
EEE if you combust, or have previously combusted, hazardous waste in an
incinerator, cement kiln, lightweight aggregate kiln, boiler, or
hydrochloric acid production furnace. You continue to be an affected
source until you cease burning hazardous waste and initiate closure
requirements pursuant to RCRA. See Sec. 63.1200(b). If you never
previously combusted hazardous waste, or have ceased burning hazardous
waste and initiated RCRA closure requirements, you are not subject to
subpart EEE. Rather, EPA has promulgated or proposed separate MACT
standards for sources that do not burn hazardous waste within the
following source categories: commercial and industrial solid waste
incinerators (40 CFR part 60, subparts CCCC and DDDD); Portland cement
manufacturing facilities (40 CFR part 63, subpart LLL); industrial/
commercial/institutional boilers and process heaters (40 CFR part 63,
proposed subpart DDDDD); and hydrochloric acid production facilities
(40 CFR part 63, subpart NNNNN). In addition, EPA considered whether to
establish MACT standards for lightweight aggregate manufacturing
facilities that do not burn hazardous waste, and determined that they
are not major sources of HAP emissions. Thus, EPA has not established
MACT standards for lightweight aggregate manufacturing facilities that
do not burn hazardous waste.
Note that non-stack emissions points are not regulated under
subpart EEE.\1\ Emissions attributable to storage and handling of
hazardous waste prior to combustion (i.e., emissions from tanks,
containers, equipment, and process vents) would continue to be
regulated pursuant to either RCRA subpart AA, BB, and CC or an
applicable MACT that applies to the before-mentioned material handling
devices. Emissions unrelated to the hazardous waste operations may be
regulated pursuant to other MACT rulemakings. For example, Portland
cement manufacturing facilities that combust hazardous waste are
subject to both subpart EEE and subpart LLL, and hydrochloric acid
production facilities that combust hazardous waste may be subject to
both subpart EEE and subpart NNNNN.\2\ In these instances subpart EEE
controls HAP emissions from the cement kiln and hydrochloric acid
production furnace stack, while subparts LLL and NNNNN would control
HAP emissions from other operations that are not directly related to
the combustion of hazardous waste (e.g., clinker cooler emissions for
cement production facilities, and hydrochloric acid product
transportation and storage for hydrochloric acid production facilities).
---------------------------------------------------------------------------

\1\ Note, however, that fugitive emissions attributable to the
combustion of hazardous waste from the combustion device are
regulated pursuant to subpart EEE.
\2\ Hydrochloric acid production furnaces that combust hazardous
waste would also be affected sources subject to subpart NNNNN if
they produce a liquid acid product that contains greater than 30%
hydrochloric acid.
---------------------------------------------------------------------------

Note that if you temporarily cease burning hazardous waste for any
reason, you remain an affected source and are still subject to the
applicable Subpart EEE requirements. However, even as an affected
source, the proposed emission standards or operating limits derived
from the hazardous waste combustors do not apply if: (1) Hazardous
waste is not in the combustion chamber and you elect to comply with
other MACT (or CAA section 129) standards that otherwise would be
applicable if you were not burning hazardous waste, e.g., the
nonhazardous waste burning Portland Cement Kiln MACT (subpart LLL); or
(2) you are in a startup, shutdown, or malfunction mode of operation.

E. What Are the Health Effects Associated With Pollutants Emitted by
Hazardous Waste Combustors?

Today's proposed rule protects air quality and promotes the public
health by reducing the emissions of some of the HAP listed in section
112(b)(1) of the CAA. Emissions data collected in the development of
this proposed rule show that metals, particulate matter, hydrogen
chloride and chlorine gas, dioxins and furans, and other organic
compounds are emitted from hazardous waste combustors. The HAP that would

[[Page 21204]]

be controlled with this rule are associated with a variety of adverse
health affects. These adverse health effects include chronic health
disorders (e.g., irritation of the lung, skin, and mucus membranes and
effects on the blood, digestive tract, kidneys, and central nervous
system), and acute health disorders (e.g., lung irritation and
congestion, alimentary effects such as nausea and vomiting, and effects
on the central nervous system). Provided below are brief descriptions
of risks associated with HAP that are emitted from hazardous waste
combustors. Note that a more detailed discussion of the risks
associated with these emissions is included in Part Four.
Antimony
Antimony occurs at very low levels in the environment, both in the
soils and foods. Higher concentrations, however, are found at antimony
processing sites, and in their hazardous wastes. The most common
industrial use of antimony is as a fire retardant in the form of
antimony trioxide. Chronic occupational exposure to antimony (generally
antimony trioxide) is most commonly associated with ``antimony
pneumoconiosis,'' a condition involving fibrosis and scarring of the
lung tissues. Studies have shown that antimony accumulates in the lung
and is retained for long periods of time. Effects are not limited to
the lungs, however, and myocardial effects (effects on the heart
muscle) and related effects (e.g., increased blood pressure, altered
EKG readings) are among the best-characterized human health effects
associated with antimony exposure. Reproductive effects (increased
incidence of spontaneous abortions and higher rates of premature
deliveries) have been observed in female workers exposed in antimony
processing facilities. Similar effects on the heart, lungs, and
reproductive system have been observed in laboratory animals.
EPA recently assessed the carcinogenicity of antimony and found the
evidence for carcinogenicity to be weak, with conflicting evidence from
inhalation studies with laboratory animals, equivocal data from the
occupational studies, negative results from studies of oral exposures
in laboratory animals, and little evidence of mutagenicity or
genotoxicity.\3\ As a consequence, EPA concluded that insufficient data
are available to adequately characterize the carcinogenicity of
antimony and, accordingly, the carcinogenicity of antimony cannot be
determined based on available information. However, IARC (International
Agency for Research on Cancer) in an earlier evaluation, concluded that
antimony trioxide is ``possibly carcinogenic to humans'' (Group 2B).
---------------------------------------------------------------------------

\3\ See ``Evaluating the Carcinogenicity of Antimony,'' Risk
Assessment Issue Paper (98-030/07-26-99), Superfund Technical
Support Center, National Center for Environmental Assessment, July
26, 1999.
---------------------------------------------------------------------------

Arsenic
Acute (short-term) high-level inhalation exposure to arsenic dust
or fumes has resulted in gastrointestinal effects (nausea, diarrhea,
abdominal pain), and central and peripheral nervous system disorders.
Chronic (long-term) inhalation exposure to inorganic arsenic in humans
is associated with irritation of the skin and mucous membranes. Human
data suggest a relationship between inhalation exposure of women
working at or living near metal smelters and an increased risk of
reproductive effects, such as spontaneous abortions. Inorganic arsenic
exposure in humans by the inhalation route has been shown to be
strongly associated with lung cancer, while ingestion or inorganic
arsenic in humans has been linked to a form of skin cancer and also to
bladder, liver, and lung cancer. EPA has classified inorganic arsenic
as a Group A, human carcinogen.
Beryllium
Beryllium is a hard, grayish metal naturally found in minerals,
rocks, coal, soil, and volcanic dust. Beryllium dust enters the air
from burning coal and oil. This beryllium dust will eventually settle
over the land and water. It enters water from erosion of rocks and
soil, and from industrial waste. Some beryllium compounds will dissolve
in water, but most stick to particles and settle to the bottom. Most
beryllium in soil does not dissolve in water and remains bound to soil.
Beryllium does not accumulate in the food chain.
Beryllium can be harmful if you breathe it. The effects depend on
how much you are exposed to and for how long. If beryllium air levels
are high enough, an acute condition can result. This condition
resembles pneumonia and is called acute beryllium disease. Long-term
exposure to beryllium can increase the risk of developing lung cancer.
Cadmium
The acute (short-term) effects of cadmium inhalation in humans
consist mainly of effects on the lung, such as pulmonary irritation.
Chronic (long-term) inhalation or oral exposure to cadmium leads to a
build-up of cadmium in the kidneys that can cause kidney disease.
Cadmium has been shown to be a developmental toxicant in animals,
resulting in fetal malformations and other effects, but no conclusive
evidence exists in humans. An association between cadmium exposure and
an increased risk of lung cancer has been reported from human studies,
but these studies are inconclusive due to confounding factors. Animal
studies have demonstrated an increase in lung cancer from long-term
inhalation exposure to cadmium. EPA has classified cadmium as a Group
B1, probable carcinogen.
Chlorine Gas
Acute exposure to high levels of chlorine in humans can result in
chest pain, vomiting, toxic pneumonitis, and pulmonary edema. At lower
levels chlorine is a potent irritant to the eyes, the upper respiratory
tract, and lungs. Chronic exposure to chlorine gas in workers has
resulted in respiratory effects including eye and throat irritation and
airflow obstruction. Animal studies have reported decreased body weight
gain, eye and nose irritation, nonneoplastic nasal lesions, and
respiratory epithelial hyperplasia from chronic inhalation exposure to
chlorine. No information is available on the carcinogenic effects of
chlorine in humans from inhalation exposure. We have not classified
chlorine for potential carcinogenicity.
Chromium
Chromium may be emitted in two forms, trivalent chromium (chromium
III) or hexavalent chromium (chromium VI). The respiratory tract is the
major target organ for chromium VI toxicity, for acute (short-term) and
chronic (long-term) inhalation exposures. Shortness of breath,
coughing, and wheezing have been reported from acute exposure to
chromium VI, while perforations and ulcerations of the septum,
bronchitis, decreased pulmonary function, pneumonia, and other
respiratory effects have been noted from chronic exposure. Limited
human studies suggest that chromium VI inhalation exposure may be
associated with complications during pregnancy and childbirth, while
animal studies have not reported reproductive effects from inhalation
exposure to chromium VI. Human and animal studies have clearly
established that inhaled chromium VI is a carcinogen, resulting in an
increased risk of lung cancer. EPA has classified chromium VI as a
Group A, human carcinogen.
Chromium III is less toxic than chromium VI. The respiratory tract
is also the major target organ for

[[Page 21205]]

chromium III toxicity, similar to chromium VI. Chromium III is an
essential element in humans, with a daily intake of 50 to 200
micrograms per day recommended for an adult. The body can detoxify some
amount of chromium VI to chromium III. EPA has not classified chromium
III with respect to carcinogenicity.
Cobalt
Cobalt is a relatively rare metal that is produced primarily as a
by-product during refining of other metals, primarily copper. Cobalt
has been widely reported to cause respiratory effects in humans exposed
by inhalation, including respiratory irritation, wheezing, asthma, and
pneumonia. Cardiomyopathy (or damage to the heart muscle) has also been
reported, although this effect is better known from oral exposure.
Other effects of oral exposure in humans are polycythemia (an
abnormally high number of red blood cells) and the blocking of uptake
of iodine by the thyroid. In addition, cobalt is a sensitizer in humans
by any route of exposure. Sensitized individuals may react to
inhalation of cobalt by developing asthma or to ingestion or dermal
contact with cobalt by developing dermatitis. Cobalt is a vital
component of vitamin B₁₂, though there is no evidence that
intake of cobalt is ever limiting in the human diet.
A number of epidemiological studies have found that exposures to
cobalt are associated with an increased incidence of lung cancer in
occupational settings. The International Agency for Research on Cancer
(IARC, part of the World Health Organization) classifies cobalt and
cobalt compounds as ``possibly carcinogenic to humans'' (Group 2B). The
American Conference of Governmental Industrial Hygienists (ACGIH) has
classified cobalt as a confirmed animal carcinogen with unknown
relevance to humans (category A3). An EPA assessment concludes that
under EPA's 1986 guidelines, cobalt would be classified as a probable
human carcinogen (group B1) based on limited evidence of
carcinogenicity in humans and sufficient evidence of carcinogenicity in
animals, as evidenced by an increased incidence of alveolar/bronchiolar
tumors in recent studies of both rats and mice. Under EPA's proposed
cancer guidelines, cobalt is considered likely to be carcinogenic to
humans.\4\
---------------------------------------------------------------------------

\4\ See ``Derivation of a Provisional Carcinogenicity Assessment
for Cobalt and Compounds,'' Risk Assessment Issue Paper (00-122/1-
15-02), Superfund Technical Support Center, National Center for
Environmental Assessment, January 15, 2002.
---------------------------------------------------------------------------

Dioxins and Furans
Exposures to 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) at
levels 10 times or less above those modeled to approximate average
background exposure have resulted in adverse non-cancer health effects
in animals. These effects include changes in hormone systems,
alterations in fetal development, reduced reproductive capacity, and
immunosuppression. Effects that may be linked to dioxin and furan
exposures at low dose in humans include changes in markers of early
development and hormone levels. Dioxin and furan exposures are
associated with altered liver function and lipid metabolism changes in
activity of various liver enzymes, depression of the immune system, and
endocrine and nervous system effects. EPA in its 1985 dioxin assessment
classified 2,3,7,8-TCDD as a probable human carcinogen. The
International Agency for Research on Cancer (IARC) concluded in 1997
that the overall weight of the evidence was sufficient to characterize
2,3,7,8-TCDD as a known human carcinogen.\5\ In 2001 the U.S.
Department of Health and Human Services National Toxicology Program in
their 9th Report on Carcinogens classified 2,3,7,8-TCDD as a known
human carcinogen.\6\
---------------------------------------------------------------------------

\5\ IARC (International Agency for Research on Cancer). (1997)
IARC monographs on the evaluation of carcinogenic risks to humans.
Vol. 69. Polychlorinated dibenzo-para-dioxins and polychlorinated
dibenzofurans. Lyon, France.
\6\ The U.S. Department of Health and Human Services, National
Toxicology Program 9th Report on Carcinogens, Revised January 2001.
---------------------------------------------------------------------------

Hydrogen Chloride/Hydrochloric Acid
Hydrogen chloride, also called hydrochloric acid, is corrosive to
the eyes, skin, and mucous membranes. Acute (short-term) inhalation
exposure may cause eye, nose, and respiratory tract irritation and
inflammation and pulmonary edema in humans. Chronic (long-term)
occupational exposure to hydrochloric acid has been reported to cause
gastritis, bronchitis, and dermatitis in workers. Prolonged exposure to
low concentrations may also cause dental discoloration and erosion. No
information is available on the reproductive or developmental effects
of hydrochloric acid in humans. In rats exposed to hydrochloric acid by
inhalation, altered estrus cycles have been reported in females and
increased fetal mortality and decreased fetal weight have been reported
in offspring. EPA has not classified hydrochloric acid for carcinogenicity.
Lead
Lead is a very toxic element, causing a variety of effects at low
dose levels. Brain damage, kidney damage, and gastrointestinal distress
may occur from acute (short-term) exposure to high levels of lead in
humans. Chronic (long-term) exposure to lead in humans results in
effects on the blood, central nervous system (CNS), blood pressure, and
kidneys. Children are particularly sensitive to the chronic effects of
lead, with slowed cognitive development, reduced growth and other
effects reported. Reproductive effects, such as decreased sperm count
in men and spontaneous abortions in women, have been associated with
lead exposure. The developing fetus is at particular risk from maternal
lead exposure, with low birth weight and slowed postnatal
neurobehavioral development noted. Human studies are inconclusive
regarding lead exposure and cancer, while animal studies have reported
an increase in kidney cancer from lead exposure by the oral route. EPA
has classified lead as a Group B2, probable human carcinogen.
Manganese
Health effects in humans have been associated with both
deficiencies and excess intakes of manganese. Chronic (long-term)
exposure to low levels of manganese in the diet is considered to be
nutritionally essential in humans, with a recommended daily allowance
of 2 to 5 milligrams per day (mg/d). Chronic exposure to high levels of
manganese by inhalation in humans results primarily in central nervous
system (CNS) effects. Visual reaction time, hand steadiness, and eye-
hand coordination were affected in chronically-exposed workers.
Manganism, characterized by feelings of weakness and lethargy, tremors,
a mask-like face, and psychological disturbances, may result from
chronic exposure to higher levels. Impotence and loss of libido have
been noted in male workers afflicted with manganism attributed to
inhalation exposures. EPA has classified manganese in Group D, not
classifiable as to carcinogenicity in humans.
Mercury
Mercury exists in three forms: elemental mercury, inorganic mercury
compounds (primarily mercuric chloride), and organic mercury compounds
(primarily methyl mercury). Each form exhibits different health
effects. Various sources may release elemental or inorganic mercury;
environmental methyl mercury is

[[Page 21206]]

typically formed by biological processes after mercury has precipitated
from the air.
Acute (short-term) exposure to high levels of elemental mercury in
humans results in central nervous system (CNS) effects such as tremors,
mood changes, and slowed sensory and motor nerve function. High
inhalation exposures can also cause kidney damage and effects on the
gastrointestinal tract and respiratory system. Chronic (long-term)
exposure to elemental mercury in humans also affects the CNS, with
effects such as increased excitability, irritability, excessive
shyness, and tremors. EPA has not classified elemental mercury with
respect to cancer.
Acute exposure to inorganic mercury by the oral route may result in
effects such as nausea, vomiting, and severe abdominal pain. The major
effect from chronic exposure to inorganic mercury is kidney damage.
Reproductive and developmental animal studies have reported effects
such as alterations in testicular tissue, increased embryo resorption
rates, and abnormalities of development. Mercuric chloride (an
inorganic mercury compound) exposure has been shown to result in
forestomach, thyroid, and renal tumors in experimental animals. EPA has
classified mercuric chloride as a Group C, possible human carcinogen.
Nickel
Nickel is a commonly used industrial metal, and is frequently
associated with iron and copper ores. Contact dermatitis is the most
common effect in humans from exposure to nickel, whether via
inhalation, oral, or dermal exposure. Cases of nickel-contact
dermatitis have been reported following occupational and non-
occupational exposure, with symptoms of itching of the fingers, wrists,
and forearms. Many studies have also demonstrated dermal effects in
sensitive humans from ingested nickel, invoking an eruption or
worsening of eczema. Chronic inhalation exposure to nickel in humans
results in direct respiratory effects, such as asthma due to primary
irritation, or an allergic response and an increased risk of chronic
respiratory tract infections.
Animal studies have reported a variety of inflammatory effects on
the lungs, as well as effects on the kidneys and immune system from
inhalation exposure to nickel. Significant differences in inhalation
toxicity among the various forms of nickel have been documented, with
soluble nickel compounds being more toxic to the respiratory tract than
less soluble compounds (e.g., nickel oxide). Animal studies have also
reported effects on the respiratory and gastrointestinal systems,
heart, blood, liver, kidney, and body weight from oral exposure to
nickel, as well as to the fetus.
EPA currently classifies nickel refinery dust and nickel subsulfide
(a major component of nickel refinery dust) as class A human
carcinogens based on increased risks of lung and nasal cancer in human
epidemiological studies of occupational exposures to nickel refinery
dust, increased tumor incidences in animals by several routes of
administration in several animal species, and positive results in
genotoxicity assays. More recently, a pair of inhalation studies
performed under the auspices of the National Toxicology Program (NTP)
of the National Institutes of Health concluded that there was no
evidence of carcinogenic activity of soluble nickel salts in rats or
mice and that there was some evidence of carcinogenic activity of
nickel oxide in male and female rats based on increased incidence of
alveolar/bronchiolar adenoma or carcinoma and increased incidence of
benign or malignant pheochromocytoma (a tumor of the adrenal gland) and
equivocal evidence in mice based on marginally increased incidence of
alveolar/bronchiolar adenoma or carcinoma in females and no evidence in
males. The Tenth Annual Report on Carcinogens classifies nickel
compounds as ``known to be human carcinogens.'' \7\ This is consistent
with the International Agency for Cancer Research (IARC) which
classifies nickel compounds as Group 1 human carcinogens.
---------------------------------------------------------------------------

\7\ Report on Carcinogens, Tenth Edition; U.S. Department of
Health and Human Services, Public Health Service, National
Toxicology Program, December 2002.
---------------------------------------------------------------------------

Organic HAP
Organic HAPs include halogenated and nonhalogenated organic classes
of compounds such as polycyclic aromatic hydrocarbons (PAHs) and
polychlorinated biphenyls (PCBs). Both PAHs and PCBs are classified as
potential human carcinogens, and are considered toxic, persistent and
bioaccumulative. They include compounds such as benzene, methane,
propane, chlorinated alkanes and alkenes, phenols and chlorinated
aromatics. Adverse health effects of HAPs include damage to the immune
system, as well as neurological, reproductive, developmental,
respiratory and other health problems.
Particulate Matter \8\
---------------------------------------------------------------------------

\8\ The discussion of PM effects is drawn from the executive
summary of the ``Fourth External Review Draft of Air Quality
Criteria for Particulate Matter,'' National Center for Environmental
Assessment, Office of Research and Development, U.S. Environmental
Protection Agency, EPA/600/P-99/002aD, June, 2003.
---------------------------------------------------------------------------

Atmospheric PM is composed of sulfate, nitrate, ammonium, and other
ions, elemental carbon, particle-bound water, a wide variety of organic
compounds, and a large number of elements contained in various
compounds, some of which originate from crustal materials and others
from combustion sources. Combustion sources are the primary origin of
trace metals found in fine particles in the atmosphere. Ambient PM can
be of primary or secondary origin.\9\
---------------------------------------------------------------------------

\9\ Secondary PM is not emitted directly but is formed in the
atmosphere by gas phase or aqueous phase reactions of emissions of
various precursor compounds.
---------------------------------------------------------------------------

[[Page 21207]]

wet and dry deposition.\10\ Acidification of surface waters can have
long-term adverse effects on aquatic ecosystems, including effects on
fish populations, macro invertebrates, species richness, and
zooplankton abundance. In the soil environment, acid deposition has the
potential to inhibit nutrient uptake, alter the ecological processes of
energy flow and nutrient cycling, change ecosystem structure, and
affect ecosystem biodiversity. In addition, ambient fine particles are
well known as the major cause of visibility impairment. Visibility
impairment (or haziness) is widespread in the U.S. and is greatest in
the eastern United States and southern California. In addition, PM
exerts important effects on materials, such as soiling, corrosion, and
degradation of surfaces, and accelerates weathering of man-made and
natural materials.
---------------------------------------------------------------------------

\10\ Nitrates and sulfates in PM are derived primarily from
emissions of SO_X and NO_X.
---------------------------------------------------------------------------

A large body of evidence exists from epidemiological studies that
demonstrates a relationship between ambient particulate matter (PM) and
mortality and morbidity in the general population and, when combined
with evidence from other studies (e.g., clinical and animal studies),
indicates that exposure to PM is a probable contributing cause to the
adverse human health effects that have been observed. For example, many
different studies report that increased cardiovascular and respiratory-
related mortality risks are significantly associated with various
measures (both long-term and short-term) of ambient PM. Some studies
suggest that a portion of the increased mortality may be associated
with concurrent exposures to PM and other criteria pollutants, such as
SO₂. Much evidence exists of positive associations between
ambient PM concentrations and increased respiratory-related hospital
admissions, emergency room, and other medical visits. Additional
findings implicate PM as likely associated with an increased occurrence
of chronic bronchitis and a contributing factor in the exacerbation of
asthmatic conditions. Recent reports from prospective cohort studies of
long-term ambient PM exposures provide substantial evidence of an
association between increased risk of lung cancer and PM, especially
exposure to fine PM or its components.
PM has other effects, beyond the health effects to human beings.
The major effect of atmospheric PM on ecosystems is indirect and occurs
through the deposition of nitrates and sulfates and the acidifying
effects of the associated hydrogen ions contained in wet and dry
deposition.\11\ Acidification of surface waters can have long-term
adverse effects on aquatic ecosystems, including effects on fish
populations, macro invertebrates, species richness, and zooplankton
abundance. In the soil environment, acid deposition has the potential
to inhibit nutrient uptake, alter the ecological processes of energy
flow and nutrient cycling, change ecosystem structure, and affect
ecosystem biodiversity. In addition, ambient fine particles are well
known as the major cause of visibility impairment. Visibility
impairment (or haziness) is widespread in the U.S. and is greatest in
the eastern United States and southern California. In addition, PM
exerts important effects on materials, such as soiling, corrosion, and
degradation of surfaces, and accelerates weathering of man-made and
natural materials.
---------------------------------------------------------------------------

\11\ Nitrates and sulfates in PM are derived primarily from
emissions of SO_X and NO_X.
---------------------------------------------------------------------------

Selenium
Selenium occurs naturally in soils, is associated with copper
refining, and several industrial processes, and has been used in
pesticides. It is an essential element and bioaccumulates in certain
plant species, and has been associated with toxic effects in livestock
(blind staggers syndrome). Soils containing high levels of selenium
(seleniferous soils can lead to high concentration of selenium in
certain plants, and pose a hazard to livestock and other species.
Bioaccumulation and magnification of selenium has also been observed in
aquatic organisms and has been shown to be toxic to piscivorous fish.
In humans, selenium partitions to the kidneys and liver, and is
excreted through the urine and feces. Selenium intoxication in humans
causes a syndrome known as selenosis. The condition is characterized by
chronic dermatitis, fatigue, anorexia, gastroenteritis, hepatic
degeneration, enlarged spleen and increased concentrations of Se in the
hair and nails. Clinical signs of selenosis include a characteristic
``garlic odor'' of excess selenium excretion in the breath and urine,
thickened and brittle nails, hair and nail loss, lowered hemoglobin
levels, mottled teeth, skin lesions and CNS abnormalities (peripheral
anesthesia, acroparesthesia and pain in the extremities). Aquatic birds
are extremely sensitive to selenium; toxic effects include
teratogenesis. Based on available data, both aquatic birds and aquatic
mammals are sensitive ecological receptors.

II. Summary of the Proposed Rule

A. What Source Categories Are Affected by the Proposed Rule?

1. Incinerators That Burn Hazardous Waste
A hazardous waste burning incinerator is defined under Sec.
63.1201(a) as a device that meets the definition of an incinerator in
40 CFR part 260.10 and that burns hazardous waste at any time.
Hazardous waste incinerators are currently subject to the emission
standards of part 63, subpart EEE.\12\ Hazardous waste incinerator
design types include rotary kilns, liquid injection incinerators,
fluidized bed incinerators, and fixed hearth incinerators. Most
incinerators have air pollution control equipment to capture
particulate matter (and nonvolatile metals) and scrubbing equipment for
the capture of acid gases. At least four incinerators are equipped with
activated carbon injection systems or carbon beds to control dioxin/
furan emissions (as well as other HAP emissions).
---------------------------------------------------------------------------

\12\ Incinerators that burn hazardous waste will also remain
subject to the RCRA hazardous waste incinerator emission limitations
pursuant to Sec. 264 subpart O until they demonstrate compliance
with the interim MACT standards and remove the emission limitations
from their RCRA permit. See Sec. 270.42 appendix I, section a.8 and
introductory paragraph to Sec. 270.62.
---------------------------------------------------------------------------

Incinerators can be further classified as either commercial or
onsite. Commercial incinerators accept and treat, for a tipping fee,
wastes that have been generated off-site. The purpose of commercial
incinerators is to generate profit from treating hazardous wastes. On-
site facilities treat only wastes that have been generated at the
facility to avoid the costs of off-site treatment. In 2003, there were
approximately 107 hazardous waste incinerators in operation, 15 of
which were commercial facilities, the remaining being on-site facilities.
2. Cement Kilns That Burn Hazardous Waste
A hazardous waste burning cement kiln is defined under Sec.
63.1201(a). Cement kilns that burn hazardous waste are currently
subject to the emission standards of part 63, subpart EEE.\13\ Cement
kilns are long, cylindrical, slightly inclined rotating furnaces that
are lined with refractory brick to protect the steel shell and retain
heat within the

[[Page 21208]]

kiln. Cement kilns are designed to calcine, or expel carbon dioxide by
roasting, a blend of raw materials such as limestone, shale, clay, or
sand to produce Portland cement. The raw materials enter the kiln at
the elevated end, and the combustion fuels generally are introduced
into the lower end of the kiln where the clinker product is discharged.
The materials are continuously and slowly moved to the lower end by
rotation of the kiln. As they move down the kiln, the raw materials are
changed to cementitious minerals as a result of increased temperatures
within the kiln.
---------------------------------------------------------------------------

\13\ Cement kilns that burn hazardous waste will also remain
subject to the RCRA Boilers and Industrial Furnace emission
limitations pursuant to Sec. 266 subpart H until they demonstrate
compliance with the interim MACT standards and remove the emission
limitations from their RCRA permit. See Sec. 270.42 appendix I,
section a.8 and introductory paragraph to Sec. 270.66.
---------------------------------------------------------------------------

Portland cement is a fine powder, usually gray in color, that
consists of a mixture of minerals comprising primarily calcium
silicates, aluminates, and aluminoferrites, to which small amounts of
gypsum have been added during the finish grinding operations. Portland
cement is the key ingredient in Portland cement concrete, which is used
in almost all construction applications.
Cement kilns covered by this proposal burn hazardous waste-derived
fuels to replace some or all of normal fossil fuels, typically coal.
Most kilns burn liquid waste; however, cement kilns also may burn
solids and small containers containing viscous or solid hazardous waste
fuels. The annual hazardous waste fuel replacement rate varies
considerably across sources from approximately 25 to 85 percent.
In 2003, there were 14 Portland cement plants in nine states
operating a total of 25 hazardous waste burning kilns. All cement kilns
use either bag houses or electrostatic precipitators to control
particulate matter emissions.
3. Lightweight Aggregate Kilns That Burn Hazardous Waste
A hazardous waste burning lightweight aggregate kiln is defined
under Sec. 63.1201(a). Lightweight aggregate kilns that burn hazardous
waste are currently subject to the emission standards of part 63,
subpart EEE.\14\ Raw materials such as shale, clay, and slate are
crushed and introduced at the upper end of the rotary kiln. In passing
through the kiln, the materials reach temperatures of 1,900-2,100 [deg]
F. Heat is provided by a burner at the lower end of the kiln where the
product is discharged. As the raw material is heated, it melts into a
semi-plastic state and begins to generate gases that serve as the
bloating or expanding agent. As temperatures reach their maximum, the
semi-plastic raw material becomes viscous and entraps the expanding
gases. This bloating action produces small, unconnected gas cells,
which remain in the material after it cools and solidifies. Lightweight
aggregate kilns are designed to expand the raw material by thermal
processing into a coarse aggregate used in the production of
lightweight concrete products such as concrete block, structural
concrete, and pavement.
---------------------------------------------------------------------------

\14\ Lightweight aggregate kilns that burn hazardous waste will
also remain subject to the RCRA Boilers and Industrial Furnace
emission limitations pursuant to Sec. 266 subpart H until they
demonstrate compliance with the interim MACT standards and remove
the emission limitations from their RCRA permit. See Sec. 270.42
appendix I, section a.8 and introductory paragraph to Sec. 270.66.
---------------------------------------------------------------------------

The lightweight aggregate kilns affected by this proposal burn
hazardous waste-derived fuels to replace some or all of normal fossil
fuels. Two of the facilities burn only liquid hazardous wastes, while
the third facility burns both liquid and solid wastes. The annual
hazardous waste fuel replacement rate is 100 percent.
In 2003, there were three lightweight aggregate kiln facilities in
two states operating a total of seven hazardous waste-fired kilns. All
lightweight aggregate kilns use baghouses to control particulate matter
and one facility also uses a venturi scrubber to control acid gas emissions.
4. Boilers That Burn Hazardous Waste
Boilers that burn hazardous waste are currently regulated under
RCRA at part 266, subpart H. We propose to use the RCRA definition of
boiler under 40 CFR 260.10 for purposes of today's rulemaking for
simplicity and continuity. This definition includes industrial,
commercial, and institutional boilers as well as thermal units known in
industry as process heaters. We propose to subcategorize boilers based
on the type of fuel that is burned, which would result in separate
emission standards for solid fuel-fired boilers and liquid fuel-fired
boilers. We discuss subcategorization options in more detail in Part
Two, Section II.
Boilers are typically described by either their design or type of
fuel burned. Hazardous waste burning boilers comprise two basic
different boiler designs--watertube and firetube. The choice of which
design to use depends on factors such as the desired steam quality,
thermal efficiency, size, economics, fuel type, and responsiveness.
Watertube boilers are those that flow the water through tubes running
the length of the boiler. The hot combustion gas surrounds these tubes,
causing the water inside to get hot. Most hazardous waste burning
boilers use this design. Watertube boilers can also burn a variety of
fuel types including coal, oil, gas, wood, and municipal or industrial
wastes. Firetube boilers are similar to watertube type, except the
placement of the water and combustion gas is reversed. Here the hot
combustion gas flows through the tubes, while the water surrounds the
tubes. This design does have some disadvantages, however, in that they
work well with only gas and liquid fuels.
Process heaters are similar to boilers (as conventionally defined),
except they heat a fluid other than water. This fluid is often an oil
or some other fluid with more suitable heating properties. Process
heaters are often used in circumstances where the amount of heat needed
is greater than what can be delivered by steam. For the purposes of
this rulemaking and consistent with current RCRA regulations, process
heaters would be classified as boilers.
Descriptions of liquid and solid fuel-fired boilers that burn
hazardous waste are provided below.
a. Liquid Fuel-Fired Boilers. A liquid fuel-fired boiler is a
device that meets the definition of a boiler under 40 CFR 260.10 and
that burns any combination of liquid and gas fuels, but no solids. See
proposed definition in Sec. 63.1201(a). A liquid fuel is defined as a
fuel that is pumpable (e.g., liquid wastes, sludges, or slurries). Most
liquid hazardous waste burning boilers co-fire natural gas, fuel oil,
or process gases to achieve the proper combustion temperatures and a
consistent steam supply.
There are approximately 104 liquid fuel-fired boilers that burn
hazardous waste, 85 of which have not installed back-end air pollution
control equipment. The rest of the liquid boilers use either a wet
scrubber, electrostatic precipitator, or fabric filter. These boilers
co-fire liquid hazardous waste with either natural gas or heating oil
at heat input rates of 10% to 100%.
b. Solid Fuel-Fired Boilers. A solid fuel-fired boiler is a device
that meets the definition of a boiler under 40 CFR 260.10 and that
burns solid fuels, including both pulverized and stoker coal.\15\ See
proposed definition in Sec. 63.1201(a). Boilers that co-fire solid
fuel with liquid or gaseous fuels are solid fuel-fired boilers.
---------------------------------------------------------------------------

\15\ Please note that the RCRA definition of boiler includes
devices defined under part 63 as boilers and process heaters.
---------------------------------------------------------------------------

There are 12 solid fuel-fired boilers that burn hazardous waste.
These boilers co-fire liquid hazardous waste with coal at heat input
rates of 6% to 33%. Nine of these boilers are stoker-fired, and three
burn pulverized coal. Two boilers are equipped with fabric filters to
control particulate matter and

[[Page 21209]]

metals, and 10 are equipped with electrostatic precipitators.
5. Hydrochloric Acid Production Furnaces That Process Hazardous Waste
Hydrochloric acid production furnaces that burn hazardous waste are
currently regulated under RCRA at part 266, subpart H. We propose to
use the RCRA definition of hydrochloric acid production furnace under
40 CFR 260.10 for purposes of today's rulemaking for simplicity and
continuity. See proposed definition in Sec. 63.1201(a).
Hydrochloric acid production furnaces burn chlorinated hazardous
wastes to make an aqueous hydrochloric acid for on-site use as an
ingredient in a manufacturing process. The hazardous waste feedstocks
have a chlorine content of over 20% by weight. The hydrochloric acid
produced by burning the chlorinated byproducts dissolves in the
scrubber water to produce an acid product containing hydrochloric acid
greater than 3% by weight. There are 17 hazardous waste burning
hydrochloric acid production furnaces currently in operation.
Chlorine-bearing feedstreams, wastes, and auxiliary fuels (usually
natural gas) are burned in these hydrochloric acid production furnaces
in a refractory lined chamber similar to a liquid waste incinerator
chamber. Combustion is maintained at a high temperature, with adequate
excess hydrogen to ensure the conversion of chlorine in the feedstreams
to hydrogen chloride in the combustion gases. Many furnaces also have
waste heat boilers, similar to those used by some incinerators, to
recover heat and return it to the production process. Others use a
water spray quench to cool the combustion gases.
The cooled combustion flue gas is routed to an acid recovery
system, consisting of multiple wet scrubbing absorption units. These
units are usually packed tower or film tray scrubbers which operate
with an acidic scrubbing solution. The scrubbing solution is recycled
to concentrate the acid until it reaches the desired concentration
level, at which point it is recovered for use as a valuable product. A
final polishing scrubber, operated with a caustic liquid solution, is
used to control emissions of hydrogen chloride and chlorine gas.

B. What HAP Are Emitted?

Incinerators, cement kilns, lightweight aggregate kilns, and
hydrochloric acid production furnaces that burn hazardous waste can
emit high levels of dioxin/furans depending on the design and operation
of the emission control equipment, and, for incinerators, whether a
waste heat recovery boiler is used. Our data base shows that boilers
that burn hazardous waste generally do not emit high levels of dioxin/furans.
All hazardous waste combustors can emit high levels of other
organic HAP if they are not designed, operated, and maintained to
operate under good combustion conditions.
Hazardous waste combustors can also emit high levels of metal HAP,
depending on the level of metals in the waste feed and the design and
operation of air emissions control equipment. Hydrochloric acid
production furnaces, however, generally feed and emit low levels of
metal HAP.
Hazardous waste combustors can also emit high levels of particulate
matter, except that hydrochloric acid production furnaces generally
feed wastes with low ash content and emit low levels of particulate
matter.\16\ The majority of particulate matter emissions from hazardous
waste combustors is in the form of fine particulate (i.e., 50% or more
of the particulate matter emitted is 2.5 microns in diameter or
less).\17\ Particulate emissions from incinerators and liquid fuel-
fired boilers depend on the ash content of the waste feed and the
design and operation of air emission control equipment. Particulate
emissions from cement kilns and lightweight aggregate kilns are not
significantly affected by the ash content of the hazardous waste fuel
because uncontrolled particulate emissions are attributable primarily
to raw material entrained in the combustion gas. Thus, particulate
emissions from kilns depend on operating conditions that affect
entrainment of raw material, and the design and operation of the
emission control equipment.
---------------------------------------------------------------------------

\16\ Emissions of particulate matter are of interest because
metal HAP, except notably for mercury, are in the particulate form
in stack gas. Thus, controlling particulate matter controls metal HAP.
\17\ Particulate size distributions are somewhat dependent on
the type of combustor. See USEPA ``Draft Technical Support Document
for HWC MACT Replacement Standards, Volume V: Emission Estimates and
Engineering Costs,'' March 2004, Chapter 7 for more information.
---------------------------------------------------------------------------

C. Does Today's Proposed Rule Apply to My Source?

The following sources that burn hazardous waste are considered to
be affected sources subject to today's proposed rule: Incinerators,
cement kilns, lightweight aggregate kilns, boilers, and hydrochloric
acid production furnaces. Affected sources do not include: (1) Sources
exempt from regulation under 40 CFR part 266, subpart H, because the
only hazardous waste they burn is listed under 40 CFR 266.100(c); (2)
research, development, and demonstration sources exempt under Sec.
63.1200(b); and (3) boilers exempt from regulation under 40 CFR part
266, subpart H, because they meet the definition of small quantity
burner under 40 CFR 266.108. See Sec. 63.1200(b).
Affected sources also do not include emission points that are
unrelated to the combustion of hazardous waste (e.g., cement kiln
clinker cooler stack emissions, hydrochloric acid production facility
emissions originating from product or waste storage tanks and transfer
operations, etc.). This is because subpart EEE only controls HAP
emission points that are directly related to the combustion of
hazardous waste. Under separate rulemakings, the Agency has or will
establish MACT standards, where warranted, to control HAP emissions
from non-hazardous waste related emission points.
Hazardous waste combustors are affected sources irrespective of
whether they are major sources or area sources. As discussed in Part
Two, Section I.A, we are proposing to subject area sources of boilers
and hydrochloric acid production furnaces to the major source MACT
standards for mercury, dioxin/furans, carbon monoxide/hydrocarbons, and
destruction and removal efficiency pursuant to section 112(c)(6). As
promulgated in the 1999 rule, both area source and major source
incinerators, cement kilns, and lightweight aggregate kilns will
continue to be subject to the full suite of Subpart EEE emission standards.

D. What Emissions Limitations Must I Meet?

Under today's proposal, you would have to comply with the emission
limits in Tables 1 and 2. Note that these emission limitations are
discussed in greater detail for each source category (and subcategory)
in Part Two, Section VII thru XII. Note also that we are proposing
several alternative emission standards: (1) You may elect to comply
with an alternative to the particulate matter standard for incinerators
and liquid fuel-fired boilers that would limit emissions of total metal
HAP; and (2) you may elect to comply with an alternative to the total
chlorine standard applicable to all source categories, except
hydrochloric acid production furnaces, under which you may establish
site-specific, risk-based emission limits for hydrogen chloride and
chlorine gas based on national

[[Page 21210]]

exposure standards. These alternative standards are discussed in Part
Two, Section XVIII and Section XIII, respectively.

Table 1.--Proposed Standards for Existing Sources
--------------------------------------------------------------------------------------------------------------------------------------------------------
Hydrochloric acid
Incinerators Cement kilns Lightweight Solid fuel-fired Liquid fuel-fired production
aggregate kilns boilers \1\ boilers \1\ furnaces \1\
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dioxin/Furans ( ng TEQ/dscm).... 0.28 for dry APCD 0.20 or 0.40 + 0.40.............. CO or THC standard 0.40 for dry APCD 0.40
and WHB sources; 400[deg]F at APCD as a surrogate. sources; CO or HC
\6\ 0.40 for inlet. standard as
others. surrogate for
others.
Mercury......................... 130 ug/dscm....... 64 ug/dscm \2\.... 67 ug/dscm \2\.... 10 ug/dscm........ 3.7E-6 lb/MMBtu 2, Total chlorine
5. standard as
surrogate
Particulate Matter.............. 0.015 gr/dscf \8\. 0.028 gr/dscf..... 0.025 gr/dscf..... 0.030 gr/dscf \8\. 0.032 gr/dscf \8\. Total chlorine
standard as
surrogate
Semivolatile Metals (lead + 59 ug/dscm........ 4.0E-4 lbs/MMBtu 3.1E-4 lb/MMBtu 170 ug/dscm....... 1.1E-5 lb/MMBtu 2, Total chlorine
cadmium). \5\. \5\ and 250 ug/ 5. standard as
dscm \3\. surrogate
Low Volatile Metals (arsenic + 84 ug/dscm........ 1.4E-5 lbs/MMBtu 9.5E-5 lbs/MMBtu 210 ug/dscm....... 1.1E-4 lb/MMBtu 4, Total chlorine
beryllium + chromium). \5\. \5\ and 110 ug/ 5. standard as
dscm \3\. surrogate
Total Chlorine (hydrogen 1.5 ppmv \7\...... 110 ppmv \7\...... 600 ppmv \7\...... 440 ppmv \7\...... 2.5E-2 lb/MMBtu 14 ppmv or
chloride + chlorine gas). \5, 7\. 99.9927% system
removal
efficiency
Carbon Monoxide (CO) or 100 ppmv CO or 10 See Part Two, 100 ppmv CO or 20 (2) 100 ppmv CO or 10 ppmv HWC
Hydrocarbons HWC. ppmv HWC. Section VIII. ppmv HWC.
Destruction and Removal 99.99% for each principal organic hazardous pollutant. For sources burning hazardous wastes F020, F021, F022, F023,
Efficiency (DRE). F026, or F027, however, 99.9999% for each principal organic hazardous pollutant.
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes:
\1\ Particulate matter, semivolatile metal, low volatile, and total chlorine standards apply to major sources only for solid fuel-fired boilers, liquid
fuel-fired boilers, and hydrochloric acid production furnaces.
\2\ Standard is based on normal emissions data.
\3\ Sources must comply with both the thermal emissions and emission concentration standards.
\4\ Low volatile metal standard for liquid fuel-fired boilers is for chromium only. Arsenic and beryllium are not included in the low volatile metal
total for liquid fuel-fired boilers.
\5\ Standards are expressed as mass of pollutant contributed by hazardous waste per million Btu contributed by the hazardous waste.
\6\ APCD denotes ``air pollution control device'', WHB denotes ``waste heat boiler''.
\7\ Sources may elect to comply with site-specific, risk-based emission limits for hydrogen chloride and chlorine gas based on national exposure
standards. See Part Two, Section XIII.
\8\ Sources may elect to comply with an alternative to the particulate matter standard. See Part Two, Section XVIII.

Table 2.--Proposed Standards for New Sources
--------------------------------------------------------------------------------------------------------------------------------------------------------
Hydrochloric acid
Incinerators Cement kilns Lightweight Solid fuel boilers Liquid fuel production
aggregate kilns \1\ boilers \1\ furnaces \1\
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dioxin/Furans ( ng TEQ/dscm).... 0.11 for dry APCD 0.20 or 0.40 + 0.40.............. Carbon monoxide 0.015 or 400[deg]F 0.40
or WHBs \5\; 0.2 400[deg]F at (CO) or at the inlet to
for others. inlet to hydrocarbon (HC) particulate
particulate as a surrogate. matter control
matter control device for dry
device. APCD; CO or HC
standard as
surrogate for
others.
Mercury......................... 8 ug/dscm......... 35 ug/dscm \2\.... 67 ug/dscm \2\.... 10 ug/dscm........ 3.8E-7 lb/MMBtu 2, Tcl as surrogate
4.
Particulate matter.............. 0.00070 gr/dscf 0.0058 gr/dscf.... 0.0099 gr/dscf.... 0.015 gr/dscf \7\. 0.0076 gr/dscf \7\ TCL as surrogate
\7\.
Semivolatile Metals (lead + 6.5 ug/dscm....... 6.2E-5 lb/MMBtu 2.4E-5 lb/MMBtu 170 ug/dscm....... 4.3E-6 lb/MMBtu 2, TCL as surrogate
cadmium). \4\. \4\. 4.
Low Volatile Metals (arsenic + 8.9 ug/dscm....... 1.4E-5 lb/MMBtu 3.2E-5 lb/MMBtu 190 ug/dscm....... 3.6E-5 lb/MMBtu in TCL as surrogate
beryllium + chromium). \4\. \4\. HW 3, 4.
Total Chlorine (Hydrogen 0.18 ppmv \6\..... 78 ppmv \6\....... 600 ppmv \6\...... 73 ppmv \6\....... 7.2E-4 lb/MMBtu 4, 1.2 ppmv or
chloride + chlorine gas). 6. 99.99937% SRE

[[Page 21211]]

Carbon monoxide CO or 100 ppmv (CO) or See Part Two, 100 ppmv CO or 20 100 ppmv CO or 10 ppmv HWC
Hydrocarbons (HWC). 10 ppmv HWC. Section VIII. ppmv HWC.
Destruction and Removal 99.99% for each principal organic hazardous pollutant. For sources burning hazardous wastes F020, F021, F022, F023,
Efficiency. F026, or F027, however, 99.9999% for each principal organic hazardous pollutant.
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes:
\1\ Particulate matter, semivolatile metal, low volatile metal, and total chlorine standards apply to major sources only for solid fuel-fired boilers,
liquid fuel-fired boilers, and hydrochloric acid production furnaces.
\2\ Standard is based on normal emissions data.
\3\ Low volatile metal standard for liquid fuel-fired boilers is for chromium only. Arsenic and beryllium are not included in the low volatile metal
total for liquid fuel-fired boilers.
\4\ Standards are expressed as mass of pollutant contributed by hazardous waste per million Btu contributed by the hazardous waste.
\5\ APCD denotes ``air pollution control device'', WHB denotes ``waste heat boiler''.
\6\ Sources may elect to comply with site-specific, risk-based emission limits for hydrogen chloride and chlorine gas based on national exposure
standards. See Part Two, Section XIII.
\7\ Sources may elect to comply with an alternative to the particulate matter standard. See Part Two, Section XVIII.

E. What Are the Testing and Initial Compliance Requirements?

We are proposing testing and initial compliance requirements for
solid fuel-fired boilers, liquid fuel-fired boilers and hydrochloric
acid production furnaces that are identical to those that are
applicable to incinerators, cement kilns, and lightweight aggregate
kilns already in place at Sec. Sec. 63.1206, 63.1207, and 63.1208.
Please note also that in Part Three of today's preamble we request
comment on, or propose revisions to, several testing and initial
compliance requirements. Any amendments to the testing and compliance
requirements that we promulgate as a result of those discussions would
be applicable to all hazardous waste combustors.
In addition, we are proposing to revise the existing initial
compliance requirements for incinerators, cement kilns, and lightweight
aggregate kilns. Under the proposed revision, owners and operators of
incinerators, cement kilns, and lightweight aggregate kilns would be
required to conduct the initial comprehensive performance test to
document compliance with the replacement standards proposed today
(Sec. Sec. 63.1219, 63.1220, and 63.1221) within 12 months of the
compliance date. Owners and operators of solid fuel-fired boilers,
liquid fuel-fired boilers and hydrochloric acid production furnaces
would be required to conduct an initial comprehensive performance test
within six months of the compliance date, and periodic comprehensive
performance tests every five years. The purpose of the comprehensive
performance test is to document compliance with the emission standards,
document that continuous monitoring systems meet performance
requirements, and establish limits on operating parameters that would
be monitored by continuous monitoring systems.
Owners and operators of liquid fuel-fired boilers equipped with a
dry air pollution control device and hydrochloric acid production
furnaces would be required to conduct a dioxin/furan confirmatory
performance test 2.5 years after each comprehensive performance test
(i.e., midway between comprehensive performance tests). The purpose of
the dioxin/furan confirmatory performance test is to document
compliance with the dioxin/furan standard when operating within the
range of normal operations. Owners and operators of solid fuel-fired
boilers, and liquid fuel-fired boilers that are not subject to a
numerical dioxin/furan emission standard (i.e., liquid fuel-fired
boilers other than those equipped with an electrostatic precipitator or
fabric filter), would be required to conduct a one-time dioxin/furan
test to enable the Agency to evaluate the effectiveness of the carbon
monoxide/hydrocarbon standard and destruction and removal efficiency
standard in controlling dioxin/furan emissions for those sources. The
Agency would use those emissions data when reevaluating the MACT
standards under section 112(d)(6) and when determining whether to
develop residual risk standards for these sources pursuant to CAA
section 112(f)(2).
Owners and operators of solid fuel-fired boilers, liquid fuel-fired
boilers and hydrochloric acid production furnaces would be required to
use the following stack test methods to document compliance: (1) Method
29 for mercury, semivolatile metals, and low volatile metals; and (2)
Method 26A for hydrogen chloride and chlorine gas; (3) either Method
0023A or Method 23 for dioxin/furans; and (4) either Method 5 or 5i for
particulate matter.
The following is a proposed time-line for testing and initial
compliance requirements for owners and operators of solid fuel-fired
boilers, liquid fuel-fired boilers and hydrochloric acid production
furnaces: (1) The compliance date is three years from publication of
the final rule; (2) you must place in the operating record a
Documentation of Compliance by the compliance date identifying that the
operating parameter limits you have determined using available
information will ensure compliance with the emission standards; (3) you
must commence the initial comprehensive performance test within six
months of the compliance date; (4) you must complete the initial
comprehensive performance test within 60 days of commencing the test;
and (5) you must submit a Notification of Compliance within 90 days of
completing the test documenting compliance with emission standards and
CMS requirements.

F. What Are the Continuous Compliance Requirements?

We are proposing continuous compliance requirements for solid fuel-
fired boilers, liquid fuel-fired boilers and hydrochloric acid
production furnaces that are identical to those already in place at
Sec. 63.1209 and applicable to incinerators, cement kilns, and
lightweight aggregate kilns. Please note, however, that in Part Three
of today's preamble we request comment on, or propose revisions to,
several continuous compliance requirements. Any amendments to the
continuous compliance requirements that we promulgate as a result of
those discussions would be applicable to all hazardous waste combustors.

[[Page 21212]]

Owners and operators of solid fuel-fired boilers, liquid fuel-fired
boilers and hydrochloric acid production furnaces would be required to
use carbon monoxide or hydrocarbon continuous emissions monitors (as
well as an oxygen continuous emissions monitor to correct the carbon
monoxide or hydrocarbon values to 7% oxygen) to ensure compliance with
the carbon monoxide or hydrocarbon emission limits.
Owners and operators of solid fuel-fired boilers, liquid fuel-fired
boilers and hydrochloric acid production furnaces would also be
required to establish limits on the feedrate of metals, chlorine, and
(for some source categories) ash, key combustor operating parameters,
and key operating parameters of the control device based on operations
during the comprehensive performance test. You must continuously
monitor these parameters with continuous monitoring systems. See Part
Two, Section XIV.C for a discussion of the specific parameters for
which you must establish limits.

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

We are proposing notification, recordkeeping, and reporting
requirements for solid fuel-fired boilers, liquid fuel-fired boilers
and hydrochloric acid production furnaces that are identical to those
already in place at Sec. Sec. 63.1210 and 63.1211 and applicable to
incinerators, cement kilns, and lightweight aggregate kilns. Please
note, however, that we are proposing a new requirement applicable to
all hazardous waste combustors that would require you to submit a
Notification of Intent to Comply and a Compliance Progress Report. See
Part Two, Section XVI.B.
The proposed notification, recordkeeping, and reporting
requirements are summarized in Part Two, Section XVI.

Part Two: Rationale for the Proposed Rule

I. How Did EPA Determine Which Hazardous Waste Combustion Sources Would
Be Regulated

A. How Are Area Sources Regulated?

We are proposing to subject area source boilers and hydrochloric
acid production furnaces to the major source MACT standards for
mercury, dioxin/furan, carbon monoxide/hydrocarbons, and destruction
and removal efficiency pursuant to section 112(c)(6).\18\ Both area
source and major source incinerators, cement kilns, and lightweight
aggregate kilns will continue to be subject to the full suite of
Subpart EEE emission standards.\19\
---------------------------------------------------------------------------

\18\ We are using carbon monoxide or hydrocarbons and
destruction and removal efficiency as surrogates for control of
polycyclic organic matter emissions.
\19\ In support of the 1999 Final Rule, EPA determined
incinerators, cement kilns, and lightweight aggregate kilns that are
area sources can emit HAP at levels that pose a hazard to human
health and the environment. Accordingly, EPA subjected area sources
within those source categories to the same emission standards that
apply to major sources. See 64 FR at 52837-38.
---------------------------------------------------------------------------

Section 112(c)(6) of the CAA requires EPA to list and promulgate
section 112(d)(2) or (d)(4) standards (i.e., standards reflecting MACT)
for categories and subcategories of sources emitting seven specific
pollutants. Four of those listed pollutants are emitted by boilers and
hydrochloric acid production furnaces: mercury, 2,3,7,8-
tetrachlorodibenzofuran, 2,3,7,8-tetrachlorodibenzo-p-dioxin, and
polycyclic organic matter. EPA must assure that source categories
accounting for not less than 90 percent of the aggregated emissions of
each enumerated pollutant are subject to MACT standards. Congress
singled out the pollutants in section 112(c)(6) as being of ``specific
concern'' not just because of their toxicity but because of their
propensity to cause substantial harm to human health and the
environment via indirect exposure pathways (i.e., from the air through
other media, such as water, soil, food uptake, etc.). Furthermore,
these pollutants have exhibited special potential to bioaccumulate,
causing pervasive environmental harm in biota and, ultimately, human
health risks.
We estimate that approximately 1,800 pounds of mercury are emitted
annually in aggregate from hazardous waste burning boilers in the
United States.\20\ Also, we estimate that hazardous waste burning
boilers and hydrochloric acid production furnaces emit in aggregate
approximately 1.1 and 1.6 grams TEQ per year of dioxin/furan,
respectively. The Agency has already counted on the control of these
pollutants from area sources in the industrial/commercial/institutional
boiler source category when we accounted for at least 90 percent of the
emissions of these hazardous air pollutants as being subject to
standards under section 112(c)(6). See 63 FR 17838; April 10, 1998.
Therefore, we are proposing to subject boiler and hydrochloric acid
furnace area sources to the major source MACT standards for mercury,
dioxin/furan, carbon monoxide/hydrocarbons, and destruction and removal
efficiency pursuant to section 112(c)(6).
---------------------------------------------------------------------------

\20\ See USEPA ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume V: Emission Estimates and Engineering
Costs,'' March, 2004, Chapter 3.
---------------------------------------------------------------------------

We are proposing that only major source boilers and hydrochloric
acid furnaces would be subject to the full suite of subpart EEE
emission standards we propose today. Section 112(c)(3) of the CAA
requires us to subject area sources to the full suite of standards
applicable to major sources if we find ``a threat of adverse effects to
human health or the environment'' that warrants such action. We cannot
make this finding for area source boilers and halogen acid production
furnaces.\21\ Consequently, area sources in these categories would be
subject to the MACT standards for mercury, dioxin/furan, carbon
monoxide/hydrocarbons, and destruction and removal efficiency standards
only to control the HAP listed under section 112(c)(6). RCRA standards
under Part 266, Subpart H for particulate matter, metals other than
mercury, and hydrogen chloride and chlorine gas would continue to apply
to these area sources unless an area source elects to comply with the
major source standards in lieu of the RCRA standards. See proposed
Sec. 266.100(b)(3) and the proposed revisions to Sec. Sec. 270.22 and
270.66.
---------------------------------------------------------------------------

\21\ We believe that two or fewer boilers are area sources. We
do not believe any hydrochloric acid production furnaces are area sources.
---------------------------------------------------------------------------

B. What Hazardous Waste Combustors Are Not Covered by This Proposal?

1. Small Quantity Burners
Boilers that are exempt from the RCRA hazardous waste-burning
boilers rule under 40 CFR 266.108 because they burn small quantities of
hazardous waste fuel would also be exempt from today's proposed rule.
Those boilers would be subject, however, to the MACT standards the
Agency has proposed for industrial/commercial/institutional boilers.
See 68 FR 1660, January 13, 2003.
The type and concentration of HAP emissions from boilers that co-
fire small quantities of hazardous waste fuel with other fuels under
Sec. 266.108 should be characterized more by the metals and chlorine
levels in the primary fuels and the effect of combustion conditions on
the primary fuels than by the composition and other characteristics of
the hazardous waste fuel. Under Sec. 266.108, boilers that burn small
quantities of hazardous waste fuel cannot fire hazardous waste at any
time at a rate greater than 1 percent of the

[[Page 21213]]

total fuel requirements for the boiler. In addition, a boiler with a
stack height of 20 meters or less cannot fire more than 84 gallons of
hazardous waste fuel a month, which would equate to an average firing
rate of 0.5 quarts per hour. Finally, the hazardous waste fuel must
have a heating value of 5,000 Btu/lb to ensure it is a bonafide fuel,
and cannot contain hazardous wastes that are listed because they
contain chlorinated dioxins/furans. Given these restrictions, we
believe that HAP emissions are not substantially related to the
hazardous waste fuels these boilers burn. Thus, these boilers are more
appropriately regulated under the MACT standards proposed at part 63,
subpart DDDDD, than the MACT standards proposed today for hazardous
waste combustors.
Boilers that burn small quantities of hazardous waste fuel under
Sec. 266.108 would become subject to part 63, subpart DDDDD, three
years after publication of the final rule for hazardous waste
combustors (i.e., the rules we are proposing today). Subpart DDDDD
exempts ``a boiler or process heater required to have a permit under
section 3005 of the Solid Waste Disposal Act [i.e., RCRA]
or covered by
40 CFR part 63, subpart EEE (e.g., hazardous waste combustors).'' See
40 CFR 63.7491(d). Boilers that burn small quantities of hazardous
waste fuel under Sec. 266.108 are exempt from the substantive emission
standards of part 266, subpart H, and the permit requirements of 40 CFR
part 270 (establishing RCRA permit requirements). In addition, owners
and operators of such boilers would not know whether they are covered
by part 63, subpart EEE, until we promulgate the final rule for
hazardous waste combustors. Thus, it is appropriate to require that
these boilers begin complying with subpart DDDDD three years after we
publish the final rule for hazardous waste combustors.
2. Sources Exempt From RCRA Emission Regulation Under 40 CFR Part
266.100(c)
Consistent with the Phase I Hazardous Waste Combustor MACT rule
promulgated in 1999, we would not subject boilers and hydrochloric acid
production furnaces to today's proposed requirements if the only
hazardous waste combusted is exempt from regulation pursuant to Sec.
266.100(c), including certain types of used oil, landfill gas, and
otherwise exempt or excluded waste. This is appropriate because HAP
emissions from sources that qualify for this exemption would not be
significantly impacted by the combustion of hazardous waste. Thus,
emissions from these sources would be more appropriately regulated by
other promulgated MACT standards that specifically address emissions
from these sources.
3. Research, Development, and Demonstration Sources
Consistent with the Phase I Hazardous Waste Combustor MACT rule
promulgated in 1999, we would not subject boilers and hydrochloric acid
production furnaces that are research, development, and demonstration
sources to today's proposed requirements. We explained at promulgation
of the Phase I MACT standards that the hazardous waste combustor
emission standards may not be appropriate for research, development,
and demonstration sources because of their typically intermittent
operations and small size. See 64 FR at 52839. Given that emissions
from these sources are addressed under RCRA on case-by-case basis
pursuant to Sec. 270.65, we continue to believe this is appropriate,
and we are today proposing the same exemption for boilers and
hydrochloric acid production furnaces.

C. How Would Sulfuric Acid Regeneration Facilities Be Regulated?

Sulfuric acid regeneration facilities burn spent sulfuric acid and
sulfur-bearing hazardous wastes or hazardous waste fuel to produce
sulfuric acid and are subject to 40 CFR part 266, subpart H, (i.e., the
RCRA Boiler and Industrial Furnace Rule) as a listed industrial
furnace. We are not proposing MACT standards for these sources because
EPA did not list sulfuric acid regeneration facilities as a category of
major sources of HAP emissions. See 57 FR 31576 (July 16, 1992). We
obtained emissions and other data on these sources and confirmed that
they emit very low levels of HAP.\22\ Accordingly, these combustors
will remain subject to RCRA regulations under part 266, subpart H.
---------------------------------------------------------------------------

\22\ See U.S. EPA, ``Draft Technical Support Document for HWC
MACT Replacement Standards, Volume II: HWC Emissions Data Base,''
March 2004.
---------------------------------------------------------------------------

II. What Subcategorization Considerations Did EPA Evaluate?

CAA section 112(d)(1) allows us to distinguish amongst classes,
types, and sizes of sources within a category when establishing floor
levels. Subcategorization typically reflects ``differences in
manufacturing process, emission characteristics, or technical
feasibility.'' See 67 FR 78058. A classic example, provided in the
legislative history to CAA 112(d), is of a different process leading to
different emissions and different types of control strategies--the
specific example being Soderberg and prebaked anode primary aluminum
processes. See ``A Legislative History of the Clean Air Act Amendments
of 1990,'' vol. 1 at 1138-39 (floor debates on Conference Report). If
we determine, for instance, that a given source category includes
sources that are designed differently such that the type or
concentration of HAP emissions are different we may subcategorize these
sources and issue separate standards.
We have determined that it is appropriate to subcategorize sources
that combust hazardous waste from those sources that do not. EPA
published an initial list of categories of major and area sources of
HAP selected for regulation in accordance with section 112(c) of the
Act on July 16, 1992 (57 FR 31576). Hazardous waste incineration,
Portland cement manufacturing, clay products manufacturing (including
lightweight aggregate manufacturing), industrial/commercial/
institutional boilers and process heaters, and hydrochloric acid
production are among the listed 174 categories of sources. Although
some cement kilns, lightweight aggregate kilns, boilers and process
heaters, and hydrochloric acid production furnaces burn hazardous
waste, EPA did not list hazardous waste burning sources as separate
source categories. Nonetheless, we generally believe that hazardous
waste combustion sources can emit different types or concentrations of
HAP emissions because hazardous waste combustors: (1) Have different
fuel HAP concentrations; (2) use different control techniques (e.g.,
feed control); and (3) have a different regulatory history given that
their toxic emissions were regulated pursuant to RCRA standards. As a
result, we believe it is appropriate to subcategorize each source
category listed above to define sources that burn hazardous waste as a
separate classes of combustors. We also assessed if further subdividing
each class of hazardous waste burning combustors is warranted using
both engineering judgement and statistical analysis. In our proposed
approach, we first use engineering information and principles to
identify potential subcategorization options. We then determine if
there is a statistical difference in the emission characteristics
between these options. See Part Two, Section VI.C for a discussion of
this statistical analysis. Finally, we review the results of the
statistical analysis to determine whether they are an appropriate basis for

[[Page 21214]]

subcategorization.\23\ We describe below the subcategorization options
we considered for each source category.
---------------------------------------------------------------------------

\23\ For example, although the statistical analysis may find a
significant difference in emission levels between potential
subcategories, the emission levels may be more a function of the
emission control equipment rather than a function of the design and
operation of the combustors within the subcategories. If differences
in emission levels are attributable to use of different emission
control devices, and if there is nothing inherent in the design or
operation of sources in both subcategories that would preclude
applicability of those control devices, subcategorization would not
be warranted.
---------------------------------------------------------------------------

A. What Subcategorization Options Did We Consider for Incinerators?

We considered whether to propose separate standards for three
hazardous waste incinerator subcategory options. First, we assessed
whether government-owned incinerator facilities had different emission
characteristics when compared to non-government facilities for the
mercury, semivolatile metal, low volatile metal, particulate matter,
and total chlorine floors. After evaluating the data, we determined
that emission characteristics from these two subcategories are not
statistically different, and, therefore are not proposing separate
emission standards.
Second, we assessed whether liquid injection incinerators emitted
significantly different levels of metals and particulate matter
compared to incinerators that feed solid wastes (e.g., rotary kilns,
fluid bed units, and hearth fired units). We define liquid injection
units as those incinerators that exclusively feed pumpable waste
streams and solid feed units as those that feed a combination of liquid
and solid wastes. We determined that emissions of metal HAP from these
potential subcategories are not statistically different.\24\ We,
therefore, are not proposing separate emission standards for metal HAP.
The statistical analysis for particulate matter shows that emissions
from liquid feed injection incinerators are higher than emissions from
solid feed injection units. However, we believe that separate standards
for particulate matter are not warranted because the difference in
emissions was more a factor of the types of back-end air pollution
devices used by the sources rather than incinerator design. We would
expect particulate emissions to be potentially higher for solid feed
units, not lower, because solid feed units have higher ash feedrates
and air pollution control device inlet particulate matter loadings.
Therefore, we must conclude that the difference is the product of less
effective back-end air pollution control.
---------------------------------------------------------------------------

\24\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume III: Selection of MACT Standards'',
March 2004, Chapter 4.
---------------------------------------------------------------------------

Third, we assessed whether incinerators equipped with dry air
pollution control devices and/or waste heat boilers have different
dioxin/furan emission characteristics when compared to other sources,
i.e., sources with either wet air pollution control or no air pollution
control devices. Our statistical analysis determined that dioxin/furan
emissions from sources equipped with waste heat boilers and/or dry air
pollution control devices are higher.\25\ We believe use of wet air
pollution control systems (and use of no air pollution control system)
can result in different dioxin/furan emission characteristics because
they have different post-combustion particle residence times and
temperature profiles, which can affect dioxin/furan surface catalyzed
formation reaction rates. As a result, we believe that it is
appropriate to subcategorize these different types of combustors.
---------------------------------------------------------------------------

\25\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume III: Selection of MACT Standards'',
March 2004, Chapter 4.
---------------------------------------------------------------------------

Note that we do not subcategorize based on the type of air
pollution control device used. See 69 FR 394 (January 5, 2004). Dioxin/
furan emission characteristics are unique in that they are not
typically fed into the combustion device, but rather are formed in the
combustor or post combustion within ductwork, a heat recovery boiler,
or the air pollution control system. Wet and dry air pollution control
systems are generally not considered to be dioxin/furan control systems
because their primary function is to remove metals and/or total
chlorine from the combustion gas. They generally do not remove dioxin/
furans from the incinerator flue gas unless they are used in tandem
with carbon injection systems or carbon beds. (In contrast, carbon
injection systems and carbon beds are considered to be dioxin/furan air
pollution control systems). Thus, the differences in dioxin formation
here reflect something more akin to a process difference resulting in
different emission characteristics, rather than a difference in
pollution-capture efficiencies among pollution control devices. We thus
are not proposing to subcategorize based on whether a source is
equipped with a dioxin/furan control system.
We also considered whether to further subcategorize based on the
presence of a waste heat boiler or dry air pollution control device.
Our analysis determined that dioxin/furan emissions from incinerators
with waste heat boilers are not statistically different from those
equipped with dry air pollution control devices.\26\ We conclude that
further subcategorization is not necessary. See Part Two, Section VII.A
for more discussion on the proposed dioxin/furan standards for incinerators.
---------------------------------------------------------------------------

\26\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume III: Selection of MACT Standards'',
March 2004, Chapter 4.
---------------------------------------------------------------------------

B. What Subcategorization Options Did We Consider for Cement Kilns?

We considered subdividing hazardous waste burning cement kilns by
the clinker manufacturing process: wet process kilns without in-line
raw mills versus preheater/precalciner kilns with in-line raw mills.
All cement kilns that burn hazardous waste use one of these clinker
manufacturing processes. Based on available emissions data, we
evaluated design and operating features of each process to determine if
the features could have a significant impact on emissions. For the
reasons discussed below, we believe that subcategorization is not warranted.
In the wet process, raw materials are ground, wetted, and fed into
the kiln as a slurry. Twenty-two of the 25 cement kilns that burn
hazardous waste use the wet process to manufacture clinker. In the
preheater/precalciner kilns, raw materials are ground dry in a raw mill
and fed into the kiln dry. The remaining three of the 25 cement kilns
burning hazardous waste use preheater/precalciner kilns with in-line
raw mills.
Combustion gases and raw materials move in a counterflow direction
inside a cement kiln for both processes. The kiln is inclined, and raw
materials are fed into the upper end while fuels are typically fired
into the lower end. Combustion gases move up the kiln counter to the
flow of raw materials. The raw materials get progressively hotter as
they travel down the length of the kiln. The raw materials begin to
soften and fuse at temperatures between 2,250 and 2,700 [deg]F to form
the clinker product.
Wet process kilns are longer than the preheater/precalciner kilns
in order to facilitate evaporation of the water from the slurried raw
material. The preheater/precalciner kilns begin the calcining process--
heating of the limestone to drive off carbon dioxide to obtain lime
(calcium oxide)--before the raw materials are fed into the kiln. This
is accomplished by routing the flue gases from the kiln up through the
preheater tower while the raw materials are passing down the preheater
tower.

[[Page 21215]]

The heat of the flue gas is transferred to the raw material as they
interact in the preheater tower. The precalciner is a secondary firing
system--typically fired with coal--located at the base of the preheater
tower.
Though not necessary in a wet process kiln, a preheater/precalciner
kiln uses an alkali bypass designed to divert a portion of the flue gas
to remove problematic volatile constituents such as alkalies (potassium
and sodium oxides), chlorides, and sulfur that, if not removed, can
lead to operating problems. In addition, removal of the alkalies is
necessary so that their concentrations are below maximum acceptable
levels in the clinker. An alkali bypass diverts between 10-30% of the
kiln off-gas before it reaches the lower cyclone stages of the
preheater tower. Without use of a bypass, the high concentration of
volatile constituents at the lower cyclone stage of the preheater tower
would create operational problems. Bypass gases are quenched and sent
to a dedicated particulate matter control device to capture and remove
the volatile constituents.
All preheater/precalciner kilns that burn hazardous waste use the
hot flue gases to dry the raw materials as they are being ground in the
in-line raw mill. Typically, the raw mill is operating or ``on''
approximately 85% of the time. The kilns with in-line raw mills must
operate both in the ``on'' mode--gases are routed through the raw mill
supporting raw material drying and preparation--and in the ``off''
mode--necessary down time for raw mill maintenance. Given that there
are few preheater/precalciner cement kilns that burn hazardous waste,
we had limited emissions data to evaluate to see if there was a
significant difference in emissions. Moreover, we do not have any data
from a preheater/precalciner kiln operating under similar operating
conditions (e.g., metals and chlorine feed concentrations) both for the
``on'' mode and ``off'' mode.
We evaluated whether there was a significant difference in HAP
emissions between wet process kilns without in-line raw mills versus
preheater/precalciner kilns with in-line raw mills. We found a
statistically significant difference in mercury emissions between wet
process kilns and preheater/precalciner kilns in the ``off'' mode.\27\
But, we conclude that there is no significant difference in emissions
of dioxin/furans, particulate matter, semivolatile metals, low volatile
metals, and total chlorine between these types of kiln systems.\28\
---------------------------------------------------------------------------

\27\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume III: Selection of MACT Standards'',
March 2004, Chapter 4.
\28\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume V: Emission Estimates and Engineering
Costs'', March 2004, Chapter 4.
---------------------------------------------------------------------------

For wet process cement kilns without in-line raw mills, mercury
remains in the vapor phase at the typical operating temperatures in the
kiln and particulate matter control equipment, and exits the kiln as
volatile stack emissions with only a small fraction partitioning to the
clinker or cement kiln dust. In the preheater/precalciner kilns with
in-line raw mill, we believe that a significant portion of the
volatilized mercury condenses on to the surfaces of the cooler raw
material in the operating raw mill. The raw material with adsorbed
mercury ends up in the raw material storage bin which will eventually
be fed to the kiln and re-volatilized. During the periods that the in-
line raw mill is ``on'', mercury is effectively captured in the raw
mill essentially establishing an internal recycle loop of mercury that
builds-up within the system. Eventually, when the in-line raw mill
switches to the ``off'' mode, the re-volatilized mercury exits the kiln
as volatile stack emissions. Notwithstanding the apparent removal of
mercury during periods that the in-line raw mill is ``on'' in a
preheater/precalciner kiln, over time the mercury is emitted eventually
as volatile stack emissions because system removal efficiencies for
mercury are essentially zero. Thus, over a longer period of time (e.g.,
one month), the mass of mercury emitted by a wet process kiln without
an in-line raw mill and a preheater/precalciner kiln with an in-line
raw mill (assuming identical mercury-containing feedstreams) would be
the same. However, at any given point in time, the stack gas
concentration of mercury of the two types of kilns could be
significantly different.
As noted above, our data base shows a significant difference in
mercury emissions between preheater/precalciner kilns when operating in
the ``off'' mode and emissions both from wet process kilns and
preheater/precalciner kilns in the ``on'' mode. In spite of this
difference, we don't believe it is technically justified to
subcategorize cement kilns for mercury.\29\
---------------------------------------------------------------------------

\29\ We note that in the September 1999 final rule we
established a provision that allows cement kilns operating in-line
raw mills to average their emissions based on a time-weighted
average concentration that considers the length of time the in-line
raw mill is on-line and off-line. See Sec. 63.1204(d).
---------------------------------------------------------------------------

In conclusion, we propose not to subcategorize the hazardous waste
burning class of cement kilns by wet process kilns and preheater/
precalciner kilns with in-line raw mills.

C. What Subcategorization Options Did We Consider for Lightweight
Aggregate Kilns?

Following promulgation of the September 1999 Final Rule, Solite
Corporation filed a Petition for Review challenging the total chlorine
standard for new kilns. For new sources, the Clean Air Act states that
the MACT floor cannot be ``less stringent than the emission control
that is achieved by the best controlled similar source.'' Solite
Corporation challenged the standard on the ground that Norlite
Corporation, another hazardous waste-burning lightweight aggregate kiln
source, should not be the best controlled similar source because they
are designed to burn for purposes of treatment hazardous wastes
containing high levels of chlorine and high mercury. Solite states that
Norlite's superior emission control equipment is designed to control
the chlorine and mercury in these wastes that are burned for treatment,
rather than primarily as fuel for lightweight aggregate production.
Thus, Solite states that Norlite's sources should be considered a
separate class of lightweight aggregate kilns.
Though we believe that subcategorizing by the concentrations of HAP
in the hazardous waste is not appropriate, we considered subdividing
hazardous waste burning lightweight aggregate kilns by the types of
hazardous waste they combust: low Btu wastes with higher concentrations
of chlorine and mercury and high Btu wastes with lower concentrations
of chlorine and mercury. We believe, however, that separate emission
standards for lightweight aggregate kilns based on the types of
hazardous waste they burn are unnecessary because the floor levels
would not differ significantly under either approach.
Analysis of available total chlorine emissions from compliance
testing indicates that the emissions are significantly different for
sources burning hazardous waste with high levels of chlorine compared
to sources burning wastes with much lower levels of chlorine. Total
chorine emissions range from 14 to 116 ppmv for sources feeding higher
concentrations of chlorine but using a venturi scrubber to control
emissions and range from 500 to 2,400 ppmv for sources feeding waste
with lower levels of chlorine and not using a wet scrubber. However,
when we identify floor levels for these potential subcategories (both
for existing and new sources), the calculated floor

[[Page 21216]]

level would be less stringent than the interim emission standard
sources are currently achieving. Because all sources are achieving the
more stringent interim standard, the interim standard becomes the
default floor level. Therefore, subdividing would not affect the
proposed floor level.
We have compliance test mercury emissions data representing maximum
emissions for only one source, and we have snap-shot mercury emissions
data within the range of normal emissions for all sources. Snap-shot
mercury emissions range from: (1) 11 to 20 ug/dscm for sources with the
potential to feed higher concentrations of mercury because they use a
venturi scrubber to control emissions; and (2) 1 to 47 ug/dscm for
sources that typically feed lower mercury containing wastes and do not
use a wet scrubber to control mercury. We performed a statistical test
and confirmed that there is no statistically significant difference in
the snap-shot mercury emissions between sources that have the potential
to feed higher levels of mercury because they are equipped with a wet
scrubber and with other sources. Therefore, it appears that
subcategorization for mercury is not warranted.\30\
---------------------------------------------------------------------------

\30\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standard, Volume III: Selection of MACT Standards'',
March 2004, Chapter 4.
---------------------------------------------------------------------------

D. What Subcategorization Options Did We Consider for Boilers?

We discuss below the rationale for proposing to subcategorize
boilers by the physical form of the fuels they burn--solid fuel-fired
boilers and liquid fuel-fired boilers. We also discuss further
subcategorization options we considered for each of those subcategories
and explain why we believe that further subcategorization is not warranted.
1. Subcategorization by Physical Form of Fuels Burned
There are substantial design differences and emission
characteristics among boilers that cofire hazardous waste primarily
with coal versus oil or gas. Because of these differences, it is
appropriate to subcategorize boilers by the physical form of the fuel
burned. We note that the Agency has already proposed that industrial/
commercial/institutional boilers and process heaters that do not burn
hazardous waste should be subcategorized by the physical form of fuels
fired.\31\
---------------------------------------------------------------------------

\31\ See 68 FR at 1670 (January 13, 2003).
---------------------------------------------------------------------------

Twelve boilers cofire hazardous waste with coal. These boilers are
designed to handle high ash content solid fuels, including the
relatively large quantities of boiler bottom ash and particulate matter
that are entrained in the combustion gas. The coal also contributes to
emissions of metal HAP. Approximately 104 boilers co-fire hazardous
waste with natural gas or fuel oil. These units are not designed to
handle the high ash loadings that are associated with coal-fired units,
and the primary fuels for these boilers contribute little to HAP
emissions. See ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume I: Description of Source Categories''
(Chapter 2.4) and ``Volume III: Selection of MACT Standards'' (Chapter
4) for a discussion of the design differences between liquid and coal
fuel-fired boilers.
Because the type of primary fuel burned dictates the design of the
boiler and emissions control systems, and can affect the concentration
of HAP, it is appropriate to subcategorize boilers by the physical form
of the fuel.
2. Subcategorization Considerations Among Solid Fuel Boilers
We considered whether to subcategorize solid fuel-fired boilers to
establish separate particulate matter standards. All 12 of the solid
fuel-fired boilers co-fire hazardous waste with coal. Three of the 12
boilers burn pulverized coal while the remaining nine are stoker-fired
boilers. Pulverized coal-fired boilers have higher uncontrolled
emissions than stoker-fired boilers because the coal is pulverized to a
talcum powder consistency and burned in suspension. Stoker-fired
boilers burn lump coal partially or totally on a grate. Thus, much more
of the coal ash is entrained in the combustion gas for pulverized coal-
fired boilers than for stoker-fired boilers.
Although the pulverized coal-fired boilers have higher uncontrolled
particulate matter emissions (i.e., at the inlet to the emission
control device), controlled emissions from the pulverized coal-fired
boilers are not statistically different than emissions from the stoker-
fired boilers, primarily because all solid fuel-fired boilers are
equipped with either a baghouse or electrostatic precipitator.\32\
Accordingly, we conclude that it is not appropriate to establish
separate particulate matter standards for pulverized coal-fired boilers
versus stoker-fired boilers. This is consistent with the proposal for
industrial/institutional/commercial boilers and process heaters that do
not burn hazardous waste.
---------------------------------------------------------------------------

\32\ See USEPA ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume III: Selection of MACT Standards,''
March 2004, Chapter 4.
---------------------------------------------------------------------------

3. Subcategorization Considerations for Liquid Fuel Boilers
We believe it is appropriate to combine liquid and gas fuel boilers
into one subcategory because emissions from gas fuel boilers are within
the range of emissions one finds from liquid fuel boilers. Also, most
of the hazardous waste burning liquid fuel boilers, in fact, burn gas
fossil fuels to supplement the liquid hazardous waste fuel. Even though
there are no hazardous waste gas burning boilers currently in
operation, today we propose to subject hazardous waste gas burning
boilers that may begin operating in the future to the standards for
liquid fuel-fired boilers. See proposed definition of liquid boiler in
Sec. 63.2101(a).
We also assessed whether liquid fuel-fired boilers equipped with
dry air pollution control devices had different dioxin/furan emission
characteristics when compared to other sources, i.e., sources with
either wet air pollution control devices or no air pollution control
device. Our statistical analysis indicated that dioxin/furan emissions
from sources equipped with dry air pollution control devices are
higher.\33\ We believe use of wet air pollution control systems (and
use of no air pollution control system) can result in different dioxin/
furan emission characteristics because they have different post-
combustion particle residence times and temperature profiles, which can
affect dioxin/furan surface catalyzed formation reaction rates. As a
result, we believe that it is appropriate to have different
subcategories for these different types of combustors. As discussed
previously for incinerators in Part Two, Section II.A, the differences
in dioxin formation here reflect something more akin to a process
difference resulting in different emission characteristics, rather than
a difference in pollution-capture efficiencies among pollution control
devices. We thus are not subcategorizing based on whether a source is
equipped with a dioxin/furan control system.
---------------------------------------------------------------------------

\33\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume III: Selection of MACT Standards'',
March 2004, Chapter 4.
---------------------------------------------------------------------------

E. What Subcategorization Options Did We Consider for Hydrochloric Acid
Production Furnaces?

Consistent with our incinerator subcategorization analysis (see
Section A of this Part), we also considered whether to establish
separate floor emission standards for dioxin/furans for

[[Page 21217]]

hydrochloric acid production furnaces equipped with waste heat recovery
boilers versus those without boilers. As discussed below, we conclude
that there is no significant statistical difference in dioxin/furan
emissions between furnaces equipped with boilers and those without
them. As a result we do not propose to have different subcategories for
these sources.
Ten of the 16 hydrochloric acid production furnaces are equipped
with waste heat recovery boilers, and all hydrochloric acid production
furnaces are equipped with wet scrubbers that quench the combustion gas
immediately after it exits the furnace or boiler. We have dioxin/furan
emissions data for eight of the ten furnaces with boilers. Two furnaces
have low dioxin/furan emissions--approximately 0.1 ng TEQ/dscm, while
the other six furnaces have emissions ranging from 0.5 to 6.8 ng TEQ/
dscm. We have dioxin/furan emissions data for five of the six furnaces
without boilers. Dioxin/furan emissions for four furnaces are below
0.15 ng TEQ/dscm. But, one furnace has dioxin/furan emissions of 1.7 ng
TEQ/dscm.
It appears that dioxin/furan emissions from hydrochloric acid
production furnaces may not be governed by whether the furnace is
equipped with a waste heat recovery boiler. We performed a statistical
test and confirmed that there is no statistically significant
difference in dioxin/furan emissions between furnaces equipped with
boilers and those without boilers.\34\ Thus, we conclude that it is not
appropriate to establish separate dioxin/furan emission standards for
furnaces with boilers and those without boilers.
---------------------------------------------------------------------------

\34\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume III: Selection of MACT Standards'',
March 2004, Chapter 4.
---------------------------------------------------------------------------

III. What Data and Information Did EPA Consider To Establish the
Proposed Standards?

The proposed standards are based on our hazardous waste combustor
data base. The data base contains general facility information, stack
gas emissions data, combustor design information, composition and feed
concentration data for the hazardous waste, fossil fuel, and raw
materials, combustion unit operating conditions, and air pollution
control device operating information. We gathered the emissions data
and information from test reports submitted by hazardous waste
combustor facilities to EPA Regional Offices or State agencies. Many of
the test reports were prepared as part of the compliance demonstration
process for the current RCRA standards, and may include results from
trial burns, certification of compliance demonstrations, annual
performance tests, mini-burns, and risk burns.

A. Data Base for Phase I Sources

The current data base for Phase I sources contain test results for
over 100 incinerators, 26 cement kilns, and 9 lightweight aggregate
kilns. In many cases, especially for cement and lightweight aggregate
kilns, the data base contain test reports from multiple testing
campaigns. For example, our data base includes results for a cement
kiln that conducted emissions testing for the years 1992, 1995, and 2000.
We first compiled a data base for hazardous waste burning
incinerators, cement kilns, and lightweight aggregate kilns to support
the proposed MACT standards in 1996 (61 FR 17358, April 19, 1996).
Based on public comments, a revised Phase I data base was published for
public comment (62 FR 960, January 7, 1997). The data base was again
revised based on public comments, and we used this data base to develop
the Phase I MACT standards promulgated in 1999 (64 FR 52828, September
30, 1999).
Following promulgation of the interim standards, we initiated a
data collection effort in early 2002 to obtain additional test reports.
The effort focused on obtaining test reports from sources for which we
had no information, obtaining data from more recent testing, and
updating the list of operating Phase I sources. Sources once identified
as hazardous waste combustors, but that have since ceased operations as
a hazardous waste combustor, were removed from the data base. This
revised data base was noticed for public comment in July 2002 (67 FR
44452, July 2, 2002) and updated based on public comments. See USEPA
``Draft Technical Support Document for HWC MACT Replacement Standards,
Volume II: HWC Emissions Data Base,'' March 2004, Appendix A for
comments and responses.
In comments on the data base notice, industry stakeholders question
whether emissions data obtained for some sources are appropriate to use
to identify MACT floor for today's proposed replacement standards.
Stakeholders suggest that it is inappropriate to use emissions data
from sources that tested after retrofitting their emission control
systems to meet the emission standards promulgated in September 1999
(and since vacated and replaced by the February 2002 Interim
Standards). Stakeholders refer to this as MACT-on-MACT: establishing
MACT floor based on sources that already upgraded to meet the 1999
standards. Stakeholders identified emissions data from only
approximately three of the Phase I sources (all incinerators) as being
obtained after the source upgraded to meet the 1999 standards. None of
these incinerator sources are consistently identified as a best
performer when establishing the proposed MACT standards.
Notwithstanding stakeholder concerns, we believe it is appropriate
to consider all of the data collected in the 2002 effort.\35\ First,
section 112(d)(3) states that floor standards for existing sources are
to reflect the average emission achieved by the designated per cent of
best performing sources ``for which the Administrator has emissions
information'' (emphasis added). Second, the motivation for a source's
performance is legally irrelevant in developing MACT floor levels.
National Lime Ass'n v. EPA, 233 F. 3d at 640. In any case, it would be
problematic to identify sources that upgraded their facilities (and
reduced their emissions) for purposes of complying with the 1999
standards versus for other purposes (e.g., normal replacement
schedule). Moreover, the MACT-on-MACT formulation is not correct.
Although the Interim Standards did result in reduction of emissions
from many sources, those standards are not MACT standards, and do not
purport to be. See February 13, 2002, Interim Standards Rulemaking, 67
FR at 7693. Finally, we note that, although we were prepared to use the
same data base for today's proposed rules as we used for the September
1999 rule to save the time and resources required to collect new data,
industry stakeholders wanted to submit new emissions data for us to
consider in developing the replacement standards. Rather than allowing
industry stakeholders to submit potentially selected emissions data,
however, we agreed to undertake a substantial data collection effort in
2002. It is unfortunate that industry stakeholders now suggest that
some portion of the new data is not appropriate for establishing MACT.
---------------------------------------------------------------------------

\35\ However, we did not consider emissions data from Ash Grove
Cement Company (Chanute, Kansas), an owner and operator of a new
preheater/precalciner kiln, because the test report is a MACT
comprehensive performance test demonstrating compliance with the new
source standards of the September 1999 final rule. We judged these
data are inappropriate for consideration for the floor analyses for
existing sources.
---------------------------------------------------------------------------

Notwithstanding our view that all of the 2002 data base should be
considered in establishing MACT standards, we

[[Page 21218]]

specifically request comment on: (1) Whether emissions data should be
deleted from the data base that were obtained from sources that owners
and operators assert were upgraded to meet the 1999 rule; and (2)
whether, because it may be problematic to identify such data, we should
identify MACT using the original 1999 data base.
Stakeholders have also raised concerns that the Agency may be
considering inappropriately emissions data in its MACT analyses based
on the language of section 112(d)(3)(A) of the Clean Air Act. Section
112(d)(3)(A) says emissions standards for existing sources shall not be
less stringent, and may be more stringent than--

the average emission limitation achieved by the best performing 12
percent of the existing sources (for which the Administrator has
emissions information), excluding those sources that have, within 18
months before the emission standard is proposed or within 30 months
before such standard is promulgated, whichever is later, first
achieved a level of emission rate or emission reduction which
complies, or would comply if the source is not subject to such
standard, with the lowest achievable emission rate (as defined by
section 171) applicable to the source category and prevailing at the
time, in the category or subcategory for categories and
subcategories with 30 or more sources,

Section 171 pertains to nonattainment areas for a particular
pollutant. The lowest achievable emission rate (LAER) for a pollutant
in a nonattainment area is the most stringent emission limitation which
is contained in the implementation plan of any State, or the most
stringent emission limitation which is achieved in practice. Given that
stakeholders neither identified any lowest achievable emission rates
for any pollutants applicable to nonattainment areas nor identified any
sources that are subject to such lowest achievable emission rates, we
conclude that there are no sources to exclude.

B. Data Base for Phase II Sources

Phase II sources are comprised of boilers and hydrochloric acid
production furnaces that burn hazardous waste. The data base for Phase
II sources was initially compiled by EPA in 1999. In developing this
data base, we collected the most recent test report available for each
source that included test results under compliance test operating
conditions. The most recent test report, however, may have also
included data used for other purposes (e.g., risk burn to obtain data
for a site-specific risk assessment), which are also included in the
data base. In nearly all instances, the dates of the test reports
collected were either 1998 or 1999.
After the initial compilation, we published the Phase II data base
for public comment in June 2000 (65 FR 39581, June 27, 2000). Since the
June 2000 notice, we have not collected additional emissions data for
Phase II sources; however, we revised the data base to address public
comments received in response to the June 2000 notice. We noticed the
Phase II data base (together with the one for Phase I sources) for
public comment in July 2002 (67 FR 44452, July 2, 2003) and revised the
data base based on comments received. The current data base for Phase
II sources contains test reports for over 115 boilers and 17
hydrochloric acid production furnaces. See USEPA ``Draft Technical
Support Document for HWC MACT Replacement Standards, Volume II: HWC
Emissions Data Base,'' March 2004.

C. Classification of the Emission Data

The hazardous waste combustor data base \36\ comprises emissions
data from tests conducted for various purposes, including compliance
testing, risk burns, annual performance testing, and research testing.
Therefore, some emissions data represent the highest emissions the
source has emitted in each of its compliance demonstrations, some data
represent normal or typical operating conditions and emissions, and
some data represent operating conditions and emissions during
compliance testing in a test campaign where there are other compliance
tests with higher emissions.
---------------------------------------------------------------------------

\36\ Though the Phase I and II data bases were developed and
titled separately, for purposes of today's proposal we are combining
both into one data base termed the ``hazardous waste combustor data base.''
---------------------------------------------------------------------------

Hazardous waste combustors generally emit their highest emissions
during RCRA compliance testing while demonstrating compliance with
emission standards. For real-time compliance assurance, sources are
required to establish limits on particular operating parameters that
are representative of operating levels achieved during compliance
testing. Thus, the emission levels achieved during these compliance
tests are typically the highest emission levels a source emits under
reasonably anticipable circumstances. To ensure that these operating
limits do not impede normal day-to-day operations, sources generally
take measures to operate during compliance testing under conditions
that are at the extreme high end of the range of normal operations. For
example, sources often feed ash, metals, and chlorine during compliance
testing at substantially higher than normal levels (e.g., by spiking
the waste feed) to maximize the feed concentration, and they often
detune the air pollution control equipment to establish operating
limits on the control equipment that provide operating flexibility. By
designing the compliance test to generate emissions at the extreme high
end of the normal range of emissions, sources can establish operating
limits that account for variability in operations (e.g., composition
and feedrate of feedstreams, as well as variability of pollution
control equipment efficiency) and that do not impede normal operations.
The data base also includes normal emissions data that are within
the range of typical operations. Sources will sometimes measure
emissions of a pollutant during a compliance test even though the test
is not designed to establish operating limits for that pollutant (i.e.,
it is not a compliance test for the pollutant). An example is a trial
burn where a lightweight aggregate kiln measures emissions of all RCRA
metals, but uses the Tier I metals feedrate limit to comply with the
mercury emission standard.\37\ Other examples of emissions data that
are within the range of normal emissions are annual performance tests
that some sources are required to conduct under State regulations, or
RCRA risk burns. Both of these types of tests are generally performed
under normal operating conditions, and would not necessarily reflect
day-to-day emission variability. However, such data may be appropriate
to use to evaluate long-term average performance.
---------------------------------------------------------------------------

\37\ A Tier 1 feedrate limit is a conservative compliance option
offered pursuant to RCRA requirements which assumes all of the
metal/chlorine that is fed to the combustion unit is emitted
(uncontrolled). Sources electing to comply with Tier 1 limits are
not required to conduct emissions testing and are not required to
establish operating parameter limits based on a compliance test. See
Sec. 266.106.
---------------------------------------------------------------------------

Other emissions tests may generate emissions in-between normal and
the highest compliance test emissions. An example is a compliance test
designed to demonstrate compliance with the particulate matter standard
where: (1) The air pollution control equipment is detuned; and (2) the
source measured lead and cadmium emissions even though it elected to
comply with RCRA Tier 1 feedrate limits for those metals and, thus,
does not spike those metals. We would conclude that lead and cadmium
emissions--together they comprise the semivolatile metals--are between
normal and the highest compliance test emissions. Emissions are not
likely to be as high as

[[Page 21219]]

compliance test emissions because the source did not use the test to
demonstrate compliance with emission standards for the metals (and so
did not spike the metals). However, emissions of the metals are likely
to be higher than normal because the air pollution control equipment
was detuned.
To distinguish between normal and compliance test data, we
classified emissions data for each pollutant for each test condition as
compliance test (CT); normal (N); in between (IB); or not applicable
(NA).\38\ These classifications apply on a HAP-by-HAP basis. For
example, some HAP measured during a test condition may be classified as
representing compliance test emissions for those HAP, while other HAP
measured during the test condition may be classified as representing
normal emissions. See USEPA ``Draft Technical Support Document for HWC
MACT Replacement Standards, Volume II: HWC Emissions Data Base,'' March
2004, Chapter 2, for additional details.
---------------------------------------------------------------------------

\38\ NA means the normal versus compliance test classification
is not applicable. Research testing data is an example of the type
of data that would get a NA rating.
---------------------------------------------------------------------------

D. Invitation To Comment on Data Base

As previously discussed, we updated the data base based on comments
received since it was last made publicly available. We believe the data
base used to determine today's proposed standards is complete and
accurate. However, given the complexity of the data base, we believe it
is appropriate to once again solicit comments on the accuracy of the
data. If you find errors, please submit the pages from the test report
that document the missing or incorrect entries and the cover page of
the test report as a reference. In addition, we identified several
sources that are no longer burning hazardous waste and removed their
emissions data and related information from the data base. We encourage
owners and operators of hazardous waste combustors to review our list
of operating combustors to ensure its accuracy. See USEPA ``Draft
Technical Support Document for HWC MACT Replacement Standards, Volume
III: Selection of MACT Standards and Technologies,'' March 2004.

IV. How Did EPA Select the Format for the Proposed Rule?

The proposed rule includes emission limits for dioxin/furans,
mercury, particulate matter, semivolatile metals, low volatile metals,
hydrogen chloride/chlorine gas, and carbon monoxide or hydrocarbons. We
also propose percent reduction standards for: (1) Destruction and
removal efficiency \39\ for organic HAP; and (2) total chlorine control
for hydrochloric acid production furnaces. Finally, sources would be
required to establish operating parameter limits under prescribed
procedures to ensure continuous compliance with the emission standards.
---------------------------------------------------------------------------

\39\ Please note that we propose today a destruction and removal
efficiency standard only for boilers and process heaters and
hydrochloric acid production furnaces. We are not reproposing the
destruction and removal efficiency standard in subpart EEE currently
in effect for incinerators, cement kilns, and lightweight aggregate kilns.
---------------------------------------------------------------------------

We discuss below the rationale for: (1) Selecting an emission limit
format rather than a percent reduction format in most cases; (2)
selecting a hazardous waste thermal emissions format for the emission
limit in some cases, and an emissions concentration format in others;
(3) selecting surrogates to control multiple HAP; and (4) using
operating parameter limits to ensure compliance with emission standards.

A. What Is the Rationale for Generally Selecting an Emission Limit
Format Rather Than a Percent Reduction Format?

Using emission limits as the format for most of the proposed
standards provides flexibility for the regulated community by allowing
a regulated source to choose any control technology or technique to
meet the emission limits, rather than requiring each unit to use a
prescribed method that may not be appropriate in each case. (See CAA
section 112(h), relating to authority to adopt work place standards).
Although a percent reduction format would allow flexibility in choosing
the control technology to achieve the reduction, a percent reduction
technology does not allow the option of achieving the standard by feed
control--minimizing the feed of metals or chlorine. Consequently, we
propose percent reduction standards only in special circumstances.
We are proposing a percent reduction standard for boilers and
hydrochloric acid production furnaces, i.e., a destruction and removal
efficiency standard for organic HAP, because all sources currently
comply with such a standard under RCRA and RCRA implementing rules.
Further, we do not have emissions data on trace levels of organic HAP
that would be needed to establish emission limits for particular compounds.
We also propose a total chlorine percent reduction standard as a
compliance option for hydrochloric acid production furnaces in lieu of
the proposed stack gas concentration limit because a stack gas
concentration limit may ultimately result in limiting the feed of
chlorine to furnaces with MACT emission control equipment. Given that
these furnaces produce hydrochloric acid from chlorinated feedstocks,
limiting the feed of chlorine is inappropriate. See Part Two, Section
VI.A and XII for more discussion on the total chlorine standard for
hydrochloric acid production furnaces.

B. What Is the Rationale for Selecting a Hazardous Waste Thermal
Emissions Format for Some Standards, and an Emissions Concentration
Format for Others?

We are proposing numerical emission limits in two formats:
hazardous waste thermal emissions, and stack gas emissions
concentrations. Hazardous waste thermal emissions are expressed as mass
of pollutant contributed by hazardous waste per million Btu of heat
contributed by hazardous waste. Emission concentration based standards
are expressed as mass of pollutant (from all feedstocks) per unit of
stack gas (e.g., [mu]g/dscm).
1. What Is the Rationale for the Hazardous Waste Thermal Emissions Format?
In the 1999 rule, we assessed hazardous waste feed control levels
for metals and chlorine by evaluating each source's maximum theoretical
emission concentration (MTEC) using the ``aggregate MTEC'' approach.
See 64 FR at 52854. MTEC is defined as the metals or chlorine feedrate
divided by the gas flow rate, and is expressed in [mu]g/dscm. We used
MTECs to assess feed control levels because it normalizes metal and
chlorine feedrates across sources of different sizes. Industry
stakeholders have claimed that use of MTECs to assess feed control
levels for energy recovery units (e.g., cement kilns) when establishing
floor standards inappropriately penalizes sources that burn hazardous
waste fuels at high firing rates (i.e., percent of heat input from
hazardous waste). This is because hazardous waste fuels generally have
higher levels of metals and chlorine than the fossil fuels they
displace, thus metal and chlorine feedrates and emissions may increase
as the hazardous waste firing rate increases.
Although we are not using the aggregate MTEC approach to evaluate
feed control in today's proposal, the SRE/Feed approach explained in
Part Two, Section VI.A, does assess each source's metal and chlorine
hazardous waste feed control levels. In order to avoid the hazardous
waste firing rate bias discussed above for energy recovery

[[Page 21220]]

units, we believe it is appropriate to instead assess feed control for
energy recovery units by ranking each source's thermal feed
concentration, which is equivalent to the mass of metal or chlorine in
the hazardous waste per million BTUs hazardous waste fired to the
combustion unit. This approach not only normalizes metal and chlorine
feedrates across sources of different sizes, but also normalizes these
feedrates across energy recovery units with different hazardous waste
firing rates. For example, a kiln that feeds hazardous waste with a
given metal concentration to fulfill 100% of its energy demand would be
an equally ranked feed control source when compared to an identical
kiln that fulfills 50% of its energy demand from coal and 50% from
hazardous waste with an identical metal concentration.
Similarly, it is our preference to express today's proposed
emission standards for metals and chlorine in units of hazardous waste
thermal emissions as opposed to expressing the standards in units of
stack gas concentrations. As previously discussed, hazardous waste
thermal emission standards are expressed as mass of HAP emissions
attributable to the hazardous waste per million Btu hazardous waste
fired to combustor. As with thermal feed concentration, thermal
emissions normalizes emissions across energy recovery units with
different hazardous waste firing rates. The hazardous waste thermal
emissions format addresses two concerns. First, it avoids the above
discussed bias against sources that burn hazardous waste fuels at high
firing rates. We prefer not to discourage energy recovery from
hazardous waste as opposed to potentially establishing standards that
effectively restrict the hazardous waste firing rate in an energy
recovery combustor. (See, for example, the requirement in CAA section
112(d)(2) to take energy considerations into account when promulgating
MACT standards, as well as the objective in RCRA section 1003(b)(6) to
encourage properly conducted recycling and reuse of hazardous waste).
Second, because the hazardous waste thermal emissions approach
controls only emissions attributable to the hazardous waste feed (see
discussion in following section), the rule can be simplified by not
including waivers for sources that cannot meet the standard because of
metals or chlorine contributed by nonhazardous waste feedstreams. To
ensure that hazardous waste combustors will be able to achieve the
standards if they use MACT control for metals and chlorine attributable
to the hazardous waste feed, but irrespective of metals and chlorine in
nonhazardous waste feedstreams, current MACT standards for cement and
lightweight aggregate kilns that burn hazardous waste provide
alternative standards that sources can request under a petitioning
procedure. See Sec. 63.1206(b)(9-10). These alternative standards
would be unnecessary under the hazardous waste thermal emissions
approach because, by definition, the approach controls only hazardous
waste-derived metals and chlorine.
2. Which Standards Would Use the Hazardous Waste Thermal Emissions Format?
We propose a hazardous waste thermal emissions format for mercury,
semivolatile metals, low volatile metals, and total chlorine (i.e., the
HAPs found in hazardous waste fuels) for source categories that burn
hazardous waste fuels where we have data to calculate a hazardous waste
thermal emissions limit. Cement kilns, lightweight aggregate kilns and
liquid-fuel fired boilers burn hazardous waste fuels and are thus
candidates for the hazardous waste thermal emission standards.
Incinerators and solid fuel-fired boilers are not candidates for
thermal emission standards because some sources within these source
categories do not combust hazardous waste for energy recovery, i.e.,
they burn low heating value hazardous waste for the purpose of treating
the waste.\40\ Consequently, these sources could not duplicate a
hazardous waste thermal emissions standard based on emissions from
sources that burn hazardous waste fuels, even though their stack gas
emission concentrations could be as low or lower than emissions from a
best performing source under the hazardous waste thermal emissions approach.
---------------------------------------------------------------------------

\40\ Three of the 13 solid fuel-fired boilers burn low heating
value hazardous waste for treatment.
---------------------------------------------------------------------------

We propose a hazardous waste thermal emissions format for all HAP
for which we can apportion emissions between the hazardous waste fuel
feed and other feedstreams. Under this approach, we apportion total
stack emissions between hazardous waste fuel and other feedstreams
using the ratio of the feedrate contribution from hazardous waste to
the total feedrate of the pollutant. Thus, the particulate matter,
metals, and total chlorine standards are candidates because we often
have data on hazardous waste and total feedrates of these pollutants.
We believe, however, that a hazardous waste thermal emissions
format is not appropriate for particulate matter for cement and
lightweight aggregate kilns because particulate matter emissions from
cement and lightweight aggregate kilns are primarily entrained raw
material, not ash contributed by the hazardous waste fuel. There is
therefore no correlation between particulate matter emissions and
hazardous waste thermal input rate.
In addition, please note that we could have expressed the proposed
particulate matter standard for liquid boilers in units of hazardous
waste thermal emissions since (unlike the case of kilns just discussed)
particulate matter emissions are attributable to the hazardous waste
fuel. However, for consistency, we elected to use the same format for
all the particulate matter standards. We invite comment as to whether
the particulate matter standard for liquid boilers should be expressed
in units of hazardous waste thermal emissions.
We do not have adequate data to establish hazardous waste thermal
emissions-based standards for several cases. An example is when we have
only normal feedrate and emissions data (e.g., the mercury standard for
cement kilns). We prefer to establish emission standards under the
hazardous waste thermal emissions format using compliance test data
because the metals and chlorine feedrate information from compliance
tests that we use to apportion emissions to calculate emissions
attributable to hazardous waste are more reliable than feedrate data
measured during testing under normal, typical operations.\41\ Thus, as
a general rule, we prefer to express emission standards for energy
recovery units using the hazardous waste thermal emissions format only
when we have sufficient compliance test feed data.\42\ These situations
are discussed below in more detail in Part Two, Sections VIII, IX, and
XI where we discuss the rationale for the proposed emission standards
for energy recovery units.
---------------------------------------------------------------------------

\41\ Feedrate data from testing during normal, typical
operations may not be as accurate as data from compliance testing
because of the sampling and analytical error associated with low
feedrates. In contrast, sources generally spike metals and chlorine
during compliance testing, so that measurement error is somewhat
masked by the higher feedrate values.
\42\ Two exceptions are the mercury and semivolatile metal
standard for liquid fuel-fired boilers. We propose to express this
standard in the hazardous waste thermal emissions format even though
it is based on normal test data because we do not use feedrate data
to apportion emissions in this case. Rather, we assume semivolatile
metal emissions from liquid fuel-fired boilers are attributable
solely to the hazardous waste given that these sources co-fire
hazardous waste with natural gas or, in a few cases, fuel oil.

---------------------------------------------------------------------------

[[Page 21221]]

3. How Are Emissions From Other Feedstreams Regulated Under the
Hazardous Waste Thermal Emissions Format?
Under the thermal emissions format, only emissions of HAP
contributed by the hazardous waste are directly regulated by today's
proposed standards. Non-mercury metal HAP emissions from raw materials
and fossil fuels would be subject to MACT standards, even though it may
not be feasible to directly control their feedrate. We are proposing
standards for particulate matter as surrogates to control these HAP
metals contributed by raw materials and fossil fuel.

C. What Is the Rationale for Selecting Surrogates To Control Multiple HAP?

HWCs can emit a wide variety of HAP, depending on the types and
concentrations of pollutants in the hazardous waste feed. Because of
the large number of HAP potentially present in emissions, we propose to
use several surrogates to control multiple HAP. This will reduce the
burden of implementation and compliance on both regulators and the
regulated community.
1. Surrogates for Metal HAP
We are proposing to control metal HAP emissions attributable to the
hazardous waste by subjecting sources to metal and particulate matter
emission limitations.\43\ We grouped metal HAP according to their
volatility because volatility is a primary consideration when selecting
an emission control technology.\44\ We then considered the following to
identify metals that would be ``enumerated'' and directly controlled
with an emission limit: (1) The amount of available data for the metal
HAP; (2) the potential for hazardous waste to contain substantial
levels of a metal; and (3) the toxicity of the metal. Other,
``nonenumerated'' metal HAP would be controlled using particulate
matter as a surrogate.
---------------------------------------------------------------------------

\43\ As discussed later, we are also propsoing particulate
matter standards to generally serve as surrogates to control
relevant metal HAP in non-hazardous waste feed streams when appropriate.
\44\ See 64 FR at 52845-47 (September 30, 1999).
---------------------------------------------------------------------------

Mercury is highly volatile, especially toxic, and may not be
controllable by the same air pollution control mechanisms as the other
HAP metals, so we are proposing a standard for mercury individually.
Two semivolatile metals can be prevalent in hazardous waste and are
particularly hazardous: lead and cadmium. We group these two metals
together and propose an emission standard for these metals, combined.
The combined emissions of lead and cadmium cannot exceed the
semivolatile metal emission limit. Three low volatile metals can be
prevalent in hazardous waste and are particularly hazardous: arsenic,
beryllium, and chromium. We group these three metals together and
propose an emission standard for these metals, combined. The combined
emissions of arsenic, beryllium, and chromium cannot exceed the low
volatile metal emission limit.
The particulate matter standard generally serves as a surrogate to
control non-enumerated metals in the hazardous waste as well as a
surrogate to control relevant metal HAP in non-hazardous waste feed
streams. We generally chose not to propose numerical metal HAP emission
standards that would have accounted for all metal HAP for two reasons
(note that such an approach would be in lieu of a proposed particulate
matter standard because particulate matter is not a listed HAP). We
generally do not have as much compliance test emissions information in
our database for the nonenumerated metal HAP compared to the enumerated
metal HAP. Thus it would be more difficult to assess the control levels
for these additional metals. We also believe that a particulate matter
standard, in lieu of emission standards that directly regulate all the
metals, simplifies compliance activities in that sources would not have
to monitor feed control levels of these nonenumerated metals on a
continuous basis.
Note that particulate matter is not an appropriate surrogate where
standards are based, in part (or in whole) on feedrate control. This is
because, unlike the case where HAP metals are controlled by air
pollution control devices, HAP metal reductions in hazardous waste
feedrate are not necessarily correlated with particulate matter
reductions, i.e., hazardous waste feedrate reductions could reduce HAP
metal emissions without a correlated reduction in particulate matter
emissions. (See National Lime, 233 F. 3d at 639 noting this
possibility.) Moreover, particulate matter that is emitted generally
contain greater percentages of HAP metals when the metal concentrations
in the hazardous waste feed increase. Thus, low particulate matter
emissions do not necessarily guarantee low metal HAP emissions,
especially in instances where the hazardous waste feeds are highly
concentrated with metal HAP.
We do not believe that the proposed emission standards for
semivolatile and low volatile metals serve as adequate surrogate
control for the nonenumerated metal HAP. Compliance with the
semivolatile and low volatile metal emission standards does not ensure
that sources are using MACT back-end control devices because they could
be achieving compliance by primarily implementing hazardous waste feed
control for the enumerated metals. Thus, if a source uses superior feed
control only for the enumerated metals, the nonenumerated metal
emissions would not be controlled to MACT levels if it were not using a
MACT particulate matter control device. The proposed semivolatile and
low volatile metal standards are also inappropriate surrogates for
controlling nonmercury metal HAP in the nonhazardous waste feedstreams
for kilns and solid fuel-fired boilers for the same reason. These
sources may comply with the proposed semivolatile and low volatile
metal emission standards by implementing hazardous waste feed control.
This would not assure that the nonmercury metal HAP emissions
attributable to the nonhazardous waste feedstreams are controlled to
MACT levels. A particulate matter standard provides this assurance.
Note that we are proposing that incinerators and liquid boilers
that emit particulate matter at levels higher than the proposed
standard but do not emit significant levels of non-mercury metal HAP
can elect to comply with an alternative standard. Under the proposed
alternative standard, these sources would be required to: (1) Limit
emissions of all semivolatile metals, including nonenumerated
semivolatile metals, to the emission limit for semivolatile metals; and
(2) limit emissions of all low volatile metals, including nonenumerated
low volatile metals, to the emission limit for low volatile metals. See
Part Two, Section XVIII for more discussion on this alternative.
2. Surrogates for Organic HAP
For Phase II sources, we propose two standards as surrogates to
control emissions of organic HAP: carbon monoxide or hydrocarbons, and
destruction and removal efficiency.\45\ Both of these standards control
organic HAP by ensuring combustors are operating under good combustion

[[Page 21222]]

practices that should result in destruction of the organic HAP. Note
that boilers and hydrochloric acid production furnaces that burn
hazardous waste are currently subject to RCRA requirements that
regulate carbon monoxide or hydrocarbon emissions and destruction and
removal efficiency standard under RCRA regulations. We propose to
control dioxin/furans by a separate standard because dioxin/furan can
also be formed post-combustion in ductwork, waste heat recovery
boilers, or dry air pollution control devices (e.g., electrostatic
precipitators and fabric filters).
---------------------------------------------------------------------------

\45\ Please note that we are proposing the organic emission
standards--carbon monoxide or hydrocarbons, and desturction and
removal efficiency--for boilers and process heaters and hydrochloric
acid production furnaces only. Requirements to comply with these
standards are currently in effect under subpart EEE for
incinerators, cement kilns, and lightweight aggregate kilns. We are
not reporposing or reopening consideration of those standards in
today's notice.
---------------------------------------------------------------------------

Hydrocarbon emissions are a direct measure of many organic
compounds, including organic HAP. Carbon monoxide emissions are a more
conservative indicator of hydrocarbon and organic HAP emissions because
the presence of carbon monoxide at elevated levels is indicative of
incomplete oxidation of organic compounds. Sources generally choose to
comply with the carbon monoxide standard because carbon monoxide
continuous emissions monitors are less expensive and easier to maintain
than hydrocarbon monitors.
We also propose to use the destruction and removal efficiency
standard to help ensure boilers and hydrochloric acid production
furnaces operate under good combustion conditions. We propose to adopt
the standard and implementation procedures that currently apply to
these sources under RCRA regulations at Sec. 266.104. We propose,
however, to require a one-time only compliance requirement for
destruction and removal efficiency, unless a source changes its design
or operation in a manner that could adversely affect its ability to
meet the destruction and removal efficiency standard. Further, previous
destruction and removal efficiency testing performed under RCRA could
be used to document the one-time compliance.

D. What Is the Rationale for Requiring Compliance With Operating
Parameter Limits To Ensure Compliance With Emission Standards?

In addition to meeting emission limits, today's proposal would
require sources to establish limits on key operating parameters for the
combustor and emission control devices. Each source would establish
site-specific limits for the parameters based on operations during the
comprehensive performance test, using prescribed procedures for
calculating the limits. The operating parameter limits would reasonably
ensure that the combustor and emission control devices continue to
operate in a manner that will achieve the same level of control as
during the comprehensive performance test.
We selected the operating parameters for which sources would
establish limits because: (1) The parameters can substantially affect
emissions of HAP; (2) they are feasible to monitor continuously; (3)
they are currently used to monitor performance under the Interim
Standards Rule for incinerators, cement kilns, and lightweight
aggregate kilns that burn hazardous waste; and (4) this is the same
general compliance approach that is currently applicable to all
hazardous waste combustion sources pursuant to the RCRA emission
standard requirements.

V. How Did EPA Determine the Proposed Emission Limitations for New and
Existing Units?

A. How Did EPA Determine the Proposed Emission Limitations for New Units?

All standards established pursuant to section 112 of the CAA must
reflect MACT, the maximum degree of reduction in emissions of air
pollutants that the Administrator, taking into consideration the cost
of achieving such emission reduction, and any non-air quality health
and environmental impacts and energy requirements, determines is
achievable for each category. The CAA specifies that the degree of
reduction in emissions that is deemed achievable for new hazardous
waste combustors must be at least as stringent as the emissions control
that is achieved in practice by the best-controlled similar unit (as
noted earlier, this specified level of minimum stringency is referred
to as the MACT floor, the term used when the statutory provision was
first introduced in Congress). However, EPA may not consider costs or
other impacts in determining the MACT floor. EPA may adopt a standard
that is more stringent than the floor (i.e., a beyond-the-floor
standard) if the Administrator considers the standard to be achievable
after considering cost, environmental, and energy impacts.

B. How Did EPA Determine the Proposed Emission Limitations for Existing
Units?

For existing sources, MACT can be less stringent than standards for
new sources, but cannot be less stringent than the average emission
limitation achieved by the best-performing 12 percent of existing
sources for categories and subcategories with 30 or more sources. EPA
may not consider costs or other impacts in determining the MACT floor.
The EPA may require a control option that is more stringent than the
floor (beyond-the-floor) if the Administrator considers the cost,
environmental, and energy impacts to be reasonable.
It has been argued that EPA is limited in the level of performance
it can evaluate in assessing which are the 12 percent existing best
performing sources to standards codified in permits, or other
regulatory limitations. The argument is based on use of the term
``emission limitation'' in section 112 (d) (3), the argument being that
``emission limitation'' is a term defined in section 302 (k) to mean
``a requirement established by the State or the Administrator which
limits the quantity, rate, or concentration of air pollutants * * *''.
EPA does not accept this argument, and indeed doubts that such an
interpretation of the statute is even permissible. In brief:
(i) Statutory text indicates that MACT floors for existing sources
is to based on actual performance. Section 112 (d) (3) (A) speaks to
the actual performance of sources, and requires that the floor for
existing sources reflect actual performance. The key statutory phrase
is not just ``emission limitation'' but ``emission limitation
achieved'', a phrase referring to actual performance, not just a limit
simply set out in a permit or regulation. The floor is to be calculated
using ``emissions information'', a reference again to actual
performance. The provision likewise states that certain sources
achieving a lowest achievable emission rate (LAER) level of performance
without being subject to LAER (a regulatory limit) are not to be
considered in assessing best performers, redundant language if only
regulatory limits could be considered.
In fact, it is clear from context when Congress used the term
``emission limitation'' to refer to regulatory limits, and when it uses
the term to refer to a level of performance actually achieved. Compare
CAA section 111(b)(1)(B) (EPA is to consider ``emissions limitations
and percent reductions achieved in practice'' when considering whether
to revise new source performance standards) with section 110(a)(2)(A)
(State Implementation Plans must contain ``enforceable emission
limitations'').
(ii) The argument leads to absurd and illegal results. The argument
that existing source MACT floors can only be based on regulatory limits
leads to results that are illegal, absurd, or both. Congress enacted
section 112 to assure technology-based control of HAP which had
heretofore gone unregulated due to the vagaries and glacial pace of

[[Page 21223]]

implementing the previous risk-based regime for HAP. 1 Legislative
History at 790, 860; 2 Legislative History at 3174-78, 3340-42. The
result, at the time of the 1990 amendments is that there were
widespread regulatory limits for only one of the 190 listed HAPs (lead,
for which there was a National Ambient Air Quality Standard) plus
NESHAPs for a half dozen other HAPs. Thus, ``emission limitations'', in
the sense used in the argument, did not exist for most HAPs. This would
lead necessarily to the result of no existing source floors because no
``emission limitations'' exist. This result is illegal. National Lime
v. EPA, 233 F. 3d 625, 634 (D.C. Cir. 2000). Where regulatory limits
are higher than actual performance levels, existing source floors
likewise would be higher than performance levels, a result both absurd
and illegal. Sierra Club v. EPA, 167 F. 3d 658, 662-63 (D.C. Cir.
1999). In fact, at the time of the 1999 rule for this source category
(hazardous waste combustion), RCRA regulatory limits were higher than
the level of performance achieved even by the very worst performing
source in the category (for some HAPs, by orders of magnitude). Yet
under the argument, the floor for existing sources would have to be
higher than even this worst performing single source.
(iii) Legislative History shows that Congress intended the existing
source floor to reflect actual best performance. The legislative
history to the MACT floor provision for existing sources likewise makes
clear that the standard was to reflect actual performance, not
regulatory limits. 2 Legislative History pp. 2887, 2898; 3353; 1
Legislative History p. 870. The legislative history to the parallel
provision for municipal waste combusters in section 129(a)(2) (which
floor requirement reads identically to section 112(d)(3)) is equally
clear, stating that the floor for such sources is to reflect emission
limitations which either have been achieved in practice or are
reflected in permit limitations, whichever is more stringent. See
Sierra Club v. EPA, 167 F. 3d at 662 (noting this legislative history.)
(iv) The argument has already been rejected in litigation. The D.C.
Circuit, in the three cases dealing with MACT floors, has held in all
three cases that the floor standard must reflect actual performance.
Sierra Club, 167 F. 3d at 162-63; National Lime, 233 F. 3d at 632;
Cement Kiln Recycling Coalition, 255 F. 3d at 865-66.
For these reasons, we reject the argument that existing source
floors are compelled to reflect only regulatory limits. Such limits may
be a permissible means of establishing existing source floors, but only
if regulatory limits ``are a reasonable means of estimating the
performance of the top 12 percent of [sources]
in each [category or
subcategory].'' Sierra Club, 167 F. 3d at 661.
Somewhat ironically, there is a regulatory limit which is relevant
in establishing floors for incinerators, cement kilns and lightweight
aggregate kilns. The interim standards fix a level of performance for
all of these sources. Thus, any floor standard can be no less stringent
than this standard (see National Lime 233 F. 3d at 640 (reason for
which a level of performance is being achieved is irrelevant in
ascertaining MACT floors)). Based on actual performance, however,
floors may be more stringent.

VI. How Did EPA Determine the MACT Floor for Existing and New Units?

We followed five basic steps to calculate the proposed MACT floors.
First, we determined which MACT methodology approach is most
appropriate to apply to the given pollutant for each source category.
Second, we selected which of the available emissions data best
represent each source's performance. Third, we evaluated whether it is
appropriate to issue separate emissions standards for various
subcategories. Fourth, we identified the best performing sources based
on the chosen methodology and data. Finally, we calculated floor levels
for new and existing sources. The following sections include a
description of each of these steps. Please note that we are also
proposing to invoke CAA section 112(d)(4) to establish risk-based
standards on a site-specific basis for total chlorine for hazardous
waste combustors (except for hydrochloric acid production furnaces).
Under the proposed approach, sources may elect to comply with either
risk-based standards or section 112(d) MACT standards. See Part Two,
Section XIII for more details.

A. What MACT Methodology Approaches Are Used To Identify the Best
Performers for the Proposed Floors, and When Are They Applied?

A MACT methodology approach is a set of procedures used to define
and identify the best performing sources consistent with CAA section
112(d)(3). We have developed and used the following three different
MACT methodologies to identify the best performing sources for the full
suite of proposed floor standards for new and existing sources: (1)
System Removal Efficiency (SRE)/Feed approach; (2) Air Pollution
Control Technology Approach; and (3) Emissions-Based approach. These
three methodologies, together with their rationales and when they are
used, are described in the following sections. Note that each
methodology described below assesses best performing sources for each
pollutant or pollutant group independently, often resulting in
different best performers for each pollutant. For a more detailed
description of these methodologies and when they are applied, see USEPA
``Draft Technical Support Document for HWC MACT Replacement Standards,
Volume III: Selection of MACT Standards,'' March 2004, Chapters 7
through 15.
1. What Is SRE/Feed Approach, and When Are We Proposing To Apply It?
The SRE/Feed MACT approach defines best performers as those sources
with the best combined front-end hazardous waste feed control and back-
end air pollution control efficiency as defined by our ranking
procedure. The approach is applicable to HAP whose emissions can be
controlled by controlling the hazardous waste feed of the HAP: metals
and chlorine.\46\
These two parameters--feedrate of metals and chlorine in hazardous
waste, and performance of the emission control device measured by
system removal efficiency \47\ determine emissions of metals and
chlorine contributed by the hazardous waste feed. Back-end air
pollution control is evaluated by assessing each source's pollutant
system removal efficiency, which is a measure of the percentage of HAP
that is emitted compared to the amount fed to the unit. In identifying
system removal efficiency as a measure of best performing, the Agency
is rejecting the notion that ``best performing'' must mean a source
with the lowest absolute rate of emission of a HAP. A source emitting
300 pounds of a HAP, but removing that HAP at a rate of 99.9% from its
emissions, can logically be considered a better performing source than
one emitting 100 pounds of the same HAP but

[[Page 21224]]

removing it at an efficiency of only 50 percent.
---------------------------------------------------------------------------

\46\ The particulate matter standard is used as a surrogate to
control nonmercury metal HAP in the nonhazardous waste feedstreams
and to control the nonenumerated metals in the hazardous waste. As
explained Part Two, Section VI.A.2.b., control of ash feed may not
be an effective technique to control metal HAP. Thus, we do not use
the SRE/Feed approach to identify floor levels for particulate
matter since ash feed control may not be a reliable indicator of
performance.
\47\ Although system removal efficiency measures primarily the
performance of the back-end emission control device, it also
measures any other internal control mechanisms, such as partitioning
of metals to the product in a cement or lightweight aggregate kiln.
---------------------------------------------------------------------------

Use of feedrate and system removal efficiency as measures of
performance is appropriate because these parameters incorporate the
effects of the myriad factors that can indirectly affect emissions,
such as level of maintenance of the combustor or emission control
equipment, and operator training, as well as design and operating
parameters that directly affect performance of the emission control
device (e.g., air to cloth ratio and bag type for a fabric filter; use
of a power controller on an electrostatic precipitator). For example,
an incinerator with a well-designed and operated fabric filter would
have a higher performance rating measured by system removal efficiency
than an identical incinerator equipped with the same fabric filter
which is, in addition, poorly maintained because of inadequate operator
training. Also, although feedrate of metals and chlorine in
nonhazardous waste feedstreams such as raw materials and fossil fuels
fed to a cement kiln can affect HAP emissions substantially, those
emissions can be feasibly controlled only by back-end control (measured
here by system removal efficiency).\48\ This is because neither fuel
switching nor raw material switching is practicable for production
facilities such as cement and lightweight aggregate kiln facilities.
Thus, feedrate of metals and chlorine contributed by the hazardous
waste--the only controllable feed parameter for these sources--is an
appropriate metric.
---------------------------------------------------------------------------

\48\ See discussion in the proposed lime production MACT
explaining why neither raw material or fossil fuel substitution are
available means of controlling the feedrate of HAP. See 67 FR at
78059-61 (Dec. 20, 2002). The rationale for lime kilns also applies
to cement and lightweight aggregate kilns. Briefly, in the context
of floor control: (1) A kiln's principle raw materials (limestone
for cement kilns and clay for lightweight aggregate kilns) are not
available to other kilns; and (2) we are not aware of raw materials,
or sources of coal or oil, that have characteristic and consistent
(low) concentrations of HAP. In the context of beyond-the-floor
control, additional issues include: (1) The cost of transporting raw
materials with lower levels of HAP (if it were feasible to identify
them) would be prohibitive; and (2) although switching from coal or
oil to natural gas would reduce the feedrate of HAP, the limitations
of the natural gas distribution infrastructure are such that natural
gas is not readily available to many sources.
---------------------------------------------------------------------------

For incinerators and solid fuel-fired boilers, feed control is
evaluated by assessing each source's hazardous waste pollutant maximum
theoretical emission concentration.\49\ Feed control for energy
recovery units (cement kilns, lightweight aggregate kilns, and liquid
fuel-fired boilers) are evaluated by assessing each source's hazardous
waste pollutant thermal feed concentration when possible (i.e., when
EPA has sufficient data to make the calculation).
---------------------------------------------------------------------------

\49\ In the 1999 rule, we developed the term maximum theoretical
emissions concentration to compare metals and chlorine feed control
levels across sources of different sizes. See 64 FR at 52854.
Maximum theoretical emissions concentration is defined as the metals
or chlorine feedrate divided by the gas flowrate, and is expressed
in terms of [mu]g/dscm. See Part Two, section IV.B.1 for more
discussion on how we normalize feedrates and emissions across sources.
---------------------------------------------------------------------------

We rank each source's pollutant hazardous waste feed control level
against all the other source's feed control level, assigning a relative
rank of 1 to the source with the lowest, i.e., best, feed control level
and assigning the highest ranking score to the source with the highest,
i.e., worst, feed control level. We do the same with each source's
system removal efficiency. We rank each source's pollutant system
removal efficiency against all the other sources' system removal
efficiencies, assigning a relative rank of 1 to the source with the
highest, i.e., best, system removal efficiency and assigning the
highest ranking score to the source with the lowest, i.e., worst,
system removal efficiency. We then add each source's feed control
ranking score and system removal efficiency ranking score to yield an
SRE/Feed aggregated score. Each source's aggregated score is arrayed
and ranked from lowest to highest, i.e., best to worst, and, for
existing sources, the best performers are the sources at the 12th
percentile aggregate score and below. Floor levels are then calculated
by using the emissions from these best performing sources. The SRE/
Feed-based standards are expressed in units of hazardous waste thermal
emissions when possible for energy recovery units.
Please note that the SRE/Feed approach can occasionally identify a
floor level for new sources that is higher than the floor level for
existing sources, as discussed below in Sections VII to XII. This is
because the source with the best SRE/Feed aggregate score, and thus,
the single best performing source under this approach, does not always
achieve the lowest emissions among the best performing sources after
accounting for emissions variability. In two cases only, the emissions
for the best performing SRE/Feed source, after accounting for emissions
variability, are higher than the average of the best performing five
(or 12%) of sources--the floor for existing sources--after considering
emissions variability.\50\ For example, the single best performing SRE/
Feed source may have both higher emissions and run variability than
other best performing sources. This source's emissions are averaged
with the other best performers to identify the floor level, and its run
variability is dampened when we calculate the floor for existing
sources by pooling run variability across the best performing sources.
When the single best performer's emissions are evaluated individually,
however, a relatively high run variability is not dampened. In those
few situations where the best performing SRE/Feed source has higher
emissions, after accounting for emissions variability (i.e., the
potential floor for new sources), than the floor for existing sources,
we default to the floor for existing sources to identify the floor for
new sources. We request comment on whether it would be more appropriate
to identify the floor for new sources under the SRE/Feed approach by
selecting the source with the lowest emissions among the best
performing existing sources, after considering run variability, rather
than the lowest SRE/Feed aggregate score.
---------------------------------------------------------------------------

\50\ This occurred for the low volatile metal standard for
cement kilns and the mercury standard for solid-fuel fired boilers.
---------------------------------------------------------------------------

The SRE/Feed methodology is generally applied only to HAP where we
can accurately assess each source's relative hazardous waste feed
control and back-end air pollution control: mercury, semivolatile
metals, low volatile metals, and total chlorine. Dioxin/furans are not
considered to be feed control HAP because they generally are not fed
into the combustor; rather, they are formed in the combustor and post
combustion. Also, whereas particulate matter (for all source
categories) and total chlorine (for hydrochloric acid production
furnaces) could be considered to be feed-controlled and back-end
controlled pollutants, we do not believe it is appropriate to assess
feed control as a control mechanism for these situations for reasons
discussed below in Section 2 (largely dealing with the inability to
control HAP in raw material feed or in fossil fuel). As a result, we
did not apply the SRE/Feed approach to these pollutants.
Finally, the SRE/Feed approach is also not applied when we do not
have sufficient compliance test data to accurately assess each source's
relative back-end control efficiency. This occurs in a limited number
of circumstances when the majority of the emissions data reflect normal
operations. The mercury and semivolatile metal standard for liquid
boilers are examples of when we do not believe we possess sufficient
data to accurately assess each source's back end control efficiency
because we are concerned that the normal feed data are too sensitive to
sampling and measurement error to provide a reliable

[[Page 21225]]

system removal efficiency that would be used reliably in the ranking
procedure. Our preference is to use system removal efficiencies that
are based on compliance testing because sources typically spike the
pollutant feeds during these compliance tests to known elevated levels,
resulting in calculated system removal efficiencies that are more reliable.
2. What Are the Air Pollution Control Technology Approaches, and When
Are They Applied?
The air pollution control technology approach is applied in two
situations where we consider it inappropriate to directly assess
hazardous waste feed control--the particulate matter standard for all
sources categories and the total chlorine standard for hydrochloric
acid production furnaces. We apply slightly different methodologies to
each of these situations, as discussed below.
a. What Methodology Was Used To Identify the Best Performing
Sources for the Particulate Matter Floors? The best performing sources
for the proposed particulate matter floor levels are determined using a
methodology that is conceptually similar to that used in the Industrial
Boiler MACT proposal. See 68 FR at 1660. We call this methodology the
``air pollution control technology'' approach because it defines best
performers as those that use the best type of back-end air pollution
control technology.
This methodology first assesses all the back-end control
technologies used by all the sources within the source category, and
ranks the general effectiveness of these control technologies from best
to worst using engineering information and principles. For example, for
particulate matter control, high efficiency particulate air filters may
be ranked as the best air pollution control device, followed by
baghouses, electrostatic precipitators, and high energy wet scrubbers.
In this example, all sources equipped with a high efficiency
particulate air (i.e., HEPA) filter would get the best ranking (e.g.,
``1''), and all sources equipped with high energy wet scrubbers would
get the worst ranking (e.g., 4).
The sources are arrayed and ranked from best to worst based on
their control technology rankings. For existing sources, MACT control
is defined as the control technology or technologies used by the best
12 percent of these sources. For example, using the previous
particulate matter control rankings, if more than 12 percent of the
sources within the source category were using high efficiency
particulate air filters, then MACT control would be defined to be high
efficiency particulate air filters. If 10 percent of all the sources
were equipped with high efficiency particulate air filters, and 4
percent were equipped with baghouses, then MACT control would be
defined as both high efficiency particulate air filters and baghouses.
After the MACT control technology or technologies are determined,
the MACT floor levels are calculated using emissions data from those
sources using MACT control. See Part Two, Section IV.D.3 for more
discussion on the ranking procedure that is used to identify the best
performing sources under this approach. Also see USEPA ``Draft
Technical Support Document for HWC MACT Replacement Standards, Volume
III: Selection of MACT Standards,'' March 2004, Chapter 9, for more
information. This methodology consequently focuses on performance of
the best pollution control device, but does not assess further control
that might result from lower HAP feedrates.\51\
---------------------------------------------------------------------------

\51\ This methodology does not, however, expand the MACT pool to
include sources with emission levels greater than those of the best
12 per cent of performers using MACT control (the approach the Court
in CKRC held was inadequately justified as representing the 12
percent of best performing sources).
---------------------------------------------------------------------------

We believe it is appropriate to identify the best performing
sources using particulate matter emissions from those using MACT back-
end control without considering hazardous waste ash feedrate control.
For cement kilns, lightweight aggregate kilns, and solid fuel-fired
boilers, particulate emissions are largely contributed by non-hazardous
waste feedstreams (i.e., entrained raw material for kilns, and
entrained coal ash for solid fuel-fired boilers). Thus, hazardous waste
feed control is an inappropriate factor to consider when assessing
particulate matter control efficiency. Assessment of, and control of,
total ash feedrate (i.e., hazardous waste plus raw materials and
nonhazardous waste fuel ash feed) would also be inappropriate because,
as discussed below, total ash feedrate may not be a reliable indicator
of a source's emission control level for metal HAP, and could
inappropriately result in a methodology that assesses (and controls)
raw material and/or nonhazardous waste fuel input.
Although particulate matter emissions for incinerators and liquid
fuel-fired boilers are more directly related to these devices'
hazardous waste ash feedrate, the hazardous waste ash feedrate for
these sources may not be a reliable indicator of a source's feedrate
(and emissions) of nonenumerated metal HAP given that the ash feed into
the combustor may contain high or low concentrations of regulated metal
HAP. A source that feeds low levels of ash thus may not be a best
performing source for metal HAP emissions if its metal concentration
levels in its ash are relatively high. Such a source could be
identified as a best performing source because its particulate matter
emissions and ash feed is low, even though its metal HAP emissions are
relatively high. This result would also inappropriately assess and
control elements of the hazardous waste ash feed that are not regulated
HAP (e.g., silica input). For these reasons, using the air pollution
control technology approach to establish particulate matter floors
without explicitly considering ash feedrate is appropriate since it
focuses on the control technology (i.e., back-end air pollution control
technology) that is known to control metal HAP emissions.\52\
---------------------------------------------------------------------------

\52\ Please note that, although we do not explicitly consider
ash feedrate when establishing the particulate matter floor, ash
feedrate is an appropriate and necessary compliance assurance
parameter for incinerators and liquid fuel-fired boilers where ash
from hazardous waste feedstreams contribute substantially (or
entirely) to particulate emissions.
---------------------------------------------------------------------------

b. What Methodology Is Used To Identify the Best Performing Sources
for the Total Chlorine Floor for Hydrochloric Acid Production Furnaces?
We apply the air pollution control technology approach to total
chlorine for hydrochloric acid production furnaces differently. For
this floor calculation, we are proposing to use the same methodology
that the Agency used for the hydrochloric acid production MACT final
rule for sources that do not burn hazardous waste. See 68 FR at 19076.
This methodology defines best performers as those sources with the best
total chlorine system removal efficiency. Each source's total chlorine
system removal efficiency is arrayed and ranked from highest to lowest,
and the best existing performers are the sources at the 12th percentile
ranking and below. We calculate the system removal efficiency floor
level using the total chlorine system removal efficiencies achieved by
these best performing sources. Consistent with the non hazardous waste
hydrochloric acid production MACT final rule, we also propose to allow
sources to comply with a total chlorine stack gas concentration limit
that is calculated by multiplying the highest hazardous waste chlorine
maximum theoretical emission concentration in the data base by 1 minus
the MACT system removal efficiency. This ensures that a source

[[Page 21226]]

complying with the alternative concentration-based standard would not
emit higher levels of total chlorine than a source equipped with wet
scrubbers that achieve MACT system removal efficiency. We believe this
alternative standard is appropriate because it gives sources the option
of complying with the floor by implementing hazardous waste feed
control.\53\
---------------------------------------------------------------------------

\53\ A source could operate with a ``less than MACT'' system
removal efficiency provided that it controls its hazardous waste
chlorine feed levels such that its emissions are lower than the
emission standard.
---------------------------------------------------------------------------

We believe this methodology is appropriate even though it does not
directly assess hazardous waste total chlorine feed control because
these sources are in the business of feeding highly chlorinated
hazardous wastes so that they can recover the chlorine for use in their
production process. Requiring these sources to minimize hazardous waste
chlorine feed would be directly regulating their raw material and would
directly affect their ability to produce their product. Again, in this
situation, we believe it is appropriate to use a methodology approach
that solely focuses on back-end control, since back-end control assures
removal of the target pollutant without inappropriately requiring a
source to control feedstreams in a manner that affects its ability to
produce its intended product.
3. What Is the Emissions-Based Approach, and When Is It Applied?
The emissions-based approach defines best performers as those
sources with the lowest emissions in our database. We array and rank
each source's pollutant emission levels from lowest to highest. The
best existing performers are the sources at the 12th percentile ranking
and below. We calculate floor levels using the emission levels from
these best performing sources. We express the emissions-based standards
in units of hazardous waste thermal emissions when possible for energy
recovery units, and use the approach whenever the SRE/Feed or air
pollution control technology approaches are not used. Specifically, we
use the emissions-based approach for the dioxin/furan floors for all
source categories, and for the mercury and semivolatile metal floors
for liquid fuel-fired boilers.
The SRE/Feed and air pollution technology-based approaches cannot
be used for the dioxin/furan floors because dioxin/furans are generated
in the combustor or post-combustion within the air pollution control
device. Since dioxin/furans are generally not fed to the units, the
SRE/Feed methodology would not properly assess dioxin/furan emission
control performance. In theory, the technology-based approach for
particulate matter could be applied to the dioxin/furan floors.
However, such a technology approach would, for the most part, identify
the same best performers as the emissions-based approach because there
is only one primary control technology being used by all the sources--
temperature control at the inlet to the dry air pollution control device.
The SRE/Feed approach cannot be used for the mercury and
semivolatile metal floors for the liquid fuel-fired boilers because we
do not have sufficient compliance test data to accurately assess each
source's back-end control efficiency. The technology-based approach is
also not appropriate because sources within this source category
control these HAP both by feed control and by back-end control. As a
result, a methodology that considers only one of the two primary
control techniques may not be appropriate.
4. Why Doesn't EPA Simply Apply the Emissions-Based Approach to All
Source Categories and HAP?
Under the most simplistic interpretation of CAA 112(d), we would
apply the emissions-based approach to all source categories and HAP in
calculating floors for existing sources. We considered proposing this
option. As described later in Part Two, Section VI.G, it was one of
three options for which we conducted a complete economics analysis. We
discuss below, however, why we believe the air pollution control
technology and SRE/Feed approaches more reasonably ascertains the
performance of the average of the best 12 percent of existing sources.
a. Why Do We Prefer the SRE/Feed Approach Over the Emissions-Based
Approach? We believe the SRE/Feed approach is a reasonable and
appropriate MACT methodology for the hazardous waste combustion source
categories because it better estimates the performance of the average
of the 12 percent best performing sources, and (as a necessary
corollary) assures that the floor standards would be achievable by such
sources. As previously discussed, we apply the SRE/Feed approach to HAP
that are actively controlled (via floor controls) by both hazardous
waste feed control and back-end air pollution control. There are only
two ways to control emissions of these HAP from these sources--limit
the feedrate of metal and chlorine and remove them prior to venting the
exhaust gas out the stack. These two control mechanisms are used
simultaneously by all sources in this category at varying levels.
We do not believe the lowest emission levels in our data base in
fact represent the full range of emissions achieved in practice by the
best performing sources. Indeed, it would be unlikely if this were the
case, since these data are necessarily ``snapshots'' of emissions from
the source, obtained in one-time testing events.\54\ Notwithstanding
that such testing seeks to encompass much of the variability in system
performance, no single test can be expected to do so. Thus, inherent
variability such as feedrate fluctuation over time due to production
process changes, uncertainties associated with correlations between
operating parameter levels and emissions, precision and accuracy
differences in different testing crews and analytical laboratories, and
changes in emission of materials (SO₂ being an example) that
may cause test method interferences. See generally 64 FR at 52857 and
52587-59.
---------------------------------------------------------------------------

\54\ One-time testing events, however, are a necessity because
Continuous Emission Monitors still do not exist for most of the HAPs
emitted by these sources.
---------------------------------------------------------------------------

An emissions-based approach for cement kilns, lightweight aggregate
kilns, and solid fuel-fired boilers that assesses performance based on
stack gas concentrations (as opposed to hazardous waste thermal
emissions) may not appropriately estimate the performance of the
average of the 12 percent best performing sources given that those best
performers may have low emissions in part because their raw material
and/or fossil fuels contained low levels of HAP during the emissions
test. We do not believe feed control of HAP in raw material and fossil
fuel should be assessed as a MACT floor control primarily because it
could result in floor levels that are not replicable by the best
performing sources, nor duplicable by other sources. See Part Two,
Section VI.A.1.
Moreover, although the emissions-based approach is not facially
inconsistent with section 112 of the Act, there are serious questions
as to whether its applicability here leads to limits that could be
achieved even by the average of the best performing sources (under the
emissions-based approach). The alternative emissions-based floor
Options 1 and 2 discussed in Part Two, Section VI.G result in floor
levels across all HAP that are achievable simultaneously by fewer than
6% of the sources for the cement kiln, incinerator, and liquid fuel-
fired boiler source

[[Page 21227]]

categories.\55\ See USEPA ``Draft Technical Support Document for HWC
MACT Replacement Standards, Volume III: Selection of MACT Standards,''
March 2004, Chapters 10 and 19, for a summary of the simultaneous
achievability analysis. A reason the floors which would result from
this methodology are so low is that there already have been at least
one and, for many of the sources, two rounds of regulatory reduction of
emissions from these sources (under the RCRA rules, and then under the
Interim Standards MACT rules for incinerators and kilns). The
emissions-based approach thus yields results more akin to new source
standards, confirmation being that the levels are not even achievable
as a whole by the average of the 12 percent best performing sources.
The simultaneous achievability of today's proposed floors, for which we
use the SRE/Feed approach for certain HAP preferentially over the
emissions-based approach, is substantially better (but not dramatically
more than 6%) for cement kilns and liquid fuel-fired boilers than the
achievability under the emissions-based approach.
---------------------------------------------------------------------------

\55\ Simultaneous achievability percentages for lightweight
aggregate kilns, solid fuel-fired boilers, and hydrochloric acid
production furnaces must be interpreted differently given that there
are significantly fewer than 30 sources within these source
categories. As a result, we believe that the emission standards
should be simultaneously achievable by at least two or three sources
for these source categories given that CAA 112(d) defines best
performing sources as the average of the best five sources.
---------------------------------------------------------------------------

There are other reasons why the emissions-based approach results in
such low simultaneous achievability percentages. If the emissions-based
approach is applied to feed-controlled HAP, the best performers are
defined as those sources that are either: (1) The lowest feeders; (2)
the best back-end controlled units; or (3) the best combination of
front-end control or back-end control. The emissions-based approach
selects the lowest emitters from the previous three categories and does
not necessarily account for the full range of emissions that are
achieved in practice by well designed and operated feed control units,
well designed and operated back-end controlled units, or well designed
and operated combination of both front-end and back-end controlled
units. As explained below, the SRE/Feed methodology better accounts for
the range of emissions from these well designed and operated
sources.\56\
---------------------------------------------------------------------------

\56\ Note, however, that many of the best performing sources for
the SRE/Feed approach are the same as those for emissions-based
approach, primarily because there is a good correlation between the
SRE/Feed aggregated ranking score and emissions in that the emission
levels generally increase as the as the aggregate ranking score increases.
---------------------------------------------------------------------------

For example, assume we have 100 sources in a hypothetical source
category, and source A is the 5th best feed controlled source and the
30th best back-end controlled source. Source B, on the other hand, is
the 30th best feed controlled source and the 5th best back-end
controlled source. The SRE/Feed ranking procedure would score these two
sources equally, even though their emissions may be different. Let's
also assume that these two sources are among the best performers for
the SRE/Feed approach. We would not expect their emission levels to be
dramatically different under the SRE/Feed approach because source A is
a superior front-end controlled source with a relatively poorer back-
end control device, and source B is a superior back-end controlled
source with relatively poorer feed control. Even though sources A and B
do not have the same emissions, they are both considered to be well
designed and operated sources because they both use a superior
combination of front-end and back-end control. The difference in
emissions merely reflects the range of emissions from well designed and
operated sources.
If the emissions-based approach was applied in the source A and B
example, the source with the higher emissions would have a worse
emission ranking, and thus may not be identified as a best performer.
Thus, even though we would consider this higher emitting source under
the SRE/Feed approach to be a well-designed and operated source, it
would not be capable of achieving the calculated floor level. We
believe this outcome may be problematic, for example, because sources
that are already operating with a well-designed and operated back-end
control unit should not have to upgrade its back-end control technology
simply because it is not achieving a floor level driven, in part, by
other sources within the source category that are implementing lower
feed control rates that are impractical for it to achieve.\57\ It may
be questionable to require these well controlled back-end units to
implement better feed control to achieve this emission-based floor
level because: (1) they may not be capable of implementing feed control
without sending/diverting the waste elsewhere--yet these units are
providing a needed and required service in treating hazardous waste;
and (2) it could be argued that hazardous waste containing high levels
of metals and chlorine should in fact be treated in the well-designed
and operated back-end controlled units (see RCRA sections 3004 (d) to
(m), requiring advanced treatment of hazardous waste before the waste
can be land disposed).
---------------------------------------------------------------------------

\57\ Moreover, the superior low metal and chlorine feedrates
that on-site incinerators and boilers are ``achieving'' may simply
reflect the composition of the waste generated by the manufacturing
operation.
---------------------------------------------------------------------------

Similarly, sources that are already achieving superior feedrate
control should not necessarily have to upgrade their feedrate control
further simply because they are not achieving a floor level driven, in
part, by sources with superior back-end control. Improving already
superior feedrate control may be problematic simply because they may
not be capable of implementing additional feed control (e.g., source
reduction) at their facility, or having generators implement further
feedrate control. EPA believes that hazardous waste feed control is an
important element of what constitutes ``best performing'' sources from
this source category, and does not wish to structure the rule to
discourage the practice by developing standards which do not directly
take this means of control into account. See CAA section 112(d)(2)(A)
(feed control is an explicit means of achieving MACT); and see also the
pollution prevention and waste minimization goals of both the CAA
(sections 112(d) (2) and 101(c) and RCRA (section 1003(b)). The SRE/
Feed approach thus better preserves the opportunity for sources to
achieve the floor levels if they are using either superior front-end
control or back-end control (or superior combination of both). At the
same time, it addresses both means by which sources in this category
can control their HAP emissions: hazardous waste feed control and back-
end air pollution capture through control technology.
The example in the previous paragraph of the source using superior
feed control is clearly applicable to incinerators and boilers that
combust hazardous waste. These are somewhat unique source categories in
that they are comprised of many different industrial sectors that may
not be capable of achieving/duplicating the same metal and chlorine
feedrate control levels of other sources within their respective source
category given that hazardous waste feed control levels are directly
influenced by amount of HAP that are generated in their specific
production process. Similarly, other sources that comprise commercial
hazardous waste combustors (i.e., kilns and commercial incinerators)
are subject to the feed control levels that are governed

[[Page 21228]]

primarily by third parties (i.e., the generators or fuel blenders). The
emissions-based approach identifies the best performers as those
sources with the lowest emissions and does not consider differences in
emission characteristics across all the industrial sectors that combust
hazardous waste. We contemplated whether we should assess if
subcategorization is appropriate based on the various industrial
sectors that combust hazardous waste. We believe, however, that such an
assessment would be difficult given the vast number of industrial
sectors that generate hazardous waste which is treated by combustion.
The emissions-based approach could be identifying a suite of floor
levels across HAP that would require sources to operate at feedrate
control levels in the aggregate that are in theory achieved by few, if
any, well-operated and designed feed controlled sources. For example,
the best performing sources for the emissions-based approach for the
incinerator semivolatile and low volatile metal floors are entirely
different. This may occur because sources have different relative feed
control levels for mercury, semivolatile metals, low volatile metals,
and total chlorine (e.g., a source could have superior semivolatile
metal feed control but only moderate low volatile metal feed control).
Finally, the emissions-based approach may result in low
simultaneous achievability percentages because a back-end control
technology for one pollutant may not control the emissions of another
pollutant as efficiently. For example, wet air pollution control
systems may control total chlorine emissions very well, but are not as
efficient at limiting particulate matter emissions when compared to a
baghouse. Thus, best performers under the emissions-based floor
approach for total chlorine could be driven by sources with wet air
pollution control systems, and the particulate matter floor could be
driven by sources equipped with baghouses, resulting in a combined set
of floors that are conceivably achieved by few sources, a result
confirmed, as noted above, in that less than 6% of existing sources
would be achieving floor standards developed using the emission-based
approach.\58,\ \59\
---------------------------------------------------------------------------

\58\ Although the SRE/Feed approach does not directly address
this issue within the methodology, the simultaneous achievability of
the SRE/Feed-based floors is substantially better (but not
dramatically more than 6%) for cement kilns and liquid fuel-fired
boilers than the achievability under the emissions-based approach.
\59\ Note that we considered using a floor methodology that
simultaneously assesses all the pollutant emissions from each
source. This methodology would define best performers as those
sources with the best aggregate emissions across all (or a subset of
all) the HAP and would perhaps more directly achieve the goal of
obtaining a full suite of emission standards that are achievable by
at least 6% of the sources. We rejected this approach in the 1999
rule, since it could potentially result in least-common denominator
source levels. See 64 FR at 52856. However, at least for
incinerators and kilns, there is less potential concern with such a
result because the Interim Standards have already reduced sources'
emissions of all HAP considerably and the Interim Standards cap the
level of floors for these sources. Nonetheless we may not have
enough complete emissions information for all HAP for many source
categories to adequately assess enough source's true ``aggregate
emissions.'' See Section VI.G.
---------------------------------------------------------------------------

We thus believe that using the SRE/Feed approach preferentially
over the emissions-based approach and technology based approach is
appropriate because use of the SRE/Feed approach results in floor
levels that better reflect the range of emissions from well-designed
and operated sources and also results in floor levels across all HAP
that are achievable simultaneously by at least 6 percent of the sources
within each source category.
b. Why Do We Prefer the Air Pollution Control Technology Approach
Over the Emissions-Based Approach? As previously discussed, we apply
the air pollution control technology approach in two situations where
we consider it inappropriate to directly assess hazardous waste feed
control using an SRE/Feed type approach: the particulate matter
standard for all source categories; and, the total chlorine standard
for hydrochloric acid production furnaces. We discuss below why the
emissions-based approach is not our preferred methodology for these
standards.
For particulate matter, the emissions-based approach identifies the
lowest emitters as best performers, irrespective of the types of
controls that were used. This would not necessarily reflect emissions
that are in fact capable of being achieved by sources using MACT back-
end control technology as defined by the air pollution control
technology approach because, as discussed above, our data are
``snapshots'' of emissions from each source, obtained in one-time
testing events. As a result, the particulate matter floors that are
based on the emissions-based approach would not necessarily account for
inherent variability such as ash feedrate fluctuation over time due to
production process changes,\60\ uncertainties associated with
correlations between operating parameter levels and emissions,
precision and accuracy differences in different testing crews and
analytical laboratories, and changes in emission of materials (SO
₂ being an example) that may cause test method
interferences. The air pollution control technology approach may better
account for this inherent variability because it assesses the emissions
ranges from those sources that utilize the defined back-end MACT
control devices, as opposed to merely selecting the lowest emitters
irrespective of the type of control it uses.
---------------------------------------------------------------------------

\60\ The emissions-based approach may not account for
particulate matter emissions variability factors that are
attributable to factors other than MACT control. For example, two
sources with identical air pollution control devices could have
different particulate matter emission concentrations merely because
they process different types and amounts of raw material and/or
nonhazardous waste fuels. From a MACT perspective, the source with
the higher emissions would not be a poorer performer because feed
control of raw material and nonhazardous waste fuels are not MACT
floor controls.
---------------------------------------------------------------------------

Also, using the emissions-based approach for incinerators and
liquid boilers (for the particulate matter standard) and hydrochloric
acid production furnaces (for the total chlorine standard) is not our
preferred approach because it assesses in part, hazardous waste ash and
chlorine feed control. As discussed above, the emissions-based approach
defines best performers as those sources with the lowest emissions, and
thus inherently accounts for and assesses hazardous waste ash and
chlorine feed control in that sources with lower ash feedrates and
chlorine feedrates may have lower emissions.\61\ This is not our
preferred way of establishing floors for these HAP for the reasons
discussed above in Section A.2.
---------------------------------------------------------------------------

\61\ The best performers identified by the air pollution
technology approach are less likely to be driven by low ash feeding
facilities for the particulate matter standard because all the
sources equipped with MACT-defined back-end control devices
typically feed high levels of ash, thus we believe particulate
matter emission levels from these sources are more a function of the
air pollution control device control efficiency rather than the ash
feed levels.
---------------------------------------------------------------------------

B. How Did EPA Select the Data To Represent Each Source When
Determining Floor Levels?

After we determine which MACT methodology is appropriate for a
given pollutant and source category, we select which of the available
emissions data to use for each source to: (1) Determine if
subcategorization is warranted; (2)

[[Page 21229]]

identify the best performing sources; and (3) calculate the floor
levels. Our emissions data base is complex because it includes, in
part, compliance test data, emissions data that is representative of
the normal operating range of the source, and, for the Phase I sources,
multiple emission test data that have been collected over a number of
years. See Part Two, Section III for more discussion on data base issues.
We follow a general ``data hierarchy'' to determine which of these
data types to use to represent each source's performance (with the
performance being reassessed for each HAP). First, we prefer to
explicitly use compliance test data rather than data representative of
normal operations because compliance test data best reflect the upper
range of emissions from each source and thus best accounts for day-to-
day emissions variability. Use of compliance test data allows us to
express emission floors as ``short-term limits'' (e.g., hourly or
twelve hour rolling averages), which is consistent with the current
interim MACT standard format for incinerators, cement kilns, and
lightweight aggregate kilns. Short-term limits are also consistent with
the RCRA emission standards currently applicable to boilers and
hydrochloric acid production furnaces. Finally, we prefer to use
compliance test data because the majority of the available data are
compliance test data.
Absent sufficient compliance test data for sources within the
source category to calculate floor levels, we default to explicitly
using data that are representative of the source's operating range
under conditions not designed to assess performance variability. Since
these so-called normal data do not typically reflect the upper range of
emissions from each source, we believe it is necessary to account for
emissions variability (in part) by expressing floors that are based on
normal data as long-term, annual average emission limits (since the
snap-shot data, by definition, do not reflect short-term variability).
We considered using all available emissions data to calculate the
floors, irrespective of whether they were normal or compliance test
data. We believe, however, that it is inappropriate to mix such
dissimilar data when calculating floor levels because it would bring
into question how to account for day-to-day emissions variability when
setting the format of the standard. For example, if a floor were
calculated using 50% percent normal data and 50% compliance data,
should the standard be expressed as a long-term limit or short-term
limit? This is critical because the averaging period associated with
the numerical emission limitation affects the stringency of the
standard. It is also unclear how mixing dissimilar data would affect
the statistical variability factor we apply to each floor to assure
that floor levels are achievable by the average of the best performing
sources. As discussed in Part Two, Section VI.E, we apply the
statistical variability factor to the floor levels to assure that the
average of the best performing sources would be able to replicate the
emission test results that were used to calculate the floor levels.
Mixing dissimilar data not only complicates the analyses, but also
could result in inconsistent evaluation of data (hence inconsistent
results), primarily because the ratio of normal data to compliance data
differs across HAP within each source and across all sources. We
therefore believe it is appropriate to assess ``like data'' explicitly
to assure results are consistent across HAP and source categories.
We prefer to use the most recent compliance test data to represent
each source in situations where we have data from multiple test
campaigns that were collected at different times. For example, we
typically have multiple test campaign emission information for cement
kilns and lightweight aggregate kilns because: (1) We conducted a
comprehensive data collection effort for these sources to update the
data base that was used to support the 1999 final rule; and (2) these
sources, prior to receiving their RCRA permit, are required to conduct
emissions tests every three years.
We believe it is appropriate to only use the most recent compliance
test data for a source because those data best reflect current
operations and emission levels. Older compliance test data may not be
representative of current emissions because: (1) Permitted feed and air
pollution control device operating levels may have been changed/
upgraded; (2) combustion unit and associated air pollution control
equipment design may have been changed/upgraded; and (3) standard
operating practices that relate to maintenance and upkeep may have been
changed/upgraded. As a result, we believe that a source's most recent
compliance data best reflect a source's upper range of emissions. We
considered using all of the sources historical compliance emissions
data to perhaps better account for day-to-day emissions variability. We
believe, however, that it is not appropriate to consider older
compliance emission test data to account for day-to-day emission
variability because: (1) The older compliance data may reflect varying
emissions merely because the source was previously operating with
poorer control levels, which is not an appropriate factor to consider
when assessing day-to-day emission variability; and (2) the most recent
compliance test data adequately accounts for day-to-day variability
because the operating levels demonstrated during the most recent
compliance test generally represent the maximum upper range of
operations and emissions.\62\
---------------------------------------------------------------------------

\62\ Operating parameter limits are established based on
compliance test operations to ensure emissions achieved during
normal operations do not exceed the emissions that were demonstrated
in the compliance test.
---------------------------------------------------------------------------

We do not apply the concept of using the most recent emissions test
information to normal emissions data (as previously discussed, we use
normal emission data to calculate floor levels only in situations where
we do not have sufficient compliance test data). We instead use all
normal emissions data that are available because we are concerned that
a source's most recent normal emissions may not be representative of
its average emissions. The most recent normal emissions data could
reflect emissions at the upper range of normal operations or the lower
end of normal operations. If we were to use only the most recent normal
emissions information, we may identify as best performers those sources
that were operating below their average levels. This would be
inappropriate because the floor level may be unachievable by the best
performing sources.
Finally, for liquid fuel-fired and solid fuel-fired boilers, we
eliminated emission test runs from the MACT analysis when we had
information that the source conducted sootblowing during that emission
test run. Boilers that burn fuels with high ash content are designed to
blow the soot off the tubes periodically to maintain proper heat
transfer. The soot can contain metal HAP, and emissions of these HAP
can increase during sootblowing. Although the current RCRA particulate
matter and metals emissions standards for these sources at Sec. Sec.
266.105 and 266.106 do not require sootblowing during compliance
testing, we have provided guidance recommending that sources blow soot
during one of the three runs of a compliance test condition and
calculate average emissions considering the frequency and duration of
sootblowing.\63\ We conclude that these sootblowing run data should not be

[[Page 21230]]

considered when establishing MACT floor, however, for several reasons.
We do not know if all sources that blow soot followed the guidance to
blow soot during a run of the test condition. If they did not, they
could be identified as a best performer but may not be able to achieve
the floor level when blowing soot. In addition, several boilers that
blew soot during a run of the test condition did not use our
recommended approach to calculate time-weighted average emissions
considering the frequency and duration of sootblowing. For these
sources, we cannot calculate time-weighted average emissions. We also
note that, for sources with emission control equipment, emissions
during sootblowing runs are not significantly higher than when not
blowing soot. This is because soot particles are relatively large and
easily controlled. For sources with no emission control equipment,
sootblowing increased particulate matter emissions for some sources,
but not others. In addition, we could not use the sootblowing run to
help address emissions variability by evaluating run variability
because the (in some cases) higher emissions during sootblowing are
unrelated to the factors affecting run variability that we are
evaluating (e.g., method precision and other largely uncontrollable
factors that affect run-to-run emissions during a test condition).
Finally, we note that the Agency did not propose to require sootblowing
to demonstrate compliance with the MACT standards for industrial,
commercial, and institutional boilers and process heaters.\64\ Although
for these reasons we conclude that it is appropriate not to consider
sootblowing run data to establish the MACT floor, we request comment on
alternative views.\65\
---------------------------------------------------------------------------

\63\ USEPA, ``Technical Implementation Document for EPA's Boiler
and Industrial Furnace Regulations'' EPA530-R-92-011, March 1992,
NTIS # PB92-154 947.
\64\ See 68 FR 1660 (January 13, 2003).
\65\ We note that a floor level considering sootblowing may be
higher than a floor level based on discounting sootblowing runs.
---------------------------------------------------------------------------

Because we do not consider sootblowing when establishing floor
levels, sootblowing would not be required during performance testing to
demonstrate compliance with the standards for particulate matter and
semivolatile and low volatile metals.\66\
---------------------------------------------------------------------------

\66\ The comparative risk assessment for this proposed rule did
not evaluate the impact of sootblowing on average emissions. To
ensure that RCRA permits are protective of human health and the
environment, regulatory officials may determine that the effect of
sootblowing on average emissions (i.e., considering the frequency
and duration of sootblowing) should be considered in some
situations, such as a source with uncontrolled or poorly controlled
particulate emissions and with relatively high particulate matter or
toxic metal emissions.
---------------------------------------------------------------------------

C. How Did We Evaluate Whether It Is Appropriate To Issue Separate
Emissions Standards for Various Subcategories?

The third step we use to calculate MACT floor levels evaluates
subcategorization options. CAA section 112(d)(1) allows us to
distinguish among classes, types, and sizes of sources within a
category when establishing floor levels. Subcategorization typically
reflects ``differences in manufacturing process, emission
characteristics, or technical feasibility.'' See 67 FR 78058.
We use both engineering principles and a statistical analysis to
assess whether it is appropriate to subcategorize and issue separate
emission standards. We first use engineering principles to determine
potential subcategory options. These subcategory options are discussed
in more detail in Part Two Section II for each source category. As
discussed in greater detail below, we then determine if there is a
statistical difference in the emission characteristics between these
potential subcategory options. Finally, we conduct a technical analysis
to determine if the statistical analysis results are consistent with
sound engineering judgement.
``Analysis of Variance'' (ANOVA) is the statistical test used to
cross-check these engineering judgements. ANOVA, a conventional
statistical method, evaluates whether there are differences in the mean
of HAP emissions levels from two or more different potential
subcategories (i.e., do the different subcategories of HAP data come
from distinctly different populations). Subcategories are considered
significantly different using a 95% confidence level. ANOVA is used in
combination with engineering principles to sequentially identify
significant differences between various different combinations of
potential subcategories. See U.S. EPA ``Draft Technical Support
Document for HWC MACT Replacement Standards, Volume III: Selection of
MACT Standards,'' March 2004, Chapter 4, for detailed steps and results
of the ANOVA evaluation process.

D. How Did We Rank Each Source's Performance Levels To Identify the
Best Performing Sources for the Three MACT Methodologies?

The fourth step used in determining the MACT floor levels involves
ranking each source's performance level to identify the best
performers. Below we discuss the general ranking procedure used for
each of the three MACT methodologies and the statistical methodology
used to perform the ranking process.
1. Emissions-Based Methodology Ranking Procedure
As previously discussed in Part Two, Section VI.A, the emissions-
based approach defines best performers as those sources with the lowest
emissions in our database. Each source's emission test runs are first
converted to an upper 99% confidence level in order to rank performance
not only on the average emission levels each source achieves, but also
on the emissions variability each source demonstrates during the
emissions tests. We believe this is appropriate because a source's
ability to consistently control its emissions below the MACT floor
levels is important in determining whether a source is in fact a well
designed and operated source.\67\ We then array and rank each source by
its 99% upper confidence emission levels from best to worst (i.e.,
lowest to highest). For existing source floors, we identify the best
performers as either sources at the 12th percentile ranking and below
or the lowest 5 ranked sources values if we have data from less than 30
sources. The best performing source for the new source floor is simply
the source with the single lowest ranked 99% upper confidence emission
level.
---------------------------------------------------------------------------

\67\ For example, a source with average emissions of 100 and
calculated variability of 10 would be ranked as a better performing
source when compared to a source with average emissions of 100 and a
calculated variability of 20.
---------------------------------------------------------------------------

2. SRE/Feed Ranking Procedure
As previously discussed, the SRE/Feed methodology approach defines
best performers as those sources with the best combined front-end
hazardous waste feed control and back-end air pollution control
efficiency as defined by our ranking procedure. The first step involves
ranking each source's feed control level. As with the emissions-based
approach, we first convert each source's feed control run levels (i.e.,
hazardous waste maximum theoretical emission concentration level or
thermal feed concentrations) to an upper 99% confidence level. We then
array each source's 99% upper confidence feed control levels from best
to worst (i.e., lowest to highest). Next we assign a feed control
ranking score to each source. The source with the lowest feed control
value gets a ranking of 1, and the source with highest feed control
value receives the highest numerical ranking.
The second step ranks each source's system removal efficiency,
which is a measure of the percent of metal or

[[Page 21231]]

chlorine that is emitted as compared to the amount fed to the
combustion unit. Again, we first convert each source's system removal
efficiency run values to an upper 99% confidence level value. We then
array each source's 99% upper confidence levels from best to worst
(i.e., highest to lowest). Next we assign a system removal efficiency
ranking score to each source. The source with the best system removal
efficiency gets a ranking of 1, and the source with the worst system
removal efficiency receives the highest numerical ranking.
As with the emissions ranking procedure discussed above, our feed
control and system removal efficiency ranking procedure measures
performance not only on the average feed control and system removal
efficiency level each source achieves, but also on the feed and system
removal efficiency variability each source demonstrates during the
emissions tests. This is appropriate because a source's ability to
consistently regulate its control mechanisms to achieve MACT emissions
is important in determining whether a source is in fact a well designed
and operated source.
Third, we add each source's feed control ranking score and system
removal efficiency ranking score together in order to calculate an
aggregated SRE/Feed score. We then array and rank each source's
aggregated score from best to worst (i.e., lowest to highest). For
existing source floors, we identify the best performers as sources at
the 12th percentile aggregate ranking and below or sources with the
lowest 5 aggregated scores if we have data from less than 30 sources.
The best performing source for the new source floor is simply the
source with the single lowest aggregated score.
3. Technology Approach Ranking Procedure for the Particulate Matter
Standard
As previously discussed in Part Two, Section VI.A.2.a, the best
performing sources for the particulate matter proposed floor levels are
determined from a pool of sources that use the MACT-defining back-end
control technology. We assess only the emissions from those sources
equipped with the MACT-defining control technology (or technologies),
and, as with the previously discussed methodologies, we convert each
source's emission run values to an upper 99% confidence level value.
Emissions information from each source is then grouped based on the
type of MACT control each source uses. The first group contains
emissions information from sources equipped with the best ranked MACT
control device; the second group includes emissions information from
sources equipped with the second best ranked MACT control technology
(if there is more than MACT control technology), and so on.
We then array and rank each source's 99% upper confidence emission
levels from best to worst (i.e., lowest to highest) within each of
these groups. If there is only one defined MACT control technology, the
best performing sources are those sources with the lowest 99% upper
confidence emission levels amongst the sources using this MACT control
technology. The lowest emitting sources are added to a list of best
performers up until the number of sources that are included in this
list is representative of 12 percent of sources within the source
category (for the existing source floor determination). If there is
more than one defined MACT control technology, the list of best
performers first considers sources with the lowest 99% upper confidence
emission levels that are equipped with the best ranked control device
up until the number of sources that are included in this list is
representative of 12 percent of sources within the sources category. If
additional sources need to be added to this list to appropriately
represent 12% of the sources within the source category, then sources
with the lowest emissions that are equipped with the second best MACT
control device are added until the appropriate number of best
performing sources are obtained.\68\ For the new source floor, the best
performer is simply the single source equipped with the best ranked
MACT control device with the lowest 99% upper confidence emission level.
---------------------------------------------------------------------------

\68\ Note that this methodolgy does not base the floor on the
highest emitting source amongst these best performers (as did the
``expanded MACT pool'' did for 1999 rule). Rather, the floor is
determined by calculating the average performance of all best
performing sources.
---------------------------------------------------------------------------

4. Technology Approach Ranking Procedure for the Total Chlorine Floor
for Hydrochloric Acid Production Furnaces
As previously discussed in Part Two, Section VI.A.2.b, the
technology approach used to determine the total chlorine floor levels
for hydrochloric acid production furnaces defines best performers as
those sources with the best total chlorine system removal efficiency.
The ranking procedure used for this methodology is identical to that
used in the emissions-based approach with the exception that system
removal efficiencies are ranked instead of emissions. Each source's
total chlorine system removal efficiency run values are first converted
to an upper 99% confidence level. We then array and rank each source's
99% upper confidence system removal efficiencies from best to worst
(i.e., highest to lowest). For existing source floors, we define best
performers as either: (1) Sources at the 12th percentile ranking and
below; or (2) sources with the lowest 5 rankings if we have data from
less than 30 sources. The best performing source for the new source
floor is simply the source with the single highest 99% upper confidence
system removal efficiency.
5. Description of the Statistical Procedures Used To Identify the 99%
Confidence Levels
As previously discussed, each source's performance level is first
converted to an upper 99% confidence level in order to rank performance
not only on the average performance level each source achieves, but
also on the emissions variability each source demonstrates during the
emissions tests. We believe this is appropriate because a source's
ability to consistently control its emissions below the MACT floor
levels is important in determining whether a source is in fact a well
designed and operated source.
Sources are ranked based on their projected ``upper 99% confidence
limit'' (or lower 99% confidence limit for system removal efficiency).
For emissions and feedrates, upper 99% confidence limits are determined
using a ``prediction limit'' calculation procedure. The prediction
limit is an estimate of the level which will capture 99 out of 100
future test condition averages (where each average comprise three
individual test runs). HAP emissions data within each source are
determined to be normally distributed. The prediction limit is
calculated for each source based on the average, standard deviation,
and number of individual test runs.
For system removal efficiencies, the lower 99% confidence limit is
determined using the ``two parameter Beta distribution''. The beta
distribution is used for modeling proportions, i.e., system removal
efficiencies, is highly robust, and appropriately bounded by zero and
1. Beta distribution modeling parameters are determined based on the
``method of moments'' using the average and standard deviation of the
individual source data. The lower 99% estimate comes directly from the
Beta distribution model. See USEPA ``Draft Technical Support Document
for HWC MACT Replacement Standards, Volume III: Selection of MACT Standards,''

[[Page 21232]]

March 2004, Chapter 8, for further discussion.

E. How Did EPA Calculate Floor Levels That Are Achievable for the
Average of the Best Performing Sources?

The emissions data we used to establish MACT floor were obtained by
manual sampling of stack gas. To ensure that the average of the best
performing sources can routinely achieve the floor during future
performance testing under the MACT standards, we must account for
emissions variability.
We account for long-term emissions variability by: (1) Using
compliance test emissions data, when available, to establish floors;
(2) when other than compliance test data must be used to establish the
floor, basing compliance on an annual average. In addition, we add a
statistically-derived variability factor to the floor to account for
run-to-run variability. This variability factor ensures that the
average of the best performing sources can achieve the floor level in
99 of 100 future tests if the best performing sources replicate the
operating conditions and other factors that affect the emissions we use
to represent the performance of those sources.
1. How Does Using Compliance Test Data Account for Variability?
We use RCRA compliance test emissions data, when available, to
establish the floors because compliance test data largely account for
emissions variability. Under RCRA compliance testing, sources must
establish operating limits based on operating conditions demonstrated
during the test. Each source designs the compliance test such that the
operating limits it establishes account for the variability of
operating parameter levels it expects to encounter during its normal
operations (e.g., feedrate of metals and chlorine; air pollution
control device operating parameters, production rate). Thus, operating
conditions during these tests generally reflect the upper range of
emissions from these sources. Using a source's compliance test
emissions to establish the floor accounts largely for long-term
emissions variability. However, this does not necessarily account for
factors that affect variability. As previously discussed, our snap-shot
data base emissions information does not necessarily account for
inherent variability such as feedrate fluctuation over time due to
production process changes, uncertainties associated with correlations
between operating parameter levels and emissions, precision and
accuracy differences that may result from using different stack
sampling crews and analytical laboratories, and changes in emission of
materials (SO₂ being an example) that may cause test method
interferences.
Use of compliance test data also does not account for run-to-run
variability. We thus use a statistically-derived variability factor to
account for the variability in emissions that would result if the best
performing sources were to replicate their compliance tests, as
discussed below.\69\
---------------------------------------------------------------------------

\69\ EPA did not statistically assess run-to-run variability in
the 1999 rule (although we noted that it existed; see 64 FR at
52857. The reason is that by using the expanded MACT pool approach
to account for variability (using surrogate sources from outside the
best performing to assess the best performing sources' variability)
we felt we had accounted for all such run-to-run variability. Id.
Since we are not proposing to expand the MACT pool here, it is
necessary to account for run-to-run variability by some other means.
---------------------------------------------------------------------------

In addition, use of compliance test data may not account for long-
term variability of particulate matter emissions from sources equipped
with a fabric filter. Accordingly, we also use a statistically-derived
variability factor to account for this variability, as discussed below.
2. How Does Using Long-Term Averaging Account for Emissions Variability
When Using Other Than Compliance Test Data?
RCRA compliance test emissions data are not available for some
metals (mercury in particular) for some source categories. In these
cases, we use other emissions test data to establish the floor. These
other test data are snap shots of emissions within the range of normal
emissions. To largely account for emissions variability when using
emissions data assumed to represent the average of normal emissions, we
propose to express the floor as a long-term, yearly, average. Sources
would comply with the floor by establishing limits on metal feedrate
and air pollution control device operating parameters. Compliance with
the metal feedrate limits would be based on an annual average feedrate,
while compliance with the air pollution control device operating limits
would be based on short-term limits (e.g., hourly rolling average). We
propose short-term averages for air pollution control device operating
parameters because the parameters may not correlate with emissions
linearly; emissions resulting when an air pollution control device
parameter is above the limit thus may not be offset by emissions
resulting when the air pollution control device parameter is below the
limit. See 1999 rule, 64 FR at 52920.
As discussed above, we also use a statistically derived variability
factor to account for the variability in emissions that would result if
the best performing sources were to replicate the emissions tests we
use to establish the floor, as discussed below.
We use the normal emissions data to represent the average emissions
from a source even though we do not know where the emissions may fall
within the range of normal emissions; the emissions may be at the high
end, low end, or close to the average emissions. It may be reasonable
to assume the emissions represent average emissions, given that we have
emissions data from several sources, and that emissions for these
sources in the aggregate could be expected to fall anywhere within the
range of normal emissions. Note that, as previously discussed, we have
not applied the concept of using the most recent emissions test
information to normal emissions data because we are concerned a
source's most recent normal emissions may not be representative of a
source's true average emissions. These emissions could reflect
emissions at the upper range of normal operations, or instead, could
reflect emissions at the lower end of normal operations. If we were to
use only the most recent normal emissions information, the MACT
standard setting process may identify best performers as those sources
that operate below their normal levels. This may be inappropriate
because the floor level may be unachievable even by the best performing
sources. We invite comment as to whether floors that are based on
normal data are in fact achievable by the best performing sources, and
whether there is perhaps a more appropriate method to identify floors
that are based on normal data.
3. What Statistical Procedures Did EPA Use To Calculate Floor Levels?
In order to calculate a floor that would be achievable by the
average of the best performing sources, we considered the variability
in emissions across runs of the test conditions of the best performing
sources. We also use statistical procedures to account for long-term
variability in particulate matter emissions for sources equipped with
fabric filters. We discuss these procedures and the rationale for using
them below.
a. Run-to-Run Variability. The MACT floor level is determined by
modeling a normally distributed population that has an average and
variability that are equal to that of the ``average'' of the best
performing MACT pool sources. The MACT floor is calculated using a

[[Page 21233]]

modified prediction limit procedure. The prediction limit is designed
to capture 99 out of 100 future three-run averages from the ``average''
of the best performing MACT sources.
Specifically, the modified prediction limit for calculating the
MACT floor is the sum of the average of the best performing sources and
the ``pooled'' variability of the best performing sources. The pooled
variability term accounts for the expected variability in future
measurements due to variations resulting from system operation and
measurement activities. The pooled variability term is based in part on
the observed variance of individual runs within test conditions from
the best performing MACT pool sources. The pooled variability term
assumes that variability from the individual best performing sources
are independent (not related), and thus are additive (and not
averaged). The pooled variability term is a function of the variances
of the individual MACT pool sources, the number of MACT pool sources,
the desired 99% confidence level, and the number of future test runs
for demonstrating compliance (assumed to be 3). See USEPA ``Draft
Technical Support Document for HWC MACT Replacement Standards, Volume
III: Selection of MACT Standards,'' March 2004, Chapter 7, for
discussion of the detailed steps and prediction limit formula used to
calculate the MACT floors.
b. Particulate Matter Variability for Fabric Filters. Compliance
test emissions of particulate matter from sources that are equipped
with a fabric filter may not account for long-term variability because
it is difficult to maximize emissions during the compliance test.\70\
Fabric filters control particulate matter emissions generally to the
same concentration irrespective of the particulate matter loading at
the inlet to the fabric filter. Because there are no operating
parameters that can be readily changed to increase emissions, it is
difficult to maximize emissions of particulate matter from a fabric
filter during compliance testing.\71\
---------------------------------------------------------------------------

\70\ We note that semivolatile and low volatile metal emissions,
however, can be maximized during compliance testing for sources
equipped with a fabric filter. Metals may be spiked in the hazardous
waste feed to levels that account for long-term feedrate
variability. Although the particulate matter emission concentration
would not be expected to increase during a metals compliance test
for a source equipped with a fabric filter, the semivolatile and low
volatile metals emissions concentrations would increase. This is
because the concentration of metals in the emitted particulate
matter would increase.
\71\ We note that this situation is unique for fabric filters.
Sources equipped with other control devices--electrostatic
precipitators, ionizing wet scrubbers, and wet scrubbers--can
readily change the device's operating conditions (e.g., power input
to an electrostatic precipitator; pressure drop across a wet
scrubber) during compliance testing to ``detune'' collection
efficiency and increase emissions. In addition, these other control
devices provide ``percent reduction'' control of pollutants whereby
as inlet loading increases, emission concentrations also increase.
Thus, increasing the inlet loading (e.g., by spiking the ash
feedrate to an incinerator) even without detuning the control device
would also increase emissions of particulate matter for devices
other than a fabric filter.
---------------------------------------------------------------------------

To address long-term variability in particulate matter emissions
for fabric filters we developed a universal variability factor (UVF).
The UVF represents the standard deviation of the pooled runs from
multiple compliance tests for a source, and is imputed as a function of
the source's emission concentration. We use the UVF to account for both
long-term and run-to-run variability to calculate the floor using the
procedures discussed above in lieu of the pooled variability term for
the most-recent test condition run variability.
To develop the data base to calculate the UVF, we considered each
best performing source that is equipped with a fabric filter and for
which we have two or more compliance tests for particulate matter. We
considered all compliance test particulate matter emissions data for
these sources, including those test conditions we previously labeled as
``IB'' (representing in-between), indicating that emissions levels are
lower than for another test condition of the compliance test campaign.
We include historical test campaign data where available for
incinerators, cement kilns, and lightweight aggregate kilns.
Considering historical compliance test data and compliance test data
labeled IB is appropriate because any differences in emission levels
(over time or among compliance test results for a test campaign) should
be indicative of emissions variability given that fabric filters
generally produce constant emission concentrations and are difficult to
detune to increase emissions for compliance testing. Finally, we
combined test conditions for multiple on-site sources where both the
combustor and fabric filter have similar design and operating
characteristics. Combining the test conditions for such sources as if
they represent emissions from a single source better accounts for
emissions variability.
To calculate the UVF, we calculated the pooled standard deviation
of the runs for each source for which we have data for two or more
compliance tests and plotted this standard deviation versus particulate
matter emission concentration for all such sources. It is reasonable to
aggregate the data for sources across all source categories given that
there is no reason to believe that the standard deviation/emissions
relationship would vary from source category to source category. We
then identified the best-fit curve for the data. The best fit curve is
a power function that achieved a R2 of 0.83, indicating a
good power function correlation between standard deviation and emission
concentration.\72\
---------------------------------------------------------------------------

\72\ The procedure we use to identify the universal variability
factor for particulate matter emissions for sources equipped with
fabric filters is discussed in detail in USEPA, ``Draft Technical
Support Document for HWC MACT Replacement Standards, Volume III:
Selection of MACT Standards,'' March 2004, Chapter 5.3. Please note
that we consider alternative approaches to identify the universal
variability factor as discussed in the technical support document,
and request comment on those alternatives.
---------------------------------------------------------------------------

We use the best-fit curve to impute a standard deviation for each
best performing source (that is equipped with a fabric filter) as a
function of the source's particulate matter emissions. We use the
source's average compliance test emissions (i.e., including historical
compliance test emissions that we label in the data base as ``WC'' and
``IB'') to represent average emissions.

F. Why Did EPA Default to the Interim Standards When Establishing Floors?

When we calculate floor levels for several standards for the Phase
I sources, the floor levels would be higher than the currently
applicable interim standards at Sec. Sec. 63.1203, 63.1204, and
63.1205. As explained earlier, we conclude that today's proposed floor
levels can be no higher than the interim standards because all sources,
not just the best performing sources, are achieving the interim
standards. The most recent emissions data in our data base are from
compliance testing in 2001 and do not represent emissions tests from
sources used to demonstrate compliance with the interim standards, thus
the data we used to calculate the proposed floor levels generally does
not reflect the control upgrades necessary for compliance with the
interim standards. The fact that we are ``capping'' the floor at the
interim standard level does not mean our proposed methodology is less
conservative than the methodology used in the 1999 rule. Our calculated
floor levels can be higher than the interim standards for several
reasons. As a result of our data collection effort, we have compiled
more emissions information from some source categories that result in
higher calculated floor levels (e.g., dioxin/furans for lightweight
aggregate

[[Page 21234]]

kilns). Some of the instances where we ``capped'' the floor at the
interim standard level occurred when the interim standard was a beyond-
the-floor standard promulgated in 1999 (e.g., semivolatile metals for
lightweight aggregate kilns). Finally, some standards are ``capped''
because we used different types of data to calculate the proposed
floors (e.g., the 1999 rule generally considered normal mercury data to
establish the mercury floor for incinerators, whereas today's proposed
approach used compliance test data to calculate the mercury floor).

G. What Other Options Did EPA Consider?

We considered five other alternative approaches to establish the
full suite of floor levels for each source category. The first two
alternative options use different combinations of the three main
methodology options to determine the proposed floors. Note that we also
conducted a complete economics and benefits analysis for these first
two alternative options. See USEPA ``Draft Technical Support Document
for HWC MACT Replacement Standards, Volume V: Emission Estimates and
Engineering Costs,'' March, 2004 for more information. The third option
identifies best performing sources by considering emissions of metals
and particulate matter simultaneously, instead of pollutant by
pollutant. The fourth option is an approach recommended by the
Environmental Treatment Council. Finally, the fifth option identifies
best performing sources as those sources with the best back-end control
efficiencies, as measured by their associated system removal
efficiencies. After review of comments we may use one or more of these
approaches in toto or part to establish final standards. We explain
below how these approaches work and the rationale for considering them.
1. What Is Alternative Option 1, and What Is the Rationale?
Under alternative option 1, we do not use the SRE/Feed methodology
to calculate any floors. We use the emissions-based approach to
establish all the floors, other than the exceptions that are explained
below. We express emission standards for energy recovery units in units
of hazardous waste thermal emissions when appropriate. All other
emission standards under this approach are expressed as stack gas
emission concentrations. The two exceptions under this option uses the
technology-based approach for the particulate matter standard (for all
source categories) and the total chlorine standard for hydrochloric
acid production furnaces, as was done for today's proposed standards.
We evaluated this option because it is simpler and more
straightforward to use than the SRE/Feed Approach. The best performing
sources simply are those with the lowest emissions in our data base,
irrespective of the level of feed control or back-end control a source
achieves. The advantages of using the air pollution control technology
approach and expressing emission standards using the hazardous waste
thermal emissions format for energy recovery units are retained.
Although we have doubts that standards based on these limits are
achievable even by the best performing sources (as noted earlier) and
that this approach could be based on unrepresentatively low hazardous
waste feedrates, we invite comment as to whether this approach is
appropriate. We present the results of using alternative option 1 to
identify floor levels for existing sources in Table 3 below. See U.S.
EPA ``Draft Technical Support Document for HWC MACT Replacement
Standards, Volume III: Selection of MACT Standards,'' March 2004,
Chapters 16, 17, and 18 for documentation of the floor levels.

Table 3.--Floor Levels for Existing Sources Under Alternative Option 1
--------------------------------------------------------------------------------------------------------------------------------------------------------
Hydrochloric acid
Incinerators Cement kilns Lightweight Solid fuel-fired Liquid fuel-fired production
aggregate kilns boilers \1\ boilers \1\ furnaces \1\
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dioxin/Furans (ng TEQ/dscm)..... 0.28 for dry APCD 0.20 or 0.40 + 0.20 or 400[deg]F CO or THC standard 3.0 or 400[deg]F CO or THC standard
and WHB 400[deg]F at APCD at kiln as a surrogate. at APCD inlet for as a surrogate.
sources,\6\ 0.20 inlet.\7\ outlet.\7\ dry APCD sources;
or 0.40 + CO or THC
400[deg]F at APCD standard as
inlet for surrogate for
others.\7\ others.
Mercury......................... 130 [mu]g/dscm \7\ 31 [mu]g/dscm \2\. 19 [mu]g/dscm \2\. 10 [mu]g/dscm..... 3.7E-6 lb/MMBtu 2, Total chlorine
5. standard as
surrogate.
Particulate Matter.............. 0.015 gr/dscf \7\. 0.028 gr/dscf..... 0.025 gr/dscf \7\. 0.063 gr/dscf..... 0.032 gr/dscf..... Total chlorine
standard as
surrogate.
Semivolatile Metals (lead 19 [mu]g/dscm..... 1.3E-4 lb/MMBtu 3.1E-4 lb/MMBtu 170 [mu]g/dscm.... 1.1E-5 lb/MMBtu 2, Total chlorine
+cadmium). \5\. \5\ and 250 [mu]g/ 5. standards as
dscm.\3\ surrogate.
Low Volatile Metals (arsenic + 14 [mu]g/dscm..... 1.1E-5 lbs/MMBtu 9.5E-5 lb/MMBtu 210 [mu]g/dscm.... 7.7E-5 lb/MMBtu 4, Total chlorine
beryllium + chromium). \5\. \5\ and 100 [mu]g/ 5. standard as
dscm.\3\ surrogate.
Total Chlorine (hydrogen 0.93 ppmv......... 41 ppmv........... 600 ppmv \7\...... 440 ppmv.......... 5.7E-3 lb/MMBtu 14 ppmv or
chloride + chlorine gas). \5\. 99.9927% system
removal
efficiency.
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes:
\1\ Particulate matter, semivolatile metal, low volatile metal, and total chlorine standards apply to major sources only for solid fuel-fired boilers,
liquid fuel-fired boilers, and hydrochloric acid production furnaces.
\2\ Standard is based on normal emissions data.
\3\ Sources must comply with both the thermal emissions and emission concentration standards.
\4\ Low volatile metal standard for liquid fuel-fired boilers is for chromium only. Arsenic and beryllium are not included in the low volatile metal
total for liquid fuel-fired boilers.
\5\ Standards are expressed as mass of pollutant contributed by hazardous waste per million Btu contributed by the hazardous waste.
\6\ APCD denotes ``air pollution control device,'' WHB denotes ``waste heat boiler.''

[[Page 21235]]

\7\ Floor level represents the ``capped interim standard level,'' which means the floor level determined by the associated methodology was less
stringent than the interim standard level.

2. What Is Alternative Option 2, and What Is the Rationale?
Under alternative option 2, we use the emissions-based approach to
establish all floors and there are no exceptions. All floor levels are
expressed in units of stack gas concentrations (we do not express any
floors for energy recovery units in terms of thermal emissions). The
best performing sources for all floors are those with the lowest
emissions, on a stack gas concentration basis.
We are not proposing this alternative option because it has the
disadvantages that the more complicated provisions of Option 1 (and to
some extent Option 2) address: (1) By not using the SRE/Feed Approach
for metals and total chlorine, it does not ensure that sources could
use either feedrate control or back-end control to achieve the floor;
(2) the approach may be inappropriately biased against sources that
burn hazardous waste fuel at high firing rates because it does not
express the standards in units of hazardous waste thermal emissions;
(3) it inappropriately considers feed control for particulate matter
and for hydrochloric acid production furnaces by not using the Air
Pollution Control Device Approach for those floors; and (4) it may not
appropriately estimate the performance of the average of the 12 percent
best performing sources given that those best performers may have low
emissions in part because their raw material and/or fossil fuels
contained low levels of HAP during the emissions test (and because we
do not believe feed control of HAP in raw material and fossil fuel
should be assessed as a MACT floor control because it could result in
floor levels that are not replicable by the best performing sources,
nor duplicable by other sources).
We invite comment as to whether this alternative approach is
appropriate, noting the doubts we have voiced above. We present the
results of using this alternative option 2 to identify floor levels for
existing sources in Table 4 below. See USEPA ``Draft Technical Support
Document for HWC MACT Replacement Standards, Volume III: Selection of
MACT Standards,'' March 2004, Chapter 16, for more information.

Table 4.--Floor Levels for Existing Sources Under Alternative Option 2
--------------------------------------------------------------------------------------------------------------------------------------------------------
Hydrochloric acid
Incinerators Cement kilns Lightweight Solid fuel-fired Liquid fuel-fired production
aggregate kilns boilers 1 boilers 1 furnaces 1
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dioxin/Furans (ng TEQ/dscm)..... 0.28 for dry APCD 0.20 or 0.40 + 0.20 or 400[deg]F CO or THC standard 3.0 or 400[deg]F CO or THC standard
and WHB sources; 400[deg]F at APCD at kiln as a surrogate. at APCD inlet for as a surrogate.
5 0.20 or 0.40 + inlet.6 outlet.\6\ dry APCD sources;
400[deg]F at APCD CO or THC
inlet for standard as
others.6 surrogate for
others.
Mercury......................... 130 [mu]g/dscm 6.. 31 [mu]g/dscm 2... 19 [mu]g/dscm 2... 10 [mu]g/dscm..... 0.47 [mu]g/dscm 2. Total chlorine
standard as
surrogate.
Particulate Matter.............. 0.0040 gr/dscf.... 0.016 gr/dscf..... 0.025 gr/dscf 6... 0.065 gr/dscf..... 0.0028 gr/dscf.... Total chlorine
standard as
surrogate.
Semivolatile Metals (lead + 19 [mu]g/dscm..... 68 [mu]g/dscm..... 130 [mu]g/dscm.... 170 [mu]g/dscm.... 8.7 [mu]g/dscm 2.. Total chlorine
cadmium). standard as
surrogate.
Low Volatile Metals (arsenic + 14 [mu]g/dscm..... 8.9 [mu]g/dscm.... 82 [mu]g/dscm..... 210 [mu]g/dscm.... 28 [mu]g/dscm 4... Total chlorine
beryllium + chromium). standards as
surrogate.
Total Chlorine (hydrogen 0.93 ppmv......... 41 ppmv........... 600 ppmv 6........ 440 ppmv.......... 2.4 ppmv.......... 2.0 ppmv.
chloride + chlorine gas).
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes:
1 Particulate matter, semivolatile metal, low volatile metal, and total chlorine standards apply to major sources only for solid fuel-fired boilers,
liquid fuel-fired boilers, and hydrochloric acid production furnaces.
2 Standard is based on normal emissions data.
3 Sources must comply with both the thermal emissions and emission concentration standards.
4 Low volatile metal standard for liquid fuel-fired boilers is for chromium only. Arsenic and beryllium are not included in the low volatile metal total
for liquid fuel-fired boilers.
5 APCD denotes ``air pollution control device'', WHB denotes ``waste heat boiler'.
6 Floor level represents the ``capped interim standard level'', which means the floor level determined by the associated methodology was less stringent
than the interim standard level.

3. What Is Alternative Option 3, and What Is the Rationale?
Under alternative option 3, we evaluated an approach to identify
the best performing sources for particulate matter, semivolatile
metals, and low volatile metals that considers how well a source is
controlling these pollutants simultaneously. Simultaneous control of
these pollutants is an appropriate consideration because these
pollutants are controlled by the same emission control device, the
particulate matter control device (e.g., a wet scrubber, electrostatic
precipitator, or fabric filter). We call this alternative approach the
Simultaneous Achievability for Particulates (SAP) Approach. See USEPA,
``Draft Technical Support Document for HWC MACT Replacement Standards,
Volume III: Selection of MACT Standards,'' March 2004, Chapters 10 and 19.

[[Page 21236]]

We evaluated semivolatile metal and low volatile metal emissions
for energy recovery sources--cement kilns, lightweight aggregate kilns,
and liquid fuel-fired boiler--under two emissions-based SAP
alternatives: hazardous waste thermal emissions, and stack gas
concentrations. The hazardous waste thermal emissions option assesses
semivolatile metal and low volatile metal thermal emissions for energy
recovery units, while assessing particulate matter using the emissions-
based stack gas concentration approach. The emissions-based stack-gas
concentration approach assesses stack gas concentrations (as opposed to
thermal emissions) for all HAP. Note that we did not evaluate
hydrochloric acid production furnaces under this SAP approach because
we propose to use the total chlorine standard as a surrogate to control
emissions of particulate matter and metals for these sources.
Under the SAP approach, we rank emissions for each pollutant across
the sources for which we have emissions data for that pollutant. For
ranking, we use the upper 99% confidence interval for the average of
the runs of the test condition for a source. For example, if we have
semivolatile metal emissions data for 15 sources, the lowest
semivolatile metal emissions level is ranked one and the highest is
ranked 15. To identify the best performing sources for all three
pollutants simultaneously, we calculate an aggregate rank score for
each source. For example, if source A has a rank of 5 for particulate
matter, a rank of 10 for semivolatile metals, a rank of 15 for low
volatile metals, the aggregate rank score for that source is 10, the
average rank across the pollutants. If we do not have emissions data
for a pollutant for a source, there is no rank score for that
pollutant, and that pollutant is not considered in the aggregate rank
score for the source.
To identify the best performing sources in the aggregate, we rank
the aggregate rank scores for the sources from lowest to highest. If we
have emissions data for all three pollutants for all sources, the 5 (or
12% if we have data for more than 30 sources) sources with the lowest
aggregate rank scores are the best performing sources. If we have
incomplete data sets for some sources for a source category, the best
performing sources for a pollutant (i.e., particulate matter,
semivolatile metals, or low volatile metals) are the sources with the
lowest aggregate rank scores and for which we have emissions data.
We present the alternative MACT floors for existing sources under
the SAP approach in Table 5 below.

Table 5.--Floor Levels for Existing Sources Under the SAP Approach
----------------------------------------------------------------------------------------------------------------
Particulate
Source category Emissions-based matter floor Semivolatile metals Low volatile metals
approach (gr/dscf) floor floor
----------------------------------------------------------------------------------------------------------------
Incinerators..................... Stack Gas Conc...... 0.0040 53 [mu]g/dscm....... 50 [mu]g/dscm.
Cement Kilns..................... Thermal Emissions... 0.027 190 lb/trillion Btu. 20 lb/trillion Btu.
Stack Gas Con....... 0.015 103 [mu]g/dscm...... 14 [mu]g/dscm.
Lightweight Aggregate Kilns...... Thermal Emissions... 0.019 300 lb/trillion Btu. 95 lb/trillion Btu.
Stack Gas Conc...... 0.019 120 [mu]g/dscm...... 89 [mu]g/dscm.
Solid Fuel-Fired Boilers......... Stack Gas Conc...... 0.090 180 [mu]g/dscm...... 230 [mu]g/dscm.
Liquid Fuel-Fired Boilers........ Thermal Emissions... 0.0039 81 lb/trillion Btu.. 180 lb/trillion
Btu.
Stack Gas Conc...... 0.0039 26 [mu]g/dscm....... 210 [mu]g/dscm.
----------------------------------------------------------------------------------------------------------------

We request comment on this alternative approach for identifying
MACT floors. If we use this approach in the final rule to identify MACT
floors, we would promulgate a beyond-the-floor standard for particulate
matter of 0.030 gr/dscf for existing solid fuel-fired boilers for the
same reasons we are proposing today a beyond-the-floor standard. See
Part Two, Section X.C for a discussion of today's proposed beyond-the-
floor particulate matter standard for solid fuel-fired boilers.
See USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume III: Selection of MACT Standards,'' March
2004, Chapters 10 and 19, for a more detailed explanation of this SAP
analysis.
4. What Is Alternative Option 4, and What Is the Rationale?
The Environmental Technology Council (ETC) recommends an approach
to calculate floor levels for metals and chlorine that uses a low
feedrate screen and addresses emissions variability differently than
the options we evaluated.\73\ We may use this approach in total or in
part to support a final rule, and therefore request comment on the approach.
---------------------------------------------------------------------------

\73\ Update on MACT Floor Evaluations Revised Data Base,
Environmental Technology Council, February 2003.
---------------------------------------------------------------------------

Under ETC's approach, test conditions are screened from further
consideration if metals or chlorine were not fed at levels that
challenge the emissions control system.\74\ Feedrates of metals and
chlorine in hazardous waste are normalized to account for size of the
combustor by converting feedrates to maximum theoretical emissions
concentrations. A low maximum theoretical emissions concentration
filter is used to screen out emissions from low feed test conditions,
where the filter is the lower 99% confidence limit of the mean of the
maximum theoretical emissions concentrations for all test conditions
for all sources within a source category.
---------------------------------------------------------------------------

\74\ This approach therefore identifies a de minimis feed
control level for each source category and does not evaluate
emissions from these de minimus feeders in the MACT analysis because
these de minimis feed control levels may not be feasible for other
sources to duplicate. The screen is performed individually by
pollutant so that if semivolatile metals were fed at rates that
challenged the emissions control system but low volatile metals were
not, only the low volatile metal emissions data for that test
condition would be screened from further analysis.
---------------------------------------------------------------------------

ETC's approach also excludes specialty units, defined as sources
that burn munitions and radiological waste (i.e., Department of Defense
and Department of Energy sources). ETC believes that these sources burn
wastes with atypical concentrations of ash and metals that may
inappropriately skew the calculation of floor levels. Under this
approach, we would either subcategorize and issue separate emission
standards for these specialty units, or omit these speciality units
from the MACT analysis and require the specialty units to comply with
the floor levels that are determined from emissions of the non-
specialty units.
After applying the low maximum theoretical emissions concentration
filter and excluding specialty units, this approach identifies the best
performing sources by ranking emissions from

[[Page 21237]]

lowest to highest.\75\ Run variability is not considered at this point.
For incinerators, cement kilns, and lightweight aggregate kilns where
we may have historical compliance test emissions from several test
campaigns for a source, test conditions from the campaign with the
lowest compliance test emissions are used to identify the best
performers.
---------------------------------------------------------------------------

\75\ This low feed screen is not applied to cement kilns and
lightweight aggregate kilns for the particulate matter standard
because ash feedrate is not considered to be a dominant factor that
influences particulate matter emissions (rather, particulate matter
emissions are more a function of the back-end control device efficiency).
---------------------------------------------------------------------------

The average of the emissions from the best performing sources are
used to calculate the floor, and an emissions variability factor is
added. For incinerators, cement kilns, and lightweight aggregate kilns
where we may have historical compliance test emissions data from
several test campaigns for a source, three approaches are considered to
select representative emissions for each best performing source: (1)
The highest compliance test emissions from any test campaign; (2) the
average of the highest compliance test emissions from all test
campaigns; and (3) the highest emissions during the most recent
compliance test campaign. By identifying the best performers based on
compliance test emissions from the test campaign with the lowest
emissions and calculating the floor using compliance test emissions
under these alternative approaches, emissions variability is addressed
in part.\76\
---------------------------------------------------------------------------

\76\ This approach for partially accounting for emissions
variability is effective only for those incinerators, cement kilns,
and lightweight aggregate kilns for which we have emissions data for
more than one test campaign.
---------------------------------------------------------------------------

Emissions variability is accounted for by adding an emissions
variability factor to the average emissions for the best performing
sources. The variability factor is a measure of the average run-to-run
variability for the test conditions for the best performing sources.
The variability factor is determined as the upper confidence limit
(calculated at the 99% confidence interval) around the mean of the runs
for each test condition for each best performer. (For sources with more
than one compliance test condition, the variability factor for each
source is first determined as the average of the variabilities
associated with each compliance test condition).\77\ The upper
confidence limits are averaged across the best performing sources, and
the average confidence limit is added to the average emissions from the
best performers to identify the floor.
---------------------------------------------------------------------------

\77\ We do not use this step in our statistical analysis because
we identify one test condition only as being representative of the
emissions for each source. Alternatively, ETC's approach includes an
option where the average of the historical compliance test
conditions is considered for Phase I sources. Under this option,
ETC's approach considers the average run-to-run variability for
those historical compliance tests.
---------------------------------------------------------------------------

We invite comment as to whether this alternative approach is
appropriate. We calculated alternative floor levels for new and
existing sources with minor adjustments.\78\ We present the results of
applying that approach in Table 6 below. See USEPA ``Draft Technical
Support Document for HWC MACT Replacement Standards, Volume III:
Selection of MACT Standards,'' March 2004, Chapters 12 and 21, for more
information on how we applied this approach to our data base.
---------------------------------------------------------------------------

\78\ Note that we modified part of ETC's suggested methodology
in some instances, which has resulted in our calculated floor levels
to differ from ETC's calculated floor levels. These modifications
are discussed in USEPA ``Draft Technical Support Document for HWC
MACT Replacement Standards, Volume III: Selection of MACT
Standards,'' March 2004, Chapter 12.

Table 6.--Floor Levels for Existing Sources Under the Modified ETC Approach
--------------------------------------------------------------------------------------------------------------------------------------------------------
Incinerators
-------------------------- Lightweight Solid fuel- Liquid fuel-
Data base Excluding Cement kilns aggregate fired fired
All speciality kilns boilers boilers
units
--------------------------------------------------------------------------------------------------------------------------------------------------------
Mercury ([mu]g/dscm)................... Avg of historical CT data 130 (308) 130 (308) 48 37 ........... ...........
\1\ \1\
Most recent CT data...... 130 (308) 130 (308) 40 31 14 4.8
\1\ \1\
Highest of historical CT 130 (308) 130 (308) 68 45 ........... ...........
data. \1\ \1\
----------------------------------------
Particulate Matter (gr/dscf)........... Avg of historical CT data 0.0043 0.0043 0.025 0.017 ........... ...........
Most recent CT data...... 0.0043 0.0043 0.025 0.017 0.11 0.0090
Highest of historical CT 0.0043 0.0043 0.030 (0.032) 0.017 ........... ...........
data. \1\
----------------------------------------
Semivolatile Metals ([mu]g/dscm)....... Avg of historical CT data 53 32 230 250 (901) \1\ ........... ...........
Most recent CT data...... 53 32 160 250 (746) \1\ 230 8.2
Highest of historical CT 53 32 300 250 (1208) \1\ ........... ...........
data.
----------------------------------------
Low Volatile Metals ([mu]g/dscm)....... Avg of historical CT data 39 46 51 110 (119) \1\ ........... ...........
Most recent CT data...... 39 36 42 110 (129) \1\ 320 52
Highest of historical CT 39 56 56 \1\ 110 (133) \1\ ........... ...........
data.
----------------------------------------
Total Chlorine (ppmv).................. Avg of historical CT data 1.4 1.8 85 600 (1655) \1\ ........... ...........
Most recent CT data...... 1.4 1.8 86 600 (1811) \1\ 410 3.2
Highest of historical CT 1.4 1.8 89 600 (1823) \1\ ........... ...........
data.
--------------------------------------------------------------------------------------------------------------------------------------------------------
Notes: ``CT'' means Compliance Test.

[[Page 21238]]

\1\ Floor would be capped by the Interim Standards. Number in parentheses represents the calculated floor level, the number preceding is the ``capped''
interim standard level.

5. What Is Alternative Option 5, and What Is the Rationale?
Alternative Option 5 would use system removal efficiency (SRE) to
identify the best performing sources for the mercury, semivolatile
metals, low volatile metals, and total chlorine floor levels. This is
similar to the approach we propose to establish the total chlorine
standard for hydrochloric acid production furnaces. See discussion in
Part Two, Section VI.A.2.b.
Floor levels would be expressed as an SRE or an emission
concentration where the emission concentration is based on the
emissions achieved by the best performing SRE sources.\79\ A source
could elect to comply with either floor. An emissions floor as an
alternative to the SRE floor is appropriate because a source may be
achieving emission levels lower than those achieved by the best
performing SRE sources even though it may not be achieving MACT floor
SRE. For example, a source may be achieving low emissions without
achieving MACT SRE by using superior feedrate control.
---------------------------------------------------------------------------

\79\ We note that an SRE option, in some form, could be added to
any of the emission-based approaches previously discussed.
---------------------------------------------------------------------------

The SRE floor is an SRE that the average of the best performing SRE
sources could be expected to achieve in 99 of 100 future tests when
operating under the conditions used to establish the SRE.\80\ The
emissions floor is a stack gas concentration, or thermal emission
concentration for source categories that burn hazardous waste fuels,
that the average of the best performing SRE sources could be expected
to achieve in 99 of 100 future tests when operating under the
conditions used to establish the SRE and emission level.
---------------------------------------------------------------------------

\80\ Note that we only considered SREs associated with emission
values designated as compliance test (CT) in the database. See USEPA
``Draft Technical Support Document for HWC MACT Replacement
Standards, Volume III: Selection of MACT Standards,'' March 2004,
Chapters 11 and 20, for more information.
---------------------------------------------------------------------------

We note that this approach is not applicable for situations where
sources in a source category do not use back-end control to control
metals or total chlorine. For example, cement kilns do not use back-end
control to control mercury or total chlorine.\81\
---------------------------------------------------------------------------

\81\ Although the alkalinity in cement kiln raw materials helps
control total chlorine emissions, we are concerned that the system
removal efficiencies achieved may not be readily reproducible.
---------------------------------------------------------------------------

This approach is also not applicable for situations where our data
base is comprised of normal emissions data. As discussed previously,
SREs calculated from normal test conditions may be unreliable because a
small error in the feedrate calculation at low feedrates can have a
substantial impact on the calculated SRE.
In situations where this SRE-based approach is not applicable, we
would use an alternative approach to identify MACT floor, such as the
Emissions approach.
Floor levels for existing sources under this approach are presented
in Table 7.
We also investigated a variation of this approach where sources
with atypically high feedrates for metals or chlorine are excluded from
the calculation of the alternative emission level. This variation may
be appropriate to ensure that sources with high feedrates do not drive
the alternative emission concentration-based floor inappropriately high
even though the source may be a best performing SRE source. Under this
variation, note that sources with high feedrates are used, however, to
identify the best performing SRE sources and MACT SRE. This is because
sources with the highest feedrates may employ the best performing back-
end control systems to meet current standards or otherwise control
emissions. As a measure of atypically high feedrates, we use the 99th
upper percentile feedrate around the mean of feedrate data in the data
set available for the analysis. To ensure that we continue to use 5
sources or 12 percent of sources to calculate the emission
concentration-based floor under this variation, we replace a best
performing SRE source that is screened out of the concentration-based
floor analysis because of high feedrates with the source with the next
best SRE.\82\
---------------------------------------------------------------------------

\82\ Since sources with atypically high feedrates may still have
low emissions, sources with hazardous waste feed control levels
above the threshold are flagged, but not immediately removed from
the data set. Sources' SREs are ranked from best to worst, initially
choosing the best ranked 5 or 12% of sources as the interim MACT
pool. The remaining sources are temporarily set aside, and the
sources comprising the interim MACT pool are ranked again from
lowest to highest emissions. Sources from the interim MACT pool that
have been flagged due to having feedrates above the upper 99th
percentile are systematically (from highest to lowest emissions)
removed from the MACT pool and replaced with sources with the next
highest ranked SREs if the emissions from the next best source
initially excluded from the interim MACT pool has lower emissions.
The sources comprising the revised interim MACT pool now become the
final MACT pool. Emissions from those sources are again used to
calculate the MACT floor, with the resulting MACT floor again
expressed as an emission standard.
---------------------------------------------------------------------------

Floor levels for existing sources under this feedrate-screened
variation are presented in Table 8.
We invite comment on these alternative floor approaches. For more
information on how the approach would work, see USEPA ``Draft Technical
Support Document for HWC MACT Replacement Standards, Volume III:
Selection of MACT Standards,'' March 2004, Chapters 13 and 22.

[[Page 21239]]

Table 7.--Floor Levels for Existing Sources Under Alternative Option 5
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
Mercury Semivolatile metals Low volatile metals Total chlorine
-------------------------------------------------------------------------------------------------------------------------------------
Emissions Emission Emission Emission
Source category -------------------- concentration concentration concentration
SRE \1\ SRE \1\ -------------------- SRE \1\ -------------------- SRE \1\ ---------------------
Stack Thermal Stack Thermal Stack Thermal Stack gas Thermal
gas \2\ \3\ gas \2\ \3\ gas \2\ \3\ \2\ \3\
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
Incinerators.................. 27 20,000 n/a \8\ 99.89 74 n/a \8\ 99.969 33 n/a \8\ 99.990 3.1 n/a \8\
\9\
-----------
Cement Kilns.................. n/a 4, 5 99.966 71 140 99.989 11 22 n/a 4, 5
-----------
Lightweight Aggregate Kilns... n/a 4, 6 99.78 330 310 99.89 100 95 n/a 4, 6
-----------
Solid Fuel-Fired Boilers...... 11 ........ n/a \8\ 99.78 180 n/a \8\ 97.9 230 n/a \8\ n/a 4, 5
-----------
Liquid Fuel-Fired Boilers..... n/a \4\
n/a \4\ 90.4 \7\ 27 \7\ 45 \7\ 99.70 25 55
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ SRE is system removal efficiency expressed as a percent.
\2\ Stack gas concentration is expressed in [mu]g/dscm for all except total chlorine, which is expressed as ppmv.
\3\ Thermal emission is expressed in lb/trillion Btu, except total chlorine which is expressed in lb/billion Btu.
\4\ Unable to determine SRE due to normal feedrate data.
\5\ No SRE due to no reliable back-end control.
\6\ Only one source has back-end control.
\7\ LVM Standards for liquid fuel-fired boilers are for Chromium, only.
\8\ Thermal emissions not appropriate for source categories with sources that do not burn hazardous waste fuels.
\9\ We believe this methodology yields inappropriate MACT mercury floors for incinerators because we have only 11 compliance test conditions, and the best
performers spiked uncharacteristically high levels of mercury during their compliance test.

Table 8.--Floor Levels for Existing Sources Under Alternative Option 5 With High Feedrate Screen
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Mercury Semivolatile metals Low volatile metals Total chlorine
-------------------------------------------------------------------------------------------------------------------------------------
Emissions Emission Emission Emission
Source category -------------------- concentration concentration concentration
SRE \1\ SRE \1\ -------------------- SRE \1\ -------------------- SRE \1\ -------------------
Stack Thermal Stack Thermal Stack Thermal Stack gas Thermal
gas \2\ \3\ gas \2\ \3\ gas \2\ \3\ \2\ \3\
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Incinerators.................. 27 7,500 n/a \8\ 99.89 64 n/a \8\ 99.969 29 n/a \8\ 99.990 1.3 n/a \8\
\9\
-----------
Cement Kilns.................. n/a 4, 5 99.966 65 130 99.989 11 18 n/a 4, 5
-----------
Lightweight Aggregate Kilns... n/a 4, 6 99.78 330 310 99.89 100 95 n/a 4, 6
-----------
Solid Fuel-Fired Boilers...... 11 ........ n/a \8\ 99.78 180 n/a \8\ 97.9 230 n/a \8\ n/a 4, 5
-----------
Liquid Fuel-Fired Boilers..... n/a \4\
n/a \4\ 90.4 \7\ 27 \7\ 110 \7\ 99.70 23 55
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ SRE is system removal efficiency expressed as a percent.
\2\ Stack gas concentration is expressed in [mu]g/dscm for all except total chlorine, which is expressed as ppmv.
\3\ Thermal emission is expressed in lb/trillion Btu, except total chlorine which is expressed in lb/billion Btu.
\4\ Unable to determine SRE due to normal feedrate data.
\5\ No SRE due to no reliable back-end control.
\6\ Only one source has back-end control.
\7\ LVM Standards for liquid fuel-fired boilers are for Chromium, only.
\8\ Thermal emissions not appropriate for source categories with sources that do not burn hazardous waste fuels.
\9\ We believe this methodology yields inappropriate MACT mercury floors for incinerators because we have
only 11 compliance test conditions, and the best performers spiked
uncharacteristically high levels of mercury during the their compliance test.

[[Page 21240]]

VII. How Did EPA Determine the Proposed Emission Standards for
Hazardous Waste Burning Incinerators?

The proposed standards for existing and new incinerators that burn
hazardous waste are summarized in the table below. See proposed Sec.
63.1219.

Proposed Standards for Existing and New Incinerators
------------------------------------------------------------------------
Emission standard \1\
Hazardous air pollutant or -------------------------------------------
surrogate Existing sources New sources
------------------------------------------------------------------------
Dioxin and furan--sources 0.28 ng TEQ/dscm.... 0.11 ng TEQ/dscm.
equipped with waste heat
boilers or dry air
pollution control system
\2\.
Dioxin and furan--sources 0.2 ng TEQ/dscm; or 0.20 ng TEQ/dscm.
not equipped with waste 0.40 ng TEQ/dscm
heat boilers or dry air and temperature at
pollution control system inlet to the
\2\. initial particulate
matter control
device <=400[deg]F.
Mercury..................... 130 [mu]g/dscm...... 8.0 [mu]g/dscm.
Particulate matter.......... 34 mg/dscm (0.015 gr/ 1.6 mg/dscm (0.00070
dscf). gr/dscf).
Semivolatile metals......... 59 [mu]g/dscm....... 6.5 [mu]g/dscm.
Low volatile metals......... 84 [mu]g/dscm....... 8.9 [mu]g/dscm.
Hydrogen chloride and 1.5 ppmv or the 0.18 ppmv or the
chlorine gas \3\. alternative alternative
emission limits emission limits
under Sec. under Sec.
63.1215. 63.1215.
Hydrocarbons \4,5\.......... 10 ppmv (or 100 ppmv 10 ppmv (or 100 ppmv
carbon monoxide). carbon monoxide).
Destruction and removal For existing and new sources, 99.99% for
efficiency. each principal organic hazardous
constituent (POHC). For sources burning
hazardous wastes F020, F021, F022, F023,
F026, or F027, however, 99.9999% for each
POHC.
------------------------------------------------------------------------
\1\ All emission standards are corrected to 7% oxygen dry basis.
\2\ A wet air pollution system followed by a dry air pollution control
system is not considered to be a dry air pollution control system for
purposes of this standard. A dry air pollution systems followed a wet
air pollution control system is considered to be a dry air pollution
control system for purposes of this standard.
\3\ Combined standard, reported as a chloride (Cl(-)) equivalent.
\4\ Sources that elect to comply with the carbon monoxide standard must
demonstrate compliance with the hydrocarbon standard during the
comprehensive performance test.
\5\ Hourly rolling average. Hydrocarbons reported as propane.

A. What Are the Proposed Standards for Dioxin and Furan?

The proposed standards for dioxin/furan for sources equipped with
dry air pollution control devices and/or waste heat boilers are 0.28 ng
TEQ/dscm for existing sources and 0.11 ng TEQ/dscm for new sources. For
incinerators using either wet air pollution control or no air pollution
control devices, the proposed standards for dioxin/furan are 0.20 ng
TEQ/dscm or 0.40 ng TEQ/dscm while limiting the temperature at the
inlet to the particulate matter control device to less than 400 [deg]F
for existing sources and 0.20 ng TEQ/dscm for new sources.
1. What Is the Rationale for the MACT Floor for Existing Sources?
Dioxin and furan emissions for existing incinerators are currently
limited by Sec. 63.1203(a)(1) to 0.20 ng TEQ/dscm; or 0.40 ng TEQ/dscm
provided that the combustion gas temperature at the inlet to the
initial particulate matter control device is limited to 400 [deg]F or
less. (For purposes of compliance, operation of a wet air pollution
control system is presumed to meet the 400 [deg]F or lower
requirement.) This standard was promulgated in the Interim Standards
Rule (See 67 FR at 6796, February 13, 2002).
Since promulgation of the September 1999 final rule, we have
obtained additional dioxin/furan emissions data. We now have dioxin/
furan emissions data for over 55 sources. The emissions in our data
base range from less than 0.001 to 34 ng TEQ/dscm.
As discussed in Part Two, Section II, we assessed whether
incinerators equipped with dry air pollution control devices and/or
waste heat boilers have statistically different emissions than sources
with either wet air pollution control or no air pollution control
equipment.\83\ Our statistical analysis indicates dioxin/furan
emissions between these types of incinerators are significantly
different. (As we explained there, these differences relate to
differences in dioxin/furan formation mechanisms, not pollution control
device efficiency.) Therefore, we believe subcategorization is
warranted for this emission standard and we are proposing separate
floor levels.
---------------------------------------------------------------------------

\83\ A source with a wet air pollution system followed by a dry
air pollution control system is not considered to be a dry air
pollution control system for purposes of this standard, while a
source with a dry air pollution system followed a wet air pollution
control system is considered to be a dry air pollution control
system. In addition, we note that a spray dryer is not considered to
be a wet air pollution control system for purposes of
subcategorization.
---------------------------------------------------------------------------

To identify the floor level for incinerators equipped with dry air
pollution control equipment and/or waste heat boilers, we evaluated the
compliance test emissions data associated with the most recent test
campaign using the Emissions Approach described in Part Two, Section
VI. The calculated floor is 0.28 ng TEQ/dscm, which considers emissions
variability. This is an emission level that the average of the best
performing sources could be expected to achieve in 99 of 100 future
tests when operating under conditions identical to the compliance test
conditions during which the emissions data were obtained. The
calculated floor level of 0.28 ng TEQ/dscm is based on five best
performing sources that achieved this floor level either by the use of
temperature control at the inlet to dry air pollution control device
and good combustion or by the use of activated carbon injection. The
single best performer is equipped with a dry air pollution control
system and a waste heat boiler, and uses activated carbon injection,
good combustion, and temperature control to control dioxin/furan
emissions. The remaining four

[[Page 21241]]

best performers are equipped with dry air pollution systems but do not
have waste heat recovery boilers. Two of these sources use activated
carbon, good combustion, and temperature control to control dioxin/
furan emissions.\84\ The other two without waste heat recovery boilers
use a combination of good combustion and temperature control to control
emissions.
---------------------------------------------------------------------------

\84\ One source uses an activated carbon injection system, and
the other uses a carbon bed.
---------------------------------------------------------------------------

We then judged the relative stringency of the calculated floor
level to the interim standard to determine if the proposed floor level
needed to be ``capped'' by the current interim standard to ensure the
proposed floor level is not less stringent than an existing federal
emission standard. A comparison of the calculated floor level of 0.28
ng TEQ/dscm to the interim standard--0.20 ng TEQ/dscm or 0.40 ng TEQ/
dscm provided that the combustion gas temperature at the inlet to the
initial particulate matter control device is limited to less than 400
[deg]F--indicates that a floor level of 0.28 ng TEQ/dscm is more
stringent than the current interim standard. This judgment is based on
our belief that the majority of these incinerators are currently
complying with the 0.40 ng TEQ/dscm and temperature limitation portion
of the interim standard.\85\ We estimate that this emission level is
being achieved by 71% of sources and would reduce dioxin/furan
emissions by 0.28 grams per year.
---------------------------------------------------------------------------

\85\ We request comment, however, on whether this judgment is
correct. If an incinerator is operated with a dry air pollution
control device inlet temperature greater than 400 [deg]F, then it
may be appropriate to instead require sources to comply with the
more stringent of the two standards, that is, 0.20 ng TEQ/dscm.
---------------------------------------------------------------------------

We also considered whether to further subcategorize based on
whether the incinerator is equipped with a waste heat recovery boiler
or dry air pollution control device. Our analysis determined that the
dioxin/furan emissions from incinerators with waste heat recovery
boilers are not statistically different from those equipped with dry
air pollution control systems. We propose, therefore, that further
subcategorization is not necessary given that incinerators using either
waste heat recovery boilers or dry air pollution control systems can
readily achieve the calculated floor level using control technologies
demonstrated by the best performing sources.
For sources with either wet air pollution control systems or no air
pollution control equipment, but are not equipped with a heat recovery
boiler, we contemplated identifying an emission limit but instead rely
on surrogates for control of organic HAP, namely good combustion
practices, to be demonstrated by complying with the carbon monoxide or
hydrocarbon emissions standard and compliance with the destruction and
removal efficiency standard.\86\ We believe that it would be
inappropriate to establish a numerical dioxin/furan floor level for
sources with wet or no air pollution control systems because the floor
emission level would not be replicable by the best performing sources
nor duplicable by other sources. Dioxin/furan formation mechanisms are
complex. Sources with wet or no air pollution control devices may have
difficulty complying with a numerical dioxin/furan limit that is based
on the lowest emitting dioxin/furan sources within this subcategory
because there is not a demonstrated floor control technology that these
sources can use to ``dial in'' to achieve a given emission level.
Moreover, dioxin/furan emissions could result from operation under poor
combustion conditions and formation on particulate matter surfaces in
duct work, on heat recovery boiler tubes, and on particulates entrained
in the combustion gas stream. As a result, we would instead identify
floor control for these sources to be operating under good combustion
practices by complying with the destruction and removal efficiency and
carbon monoxide/hydrocarbon standards.
---------------------------------------------------------------------------

\86\ Use of ``good combustion practices'' does not necessarily
preclude significant dioxin/furan formation. Our data base suggests,
however, that incinerators using wet air pollution control systems
achieve dioxin/furan emissions less than 0.40 ng TEQ/dscm. See
USEPA, ``Draft Technical Support Document for HWC MACT Replacement
Standards, Volume III: Selection of MACT Standards,'' March 2004,
Chapter 2.
---------------------------------------------------------------------------

Though MACT floor for these units is operating under good
combustion practices, there is a regulatory limit which is relevant in
identifying the floor level. Hazardous waste incinerators are complying
with an interim standard for dioxin/furan--an emission limit of 0.20 ng
TEQ/dscm or, alternatively, 0.40 ng TEQ/dscm provided that the
combustion gas temperature at the inlet to the initial particulate
matter control device is limited to 400 [deg]F or less--that fixes a
level of performance for the source category. Given that all sources
are meeting this interim standard and that the interim standard is
judged as more stringent than a MACT floor of ``good combustion
practices,'' the dioxin/furan floor level can be no less stringent than
the current regulatory limit.\87\ Therefore, the proposed floor level
for incinerators with either wet air pollution control systems or no
air pollution control equipment that are not equipped with a heat
recovery boiler is either 0.20 ng TEQ/dscm or 0.40 ng TEQ/dscm provided
that the combustion gas temperature at the inlet to the initial
particulate matter control device is limited to 400 [deg]F or less.
This emission level is currently being achieved by all sources because
the interim standard is an enforceable standard currently in effect.
---------------------------------------------------------------------------

\87\ Even though all sources have recently demonstrated
compliance with the interim standards, the dioxin/furan data in our
data base preceded the compliance demonstration.
---------------------------------------------------------------------------

2. EPA's Evaluation of Beyond-the-Floor Standards for Existing Sources
We evaluated beyond-the-floor standards based on the use of control
technology which removes dioxin/furan, namely use of an activated
carbon injection system or a carbon bed system as beyond-the-floor
control for further reduction of dioxin/furan emissions. Activated
carbon is currently used at three incinerators to control dioxin/furan.
We evaluated a beyond-the-floor level of 0.10 ng TEQ/dscm for all
incinerators, which represents a 65-75% reduction in dioxin/furan
emissions from the floor level. We selected this level because it
represents a level that is considered routinely achievable with
activated carbon.\88\
---------------------------------------------------------------------------

\88\ USEPA, ``Draft Technical Support Document for HWC MACT
Replacement Standards, Volume V: Emissions Estimates and Engineering
Costs,'' March 2004, Chapter 4.3.
---------------------------------------------------------------------------

For incinerators equipped with dry air pollution control equipment
and/or waste heat boilers, the national incremental annualized
compliance cost for these sources to meet the beyond-the-floor level
rather than comply with the floor controls would be approximately $2.2
million and would provide an incremental reduction in dioxin/furan
emissions beyond the floor level controls of 0.5 grams TEQ per year.
Nonair quality health and environmental impacts and energy effects were
evaluated to estimate the impacts between activated carbon injection
and carbon beds and controls likely to be used to meet the floor level.
We estimate that this beyond-the-floor option would increase the amount
of hazardous waste generated by 1,500 tons per year in addition to
using an additional 3 million kW-hours per year beyond the requirements
to achieve the floor level. The costs associated with these hazardous
waste treatment/disposal and energy impacts are accounted for in the
national annualized compliance cost estimates. Therefore, based on
these factors and costs of approximately $4.4 million per

[[Page 21242]]

additional gram of dioxin/furan removed, we are not proposing a beyond-
the-floor standard based on activated carbon injection and carbon bed
systems.
For sources with either wet air pollution control systems or no air
pollution control equipment that are not equipped with a heat recovery
boiler, the national incremental annualized compliance cost for these
sources to meet the beyond-the-floor level would be approximately $3.9
million and would provide an incremental reduction in dioxin/furan
emissions beyond the MACT floor controls of 0.35 grams TEQ per year.
Nonair quality health and environmental impacts and energy effects were
also evaluated. We estimate that this beyond-the-floor option would
increase the amount of hazardous waste generated by 700 tons per year.
The option would also require sources to use an additional 2 million
kW-hours per year and 70 million gallons of water beyond the
requirements to achieve the floor level. Therefore, based on these
factors and costs of approximately $11 million per additional gram of
dioxin/furan removed, we are not proposing a beyond-the-floor standard
based on activated carbon injection and carbon bed systems.
3. What Is the Rationale for the MACT Floor for New Sources?
Dioxin and furan emissions for new incinerators are currently
limited by Sec. 63.1203(b)(1) to 0.20 ng TEQ/dscm. This standard was
promulgated in the Interim Standards Rule (See 67 FR at 6796, February
13, 2002).
For incinerators equipped with dry air pollution control equipment
and/or waste heat boilers, the calculated floor level is 0.11 ng TEQ/
dscm, which considers variability. This is an emission level that the
single best performing source identified using the Emissions Approach
could be expected to achieve in 99 out of 100 future tests when
operating under conditions identical to the compliance test conditions
during which the emissions data were obtained.
For sources with either wet air pollution control systems or no air
pollution control equipment that are not equipped with a heat recovery
boiler, as previously discussed for existing sources, we believe that
it would be inappropriate to establish numerical dioxin/furan emission
for these sources. We would instead identify floor control for these
sources to be operating under good combustion practices by complying
with the destruction and removal efficiency and carbon monoxide/
hydrocarbon standards.
Though MACT floor for these units is operating under good
combustion practices, there is a regulatory limit which is relevant in
identifying the floor level. New hazardous waste incinerators are
subject to an interim emission standard for dioxin/furan of 0.20 ng
TEQ/dscm. Given that the interim standard is judged more stringent than
a MACT floor of ``good combustion practices,'' the dioxin/furan floor
level can be no less stringent than the current regulatory limit.
Therefore, the proposed floor level for incinerators with either wet
air pollution control systems or no air pollution control equipment
that are not equipped with a heat recovery boiler is 0.20 ng TEQ/dscm.
Therefore, we are proposing the current interim standard of 0.20 ng
TEQ/dscm as the floor level for new sources.
4. EPA's Evaluation of Beyond-the-Floor Standards for New Sources
We evaluated beyond-the-floor standards based on the use of a
carbon bed system to achieve additional removal of dioxin/furan. Given
the relatively low dioxin/furan levels at the floor, we made a
conservative assumption that the use of a carbon bed will provide an
additional 50% dioxin/furan control. We applied this removal efficiency
to the dioxin/furan floor levels to identify the beyond-the-floor
levels.
For a new incinerator with average gas flowrate equipped with dry
air pollution control equipment and/or a waste heat boiler, the
national incremental annualized compliance cost to meet the beyond-the-
floor level of 0.06 ng TEQ/dscm rather than comply with the floor
controls would be approximately $0.22 million and would provide an
incremental reduction in dioxin/furan emissions beyond the floor level
controls of 0.013 grams TEQ per year. Nonair quality health and
environmental impacts and energy effects were evaluated. Therefore,
based on these factors and costs of approximately $17 million per
additional gram of dioxin/furan removed, we are not proposing a beyond-
the-floor standard based on activated carbon bed systems.
For a source with either a wet air pollution control system or no
air pollution control equipment that is not equipped with a heat
recovery boiler, the national incremental annualized compliance cost
for a new incinerator with an average gas flowrate to meet a beyond-
the-floor level of 0.10 ng TEQ/dscm would be approximately $0.22
million and would provide an incremental reduction in dioxin/furan
emissions beyond the MACT floor controls of 0.024 grams TEQ per year.
Considering the nonair quality health and environmental impacts and
energy effects in addition to costs of approximately $9.3 million per
additional gram of dioxin/furan removed, we are not proposing a beyond-
the-floor standard based on a carbon bed system.

B. What Are the Proposed Standards for Mercury?

We are proposing to establish standards for existing and new
incinerators that limit emissions of mercury to 130 [mu]g/dscm and 8
[mu]g/dscm, respectively.
1. What Is the Rationale for the MACT Floor for Existing Sources?
Mercury emissions for existing incinerators are currently limited
to 130 [mu]g/dscm by Sec. 63.1203(a)(2). This standard was promulgated
in the Interim Standards Rule (See 67 FR at 6796).
We have both normal and compliance test emissions data for over 50
sources. For several sources, we have emissions data from more than one
test campaign. The mercury stack emissions in our data base range from
less than 1 to 35,000 [mu]g/dscm, which are expressed as mass of
mercury per unit volume of stack gas.
To identify the floor level, we evaluated the compliance test
emissions data associated with the most recent test campaign using the
SRE/Feed Approach. The calculated floor is 610 [mu]g/dscm, which
considers emissions variability. Even though all sources have recently
demonstrated compliance with the interim standard of 130 [mu]g/dscm,
all the mercury emissions data in our data base precede initial
compliance with these interim standards. As a result, the calculated
floor level of 610 [mu]g/dscm is less stringent than the interim
standard, which is a regulatory limit relevant in identifying the floor
level (so as to avoid any backsliding from a current level of
performance achieved by all incinerators, and hence, the level of
minimal stringency at which EPA could calculate the MACT floor).
Therefore, we are proposing the floor level as the current emission
standard of 130 [mu]g/dscm. This emission level is currently being
achieved by all sources.
We invite comment on an alternative approach to identify the floor
level using available normal emissions data instead of the compliance
test data. For reasons we discussed above in Part Two, our floor-
setting methodology favors compliance test data over normal emissions
data. However, there are available more mercury emissions data

[[Page 21243]]

characterized as normal--over 40 test conditions--than the eleven
compliance test results. Given that the data base includes considerably
more normal emissions than compliance test data, we invite comment on
whether the floor analysis should be based on the normal emissions data
instead of the compliance test data. The floor level considering the
normal data using the Emissions Approach is 7.8 [mu]g/dscm, which
considers emissions variability. If we were to adopt such an approach,
we would require sources to comply with the limit on an annual basis
because the floor analysis is based on normal emissions data. Under
this approach, compliance would not be based on the use of a total
mercury continuous emissions monitoring system because these monitors
have not been adequately demonstrated as a reliable compliance
assurance tool at all types of incinerator sources. Instead, a source
would maintain compliance with the mercury standard by establishing and
complying with short-term limits on operating parameters for pollution
control equipment and annual limits on maximum total mercury feedrate
in all feedstreams.
2. EPA's Evaluation of Beyond-the-Floor Standards for Existing Sources
We identified two potential beyond-the-floor techniques for control
of mercury: (1) Activated carbon injection; and (2) control of mercury
in the hazardous waste feed.
Use of Activated Carbon Injection. We evaluated activated carbon
injection as beyond-the-floor control for further reduction of mercury
emissions. Activated carbon injection is currently being used at three
incinerators and has been demonstrated for controlling mercury and has
achieved efficiencies ranging from 80% to greater than 90% depending on
various factors such as injection rate, mercury speciation in the flue
gas, flue gas temperature, and carbon type. Given the limited
experience at hazardous waste combustion systems, we made a
conservative assumption that the use of activated carbon will provide
70% mercury control. We evaluated a beyond-the-floor level of 39 [mu]g/
dscm.
The national incremental annualized compliance cost for
incinerators to meet this beyond-the-floor level rather than comply
with the floor controls would be approximately $7.1 million and would
provide an incremental reduction in mercury emissions beyond the MACT
floor controls of 0.39 tons per year. Nonair quality health and
environmental impacts and energy effects were evaluated to estimate the
impacts between activated carbon injection and controls likely to be
used to meet the floor level. We estimate that this beyond-the-floor
option would increase the amount of hazardous waste generated by 1,800
tons per year and would require sources to use an additional 5.8
million kW-hours per year beyond the requirements to achieve the floor
level. The costs associated with these hazardous waste treatment/
disposal and energy impacts are accounted for in the national
annualized compliance cost estimates. Therefore, based on these factors
and costs of approximately $18 million per additional ton of mercury
removed, we are not proposing a beyond-the-floor standard based on
activated carbon injection.
Feed Control of Mercury in the Hazardous Waste. We also evaluated a
beyond-the-floor level of 100 [mu]g/dscm, which represents a 20%
reduction from the floor level. We chose a 20% reduction as a level
that represents the practicable extent that additional feedrate control
of mercury in hazardous waste (beyond feedrate control that may be
necessary to achieve the floor level) can be used and still achieve
modest emissions reductions.\89\ The national incremental annualized
compliance cost for incinerators to meet this beyond-the-floor level
rather than comply with the floor controls would be approximately $1.8
million and would provide an incremental reduction in mercury emissions
beyond the MACT floor controls of 0.11 tons per year. Nonair quality
health and environmental impacts and energy effects were also
evaluated. Therefore, based on these factors and costs of approximately
$17 million per additional ton of mercury removed, we are not proposing
a beyond-the-floor standard based on feed control of mercury in the
hazardous waste.
---------------------------------------------------------------------------

\89\ Ideally, a methodology to estimate costs of feed control
should consider lost revenues associated with hazardous wastes not
fired and costs to implement feed control of metals and chlorine. We
attempted to conduct such an analysis; however, we concluded that
there are too many uncertainties to do this analysis. Instead, we
developed an alternative approach to cost feed control of metals and
chlorine in the hazardous waste based on the assumption that a
source would not implement a feed control strategy if the costs
exceed the costs to retrofit an existing air pollution control
device. Thus, our cost estimates of feed control represent an upper
bound estimate on likely costs to control metals or chlorine in
hazardous waste. See USEPA, ``Draft Technical Support Document for
HWC MACT Replacement Standards, Volume V: Emission Estimates and
Engineering Costs,'' March 2004, Chapter 4.
---------------------------------------------------------------------------

For the reasons discussed above, we propose a mercury emissions
standard of 130 [mu]g/dscm for existing incinerators.
3. What Is the Rationale for the MACT Floor for New Sources?
Mercury emissions from new incinerators are currently limited to 45
[mu]g/dscm by Sec. 63.1203(b)(2). This standard was promulgated in the
Interim Standards Rule (See 67 FR at 6796).
The MACT floor for new sources for mercury would be 8 [mu]g/dscm,
which considers emissions variability. This is an emission level that
the single best performing source identified with the SRE/Feed Approach
considering compliance test data could be expected to achieve in 99 of
100 future tests when operating under conditions identical to the test
conditions during which the emissions data were obtained.
As we did for existing sources, we also invite comment on basing
the floor analysis on the normal emissions data using the Emissions
Approach. The floor level using the normal data is 0.70 [mu]g/dscm,
which considers emissions variability. If we were to adopt such an
approach, we would require sources to comply with the limit on an
annual basis because it is based on normal emissions data.
4. EPA's Evaluation of Beyond-the-Floor Standards for New Sources
We identified two potential beyond-the-floor techniques for control
of mercury: (1) Use of a carbon bed; and (2) control of mercury in the
hazardous waste feed.
Carbon Bed System. We evaluated a carbon bed system as beyond-the-
floor control for further reduction of mercury emissions. Given the
relatively low floor level, we made a conservative assumption that the
use of a carbon bed system would provide 50% mercury control. The
incremental annualized compliance cost for a new incinerator with
average gas flow rate to meet a beyond-the-floor level of 4 [mu]g/dscm,
rather than comply with the floor level, would be approximately $0.22
million and would provide an incremental reduction in mercury emissions
of approximately 2.1 pounds per year. Nonair quality health and
environmental impacts and energy effects are accounted for in the
national annualized compliance cost estimates. Therefore, based on
these factors and costs of approximately $200 million per additional
ton of mercury removed, we are not proposing a beyond-the-floor
standard based on a carbon bed system.
Feed Control of Mercury in the Hazardous Waste. We also believe
that the expense for a reduction in mercury emissions based on further
control of mercury concentrations in the

[[Page 21244]]

hazardous waste is not warranted. A beyond-the-floor level of 6.4
[mu]g/dscm, which represents a 20% reduction from the floor level,
would result in a small incremental reduction in mercury emissions. For
similar reasons discussed above for existing sources, we likewise
conclude that a beyond-the-floor standard based on controlling the
mercury in the hazardous waste feed would not be justified because of
the costs and emission reductions. Therefore, we propose a mercury
standard of 8 [mu]g/dscm for new sources.

C. What Are the Proposed Standards for Particulate Matter?

We are proposing to establish standards for existing and new
incinerators that limit emissions of particulate matter to 0.015 and
0.00070 gr/dscf, respectively.
1. What Is the Rationale for the MACT Floor for Existing Sources?
Particulate matter emissions for existing incinerators are
currently limited to 0.015 gr/dscf (34 mg/dscm) by Sec. 63.1203(a)(7).
This standard was promulgated in the Interim Standards Rule (See 67 FR
at 6796). The particulate matter standard is a surrogate control for
the hazardous air pollutant metals antimony, cobalt, manganese, nickel,
and selenium.
We have compliance test emissions data for most incinerators. For
some sources, we have compliance test emissions data from more than one
compliance test campaign. Our data base of particulate matter stack
emission concentrations range from 0.0002 to 0.078 gr/dscf.
To identify the MACT floor for incinerators, we evaluated the
compliance test emissions data associated with the most recent test
campaign using the Air Pollution Control Technology Approach. The
calculated floor is 0.020 gr/dscf (46 mg/dscm), which considers
emissions variability. This is an emission level that the average of
the best performing sources could be expected to achieve in 99 of 100
future tests when operating under conditions identical to the
compliance test conditions during which the emissions data were
obtained. The calculated floor level of 0.020 gr/dscf is less stringent
than the interim standard of 0.015 gr/dscf, which is a regulatory limit
relevant in identifying the floor level (so as to avoid any backsliding
from a current level of performance achieved by all incinerators, and
hence, the level of minimal stringency at which EPA could calculate the
MACT floor). Therefore, we are proposing the floor level as the current
emission standard of 0.015 gr/dscf. This emission level is currently
being achieved by all sources.
2. EPA's Evaluation of Beyond-the-Floor Standards for Existing Sources
We evaluated improved particulate matter control to achieve a
beyond-the-floor standard of 17 mg/dscm (0.0075 gr/dscf). For an
existing incinerator that needs a significant reduction in particulate
matter emissions, we assumed and costed a new baghouse to achieve the
beyond-the-floor level. If little or modest emissions reductions were
needed, then improved control was costed as design, operation, and
maintenance modifications of the existing particulate matter control
equipment.
The national incremental annualized compliance cost for
incinerators to meet this beyond-the-floor level rather than comply
with the floor controls would be approximately $3.9 million and would
provide an incremental reduction in particulate matter emissions beyond
the MACT floor of 48 tons per year. Nonair quality health and
environmental impacts and energy effects were evaluated to estimate the
nonair quality health and environmental impacts between further
improvements to control particulate matter and controls likely to be
used to meet the floor level. We estimate that this beyond-the-floor
option would increase the amount of hazardous waste generated by 48
tons per year and would also require sources to use an additional 2.7
million kW-hours per year beyond the requirements to achieve the floor
level. The costs associated with these impacts are accounted for in the
national annualized compliance cost estimates. Therefore, based on
these factors and costs of approximately $81,000 per additional ton of
particulate matter removed, we are not proposing a beyond-the-floor
standard based on improved particulate matter control.
3. What Is the Rationale for the MACT Floor for New Sources?
Particulate matter emissions from new incinerators are currently
limited to 0.015 gr/dscf (34 mg/dscm) by Sec. 63.1203(b)(7). This
standard was promulgated in the Interim Standards Rule (See 67 FR at
6796).
The MACT floor for new sources for particulate matter would be 1.6
mg/dscm (0.00070 gr/dscf), which considers emissions variability. This
is an emission level that the single best performing source identified
with the Air Pollution Control Technology Approach could be expected to
achieve in 99 of 100 future tests when operating under operating
conditions identical to the test conditions during which the emissions
data were obtained.
As discussed in Part Two, Section II, we considered whether to
propose separate standards (subcategorize) for particulate matter for
several different potential subcategories such as government-owned
versus non-government incinerators and liquid injection versus solid
fuel-fired incinerators. We determined that the emission
characteristics from these potential subcategories are not
statistically different, and, therefore, separate standards for
particulate matter are not warranted. We request comment on whether
these subcategorization considerations capture the appropriate
differences in manufacturing process, emission characteristics, or
technical feasibility for particulate matter. We note, for example, the
single best performing source, which is the basis of the floor level
for new incinerators, is an incinerator used to decontaminate scrap
metal. Though we believe these sources are best performers because they
use highly efficient baghouses for the capture of particulate matter,
and, therefore, appropriate for inclusion in the analysis, we invite
comment on whether we have considered the appropriate subcategories for
particulate matter. We note that a floor level based on the second best
performing incinerator source would be 0.0021 gr/dscf.
4. EPA's Evaluation of Beyond-the-Floor Standards for New Sources
We evaluated improved emissions control based on a state-of-the-art
baghouse using a high quality fabric filter bag material to achieve a
beyond-the-floor standard of 1.2 mg/dscm (0.0005 gr/dscf). The
incremental annualized compliance cost for a new incinerator to meet
this beyond-the-floor level, rather than comply with the floor level,
would be approximately $80,000 and would provide an incremental
reduction in particulate matter emissions of approximately 0.15 tons
per year. Nonair quality health and environmental impacts and energy
effects were also evaluated and are accounted for in the national
annualized compliance cost estimates. We estimate that this option
would require a new source to use an additional 0.2 million kW-hours
per year. For these reasons and a cost-effectiveness of $0.53 million
per ton of particulate matter removed, we are not proposing a beyond-
the-floor standard based on improved particulate matter control for new
incinerators. Therefore, we propose a particulate

[[Page 21245]]

matter standard of 1.6 mg/dscm for new sources.

D. What Are the Proposed Standards for Semivolatile Metals?

We are proposing to establish standards for existing and new
incinerators that limit emissions of semivolatile metals (cadmium and
lead) to 59 ug/dscm and 6.5 ug/dscm, respectively.
1. What Is the Rationale for the MACT Floor for Existing Sources?
Semivolatile metals emissions from existing incinerators are
currently limited to 240 ug/dscm by Sec. 63.1203(a)(3). This standard
was promulgated in the Interim Standards Rule (See 67 FR at 6796).
Incinerators control emissions of semivolatile metals with air
pollution control equipment and/or by controlling the feed
concentration of semivolatile metals in the hazardous waste.
We have compliance test emissions data for nearly 30 incinerators.
Semivolatile metal stack emissions range from approximately 4 to 29,000
ug/dscm. These emissions are expressed as mass of semivolatile metals
per unit volume of stack gas. Lead was usually the most significant
contributor to semivolatile emissions during compliance test
conditions.
To identify the MACT floor, we evaluated the compliance test
emissions data associated with the most recent test campaign using the
SRE/Feed Approach. The calculated floor is 59 ug/dscm, which considers
emissions variability. This is an emission level that the average of
the best performing sources could be expected to achieve in 99 of 100
future tests when operating under conditions identical to the
compliance test conditions during which the emissions data were
obtained. We estimate that this emission level is being achieved by 52%
of sources. The floor level would reduce semivolatile metals emissions
by 0.43 tons per year.
2. EPA's Evaluation of Beyond-the-Floor Standards for Existing Sources
We identified two potential beyond-the-floor techniques for control
of semivolatile metals: (1) Improved particulate matter control; and
(2) control of semivolatile metals in the hazardous waste feed.
Improved Particulate Matter Control. Controlling particulate matter
also controls emissions of semivolatile metals. We evaluated a beyond-
the-floor level of 30 [mu]g/dscm, which is a 50% reduction from the
floor level, based on additional reductions of particulate matter
emissions by operating and maintaining existing control equipment to
have improved collection efficiency. The national incremental
annualized compliance cost for incinerators to meet this beyond-the-
floor level rather than comply with the floor controls would be
approximately $3.0 million and would provide an incremental reduction
in semivolatile metals emissions beyond the MACT floor controls of 190
pounds per year. Nonair quality health and environmental impacts and
energy effects were evaluated to estimate the impacts between further
improvements to control particulate matter and controls likely to be
used to meet the floor level. We estimate that this beyond-the-floor
option would increase the amount of hazardous waste generated by 50
tons per year and would require sources to use an additional 3.4
million kW-hours per year beyond the requirements to achieve the floor
level. The costs associated with these hazardous waste treatment and
energy impacts are accounted for in the national annualized compliance
cost estimates. Therefore, based on these factors and costs of
approximately $31 million per additional ton of semivolatile metals
removed, we are not proposing a beyond-the-floor standard based on
improved particulate matter control.
Feed Control of Semivolatile Metals in the Hazardous Waste. We also
evaluated a beyond-the-floor level of 47 [mu]g/dscm, which represents a
20% reduction from the floor level. We chose a 20% reduction as a level
that represents the practicable extent that additional feedrate control
of semivolatile metals in the hazardous waste can be used and still
achieve modest emissions reductions. The national incremental
annualized compliance cost for incinerators to meet this beyond-the-
floor level rather than comply with the floor controls would be
approximately $1.7 million and would provide an incremental reduction
in semivolatile metals emissions beyond the MACT floor of 90 pounds per
year. Nonair quality health and environmental impacts and energy
effects were also evaluated and are accounted for in the national
annualized compliance cost estimates. For these reasons and costs of
approximately $39 million per additional ton of semivolatile metals
removed, we are not proposing a beyond-the-floor standard based on feed
control of semivolatile metals in the hazardous waste.
For the reasons discussed above, we propose to establish the
emission standard for existing incinerators at 59 [mu]g/dscm.
3. What Is the Rationale for the MACT Floor for New Sources?
Semivolatile metals emissions from new incinerators are currently
limited to 120 [mu]g/dscm by Sec. 63.1203(b)(3). This standard was
promulgated in the Interim Standards Rule (See 67 FR at 6796).
The MACT floor for new sources for semivolatile metals would be 6.5
[mu]g/dscm, which considers emissions variability. This is an emission
level that the single best performing source identified with the SRE/
Feed Approach could be expected to achieve in 99 of 100 future tests
when operating under conditions identical to the test conditions during
which the emissions data were obtained.
4. EPA's Evaluation of Beyond-the-Floor Standards for New Sources
We identified two potential beyond-the-floor techniques for control
of semivolatile metals: (1) Improved control of particulate matter; and
(2) control of semivolatile metals in the hazardous waste feed.
Improved Particulate Matter Control. We evaluated a standard of 3.3
[mu]g/dscm, which is a 50% reduction from the floor level, based on a
state-of-the-art baghouse using a high quality fabric filter bag
material as beyond-the-floor control for further reductions in
semivolatile metals emissions. The incremental annualized compliance
cost for a new incinerator with an average gas flow rate to meet this
beyond-the-floor level, rather than comply with the floor level, would
be approximately $80,000 and would provide an incremental reduction in
semivolatile metals emissions of approximately 2 pounds per year.
Nonair quality health and environmental impacts and energy effects were
also evaluated and are included in the cost estimates. We estimate that
this option would require a new source to use an additional 0.2 million
kW-hours per year. For these reasons and costs of $94 million per ton
of semivolatile metals removed, we are not proposing a beyond-the-floor
standard based on improved particulate matter control for new sources.
Feed Control of Semivolatile Metals in the Hazardous Waste. We also
believe that the expense for a reduction in semivolatile metals
emissions based on further control of semivolatile metals
concentrations in the hazardous waste is not warranted. A beyond-the-
floor level of 5.2 [mu]g/dscm, which represents a 20% reduction from
the floor level, would result in little additional semivolatile metals
reductions. For similar reasons discussed above for existing sources, we

[[Page 21246]]

judge that a beyond-the-floor standard based on controlling the
semivolatile metals in the hazardous waste feed would not be justified
because of the costs and expected emission reductions. Therefore, we
propose a semivolatile metals standard of 6.5 [mu]g/dscm for new sources.

E. What Are the Proposed Standards for Low Volatile Metals?

We are proposing to establish standards for existing and new
incinerators that limit emissions of low volatile metals (arsenic,
beryllium, and chromium) to 84 [mu]g/dscm and 8.9 [mu]g/dscm, respectively.
1. What Is the Rationale for the MACT Floor for Existing Sources?
Low volatile metals emissions from existing incinerators are
currently limited to 97 [mu]g/dscm by Sec. 63.1203(a)(4). This
standard was promulgated in the Interim Standards Rule (See 67 FR at
6796). Incinerators control emissions of low volatile metals with air
pollution control equipment and/or by controlling the feed
concentration of low volatile metals in the hazardous waste.
We have compliance test emissions data for nearly 30 incinerators.
Low volatile metal stack emissions range from approximately 1 to 4,300
[mu]g/dscm. These emissions are expressed as mass of low volatile
metals per unit volume of stack gas.
To identify the MACT floor, we evaluated the compliance test
emissions data associated with the most recent test campaign using the
SRE/Feed Approach. The calculated floor is 84 [mu]g/dscm, which
considers emissions variability. This is an emission level that the
average of the best performing sources could be expected to achieve in
99 of 100 future tests when operating under conditions identical to the
compliance test conditions during which the emissions data were
obtained. We estimate that this emission level is being achieved by 85%
of sources and would reduce low volatile metals emissions by 56 pounds
per year.
2. EPA's Evaluation of Beyond-the-Floor Standards for Existing Sources
We identified two potential beyond-the-floor techniques for control
of low volatile metals: (1) Improved particulate matter control; and
(2) control of low volatile metals in the hazardous waste feed.
Improved Particulate Matter Control. Controlling particulate matter
also controls emissions of low volatile metals. We evaluated a beyond-
the-floor level of 42 [mu]g/dscm, which is a 50% reduction from the
floor level, based on additional reductions of particulate matter
emissions by operating and maintaining existing control equipment to
have improved collection efficiency. The national incremental
annualized compliance cost for incinerators to meet this beyond-the-
floor level rather than comply with the floor controls would be
approximately $0.88 million and would provide an incremental reduction
in low volatile metals emissions beyond the MACT floor controls of 365
pounds per year. Nonair quality health and environmental impacts and
energy effects were evaluated to estimate the impacts between further
improvements to control particulate matter and controls likely to be
used to meet the floor level. We estimate that this beyond-the-floor
option would increase the amount of hazardous waste generated by 100
tons per year and would require sources to use an additional 0.7
million kW-hours per year beyond the requirements to achieve the floor
level. The costs associated with these impacts are accounted for in the
national annualized compliance cost estimates. Therefore, based on
these factors and costs of approximately $4.8 million per additional
ton of low volatile metals removed, we are not proposing a beyond-the-
floor standard based on improved particulate matter control.
Feed Control of Low Volatile Metals in the Hazardous Waste. We also
evaluated a beyond-the-floor level of 67 [mu]g/dscm, which represents a
20% reduction from the floor level. We chose a 20% reduction as a level
that represents the practicable extent that additional feedrate control
of low volatile metals in the hazardous waste can be used and still
achieve modest emissions reductions. The national incremental
annualized compliance cost for incinerators to meet this beyond-the-
floor level rather than comply with the floor controls would be
approximately $0.25 million and would provide an incremental reduction
in low volatile metals emissions beyond the MACT floor controls of 0.11
tons per year. Nonair quality health and environmental impacts and
energy effects were also evaluated and are accounted for in the
national annualized compliance cost estimates. Therefore, based on
these factors and costs of approximately $2.2 million per additional
ton of low volatile metals removed, we are not proposing a beyond-the-
floor standard based on feed control of low volatile metals in the
hazardous waste.
For the reasons discussed above, we propose to establish the
emission standard for existing incinerators at 84 [mu]g/dscm.
3. What Is the Rationale for the MACT Floor for New Sources?
Low volatile metal emissions from new incinerators are currently
limited to 97 [mu]g/dscm by Sec. 63.1203(b)(4). This standard was
promulgated in the Interim Standards Rule (See 67 FR at 6796).
The MACT floor for new sources for low volatile metals would be 8.9
[mu]g/dscm, which considers emissions variability. This is an emission
level that the single best performing source identified with the SRE/
Feed Approach could be expected to achieve in 99 of 100 future tests
when operating under conditions identical to the test conditions during
which the emissions data were obtained.
4. EPA's Evaluation of Beyond-the-Floor Standards for New Sources
We identified two potential beyond-the-floor techniques for control
of low volatile metals: (1) Improved control of particulate matter; and
(2) control of low volatile metals in the hazardous waste feed.
Improved Particulate Matter Control. We evaluated a standard of 4.5
[mu]g/dscm, which is a 50% reduction from the floor level, based on a
state-of-the-art baghouse using a high quality fabric filter bag
material as beyond-the-floor control for further reductions in low
volatile metals emissions. The incremental annualized compliance cost
for a new incinerator with average gas flowrate to meet this beyond-
the-floor level, rather than comply with the floor level, would be
approximately $80,000 and would provide an incremental reduction in low
volatile metals emissions of approximately 2.3 pounds per year. Nonair
quality health and environmental impacts and energy effects were also
evaluated and are included in the cost estimates. For these reasons and
costs of $69 million per ton of low volatile metals removed, we are not
proposing a beyond-the-floor standard based on improved particulate
matter control for new sources.
Feed Control of Low Volatile Metals in the Hazardous Waste. We also
believe that the expense associated with a reduction in low volatile
metals emissions based on further control of low volatile metals
concentrations in the hazardous waste is not warranted. A beyond-the-
floor level of 7.1 [mu]g/dscm, which represents a 20% reduction from
the floor level, would result in little additional low volatile metals
reductions. For similar reasons discussed above for existing sources, we

[[Continued on page 21247]]

Notices For
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994

National Emission Standards for Hazardous Air Pollutants: Proposed Standards for Hazardous Air Pollutants for Hazardous Waste Combustors (Phase I Final Replacement Standards and Phase II)

Local Navigation