Friday
April 19, 1996
Part II
Environmental
Protection Agency
40 CFR Part 60, et al.
Hazardous Waste Combustors; Revised
Standards; Proposed Rule
-------
-------
17358
Federal Register /Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
ENVIRONMENTAL PROTECTION
AGENCY
40 CFR Parts 60,63, 260, 261, 264, 265,
266, 270, and 271
[FRL-5447-2]
RIN2050-AF01
Revised Standards for Hazardous
Waste Combustors
AGENCY: Environmental Protection
Agency.
ACTION: Proposed rule.
SUMMARY: The Agency is proposing
revised standards for hazardous waste
incinerators i hazardous waste-burning
cement kilns, and hazardous waste-
burning lightweight aggregate kilns.
These standards are being proposed
under joint authority of the Clean Air
Act (CAA) and Resource Conservation
and Recovery Act (RCRA). The
standards limit emissions of chlorinated
dioxins and furans, other toxic organic
compounds, toxic metals, hydrochloric
acid, chlorine gas, and particulate
matter. These standards reflect the
performance of Maximum Achievable
Control Technologies (MACT) as
specified by the Clean Air Act. The
MACT standards also should result in
increased protection to human health
and the environment over existing
RCRA standards. The nature of this
proposal requires that the following
actions also be proposed: proposing the
addition of hazardous waste-burning
lightweight aggregate kilns to the list of
source categories in accordance with
112(c)(5) of the Act; exempting from
RCRA emission controls secondary lead
facilities subject to MACT; considering
an exclusion for certain "comparable
fuels"; and revising the small quantity
burner exemption under the BIF rule.
DATES: EPA will accept public
comments on this proposed rule until
June 18,1996V
ADDRESSES: Commenters must send an
original and two copies of their
comments referencing docket number
F-96-RCSP-FFFFF to: RCRA Docket
Information Center, Office of Solid
Waste (5305W), U.S. Environmental
Protection Agency Headquarters (EPA,
HQ), 401 M Street, SW., Washington,
DC 20460. Deliveries of comments
should be made to the Arlington, VA,
address listed below. Comments may
also be submitted electronically through
the Internet to: RCRA-
Docket@epamail.epa.gov. Comments in
electronic format should also be
identified by the docket number F-96-
RCSP-FFFFF. All electronic comments
must be submitted as an ASCII file
avoiding the use of special characters
and any form of encryption.
Commenters should not submit
electronically any Confidential Business
Information (CBI). An original and two
copies of CBI must be submitted under
separate cover to: RCRA CBI Document
Control Officer, Office of Solid Waste
(5305W), U.S. EPA, 401 M Street, SW,
Washington, DC 20460.
Public comments and supporting
materials are available for viewing in
the RCRA Information Center (RIG),
located at Crystal Gateway One, 1235
Jefferson Davis Highway, First Floor,
Arlington, VA. The RIG is open from 9
a.m. to 4 p.m., Monday through Friday,
excluding federal holidays. To review
docket materials, the public must make
an appointment by calling (703) 603-
9230. The public may copy a maximum
of 100 pages from any regulatory docket
at no charge. Additional copies cost
$.15/page. The index and some
supporting materials are available
electronically. See the "Supplementary
Information" section for information on
accessing them.
A public hearing will be held, if
requested, to discuss the proposed
standards for hazardous waste
combustors, in accordance with section
307(d)(5) of the Act. Persons wishing to
make an oral presentation at a public
hearing should contact the EPA at the
address given in the ADDRESSES section
of-this preamble. Oral presentations will
be limited to 5 minutes each, unless
additional time is feasible. Any member
of the public may file a written
statement before, during, or within 30
days after the hearing. Written
statements should be addressed to the
RCRA Docket Section address given in
the ADDRESSES section of this preamble
and should refer to Docket No. F-96-
RCSP-FFFFF. A verbatim transcript of
the hearing and written statements will
be available for public inspection and
copying during normal working hours at
the EPA's RCRA Docket Section in
Washington, D.C. (see ADDRESSES
section of this preamble).
FOR FURTHER INFORMATION CONTACT: For
general information, contact the RCRA
Hotline at 1-800-424-9346 or TDD 1-
800-553-7672 (hearing impaired). In
the Washington metropolitan area, call
703-412-9810 or TDD 703-412-3323.
For more detailed information on
specific aspects of this rulemaking,
contact Larry Denyer, Office of Solid
Waste (5302W), U.S. Environmental
Protection Agency, 401 M Street, SW.,
Washington, DC 20460, (703) 308-8770,
electronic mail:
Denyer.Larry@epamail.epa.gov. For
more detailed information on
implementation of this rulemaking,
contact Val de la Fuente, Office of Solid
Waste (5303W), U.S. Environmental
Protection Agency, 401 M Street, SW.,
Washington, DC 20460, (703) 308-7245,
electronic mail:
DeLaFuente.Val@epamail.epa.gov. For
more detailed information on regulatory
impact assessment of this rulemaking,
contact Gary Ballard, Office of Solid
Waste (5305), U.S. Environmental
Protection Agency, 401 M Street, SW.,
Washington/DC 20460, (202) 260-2429,
electronic mail:
Ballard.Gary@epamail.epa.gov. For
more detailed information on risk
analyses of this rulemaking, contact
David Layland, Office of Solid Waste
(5304), U.S. Environmental Protection
Agency, 401 M Street, SW., Washington,
DC 20460, (202) 260-4796, electronic
mail: Layland.David@epamail.epa.gov.
SUPPLEMENTARY INFORMATION: The index
and the following supporting materials
are available on the Internet: (List
documents) Follow these instructions to
access the information electronically:
Gopher: gopher.epa.gov
WWW: http://www.epa.gov
Dial-up: (919) 558-0335.
This report can be accessed off the
main EPA Gopher menu, in the
directory: EPA Offices and Regions/
Office of Solid Waste and Emergency
Response (OSWER)/Office of Solid
Waste (RCRA)/(consult with
Communication Strategist for precise
subject heading)
FTP: ftp.epa.gov
Login: anonymous
Password: Your Internet address
Files are located in /pub/gopher/
OSWRCRA
The official record for this action will
be kept in paper form. Accordingly, EPA
will transfer all comments received
electronically into paper form and place
them in the official record, which will
also include all comments submitted
directly in writing. The official record is
the paper record maintained at the
address in ADDRESSES at the beginning
of this document.
EPA responses to comments, whether
the comments are written or electronic,
will be in a notice in the Federal
Register or in a response to comments
document placed in the official record
for this rulemaking. EPA will not
immediately reply to Commenters
electronically other than to seek
clarification of electronic comments that
may be garbled in transmission or
during conversion to paper form, as
discussed above.
Glossary of Acronyms
APCD—Air Pollution Control Device
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17359
BDAT—Best Demonstrated Available
Technology
BIFs—Boilers and Industrial Furnaces
BTF—Beyond-the-Floor
CAA—Clean Air Act
Clz—Chlorine
CO—Carbon Monoxide
D/F—Dioxins/Furans
D/O/M—Design/Operation/Maintenance
ESP—ElectrostaticTPrecipitator
EU—European Union
FF—Fabric Filter
HAP—Hazardous Air Pollutant
HC—Hydrocarbons
HC1—Hydrochloric acid
Hg—Mercury
HHE—Human Health and the
Environment
HON—Hazardous Organic NESHAPs
HSVVA—Hazardous and Solid Waste
Amendments
HVVC—Hazardous Waste Combustion/
Combustor
ICR—Information Collection Request
LDR—Land Disposal Restrictions
LVM—Low-volatile Metals
LWAK—Lightweight Aggregate Kiln
MACT—Maximum Achievable Control
Technology
MTEC—Maximum Theoretical Emission
Concentration
NESHAPs—National Emission
Standards for Hazardous Air
Pollutants
PM—Particulate Matter
PICs—Products of Incomplete
Combustion
RCRA—Resource Conservation and
Recovery Act
RIA—Regulatory Impact Assessment
SVM—Semivolatile Metals
TCLP—Toxicity Characteristic Leaching
Procedure
UTS—Universal Treatment Standards
Port One: Background
I. Overview
II. Relationship of Today's Proposal to
EPA's Waste Minimization National Plan
Part Two: Devices That Would Be Subject To
The Proposed Emission Standards
I. Hazardous Waste Incinerators
A. Overview
B. Summary of Major Incinerator Designs
C. Number of Incinerator Facilities
D. Typical Emission Control Devices For
Incinerators
II. Hazardous Waste-Burning Cement Kilns
A. Overview of Cement Manufacturing
B. Summary of Major Design and Operating
Features of Cement Kilns
C. Number of Facilities
D. Emissions Control Devices
III. Hazardous Waste-Burning Lightweight
Aggregate Kilns
A. Overview of Lightweight Aggregate
Kilns (LWAKs)
B. Major Design and Operating Features
C. Number of Facilities
D. Air Pollution Control Devices
Part Three: Decision Process for Setting
National Emission Standards for
Hazardous Air Pollutants (NESHAPs)
I. Source of Authority for NESHAP
Development
II. Procedures and Criteria for Development
of NESHAPs
III. List of Categories of Major and Area
Sources
A. Clean Air Act Requirements
B. Hazardous Waste Incinerators
C. Cement Kilns
D. Lightweight Aggregate Kilns
IV. Proposal to Subject Area Sources to the
NESHAPs under Authority of Section
V. Selection of MACT Floor for Existing
Sources
A. Proposed Approach: Combined
Technology-Statistical Approach
B. Another Approach Considered But Not
Used
C. Identifying Floors as Proposed in
CETRED
D. Establishing Floors One HAP or HAP
' Group at a Time
VI. Selection of Beyond-the-Floor Levels
for Existing Sources
VII. Selection of MACT for New Sources
VIII. RCRA Decision Process
A. RCRA and CAA Mandates to Protect
Human Health and the Environment
B. Evaluation of Protectiveness
C. Use of Site-Specific Risk Assessments
under RCRA
Part Four: Rationale for Selecting the
Proposed Standards
I. Selection of Source Categories and
Pollutants
A. Selection of Sources and Source
Categories
B. Selection of Pollutants
C. Applicability of the Standards Under
Special Circumstances
II. Selection of Format for the Proposed
Standards
A. Format of the Standard
B. Averaging Periods
III. Incinerators: Basis and Level for the
Proposed NESHAP Standards for New
and Existing Sources
A. Summary of MACT Standards for
Existing Incinerators
B. Summary of MACT Standards For New
Incinerators
C. Evaluation of Protectiveness
IV. Cement Kilns: Basis and Level for the
Proposed NESHAP Standards for New
and Existing Sources
A. Summary of Standards for Existing
Cement Kilns
B. MACT for New Hazardous Waste-
Burning Cement Kilns
C. Evaluation of Protectiveness
V. Lightweight Aggregate Kilns: Basis and
Level for the Proposed NESHAP
Standards for New and Existing Sources
A. Summary of MACT Standards for
Existing LWAKs
B. MACT for New Sources
C. Evaluation of Protectiveness
VI. Achievability of the Floor Levels
VII. Comparison of the Proposed Emission
Standards With Emission Standards for
Other Combustion Devices
VIII. Alternative Floor (12 Percent) Option
Results
A. Summary of Results of 12 Percent
Analysis
B. Summary of MACT Floor Cost Impacts
and Emissions Reductions
C. Alternative Floor Option: Percent
Reduction Refinement
IX. Additional Data for Comment
Part Five: Implementation
I. Selection of Compliance Dates
A. Existing Sources
B. New Sources
C. One year extensions for Pollution
Prevention/Waste Minimization
II. Selection of Proposed Monitoring
Requirements
A. Monitoring Hierarchy
B. Use of Comprehensive Performance Test
Data to Establish Operating Limits
C. Compliance Monitoring Requirements
D. Combustion Fugitive Emissions
E. Automatic Waste Feed Cutoff (AWFCO)
Requirements and Emergency Safety
Vent (ESV) Openings
F. Quality Assurance for Continuous
Monitoring Systems
III. MACT Performance Testing and
Related Issues
A. MACT Performance Testing
B. RCRA Trial Burns
C. Waiver of MACT Performance Testing
for HWCs Feeding De Minimis Levels of
Metals or Chlorine
D. Relative Accuracy Tests for GEMS
IV. Selection of Manual Stack Sampling
Methods
V. Notification, Recordkeeping, Reporting,
and Operator Certification Requirements
A. Notification Requirements
B. Reporting Requirements
C. Recordkeeping Requirements
VI. Permit Requirements
A. Coordination of RCRA and CAA
Permitting Processes
B. Permit Application Requirements
C. Clarifications on Definitions and Permit
Process Issues
D. Pollution Prevention/Waste
Minimization Options
E. Permit Modifications Necessary to Come
Into Compliance With MACT Standards
VII. State Authorization
A. Authority for Today's Rule
B. Program Delegation Under the Clean Air
Act
C. RCRA State Authorization
VIII. Definitions
A. Definitions Proposed in §63.1201
B. Conforming Definitions Proposed in
§§ 260.10 and 270.2
C. Clarification of RCRA Definition of
Industrial Furnace
Part Six: Miscellaneous Provisions and Issues
I. Comparable Fuel Exclusion
A. EPA's Approach to Establishing
Benchmark Constituent Levels
B. Sampling, Analysis, and Statistical
Protocols Used
C. Options for the Benchmark Approach
D. Comparable Fuel Specification
E. Exclusion of Synthesis Gas Fuel
F. Implementation of the Exclusion
G. Transportation and Storage
H. Speculative Accumulation
I. Regulatory Impacts
II. Miscellaneous Revisions to the Existing
Rules
A. Revisions to the Small Quantity Burner
Exemption under the BIF Rule
B. The Waiver of the PM Standard under
the Low Risk Waste Exemption of the
-------
17360
Federal Register 7 VpL 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
BIF Rule Would Not Be Applicable to
HWCs
C. The "Low Risk Waste" Exemption from
the Emission Standards Provided by the
Existing Incinerator Standards Would Be
Superseded by the MACT Rules
D. Bevill Residues
E. Applicability of Regulations to Cyanide
Wastes
F. Shakedown Concerns
G. Extensions of Time Under Certification
of Compliance
H. Technical Amendments to the BIF Rule
I. Clarification of Regulatory Status of Fuel
Blenders
J. Change in Reporting Requirements for
Secondary Lead Smelters Subject to
MACT
Part Seven: Analytical and Regulatory
Requirements
I. Executive Order 12866
II. Regulatory Options
III. Assessment of Potential Costs and
Benefits
A. Introduction
B. Analysis and Findings
C. Total Incremental Cost per Incremental
Reduction in HAP Emissions
D. Human Health Benefits
E. Other Benefits
IV. Other Regulatory Issues
A. Environmental Justice
B. Unfunded Federal Mandates
C. Regulatory Takings
D. Incentives for Waste Minimization and
Pollution Prevention
V. Regulatory Flexibility Analysis
VI. Paperwork Reduction Act
VII. Request for Data
Appendix—Comparable Fuel Constituent
and Physical Specifications
PART 60—STANDARDS OF
PERFORMANCE FOR NEW
STATIONARY SOURCES
PART 63—NATIONAL EMISSION
STANDARDS FOR HAZARDOUS AIR
POLLUTANTS FOR SOURCE
CATEGORIES
PART 260—HAZARDOUS WASTE
MANAGEMENT SYSTEM: GENERAL
PART 261—IDENTIFICATION AND LISTING
OF HAZARDOUS WASTE
PART 264—STANDARDS FOR OWNERS
AND OPERATORS OF HAZARDOUS
WASTE TREATMENT, STORAGE, AND
DISPOSAL FACILITIES
PART 265—INTERIM STATUS STANDARDS
FOR OWNERS AND OPERATORS OF
HAZARDOUS WASTE TREATMENT,
STORAGE, AND DISPOSAL FACILITIES
PART 266—STANDARDS FOR THE
MANAGEMENT OF SPECIFIC
HAZARDOUS WASTES AND SPECIFIC
TYPES OF HAZARDOUS WASTE
MANAGEMENT FACILITIES
PART 270—EPA ADMINISTERED PERMIT
PROGRAMS: THE HAZARDOUS
WASTE PERMIT PROGRAM
PART 271—REQUIREMENTS FOR
AUTHORIZATION OF STATE
HAZARDOUS WASTE PROGRAMS
PART ONE: BACKGROUND
I. Overview
The U.S. Environmental Protection
Agency (EPA) is proposing to revise
standards for hazardous waste
incinerators and hazardous waste-
burning cement kilns and lightweight
aggregate kilns (LWAKs) under joint
authority of the Clean Air Act, as
amended, (CAA) and the Resource
Conservation and Recovery Act, as
amended (RCRA). The emission
standards in today's proposal have been
developed under the CAA provisions
concerning the maximum level of
achievable control over hazardous air
pollutants (HAPs), taking into
consideration the cost of achieving the
emission reduction, any non-air quality
health and environmental impacts, and
energy requirements. These maximum
achievable control technology (MACT)
standards, also referred to as National
Emission Standards for Hazardous Air
Pollutants (NESHAPs), are proposed in
today's rule for the following HAPs:
dioxins/furans, mercury, two
semivolatile metals (lead and cadmium),
four low volatility metals (antimony,
arsenic, beryllium, and chromium),
particulate matter, and hydrochloric
acid/chlorine gas. Other toxic organic
emissions are addressed by standards
for carbon monoxide (CO) and
hydrocarbons (HC).
This action is being taken for several
reasons. First, this proposal is consistent
with the terms of the 1993 settlement
agreement between the Agency and a
number of groups who challenged EPA's
final RCRA rule entitled "Burning of
Hazardous Waste in Boilers and
Industrial Furnaces" (56 FR 7134, Feb.
21,1991). These groups include the
Natural Resources Defense Council,
Sierra Club, Inc., Hazardous Waste
Treatment Council (now the
Environmental Technology Council),
National Solid Waste Management
Association, and a number of local
citizens' groups. Under this settlement
agreement, the Agency is to propose this
rulemaking by September-November,
1995, and finalize it by December 1996.
Second, EPA has scheduled
rulemakings to develop maximum
achievable control technology (MACT)
standards for hazardous waste •
incinerators and cement kilns. To
minimize the burden on the Agency and
the regulated community, the Agency
has combined its efforts under the CAA
and RCRA into one rulemaking to
establish MACT standards, which also
would satisfy the RCRA settlement
agreement obligations.
Third, the Agency's Hazardous Waste
Minimization and Combustion Strategy,
first announced in May 1993, in
addition to stressing waste
minimization, also made a commitment
to upgrade the emission standards for
hazardous waste-burning facilities. The
three categories of facilities covered in
this proposal burn over 80 percent of
the total amount of hazardous waste
being combusted each year. [The
remaining 15—20 percent is burned in
industrial boilers and other types of
industrial furnaces, which are to be
addressed in the next rulemaking for
which a proposal is to be issued by
December 1998 or sooner.]
Finally, as relates to the development
of revised standards under concurrent
Clean Air Act and RCRA authority, most
of these hazardous waste combustion
facilities are major sources of HAP
emissions. They therefore must be
regulated under section 112(d) of the
Clean Air Act. In addition, EPA noted,
when promulgating the RCRA rules for
boilers and industrial furnaces in 1991
and in a proposal to revise the
incinerator rules, that existing standards
did not fully consider the possibility of
exposure via indirect (non-inhalation)
exposure pathways. 56 FR at 7150,
7167, 7169-70 (Feb. 21,1991); 54 FR at
43720-21, 43723, 43757 (Oct. 26, 1989).
The Agency reiterated these concerns in
the Combustion Strategy announced in
1993 as one of the major factors leading
to its decision to undertake revisions to
the standards for hazardous waste
combustors. As also noted in the
Combustion Strategy and elsewhere,
site-specific RCRA omnibus authority,
whereby permit writers can impose
additional conditions as are necessary to
protect human health and the
environment, can be used to buttress the
existing regulations. See, e.g., 56 FR
7145, at n.8. Nevertheless, this process
is expensive, time-consuming, and not
always sufficiently certain in result. The
Agency thus indicated, in the
Combustion Strategy, that technology-
based standards could provide a
superior means of control by providing
certainty of operating performance.
Because of the joint authorities under
which this rule is being proposed, the
proposal also contains an
implementation scheme that is intended
to harmonize the RCRA and CAA
programs to the maximum extent
permissible by law. In pursuing a
common-sense approach towards this
objective, the proposal seeks to establish
a framework that: (1) Provides for
combined (or at least coordinated) CAA
and RCRA permitting of these facilities;
(2) allows maximum flexibility for
regional, state, and local agencies to
determine which of their resources will
be used for permitting, compliance, and
enforcement efforts; and (3) integrates
the monitoring, compliance testing, and
recordkeeping requirements of the CAA
and RCRA so that facilities will be able
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17361
to avoid two potentially different
regulatory compliance schemes.
In addition, mis proposal addresses
the variety of issues, to the extent
appropriate at this time, raised in
several petitions filed with the Agency.
These petitions are from the Cement
Kiln Recycling Coalition (Jan. 18,1994),
the Hazardous Waste Treatment Council
(May 18,1994), and the Chemical
Manufacturers Association (Oct. 14,
1994).
II. Relationship of Today's Proposal to
EPA's Waste Minimization National
Plan
EPA believes that today's proposed
rule will create significant incentives for
source reduction and recycling by waste
generators that would, in turn, help
facilities achieve compliance with the
MACT standards. RCRA, as well as the
Pollution Prevention Act of 1990 (PPA),
encourage pollution prevention at the
source, and the Clean Air Act mentions
pollution prevention as a specific means
of achieving MACT. In § 112(d)(2) of the
CAA, Congress expressly defined MACT
as the "application of measures,
processes, methods, systems, or
techniques including, but not limited to,
measures which reduce the volume of,
or eliminate emissions of, such
pollutants through process changes,
substitution of materials and other
modifications."
In addition, hi the Hazardous and
Solid Waste Amendments of 1984
(HSWA) to RCRA, Congress established
a national policy for waste
minimization. Section 1003 of RCRA
states that, whenever feasible, the
generation of hazardous waste is to be
reduced or eliminated as expeditiously
as possible. Section 8002(r) requires
EPA to explore the desirability and
feasibility of establishing regulations or
other incentives or disincentives for
reducing or eliminating the generation
of hazardous waste. In 1990, the PPA
reinforced these policies by declaring it
"to be the national policy of the United
States that pollution should be
prevented at the source whenever
feasible" and, when not feasible, waste
should be recycled, treated, or disposed
of—in that order of preference.
Although the Agency has devoted
significant effort to evaluation and
promotion of waste minimization in the
past', the Hazardous Waste
Minimization and Combustion Strategy,
first announced in May 1993, recently
provided a new impetus to this effort.
1 For example, EPA prepared a report to Congress,
"Minimization of Hazardous Wastes" (October
1UBG), that summarized existing waste
minimization activities and evaluated options for
promoting waste minimization.
The Strategy had several components,
among which was reducing the amount
and toxicity of hazardous waste
generated in the United States. Other
components of the Strategy included
strengthening controls on emissions
from hazardous waste combustion units;
enhancing public participation in
facility .permitting; establishing risk
assessment policies with respect to
facility permitting; and continued
emphasis on strong compliance and
enforcement.
EPA held a National Roundtable and
four Regional Roundtables throughout
the nation in 1993-94 to facilitate a
broad dialogue on the spectrum of waste
minimization and combustion issues.
The major messages from these
Roundtables became the building blocks
for EPA's further efforts to promote
source reduction and recycling and
specifically for EPA's Waste
Minimization National Plan, released in
November 1994.
The Waste Minimization National
Plan focuses on the goal of reducing
persistent, bioaccumulative, and toxic
constituents in hazardous waste
nationally by 25 percent by the year
2000 and 50 percent by the year 2005.
The central themes of the National Plan
are: (1) Developing a framework for
setting national priorities for the
minimization of hazardous waste; (2)
promoting multimedia environmental
benefits and preventing cross-media
transfers; (3) demonstrating a strong
preference for source reduction by
shifting attention to hazardous waste
generators to reduce generation at its
source; (4) defining and tracking
progress in minimizing the generation of
wastes; and (5) involving citizens in
waste minimization implementation
decisions. The Agency intends to
continue its pursuit of hazardous waste
minimization under the National Plan
and other Agency initiatives in concert
with the actions proposed in today's
rule.
Of the 3.0 million tons of hazardous
waste combusted in 1991,
approximately two-thirds of that
amount were combusted at on-site
facilities (i.e., the same facilities at
which the waste was generated).
Combustion at an on-site facility
therefore presents a situation in which
the same facility owners and operators
may have some measure of control over
generation of wastes at its source and its
ultimate disposition. Although close to
400 industries generated wastes
destined for combustion in 1991, much
of the quantity was concentrated in a
few sectors. As a companion to this
proposed rule, EPA is focusing its waste
minimization efforts on reducing the
generation and subsequent release to the
environment of the most persistent,
bioaccumulative, and toxic constituents
in hazardous wastes (i.e., metals,
halogenated organics).
Analysis of waste minimization
potential suggests that generators
currently burning wastes may have a
number of options for eliminating or
reducing these wastes. We believe that
roughly 15 percent of all combusted
wastes may be amenable to waste
minimization. Three waste generating
processes appear to have the most
potential in terms of tonnage reduction:
(1) Solvent and product recovery/
distillation procedures, primarily in the
organic chemicals industry, (2) product
processing wastes, and (3) process waste
removal and cleaning. In addition,
preliminary analyses of Toxics Release
Inventory and hazardous waste stream
data indicate that over 3 million pounds
of hazardous metals are contained in
waste streams being combusted. The top
5 ranking metals (with respect to health
risk considering persistence,
bioaccumulation, and toxicity) are
mercury, cadmium, lead, copper, and
selenium. Additional analyses are
underway to identify the industry
sectors and production processes that
are chief sources of these and other high
priority hazardous constituents.2
In today's rule, EPA is soliciting
comment on two options to promote the
use of pollution prevention/waste
minimization measures as methods for
helping meet MACT standards. These
options (regarding feed stream analysis
and permitting requirements) are
described in Part Five, Section VI,
Subsection D of this preamble. EPA is
also seeking comment on a proposal to
consider, on a case-by-case basis,
extending the compliance deadlines for
this rule by one year if a facility can
show that extra time is needed to
implement pollution prevention/waste
minimization measures in order for the
facility to meet the MACT standards and
that implementation cannot be
practically achieved within the allotted
three-year period after promulgation of
this rule (see Part V, Section 1,
Subsection C).
PART TWO: DEVICES THAT WOULD
BE SUBJECT TO THE PROPOSED
EMISSION STANDARDS
I. Hazardous Waste Incinerators
A. Overview
A hazardous waste incinerator is an
enclosed, controlled flame combustion
2USEPA, Office of Solid Waste, "Setting
Priorities for Hazardous Waste Minimization", July
1994.
-------
17362
Federal Register / Vol. 61, ,No. 77 / Friday, April 19, 1996 / Proposed Rules
device, as defined in 40 CFR 260.10,
and is used to treat primarily organic
and/or aqueous wastes. These devices
may be in situ (fixed), or consist of
mobile units (such as those used for site
remediation and superfund clean-ups)
or may consist of units burning spent or
unusable ammunition and/or chemical
agents that meet the incinerator
definition.
B. Summary of Major Incinerator
Designs
The following is a brief description of
the typical incinerator designs used in
the United States.3
1. Rotary Kilns
Rotary kiln systems typically contain
two incineration chambers: the rotary
kiln and an afterburner. The kiln itself
is a cylindrical refractory-lined steel
shell 10—20 feet in diameter, with a
length-to-diameter ratio of 2 to 10. The
shell is supported by steel trundles that
ride on rollers, allowing the kiln to
rotate around its horizontal axis at a rate
of 1-2 revolutions per minute. Wastes
are fed directly at one end of the kiln
and heated by primary fuels. Waste
continues to heat and bum as it travels
down the inclined kiln. Combustion air
is provided through ports on the face of
the kiln. The kiln typically operates at
50—200 percent excess air and
temperatures of 1600-1800°F. Flue gas
from the kiln is routed to an afterburner
operating at 2000-2500°F and 100-200
percent excess air where unbumt
components of the kiln flue gas are more
completely combusted. Auxiliary fuel
and/or pumpable liquid wastes are
typically used to maintain the
afterburner temperature.
Some rotary kiln incinerators, known
as slagging kilns, operate at high enough
temperatures such that residual
materials leave the kiln in a molten slag
form. The molten residue is then water-
quenched. Another kiln, an ashing kiln,
operates at a lower temperature,
producing a residual ash, which leaves
as a dry material.
2. Liquid Injection Incinerators
A liquid injection incinerator system
consists of an incineration chamber,
waste burner and auxiliary fuel system.
The combustion chamber is a
cylindrical steel shell lined with
refractory materialand mounted
horizontally or vertically. Liquid wastes
are atomized as they are fed into the
combustion chamber through waste
burner nozzles. Typical combustion
3 For a more detailed description of incineration
technology, see "Combustion Emissions Technical
Resource Document (CETRED)", USEPA EPA530-
R-94-014, May 1994.
chamber temperatures are 1300-3000°F
and residence times are from 0.5 to 3
seconds.
3. Fluidized Bed Incinerators
A fluidized bed system is essentially
a vertical cylinder containing a bed of
granular material at the bottom.
Combustion air is introduced at the
bottom of the cylinder and flows up
through the bed material, suspending
the granular particles. Waste and
auxiliary fuels are injected into the bed,
where they mix with combustion air
and burn at temperatures from 840—
1500°F. Further reaction occurs in the
volume above the bed at temperatures
up to 1800°F.
4. Fixed Hearth Incinerators
Fixed hearth incinerators typically
contain two furnace chambers: a
primary and a secondary chamber.
Some designs have two or three step
hearths on which ash and waste are
pushed with rams through the system.
A controlled flow 'underfire'
combustion air is introduced up through
the hearths. The primary chamber
operates in "starved air" mode and the
temperatures are around 1000°F. The
unburnt hydrocarbons reach the
secondary chamber where 140-200
percent excess air is supplied and
temperatures of 1400-2000°F are
achieved for more complete
combustion.
C. Number of Incinerator Facilities
Currently, 162 permitted or interim
status incinerator facilities, having 190
units, are in operation in-the U.S.
Another 26 facilities are proposed4 (i.e.,
new facilities under construction or
permitting). Of the above 162 facilities,
21 facilities are commercial facilities
that burn about 700,000 tons of
hazardous waste annually. The
remaining 141 are on-site or captive
facilities and burn about 800,000 tons of
waste annually.
D. Typical Emission Control Devices for
Incinerators
Incinerators are equipped with a wide
variety of air pollution control devices
(APCDs), which range from no control
(for devices burning low ash and low
chlorine wastes) to sophisticated state-
of-the-art units providing controller
several pollutants. Hot flue gases from
the incinerators are cooled and cleaned*
of the air pollutants before they exit the
stack. Cooling is mostly done by water
quenching, wherein atomized water is
sprayed directly into the hot gases. The
4DSEPA "List of hazardous waste incinerators,"
November 1994.
cooled gases are passed through various
pollution control devices to control PM,
metals and organic emissions to desired
or required levels. Most incinerators use
wet APCDs to scrub acid emissions (3
facilities use dry scrubbers). Typical
APCDs used include packed towers,
spray dryers, or dry scrubbers for acid
gas (e.g., HC1, C12) control, and venturi-
scrubbers, wet or dry electrostatic
precipitators (ESPs) or fabric filters for
particulate control.
Activated carbon injection for
controlling dioxin and mercury is being
used at only one incinerator. Newer
APC technologies (such as catalytic
oxidizers and dioxin/furan inhibitors)
have recently emerged, but have not
been used on any full scale facilities in
the U.S. For detailed description of
APCDs, see Appendix A of
"Combustion Emissions Technical
Resource Document (CETRED)," US
EPA Document #EPA530-R-94-014,
May 1994.
II. Hazardous Waste-Burning Cement
Kilns
A. Overview of Cement Manufacturing
Cement refers to the commodities that
are produced by heating mixtures of
limestone and other minerals or
additives at high temperature in a rotary
kiln, followed by cooling, grinding-, and
finish mixing. This is the manner in
which the vast majority of
commercially-important cementitious
materials are produced in the United
States. Cements are used to chemically
bind different materials together. The
most commonly produced cement type
is "Portland" cement, though other
standard cement types are also
produced on a limited basis (e.g.,
sulfate-resisting, high-early-strength,
masonry, waterproofed). Portland
cement is a hydraulic cement, meaning
that it sets and hardens by chemical
interaction with water. When combined
with sand, gravel, water, and other
materials, Portland cement forms
concrete, one of the most widely used
building and construction materials in
the world. Cement produced and sold in
the U.S. must meet specifications
established by the American Society for
Testing and Materials (ASTM). Each
type requires specific additives or
changes in the proportions of the raw
material mix to make products for
specific applications.
B. Summary of Major Design and
Operating Features of Cement Kilns
Cement kilns are horizontally
inclined rotating cylinders, refractory-
brick lined, and internally-fired, that
calcine a blend of raw materials
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17363
containing calcium (typically
limestone), silica and alumina (typically
clay, shale, slate, and/or sand), and iron
(typically steel mill scale or iron ore) to
produce Portland cement. Generally,
there is a wet process and a dry process
for producing cement. In the wet
process, the limestone and shale are
ground up, wetted and fed into the kiln
as a slurry. In the dry process, raw
materials are ground dry and fed into
the kiln dry. Wet process kilns are
typically longer than dry process kilns
in order to facilitate water evaporation
from the slurried raw material. Wet
kilns can be more than 450 feet in
length. Dry kilns are more thermally
efficient and frequently use preheaters
or precalciners to begin the calcining
process (i.e., the essential function of
driving COj from raw materials) before
the raw materials are fed into the kiln.
Combustion gases and raw materials
move in a counterflow direction, with
respect to each other, inside a cement
kiln. The kiln is inclined, and raw
materials are fed into the upper end
(i.e., the "cold" end) while fuels are
normally fired into the lower end (i.e.,
the "hot" 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
eventually begin to soften and fuse at
temperatures between 2,250 and 2,700
"F to form the clinker product. Clinker
is then cooled, ground, and mixed with
other materials, such as gypsum, to form
cement.
Combustion gases leaving the kiln
typically contain from 6 to 30 percent of
the free solids as dust, which are often
recycled to the kiln feed system, though
the extent of recycling varies greatly
among cement kilns.
Dry kilns with a preheater (PH) or
precalciner (PC) often use a by-pass duct
to remove from 5 to 30 percent of the
kiln off-gases from the main duct. The
by-pass gas is passed through a separate
air pollution control system to remove
particulate matter. Collected by-pass
dust is not reintroduced into the kiln
system to avoid a build-up of metal salts
that can affect product quality.
Some cement kilns burn hazardous ,
waste-derived fuels to replace from 25
to 100 percent of normal fossil fuels
(e.g., coal). Most kilns burn liquid waste
fuels but several also burn bulk solids
and small (e.g., six gallon) containers of
viscous or solid hazardous waste fuels.
Containers are introduced either at the
upper, raw material end of the kiln or
at the midpoint of the kiln. EPA has also
found that hazardous waste-fired
precalciners can still be considered part
of the cement kiln and, thus, would be
part of an industrial furnace (per the
definition in 40 CFR 260.10). See 56 PR
at 7184-85 (February 21,1991). This
finding is codified at
§ 266.103(a)(5)(I)(c). This is the only
time (and the only rulemaking) in which
the Agency found that a device not
enumerated in the list of industrial
furnaces in § 260.10 can be considered
part of the industrial furnace when it
burns hazardous wastes separate from
those burned in the main combustion
device.
C. Number of Facilities
The Agency has emissions data from
26 facilities representing 49 cement
kilns in the U.S. It should be noted that
some facilities no longer burn or process
hazardous waste since they were
required to certify compliance with the
BIF regulations in August 1992.
Of the hazardous waste-burning kilns
for which we have emissions data, 14
facilities use a wet process, 5 facilities
use a dry process, and the remaining 7
facilities employ either preheaters or
preheater/precalciners in the cement
manufacturing process.
D. Emissions Control Devices
All hazardous waste-burning cement
kilns either use fabric filters (baghouses)
or electrostatic precipitators (ESPs) as
air pollution control devices. ESPs have
traditionally been employed in the
cement industry and are currently used
at 17 of the facilities. Nine facilities use
fabric filters. A detailed description of
these and other air pollution control
devices is contained in the technical
support document.5
III. Hazardous Waste-Burning
Lightweight Aggregate Kilns
A. Overview of Lightweight Aggregate
Kilns (LWAKs)
The term lightweight aggregate refers
to a wide variety of raw materials (such
as clay, shale, or slate) which after
thermal processing can be combined
with cement to form concrete products.
Lightweight aggregate concrete is
produced either for structural purposes
or for thermal insulation purposes. A
lightweight aggregate plant is typically
composed of a quarry, a raw material
preparation area, a kiln, a cooler, and a
product storage area. The material is
taken from the quarry to the raw
material preparation area and from there
is fed into the rotary kiln.
s USEPA, "Draft Technical Support Document for
HWC MACT Standards, Volume I: Description of
Source Categories", February 1996.
B. Major Design and Operating Features
A rotary kiln consists of a long steel
cylinder, lined internally with refractory
bricks, which is capable of rotating
about its axis and is inclined at an angle
of about 5 degrees to the horizontal. The
length of the kiln depends in part upon
the composition of the raw material to
be processed but is usually 30 to 60
meters. The prepared raw material is fed
into the kiln at the higher end, while
firing takes place at the lower end. The
dry raw material fed into the kiln is
initially preheated by hot combustion
gases. Once the material is preheated, it
passes into a second furnace zone where
it melts to a semiplastic state and begins
to generate gases which serve as the
bloating or expanding agent. In this
zone, specific compounds begin to
decompose and form gases such as SO2,
CO2, SOs, and Oa that eventually trigger
the desired bloating action within the
material. As temperatures reach their
maximum (approximately 2100°F), the
semiplastic 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. The product exits the kiln
and enters a section of the process
where it is cooled with cold air and then
conveyed to the discharge.
Kiln operating parameters such as
flame temperature, excess air, feed size,
material flow, and speed of rotation vary
from plant to plant and are determined
by the characteristics of the raw
material. Maximum temperature in the
rotary kiln varies from 2050 °F to
2300 °F, depending on the type of raw
material being processed and its
moisture content. Exit temperatures may
range from 300 °F to 1200 °F, again
depending on the raw material and on
the kiln's internal design.
Approximately 80 to 100 percent excess
air is forced into the kiln to aid in
expanding the raw material.
C. Number of Facilities
EPA has identified 36 lightweight
aggregate kiln locations in the United
States. Of these, EPA has identified
seven facilities that are currently
burning hazardous waste in a total of 15
kilns.
D. Air Pollution Control Devices
Lightweight aggregate kilns use one or
a combination of air pollution control
devices, including fabric filters, venturi
scrubbers, spray dryers, cyclones and
wet scrubbers. All of the facilities utilize
fabric filters as the main type of
emissions control, although one facility
uses a spray dryer, venturi scrubber and
-------
17364 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
wet scrubber in addition to a fabric
filter. For detailed descriptions of these
and other air pollution control devices,
please see Appendix A of the draft EPA
document Combustion Emissions
Technical Resource Document
(CETRED).^
PART THREE: DECISION PROCESS
FOR SETTING NATIONAL EMISSION
STANDARDS FOR HAZARDOUS AIR
POLLUTANTS (NESHAPs)
I. Source of Authority for NESHAP
Development
The 1990 Amendments to the Clean
Air Act significantly revised the
requirements for controlling emissions
of hazardous air pollutants. EPA is now
required to develop a list7 of categories
of major and area sources8 of the
hazardous air pollutants (HAPs)
enumerated in section 112 and to
develop technology-based performance
standards for such sources over
specified time periods. See Clean Air
Act (the Act or CAA) §§ 112(c) and
112(d). Section 112 of the Act replaces
the previous system of pollutant-by-
pollutant health-based regulation that
proved ineffective at controlling the
high volumes, concentrations, and
threats to human health and the
environment posed by HAPs in air
emissions. See generally S. Rep. No.
228,101st Cong. 1st Sess. 128-32
(1990).
Section 112(f) also requires the
Agency to report to Congress by the end
of 1996 on estimated risk remaining
after imposition of technology-based
standards and to make
recommendations as to legislation to
address such risk. CAA § 112(f)(l). If
Congress does not act on the
recommendation, then EPA must
address any significant remaining
residual risks posed by sources subject
to the section 112(d) technology-based
standards within 8 years after
promulgation of these standards. See
§ 112(f)(2). The Agency is required to
impose additional controls if such
controls are needed to protect public
health with an ample margin of safety,
or to prevent adverse environmental
effects. Id. In addition, if the
6USEPA, "Draft Combustion Emission Technical
Resource Document (CETRED)", EPA 530-R-94-
014, May 1994.
'The Agency published an initial list of
categories of major and area sources of HAPs on
July 16,1992. See 57 FR 31576.
8 See Part Three, Section in of today's proposal
for a discussion of major and area sources.
Generally, a major source is a stationary source that
emits, or has the potential to emit considering
controls, 10 tons per year of a HAP or 25 tons per
year of a combination of HAPs. CAA § 112(a)(l). An
area source is generally a stationary source that is
not a major source. Id. § 112(a)(2).
technology-based standards for
carcinogens do not reduce the lifetime
excess cancer risk for the most exposed
individual to less than one in a million
(1x10 ~6), then the Agency must
promulgate additional standards. See
§112(f)(2)(A).
II. Procedures and Criteria for
Development of NESHAPs
NESHAPs are developed in order to
control HAP emissions from both new
and existing sources according to the
statutory directives set out in § 112. The
statute requires a NESHAP to reflect the
maximum degree of reduction of HAP
emissions that is achievable taking into
consideration the cost of achieving the
emission reduction, any non-air quality
health and environmental impacts, and
energy requirements. § 112(d)(2). In
regulatory parlance, these are often
referred to as maximum achievable
control technology (or MACT)
standards.
The Clean Air Act establishes
minimum levels, usually referred to as
MACT floors, for the emission
standards. Section 112(d)(3) requires
that MACT floors be determined as
follows: for existing sources in a
category or sub-category with 30 or
more sources, the MACT floor cannot be
less stringent than the "average
emission limitation achieved by the best
performing 12 percent of the existing
sources * * *"; for existing sources in
a category or sub-category with less than
30 sources, then the MACT floor cannot
be less stringent than the "average
emission limitation achieved by the best
performing 5 sources * * *"; for new
sources, the MACT floor cannot be "less
stringent than the emission control that
is achieved by the best controlled
similar source * * *". See §112(d)(3)
(A) and (B).
EPA must, of course, consider in all
cases whether to develop standards that
are more stringent than the floor
("beyond the floor" standards). To do
so, however, EPA must consider the
enumerated statutory criteria such as
cost, energy, and non-air environmental
implications.
Emission reductions may be
accomplished 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,
substitution 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
emissions point; (4) design, equipment,
work practice, or operational standards
(including requirements for operator
training or certification); or (5) any
combination of the above. See
§112(d)(2).
Application of techniques (1) and (2)
of the previous paragraph are consistent
with the definitions of pollution
prevention under the Pollution
Prevention Act and the definition of
waste minimization under RCRA/
HSWA. These terms have particular
applicability in the discussion of
pollution prevention/waste
minimization options presented in the
permitting and compliance sections,of
today's proposal.
To develop a NESHAP, the EPA
compiles available information and in
some cases collects additional
information about the industry,
including information on emission
source quantities, types and
characteristics of HAPs, pollution
control technologies, data from HAP
emissions tests (e.g., compliance tests,
trial burn tests) at controlled and
uncontrolled facilities, and information
on the costs and other energy and
environmental impacts of emission
control techniques. EPA uses this
information in analyzing and
developing possible regulatory
approaches. EPA, of course, does not
always have or collect the same amount
of information per industry, but rather
bases the standard on information
practically available.
Although NESHAPs are normally
structured in terms of numerical
emission limits—the preferred means of
establishing standards—alternative ;
approaches are sometimes necessary
and appropriate. In some cases, for
example, physically measuring
emissions from a source may be
impossible, or at least impractical,
because of technological and economic
limitations. Section 112(h) authorizes
the Administrator to promulgate a
design, equipment, work practice, or
operational standard, or a combination
thereof, in those cases where it is not
feasible to prescribe or enforce an
emissions standard.
EPA is required to develop emission
standards based on performance of
maximum achievable control
technology for categories or sub-
categories of major sources of hazardous
air pollutants. § 112(d)(l). As explained
more fully in the following section, a
major source emits, or has the potential
to emit considering controls, either 10
tons per year of any hazardous air
pollutant or 25 tons or more of any
combination of those pollutants.
§ 112(a)(l). EPA also can establish lower
thresholds where appropriate. Id. EPA
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17365
may in addition require sources
emitting particularly dangerous
hazardous air pollutants (such as
particular chlorinated dioxins and
furans) to be regulated under the MACT
standards for major sources. § 112(c)(6).
Area sources are any source which is
not a major source. Such sources must
be regulated by technology-based
standards if they are listed, pursuant to
§ 112(c)(3), based on the Agency's
finding that these sources (individually
or in the aggregate) present a threat of
adverse effects to human health or the
environment warranting regulation.
After such a determination, the Agency
has a further choice as to require
technology-based standards based on
MACT or on generally achievable
control technology (GACT). § 112(d)(5).
In this rulemaking, EPA is proceeding
pursuant to § 112(c)(6) (i.e., imposing
MACT controls on area sources),
because these hazardous waste
combustion units emit a number of the
HAPs singled out in that provision,
including the enumerated dioxins and
furans, mercury, and polycyclic organic
matter. (See discussion below.)
III. List of Categories of Major and Area
Sources
A. Clean Air Act Requirements
As just discussed, Section 112 of the
CAA requires that the EPA promulgate
regulations requiring the control of
hazardous air pollutants emissions
associated with categories or
subcategories of major and area sources.
These source categories and
subcategories are to be listed pursuant
to § 112(c)(l). EPA published an initial
list of 174 categories of such major and
area sources in the Federal Register on
July 16,1992 (57 FR 31576).
B. Hazardous Waste Incinerators
"Hazardous waste incinerators" is one
'of the 174 categories of sources listed.
The category consists of commercial and
on-site (including captive) incinerating
facilities. The listing was based on the
Administrator's determination that at
least one hazardous waste incinerator
may reasonably be anticipated to emit
several of the 189 listed HAPs in
quantities sufficient to designate them
as major sources. EPA used two
emission rate values to evaluate the
available hazardous waste incinerator
emissions data: the maximum emission
rate measured during the compliance
test, and the average emission rate. The
data indicate that approximately 30
percent of the facilities meet the major
source criteria when using the
maximum emissions rate value. When
using the average emissions rate value
approximately 15 percent of facilities
meet the major source criteria.9 Those
facilities meeting the major source
criteria do so for HC1 and C12 emissions,
and one facility is also a major source
for antimony emissions.
It should be noted that a major source ,
and boundary for considering whether a
source is a major includes all potential
emission points of HAPs at that
contiguous facility, including storage
tanks, equipment leaks, and other
hazardous waste handling facilities. The
above calculations for incinerators on
whether a source is a major source
under § 112 do not reflect these
potential emission points.
Notwithstanding the fact that most
HW incinerators are not likely to meet
the HAP emission thresholds for major
sources, the Agency is proposing to
subject all HWCs to regulation under
MACT as major sources, under the
authority of § 112(c)(6). See Section IV
below.
C. Cement Kilns
Another of the 174 categories of major
and area sources of HAPs is Portland
Cement Manufacturing (cement kilns).
In evaluating the emissions data for the
hazardous waste-burning cement kilns,
85 percent of the cement kilns were
determined to meet the major source
criteria when using the maximum
emission rate value. Using the average
emission rate value, just over 80 percent
of the hazardous waste-burning cement
kilns meet the major source criteria.10
Those facilities meeting the major
source criteria do so for HC1 and Cla
emissions, and one facility is also a
major source for organic emissions. It
should be noted that the calculation on
whether a cement kiln is a major source
did not include potential emission
points of HAPs at that contiguous
facility.
Notwithstanding the fact that some
hazardous waste-burning cement kilns
may not meet the definition of major
source, the Agency is proposing to
subject all HWCs to regulation under
MACT, as major sources, under the
authority of § 112(c)(6). See Section IV
below.
D. Lightweight Aggregate Kilns
Section 112(c)(5) authorizes EPA to
amend the source category list at any
time to add categories or subcategories
that meet the listing criteria. EPA is
proposing to exercise that authority by
adding HW-burning lightweight
aggregate kilns to the list of source
categories.
In analyzing the emissions data, EPA
found that all hazardous waste-burning
LWAKs met the major source criteria for
two HAPs, HC1 and C12, using either the
average or maximum emission rate
value.'' It should be noted that the
calculation on whether a LWAK is a
major source did not include potential,
emission points of HAPs at that
contiguous facility. EPA is therefore
proposing today the addition of
hazardous waste-burning LWAKs as a
source category in accordance with
section 112(c)(5) of the Act. In addition,
as discussed below, even if a LWAK
would otherwise be an area source, EPA
is proposing to subject it to the same
NESHAPS as major LWAK sources.
IV. Proposal To Subject Area Sources to
the NESHAPs Under Authority of
Section 112(c)(6)
EPA is today proposing to subject all
hazardous waste incinerators, hazardous
waste-burning cement kilns, and
hazardous waste-burning lightweight
aggregate kilns (i.e., both area and major
sources) to regulation as major sources
pursuant to CAA § 112(c)(6). That
provision states that, by November 15,
2000, EPA must list and promulgate
§ 112 (d)(2) or (d)(4) standards (i.e.,
standards reflecting MACT) for
categories (and subcategories) of sources
emitting specific pollutants, including
the following HAPs emitted by HWCs:
polycyclic organic matter, mercury,
2,3,7,8-tetrachlorodibenzofuran, and
2,3,7,8-tetrachlorodibenzo-p-dioxin.
(Although the Agency has not prepared
the list, it is the Agency's intention to
include hazardous waste combustors.)
EPA must assure that sources
accounting for not less than 90 percent
of the aggregate emissions of each
enumerated pollutant are subject to
MACT standards.
The chief practical effect of invoking
§ 112(c)(6) for this rulemaking is to
subject area sources that emit 112(c)(6)
pollutants to the same MACT standards
as major sources, rather than to the
potentially less stringent 112(d)(5) or
"GACT" ("generally achievable control
technology") standards.12 Today's
proposal constitutes one of many EPA
actions to assure that sources
accounting for at least 90 percent of
'For further details see USEPA, "Draft Technical
Support Document for HWC MACT Standards,
Volume I: Description of Source Categories",
February 1996.
"Ibid.
12 EPA also solicits comment on an alternative
reading of § 112(c)(6), whereby the provision would
require MACT control for the enumerated
pollutants but not necessarily for other HAPs
emitted by the source, which HAPs are not
enumerated in § 112(c)(6).
-------
17366
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
emissions of § 112(c)(6) pollutants are
subject to MACT standards.
Although § 112(c)(6) requires the
Agency to regulate source categories
that emit not less than 90 percent of the
aggregate emissions of the high priority
HAPs, the Agency will use its discretion
to avoid regulating area source
categories with trivial aggregate
emissions of specific § 112(c)(6) HAPs.
However, as an example of the
emissions that are possible from the
HWC source categories, it is estimated
that HWCs presently emit in aggregate
11.1 tons of mercury per year. Of this
quantity, 4.6 tons per year can be
attributed to hazardous waste
incinerators and 6.5 tons per year to
hazardous waste-burning cement and
lightweight aggregate kilns. Also, it is
estimated that HWCs presently emit in
aggregate 122 pounds of dioxins/furans
(or 2.15 pounds TEQ) per year. Of this
quantity, 9 pounds (or 0.2 pounds TEQ)
per year can be attributed to hazardous
waste incinerators and 113 pounds (or
1.95 pounds TEQ) per year to hazardous
waste-burning cement and lightweight
aggregate kilns. To show an example of
how today's proposal constitutes an
action to assure that sources accounting
for at least 90 percent of emissions of
§ 112(c)(6) pollutants are subject to
MACT standards, the document
Estimating Exposure to Dioxin-Like
Compounds, Vol. II: Properties, Sources,
Occurrence and Background Exposures
(EPA, 1994) estimates (on p. 29) that
national emissions of dioxins and furans
(D/F) total 4.18 pounds TEQ per year.
Based on this estimation, HWCs account
for 51 percent of the annual national
emissions of D/F. (Consequently, EPA
expects these source categories to be
included in the list of sources to be
controlled to achieve the requisite 90
percent reduction in aggregate
emissions of section 112(c)(6)
pollutants.)
Congress singled out the HAPs
enumerated in § 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.). S. Rep.
No. 228,101st Cong. 1st Sess., pp. 155,
166. These pollutants have exhibited
special potential to bioaccumulate,
causing pervasive environmental harm
in biota (and, ultimately, human health
risks). Id. Indeed, as discussed later, the
data appear to show that much of the
human health risk from emissions of
these HAPs from HWCs comes from
these indirect exposure pathways. Id. at
p. 166. Congress' express intention was
to assure that sources emitting
significant quantities of § 112(c)(6)
pollutants received a stricter level of
control. Id.
V. Selection of MACT Floor for Existing
Sources
The starting point in developing
MACT standards is determining floor
levels, i.e. the minimum (least stringent)
level at which the standard can be set.
All of the hazardous waste
combustion units subject to this
proposed rule are already subject to
RCRA regulation under 40 CFR Parts
264, 265, or 266. As a result, the Agency
has a substantial amount of data
reflecting performance of these devices.
These data consist largely of trial burn
data for hazardous waste incinerators
and data from certifications of
compliance for hazardous waste-
burning cement kilns and LWAKs
obtained pursuant to 266.103(c). These
data consist of at least three runs for any
given test condition.
In using these "short term" test data
to establish a MACT floor, the Agency
has developed an approach that ensures
the standards are achievable, i.e. reflect
the performance over time of properly
designed and operated air pollution
control devices (or operating practices)
taking into account intrinsic operating
variability.
In addition, the Agency notes that the
.. floor calculations were performed on
individual HAPs or, in the case of
metals, in two groups of HAPs that
behave similarly (i.e., separate floor
levels for each hazardous air pollutant
or group of metal pollutants). However,
for HAPs that are controlled by the same
type of air pollution control device
(APCD), EPA has ensured that all HAP
floors are simultaneously achievable by
identifying the APCD and APCD
treatment train that can be used to meet
all floor levels. The ultimate floor levels
thus derived can be achieved using the
identified technology. This approach is
consistent with methods used by EPA in
other rules to.calculate MACT
requirements where the HAP species
present must be treated by a treatment
train. See, e.g., MACT Rules for
Secondary Lead Smelters. 60 FR 32589
(June 23,1995).
The Agency is not, however, treating
hazardous waste-burning incinerators,
cement kilns, and LWAKs as a single
source category for purposes of
developing the MACT floor (or for any
other purpose). The Agency's initial
view is that there are technical
differences in performance for particular
HAPs among the three source categories,
and therefore that the technology-based
floors must reflect these operating
differences.
A. Proposed Approach: Combined
Technology-Statistical Approach
This analysis first identified the best
performing control technology(ies) for
each source category (i.e., incinerators,
cement kilns, 'and lightweight aggregate
kilns) and each HAP of concern by
arraying from lowest to highest all the
particular HAP emissions data from
existing units within the source category
by test condition averages. These
technologies comprise MACT floor. In
cases where a source had emissions data
for a HAP from several different test
conditions of a compliance test, the
Agency arrayed each test condition
separately. The Agency then identified
the emission control technology or
technologies (and normalized feedrate
of metals and chlorine in hazardous
waste) used by sources with emissions
levels at or below the level emitted by
the median of the best performing 12
percent of sources. The sources are
termed "the best performing 6 percent"
of the sources, or "MACT pool", and the
controls they use comprise MACT floor.
The next step was to identify an
emissions level that MACT floor control
could achieve. Thus, emissions data
from all sources (in the source category)
that use MACT floor control were
arrayed in ascending order by average
emissions. [This is referred to as the
"expanded MACT pool" or "expanded
universe".] The Agency evaluated the
control technologies used by the
additional sources within the
"expanded universe" as available data
allowed to ensure that they were in fact
equivalent in design to MACT floor. The
Agency then selected the test condition
in the expanded MACT pool with the
highest mean emissions to identify the
emission level that MACT floor could
achieve.
Because the emissions database was
comprised of "short-term" test data, the
Agency used a statistical approach to
identify an emission level that MACT
floor could achieve routinely. The
Agency then identified the test
condition in the expanded MACT pool
with the highest mean emissions to
statistically calculate a "design level"
and a floor standard. The design level
was calculated as the log mean of the
emissions for the test condition. The
standard was calculated as a level that
a source (that is designed and operated
to routinely meet the design level) could
meet 99 percent of the time if it has the
average within-test-condition emissions
variability of the expanded MACT pool.
Although the Agency evaluated 90th
and 95th percentile limits, the 99th
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17367
percentile limit was chosen to: (1) More
accurately reflect the variability that
could be present in emissions data, and
(2) appropriately characterize this
variability in light of the consequence of
failing to achieve the emissions
standards. Additional information on
how MACT floor levels were identified
is provided in the "Draft Technical
Support Document for HWC MACT
Standards, Volume HI: Selection of
Proposed MACT Standards and
Technologies".
In accounting for operating
variability, the Agency solicits comment
on whether it may have
overcompensated so that the identified
floor levels are unduly lenient. The test
data on which the proposal is based to
some extent reflect worst-case
performance conditions because RCRA
sources try to obtain maximum
operating flexibility by conducting test
burns at extreme operating conditions.
For example, many sources spike wastes
with excess metals and chlorine during
compliance testing. In addition, sources
operate their emissions control devices
under low efficiency conditions (while
still meeting emission standards) to
ensure lenient operating limits. It thus
may be that the Agency's emissions
database is so inflated that separate
consideration of emissions variability
may not be warranted. A floor level
could be the highest mean of the test
conditions in the expanded MACT pool.
The Agency emphasizes that it would
be preferable, for purposes of setting
these MACT standards, to have
operational and emissions data that
better reflect long-term, more routine
day-to-day facility operations from all of
the source categories. We believe that
this type of data would enable the
MACT process to articulate a set of HAP
standards that would not create some of
the issues raised in subsequent sections
of this preamble (such as the most
appropriate resolution of a variability
factor, the optimum approach for
considering the contribution of cement
and lightweight aggregate kiln raw
material feed to HAP emissions, and
better identification among sources that
are now in an expanded MACT pool but
which, with better data, would be
determined not to be employing the
identified floor controls). As noted in
these subsequent sections, the Agency
urges commenters to submit these types
of data.
B, Another Approach Considered but
not Used
Although the Agency believes the
proposed approach reflects a reasonable
interpretation of the statute, there are
other possible interpretations. One of
these interpretations, termed the "12
percent approach", was raised and, in
fact, evaluated during the process
already outlined. This approach is
presented here, along with the results of
the process in Part Four, Section VIII,
for public inspection.
This "12 percent approach" was
evaluated in a like manner to the
Agency's preferred approach just
described. Again, the best performing
control technology(ies) for each source
category and each HAP were identified
by arraying the data by test condition
averages. However, the Agency
identified the technology or
technologies used by the best
performing 12 percent of the sources.
After arraying emissions data from all
facilities in the source category that use
the identified MACT floor
technology(ies) (i.e., the expanded
MACT pool), the Agency selected an
emissions floor level based on the
statistical average of the 12 percent
MACT pool, to which was added the
average within-test condition variability
within the expanded MACT pool. The
emissions floor was then calculated at a
level that a source with average
emissions variability would be expected
to achieve 99 percent of the time. The
approach was not proposed because it
could not be demonstrated that sources
within the expanded MACT pool using
MACT floor controls could achieve the
floor levels. Again, the details of the
statistical methods employed are
presented in the "Draft Technical
Support Document for HWC MACT
Standards, Volume HI: Selection of
Proposed MACT Standards and
Technologies".
C. Identifying Floors as Proposed in
CETRED
The discussion in the Draft
Combustion Emissions Technical
Resource Document (CETRED) (U.S.
EPA, EPA530-R-94-014, May 1994)
presented one methodology for
establishing particulate matter (PM) and
dioxin/furan (D/F) technology-based
emission levels for hazardous waste
combustors (HWCs). The document
presented a procedure for establishing
numerical levels which took into
account the natural variability that was
present in the Agency's PM and D/F
emissions data. EPA received numerous
comments on the document.
The approaches outlined in CETRED
were an initial and preliminary attempt
to apply the process by which the
NESHAPs are to be established for the
existing types of hazardous waste
combustors. The approaches in CETRED
focused solely on the performance of
MACT and how to establish the "floor"
emission level under the MACT process.
In CETRED, determination of the
MACT floor involved: (1) screening
unrepresentative data; (2) ranking all
HWC sources based on the data average,
considering variability; (3) identifying
the top 12 percent of sources as the
MACT pool; and (4) statistically
evaluating the MACT pool to set the
MACT floor. These elements and
considerations are described in further
detail in CETRED and the "Draft
Technical Support Document for HWC
MACT Standards, Volume III: Selection
of Proposed MACT Standards and
Technologies". The Agency specifically
indicated the preliminary nature of the
CETRED approaches and, in light of
further deliberations and comments
received, has considered and adopted
other approaches for this proposal. The
comments received are found in the
docket.
In considering the use of a purely
statistical approach to setting MACT
floors, the Agency recognized that
whether sources could actually achieve
a statistically-derived MACT floor level
on a regular basis was significant in
determining whether a purely statistical
approach could be appropriate or not.
The Agency encountered difficulties in
identifying an appropriate purely
statistical model for the combined
source category (HW incinerators, HW-
burning cement kilns, and HW-burning
lightweight aggregate kilns) emissions
database. Consequently, the Agency
abandoned a purely statistical approach
and examined an approach—referred to
here as the "technology approach"—
that used demonstrated technological
capabilities as a key factor in selecting
MACT floor levels.
D. Establishing Floors One HAP or HAP
Group at a Time
EPA believes it is permissible to
establish MACT floors separately for
individual HAPs or group of HAPs that
behave the same from a technical
standpoint (i.e., based on separate
MACT pools and floor controls),
provided the various MACT floors are
simultaneously achievable. As set out
below, Congress has not spoken to this
precise issue. An interpretation that
allows this approach is consistent with
statutory goals and policies, as well as
established EPA practice in developing
MACT standards.
As described earlier, Congress
specified in section 112(d)(3) the
minimum level of emission reduction
that could satisfy the requirement to
adopt MACT. For new sources, this
floor level is to be "the emission control
that is achieved in practice by the best
-------
17368 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
controlled similar source". For existing
sources, the floor level is to be "the
average emission limitation achieved by
the best performing 12 percent of the
existing sources" for categories and
subcategories with 30 or more sources,
or "the average emission limitation
achieved by the best performing 5
sources" for categories and
subcategories with fewer than 30
sources. An "emission limitation" is "a
requirement * * * which limits the
quantity, rate, or concentration of
emissions of air pollutants" (section 302
(k)) (although the extent, if any, the
section 302 definitions need to apply to
the terms used in section 112 is not
clear).
This language does not expressly
address whether floor levels can be
established HAP-by-HAP. The existing
source MACT floor achieved by the
average of the best performing 12
percent can reasonably be read as
referring to the source as a whole or
performance as'to a particular HAP. The
statutory definition of "emission
limitation" (assuming it applies)
likewise is ambiguous, since
"requirements limiting quantity, rate, or
concentration of pollutants" could
apply to particular HAPs or all HAPs.
The reference in the new source MACT
floor to "emission control achieved by
the best controlled similar source" can
mean emission control as to a particular
HAP or achieved by a source as a whole.
Here, Congress has not spoken to the
precise question at issue, and the
Agency's interpretation effectuates
statutory goals and policies in a
reasonable manner. See Chevron v.
NRDC, 467 U.S. 837 (1984) (indicating
that such interpretations must be
upheld). The central purpose of the
amended air toxics provisions was to
apply strict technology-based emission
controls on HAPs. See, e.g., H. Rep. No.
952,101st Cong. 2d sess. 338. The
floor's specific purpose was to assure
that consideration of economic and
other impacts not be used to "gut the
standards". While costs are by no means
irrelevant, they should by no means be
the determining factors. There needs to
be a minimum degree of control in
relation to the control technologies that
have already been attained by the best
existing sources. Legislative History of
the Clean Air Act Vol. II at 2897
(statement of Rep. Collins).
Furthermore, an alternative
interpretation would tend to result in
least common denominator floors where
multiple HAPs are emitted, whereby
floors would no longer be reflecting
performance of the best performing
sources. For example, if the best
performing 12 percent of facilities for
HAP metals did not control organics as
well as a different 12 percent of
facilities, the floor for organics and
metals would end up not reflecting best
performance. Indeed, under this
reading, the floor would be no control,
because no plant is controlling both
types of HAPs.
EPA is convinced that this result is
not compelled by the statutory text, and
does not effectuate the evident statutory
purpose of having floor levels reflect
performance of an average of a group of
best-performing sources. Conversely,
using a HAP-by-HAP approach (or an
approach that groups HAPs based on
technical factors) to identify separate
floors for metals and organics in this
example promotes the stated purpose of
the floor to provide a minimum level of
control reflecting what best performing
existing sources have already
demonstrated an ability to do.
EPA notes, however, that if optimized
performance for different HAPs is not
technologically possible due to
mutually inconsistent control
technologies (for example, metals
performance decreases if organics
reduction is optimized), then this would
have to be taken into account in
establishing a floor (or floors).
(Optimized controls for both types of
HAPS would not be MACT in any case,
since the standards would not be
mutually achievable.) The Senate Report
indicates that in such a circumstance,
EPA is to optimize the part of the
standard providing the most
environmental protection. S. Rep. No.
228,101st Cong. 1st sess! 168. It should
be emphasized, however, that "the fact
that no plant has been shown to be able
to meet all of the limitations does not
demonstrate that all the limitations are
not achievable". Chemical
Manufacturers Association v. EPA, 885
F. 2d at 264 (upholding technology-
based standards based on best
performance for each pollutant by
different plants, where at least one plant
met each of the limitations but no single
plant met all of them).
All available data for HWCs indicate
that there is no technical problem
achieving the floor levels for each HAP
or HAP metal group simultaneously,
using the MACT floor technology. In the
case of metals and PM, the
characteristics of the MACT floor
technology associated with the hardest-
to-meet floor (e.g., the fabric filter with
lowest air-to-cloth ratio) would define
the MACT floor technology for purposes
of determining achievability of floors
and for purposes of costing out the
impact of the standards. Existing data
show that approximately 9 percent of
existing hazardous waste incinerators,
approximately 8 percent of hazardous
waste-burning cement kilns, and
approximately 25 percent of hazardous
waste-burning LWAKs are already
achieving the proposed floor standards
for all HAPs.
Finally, EPA notes that the HAP-by- .
HAP or HAP group approach to
establishing MACT floor levels is not
unique to this rule. For example, the
Agency has adopted it for the NESHAP
for the secondary lead source category
(60 FR 32589 (June 23, 1995)) and
proposed the same approach for
municipal waste combustors (59 FR
48198 (September 20,1994)).
As discussed above, EPA has the
authority to establish MACT floors on a
HAP group by HAP group basis and has
done so in this case. In doing so, EPA
will ensure that such floors, taken as a
whole, are reasonably achievable for
facilities subject to the MACT standards.
VI. Selection of Beyond-the-Floor
Levels for Existing Sources
As discussed in Section V above, the
MACT floor defines the minimum level
of emission control for existing sources,
regardless of cost or other
considerations. The process of
. considering emissions levels more
stringent than the MACT floor for
existing sources is called a "beyond-the-
floor" (BTF) analysis and involves
consideration of certain additional
factors, including cost, any non-air
quality health and environmental
impacts and energy requirements,
technologies currently in use within
these industry sectors, and also other
more efficient and appropriate
technologies that have been
demonstrated and are available on the
market (e.g., carbon bed for dioxin/furan
control).
Because there are virtually unlimited
BTF emissions levels that the Agency
could consider, the Agency used several
criteria in this proposal to identify when
to examine a particular beyond-the-floor
emissions level in detail, and also
whether to propose a MACT standard
based on the beyond-the-floor emissions
levels for existing sources.
The primary factor is the cost-
effectiveness of setting MACT standards
based upon a more efficient technology
than the MACT floor technology(ies). If
the Agency's economic analysis
suggested that BTF levels could be cost-
effectively achieved (particularly if
significant health benefits would result
from a lower emission level), then an
applicable BTF emission level control
technology was identified to achieve
that level. The associated costs were
then weighed along with the other
criteria. Dioxin/furans is an example
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17369
xvhere the Agency considered a BTF
level because a beyond-the-floor
emission level can be achieved in a cost-
effective manner, achieving, in addition,
significant non-air quality
environmental benefits.
VII. Selection of MACT for New
Sources
For new sources, the standards for a
source category (or sub-category) cannot
be less stringent than the emission
control that is achieved in practice by
the best-controlled similar source. See
§ 112(d)(3). The following discussion
summarizes the methodology used by
the Agency in developing today's
proposed emissions standards for new
HWC sources.
The approach used to identify MACT
for new sources parallels in most ways
the approach used to determine the
MACT floor for existing sources. For
each HAP, the Agency identified the
technology associated with the single
best performing source (for each source
category). The Agency used this best
performing technology then looked at
all facilities operating the control
technology, and determined the
achievable emission levels that
represent "the emission control that is
achieved in practice by the best
controlled similar source" by using the
maximum value achieved by properly-
operated technology (adjusted upwards
by a statistically derived variability
factor). For further details, see the
technical background documents13
supporting today's proposal.
Since MACT for new sources is to
reflect optimized achievable
performance and is not necessarily
limited to performance levels currently
achieved, the Agency also considered
several other factors in selecting the
MACT new emissions limit. These
factors included: (1) Comparisons to
other emissions standards which may
indicate that a technology is
demonstrated and its level of
performance (e.g., proposed municipal
xvaste combustors and medical waste
incinerators regulations and the
European Union waste incineration
standards); and (2) test condition
emissions variability.
As mentioned earlier, the Agency
believes that it is appropriate to
compare the proposed emissions
standards for new sources to other
existing or recently proposed standards
applicable to hazardous waste
combustors or similar devices as a type
of "reality check" that we are
developing the most rigorous emissions
limits for new sources based upon the
best technologies available today.
The extracted data and data plots are
presented in the background
document14 located in the docket.
VIII. RCRA Decision Process
It is EPA's intention to eliminate
duplicative or potentially duplicative
regulation wherever possible. In this
section, we discuss: (1) The RCRA
mandate to ensure protection of human
health and the environment and how
that mandate relates to the CAA
technology-based MACT standards; (2)
how, for RCRA purposes, we evaluated
the protectiveness of the proposed
MACT standards; (3) how, for RCRA
purposes, the Agency intends to
continue its policies with respect to site-
specific risk assessments and permitting
so that, in appropriate situations,
additional RCRA permit conditions can
be developed as necessary to protect
human health and the environment; and
(4) how waste minimization
opportunities may be considered at
individual facilities during the
permitting process.
A. RCRA and CAA Mandates To Protect
Human Health and the Environment
The Agency is proposing emission
standards for HWCs under joint
authority of the Clean Air Act
Amendments of 1990 and the Resource
Conservation and Recovery Act (RCRA).
As noted earlier, section 3004(a) of
RCRA requires the Agency to
promulgate standards for hazardous
waste treatment, storage, and disposal
facilities as necessary to protect human
health and the environment. The
standards for incinerators generally rest
on this authority. In addition, § 3004(q)
requires the Agency to promulgate
standards as necessary to protect human
health and the environment specifically
for facilities that burn hazardous waste
fuels (e.g., cement and light-weight
aggregate kilns). Using RCRA authority,
the Agency has historically established
emission (and other) standards for
HWCs that are either entirely risk-based
(e.g., site-specific standards for metals
under the BIF rule), or are technology-
based but determined by a generic risk
assessment to be protective (e.g., the
DRE standard for incinerators and BIFs).
The MACT standards proposed today
implement the technology-based regime
of CAA § 112. There is, however, a
residual risk component to air toxics
"USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume IE: Selection of
Proposed MACT Standards and Technologies",
February 1990.
« USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume ID: Selection of
Proposed MACT Standards and Technologies",
February 1996.
standards. Section 112(f) of the Clean
Air Act requires the Agency to impose,
within eight years after promulgation of
the technology-based standards
promulgated under § 112(d) (i.e., the
authority for today's proposed
standards), additional controls if needed
to protect public health with an ample
margin of safety or to prevent adverse
environmental effect. (Cost, energy, and
other relevant factors must be
considered in determining whether
regulation is appropriate in the case of
environmental effects.)
As noted earlier, EPA's express intent
is to avoid regulatory duplication. RCRA
§ 1006 directs that EPA "integrate all
provisions of [RCRA] for purposes of
administration and enforcement and
* * * avoid duplication, to the
maximum extent possible, with the
appropriate provisions of the Clean Air
Act* * *." The overall thrust of the
proposed rule is to have the CAA
standards supplant independent RCRA
standards wherever possible (i.e., to
have the CAA standards, wherever
possible, also serve to satisfy the RCRA
mandate so that additional RCRA
regulation is unnecessary).
Under RCRA, EPA must promulgate
standards "as may be necessary to
protect human health and the
environment." RCRA § 3004(a) and (q).
Technology-based standards developed
under CAA § 112 do not automatically
satisfy this requirement, but may do so
in fact. See 59 FR at 29776 (June 6,
1994) and 60 FR at 32593 (June 23,
1995) (RCRA regulation of secondary
lead smelter emissions unnecessary at
this time given stringency of
technology-based standard and
pendency of § 112(f) determination). If
the MACT standards, as a factual matter,
are sufficiently protective to also satisfy
the RCRA mandate, then no
independent RCRA standards are
required. Conversely, if MACT
standards are inadequate, the RCRA
authorities would have to be used to fill
the gap.
It should be noted that this RCRA risk
evaluation can inform the MACT
decision process as well. For example,
the RCRA risk evaluations indicate the
potential for significant risk via indirect
pathways from dioxins and furans
originating in today's baseline air
emissions for HWCs. EPA is explicitly
authorized to consider non-air
environmental impacts (such as
exposure to HAPS which, after
emission, enter into the food chain and
are eventually consumed by humans
and other biota) in determining whether
to adopt standards more stringent than
the MACT floor. Thus, EPA can
consider benefits from curbing these
-------
27370
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 /Proposed Rules
indirect exposures as part of its beyond-
the-floor determinations.
As discussed below, the Agency has
conducted an evaluation, for the
purposes of satisfying the RCRA
statutory mandates, of the degree of
protection afforded by the MACT :
standards being proposed today.
However, the Agency's current RCRA
evaluation is not intended to have any
bearing on what we may or may not
determine is necessary in several years
to satisfy the § 112(f) provisions.
B. Evaluation of Protectiveness
To determine whether the MACT
standards are consistent with the
Agency's mandate under RCRA to
establish standards for hazardous waste
management facilities and to issue
permits that are protective of human
health and the environment, the Agency
conducted two types of analyses to
assess the extent to which potential
risks from current hazardous waste
combustion emissions would be
reduced through implementation of
MACT standards.
The first of these analyses was
designed to assess the potential risks to
individuals living near hazardous waste
combustion facilities and to nearby
aquatic ecosystems. The procedures
used in this analysis are discussed in
detail in the background document
contained in the docket for today's
proposal.15 The results are summarized
in Part Four of today's notice,
"Rationale for Selecting Proposed
Standards".
The second analysis of potential risk
reduction was a more qualitative
evaluation of risks at the national level,
for those two constituents (dioxins and
mercury) which the Agency believes
pose significant health risks at the
national level and which are found at
significant concentrations in hazardous
waste combustor emissions. The results
of this analysis are presented in Section
Seven, "Regulatory and Administrative
Requirements", as part of the discussion
of potential costs and benefits required
under Executive Order 12866.
1. Individual Risk Analysis
The Agency assessed potential risks to
individuals from both direct inhalation
of emissions (after dispersion in the
ambient air) and indirect exposure to
emissions through deposition onto soils
and vegetation and subsequent uptake
through the food chain. The analysis
focussed primarily on dioxins and
15 "Risk Assessment Support to the Development
of Technical Standards for Emissions from
Combustion Units Burning Hazardous Wastes:
Background Information Document," February 20,
1996.
related compounds since these have
been of major concern to the Agency
from a risk perspective and because
there is enough information about the
properties of these constituents to allow
for a quantitative analysis. The
individual risk analysis did also include
risks from inhalation of metals,
hydrogen chloride, and chlorine (Ck).
The Agency conducted an evaluation
of risks from metals through indirect
exposure routes. With the exception of
mercury, most of the metals are not
expected to accumulate significantly in
the food chain, and the risks from other
indirect exposure routes (such as
deposition on soil and incidental
ingestion of the soil) are not projected
to be significant, even with conservative
assumptions.
With respect to mercury, the Agency
suspects that there may be significant
individual risks near hazardous waste
combustion facilities, primarily through
deposition, erosion to surface waters,
and accumulation in fish which are then
consumed. However, the current state of
knowledge concerning the behavior of
mercury in the environment does not
allow for a meaningful quantitative risk
assessment of emission sources which is
precise enough to support regulatory
decisions at the national level..
Specifically, there is insufficient
information with respect to speciation
of the mercury into various forms in .
emissions and with respect to the
deposition and cycling of mercury
species in the environment to conduct
a defensible national quantitative
assessment of mercury deposition,
erosion to surface waters, and
bioaccumulation in fish. The Agency
solicits comment and information on
the issue of the risks posed by mercury
emissions from hazardous waste
combustion facilities.
The Agency also considered potential
risks from emissions of non-dioxin •
semi-volatile organics that are products
of incomplete combustion (PICs).
However, the Agency was not able to
conduct an appropriate analysis for
several reasons. First, the limited
emissions data now available to the
Agency on non-dioxin PICs are not
sufficiently reliable to conduct an
adequate assessment of risk. Second,
there is not a universally accepted set of
parameter values for some non-dioxin
PICs with which to assess potential
exposures (e.g., the use of octanol-water
partition coefficients (Kow) to predict
bioaccumulation versus the use of
empirical data and the extent to which
bioaccumulation of compounds such as
phthalates and polycyclic aromatic
hydrocarbons (PAHs) occurs in
domestic animals). The Agency solicits
comment on these issues and, in
particular, requests data on
bioaccumulation of PAHs, phthalates,
and other non-dioxin PICs in farm
animals used for food production and in
other mammals and birds. The Agency
also intends to obtain a better set of data
relating to the non-dioxin PIC emissions
from hazardous waste combustion
facilities.
2. Individual Risks From Dioxins
In order to evaluate potential risks
from dioxins to individuals living near
hazardous waste combustion facilities,
the Agency selected eleven example
facility locations, consisting of areas in
which five actual cement kilns, four
incinerators, and two lightweight
aggregate kilns are located. The example
facility locations represent a variety of
environmental settings and facility
characteristics. The purpose of using
example facilities was to incorporate as
much realism as possible into the
Agency's risk assessment and to reduce
the reliance on hypothetical,
conservative assumptions about either
location or source type characteristics.
Site-specific characteristics considered
in the analysis include meteorological
conditions, topography, and land use as
well as stack height and gas flow rates.
However, the stack gas concentrations
used in the modeling of the example
facilities were derived from national
emissions data. Therefore, while the
example facility analyses are useful for
providing information to evaluate
national standards on a generic basis,
they are not site-specific assessments of
any individual facility and cannot be
regarded as such.
The Agency has identified a number
of indirect exposure pathways which
are most likely to present significant
risks. These include: consumption of
locally-produced meat, eggs, and dairy
products and consumption of fish from
local waterways. Contamination of food
occurs from deposition of toxic
emissions onto plants and soil with
subsequent ingestion by farm animals
or, in the case of fish contamination,
from deposition directly into water
bodies or onto soil and runoff into
surface waters with subsequent uptake
in fish.
In assessing risks to the more highly
exposed individuals, the Agency
assumed that certain segments of the
population subsisted in part on home-
produced foods or fish obtained from
nearby lakes or streams. In addition, the
Agency assumed that these individuals
were exposed in the farming and fishing
areas most affected by the example
facilities' emissions. In its analysis of
the eleven example facilities, the
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17371
Agency attempted to identify the actual
location of farms and water bodies
where subsistence activities might be
expected to occur. For dioxins, the
highest exposures are expected to occur
for individuals whose diets include
significant amounts of home-produced
meat and eggs or locally caught fish.
Individuals likely to have high
exposures include subsistence farmers
that raise beef cattle, dairy cows, or
chickens along with their families as
well as subsistence fishers and
recreational anglers and their families.
In evaluating individual risks, the
Agency projected both "high end" and
"central tendency" estimates of risks to
the individuals of concern in the
analysis. The central tendency estimates
were derived by setting all emission
rates, fate and transport parameters, and
exposure assumptions at central
tendency values, as described in the risk
assessment background document. To
derive high end risk estimates, the
Agency set the emission levels at the
90th percentile of the distribution of
available dioxin concentrations and, for
most exposure scenarios, set one
exposure parameter to a high end value
while keeping all other parameters at
central tendency values. For purposes of
evaluating the protectiveness of the
standards, the Agency used a target risk
level of 10-5 for the high end individual
risk, which is consistent with the
approach taken in the 1991BIF rule.
3. Uncertainties in the Individual
Dioxin Risk Estimates
Much of the information used to
derive the individual risk estimates for
dioxins was taken from the Agency's
draft Dioxin Reassessment documents16
17 to. Those documents discuss in
considerable detail a number of the
uncertainties associated with both the
cancer slope factor (the dose-response
descriptor) and the many parameters
used in the exposure assessment. Some
of these uncertainties are also discussed
in the risk assessment background
document for today's proposal.
In addition, there have oeen a large
number of public comments on the
Dioxin Reassessment, which the Agency
is now considering. If the Agency
decides to revise its assessment of either
the toxicity or exposure associated with
'""Health Assessment Document for 2,3,7,8-
Tetrachlorodibenzo-p-Dioxin (TCDD) and Related
Compounds Volume I and II", Office of Research
and Development, June 1994.
""'Health Assessment Document for 2,3,7,8-
Totrachlorodibenzo-p-Dioxln (TCDD) and Related
Compounds Volume III", Office of Research and
Development, August 1994.
18"EstImating Exposure to Dioxin-Like
Compounds Volumo I, II, and ID", Office of
Research and Development, June 1994.
dioxins prior to the final promulgation
of this rule, those revisions will be
considered in the development of the
final rule.
The Agency is also conducting an '•
external peer review of its risk analysis
supporting today's proposal. The results
of this peer review, which are expected
during the comment period, will be
available in the public record for this
rule and will be considered in
developing the final rule.
4. Qualitative Assessments of National
Risks
While the individual risk assessment
discussed above provides a quantitative
measure of the protectiveness of the
proposed MACT standard, there are
other ways of evaluating potential
impacts of reducing emissions of
hazardous constituents. One approach
taken by the Agency is to describe to the
extent practicable what is known about
the national extent of risks from
constituents such as dioxins and
mercury. To put that information in
context with respect to this rule, the
relative contribution of hazardous waste
combustion to other known air releases
of these constituents to the environment
is then presented. The Agency
recognizes that it is not appropriate to
quantitatively correlate emissions with
risk on a national scale; nevertheless,
this type of information is useful for
qualitatively evaluating the potential
impact of the proposed MACT rule.
C. Use of Site-Specific Risk Assessments
Under RCRA
As part of the Agency's Hazardous
Waste Minimization and Combustion
Strategy, EPA currently has a national
RCRA policy of strongly recommending
to all federal and state RCRA permit
writers that, under the omnibus permit
provisions of RCRA § 3005(c)(3), site-
specific risk assessments be performed
as part of the RCRA permitting process
if necessary to protect human health
and the environment. Regions and
authorized states have been
implementing this national policy since
mid-1993 under the aegis of the
omnibus and other applicable
authorities.
The Combustion Strategy announced
this policy encouraging site-specific risk
assessments as part of the overall effort
to ensure that, under appropriate legal
authorities, all RCRA combustion
permits being issued are sufficiently
protective. Specifically, these site-
specific risk assessments were intended
to address potential concerns about a
suite of hazardous air pollutants, among
them dioxins, furans, metals, and non-
dioxin PICs, during the time it took for
the Agency to upgrade the technical
standards for hazardous waste
incinerators, boilers, and industrial
furnaces. This proposal is the first
rulemaking that the Agency has issued
in the upgrading effort.
The question has arisen as to the
status of the Agency's current policy
with respect to site-specific risk
assessments, particularly with respect to
the HAPs for which standards are being
proposed today as well as for other non-
dioxin PICs. As noted above, the Agency
has conducted a risk evaluation under
RCRA of the degree of protection
afforded by the proposed MACT
standards for the HAPs addressed in
today's rule. However, with respect to
mercury and non-dioxin PICs, the
Agency does not at this time have
sufficient reliable data to be able to
assess, on a national basis, the
magnitude of the risks that can routinely
be expected from burning hazardous
waste in HWCs. Although the Agency
has plans to obtain extensive and
detailed PIC emissions data from
hazardous waste combustors in the
coming months, it may be some time
before the Agency is in a proper
position to make any type of regulatory
and policy judgment about the need, if
any, for additional national standards
for these toxic organics. Indeed, at
several sites, the levels of some non-
dioxin PICs have not previously been
shown to be of concern, at least to the
extent that site-specific testing revealed
their presence and to the extent
evaluated in site-specific risk
assessments.
The Agency is continuing its policy of
recommending that, if necessary to
protect human health and the
environment, site-specific risk
assessments be conducted as part of
RCRA permitting for all hazardous
waste combustors (incinerators, boilers,
and industrial furnaces alike) until
national standards for HAPs of concern
are in place. We expect that, in most
situations prior to actual
implementation of facility measures to
appropriately control the HAPs
addressed in this rule, the EPA regional
and authorized state permitting officials
will find there is a necessity to conduct
site-specific risk assessments prior to
final permit determinations. We also
note that the remaining uncertainties
about the risks from non-dioxin PICs
and mercury would likely bear upon
implementation of the national policy.
However, small on-site facilities are not
likely to present the same level of
potential risk as other facilities. This
industry segment may not warrant site
specific risk assessments with the same
frequency as the large on-site or
-------
17372
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
commercial facilities. Among the factors
that the regions and states should
consider in their evaluation of the
necessity for a site-specific risk
assessment are: (1) The current level of
HAPs being emitted hy a facility,
particularly in comparison to the MACT
standards being proposed and in
comparison to the emissions
assumptions and exposure scenarios
used in the RCRA risk evaluation of the
proposed MACT standards (detailed in
the Background Document); (2) whether
the facility is exceeding the proposed
HAP standards, particularly for dioxins/
furans and mercury, what immediate
measures could be instituted to reduce
those emissions; (3) the scope of waste
minimization efforts at the facility with
respect to the HAPs of concern and the
status of implementation of any facility
waste minimization plan; (4) particular
site-specific considerations such as
proximity to receptors, unique
dispersion patterns, etc.; (5) the PICs
most likely to be found and those most
likely to pose significant risk; (6) the
presence or absence of other sources of
HAPs in sufficient proximity as to exert
a significant influence on interpretation
of a facility-specific risk assessment; (7)
the presence or absence of significant
ecological considerations, including for
example high background levels of a
particular contaminant or proximity of a
particularly sensitive ecological area;
and (8) the volume and types of wastes
being burned. This list is by no means
exhaustive, but is meant only to suggest
significant factors that have thus far
been identified. Others may be equally
or more important.
Continuation of the site-specific risk
assessment policy rests primarily on the
RCRA requirement to ensure that all
permits are protective of human health
and the environment. Until the Agency
is in a position to determine, on a
national basis, whether additional
standards are needed to address toxic
emissions, we anticipate this policy will
remain in effect. EPA's intention is to
make that determination, if sufficient
data is in hand, by the time of the final
rule, now scheduled for issuance in
December 1996. In that respect, we
emphasize the importance of the
submission of detailed data on non-
dioxin PICs from commenters.
In the meantime, the omnibus
provision in § 3005(c)(3) provides the
regions and authorized states with the
proper site-by-site authority to ensure
that these risk assessments are
completed as part of the permitting
process. Other RCRA statutory and
regulatory provisions may apply as well.
Furthermore, we encourage individual
facilities to work with their local
communities in designing these risk
assessments and in carrying out the
testing and analysis, so that the
confidence of local communities is
maximized.
In addition, EPA strongly urges
companies to explore waste
minimization opportunities as a means
to reduce risks from combustion
emissions, particularly with respect to
the HAPs of concern. Nearly every state
provides free pollution prevention/
waste minimization technical
assistance. Further information on how
to obtain this assistance can be
furnished by state permitting agencies
or by contacting the National Pollution
Prevention Roundtable at (202) 466-
7272. Other sources of information
include Enviro$ense, an electronic
library on pollution prevention,
technical assistance, and environmental
compliance. Access is via a system
operator (703) 908-2007, via modem at
(703) 908-2092, or via Internet at http:/
/wastenot.inel.gov/enviro-sense.
PART FOUR: RATIONALE FOR
SELECTING THE PROPOSED
STANDARDS
This part describes the Agency's
rationale for today's proposed standards
and other options under consideration.
I. Selection of Source Categories and
Pollutants
A. Selection of Sources and Source
Categories
The Agency is proposing emissions
standards for three source categories:
hazardous waste incinerators, hazardous
waste-burning cement kilns, and
hazardous waste-burning lightweight
aggregate kilns. The Agency is not
proposing to regulate emissions from
CKs (in this notice) or LWAKs that do
not burn hazardous waste.
In this section, we discuss the
Agency's analysis of subdividing
incinerators by size (i.e., small and large
sources) and subdividing cement kilns
by process type (i.e., wet and dry). We
also discuss the scope of the MACT
standards for cement kilns, and the
existing RCRA standards that control
emissions of HAPs from equipment
leaks and tanks which are used to
manage hazardous waste.
1. Consideration of Subdividing
Incinerators by Size
Section 112(d) allows the
Administrator to distinguish among
classes, types, and sizes of sources
within a source category in establishing
MACT floor levels. Given that the size
of incinerators, as measured by gas flow
rate in actual cubic feet per minute
(acfm), varies substantially (i.e., from
1,000 acfm to 180,000 acfm), the Agency
considered subdividing incinerators by
size.
The basis for distinguishing between
small and large incinerators as well as
the preliminary estimates of the
resultant floor levels for each category
are presented in the docket and
summarized below. The Agency is not
proposing separate standards (at the
floor)19 for incinerators because: (1) the
types and concentrations of
uncontrolled HAP emissions are similar
for large and small incinerators; (2) the
same types of emission control devices
are applicable to both small and large
incinerators; and (3) the floor levels
would be generally unchanged20
(several floor levels would decrease
somewhat), with the exception that the
LVM standard for large incinerators
would increase by more than a factor of
four. We believe that the higher LVM
floor level for large incinerators would,
not be appropriate given that
approximately 80 percent of
incinerators already are meeting the
LVM floor without subdividing.
The Agency invites comment on its
determination that subdividing
incinerators by size would not be
warranted. We also invite comment on
whether subdividing incinerators by
other classifications (e.g., commercial
versus on-site units) would be
appropriate for establishing MACT floor
levels. Commenters should provide data
and information on, in particular: (1)
how the types and concentrations of
uncontrolled HAP emissions are
different for the suggested categorization
of sources; (2) whether and why MACT
emission control technology would not
be applicable to a category of sources;
and (3) other appropriate factors.
To investigate the effect on MACT
floor levels of subdividing incinerators
by size, the Agency identified a gas flow
rate of 23,127 acfm as a reasonable and
appropriate demarcation between small
and large incinerators. This value was
determined using a slope analysis
approach whereby gas flow rates for
each source (for which the Agency had
data) were plotted in ascending order.
The Agency chose the point at which
the slope markedly changed as the point
of demarcation between small and large
incinerators. Approximately 57 percent
of incinerators for which we have gas
flow rate data would be classified as
small using this approach.
19 Note that we discuss in Part Four, Section in
in the text whether beyond-the-floor standards for
D/F, Hg, and PM (as currently proposed for all
incinerators) are appropriate for small incinerators.
20 And therefore, a level of complexity would be
added to the rule without substantial benefit.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17373
Projected MACT floor levels for small
and large incinerators are compared to
floor levels for combined incinerators
(i.e., without subdividing) in the table
below:
D/F (ng/dscm)
PM (mg/dscm)
Ha (uo/dscm)
SVM (jig/dscm)
LVM (ng/dscm)
HC1 + Clj (ppmv)
CO (ppmv)
HC (ppmv) .
Small incinerators
Floor level
0.2 TEQ or <400 °F
180
110
230
160
280
100
12
Large incinerators
Floor level
0.2 TEQ or <400 °F
180
130
270 .
880
260
100
12
Floor levels for all
incin-
erators combined
0.2TEQor<400°
180
130
270
210 -•
280
100
12
F.
2, Consideration of Subdividing Cement
Kilns by Manufacturing Process
The Agency also considered whether
to subdivide the cement kiln source
category into wet and dry process kilns
given that these types of kilns are
designed and operated differently; (See
discussion in Part Two, Section II.)
MACT floor levels for wet and dry kilns
are compared to floor levels for
combined cement kilns (i.e., without
subdividing) in the table below:
Pnlltitant
D/F (ng/dscm)
PM (mg/dscm)
Ha (ua/dscm)
SVM (ng/dscm)
LVM (ug/dscm)
HCI + C\i (ppmv)
Wet process kilns
Floor level
0.2 TEQ or 41 8 °F
69
83
870
220
460
Dry process kilns
Floor level
0 2 TEQ or 547 °F .
69
150
57
49
340
Floor levels for all kilns
combined
02 TEQ or 418 °F
69
130
57
130
640
Subdividing cement kilns by process
type would result in a mix of impacts
with varying degrees of significance. For
wet kilns, the main impact would be an
increase in the SVM floor from 57 to 870
ug/dscm. The mercury floor, on the
other hand, would drop from 130 to 83
Ug/dscm. The remainder of the floors
would remain roughly the same. For dry
cement kilns, the main impact would be
that the LVM floor drops from 130 to 49
Ug/dscm. The dioxin/furan floor would
change by allowing a higher APCD
temperature—547 °F rather than 418 °F.
The Agency is not proposing separate
standards for wet and dry process kilns
because: (1) The types and
concentrations of uncontrolled HAP
emissions are similar for both types of
kilns; (2) the same types of emission
control devices are applicable to both
types of kilns; (3) for dry process kilns,
the LVM floor level would drop to an
extremely low level that may be difficult
for many kilns to achieve because of the
presence of these metals in raw
materials; and (4) for wet kilns, the SVM
floor would increase to 870 ug/dscm, a
level much higher than the industry can
achieve.21 There may also be other
factors that should be considered, and
the Agency invites comment on those in
addition to the factors noted above.
We note that the cement industry has
asserted that it is not feasible to use a
FF on wet kilns in cold climates because
the "high moisture content of the gas
will clog the fabric with cement-like
dust and ice." 22 This is not consistent
with the Agency's understanding.
Although wet kilns located in cold
climates that operate at low flue gas
temperatures (e.g., 350-400 °F) in order
to minimize formation of D/F and
improve performance of activated
carbon injection systems may be
required to improve insulation or take
other measures to minimize cold spots
in the baghouse to limit corrosion, we
believe that appropriate measures can
be readily taken. The Agency is aware
of two wet kilns that currently operate
fabric filters in cold climates
(Thomaston, Maine, and Dundee,
Michigan) at flue gas temperatures
21 Sen letter from Craig Campbell, CKRC, to James
Borlow, USEPA, undated but received February 20,
1096. Wo nota that, although the Agency is
proposing a SVM standard of 57 ug/dscm, we invite
comment on an alternative (and potentially
preferable) approach to identify MACT floor
technology which would result in a floor-based
standard of 160 ug/dscm. See Part Four, Section IV
in the text. Because we identified the alternative
approach late in the rule development process, we
are inviting comment on the higher standard rather
than proposing it.
22 See letter from Micheal O'Bannon, EOF Group,
to Elliot Laws, USEPA, dated February 14,1996, p.
3 of Attachment.
below 400 °F.23 In addition, a wet kiln
burning hazardous waste in Paulding,
Ohio, is currently upgrading its PM
control system to replace an ESP with
aFF.
The Agency invites comment on the
appropriate criteria to be used and upon
its determination that subdividing
cement kilns by process type is not
warranted. Commenters should provide
data and information on, in particular:
(1) Whether the types and
concentrations of uncontrolled HAP
emissions are different for wet and dry
kilns; (2) whether and why MACT
emission control technology(ies) would
not be applicable to a wet or dry kiln;
and (3) other appropriate factors.
3. Scope of the MACT Standards for
Cement Kilns
The proposed NESHAP for cement
kilns addresses only exhaust
combustion gas emissions from main
stack(s), bypass stack(s), and fugitive
combustion emissions (e.g., leaks from
kiln seals). The cement kiln standards
would not apply to process or fugitive
emissions that are not affected
23 See USEPA, "Draft Technical Support
Document For HWC MACT Standards, Volume IE:
Selection of Proposed MACT Standards and
Technologies", February, 1996, for further
information.
-------
17374 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
by burning hazardous waste (such as
emissions from raw material processing
or clinker cooler emissions).24
4. Current RCRA Controls on Equipment
Leaks and Tanks
We note that the Agency has
promulgated ah- emission standards
regulating fugitive emissions from
equipment leaks (e.g., pumps,
compressors, valves) and tanks which
are used to manage hazardous waste.
Accordingly, these devices are not
addressed by today's proposal. (Tanks
and equipment leaks from HW
management activities at HWCs are
regulated under RCRA standards. See,
e.g., 40 CFRParts 264 and 265, Subparts
AA, BB, and CC. These controls are
expected to be consistent with MACT
and are not being reevaluated here.)
B. Selection of Pollutants
As noted earlier, section 112(b) of the .
Clean Air Act contains a list of 189
hazardous air pollutants for which the
Administrator must promulgate
regulations establishing emissions
standards for designated major and area
sources. The list of 189 HAPs is
comprised of metallic, organic, and
inorganic compounds.
Hazardous waste incinerators and
hazardous waste-burning cement kilns
and LWAKs emit many of the listed
HAPs. Data available to the Agency
indicate that metal HAP emissions
include antimony, arsenic, beryllium,
cadmium, chromium, lead, mercury,
nickel, and selenium compounds.
Organic HAPs emitted include
chlorinated dioxin and furan, benzene,
carbon disulfide, chloroform,
chloromethane, hexachlorobenzene,
methylene chloride, naphthalene,
phenol, toluene, and xylene.
Hydrochloric acid and chlorine gas are
prevalent inorganic compounds found
in stack emissions because of high
chlorine content of many hazardous
wastes.
Today, the Agency is proposing eight
emissions standards for individual
HAPs, group of HAPs, or HAP
surrogates. These emission standards
cover dioxin/furan, mercury, particulate
matter, semivolatile HAP metals (lead
and cadmium), low-volatile HAP metals
24 Today's proposal applies only to those kilns, .
that burn or process hazardous waste irrespective
of the purpose of burning or processing. The term
"burn" means burning for energy recovery or
destruction, or processing as an ingredient. The
Agency is developing a NESHAP for cement kilns
that do not process hazardous waste in a separate
rulemaking. That NESHAP will also regulate those
hazardous waste-burning cement kiln process and
fugitive emissions that would not be subject to
today's rule (i.e., emission sources other than the
main or by-pass stack).
(antimony, arsenic, beryllium, and
chromium), carbon monoxide,
hydrocarbons, and total chlorides. The
following discussion presents the
Agency's rationale for proposing
NESHAPs for these individual HAPs,
group of HAPs, or HAP surrogates.
1. Toxic Metals-
In developing today's proposed rule,
the Agency considered 14 toxic metals
that may pose a hazard to human health
and the environment when they are
components of emissions from
hazardous waste combustion sources.
Section 112(b) of the Act contains a list
of 11 metal HAPs: antimony, arsenic,
beryllium, cadmium, chromium, cobalt,
lead, manganese, mercury, nickel, and
selenium. The list of hazardous
constituents under RCRA25 specifies
three additional metals: barium, silver,
and thallium. Five of these metals (or
their compounds) are known or
suspected carcinogens: arsenic,
beryllium, cadmium, hexavalent
•chromium, and nickel.
To develop an implementable
approach for controlling the metal HAP
emission levels, the Agency grouped
metal HAPs by their relative volatility
and is proposing an emissions limit for
the each volatility group (i.e., the sum
of emissions from the metals in the
group cannot exceed the limit). We
selected the following three groups: (1)
A high-volatile group comprised of only
mercury, (2) a semivolatile group
• comprised of lead and cadmium, and (3)
a low-volatile group consisting of
antimony, arsenic, beryllium, and
chromium. The Agency's proposal not
to include the remaining seven toxic
. metals in these volatility groupings is
discussed later in this section.
Our data indicate that mercury is
generally in the vapor form in and
downstream of the combustion
chamber, including at the air pollution
control device (APCD). Thus, the level
of emissions is a function of the feedrate
of mercury and the use of APCDs that
can control Hg in the vapor form (e.g.,
carbon injection, wet scrubbers for some
control of soluble HgCl). The
semivolatile group metals typically
vaporize at combustion temperatures, ,
then condense onto fine particulate
before entering the APCD. Thus,
emissions of semivolatile metals are a
function not only of the feedrate of the
metal, but also of the efficiency of the
particulate matter (PM) control device.
Low-volatile metals are less apt to
vaporize at combustion temperatures
and therefore partition primarily to the
bottom ash, residue, or clinker (in the
case of cement kilns) or adsorb onto
large, easy-to-control particles in the
combustion gas. Thus, low-volatile
metal emissions are more strongly
related to the operation of the PM APCD
than to the feedrate.26
We note that the dynamics associated
with the fate of metals in a combustion
device are much more complex than
presented here. Numerous factors
impact metals' behavior such as the
presence of chlorine (higher metal
volatility associated with metal
chlorides than metal oxides),
combustion conditions within the
device (e.g., temperature profile), inter-
metal relationships, physical and
chemical form the metal exhibits when
introduced to the device (e.g., valence
state and solid versus liquid), type and
efficiency of the particulate control
device, and differences in the design
and operation of sources (e.g., cement
kiln dust recycling rate). See the
technical background document
supporting today's proposal for more
details.27
Setting an emission level for a number
of grouped metals has several
advantages and disadvantages. One
advantage is that fewer individual
standards are involved, which helps
implementability. Moreover, grouping
allows a facility more flexibility in
complying with an emissions standard
based on facility-specific characteristics
(e.g., special characteristic waste
streams) and operation requirements
(e.g., reduced spiking of numerous
metals). On the other hand, a
disadvantage of a group emission limit
is that it potentially allows higher
emissions of the more toxic metals
within a group (than if an individual
metal limit were established).28
The Agency is proposing not to
regulate directly emissions of the
remaining four metal HAPs (i.e., cobalt,
manganese, nickel, and selenium).29 The
25 The list of hazardous constituents is contained
in appendix VIII of Part 261. Cobalt and manganese
are not hazardous constituents.
26 Although, at a given PM emission rate at a
source, emissions of LMV will be affected by LVM
feedrate.
27USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume VII:
Miscellaneous Technical Issues", February 1996.
28 We note that, for the risk assessment used to
determine if RCRA concerns would be adequately
addressed by the proposed MACT standards, we
assumed that each metal in a volatility was emitted
in turn at the emission limit for that volatility
group.
29 The Agency acknowledges that three metals
(barium, silver and thallium), currently regulated by
the BIF rule, would not be regulated under this
MACT proposal. EPA notes that these three metals
are not HAPs. The Agency believes that the
combination of the proposed particulate and metals
standards, would adequately control emissions of
these three metals.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17375
Agency's rationale is based upon a
combination of factors: (1) Inadequate
emissions data for Co, Mg, Ni, and Se;
and (2) relatively low toxicity of Co and
Mn. The Agency specifically requests
comment on whether these four metals
would be adequately controlled under
the MACT standards that would be
provided by today's proposal.
The Agency is aware of two other
approaches to group toxic metals. First,
the European Union has established
three groupings to control metal
emissions from hazardous waste
incineration units. One "group"
includes only mercury, a second group
consists of cadmium and thallium, and
the third group includes antimony,
arsenic, chromium, cobalt, copper, lead,
manganese, nickel, tin, and vanadium.
Section VII of this Part summarizes the
European Union emission standards.
A rulemaking petition30 submitted to
the Agency by the Cement Kiln
Recycling Coalition (CKRC) contained a
report31 (appendix D of the petition)
prepared by a technical advisory board
to the CKRC. Their analysis of stack
emissions and cement kiln dust data
suggests three volatility groupings based
on metal volatility demonstrated in
cement kilns. The groupings are: (1)
Volatile metals including mercury and
thallium; (2) semivolatile metals
consisting of antimony, cadmium, lead,
and selenium; and (3) low-volatile
metals comprising barium, beryllium,
chromium, arsenic, nickel, manganese,
and silver. See the technical background
document for further discussion on
grouping metals by volatility.32 The
Agency requests comments on the
appropriateness of grouping metals by
volatility and requests supporting
information and data on the appropriate
•w CKRC's rulemaking petition proposes to
establish now technology-based combustion
omissions standards and was submitted to EPA on
January 18,1994. CKRC's petition consists of four
basic components. First, the stringency of current
DIP Rulo toxic metal limits should be increased by
factors of 5 to 10 and applied to all combustion
devices (i.e., both BIFs and incinerators). Second,
now regulatory efforts for dioxin/furan standards
should focus on a toxic equivalency approach
(TEQ) rather than on a total congener approach.
Third, tho implementation of the new metals and
dioxin/furan standards should be applied uniformly
to all types of hazardous waste combustors (HWCs)
and imposed at the same time. Finally, EPA should
conduct a rulemaking on indirect exposure risk
assessments before requiring their use. CKRC's
petition has been placed in the docket supporting
today's proposal.
" "Scientific Advisory Board on Cement Kiln
Recycling (Process Technology Workgroup),
Evaluation of the Origin, Emissions and Control of
Organic and Metal Compounds From Cement Kilns
Co-Fired With Hazardous Wastes," June 8,1993.
M USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume VH:
Miscellaneous Technical Issues," February 1996.
composition of metal volatility groups
(i.e., for the metals discussed above).
2. Toxic Organic Compounds
Burning hazardous waste that
contains toxic organic compounds
under poor combustion conditions can
result in substantial emissions of HAPs
originally present in the waste as well
as other compounds, due to the partial
but incomplete combustion of the
constituents in the waste (known as
products of incomplete combustion, or
PICs). PICs can be unburned organic
compounds that were present in the
waste, thermal decomposition products
resulting from organic constituents in
the waste, or compounds synthesized
during or immediately after combustion.
The quantity of toxic organic
compounds emitted depends on such
factors as the combustion conditions
under which the waste is burned
(including time, temperature, and
turbulence), the concentrations of the
toxic compounds in the waste, and the
waste firing rate.
Since the majority of the 189
enumerated HAPs are organics, the
Agency has concluded (for today's
proposal) that establishing individual
emission limits for each of the organic
HAP compounds emitted from these
combustion sources would be
impractical and not implementable.
Measuring each compound would be
very costly and would pose
unreasonable compliance and
monitoring burden on the regulated
community while achieving little, if
any, emission reduction from the
approach presented in today's proposal.
In addition, EPA and state compliance
oversight and enforcement efforts would
also be unreasonably costly without
concurrent benefits. Also, the Agency
does not have adequate emissions data
to support development of individual
organic emission limits 33 at this time.
Therefore, the Agency is proposing a
multi-faceted approach to control the
toxic organic HAPs to be addressed
under § 112: (1) Emissions limits for
dioxin and furan on a toxicity
equivalents (TEQ) basis; (2) limits on
flue gas concentrations of hydrocarbons
(HC) as a HAP surrogate; (3) limits on
flue gas concentrations of carbon
monoxide (CO) also as a HAP surrogate;
and (4) emission limits for particulate
matter (PM) to control adsorbed
semivolatile organic HAPs (see separate
discussion on PM below).
First, given the high toxicity of some
dioxin and furan congeners and the fact
that standards ensuring good operating
conditions alone (i.e., temperature at the
inlet of the APCD) will not always
control emissions of dioxin/furans
(D/F), the Agency has determined that
proposing an emission standard
specifically for D/F is a necessary
component to the multi-faceted
approach for toxic organics emissions
control. The D/F standard proposed
today is based on TEQ (Toxicity
Equivalents).34 TEQ is a method for
assessing the risks associated with
exposures to complex mixtures of
chlorinated dibenzo-p-dioxin and
dibenzqfurans (CDDs and CDFs). The
method relates the toxicity of the 209
structurally related chemical pollutants
to the toxicity of 2,3,7,8-
tetrachlorodibenzo-p-dioxin (2,3,7,8-
TCDD).
Second, the Agency is proposing to
use carbon monoxide (CO) and
hydrocarbons (HC) as surrogates to
control emissions of non-D/F organic
HAPs. We note that limiting CO and HC
emissions to levels ensuring good
combustion conditions would also help
minimize D/F precursors. CO and HC
emissions are both recognized
indicators of combustion intensity and
completeness. Low CO flue gas levels
are indicative of a combustion device
operating at high combustion efficiency
(56 FR at 7149-54). Operating at high
combustion efficiency helps ensure
minimum emissions of unburned (or
incompletely burned) organics.
However, limiting CO may not by itself
absolutely minimize PIC emissions.
This is because PICs can result from
small pockets within the combustion
zone where adequate time, temperature,
turbulence, and oxygen have not been
provided to completely oxidize these
organics.35 As combustion becomes less
efficient or less complete, at some point,
the emissions of total organics
(measured as HC) will increase. A
33 The number of organic HAPs measured at each
facility varies widely with some facilities reporting
measurements for a large number of HAPs while
other facilities measuring only a few HAPs.
34 The TEQ approach used for today's proposal is
the I-TEQ/89 approach defined in USEPA, "Interim
Procedure for Estimating Risks Associated With
Exposures to Mixtures of Chlorinated Dibenzo-p-
Dioxin and -Dibenzofurans (CDDs and CDFs) and
1989 Update," March 1989. For a discussion of
establishing D/F limits based on TEQ versus total
congeners, see USEPA, "Combustion Emissions
Technical Resource Document (CETRED)," May
1994, pp. 4-21.
35 We note that there are emissions data
indicating that even though CO levels are below 100
ppmv, HC emissions can exceed 5 ppmv (measured
as propane with a heated sampling system), the
upper HC level that is generally representative of
operating under good combustion conditions. See
56 FR 7154, note 26 (February 21,1991), and
Energy and Environmental Research Corporation,
"Surrogate Evaluation of Thermal Treatment
Systems," Draft Report dated October 17,1994,
Figure 2-1.
-------
17376 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
portion of the HC emission is comprised
of organic HAPs. Thus, CO levels
provide an indication of the potential
for organic HAP emissions and CO
limits are therefore proposed as a
measure to help prevent these
emissions. HC limits are proposed to
document actual emissions of organic
HAPs.36
Notwithstanding today's proposal to
establish MACT standards for both CO
and HC emissions for HWIs and LWAKs
(CKs would be required to comply with
either a CO or HC standard for technical
reasons discussed in Section IV below),
the Agency invites comment on whether
standards for both CO and HC (coupled
with the D/F and PM standards to also
control organic HAPs) are unnecessarily
redundant. Commenters should provide
data and information on how either CO
or HC alone (but in conjunction with
D/F and PM standards) would ensure
proper control of organic HAPs. In
particular, commenters should address
the fact that the Agency's database
indicates that HC levels can exceed
good combustion condition levels when
CO levels are below 100 ppmv (thus
suggesting that controls on both CO and
HC are needed). In addition,
commenters should address how the
MACT standards proposed today for HC
would or could ensure that sources
operate under good combustion
conditions and thus minimize emissions
of organic HAPs.
If based on review of comments and
further analysis the Agency determines
that standards for both CO and HC are
not warranted, we would consider,
among other potential options, the
following alternative regulatory
approaches: (1) Give each source the
option of complying with either the CO
or HC standard (as proposed today for
technical reasons for by-pass duct gas
for cement kilns); or (2) establish a
national standard for either CO or HC,
but not both (the Agency would
determine which parameter is more
appropriate and establish a standard for
that parameter). The Agency invites
comment on these alternative regulatory
approaches or others that would ensure
proper control of organic HAP
emissions.
3. Hydrochloric Acid (HC1) and
Chlorine (C12)
Both hydrochloric acid and chlorine
are designated HAPs that are present in
HWC emissions. However, the test
method used to determine HCl and Cla
emissions (BIF methods 0050, 0051, and
9057, commonly referred to as "Method
26A")37 may not be able to distinguish
between HCl and C12 in all situations.38
Therefore, EPA proposes combining the
two HAPs into a single HCl and C12
standard. We believe this is appropriate
because emissions of both of these HAPs
can be controlled by limiting feedrate of
chlorine in hazardous waste and wet
scrubbing.39
4. Particulate Matter (PM)
EPA is proposing to use particulate
matter (PM) as a surrogate for non-D/F
organic HAPs (that are adsorbed onto
the PM) and for the metal HAPs which
are not specified in the metals standards
(i.e., Co, Mn, Ni, and Se).40 More than
40 semivolatile organic HAPs can be
adsorbed onto PM and can, thus, be
controlled by a MACT standard for
PM.41 The metal HAPs that are not
directly controlled by the MACT
standards for metals can also be
controlled (at least partially) by a PM
standard. The low volatility metals are
likely to be entrained in larger
particulates and the semivolatile metals
36 We note that virtually all HWCs are already
equipped with a CO monitor because of RCRA
requirements. In addition, several incinerators,
cement kilns and lightweight aggregate kilns are
also equipped with a HC monitor because of RCRA
or state requirements or voluntary initiative.
37 We note that owners and operators of cement
kilns have argued that this method provides
measurements that are biased high because metallic
salts penetrate the filter and the chloride is
incorrectly reported as HCl. EPA has considered
this concern and continues to believe that metallic
salts do not significantly bias the results.
Nonetheless, we invite comment on this issue. If,
in fact, metallic salts can bias the results, we invite
comment particularly on how or whether the
proposed MACT standards could be adjusted given
the inflated emissions database, and how
compliance with an adjusted standard could be
demonstrated.
38 In the presence of other halogens (e.g., fluorine
and bromine) that are often constituents of
hazardous waste, fossil fuels or kiln raw materials,
EPA is concerned that reactions can occur in the
impinger solutions used by the stack sampling
method that cause a portion of the C12 to be
reported as HCl. Thus, the HCl levels could be
biased high, and the C12 levels could be biased low.
Nonetheless, the method does continue to give an
accurate determination of combined HCl and C12
levels in the presence of other halogens.
39 We also note that, for purposes of determining
whether the proposed MACT standard would
satisfy RCRA concerns, we evaluated the level of
protection that would be provided assuming
(conservatively) that 10 percent of the HCl/Ck
standard would be emitted as the more toxic C12.
*> We note that PM 10 is a criteria pollutant under
the Clean Air Act. PM can also have adverse effects
on human health even if toxics are not adsorbed on
the PM. Although EPA cannot control PM in and
by itself under § 112(d) (it must be a surrogate for
HAP control), EPA may consider reductions in
criteria pollutants in assessing cost-effectiveness of
MACT controls. See S. Rep. No. 228,101st
Congress, 1st Session, p. 172.
41 See memo from Larry Gonzalez, EPA, to the
docket for this rule (F-96-RCSP-FFFFF), entitled
"Semi-volatile Organic HAPs that Can Be Adsorbed
onto PM", dated February 22,1996.
are likely to be condensed onto small
particulates.
The Agency notes that we are
proposing to use PM also as a
compliance parameter to ensure
compliance with the SVM, LVM, and D/
F standards. As discussed in Part V,
Section n, of the preamble, a site-
specific PM operating limit would be
established as a surrogate for the PM
control device collection efficiency.
Given that we are also proposing a PM
MACT emission standard, the site-
specific operating limit for PM could
not exceed the PM standard.
C. Applicability of the Standards Under
Special Circumstances
In this section, we discuss the
applicability of the proposed MACT
standards under the following
circumstances: (1) When a regulated
metal or chlorine is not present in the
hazardous waste at detectable levels; (2)
when the source temporarily ceases
hazardous waste burning; and (3) when
the source terminates hazardous waste
burning.
1. Nondetect Levels of Metals or
Chlorine in All Feedstreams
If no feedstreams to a HWC (e.g., on-
site incinerator) contain detectable
levels of Hg, SVM, LVM, or chlorine, the
source would not be subject to the
emission standard associated with the
metal or chlorine (e.g., if no feedstreams
contain detectable levels of chlorine, the
HCl/Cla standard would be waived). In
addition, performance testing,
monitoring, notification, and
recordkeeping requirements ancillary to
the waived standard would also be
waived. We believe that this waiver is
appropriate because the source would
be incompliance with the emission
standard by default if it was not feeding
the metal or chlorine.
To be eligible for the waiver, the
source must develop and implement a
feedstream sampling and analysis plan
to document that no feedstream
contains detectable levels of the metal
or chlorine (for which a waiver is
claimed).
The Agency invites comment on
whether it is necessary to specify
minimum detection levels (or to take
other measures) to ensure that
appropriate analytical procedures are
used to document levels of metal or
chlorine in feedstreams.
2. Nondetect Levels of Metals or
Chlorine in the Hazardous Waste Feed
The proposed MACT standards for
mercury, SVM, LVM, or chlorine would
apply even if these constituents are not
present at detectable levels in the
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17377
hazardous waste. This issue is relevant
for cement kilns and light-weight kilns
because, if these sources were not
burning hazardous waste, the proposed
MACT standards would not apply.
Cement kilns (CKs) that do not burn
hazardous waste would be subject to
separate MACT standards that the
Agency is developing for those sources,
and light-weight aggregate kilns
(LVVAKs) that do not burn hazardous
waste would not be subject to any
MACT standards.
It could be argued that a CK or LWAK
that burns hazardous waste with
nondetect levels of Hg, SVM, LVM, or
chlorine is not burning hazardous waste
with respect to that metal or the HC1/C12
standard. Accordingly, regulation
should revert to any applicable MACT
standard for the source when not
burning hazardous waste. The Agency
rejects this argument, however. A source
cannot be subject to regulation under
two MACT source categories. Further,
such an approach would be extremely
difficult to implement and enforce for
CKs given that compliance procedures
would be different for the two source
categories.
3. Sources That Temporarily Cease
Burning Hazardous Waste
Sources that temporarily cease
burning hazardous waste would remain
subject to today's proposed standards.
Similar to the discussion above, such
sources could argue that in the interim
\vhen hazardous waste is not burned,
MACT regulation should revert to the
MACT standards applicable to CKs or
LVVAKs that do not burn hazardous
waste.
The Agency rejects this argument as
well and for the same reasons discussed
above: a source cannot be intermittently
subject to MACT regulation under two
source categories, and implementation
and enforcement would be extremely
complicated. See the discussion below
regarding how to define temporary
interruptions in waste burning versus
termination of waste burning.
4. Sources That Terminate Hazardous
Waste Burning
A source that terminates hazardous
waste burning would no longer be
subject to today's proposed rules. A
source has terminated hazardous waste
burning when it: (1) ceases burning
hazardous waste (i.e., hazardous waste
is not fed and hazardous waste does not
remain in the combustion chamber); and
(2) stops complying with the proposed
standards and begins complying with
other applicable MACT standards (i.e.,
cement kilns must comply with the
MACT standards, when promulgated,
for kilns that do not burn hazardous
waste). In addition, today's rule would
require sources that terminate
hazardous waste burning to notify the
Administrator in writing within 5 days
of the termination.
Such sources could begin burning
hazardous waste again under the
following conditions: (1) The source
must comply with the MACT standards
applicable to new sources; (2) the source
must submit a notification of
compliance with the standards (based
on a comprehensive performance test);
and (3) prior to submitting the
notification of compliance, the source
cannot burn hazardous waste for more
than a total of 720 hours, and hazardous
waste may be burned only for purposes
of emissions pretesting (i.e., in
preparation for the comprehensive
performance test) or comprehensive
performance testing.
We are taking this position regarding
termination of waste burning to avoid
the implementation and enforcement
complications that could result if a
source could claim that it was not
subject to the proposed regulations
during those periods of time that it was
not burning hazardous waste. Without
these requirements, a source could
vacillate at will between being regulated
and unregulated (or for CKs, between
being subject to regulation as a
hazardous waste-burning kiln versus a
non-hazardous waste-burning kiln). We
invite comment on whether these
requirements are reasonable and
appropriate to address the Agency's
implementation and enforcement
concerns.
II. Selection of Format for the Proposed
Standards
A. Format of the Standard
When EPA regulates a source, it must
determine on a case-by-case basis what
format the standards are. This section
explains the reasons why EPA chose the
format it did for this specific source
category. Due to differing situations in
other cases, other formats may be
chosen for other source categories.
1. Units
EPA investigated four formats for use
in
stand
calculated mass-based emissions;
percent reduction; and concentration-
based. The Agency ultimately selected
concentration-based standards for the
reasons discussed below.
The mass-based approach would set a
limit of mass emissions per unit time,
i.e., kg/hr, Ib/hr, etc. This approach was
rejected because it is inherently
o
expressing today's proposed
idards: mass-based emissions;
incompatible with technology based
standards for several reasons. First, a
mass-based standard does not assure
good control at small facilities. Small
facilities have lower flow rates, would
be allowed higher concentration of
emissions, and thus could meet a
standard with no or minimal
technological control. Also, it produces
an undue burden on larger facilities in
that they would have to install controls
and small facilities would not. One
potential consequence is that it would
cause an incentive for more small
facilities, causing an increase in
emissions nationally. For these reasons,
this option.was not chosen.
An alternate to the mass-based
approach is the calculated mass-based
approach. This would involve EPA
determining some appropriately low
level of metals and chlorine feed,
multiplying that by a system removal
efficiency factor, and issuing the result
as a mass-based limit. One concern with
this approach is EPA does not know
what feedrate would be appropriate.
Any feedrate could be construed as
arbitrary. Also, the approach would
result in a mass-based limit which does
not address concerns described in the
preceding paragraph. It also does not
address how to set the other standards:
CO, HC, PM, and dioxin/furans. For
these reasons, this option was not
chosen.
A third approach is to set the,
standards based on a specified percent
reduction. This comports well with a
technology-based approach because it
deals directly with determining what
technology performs most efficiently.
However, there are problems with this
approach. First, it is difficult to
determine where the percent reduction
should be applied: feed to stack, across
the APCD train, or across a specific
control device. Use of feed to stack
percent reductions present a difficulty
due to the measurement variability of
feed samples and stack emissions.
APCD train or device specific percent
reductions would be difficult to
implement. Facilities are not configured
to sample inlet emissions to the APCD
train or to a specific APCD. Thus,
facilities would have to be reconfigured
to allow inlet sampling. Stack sampling
would be required at both the outlet
and, possibly, multiple inlet points.
This would significantly increase the
testing burden. In addition,
implementation of any approach based
on percent reduction would involve
substantial and expensive monitoring of
operating parameters to ensure that the
specified percent reduction occurs
during operation. For these reasons, this
approach was not chosen.
-------
27378- Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
The approach that was chosen for
these source categories is to set
concentration-based standards. This
approach is consistent with how EPA
has historically hased air emission
standards. It favorably addresses the
problems of the other options. However,
it does allow larger facilities to emit
higher mass emissions of HAPs. But
mass-based levels would result in
higher emissions nationally by
encouraging more smaller facilities (see
previous paragraph). This tradeoff>
having higher mass emissions at larger
facilities but lower emissions nationally,
was considered acceptable for this
proposal. Concentration based
approaches are also easier to implement
and do not necessarily rely on the
setting of operating limits. For this
reason, concentration-based standards
are regarded as preferable to the other
options, and was chosen on that basis.
It is possible that other units could be"
chosen for other source categories. As
explained in the introductory paragraph
this is consistent because other units
might be more appropriate for other
source categories.
2. Correction to 7 Percent Oxygen and ,
20° C ,
All standards are corrected to 7
percent oxygen and 20° C. This is
because the data EPA used to derive the
standards were corrected in this
manner. This is also consistent with the
correction used for BIFs, hazardous
waste incinerators, MWCs, and MWIs.
3. Significant Figures and Rounding
All standards proposed here are
expressed to two significant figures.
For the purposes of rounding, we
propose to require the use of ASTM
procedure E—29-90 or its successor. •
This procedure is the American
standard for rounding. Rounding shall
be avoided prior to rounding for the
reported result. ' '
B. Averaging Periods . ;
Averaging periods are the time
periods over which emissions or
feedstream and operating parameters are
set. These periods require consideration
because of the inherent variability
associated with the operation of
complying (i.e>, properly designed and
operated) MACT devices. As noted
above, facilities normally operate within
certain limits but do have emissions
above and below these normal levels
due to the natural variability associated
with the operation of a facility. EPA
must account for this variability when
promulgating technology-based
standards. See, e.g., FMC Corp. v. Train,
538 F.2d 973, 986 (4th Cir. 1976). If EPA
were to establish a "not-to-be-exceeded"
limit, that limit would invariably be
higher than if the limit were expressed
as an average emission level. That
would tend to encourage higher
emitting, but low variability devices
since they could meet the not-to-exceed
standard.
For instance, say EPA is considering
establishing a standard on: an
instantaneous basis; a one hour average;
and a 12-hour average. Also, assume
that the complying MACT facility .has
average emissions of 5 and short-term
perturbations as high as 300. In this case
equally stringent emissions levels could
be: 300 on an instantaneous basis; on
the order of 10 for an hourly average; or
closer to 5 for the 12-hour average. If the
limit were established at 300 on an
instantaneous basis, this could
significantly favor a facility that has
high perturbations less than 300, but
average emissions of 250 (assuming the
facility with average emissions of 250
could meet the instantaneous limit, 300,
with fewer controls.) This facility would
emit 50 times more of that HAP than a
facility operating at an emission average
of 5, but would still comply with the
standard. To address the problem of
setting limits on an instantaneous basis,
emissions and feedstream and operating
limits are established on the average
with specified averaging periods.
1. Manual Methods
The MACT standards for HWCs
(except those for HC and CO) were
based on the average of data from three
test runs during which emissions were
measured by manual methods. EPA thus
proposes that compliance be based on
the average of three manual methods
test runs to be consistent with data used
to establish the standards. Chemical
Waste Management v. EPA, 976 F.2d 2,
34 (B.C. Cir. 1992) (Noting that this is
an inherently reasonable approach and
is consistent with the standard approach
for compliance under the Part 63 MACT
standards.)
The standard could be set in such a
way as to require all three runs to be
less than the standard. Such a standard
would be derived by choosing the
highest data point from three manual
test runs and would result in an
emission level higher than those
proposed. The "not-to-be-exceeded"
approach was considered problematic
for reasons just described, so averaging
was chosen.
Manual methods sample facility
exhaust emissions for a period of time.
The minimum length of time required to
sample is specified indirectly by the
manual method in the form of collection
or gas flow specifications. The results of
the manual method test are reported as
an average over the sampling period.
Therefore for manual method test runs,
the averaging period is the sampling
period over which the sample was
collected.
EPA proposes no specific averaging
period here for manual method test
runs, with one caveat discussed below.
Instead EPA proposes to rely on the
minimum sampling volumes or
collected sample (whichever the method
requires) specified by the manual
methods. EPA invites comment on
whether minimum sampling periods for
manual methods should be specified
directly.
EPA is proposing a three hour
minimum sampling time for method
0023A. Three hours is also the
minimum sampling period stated in
method 23 to Part 60, appendix A. EPA
is proposing a minimum sampling time
in order to ensure that each D/F run
samples long enough to obtain adequate
samples of the various congeners to
determine compliance with the TEQ
standard. This issue is important here
because there is an inconsistency
between air rules and RCRA rules
regarding how to treat nondetected
congeners when calculating the TEQ.
The document which defines the TEQ
calculation, "Interim Procedures for
Estimating Risks Associated with
Exposures to Mixtures of Chlorinated
Dibenzo-p-Dioxins (CDDs and CDFs)
and 1989 Update" (EPA/625/3-89/016,
March 1989), uses in its examples the
assumption that all non-detects are zero.
Also, Method 23 of Part 60 Appendix A,
the method used by air programs for
.determining total D/F congeners,
similarly states in Section 9, titled
Calculations:
Any PCDD's or PCDF's that are reported as
nondetected (below the MDL) shall be
counted as zero for the purpose of calculating
the total concentration of PCDD's and PCDF's
in the sample.
Therefore, many assume that nondetects
are zero for the purposes of calculating
site specific TEQs.
Unfortunately, RCRA programs in
most instances use the nondetect value,
not zero, in the calculation of the TEQ.
(See BIF method 23 found in Part 266,
Appendix IX, section 3.4.) Since this
rule would be promulgated under both
RCRA and CAA authority, this issue
needs to be resolved.
The Agency believes a facility will
have to measure for 20 minutes per run
using SW-846 method 0023a to obtain
enough sample to be useful for the TEQ
calculation. This leads EPA to believe
that enough sample will be collected
during a three hour run to assure that
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17379
nondetected congeners are indeed not
present. If a source complies with the
minimum sampling period and still has
non-detects, then EPA proposes
allowing non-detects to be assumed to
be zero.
This would also apply to other
methods which have passed the Method
301 validation procedures and EPA has
agreed are acceptable. In the case of
other methods, the facility would
assume that non-detects are zero if the
method accumulates the same amount
or more sample than Method 0023 A
would in a three hour run. If a source
chooses not to comply with the three
hour minimum, EPA would mandate
that non-detected congeners be assumed
to be present at the detection level for
the purposes of the TEQ calculation.
EPA specifically invites comments on
the selection of the proposed minimum
sampling time for the D/F method and
the assumed concentration of
nondetected congeners in the
calculation of the TEQ.
2. Continuous Emissions Monitoring
Systems (GEMS)
EPA is proposing to require the use of
five GEMS—CO, HC, O2, Hg, and PM—
and to allow the use of GEMS for SVM,
LYM, HC1, and C12. Presently, for
cement kilns and LWAKs, continuous
emission monitoring of O2 and CO (or
HC) is required under the BIF rule (40
CFR 266.103(c)(l)(v)). Emission limits
and their associated averaging period
must be established for all of these
pollutants (except for O2) in keeping
with the nature of compliance with a
GEMS. (The O2 GEMS is used to
continuously correct the GEMS readings
for the other pollutants to 7 percent O2.
There is no emission limit specific to
O2.) Hourly rolling average emissions
data are available to establish emission
limits for CO and HC on an hourly-
rolling average.
Only manual method stack emissions
data, however, are available to establish
appropriate emission limits and
averaging periods for the other
standards: Hg, PM « SVM, LVM, and
HC1 and C12. This presents a unique
issue for the Agency to resolve since, in
most cases, EPA promulgates GEMS
standards by collecting GEMS emissions
data from facilities run under "normal"
conditions. The Agency would use this
GEMS data to calculate a statistically
based GEMS emission standard,
assuming some confidence interval and
number of annual exceedances. Since
no "normal" GEMS data exists, but
worst-case manual test data from trial
burns and compliance tests does, an .
alternate approach must be developed to
derive a GEMS emission standard an its
associated averaging period.
a. Approach to Establishing Averaging
Periods for Hg, PM « SVM, LVM, HC1
and C12 GEMS. One important issue
concerning the data is that it was
obtained from trials burn and
compliance test results (similar to the
comprehensive performance test,
described in section III of Part Five).
These are generally worst-case tests
facilities used to establish operating
limits under the BIF and Incinerator
rules. Facilities must be in compliance
with all standards at all times they are
burning hazardous waste. Therefore, the
emissions represented by this data are
the highest emissions the facility could
experience and be in compliance with
the current BIF and incinerator rules. In
other words, the emissions data
represents a not-to-be-exceeded
emission level for the given facility.
Now, let us examine how a facility
would comply with today's proposed
emission standards if they were not to
use a GEMS, but by performing a
comprehensive performance test and
complying with the standards using
operating parameter limits. As a result
of today's proposed rule and as was the
case in the BIF and incinerator rules,
EPA believes facilities will conduct a
comprehensive performance test in the
same way current trial burns and
compliance tests are conducted. That is
they will attempt to get the widest
operating envelope possible by
intentionally running the facility under
conditions which will maximize
emissions (by practices such as
maximizing feed-rates, running control
devices less effectively, etc.) and yet not
exceed any applicable emission
standards. Facilities will use the
operating data from the comprehensive
test to establish and continuously
monitor operating limits for feedrate
and device parameters. This defines the
facility's operating envelope. During
normal operation, owner/operators will
operate in such a way that the facility
is performing better than the operating
limits established during the
comprehensive performance test. Since
exceedances of operating limits
established during the comprehensive
performance test are a de facto violation
of the corresponding standard, this
means that the emissions during normal
operation will at all times be lower than
those during the comprehensive test.
When complying with today's
proposed standards using a GEMS, it is
important that facilities using a GEMS
not be at a disadvantage relative to
facilities using operating parameter
limits. There are two ways a
disadvantage could occur: when the
emission standard is numerically less
and/or the averaging period is shorter.
In the case of manual stack tests, the
averaging period is the stack sampling
time. Therefore, the GEMS emission
limit would be equal in stringency to
the manual stack test limit if they both
had the same numerical value and the
GEMS averaging period were equal to
the sampling period for the manual
method.
Also, EPA believes facilities have a
number of advantages using GEMS.
First, the assumptions to assure
compliance are fewer and less
conservative (direct measure of the
standard is the top of the monitoring
hierarchy; see section II.A. of Part Five.)
GEMS are less intrusive on the facility
than operating parameter limits. Most
importantly, GEMS mean facilities need
to monitor only one emissions
parameter to assure compliance rather
than multiple operating limits, often
relevant to more than one standard.44
In summary, regardless of whether
GEMS or operating limits are used, both
continually assure that the facility is
meeting the standard(s) at all times.
GEMS are an alternate, more direct,
method of confirming a state of
performance than are continuously
monitored operating parameter limits
established through a comprehensive
test. A facility which complies with the
standards in today's proposed rule
would experience its highest emissions
during a comprehensive performance
test, when the facility establishes its
operating envelope to ensure it is in
compliance with the standards at all
times. Therefore, a GEMS limit is
equally stringent to a standard for a
comprehensive performance test if it is
numerically equal and has the same
averaging period. For comprehensive
performance tests, the averaging period
is the sampling time for the manual
method. Therefore, it is proposed that
the GEMS standards be the same
numerical limits established for manual
metiiod comprehensive performance
tests with the averaging period equal to
•° Note that the PM GEM is also used as an
operating parameter for PM APCD efficiency and
that additional averaging periods apply during
normal operation. See Part Five, Section II.C.7.
titled "Particulate Matter" for more information.
43 Note that the PM GEM is also used as an
operating parameter for PM APCD efficiency and
that additional averaging periods apply during
normal operation. See Part Five, Section II.C.7.
titled "Particulate Matter" for more information.
44For example, an exceedance of an operating
parameter limit Used to ensure compliance with the
dioxin, mercury, SVM, LVM, and HC1 and C12
standards would be a violation of all those
standards. If a GEM were used for one or more of
these standards, a violation would only occur if the
CEM limit were exceeded.
-------
17380 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 /Proposed Rules
the sampling period for three manual
method test runs.
b. Averaging Periods for CO and HC
GEMS. As stated previously, the data
used to derive today's proposed CO and
HC standards proposed are not manual
methods data, hut continuous emissions
data hased on a one-hour rolling
average. To be consistent with the data
used to derive the standards, it is
proposed that the averaging periods for
CO and HC GEMS standards remain
one-hour.
c. Averaging Periods for Other GEMS.
Based on the discussion of subsection I
above, EPA proposes the following
GEMS averaging periods for GEMS. The
numerical standard is the same as those
proposed in sections III through V of
this part.
Three main assumptions were used in
determining how long a facility would
have to sample to achieve the minimum
levels specified in the manual methods.
They are assumptions for: sample flow
rate; flue gas oxygen content; and the
detection limit or specified sample
collection specified in the method. For
sample flow rate, EPA assumed a flow
rate of 0.5 scfm because this is either
what is directly stated as the flow rate
in the methods or it is used by
convention.
The Agency also assumed that the
oxygen concentration in the flue gas was
7 percent, the basis of today's standards.
Oxygen concentrations in the flue gas
can change greatly, but EPA believes
that the derived sampling time is elastic
relative to the assumed oxygen
concentration. In other words, the
sampling times would change roughly
five to ten per cent over the range of
oxygen concentrations experienced by
HWCs. This is not significant relative to
other assumptions made here, so a 7
percent oxygen concentration was
assumed.
Finally, each method specifies a
minimum analytical detection limit or
sample collection. We assumed that a
test operator would collect three times
what is prescribed in the method to
account for facility variability,
unknowns at a given site, etc. This is a
conventional approach used by testing
contractors. This will be referred to
below as the "collected sample."
There are other issues which need to
be addressed as well. One GEMS can be
used to comply with more than one
standard and standards can vary from
subcategory to subcategory. Therefore,
EPA proposes that the sampling time
used to derive the averaging period be
the longest sampling time which relates
to the GEM averaging period. For an
example, see the discussion on the Hg
and multi-metals GEM standards, below.
Manual methods tests do not run on-
the-hour, so an averaging periods with
some fraction of an hour would result if
rounding were not used. EPA believes it
is reasonable and simpler to have
integer value hourly averages. Since the
direct measure of a standard at the stack
is at the top of the monitoring hierarchy,
a less conservative approach is
warranted in this case, so EPA proposes
that averaging periods for GEMS be
rounded up to the nearest hour. (See
section II. A. of Part Five for more .
information on the monitoring
hierarchy.)
Also, a resulting averaging period may
be inappropriately short, i.e., less than
one hour. In this case EPA would
establish an averaging period of one-
hour. This is reasonable since the
averages for operating parameters to
control average emissions are one-hour.
(See section II.B.l. of Part Five for a
discussion of averages for operating
parameters.) Monitoring of a standard
continuously at the stack is at the top of
the monitoring hierarchy, while
establishing operating parameter limits
is at the bottom. It would be
inconsistent if an averaging period for
GEMS were less than those for operating
parameter limits, so a one-hour average
will be proposed in this case.
For mercury (Hg) and multi-metal
GEMS, it is proposed that the averaging
period be ten hours. SW—846 method
0060 would be the manual method used
to comply with these standards if a CEM
were not used. Emission standards for
these HAP categories vary greatly from
HAP-to-HAP and within a HAP, from
subcategory-to-subcategory. But the
proposed SVM standard for LWAKs
results in the longest sample collection
time. EPA believes that an LWAK will
have to sample for approximately 200
minutes per run to collect 15 ug of
sample to be in compliance with the
LWAK SVM standard. Three runs of 200
minute duration is 600 minutes, or ten
bom's.
For the HC1 and C12 standard, it is
proposed that the GEMS averaging
period be one hour. In this case, EPA
has determined that a facility would
have to sample less than ten minutes
per run to collect the minimum amount,
300 u.g, of sample specified by the
method. If three times this sampling
time were used to establish the
averaging time, it would result in one of
roughly 30 minutes. This is
unreasonable for a GEMS averaging
period, so EPA is proposing that the
averaging period be one hour.
Finally, it is proposed that the PM
GEMS averaging period be two hours.
This is because a facility would have to
sample for roughly 30 minutes per run
to collect the minimum amount, 30 mg,
of particulate specified by the method.
Three times this sampling time is 1.5
hours, so after rounding an averaging
period of two hours is proposed.
Table IV.2.1 summarizes the GEMS
averaging period for the various GEMS
emission standards.
TABLE IV.2.1 .—AVERAGING PERIODS
FOR GEMS STANDARDS
HAP or standard
PM
Mercury (Hg)
SVM
LVM
HCl and CI2
CO
HC
GEMS
averaging
period
2 hours.
10 hours.
10 hours.
10 hours.
1 hour.
1 hour.
1 hour.
d. All Averages are Rolling Averages.
All GEMS averaging periods are on a
rolling-basis. In other words, each time
a sample is recorded, a new rolling
average is calculated using the new
sample and all previous samples
obtained during the specified averaging
period. If sample results are recorded
every minute and the averaging period
is one hour, then the most recent sample
is averaged together with the results of
the previous 59 samples to obtain the
hourly rolling average. When there are
not enough data to obtain a rolling
average, one of two approaches would
be used. We propose that for short-term
interruptions of the rolling average that
the rolling average "pick-up" where it
left off, i.e., consider the one-minute
average immediately prior to the
interruption to be the one minute
average that occurred prior to the
current one-minute average. For longer
term interruptions, all available one
minute averages would be averaged
together until the time period since the
start of the rolling average equals the
averaging period for that parameter.
Then there is enough data to perform
the rolling average as usual, and the
rolling average would continue as
normal. For more information on the
use of GEMS and the rolling average, see
Part Five, Section II.C. "Compliance
Monitoring Requirements" and the
proposed regulations, Appendix J to
Part 60.
3. Feedstream and Operating Limits
Today, EPA is proposing specific
monitoring requirements to ensure
facilities are in compliance with the
standards during normal operation.
Some of these monitoring requirements
require setting limits on feedstream or
operating parameters. These limits will
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17381
be set on an average. Other limits would
be instantaneous limits, such as those
for fugitive process emissions.
It is proposed that four averaging
periods be used for feedstream and
operating limits: twelve hour, one hour,
ten minutes, and instantaneous. All
averages would be calculated on a
rolling-average basis with measurements
taken every 15 seconds to obtain a one
minute average. The one minute
averages are used to obtain the twelve
hour, one hour or ten minute rolling
average. The use of one-minute
averages, i.e., the average of the
previous 15 second averages within that
minute, is the current practice for
HWCs. "Instantaneous" limits are just
that, values not to be exceeded at any
time. Averaging does not occur for
"instantaneous" values. These
definitions supersede requirements in
the Part 63 general provisions, which
are less stringent. Consult chapter 5,
volume IV of the Technical Background
Document for more information
regarding EPA's choice of the time
duration for averaging periods.
For discussion on what operating
limits EPA is proposing and what the
averaging period will be for particular
operating limits, see section II of Part
Five of this preamble.
III. Incinerators: Basis and Level for the
Proposed NESHAP Standards for New
and Existing Sources
Today's proposal would establish
maximum achievable control
technology (MACT) emission standards
for dioxins/furans, mercury,
semivolatile metals (cadmium and lead),
low volatile metals (arsenic, beryllium,
chromium and antimony), hydrochloric
acid and chlorine (combined),
particulate matter, carbon monoxide,
and hydrocarbons from existing and
new hazardous waste incinerators
(HVVIs). See proposed § 63.1203. The
following discussion addresses how
MACT floor and beyond-the-floor (BTF)
levels were established for each HAP,
and EPA's rationale for the proposed
standards. The Agency's overall
procedural approach for MACT
determinations has been discussed in
Part Three, Sections V and VI for
existing sources and in Section VII for
new sources.
To conduct the MACT floor analyses
presented today, the Agency compiled
available data from hazardous waste-
burning incinerators: both commercial
as well as on-site facilities. As discussed
earlier, the vast majority of these data
were generated during trial burns to
demonstrate compliance with existing
RCRA standards at 40 CFR Part 264,
Subpart O. Therefore, the data were
obtained under proper QA/QC
procedures. These emissions data,
however, represent worse-case
emissions that cannot be exceeded
(because limits on operating parameters
are based on operations during the trial
burn). As noted earlier, the Agency
invites commenters to submit data that
reflect more normal, day-to-day
operations and emissions. This will
enable the Agency, among other things,
to be better able to distinguish among
facilities that are now included in the
expanded MACT floor pool but which,
upon closer inspection and with better
data, may not be actually employing the
identified floor controls.
A. Summary of MACT Standards for
Existing Incinerators
This section summarizes EPA's
proposed emission levels for existing
incinerators for each HAP, HAP group,
or HAP surrogate. The proposed
emission standards for HWIs are
presented in the table below:
TABLE IV.3.A.1.—PROPOSED MACT
STANDARDS FOR EXISTING INCINER-
ATORS
HAP or HAP surrogate
Dioxin/furans
Particulate Matter
Mercury i
SVM [Cd, Pb]
LVM [As, Be, Cr, Sb] ....
HCI + CI2
CO . .
HC
Proposed stand-
ards1
0 20 ng/dscm TEQ
0.030 gr/dscf
(69 mg/dscm).
50 ng/dscm
270 ng/dscm
210 ng/dscm.
280 ppmv
1 00 ppmv
1 2 DDimv.
1 All emission levels are corrected to 7 per-
cent O2.
1. Dioxins and Furans (D/Fs)
a. MACT Floor. The Agency's analysis
of dioxin/furan (D/F) emissions from
HWCs and other combustion devices
(e.g., municipal waste combustors and
medical waste incinerators) indicates
that temperature of combustion gas at
the inlet to the particulate matter (PM)
control device can have a major effect
on D/F emissions.45 D/F emissions
generally decrease as the gas
temperature of the PM control device
decreases, and emissions are lowest
when the gas temperature of the PM
control device is below the optimum
temperature window for D/F
formation—450 to 650 "P.46 Given that
45 USEPA, "Draft Technical Support Document
For HWC MACT Standards, Volume HI: Selection
of Proposed MACT Standards and Technologies",
February 1996.
46 For example, during compliance testing of a
cement kiln, D/F emissions exceeded 1.7 ng/dscm
(TEQ) at a ESP temperature of 435° F.
incinerators are equipped with both wet
and dry PM control devices that operate
under a range of temperatures, the
Agency is identifying a MACT floor for
D/F based on temperature control at the
inlet to the PM control device.
Incinerators emitting D/F at or below
levels emitted by the median of the best
performing 12 percent of incinerators
have combustion gas temperatures
below 400° F. These best performing
sources were equipped with venturi
scrubbers to control PM. The gas
temperature of the wet air pollution
control system for one source was 163°
F; gas temperature data for the other
best performing sources were not
available. Although gas temperatures at
a wet PM control device would
normally be less than 200° F,
temperatures could be higher in the
presence of acid gases such as HCI and
SO2. Consequently, the Agency believes
that it would be reasonable and
appropriate to generalize that gas
temperatures of wet PM control devices
are less than 400° F.
The Agency evaluated D/F emissions
from all incinerators that are equipped
with wet PM control systems. Average
D/F emissions for test conditions ranged
from 0.01 ng/dscm,(TEQ) to 39 ng/dscm
(TEQ). D/F emissions were as high as
3.5 ng/dscm (TEQ) for incinerators that
were not burning substantial levels of
known D/F precursors or were not
equipped with a waste heat boiler
(WHB). (It is hypothesized that WHB-
equipped incinerators may have high
(uncontrolled) D/F emissions because
D/F may be formed on particulate
' attached to boiler tubes as combustion
gases pass through the optimum
temperature window (450^650° F) for
D/F formation.) WHB-equipped
incinerators using wet PM control
devices had D/F emissions ranging from
0.4 to 8 ng/dscm (TEQ), and an
incinerator equipped with a wet PM
control device burning waste comprised
of approximately 30 percent PCBs had
D/F emissions of 39 ng/dscm (TEQ).
The Agency is consequently
identifying temperature control to below
400° F at the PM control device as the
MACT floor. Given that approximately
45 percent of test conditions in our
database have average D/F emissions
below 0.20 ng/dscm (TEQ), we believe
that it is appropriate to express the floor
as "0.20 ng/dscm (TEQ), or temperature
at the PM control device not to exceed
400° F". This would allow sources that
operate at temperatures above 400° F
but that achieve the same D/F emissions
as 45 percent of sources that operate
below 400° F to meet the standard
without incurring the expense of
-------
27382
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
lowering the PM control device gas
temperature. ' • -
EPA estimates that 75 percent of
incinerators are currently meeting the
floor level. The annualized cost for the
remaining incinerators to reduce D/F
emissions to 0.20 ng/dscm (TEQ) or
control gas temperature at the PM
control device to below 400° F would be
$3.0 million. Achievement of the floor
levels would reduce D/F TEQ emissions
nationally by. 35 g/yr.
b. Beyond-the-Floor (BTF)
Considerations. The Agency has
identified activated carbon injection (CI)
operated at gas temperatures less than
400° F as BTF control for D/F for
incinerators.47 CI is currently used by a
Commercial hazardous waste
incinerators to achieve emission levels
routinely [based on quarterly stack
testing) of less than 0.20 ng/dscm (TEQ).
CI is also used to reduce D/F emissions
from several municipal and medical
waste incinerators (MWIs) in a similar
manner.
CI has been demonstrated to be, ,
routinely effective at removing greater
than 95 percent of D/F and some tests
have demonstrated a removal efficiency
exceeding 99 percent at gas
temperatures of400b F or below.48 To
determine a BTF emission level, the
Agency considered the emission levels
that could result from gas temperature
control to less than 400° F combined
withCI.
To estimate D/F emissions with
temperature control combined with CI,
the Agency considered the range of
emissions from sources in the MACT
floor database, as discussed above.
Incinerators that are not equipped with
a WHB and not burning high levels of
D/F precursors (the vast majority of
incinerators) could be expected to
achieve D/F emissions of less "than 3.5
ng/dscm (TEQ) with temperature
control only. These sources could be
expected to achieve D/F emissions of
below 0.18 ng/dscm'(TEQ) when using
CI assuming a fairly conservative
removal efficiency of 95 percent.
There are three sources in our
database equipped with WHBs. One
currently uses CI to achieve D/F
emissions below 0.20 ng/dscm (TEQ)
when controlling PM with an ESP
operating below 400° F. Another source
47 We note that incinerators using wet PM control
systems would need to reheat the combustion gas
before injecting the carbon. This is because CI is not
efficient at D/F (or Hg) removal at gas temperatures
below the dew point. Gas reheating in these
situations was considered in estimating the cost of
compliance with the proposed standards.
•"USEPA, "Draft Technical Support Document
For HWC MACT Standards, Volume III: Selection
of Proposed MACT Standards and Technologies",
February 1996.
had D/F emissions of 0.56 ng/dscm
(TEQ) when controlling PM with a wet
system. This source could be expected
to achieve D/F emissions below 0.03 ng/
dscm (TEQ) using CI at a removal
efficiency of 95 percent. The third
WHB-equipped incinerator in our
database had D/F emissions of 8.0 ng/
dscm (TEQ) when controlling PM with
a wet system. This source could be
expected to achieve D/F emissions
below 0.40 ng/dscm using CI at a
removal efficiency of 95 percent. We
note, however, that the feed to this
source during testing comprised
approximately 10 percent
hexachlorophenol, a D/F precursor.-
Finally, one incinerator in the
database that controlled PM with a wet
system had D/F emissions of 39 ng/
dscm (TEQ). This source could be
expected to achieve D/F emissions
below 2 ng/dscm (TEQ) when using CI
at 95 percent efficiency. We note,
however, that the feed to this source
during testing comprised approximately
30 percent PCBs, known D/F precursors.
The Agency has considered^ this
information and determined that it
would be reasonable and appropriate to
establish 0.20 ng/dscm (TEQ) as an
emission level that is achievable with
BTF control. Although two sources in
our database that fed (during testing)
high levels of D/F precursors may not
have been able to achieve that level if
they had been equipped with CI, we
believe that those sources could achieve
a level of 0.20 ng by reducing the
feedrate of D/F precursors.
We note that, oecause we have
assumed a fairly conservative CI
removal efficiency of 95 percent to
identify the 0.20 ng/dscm BTF level, we
believe that this adequately accounts for
emissions variability that would be
experienced at a given source
attempting to operate under constant
conditions (e.g., as during a
performance test). That is, because CI
removal efficiency is likely to be up to
or greater than 99 percent, we believe
that it is not necessary to add a
statistically-derived variability factor to
the 0.20 ng/dscm BTF level to account
for emissions variability. Accordingly,
the 0.20 ng/dscm (TEQ) BTF level is
proposed as the emission standard.
We invite comment on this issue, and
note that if a statistically-derived
variability factor were deemed
appropriate, the BTF level of 0.20 ng/
dscm would be expressed as a standard
of 0.31 ng/dscm (TEQ). We note,
however, that under this approach, it
may be appropriate to use a less
conservative CI removal efficiency (i.e.,
because emissions variability would be
accounted for using statistics rather than
in the engineering decision to use a
conservative CI removal efficiency),
thus lowering the 0.20 ng/dscm level to
approximately 0.1 ng/dscm (TEQ). If so,
the BTF standard would be
approximately 0.21 ng/dscm (TEQ) (i.e.,
virtually identical to the proposed
standard) after considering a
statistically-derived variability factor.
EPA estimates that 50 percent of
incinerators are currently meeting a BTF
level of 0.20 ng/dscm (TEQ). The
incremental annualized cost for the
remaining incinerators to meet this BTF
level rather than comply with the floor
controls would be $26.2 million, and
would provide an incremental national
reduction of 38 g/yr in D/F TEQ
emissions over the floor level. This
represents an overall reduction of about
95 percent compared to baseline D/F
emissions of 77 g/year.
EPA has determined that proposing a
BTF MACT standard is warranted and a
number of factors support the proposed
BTF level of 0.20 ng/dscm (TEQ). D/F
are some of the most toxic compounds
known due to their bioaccumulation
potential and wide range of health
effects at exceedingly low doses,
including carcinogenesis. Exposure via
indirect pathways was in fact a chief
reason Congress singled out D/F for
priority MACT control in section
112(c)(6). See S. Rep. No. 228,101st
Cong. 1st Sess. at 154-155 (1990). As
discussed elsewhere in today's
preamble (and as qualified by the
discussion below regarding small
incinerators), EPA's risk analysis
developed for purposes of RCRA in fact
shows that D/F emissions from
hazardous waste incinerators could pose
significant risks by indirect exposure
pathways and that these risks would be
reduced by BTF controls. EPA is
expressly authorized to consider this
non-air environmental benefit in
determining whether to adopt a BTF
level. CAA section 112(d)(2).
As discussed in Part Seven of the
preamble, the cost-effectiveness of the
BTF level for small on-site incinerators
may be high. This is because on-site
incinerators are generally smaller than
commercial incinerators, have lower gas
flow rates, and therefore have lower
mass emission rates of D/F. Thus, the
cost per gram of D/F TEQ removed for
small incinerators is greater than for
large (on-site and commercial)
incinerators. Accordingly, the Agency
invites data and comment on: (1)
whether the BTF level is cost-effective
for small incinerators; and (2) whether
the final rule should establish MACT
standards at the floor level (i.e., 0.20 ng/
dscm (TEQ), or 400° F) for these small
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17383
incinerators.4950 Under this approach,
the Agency would use the same
definition of small incinerator used to
identify incinerators subject to less
frequent performance testing—
incinerators with gas flow rates less
than 23,127 acfin.51
EPA notes further that the control
technology on which the proposed BTF
standard is based, carbon injection, also
controls mercury. The ability and
efficiencies of controlling two such high
toxicity HAPs with the same highly-
efficient control technology is an
important factor in the Agency's
decision to propose a BTF standard. The
Agency notes further that the absolute
cost of achieving the proposed standard
is relatively low, particularly
considering the toxicity of D/F (as well
as mercury, which, as just noted, would
also be controlled). For example, the
proposed BTF levels would result in
annualized costs of $27 million to all
HWIs or $15 per ton of hazardous waste
burned.
Finally, EPA's initial view is that it
may be necessary to adopt further
controls under RCRA to control D/F if
it did not adopt the BTF level. This
would defeat one of the purposes of this
proposal—to avoid imposing emission
standards under both statutes for these
sources wherever possible. These risks
would, however, be reduced to
acceptable levels if emission levels are
reduced to the proposed BTF level of
0.20 ng/dscm (TEQ).
2. Particulate Matter
a. MACT Floor. The Agency has a
database for PM emissions from 74
HWIs that indicates a range (by test
condition average) from 0.0003 gr/dscf
to 1.9 gr/dscf. For MACT determination,
the median of the best performing 12
percent of the HWIs in the MACT pool
were analyzed and found to be using the
following APCDs to control PM: (1) A
fabric filter (with an air to cloth ratio of
less than 10.0 acfm/ft2); and (2) an
ionizing wet scrubber (IWS) in
•"See also discussion in Part Four, Section I
(Selection of Source Categories and Pollutants),
regarding whether the Agency should subdivide
Incinerators by size and promulgate separate floor
standards (and BTF standards, if warranted).
50 If after review of comments and further analysis
t'no Agency determines that subdividing
incinerators is not appropriate but, because of cost-
offoctlvencss considerations, BTF levels are not
warranted for all types of incinerators, the Agency
Invites comment on whether such cost-effectiveness
and BTF decisions should be based on incinerator
size or whether the incinerator is a commercial or
on-silo unit.
91 We also use this definition to request
(elsewhere in the text) comment on whether the
requirement to use Hg and PM GEMS for
compliance monitoring should be relaxed or waived
for small incinerators.
combination with a venturi-scrubber.
Accordingly, these APCDs were
tentatively designated as the MACT
floor technologies. To identify an
emission level that these technologies
could be expected to achieve routinely,
the Agency examined the emissions
from all incinerators (in the database)
that were equipped with these PM
control devices. A MACT floor level of
240 mg/dscm (0.107 grains/dscf)
resulted from the analysis based on
considerations discussed in Part Three,
Section V, above.
This level, however, is higher than the
current federal standard of 180 mg/dscm
(0.08 grains/dscf).52 Thus, the Agency is
not proposing to use the statistically-
derived approach to identify the MACT
floor emission level. The Agency has
regulated PM emissions from hazardous
waste incinerators under RCRA (40 CFR
264.343(c)) since 1981 and all RCRA-
permitted incinerators have been
required to meet the federal standard of
0.08 gr/dscf (180 mg/dscm). The
Agency, therefore, is identifying the
MACT floor at the regulated level of 180
mg/dscm.
The APCDs commonly used at HWIs
to control PM to the current RCRA
standard are fabric filters, ESPs, IWSs,
and venturi-scrubbers. Accordingly, we
have designated these technologies as
MACT floor for PM control.
Approximately 95 percent of all test
conditions in our database have lower
average levels (average over all runs of
the test condition) than the MACT floor
level of 180 mg/dscm.53 This MACT
floor level will not impose any
incremental burden on HWIs (except
compliance and related permitting
costs) since it is the currently
enforceable level.
b. Beyond-the-Floor Considerations.
The Agency considered two levels of
more stringent BTF PM standards, 69
and 34 mg/dscm (0.03 and 0.015 gr/
dscf), since well designed and well
operated ESPs, IWSs, and fabric filters
can routinely achieve PM control at the
69 mg/dscm level,54 while state-of-the-
art ESPs, IWSs and FFs can achieve 34
mg/dscm level. The Agency is
52 This anomalous result is apparently
attributable to: (1) inability to consider emissions
from only those HWIs truly using MACT floor
control (because of inadequate data to properly
characterize the design, operation, and maintenance
of the control device); and (2) use of a variability
factor that is based on emissions variability (during
trial burn testing) that may be much higher than
many sources actually experience.
53 We presume that those few test conditions that
exceeded the 180 mg/dscm standard occurred
during failed trial burn tests.
54 We note also that, as discussed in the next
section, cement kilns with much higher inlet
particulate loadings are currently required to meet
a 69 mg/dscm standard.
proposing a BTF standard of 69 mg/
dscm (0.03 grains/dscf) based on
engineering evaluation of the emissions
data from HWIs. (We note that, as
discussed in Sections IV and V below,
it also is consistent with the proposed
standards for cement kilns and LWAKs).
Most of the HWIs having PM emissions
between 69 to 180 mg/dscm (0.03 to
0.08 gr/dscf) range are likely to be using
older APCDs that can be upgraded to
provide better PM control. Only 30
percent of all test conditions 55 in our
database were found to have PM
emissions greater than the proposed
BTF level of 69 mg/dscm (0.03 gr/dscf).
Analysis of the test data appeared to
indicate that some sources operated
under poor, non-normal conditions
during one test condition resulting in
high PM levels, while much lower PM
emissions were achieved during other
test conditions. As noted elsewhere, the
Agency is specifically concerned that
the nature of these test data (and the
absence of more detailed, routine
operations and emissions data) has
interfered with our ability to derive
MACT standards that appropriately
reflect the lower, day-to-day emissions
achievements of the best performing
facilities. The Agency will continue to
refine its analysis in this regard, and we
specifically invite data and comments
on this issue.
The Agency estimates that 9 percent
of existing incinerators can achieve the
proposed BTF levels using design,
operation and maintenance upgrades of
their APCDs, while 11 percent facilities
would require installation of new fabric
filters or other equivalent APCD (e.g.,
ESP or IWS). The national annualized
cost to HWIs to comply with the
proposed BTF level would be $2.7
million and would provide an
incremental reduction of PM emissions
of 839 tons/year (52 percent) from the
baseline emissions level of 1606 tons/
year. Accordingly, the Agency believes
that a BTF level of 69 mg/dscm (0.03 gr/
dscf) is appropriate.
The performance of many APCDs can
be improved to achieve a more stringent
PM BTF level of 34 mg/dscm by
adopting good D/O/M practices; in other
cases, the APCD may have to be
upgraded or replaced. Upgrades include
techniques for ESPs such as
humidification or increasing the plate
area or power input, and for FFs,
increasing cloth to air ratio and pressure
drop across bags, or retrofits to modern
fabrics like heavy woven fiberglass. The
Agency is concerned, however, that the
cost of such retrofitting to achieve PM
levels of 34 mg/dscm (0.015 gr/dscf)
55 Representing 20 percent of the sources.
-------
17384
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
could be substantial. We also note that
PM is not a HAP, but rather a surrogate
for non-dioxin/furan HAPs adsorbed on
to PM and for metal HAPs not directly
controlled by a MACT standard. These
HAPS would be controlled to some
extent by other proposed standards (e.g.,
metal-specific standards; CO and HC
limits to control organic HAPs). For
these reasons, we believe that
controlling PM to the proposed BTF
level of 69 mg/dscm (0.03 gr/dscf) is
appropriate. In addition, we also note
that the Agency has no information that
a lower PM standard would be needed
to satisfy RCRA requirements.
3. Mercury
a. MACT floor for mercury. Mercury.
(Hg) emissions from incinerators are
currently controlled by controlling the
feedrate of Hg and by using wet
scrubbers (although such scrubbers are
used primarily for acid gas control). Wet
scrubbers can remove soluble forms of
mercury species (e.g., HgCl).
The Agency's Hg emissions database
from 29 HWIs indicates that baseline Hg
emissions range from 0.05 |ig/dscm to a
high of 1,3.60 ng/dscm. To identify
MACT floor control, EPA determined
that sources with Hg emissions at or
below the level emitted by the median
of the best performing 12 percent of
sources were controlling Hg using
either: (1) Hg feedrate control expressed
as a maximum theoretical emission
concentration (MTEC)56 of 19 ng/dscm;
or (2) wet scrubbers coupled with an
MTEC of 51 ng/dscm. Analysis of
emissions from all incinerators in the
database using these or better controls
(i.e., lower Hg feedrates expressed as
lower MTECs) resulted in a MACT floor
level of 130 ng/dscm.57 To meet this
floor level 99 percent of the time, EPA
estimates that a source with average
emissions variability must be designed
and operated to routinely meet an
emission level of 5 7 ng/dscm.
EPA estimates that approximately 70
percent of incinerators currently meet
the floor level. The annualized cost for
the remaining incinerators to meet the
floor level is estimated to be $29.5
million, and would reduce Hg emissions
nationally by 7,166 Ibs per year from the
baseline emissions level of 9,193 Ibs per
year.
b. Beyond-the-Floor Considerations.
The Agency has considered two
alternative beyond-the-floor (BTF)
controls for unproved Hg control: flue
gas temperature reduction to 400° F or
less followed by either activated carbon
injection (CI) or carbon bed (CB). (As
discussed in the D/F section, we note
that incinerators with PM control
devices operating below the dew point
(e.g., venturi-scrubbers, ionizing wet
scrubbers) would have to reheat the
combustion gas before using CI, and
would need to add a FF or other PM
control device to remove the injected
carbon.) EPA believes that Cl-controlled
systems can routinely achieve Hg
emission reductions of 90 percent or
better and that CB-controlled systems
can routinely achieve Hg emissions of
99 percent or better.58
For Cl-controlled systems, EPA has
identified a BTF emission standard of
50 jig/dscm, assuming first that a source
has controlled its Hg emissions to only
300 [ig/dscm using a wet scrubber and/
or feed control, and second, a CI
removal efficiency of 90 percent. (The
BTF emission standard corresponds to a
design level of 30 (ig/dscm, i.e., a level
that the device is designed and operated
to achieve routinely.)59 For CB systems,
the BTF standard would be 5.0 jig/dscm
(assuming 99 percent removal
efficiency).
We note that another option for
identifying BTF levels would be to
consider the CI or CB system as an add
on to the floor controls identified above.
Under this option, emission levels prior
to CI would be assumed to be the floor
level, 130 ng/dscm. Thus, a CI system at
90 percent removal could be expected to
achieve a standard of approximately 13
Hg/dscm. A CB system at 99 percent
removal could be expected to achieve a
standard of approximately 1.3 ng/dscm.
We specifically request comment on
whether this approach of applying BTF
reductions to the floor levels is
appropriate.
We also note that an alternative
approach to using a statistically-derived
variability factor to account for
emissions variability would be to
assume a conservative control efficiency
for the CI or CB BTF technology. We
believe that using a conservative
removal efficiency could adequately
account for emissions variability. Under
=6 MTEC is the Hg feedrate divided by the gas
flow rate, and is an approach to normalize Hg •
feedrate across sources.
57 As discussed above in the text, we added a
within-test condition emissions variability factor to
the log-mean of the runs for the test condition in
the expanded MACT pool with the highest average
5BUSEPA, "Draft Technical Support Document
For HWC MACT Standards, Volume HI: Selection
of Proposed MACT Standards and Technologies",
February 1996. See also memo from Shiva Garg,
EPA, to the Docket (No. F-96-RCSP-FFFFF), dated
February 22,1996, entitled "Performance of
Activated Carbon Injection On Dioxin/Furan and
Mercury Emissions."
59 To achieve a standard of 50 ug/dscm 99 percent
of the time, a source with average emissions
variability must be designed and operated to
achieve an emission level of 30 ug/dscm.
this approach, we would conservatively
assume that Cl-controlled systems could
achieve a removal efficiency of 80
percent and that CB-controlled systems
could achieve an efficiency of 90
percent. When these removal
efficiencies are applied to the floor level
of 130 |ig/dscm (corresponding to a
design level of 57 ng/dscm), this would
result in emission standards of 11 |ig/
dscm for Cl-controlled systems, and 5.7
ug/dscm for CB-controlled systems.60
We invite comment on this alternative
approach to account for emissions
variability among runs within a test
condition.
For the reasons discussed below, EPA
believes that a BTF level based on use
of CI is warranted and is proposing a
MACT standard of 50 ng/dscm. The
proposed standard would result in
nationwide Hg emissions reductions of
757 Ibs per year above the floor level
and 7,922 Ibs per year from baseline
levels, and the incremental annualized
cost to achieve the BTF level over the
floor level would be $7.7 million.
EPA has considered costs in relation
to emissions reductions and the special
bioaccumulation potential that Hg poses
and determined that proposing a BTF
limit is warranted. Hg is one of the more
toxic metals known due to its
bioaccumulation potential and the
adverse neurological health effects at
low concentrations especially to the
most sensitive populations at risk (i.e.,
unborn children, infants and young
children). Congress has singled out
mercury in CAA section 112(c)(6) for
prioritized control. A more detailed
discussion of human health benefits for
mercury can be found in Part Seven of
today's proposal. The chief means of
control, activated carbon injection, also
controls D/F so that there are distinct
efficiencies in control.61
The Agency evaluated a more
stringent standard of 8 (ig/dscm for Hg
emissions based on CB technology. This
standard would result in additional
national Hg reductions of 960 Ibs per
year over the proposed standard of 50
60The same approach could be applied to the
previously discussed approach of applying the BTF
control to an assumed emission level of 300 ug/
dscm. When assuming the conservative removal
efficiencies of 80 percent for CI and 90 percent for
CB, this would result in BTF standards of 60 ug/
dscm for Cl-controlled systems and 30 ug/dscm for
CB-controlled systems. A statistically-derived
variability factor would not be added because
emissions variability is accounted for by assuming
conservative (i.e., lower-than-expected) removal
efficiencies for CI and CB systems.
61 As discussed for D/F, we invite comment on
whether the final rule should establish floor levels,
rather than BTF levels, for Hg for small incinerators.
This is because the Agency is concerned about the
cost-effectiveness of the BTF levels for small
incinerators.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17385
Hg/dscm at an incremental annualized
national cost of $20 million. The
Agency does not believe that a CB-based
emission level of 8 ug/dscm would be
appropriate.
4. Semivolatile Metals (SVM) (Cadmium
and Lead)
a. MACT Floor. Emissions of SVMs
from HWIs are currently controlled by
PM control devices. In addition, some
incinerators have specific emission
limits for these metals established under
RCRA omnibus permit authority. The
Agency has a database for SVM
emissions from 42 HWIs, which
indicates a range (by test condition
average) from a low of 1.46 to a high of
29,800 ng/dscm. For the MACT analysis,
the median of the best performing 12
percent of HWIs were found to be using:
(1) a venturi-scrubber (VS)62 with a
MTEC level of 170 ug/dscm; (2) a
combination of ESP and WS with a
MTEC level of 5,800 ug/dscm; and (3) a
combination of VS and IWS with a
MTEC of 49,000 ug/dscm.63
Accordingly, we identified these
technologies as MACT floor.
To identify an emission level that
these technologies could routinely
achieve, we evaluated the emission
levels from all HWIs equipped with
these contrdls.64 We identified the test
condition in this expanded MACT pool
with the highest average emission and
used procedures discussed above in Part
Three, Section V, (i.e., addition of a
within-test condition emissions
variability factor to the log mean of the
runs for this test condition) to identify •
a MACT floor level 270 ug/dscm.
We estimate that approximately 65
percent of all incinerators currently
meet this MACT floor level. Sources not
already meeting the floor level can
readily achieve it by making design,
operation, or maintenance
improvements to their existing PM
control system or by retrofitting with a
new PM control device.
The national annualized cost to HWIs
to comply with the proposed floor level
is estimated to be $9.9 million, and
MBecause virtually all other PM control devices
(e.g., ESP, FF, 1VVS) would be expected to have a
SVM collection efficiency equivalent to or better
than a VS, a source equipped with any PM control
device and having a MTEC less than 170 ug/dscm
was considered to bo using MACT floor control.
M Wo considered a FF to have equivalent (or
bettor) SVM removal efficiency compared to an
IWS. Thus, wo considered a source equipped with
a FF and any wet scrubber (ahead of the FF) and
having a MTEC less than 49,000 ug/dscm to be
using MACT floor control. A FF alone may not
provide equivalent control of SVM because SVM
can bo volatile In stack emissions.
64 Sources with belter controls (MACT technology
and lower foedrate expressed as MTEC) were also
included in tho expanded MACT pool.
would provide a reduction in Cd and Pb
emissions of 50 tons/year, a 94 percent
reduction in emissions.
b. Beyond-the-Floor Considerations.
The Agency is not proposing a more
stringent BTF standard for SVM. We
note that the floor level alone would
provide for a 94 percent reduction in
emissions, and emissions at the floor are
not likely to trigger the heed for
additional control for these sources
under RCRA.
5. Low Volatile Metals (Arsenic,
Beryllium, Chromium and Antimony)
a. MACT floor. The Agency has a
database for LVM emissions from 41
HWIs, which indicates a range (by test
condition average) from a low of 3.5 to
a high of 133,000 ug/dscm. For MACT
analysis, the median of the best
performing 12 percent of HWIs achieved
the LVM emission levels using: (1) a
venturi-scrubber (VS) for MTECs up to
1,000 ag/dscm; and (2) an ionizing wet
scrubber (IWS) for MTECs up to 6,200
Ug/dscm. Accordingly, we identified
these technologies as MACT floor.
In addition, we consider any PM
control device to provide equivalent
LVM control to a VS. We therefore
identified an ESP, IWS, or FF with a
MTEC up to 1,000 ug/dscm as MACT
floor control. Similarly, we consider a
FF or ESP as equivalent technology to
a IWS. Thus, a FF or ESP coupled with
a MTEC up to 6,200 ug/dscm is also
considered MACT floor control.
To identify an emission level that
these technologies could routinely
achieve, we considered the emissions
from all HWIs in our database equipped
with MACT floor control. We identified
the test condition in this expanded
MACT pool with the highest average
emissions and added a within-test
condition emissions variability factor to
the log-mean of the test condition runs.
See Part Three, Section V, above.
Accordingly, we have identified a
MACT floor level of 210 ug/dscm.
Approximately 80 percent of all test
conditions in our database achieved the
MACT floor level even though many
HWIs were equipped with different
APCDs or had higher MTECs. EPA
believes that most HWIs would be able
to achieve the proposed MACT floor
without installing an add-on control
system. The control technologies
necessary to achieve the MACT floor
level are already being used by many
HWIs for PM and acid gas control.
The national annualized cost to HWIs
to comply with the floor level would be
$7.7 million and would provide an
incremental reduction in LVM
emissions of 25 tons/year (91 percent)
from the baseline emissions level of 27.3
tons/year.
b. Beyond-the-Floor Considerations.
The Agency is not proposing a more
stringent LVM standard using BTF
controls (i.e., better performing PM
control equipment). We note that the:
floor level alone would provide for a 91
percent reduction in emissions, and
emissions at the floor are not likely to
trigger the need for additional control
for these sources under RCRA.
6. Hydrochloric Acid and Chlorine
a. MACT floor for HC1/C12. The
Agency's database for HC1/C12 emissions
from 59 HWIs indicates a range (by test
condition average) from a low of 0.1 to
a high of 1068 ppmv (expressed as HC1
equivalents). For MACT analysis, the
median of the best performing 12 (
percent of HWIs achieving the lowest
HCl/Cla emission levels were found to
be using some kind of scrubbing using
combinations of absorber, ionizing wet
scrubber, VS, packed bed scrubber
(PBS), or generic wet scrubber. In
addition, the best performing sources
had a chlorine feedrate of up to 2.1E7
ug/dscm, expressed as a MTEC.
Accordingly, we identified MACT floor
•control as wet scrubbing coupled with
a chlorine MTEC up to 2.1E7 ug/dscm.
To identify an emission levelthat wet
scrubbing with an MTEC up to 2.1E7 ug/
dscm could routinely achieve, we
considered the emissions from all HWIs
in our database equipped with these
controls. We identified the test
condition in this expanded MACT pool
with the highest average emissions and
added a within-test condition emissions
variability factor to the log-mean of the
test condition runs. See Part Three,
Section V, above. Accordingly, we have
identified a MACT floor level of 280
ppmv.
Over 90 percent of all test Conditions
in our database achieve this MACT floor
level. At current baseline levels, HWIs
emit 1712 tons/year of HC1/C12, and at
today's proposed MACT standard, these
emissions would be reduced by 592
tons/year, a reduction of 35 percent. The
estimated annualized national cost to
the industry to meet the proposed
MACT standard would be $4.5 million.
b. Beyond the-Floor Considerations.
The Agency considered whether to
propose a BTF level and determined
that it would not be warranted. We note
that emissions at the floor are not likely
to trigger the need for additional control
for these sources under RCRA.
7. Carbon Monoxide and Hydrocarbons
As discussed in Section I above, the
Agency believes that establishing
emission limits and continuous
.
-------
17386
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
monitoring of two surrogate compounds
(hydrocarbons (HC) and carbon
monoxide (CO)) will help control
emissions of non-dioxin organic HAPs
(in combination with PM control to
control absorbed organic HAPs).
a. MACT Floor for HC. The Agency's
database for HC emissions from 31
HWIs indicates a range (by test
condition average) from a low of 0.2 to
a high of 35.8 ppmv. Unlike certain
cement kilns and LWAKs; incinerators
are not required to monitor HC under
RCRA regulations. Facilities generally
obtained HC emissions data for their
own information and often used an
unheated FID detector, in which soluble
volatiles and semivplatiles are
condensed out before entering the
detector. Also much of the data were .
based on run averages (as opposed to .
the maximum hourly rolling average
format proposed today).65
Notwithstanding these shortcomings,
the Agency used these data to identify
a MACT floor level.
The Agency identified MACT control
for HC as operating under good
combustion practices (GCPs). GCPs
include techniques such as thorough air,
fuel, and waste mixing, provision of
adequate excess oxygen, maintenance of
high temperatures to destroy organics,
design of the facility to provide high
enough residence times for destruction
of organics, operation of the facility by
qualified and certified operators, and
periodic equipment maintenance to
manufacturer-recommended standards.
To identify the MACT floor level, the
Agency conducted a quantitative
evaluation of the data combined with
engineering judgment to identify test
conditions that appear to be conducted
under good combustion conditions.
Since it is not possible to say with
certainty which test conditions were
conducted using GCPs absent a detailed
examination of all test conditions, we
conducted the analysis by arraying the
entire HC database from the lowest to
the highest emission levels. We then
assumed that test conditions beyond a
clear break-point were not operated
under GCPs. Based on the above
analysis and a statistical evaluation of
the level that the average source can
achieve 99 percent of the time, the
65 The average bf emissions over a run is lower
than the maximum hourly rolling average for the
run. In addition, unheated FIDs report lower HC
levels than a heated FID that would be required
under today's proposal. Both of these factors would
lead the Agency to underestimate the cost of
compliance! On the other hand, the HC levels in the
database were measured during worst-case, trial
burn conditions. Thus, these emissions are likely to
be much higher than during normal operations.
This factor has lead the Agency to overestimate
compliance costs.
Agency identified a MACT floor level of
12 ppmv.
We estimate that the annualized
burden on HWIs to meet this floor level
would be $8.5 million. An annual
reduction of 49 tons of HC emissions (20
percent) is expected from the baseline
levels of 239 tons/year.
EPA specifically invites comment on
the approach used to identify the MACT
floor level and requests HC data on a
hourly rolling average basis, using
heated FID monitors.
b. MACT floor for CO. RCRA
regulations for HWIs were promulgated
in 1981 and limit CO emissions to levels
achieved during the trial burn. (As
noted elsewhere, facilities typically
design trial bums to maximize CO in
order to provide operational flexibility.)
Most of our database for CO (from 59
facilities) is based on run-averages
during trial burns (rather than an hourly
rolling average-basis; see discussion
below). The CO levels in our database
that are on a run-average basis range
from 0.3 to 10,400 ppmv.
We are proposing today a maximum
hourly rolling average (MHRA) format
for CO (and HC), which is the same
format in which a standard of 100 ppmv
(Tier 1) was proposed in 1990 for HWIs
(see 55 FR17862 (April 7,1990)) and
promulgated for CKs and LWAKs in
1991 (see 56 FR 7134 (February 21,
1991)).
Although the Agency did not
promulgate a final rule for CO emissions
from HWIs (because of Agency resource
constraints), the Agency published a
guidance document66 wherein a Tier 1
CO limit of 100 ppmv HRA was
recommended for control of PIC
emissions if warranted on a site-specific
basis. Accordingly, subsequent trial
burns for HWIs have been conducted
using a HRA format for CO. Our CO
database in the HRA format is
comprised of 17 test conditions and has
a range of 10 to 1,500 ppmv.
For MACT determination, the Agency
conducted an analysis similar to that
described above for HC and a CO MACT
floor level of 120 ppmv resulted (e.g.,
MACT floor control is GCPs, and a
break-point analysis was used to
identify sources likely to be truly using
GCPs). Nonetheless, since the Agency
has previously proposed a CO limit of
100 ppmv and since this level is readily
achievable by well-designed and well-
operated HWIs, the Agency is proposing
100 ppmv HRA as the MACT floor.
We note that this floor level compares
favorably with CO standards for other
^USEPA, "Guidance on PIC Controls For
Hazardous Waste Incinerators", April 1990, EPA/
530-SW-90-040.
types of incinerators such as medical
waste incinerators for which the
proposed standard is 50 ppmv (60 FR
10654, February 27,1995), and mass
burn and fluidized bed municipal waste
incinerators for which the promulgated
CO standard is 100 ppmv (60 FR 65382,
December 19,1995).
The Agency estimates that at a 100
ppmv standard, national CO emission
reductions of 13,200 tons/year could be
achieved from the baseline level of
14,080 tons/year at an annualized
national cost of $17.4 million.
c. Beyond-the-Floor Considerations.
The Agency considered more stringent
BTF limits for CO and HC. Although
state-of-the-art HWIs operating under
GCPs should be able to routinely
achieve levels below 100 ppmv HRA for
CO and 12 ppmv HRA for HC, the
Agency is concerned that the
incremental compliance cost may not
warrant more stringent standards.
EPA invites comments specifically on:
(1) the use of CO and HC as surrogates
for non-dioxin organic emissions; and
(2) data and information and
suggestions on an approach to identify
a lower floor level for HC that more
accurately reflects the levels that are
being routinely achieved by HWIs
operating under GCPs.
8. MACT Floor and BTF Cost Impacts
The annualized national cost to
achieve the proposed standards is
estimated at $486,000 for each on-site
incinerator unit and $731,000 for each
commercial unit. The total (pre-tax)
national annualized cost is estimated to
be $90 million for on-site and $25
million for commercial incinerators.
These costs include a GEMS cost of
$130,000 per source annually. The most
expensive HAPs would be dioxins and
mercury, for which BTF levels have
been proposed, and would cost $3.0
million and $30 million respectively
nationally at MACT floor levels, and
$29.2 million and $37.2 million
respectively at BTF levels. These costs
include maintenance and operation of
the equipment and GEMS. GEMS
account for 18 percent of the total
compliance,cost. Details of these cost
estimates have been provided in
"Second Addendum to the Regulatory
Impact Assessment for Proposed
Hazardous Waste Combustion
Standards" and are based on no market
exit by any HWI and assuming that the
facilities have only a limited ability to
pass through the costs of the rule to
generators.
The Agency, however, estimates that
perhaps 4 of the 34 commercial facility
units and up to 51 of the 184 on-sife™-~
facility units would elect to cease
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17387
burning hazardous wastes as a result of
today's proposals. Most of these
facilities burn small quantities of
hazardous wastes. These facilities
would likely find it more economical to
transport the hazardous wastes to other
facilities, while perhaps continuing to
burn other non-hazardous and
industrial wastes, in lieu of incurring
expenditures to upgrade their units to
continue to burn that small quantity of
HVV under MACT standards. As such,
the total quantity of wastes burned
would not be affected since those wastes
would be burned by other HWCs, for
which there appears to be sufficient
capacity available.
B. Summary of MACT Standards For
New Incinerators
1. Basis for MACT New
According to Section 112 of CAA, the
degree of reduction in emissions
deemed achievable for new facilities
may not be less stringent than the
emissions control achieved in practice
by the best controlled similar unit. This
section summarizes EPA's rationale for
establishing MACT standards for new
HWIs. The methodology for determining
the standards for new incinerators is
similar to that for existing sources,
except that MACT floor control is based
on the single best performing
technology, and the MACT pool is
expanded to consider emissions from
any source using that technology. For
more details see "Draft Technical
Support Document for HWC MACT
Standards, Volume HI: Selection of
Proposed MACT Standards and
Technologies".
The Agency is proposing the
following standards for new HWIs:
TABLE IV.3.B.1—PROPOSED MACT
STANDARDS FOR NEW INCINERATORS
HAP or HAP surrogate
Dioxins/furans
Particulate matter
Mercury
SVM [Cd, Pb]
LVM [As. Be, Cr, Sb]
HCI + CI2
CO
HC
Proposed stand-
ard"
0.2 ng/dscm
TEQ.
69 mg/dscm
(0.030 gr/dscf).
50 ng/dsom.
62 ng/dscm.
60 ng/dscm.
67 ppmv.
100 ppmv.
12 ppmv.
•All emission levels are corrected to 7 per-
cent ©2.
2. MACT New for Dioxin/Furans
a. MACT New Floor. EPA examined
its emissions database and identified the
single best performing existing source,
and found mat the test condition with
the lowest PCDD/F TEQ emissions had
a test-condition average of 0.005 ng/
dscm. This facility employs a water
quench and wet scrubbing air pollution
control systems (APCSs). The D/F
emission control by this source is being
achieved by inhibiting the formation of
D/F in the APCD by rapid quench of the
hot gases from the combustion chamber.
Therefore, the Agency selected wet
scrubbing and low APCD inlet
temperature (400° F) as the MACT floor
control.
To determine an emission level that
this the floor control could be expected
to achieve, the Agency considered data
from all HWIs using the MACT floor
control. Using the same methodology as
used for identifying the floor level for
existing sources, the Agency identified
a MACT floor level of 0.20 ng/dscm
TEQ or an APCD inlet temperature of
400° F.
b. Beyond-the-Floor (BTF)
Considerations. As discussed above for
existing sources, the Agency selected
activated carbon injection (ACI) as the
BTF technology. ACI is routinely
effective in removing greater than 95
percent of D/F from flue gases. The
Agency had identified a BTF level of 0.2
ng/dscm TEQ for the same reasons
discussed above for the BTF standard
for existing sources.
The Agency also consider a carbon
bed as a BTF technology to achieve
lower emission levels. As discussed for
existing sources, however, the Agency is
concerned that the cost of carbon beds
may not be warranted given the
incremental emissions reduction over a
ACI-based BTF standard.
3. PM Standard for New HWIs
The single best performing source in
our database for PM emissions was a
source equipped with a FF having an air
to cloth ratio of 3.8 acfm/ft2. Thus, this
technology represents MACT new floor
control. When we considered emissions
data from all sources equipped with this
level of control (or better), we identified
a floor level of 0.039 gr/dscf.
The Agency considered more efficient
PM control (e.g., lower air-to-cloth ratio,
better bags) as BTF control that could
achieve alternative BTF levels of 0.03 or
0.015 gr/dscf. These are the same
controls investigated for BTF
considerations for existing sources.
The Agency is proposing the same
BTF standard for new sources as it is
proposing for existing sources—(69 mg/
dscm or 0.03 gr/dscf). This standard is
readily achievable. The Agency is not
proposing a 0.015 gr/dscf standard
because, as discussed for existing
sources, it is not clear that the
additional cost is warranted considering
the incremental reduction in PM.
4. Mercury Standard for New HWIs
a. MACT New Floor. The single best
performing source in our database for
Hg emissions was a source equipped
with a wet scrubber (WS) and having a
MTEC of 51 ng/dscm. The Agency
considered any wet scrubbing device an
equivalent control technology (when
coupled with a MTEC up to 51 u,g/dscm)
because of the ability to scrub soluble
forms of mercury species. Thus, the
Agency identified MACT new floor
control as any wet scrubber coupled
with a MTEC up to 51 ng/dscm. When
we considered emissions data from all
sources equipped with this level of
control, we identified a floor level of
115 ug/dscm.
b. Beyond-the-Floor Considerations.
As for existing sources, the Agency
considered the use of both activated
carbon injection (ACI) and carbon bed
(CB) as alternative BTF technologies.
We are proposing a BTF standard of 50
ug/dscin for new sources based on use
of ACI for the same reasons we are
proposing this standard for existing
sources.
5. Semivolatile Metals Standard for New
HWIs
a. MACT New Floor. The single best
performing source in our database for
SVM emissions was a source equipped
with a VS in combination with a IWS,
and having a MTEC of 49,000 ng/dscm.
The Agency considered a wet scrubber
in combination with a FF (coupled with
a MTEC up to 49,000 ng/dscm) to
provide equivalent or better control of
SVM. Thus, these technologies
represent MACT new floor control.
When we considered emissions data
from all sources equipped with this
level of control, we identified a floor
level of 240 u,g/dscm.
b. Beyond-the-Floor Considerations.
The Agency believes that state-of-the-art
FFs can achieve much lower emissions
of SVM. For example, the Agency has
determined that MWCs equipped with a
FF can achieve more than a 99 percent
reduction in SVM. See 59 FR 48198
(September 20,1994). Given that we
have identified a MACT new floor
(design) level for cement kilns of 35 ng/
dscm (see discussion in Section IV
below), we believe that a design level of
35 ng/dscm for HWIs is achievable,
reasonable, and appropriate. To ensure
that a source that is designed to meet a
SVM level of 35 ng/dscm can meet the
standard 99 percent of the time
(assuming the source has average
within-test condition emissions
variability for sources equipped with
-------
17388
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
ESPs and FFs), the Agency has
established a standard of 62 jig/dscm.
We note that SVM emissions at this
level are not likely to result in
additional regulation of these sources to
satisfy RCRA health risk concerns.
6. Low Volatile Metals Standard for
NewHWIs
a. MACT New Floor. The single best
performing source in our database for
LVM emissions was a source equipped
with a VS with an MTEC of 1,000 ng/
dscm. Given the LVM collection
efficiency of a VS, the Agency
considered any PM control device (e.g.,
ESP, IWS, FF) to provide equivalent or
better collection efficiency. Thus, these
technologies represent MACT new floor
control. When we considered emissions
data from all sources equipped with this
level of control, we identified a floor
level of 260 ug/dscm. (We note that this
floor level for new sources is higher
than the floor level proposed for
existing sources. Although the
statistically-derived emissions
variability factor was added to the same
test condition for both MACT existing
floor and MACT new floor, the
variability factor was greater for test
conditions in the MACT new expanded
pool.)
b. Beyond-the-Floor Considerations.
The Agency believes that state-of-the-art
PM control devices (e.g., ESPs, IWS,
FFs) can achieve LVM emission levels
well below the floor level. Given that we
have identified a floor (design) level67
for new CKs and new LWAKs of 35 jig/
dscm and 26 ug/dscm, respectively (see
discussion in Sections IV and V below),
we believe that a BTF design level of 3.5
ug/dscm is achievable, reasonable, and
appropriate for new HWIs. To ensure
that a source that is designed to meet a
LVM level of 35 ng/dscm can meet the
standard 99 percent of the time
(assuming the source has average
within-test condition emissions
variability for sources equipped with
ESPs and FFs), the Agency has
established a standard of 60 ug/dscm.
We note that LVM emissions at this
level are not likely to result in
additional regulation of these sources to
satisfy RCRA health risk concerns.
As discussed elsewhere in today's
proposal, we are encouraging but not
requiring sources to document
compliance with the metals standard
using a multi-metal continuous
monitoring system (GEMS). Given that
available information indicates that a
multi-metal GEMS could not effectively
detect LVM emissions below 80 ug/
dscm, we are proposing an alternative
standard of 80 |ig/dscm for sources that
elect to document compliance with a
GEMS.
7. HC1 and C12 Standards for New HWIs
a. MACT New Floor. The single best
performing source in our database for
HC1 and Cla emissions was a source
equipped with a wet scrubber with a
MTEC of 1.7E7 ug/dscm. The Agency
considered any wet scrubber to be
equivalent technology. Thus, MACT
new floor control is defined as wet
scrubbing with a MTEC up to 1.7E7 pg/
dscm. When we considered emissions
data from all sources equipped with this
level of control, we identified a floor
level of 280 ppmv.
b. Beyond-the-Floor Considerations.
The Agency believes that state-of-the-art
wet scrubbers can readily achieve better
than 99 percent removal of HC1 and C12.
Applying this removal efficiency to the
test condition in our database with the
highest average emission (i.e., 1,100
ppmv; no emission control device)
results in an emission of 11 ppmv. We
do not believe, however, that it is
necessary to establish a BTF (design)
level68 this low for HC1 and C12.
Accordingly, we believe that it is
reasonable and appropriate to establish
a design level of 25 ppmv which
corresponds to a statistically-derived
standard of 67 ppmv.69
We note that this level is consistent
with the levels we are proposing for
new CKs (67 ppmv BTF level) and new
LWAKs (62 ppmv floor level). Further,
we note that HC1 and Cla emissions at
this level are not likely to result in
additional regulation of these sources to
satisfy RCRA health risk concerns.
8. Carbon Monoxide and Hydrocarbon
Standards for New HWIs
As with existing sources, CO and HC
in conjunction with PM remain the
parameters of choice to monitor
continuously for controlling nbn-dioxin
organics. Current regulations require
continuous monitoring of CO, but not of
HC, and so the database of CO from
incinerators is quite extensive.
However, the format of our CO data is
mostly on a run average basis as
explained above. The CO levels of the
best performing facility in this database
67That is, the log mean of runs for the test
condition in the expanded MACT pool with the
highest average emission. A within-test condition
emissions variability factor (based on test
conditions in the expanded MACT pool) is added
to the log-mean for this test condition to derive the
standard.
68 An emissions variability factor would be added
to the log-mean of the runs of this test condition
to derive a standard.
69The variability factor is based on within-test
condition emissions variability for incinerators
equipped with wet scrubbers.
are less than 10 ppmv hourly rolling
average (HRA). The technology to
achieve low level of non-dioxin organics
is "Good Combustion Practices", which
is the same as for existing sources.
As such, we are proposing the same
MACT standards for CO and HC as for
existing sources, but request comments
on whether more stringent standards
would be more appropriate for new
sources. The promulgated standard for
new large MWCs ranges from 50 to 150
ppmv based on type of the device and
the Agency would like to consider more
stringent levels for CO and HC that are
representative of good combustion
practices in new HWIs in the final rule.
9. MACT New Cost Impacts
The annualized incremental costs
(capital, operation and maintenance) for
a small, medium and large HWI based
on today's proposed control levels are
estimated at S336K, $514K and $772K,
respectively. Major increases are due to
installing FF, activated carbon injection
(for D/F and Hg control) and scrubbing
devices (for acid gas control). For this
analysis, it was assumed that baseline
facilities can comply with existing
regulations using a wet scrubber and
venturi-scrubber. Since the number of
new facilities starting construction
every year is uncertain, total annualized
incremental cost for all the new HWIs
in the U.S. due to today's proposal
cannot be estimated. The above costs
include increased costs of APCS'
needed above baseline levels, and do
not include costs of the main incinerator
system or the ancillary systems like
fans, stack etc. Details of these costs
have been provided in the "Regulatory
Impact Assessment for the Proposed
Hazardous Waste Combustion MACT
Standards".
C. Evaluation of Protectiveness
In order to satisfy the Agency's
mandate under the Resource
Conservation and Recovery Act to
establish standards for facilities that
manage hazardous wastes and issue
permits that are protective of human
health and the environment, the Agency
conducted an analysis to determine if
the proposed MACT standards satisfy
RCRA requirements, or whether
independent RCRA standards would be
needed. These analyses were designed
to assess both the potential risks to
individuals living near hazardous waste
combustion facilities who are highly
exposed and risks to other less exposed
individuals living near such facilities.
The Agency evaluated potential risks
both from direct inhalation exposures
and from indirect exposures through
deposition onto soils and vegetation and
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17389
subsequent uptake through the food
chain. The Agency evaluated a variety
of exposure scenarios representing
various populations of interest,
including subsistence farmers,
subsistence fishers, recreational anglers,
and home gardeners.70 In characterizing
the risks within these populations of
interest, both high-end and central
tendency exposures were considered.
The primary exposure parameter
considered in the high-end
characterization was exposure duration.
For the baseline, 90th percentile stack
gas concentrations were also included
in the high-end characterization to
reflect the variability in current
emissions. For dioxins at the floor, the
high-end characterization also included
90th percentile stack gas concentrations
to reflect the large variation in dioxin
emissions using the floor technology
(i.e., temperature control). For the
MACT standards, the Agency used the
design value which is the value the
Agency expects a source would have to
design in order to be assured of meeting
the standard on a daily basis and hence
is always a lower value than the actual
standard for all HAPs controlled by a
variable control technology.71 The
procedures used in the Agency's risk
analyses are discussed in detail in the
background document for today's
proposal.72
The risk results for hazardous waste
incinerators are summarized in Table
III.C.l for cancer effects and Table
III.C.2 for non-cancer effects for the
populations of greatest interest, namely
subsistence farmers, subsistence fishers,
recreational anglers, and home
gardeners. The results are expressed as
a range where the range represents the
variation in exposures across the
example facilities (and example water
bodies for surface water pathways) for
the high-end and central tendency
exposure characterizations across the
exposure scenarios of concern. For
example, because dioxins
bioaccumulate in both meat and fish,
the subsistence farmer and subsistence
fisher scenarios are used to determine
the range.73
TABLE III.C.1.—INDIVIDUAL CANCER RISK ESTIMATES FOR INCINERATORS"
Dioxins
Semi-volatile metals z
Low volatile metals 3
Existing Sources
BTF
2E-9 to 9E-5
3E-9to5E-54 ....
3E-9to2E-6s.
4E-9 to 7E-7
5E-8 to 5E-7
2E-10to4E-6
5E-8 to 8E-6
New Sources
Floor
BTF
GEM Option6
3E-9to5E-54
3E-9to2E-65.
5E-8 to 5E-7
2E-8 to 2E-7
5E-8 to 8.E-6
4E-8 to 6E-6
1 Lifetime excess cancer risk.
2 Carcinogenic metal: cadmium.
3 Carcinogenic metal: arsenic, beryllium, and chromium (VI).
4 Based on 20 ng/dscm TEQ, the highest level known to be emitted at the floor.
s Based on 0.20 ng/dscm TEQ. ...
8 Based on SVM standard of 60 ng/dscm and LVM standard of 80 ng/dscm (applicable only if the source elects to document compliance using
a multi-metals GEM).
TABLE III.C.2.—INDIVIDUAL NON-CANCER RISK ESTIMATES FOR INCINERATORS 1
Semi-volatile metals2
Low volatile metals3
Hydrogen chloride
Chlorine
Existing Sources
Floor
BTF
OEM Option 6
<0.001 to 0.02
<0.001 to 0.01
Nev
<0 001 to 0.01
<0 001 to 0.003
<0.001 to 0.004
<0.001 to 0.2
<0.001 to 0.09
v Sources
<0.001 to 0.09
<0.001 to 0.03
<0.001 to 0.06.
0.001 to 0.05
0.02 to 0.054
0.02 to 0.05 4
0.004 to 0.01 4
0.008 to 0.7
0.07 to 0.3s
0.07 to 0.3s
0.02 to 0.07s
1 Hazard quotient.
2 Cadmium and lead.
3 Antimony, arsenic, beryllium, and chromium.
4 HCI+Cl z assuming 100 percent HCI.
5 HCI+C12 assuming 10 percent Cl 2. , ...
8 Based on SVM standard of 60 (ig/dscm and LVM standard of 80 ng/dscm (applicable only if the source elects to document compliance using
a, multi-metals GEM).
™ In addition, the Agency evaluated a "most
exposed individual" for the purpose of assessing
inhalation risks. A most exposed individual (MEI)
is operationally defined as an individual who
resides at the location of maximum predicted
ambient air concentration.
71 For the semi-volatile and low volatility metals
categories, the Agency assumed the source could
emit up to the design value for each metal in the
category for the purpose of assessing protectiveness.
72 "Risk Assessment Support to the Development
of Technical Standards for Emissions from
Combustion Units Burning Hazardous Wastes:
Background Information Document," February 20,
1996.
73 For the semi-volatile and low volatility metals
categories, the inhalation MEI scenarios are also
used. For hydrogen chloride and chlorine (Clz) only
the inhalation MEI scenarios are used.
-------
17390
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
The risk analysis indicates that for the
semi-volatile and low volatility metals
category, the MACT standards for
incinerators are protective at the floor
for both existing and new sources. The
analysis indicates that the CEM
compliance option for new sources is
also protective. For hydrogen chloride
and chlorine (CLz), the MACT standards
for incinerators are also protective at the
floor for both existing and new sources.
However, the analysis indicates that for
dioxins the proposed beyond the floor
standards, rather than the floor levels,
are protective.
IV. Cement Kilns: Basis and Level for
the Proposed NESHAP Standards for
New and Existing Sources
Today's proposal would establish new
emission standards for dioxins/furans,
mercury, semivolatile metals (cadmium
and lead), low volatile metals (arsenic,
beryllium, chromium and antimony),
particulate matter, acid gas emissions
(hydrochloric acid and chlorine),
particulate matter (PM), hydrocarbons,
and carbon monoxide (for the by-pass
duct) from existing and new hazardous
waste-burning cement kilns. See
proposed § 63.1204. The following
discussion addresses how MACT floor
and beyond-the-floor (BTF) levels were
established for each HAP, and EPA's
rationale for the proposed standards.
The Agency's overall methodology for
MACT determinations has been
discussed in Part Three, Sections V and
VI for existing sources and in Section
VII for new sources.
To conduct the MACT floor analyses
presented today, the Agency compiled
all available emissions data from
hazardous waste-burning cement kilns.
As noted earlier, the vast majority of
this database is comprised of
compliance test emissions data
generated as a.result of Boiler and
Industrial Furnace (BIF) rule
requirements.74 The Agency is also
aware that additional emissions data
will become available. Sources of new
data include test reports generated from
compliance recertification testing
(required every three years under the
BIF rule for interim status facilities; see
§ 266.103(d)), results from voluntary
industry initiatives and testing
programs, supplemental emissions
testing conducted by individual
companies, and data from pilot-scale
research by EPA's Office of Research
and Development. As timely and
appropriate, notice of these additional
data, if used as a basis for standards in
this rulemaking, will be published to
allow for review. However, we
emphasize again that, for purposes of
setting MACT standards, it is preferable
to have data that reflect the normal, day-
to-day operations and emissions. In
addition, the Agency believes that this
type of data will substantially assist in
the appropriate resolution of some of
the issues (e.g., variability, proper
identification of sources in MACT floor
pools, raw material feed contributions to
emissions) that are raised in the
following sections. We invite
commenfers to submit this type of data
and to discuss these issues in their
comments.
In addition, the Agency requests
comments on whether we should use
emissions data from cement kilns that
no longer burn hazardous waste for
MACT floor determinations.75 Even
though these cement kilns subsequently
decided to stop burning waste, we
believe that their emissions data
represent the level of emission control
achieved at a kiln burning hazardous
waste and are therefore appropriate for
use in a MACT analysis. Moreover, the
air pollution control equipment
employed by these facilities is similar in
type, design and operation to equipment
employed by the waste-burning industry
as a whole.
The Agency conducted a preliminary
analysis of the effect on MACT floor
levels of removing these emissions data
from consideration, and found no
significant impacts (see discussion later
in this section on MACT floor levels)
other than for semivolatile metals and
hydrocarbons in the by-pass duct. The
SVM floor would rise from 57 ug/dscm
(today's proposed floor level) to
approximately 1200 ug/dscm.76 This
level is much higher than the cement
industry can achieve.77 Also, the
74By August 21,1992, or by the applicable date
allowed by an extension by the Regional
Administrator, owners and operators of BIF
facilities burning hazardous waste were required to
conduct compliance testing and submit'a
certification of compliance with the emissions
standards for individual toxic metals, HG1, Cl 2,
particulate matter, and CO, and where applicable,
HC and dioxin/furans. See 40 GFR § 266.103(c).
75 Cement kilns no longer burning hazardous
waste include three Southdown plants (Fairborn,
OH, Knoxville, TN, and Kosmosdale, KY) and North
Texas Cement (Midlothian, TX).
76 The Agency notes that we are also taking
comment on a SVM floor level of 160 ug/dscm
(using an alternative approach discussed later in
this section). A SVM floor level of 1200 ug/dscm
appears unnecessarily high considering our
proposed floor analysis and that of others (e.g., see
Part Four, section 9).
77 See letter from Craig Campbell, CKRC, to James
Berlow, EPA, undated but received February 20,
1996. We note that, although the Agency is
proposing a SVM standard of 57 ug/dscm, we invite
comment on an alternative (and potentially
preferable) approach to identify MACT floor
technology which would result in a floor-based
standard of 160 ug/dscm. See discussion on SVM
Agency notes that a SVM floor of 1200
ug/dscm may necessitate the need to
consider adopting further controls
under RCRA to address potential risks
that SVMs (especially cadmium) may
pose.78
In addition, the by-pass duct HC floor
would be affected because two-thirds of
the HC data available to the Agency
were generated by these cement plants
and would no longer be considered in
the analysis. This may make calculation
of the HC MACT floor problematic using
the current MACT approach due to the
limited remaining emissions data. The
remainder of the HAP floors would
remain roughly at today's proposed
levels.
If EPA were to decide to exclude data
from cement kilns that no longer burn
hazardous waste, the Agency then
believes that emission data from cement
kilns that have made significant
modifications or retrofits to their
manufacturing process (e.g., replacing a
raw material with one with different
characteristics, installing new control
equipment) since the earlier emissions
data were generated must also be
considered for exclusion from MACT
analysis. The Agency requests comment
on whether we should use these
emissions data (i.e., the data generated
prior to significant process changes) in
MACT analysis. The commenter should
also address how the Agency could
identify cement kilns that have made
significant process changes and the
scope of modifications or retrofits that
would significantly impact emissions.
Finally, since changes can affect some
HAP emissions and not others, the
commenter should address whether this
issue should be decided on an
individual HAP basis.
A. Summary of Standards for Existing
Cement Kilns
This section summarizes EPA's
rationale for identifying MACT for
existing cement kilns that burn
hazardous waste and the proposed
emission limits. The discussion of
MACT includes discussions of "floor"
controls and considerations of "beyond-
the-floor" controls. Table IV.4.A.1
summarizes the proposed emission
limits.
floor later in this section. Because we identified the
alternative approach late in the rule development
process, we are inviting comment on the higher
standard rather than proposing it.
78 The Agency doubts that a MACT beyond-the-
floor level would be warranted.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17391
TABLE IV.4.A.1 .—PROPOSED EMIS-
SION STANDARDS FOR EXISTING CE-
MENT KILNS
HAP or HAP surro-
gate
Dioxin/furans (TEQ)
Particulate Matter ....
Mercury
SVM (Cd, Pb)
LVM (As, Be, Cr, Sb)
HCI+CI2(totaI
chlorides).
Hydro-carbons:
Main Stack*1
By-pass Stackc
Carbon Monoxide:
Main Stack
By-pass Stack <=
Proposed standard3
0.20 ng/dscm (TEQ).
69 mg/dscm (0.030
gr/dscf).
50 [ig/dscm.
57 ng/dscm.
130 ng/dscm.
630 ppmv.
20 ppmv.
6.7 ppmv.
N/A.
100 ppmv.
•All emission levels are corrected to 7 per-
cent O2.
b Applicable only to long wet and dry proc-
ess cement kilns (i.e., not applicable to pre-
heater and/or precalciner kilns).
c Emissions standard applicable only for ce-
ment kilns configured with a by-pass duct
(typically preheater and/or precalciner, kilns).
Source must comply with either the HC or CO
standard in the by-pass duct. A long wet or
long dry process cement kiln that has a by-
pass duct has the option of meeting either the
HC level in the main stack or the HC or CO
limit in the by-pass duct.
1. Dioxin/Furans
a. MACT Floor. The Agency's analysis
of dioxin/furan (D/F) emissions from
HWCs and other combustion devices
(e.g., municipal waste combustors and
medical waste incinerators) indicates
that temperature of flue gas at the inlet
of the PM control device can have a
major effect on D/F emissions.79 D/F
emissions generally decrease as the gas
temperature of the PM control device
decreases, and emissions are lowest
when the gas temperature of the PM
control device are below the optimum
temperature window for D/F
formation—450 °F to 650 °F.80 Given
that CKs operate their ESPs and FFs
under a range of temperatures (i.e., from
350 °F to nearly 750 °F), the Agency is
identifying MACT floor for D/F based
on temperature control at the inlet to the
ESPorFF.8'
'"USEPA, "Draft Technical Support Document
For HWC MACT Standards, Volume HI: Selection
of Proposed MACT Standards and Technologies",
February 1996.
80For example, consider kiln #1 at the Ash Grove
Cement Company in Chanute, Kansas. During BIF
certification of compliance testing in 1992, Ash
Grovo dioxins/furans emissions exceeded 1.7 ng/
dscm (TEQ) at a control device temperature of 435
*F. Testing in 1994 at a temperature of
approximately 375 °F resulted in emissions less
than 0.05 ng/dscm (TEQ).
»' Tho Agency notes, however, that other factors
can affect D/F emissions including presence of
precursors in the feed or as a result of incomplete
combustion and presence of compounds thought to
The emissions data for CKs includes
results from 58 test conditions collected
from 19 cement plants, with a total of
28 kilns being tested. The Agency's
database shows that the average test
condition D/F emissions ranged from
0.004 to nearly 50 ng/dscm (TEQ).
Kilns emitting D/F at or below levels
emitted by the median of the best
performing 12 percent of kilns had flue
gas temperatures at or below 418°F at
the inlet to the ESP or FF, while inlet
temperatures for other kilns ranged to
nearly 750°F. The Agency then
evaluated D/F emissions from all kilns
that operated the ESP or FF at 418°F or
less and determined that 75 percent had
D/F emissions less than 0.2 ng/dscm
(TEQ). The other 25 percent of kilns
generally had TEQs less than 0.8 ng/
dscm (TEQ), although one kiln emitted
4.7 ng/dscm (TEQ).
The Agency is, therefore, identifying
temperature control at the inlet to the
ESP or FF at 418 °F as the MACT floor
control. Given that 75 percent of sources
achieve D/F emissions of 0.20 ng/dscm
(TEQ) at that temperature, the Agency
believes that it is appropriate to express
the floor as "0.20 ng/dscm (TEQ), or
(temperature at the inlet to the ESP or
FF not to exceed) 418 °F". This would
allow sources that operate at
temperatures above 418 °F but that
achieve the same D/F emissions as the
majority of sources that operate below
418 °F (i.e., 0.20 ng/dscm (TEQ)) to meet
the standard without incurring the
expense of lowering the temperature at
the ESP or FF.
EPA estimates that over 50 percent of
CKs currently are meeting the floor
level. The national annualized
compliance cost82 for CKs to reduce D/
F emissions to 0.20 ng/dscm (TEQ) or
control ESP or FF inlet temperature to
below 418 °F would be $7.3 million for
the entire hazardous waste-burning
cement industry, and would reduce D/
F TEQ emissions nationally by 830
grams/year (TEQ) or 96 percent from
current baseline emissions.
b. Beyond-the-Floor (BTF)
Considerations. The Agency has
inhibit surface-catalyzed formation of D/F such as
sulfur. Thus, D/F emissions may be low (e.g., 0.2
ng TEQ per dcsm) even though the temperature of
stack gas at the inlet to the ESP or FF may exceed
400-450 °F, and D/F emissions may be relatively
high (e.g., 0.3-0.5 ng TEQ per dscm) even though
the temperature may be below that range.
82 Total annual compliance costs are before
consolidation and do not incorporate market exit
resulting from the proposed rule. Also, GEM costs
assume that no facilities currently have a HC
analyzer in place. Thus, these compliance costs
may result in overstated annual compliance costs.
See the "Second Addendum to the Regulatory
Impact Assessment for Proposed Hazardous Waste
Combustion MACT Standards", February 1996, for
details.
identified activated carbon injection (CI)
at less than 400 °F as a BTF control for
D/F for cement kilns because CI is
currently used in similar applications
such as hazardous waste incinerators,
municipal waste combustors, and
medical waste incinerators. The Agency
is not aware of any CK flue gas
conditions that would preclude the
applicability of CI or inhibit the
performance of CI that has been
demonstrated for other waste
combustion applications.
Carbon injection has been
demonstrated to be routinely effective at
removing greater than 95 percent of D/
F for MWCs and MWIs and some tests
have demonstrated a removal efficiency
exceeding 99 percent at gas
temperatures of 400 °F or less.83 To
determine a BTF emission level, the
Agency considered the emission levels
that would be expected to result from
gas temperature control to less than 400
°F combined with CI.
To estimate emissions with
temperature control only, the Agency
considered the MACT floor database
that indicates, as noted above, 25
percent of CKs operating the ESP or FF
at temperatures above 418°F could be
expected to emit D/F at levels above 0.2
ng/dscm (TEQ). Although the majority
could be expected to emit levels of 0.8
ng/dscm (TEQ) or below, some could be
expected to emit levels as high as 4.7 ng
TEQ.
When CI is used in conjunction with
temperature control, an additional 95
percent reduction in emissions could be
expected. Accordingly, emissions with
these BTF controls could be expected to
be less than a range of 0.04 to 0.24 ng/
dscm (TEQ) (i.e., 95 percent reduction
from 0.8 ng and 4.7 ng, respectively).
Given that CI reductions greater than 95
percent are readily feasible, the Agency
believes that it is appropriate to identify
0.20 ng/dscm (TEQ) as a reasonable BTF
level that could be routinely achieved.
The Agency notes that, because we
have assumed a fairly conservative
carbon injection removal efficiency of
95 percent to identify the 0.20 ng/dscm
(TEQ) level, we believe that this
approach adequately accounts for
emissions variability at an individual
kiln because CI removal efficiency is
likely to be up to or greater than 99
percent. EPA thus believes that it is not
necessary to add a statistically-derived
variability factor to the 0.20 ng/dscm
(TEQ) level to account for emissions
variability at an individual kiln. Thus,
83USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume III: Selection of
Proposed MACT Standards and Technologies",
February 1996.
-------
17392
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
the 0.20 ng/dscm (TEQ) BTF level
represents the proposed emission
standard.
EPA solicits comment on this
approach, and notes that if a
statistically-derived variability factor
were deemed appropriate with the
assumed conservative CI removal
efficiency, the BTF .level of 0.20 ng/
dscm (TEQ) would be expressed as a
standard of 0.31 ng/dscm (TEQ). We
note, however, that under this approach,
it may be more appropriate to use a less
conservative, higher CI removal
efficiency of 99 percent (i.e., because
emissions variability would be
accounted for using statistics rather than
in the engineering decision to use a
conservative CI removal efficiency).
Doing so would lower the 0.20 ng/dscm
(TEQ) level to approximately 0.04 ng/
dscm (TEQ) (i.e., 99 percent reduction
from 0.8 ng and 4.7 ng results in levels
of 0.008 ng to 0.047 ng/dscm (TEQ), ,
respectively, and 0.04 ng is a reasonable
value within this range). If so, the D/F
standard would be about 0.15 ng/dscm
(TEQ) (i.e., 0.04 ng/dscm TEQ plus the
variability factor of 0.11 ng/dscm TEQ).
We note that although CI is normally
a relatively inexpensive control
technology to add to sources (with flue
gas above the dew point) that already
have PM controls at the 69 mg/dscm ,
level, CKs present a special situation.
This is because: (1) CI will remove Hg
as well as D/F (see discussion below
regarding BTF control for Hg); (2) CKs
recycle as much collected PM as
possible because it is useful raw
material and doing so reduces cement
kiln dust (CKD) management cost; (3)
some CKs recycle the CKD by injecting
it at the raw material feed end of the
kiln where the D/F may not be
destroyed; and (4) to remove Hg from
the recycling system to ensure
compliance with the Hg standard, a
portion of the CKD would have to be
wasted.84
Accordingly, EPA has assumed that
CKs that have to use CI to meet the BTF
standard (i.e., those that cannot achieve
the standard with temperature control
alone) would install the CI system after
the existing ESP or FF and add a FF to
remove the injected carbon with the
adsorbed D/F (and Hg). Although
adding a new FF in series is an
expensive approach, it would enable
CKs to meet both the proposed D/F and
Hg standards (as well as the PM, SVM,
and LVM standards). Thus, the cost of
the CI and FF systems have been
apportioned among these proposed
standards.
EPA estimates that 40 percent of CKs
are currently meeting this BTF level.
The national incremental annualized
compliance cost for the remaining CKs
to meet this BTF level85 rather than
comply with the floor controls would be
$6.6 million for the entire hazardous
waste-burning cement industry, and
would provide an incremental reduction
in D/F (TEQ) emissions nationally
beyond the MACT floor controls of 20
grams/year (TEQ).
EPA has considered costs in relation
to emissions reductions and the special
bioaccumulation potential that D/F pose
and determined that proposing a BTF
limit is warranted.86 D/F are some of the
most. toxic compounds known due to
their bioaccumulation potential and
wide range of health effects at
exceedingly low doses, including
carcinogenesis. Further, as discussed
elsewhere in today's preamble, EPA's
risk analysis developed for purposes of
RCRA shows that emissions of these
compounds from hazardous waste-
burning cement kilns could pose
significant risks by indirect exposure
pathways, and that these risks would be
reduced by BTF controls. Finally, EPA
is authorized to consider this non-air
environmental benefit in determining
whether to adopt a BTF level. As noted
earlier, exposure via these types of
indirect pathways was in fact a chief
reason Congress singled out D/F for
priority MACT control in section
Finally, EPA's initial view is that it
may need to adopt further controls
under RCRA to control D/F if it did not
adopt the BTF MACT standard. This
would defeat one of the purposes of this
proposal, to avoid regulation of
emissions under both statutes for these
sources wherever possible. These risks
would, however, be reduced to
acceptable levels if emissions levels are
reduced to 0.20 ng/dscm (TEQ).
For these reasons, the Agency is
proposing a.BTF level of 0.20 ng/dscm
84 We note that most CKs currently dispose of a
portion of CKD to control clinker quality (i.e., to
control alkali salts). Nonetheless, the economics of
CKD management are uncertain at this time given
impending Agency action to ensure proper
management. Thus, we believe that CKs will
increase efforts in the future to minimize the
amount of CKD that is disposed.
85 We note that not every source with D/F
emissions currently exceeding 0.20 ng TEQ per
dscm would need to install CI to meet the standard.
As noted previously in the text, 75 percent ,of
sources could be expected to meet the standard
with temperature control only. In estimating the
cost of compliance with the standard, EPA
considered the magnitude of current emissions and
current operating temperatures to project whether
the source could comply with the standard with
temperature control only.
86 We note that the D/F BTF control technology,
CI, would also be used to control mercury
emissions beyond the floor.
(TEQ) for D/F emitted from hazardous
waste-burning cement kilns.
2. Particulate Matter
a. MACT Floor. Cement kilns have
high particulate inlet loadings to the
control device due to the nature of the
cement manufacturing process; that is, a
significant portion of the finely
pulverized raw material fed to the kiln
is entrained in the flue gas entering the
control device. CKs use ESPs or FFs to
control PM to a 0.08 gr/dscf standard
under the BIF rule, unless the kiln is
subject to the more stringent New
Source Performance Standard (NSPS)
(see 40 CFR 60.60 (Subpart F)) of 0.3 lb/
ton of raw material feed (dry basis) to
the kiln,87 which is generally equivalent
to 69 mg/dscm or 0.03 gr/dscf.
The PM emissions data for CKs
includes results from 54 test conditions
collected from 26 facilities, with a total
of 34 units being tested. The Agency
analyzed all available PM emissions
data and determined that sources with
emission levels at or below the level
emitted by the median of the best
performing 12 percent of sources used
fabric filters with air-to-cloth (A/C)
ratios of 2.3 acfm/ft2 or less. Analysis of
emissions data from all CKs using FFs
with the 2.3 acfm/ft2 A/C ratio or less
resulted in a level of 0.065 gr/dscf.
Because the NSPS is a federally
enforceable limit that many cement
kilns are currently subject to, the
Agency has chosen the existing NSPS
standard, not the statistically-derived
limit discussed above, as MACT for
existing hazardous waste-burning CKs.
Thus, the Agency is identifying a MACT
floor for PM and is identifying the floor
level as the NSPS limit of 69 mg/dscm
(0.03 gr/dscf). Given that the NSPS
standard was promulgated in 1971, the
Agency believes that it is reasonable to
consider it as the MACT floor level. We
note further that 30 percent of cement
kiln test conditions currently meet the
69 mg/dscm floor level.
As mentioned above, the NSPS
standard for PM is expressed as 0.3 lb/
ton of raw material (dry basis) feed to
the kiln. Although we are proposing to
establish the floor level as the MACT
standard (see BTF discussion below)
expressed as 69 mg/dscm (0.03 gr/dscf),
we specifically invite comment on
whether the standard should be
expressed in terms of raw material feed.
We are proposing a "mg/dscm" basis for
the standard because a PM
concentration in stack gas is commonly
used for waste combustors-hazardous
waste incinerators, municipal waste
87 See § 60.62 Standard for particulate matter for
further details.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17393
combustors, and medical waste
incinerators. We note, however, that
using a "mg/dscm" basis for the CK
standard would penalize the more
thermally efficient dry kilns (generally
preheater and precalciner kilns). This is
because these kilns have lower stack gas
flow rates per ton of raw material feed
because they do not need to provide
additional heat (by burning hazardous
waste and/or fossil fuel) to evaporate the
water in the raw material slurry. Thus,
wet kilns have higher gas flow rates per
ton of raw material than dry kilns
because of increased combustion gas
and water vapor. This higher stack gas
flow rate dilutes the PM emissions and
effectively makes a concentration-based
standard less stringent for wet kilns.
Consequently, the Agency will consider
whether the final rule should express
the floor standard as 0.3 Ib/ton of raw
material (dry basis) feed to the kiln.
EPA estimates that 30 percent of
cement kiln test conditions (in our
database) are currently meeting the floor
level. The national annualized
compliance cost for the remaining CKs
to reduce PM emissions to the floor
level would be $6.5 million for the
entire hazardous waste-burning cement
industry, and would reduce PM
emissions nationally by 2400 tons per
year.
b. Beyond-the-Floor Considerations.
EPA considered but is not proposing a
more stringent beyond-the-floor level
(e.g., 35 mg/dscm (0.015 gr/dscf)) for
cement kilns. For this analysis, EPA
determined that it does not have
adequate data to ensure that, given the
high inlet grain loading caused by
entrained raw material, CKs can
routinely achieve that emission level
day-in and day-out with a single PM
control device—ESP or FF. We note
that, to ensure compliance with a 35
mg/dscm standard 99 percent of the
time, a source with average emissions
variability must be designed and
operated to achieve an emission level of
approximately 18 mg/dscm (or 0.008 gr/
dscf). EPA estimates that 15 percent of
CKs currently have average PM
emissions below 18 mg/dscm.
Reducing the floor level from 69 mg/
dscm to a BTF level of 35 mg/dscm
would require an improved technology
such as the use of more expensive fabric
filter bags (e.g., bags backed with a
teflon membrane) or the addition of a FF
for kilns with ESPs. The addition or
upgrade of FFs to all kilns could
potentially be cost effective, since to
meet the proposed floor for SVM and
LVM, as well as the proposed BTF for
D/Fs and Hg, addition of a new FF is
projected for a majority of the kilns
(about 80 percent). Thus, a PM BTF
level of 18 mg/dscm may be the
incremental cost between a fabric filter
with conventional fiberglass bags and
state-of-the-art membrane-type bags for
those kilns currently employing FFs; the
addition of new FFs with membrane
bags for those kilns with ESPs; or new
FFs with membrane bags for the
remaining facilities which are not
projected to need upgrades to meet the
floor and proposed BTF levels.
At first glance it may seem cost
effective, primarily since an improved
BTF PM level would lead to added
benefits with reduced SVM, LVM, and
condensed organics emissions.
However, the Agency is uncertain how
facilities will meet the proposed SVM,
LVM, D/FS, and Hg levels. For example,
kilns could meet the mercury BTF level
with feedrate control or carbon injection
without addition of a new FF
(potentially incurring the penalty of
reduced or eliminated kiln dust
recycle). Additionally, CKs could meet
the D/F BTF level with PM control
device temperature reduction instead of
carbon injection with an add-on FF.
Finally, kilns could meet the SVM and
LVM floor levels with feedrate control.
Therefore, many of the kilns may not
add new FFs to comply with proposed
floor (e.g., SVM, LVM) or proposed BTF
levels (e.g., D/FS, Hg) and EPA's
estimated engineering cost to meet the
floor has been conservatively overstated.
Thus, it may not be accurate to conclude
that the BTF for PM is close to the
incremental cost between FF fabric
types. Under this circumstance, the
incremental cost is more accurately the
cost of many new FF unit additions
which the Agency believes would not be
cost effective. For these reasons the
Agency believes it is not appropriate to
propose a BTF PM standard of 35 mg/
dscm for existing CKs. EPA specifically
invites comment on whether the final
rule should establish a BTF standard for
PM of 35 mg/dscm (or 0.15 Ib/ton of raw
material (dry basis) feed into the kiln).
3. Mercury
a. MACT Floor. Mercury emissions
from CKs are currently controlled by the
BIF rule, and CKs have elected to
comply with the BIF standard by
limiting the feedrate of Hg in the
hazardous waste feed.88 Thus, the
MACT floor level is based on hazardous
waste feed control.
Mercury emissions from cement kilns
range from 3 ug/dscm to an estimated
600 [ig/dscm. The Agency has Hg
emissions data from 42 test conditions
collected from 21 cement plants, with a
total of 28 kilns being tested. Since
mercury is a volatile compound at the
typical operating temperatures of ESPs
and baghouses, collection of mercury by
these control devices is highly variable
(e.g., Hg removal efficiencies ranged
from zero to more than 90 percent).
Most of the mercury exits the kiln
system as volatile stack emissions, with
only a small fraction partitioning to the
clinker product or CKD.
To identify the floor level for
hazardous waste feed control, the
Agency determined that sources with
Hg emissions at or below the level
emitted by the median of the best
performing 12 percent of sources had
normalized hazardous waste Hg
feedrates, or MTECs, (i.e., maximum
theoretical emission rates 89) of 110 |ig/
dscm or less. Analysis of all existing
cement kiln sources using this
hazardous waste feedrate control
resulted in a MACT floor level of 130
Hg/dscm. To meet this standard 99
percent of the time, EPA estimates that
a source with average emissions
variability 90 must be designed and
operated to routinely achieve an
emission level of 81 ug/dscm.
We note that raw materials and fossil
fuels also contribute to cement kiln Hg
feedrates and emissions. Given that all
sources must be able to meet the floor
level using the floor control, we
investigated whether all CKs could meet
the floor level by only controlling
hazardous waste Hg feedrate to the
MACT MTEC of 110 ug/dscm. We have
determined that all CKs in the Hg
emissions database, except for one kiln
with apparently anomalous data on
mercury in raw material, would be able
to meet the floor level using floor
control.91 The one kiln reported
substantially higher Hg feedrates in the
raw material than other kilns. We
believe that this data may either be
erroneous or the kiln may have spiked
Hg into the raw material during BIF
compliance testing. We specifically
invite data and comment on the issue of
normal Hg content in raw material.
EPA estimates that nearly 80 percent
of CKs could currently comply with the
floor level. The total annualized
compliance cost for the remaining kilns
88 BIF Hg emission limits are implemented by
establishing limits, in part, on the maximum feed
rate of Hg in total feedstreams. Feedstream sources
of mercury include hazardous waste, Hg spiking
during compliance testing, raw material, coal and
other fuels.
89 MTEC is the hazardous waste Hg feedrate
divided by the gas flow rate. .
""This represents the variability of emissions
among runs within a test condition included within
the expanded MACT pool.
91USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume III: Selection of
Proposed MACT Standards and Technologies",
February 1996.
-------
17394
Federal Register / Vol. 61, No. 77 /Friday, April 19, 1996 / Proposed Rules
to reduce Hg emissions to the floor level
is estimated to be up to $7.5 million for
the entire cement industry, and would
reduce Hg emissions nationally by 7,200
Ibs per year, or by 58 percent from
baseline emissions.
b. Beyond-the-Floor Considerations.
The Agency has considered two BTF
control options for improved Hg control:
flue gas temperature reduction to 400°F
or less followed by either carbon
injection (CI) or carbon bed (CB). Either
control option would be implemented
in conjunction with hazardous waste
feedrate control of Hg. Due to the
uncertainty surrounding the actions that
cement kilns will undertake in
achieving increased Hg control (i.e.,
with respect to reducing the Hg content
of the hazardous waste received at the
kiln versus installing the carbon
injection technology to capture
volatilized mercury without reducing
Hg content in the hazardous waste feed),
the Agency assumed a conservative
emissions level attributable to feedrate
control to which the Agency applied the
BTF control technology (i.e., 300 ug/
dscm). EPA believes that CI systems can
routinely achieve Hg emission
reductions of 80 to 90 percent or
better92 and that CB systems can
routinely achieve Hg emissions of 90 to
99 percent or better.93
The BTF level under the Cl-controlled
option would, therefore, be 50 ug/dscm
(corresponding to a design level of 30
ug/dscm), based on 90 percent reduction
after the source has controlled its Hg
emissions to 300 ug/dscm by limiting
Hg in the hazardous waste. As discussed
later, EPA is proposing a 50 ug/dscm
based on this BTF option.94
The BTF level under the CB-
controlled option would be 8 ug/dscm
(corresponding to a design level of 5 ug/
dscm), based on 99 percent reduction
after the source has controlled its Hg
emissions to 300 ug/dscm by limiting
Hg in the hazardous waste.
We note that another control option
for identifying BTF levels would be to
consider the floor hazardous waste
feedrate control—MTEC of 110, ug/dscm
or less—an initial component of BTF
control followed by either CI or CB.
Under this approach, BTF emission
levels would be identified by first
assuming sources would impose only
feedrate controls to meet the floor level
of 13Q ug/dscm (corresponding to a
design le,vel of 81 ug/dscm). Thus, a CI
injection system at 90 percent removal
could be expected to achieve a standard
of 13 ug/dscm (corresponding to a
design level of 8.1 ug/dscm); A CB
system at 99 percent removal could be
expected to achieve a design level of 0.8
ug/dscm to which an emissions
variability factor would be added to
identify the standard. EPA solicits
comment on whether this option of
applying BTF reduction based on CI or
CB to the floor levels should be adopted.
We also note that an alternative
approach to using a statistically-derived
variability factor to account for
emissions variability would be to
assume a more conservative control
efficiency for the CI or CB BTF
technology. We believe that using a
more conservative removal efficiency
could be a means to adequately account
for emissions variability given that
actual emissions using the BTF control
would be expected to be lower than the
assumed emission level. Under this
approach, we would more
conservatively assume that CI-
controlled'systems could achieve a
removal efficiency of 80 percent and
that CB:controlled systems could
achieve an efficiency of 90 percent.
When these removal efficiencies are
applied, this would result in emission
standards of 16 ug/dscm for Cl-
controlled systems, and 8 ug/dscm for
CB-controlled systems 95. We invite
comment on these alternative
approaches to account for emissions
variability at an individual plant.
EPA believes that CI is a cost-effective
BTF control, and is proposing a 50 ug/
dscm Hg emission standard based on •
that control in conjunction with a
preceding estimated hazardous waste
feedrate control resulting in an
emissions level of 300 ug/dscm prior to
the CI control. We estimate that 57
percent of CKs are currently meeting
this level. The incremental national
annualized compliance cost for the
remaining CKs to meet this level rather
than comply with the floor controls
would be $7.8 million, and would
92 Memorandum from Frank Behan, USEPA, to
RCRA Docket. Discussion of mercury removal
efficiency with activated carbon injection during an
emissions test at a Lafarge Corporation cement kiln.
February 26,1996.
93 USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume III: Selection of
Proposed MACT Standards and Technologies",
February 1996.
94 To achieve a standard of 50 ug/dscm 99 percent
of the time, a source with average emissions
variability must be designed and operated to
achieve an emission level of 30 ug/dscm.
95The same approach could also be utilized with
the previously discussed approach of applying the
BTF control to an assumed emission level of 300
ug/dscm. When assuming the conservative removal
efficiencies of 80 percent for CI and 90 percent for
CB, this would result in BTF standards of 60 ug/
dscm for Cl-controlled systems and 30 ug/dscm for
CB-controlled systems. Again a statistically-derived
variability factor would not be added because
emissions variability is accounted for by assuming
conservative removal efficiencies for CI and CB
systems.
provide an incremental reduction in Hg
emissions of 2100 Ibs per year .
nationally beyond the MACT floor
controls.
We specifically are interested in.
comment on whether CB is a cost
effective BTF control96. The CB-based
BTF emission level would be 8 ug/dscm
(assuming 90 percent removal ••,
efficiency). We estimate that 22 percent
of CKs are currently meeting this level.
The incremental national annualized
compliance cost for the remaining CKs
to meet this level rather than comply
with the floor controls (and proposed
Cl-based level of 50 ug/dscm) is
estimated to be $34.8 million and would
provide an incremental reduction in Hg
emissions nationally of 5,100 Ibs per
year from the floor.
The Agency also invites comment on
whether special consideration should be
given to kilns that may burn hazardous
waste with non-detect levels of Hg.97
Such kilns could be considered to be
appropriately regulated, with respect to
Hg emissions, by only the standards the
Agency is developing for cement kilns
that do not burn hazardous waste. Thus,
today's proposed Hg standards for
waste-burning kilns would be waived.
To minimize implementation confusion
and difficulties and to accommodate
enforcement concerns, if a CK at any
time burns hazardous waste with
detectable levels of Hg, the kiln would
be subject to today's proposed rules at
all times, even if it subsequently burned
waste with non-detect levels of Hg.
Under the waiver, the owner and
operator would be required to sample
and analyze the hazardous waste as
necessary to document that it continues
to contain non-detect levels of Hg. We
invite comment on whether such a
deferral to another MACT standard (yet
to be proposed for non-hazardous waste-
burning CKs) is workable, given the
potential for piece-meal permitting and
enforcement.
EPA has considered costs in relation
to emissions reductions and the special
bioaccumulation potential that Hg poses
and determined that proposing a BTF
limit is warranted. Hg is one of the more
toxic metals known due to its ;
bioaccumulation potential and the
adverse neurological health effects at
low concentrations especially to the
most sensitive populations at risk (i.e.,
96 We also note that, while the Agency does not
have information to conclude that application of the
carbon bed technology would be problematic for
cement kilns, carbon beds have never been tested
at a full-scale cement kiln. Thus, we invite
comment on the technical feasibility of CB control
of Hg emissions from CKs.
97 We also invite comment on what minimum
detection levels would be acceptable.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17395
unborn children, infants and young
children). A more detailed discussion of
human health benefits for mercury can
be found in Part Seven of today's
proposal. The indirect exposure
pathway resulting from airborne
deposition of Hg is of particular
concern, and a particular reason that
Congress singled out Hg for priority
regulation in section 112(c)(6). See S.
Rep. No. 228,101st Cong. 1st Sess. at
153-55,166. EPA is specifically
authorized to take into account such
non-air environmental benefits in
assessing when to adopt BTF standards.
As noted below, hazardous waste-
burning cement kilns are a significant
source of Hg emissions, and the BTF
option \vill control those emissions from
75 percent over baseline and 47 percent
over the floor. EPA believes the cost of
controlling this especially dangerous
HAP to be warranted in light of the
extent of control, magnitude of
emissions, limited effect on cost of
treating hazardous waste (and no net
effect on the cost of cement), and the
fact that the control technology, carbon
injection, will also control dioxins and
furans. Finally, EPA notes that control
of Hg at the BTF level should eliminate
the uncertainty presently involved in
individual RCRA permitting decisions
where permit writers may develop site-
specific permit limits beyond those
required by current regulations if
necessary to protect human health and
the environment.
4. Semivolatile Metals
a. MACT Floor. Emissions of SVM
from CKs are currently controlled under
the BIF rule. Kilns use a combination of
hazardous waste feedrate control and
PM control to comply with those
standards. Accordingly, MACT floor
control is based on a combination of
hazardous waste feedrate control and
PM control.
The SVM emissions data for CKs
includes results from 45 test conditions
collected from 26 cement plants, with a
total of 34 kilns being tested. Baseline
emissions of the semivolatile metals
group (consisting of cadmium and lead)
ranged from 3 ug/dscm to slightly over
6,000 ug/dscm. Cadmium and lead are
volatile at the usual high temperatures
within the cement kilns itself, but
typically condense onto the fine
particulate at baghouse and ESP
temperatures, where they are collected.
As a result, control of semivolatile
emissions is associated with PM control.
However, because of the potential for
adsorption for these two metals onto the
fine PM that is less effectively collected
than larger-sized PM, the control
efficiency for semivolatile metals is
likely to be lower than that for total PM.
As discussed earlier, all cement plants
currently use either baghouses or ESPs
to control particulate emissions.
The Agency analyzed all available Cd
and Pb emissions data and determined
that sources with emission levels at or
below the level emitted by the median
of the best performing 12 percent of
sources used fabric filters with air-to-
cloth (A/C) ratios of 2.1 acfm/ft2 or less
for a kiln system with a hazardous waste
MTEC of 84,000 ug/dscm or less.
Analysis of emissions data from all CKs
using FFs with the 2.1 acfm/ft2 A/C ratio
and with a HW MTEC of 84,000 ^g/
dscm or less resulted in a floor level of
57 ug/dscm.
EPA notes that raw materials and
fossil fuels also contribute to cement
kiln SVM feedrates and emissions.
Given that all sources must be able to
meet the floor level using the floor
control, EPA investigated whether all
CKs could meet the floor level
employing the MACT technologies
without being forced to substitute raw
materials. Our preliminary evaluation
determined that about 10 percent of
sources had raw material containing Cd
and Pb in greater concentrations than
sources in the expanded MACT pool;
thus, these sources may not be able to
achieve the floor with MACT alone.98
Before we reach any final conclusions
on this point, the Agency believes that
further data are needed on the normal,
day-to-day levels of Pb and Cd in raw
material feed.
In addition, one approach to address
this issue (of sources with higher levels
of SVM metals in their raw materials
than sources in the expanded MACT
pool and that, therefore, cannot meet the
floor level using floor control) is to: (1)
identify the source with the highest
normalized (by MTEC) feedrate of
metals in raw material; (2) assume the
source is also feeding hazardous waste
with the floor control MTEC level of the
metals; and (3) project SVM emissions
from the source based on combined raw
material and hazardous waste MTECs
using a representative system removal
efficiency (SRE) from the expanded
MACT pool considering an appropriate
variability factor (e.g., variability of
emissions among runs within a test
condition in the expanded MACT pool).
The Agency has not yet conducted this
type of analysis, but intends to do so.
Again, we also believe that data
reflecting normal, day-to-day levels of
Cd and Pb in raw material feed is
«8USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume III: Selection of
Proposed MACT Standards and Technologies",
February 1996.
important in pursuing this avenue of
analysis. We invite comment on this
approach.
The Agency also notes that the MACT
pool for SVM consists entirely of CKs
employing FF controls; that is, no
cement plants with ESPs are in the
MACT pool or expanded MACT pool.
EPA believes that well designed,
operated, and maintained ESPs can
achieve good control of SVMs. In fact
several CKS employing ESPs in our
database currently achieve the floor
level of 57 ug/dscm. Because the Agency
is concerned that the SVM floor analysis
may be overly exclusive (because
comparably designed and operated ESPs
were not considered in the MACT floor
analysis) in identifying the floor MACT
level and technology, EPA specifically
requests comment on the merits of the
following alternative floor approach.
This approach identifies comparably
designed and operated ESPs (in our
SVM database) equivalent to the MACT
FF (and at the MACT MTEC) and
includes these sources in the analysis as
an "equivalent technology" of MACT.
The Agency has identified an ESP with
an SCA of 500 ft2/kacfm or better as an
equivalent technology to the MACT FF
with an A/C ratio of 2.1 acfm/ft2. The
Agency conducted this analysis and
determined that the floor level would
increase from 57 to 160 ug/dscm using
this approach. To meet this standard 99
percent of the time, EPA estimates that
a source with average emissions
variability must be designed and
operated to routinely achieve an
emission level of 99 p.g/dscm. EPA
investigated whether all CKs could meet
the floor level employing the MACT
technologies without being forced to
substitute raw materials and determined
that all CKs (in the SVM emissions
database) with the exception of one kiln
would be able to meet the 160 ng/dscm
level using this less restrictive MACT
definition. The Agency specifically
requests comment on this alternative
floor approach and floor level.
EPA recognizes that PM, SVM, and
LVM emissions from cement kilns are
similarly controlled, in part, by a good
PM control (e.g., ESP, FF). The floor
control for SVM (FF with an A/C ratio
of 2.1 acfm/ft2) offers slightly more
control than the floor control for LVM
(FF with an A/C ratio of 2.3 acfm/ft2 or
an ESP with a SCA of 350 ft2/kacfm).
Thus, the controls necessary to achieve
the SVM MACT floor level would
appear to be governing for control of
these HAPs.
EPA estimates that 33 percent of CKs
are currently meeting the floor level of
57 ug/dscm. The national annualized
compliance cost for the cement kilns to
-------
17396
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
reduce SVM emissions to the floor level
would be $13.1 million, and would
reduce national Pb and Cd emissions by
29 tons per year or 94 percent from
current baseline emissions.
b. Beyond-the-Floor Considerations.
The Agency considered whether to
propose a more stringent level than the
floor of 57 ug/dscm, but believes that it
would not be appropriate. Since control
of SVM emissions is associated with PM
control, a more stringent BTF level
would require CKs to upgrade to more
expensive fabric filter bags (e:g., bags
backed with a teflon membrane) or the
addition of a FF for kilns with ESPs.
Even though the engineering costs to
comply with a BTF SVM level would be
modest for CKs, the resulting
incremental reduction in SVM
emissions from the floor level would be
minimal. Thus, the Agency believes that
lowering the SVM proposed standard is
not warranted based on the minimal
impact on overall SVM emissions; the
floor already provides substantial
control by reducing baseline SVM
emissions by 94 percent. Thus, the
Agency is proposing a MACT floor SVM
standard of 57 ug/dscm for existing
cement kilns.
5. Low-Volatile Metals
a. MACT Floor. Emissions of LVM
from CKs are also currently controlled
under the BIF rule. Kilns use a
combination of hazardous waste
feedrate control and PM control to
comply with those standards.
Accordingly, MACT floor control is
based on a combination of hazardous
waste feedrate control and PM control.
The Agency has LVM emissions data
which consists of 45 test conditions
collected from 26 cement plants, with a
total of 35 kilns being tested. Average
emissions of the low volatility metals
group (arsenic, antimony, beryllium,
and chromium) ranged from 4 ug/dscm
to 520 ug/dscm. Due to the relatively
low volatility of these metals, more than
70 percent of these metals typically
partition to the clinker product while
the remainder typically condense onto
particulate and are collected in the
APCD (in this case either an ESP or
baghouse). Thus, performance of the
control devices is an important factor in
controlling LVM emissions.
To identify MACT floor, EPA
characterized the LVM controls used by
kilns emitting LVM at levels at or below
the level emitted by the median of the
best performing 12 percent of sources.
MACT floor control is thus defined as:
(1) a baghouse (i.e., fabric filter) with an
air-to-cloth ratio of 2.3 acfm/ft2 or less
with a hazardous waste (HW) MTEC less
than 140,000 ug/dscm; or (2) an ESP
with specific collection area of 350 ft2/
kacfm with a HW MTEC less than
140,000 ug/dscm. Analysis of available
emissions data for all CKs employing
either of these controls resulted in a
floor emissions level of 130 ug/dscm.
EPA notes that raw materials and
fossil fuels also contribute to cement
kiln LVM feedrates and emissions.
Given that all sources must be able to
meet the floor level using the floor
control, EPA investigated whether all
CKs could meet the floor level
employing the MACT controls without
being forced to substitute raw material
feed. EPA determined that all CKs
would be able to meet the floor level
using floor control without switching
raw materials."
EPA estimates that 80 percent of CKs
are currently meeting the floor level.
The national annualized compliance
cost for the cement kilns to reduce LVM
emissions to the floor level would be
$2.8 million for the entire hazardous
waste-burning cement industry, and
would reduce LVM national emissions
by 1.7 tons per year or 49 percent from
current baseline emissions.
b. Beyond-the-Floor Considerations.
The Agency considered whether to
propose a more stringent level than the
floor of 130 ug/dscm. We determined
that proposing such a BTF level is not
warranted for several reasons: (1) It
would not likely be cost effective; (2)
LVM are not of particular concern
because they are not bioaccumulative;
and (3) establishing the MACT standard
at the floor would not trigger the need
for a more stringent RCRA standard.
Since control of LVM emissions is
associated with PM control, a more
stringent BTF level would require CKs
to either install new control equipment
or to upgrade existing control
equipment (e.g., install more expensive
FF bags). Even though the engineering
costs to comply with a lower LVM BTF
level would be moderate, the resulting
reduction in LVM emissions is minimal
since CK LVM national emissions are
estimated to be 1.7 tons/year for the
entire industry at the floor. Thus, a LVM
BTF standard is not believed to be
warranted based on this limited
reduction in LVM emissions.
6. Hydrochloric Acid and Chlorine
a. MACT Floor. HC1 and C12 (also
referred to as total chlorine) emissions
from CKs are currently regulated by the
BIF rule. CKs use the natural alkalinity
of the limestone raw material and
hazardous waste feedrate control (of
total chlorine and chloride) to comply
with those standards. No hazardous
waste-burning cement kiln currently
employs a dedicated control device
(e.g., wet scrubber, venturi scrubber)
designed specifically to remove HCl/CU
from the flue gas. Accordingly, MACT
floor is based on hazardous waste
feedrate control.100
The Agency has HC1 and C12
emissions data consists of 52 test
conditions collected from 26 cement
plants, with a total of 35 kilns being
tested. Total chlorine emissions from
cement kilns range from less than 0.1
ppmv to 220 ppmv. To identify MACT
floor, EPA identified the highest
hazardous waste feed MTEC (i.e.,
normalized hazardous waste feedrate of
total chlorine) used by kilns emitting
HC1/C12 at levels at or below the level
emitted by the median of the best
performing 12 percent of sources—1.6 g/
dscm. The analysis of all available
emissions data for kilns with a
hazardous waste MTEC for total
chlorine of 1.6 g/dscm or less resulted
in a floor emissions level of 630 ppmv.
Our data indicate that 100 percent of the
test conditions in the Agency's database
are achieving this floor value.
This determination is confounding
given that the highest average emissions
from any test condition in the entire
database, irrespective of hazardous
waste MTEC for total chlorine, was 220
ppmv. This anomalous finding is
apparently attributable to: (1) The data
set having very high average within-test-
condition variability; and (2) adding the
average variability factor to the log mean
rather than the arithmetic mean of the
single test condition with the highest
arithmetic mean within the expanded
MACT pool (those sources using MACT
floor control). If that source had
unusually high emissions variability,
then the log mean could be substantially
higher than the arithmetic mean,
resulting in an unusually high emission
level to which the variability factor was
added.
Because of these concerns, the Agency
invites comme'nt on alternative
approaches that may identify a more
reasonable floor level. One approach
could be to add the average variability
factor for the data set to the arithmetic
mean, rather than the log mean, of the
highest test condition in the expanded
»USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume ffl: Selection of
Proposed MACT Standards and Technologies",
February 1996.
""Although owners and operators normally have
no control over the control provided by raw
material alkalinity, we note that kilns equipped
with FFs appear to provide better control than kilns
equipped with ESPs. This may be due to the longer
time of contact between the gas stream and the
alkaline dust as the gases pass through the dust bed
on the bags.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17397
MACT pool. In addition, if this still
resulted in a calculated floor level
greater than any emission level in the
database, irrespective of hazardous
waste MTEC for total chlorine, the floor
level could be capped at the highest
emission level in the database—220
ppmv.
As for the metals EPA notes that raw
materials and fossil fuels also contribute
to cement kiln chlorine feedrates and
emissions. Given that all sources must
be able to meet the floor level using
floor control, EPA investigated whether
all CKs could meet the floor level
employing the MACT controls without
being forced to substitute raw material.
As discussed above, all CKs would be
able to meet the floor level using floor
control \vithout switching raw
materials.
Sources would not incur cost to
comply \vith the proposed floor level
because it is higher than any baseline
emission levels in the entire database,
and there would be no emissions
reductions at the floor level.
b. Beyond-the-Floor Considerations.
The neutralization provided naturally
by alkaline raw materials essentially
acts as a dry scrubber to help control
HCl/Cla emissions. Therefore, we do not
believe that substantial further
reductions could be achieved with the
use of dry scrubber systems. Wet
scrubbers, however, could be expected
to provide 99 percent or greater removal
ofHCl/C!2.
BTF control is therefore being defined
as a wet scrubber in conjunction with
the floor control for hazardous waste
chlorine feedrate (defined by a MTEC of
1.6 g/dscm). Given that the proposed
floor level based on hazardous waste
chlorine feedrate control only would be
630 ppmv, the resulting BTF level
would be 6.3 ppmv (at 99 percent
removal).
Selecting a more effective control
technology such as a wet scrubber
would be expensive and the Agency
believes that a BTF level would not be
appropriate. For example, in one
alternate investigation, we evaluated a
25 ppmv HC1 level. The Agency
estimated in that case the national
incremental annualized compliance cost
to meet this level would be $17 million.
This represents HC1/C12 emissions
reductions of 1,900 tons per year or a 71
percent reduction from baseline
emissions. The Agency believes that the
total incremental costs associated with a
standard of 6.3 ppmv would be
approximately equal to the incremental
costs at a BTF level of 25 ppmv. We also
note that, at a MACT floor standard of
630 ppmv, the Agency would not be
required to establish a more stringent
standard under RCRA to ensure
protection of human health and the
environment.
In summary, the Agency is proposing
a MACT floor HC1/C12 standard of 630
ppmv for existing cement kilns.
7. Carbon Monoxide and Hydrocarbons
a. MACT Floor. As discussed in
Section I above, the Agency believes
that control of non-dioxin organic HAP
emissions can be achieved, in part, by
establishing emissions limits on two
surrogate compounds: (1) Carbon
monoxide, and (2) hydrocarbons, and
also by the presence of controls for D/
F. Both CO and HCs are not listed HAPs,
but the Agency is using them as
surrogates for the enumerated organic
HAPs of § 112(b)(l) which can be non-
D/F products of incomplete combustion
(PICs). The Agency is not proposing
main stack MACT standards on carbon
monoxide for existing cement kilns for
reasons discussed below; however,
those kilns with by-pass ducts would be
required to either comply with a
separate CO or HC limit in the by-pass
duct.
i. Carbon Monoxide in the Main
Stack. The Agency is not proposing a
main stack CO limit because CO is not
a universally reliable indicator of
combustion intensity and efficiency in
cement kilns due to CO generation by
process chemistry and evolution from
the trace organics in the raw material
feedstocks.101 These feedstocks can
generate large quantities of CO
emissions which are unrelated to the
combustion efficiency of burning the
waste and fuel. Whereas all the CO from
incinerators is combustion-generated,
the bulk of the CO from cement kilns
can be the result of process events
unrelated to the combustion conditions
at the burner where the wastes are
introduced, or CO can be produced from
CO2 (contained in the limestone) by
dissociation at high sintering
conditions. As a result, few cement
kilns were able to certify compliance
with the CO standard in the BIF rule
(§ 266.104(b)), but instead complied
with the alternative carbon monoxide
standard of § 266.104(c) that allowed CO
to exceed the 100 ppmv limit provided
that stack gas concentrations of HCs did
not exceed 20 ppmv. Thus, the Agency
believes it inappropriate to establish a
CO standard measured in the main stack
for all cement kilns.
ii. Hydrocarbons in the Main Stack.
CKs emit hydrocarbon (HC) emissions
that result from incomplete combustion
of fuels and desorption of trace levels or
101 See 56 FR at 7150, 7153-55 (February 21,
1991).
organic compounds from raw materials.
These HC emissions contain organic
HAPs. Organics in the raw materials are
believed to be primarily from kerogen in
the shale and limestone which has a
porous structure allowing for organic
deposits. These organics cause HC ,
emissions because they are largely not
destroyed given that combustion gases
flow counter-current to the raw-
materials (i.e., fuels are generally fired
at the opposite end from where the raw
materials are fed).
Even when a CK is operated under
good combustion conditions (and thus
is generating low or insignificant levels
of fuel-related HC), HC levels resulting
from organics in the raw materials can
range from 10 to 400 ppmv. This makes
it problematic to use HC as the only or
the principal means to ensure good
combustion efficiency of hazardous
waste fuels to minimize emissions of
toxic PICs (i.e., non-D/F organic HAPs).
Wet Process Kilns and Long Dry
Process Kilns. The BIF rule currently
limits HC levels in the main stack (i.e.,
the only kiln off-gas stack) of wet and
long dry kilns to 20 ppmv. EPA is aware
of five kilns that initially had stack HC
levels exceeding the 20 ppmv limit.
Four of the kilns changed the source of
shale used as raw material to use a shale
with lower organic content. (Shale
comprises a small fraction of raw
material feed.) The fifth kiln feeds
limestone with (relatively) high levels of
organic matter and has indicated that
transporting an alternative source of
limestone to the site may be
prohibitively expensive. Other potential
options, such as installing an
afterburner to destroy organics or
reconstructing the kiln system to
otherwise destroy HC desorbed from the
limestone, may likewise be
prohibitively expensive approaches.
EPA has determined that MACT floor
for HC control for wet and long dry
kilns should be control based on the
current federally-enforceable BIF
standards (i.e., control of organics in
raw materials coupled with operating
under good combustion practices to
minimize fuel-related HC), and the floor
level should be the BIF limit of 20 ppmv
HC for such kilns. We note further that
the source could stop burning
hazardous waste and avoid having to
comply with the HC floor level.
Cement Kilns with By-pass Ducts.
Kilns that are equipped with a by-pass
duct (typically preheater or precalciner
kilns) to divert a portion of the kiln off-
gas to a separate PM control device
monitor fuel-related HC separately from
raw material-related HC. This is because
the by-pass duct diverts the kiln gas
before it enters the calcining zone where
-------
17398 Federal Register 7 Vol. 61, No. 77 I Friday, April 19, 1996 / Proposed Rules
the organics from the raw material are
desorbed. Thus, in general, fuel-related
HC can be monitored in the by-pass
duct, and raw material-related HC can
be monitored in the main stack. We
invite comment on whether hazardous
waste fuel combustion by-products (e.g.,
chlorine) can react with organic
compounds desorbed from raw material
to form organic HAPs. If the Agency
determines that hazardous waste firing
can substantially (adversely) affect
emissions of organic HAPs from the
main stack, then we will consider
limiting HC to 20 ppmv. This is the
limit we are proposing today for long
kilns without a by-pass duct.
Monitoring HC in the by-pass is
discussed later in this section.
The Agency's RCRA BIF rule does not
control HC in the main stack of cement
kilns that comply with the BIF HC limit
in the by-pass duct because, under the
RCRA rule, the Agency was concerned
about PICs derived from hazardous
waste combustion rather than toxic
organics desorbed from raw materials.
Therefore, any MACT standard for HC
in the main stack of these types of kilns
must be a BTF standard since the floor
for these sources is uncontrolled, and
these CKs do not otherwise control
organic HAPs in their stack emissions.
The Agency is concerned that main
stack HC emissions contain HAPs for
several reasons: (1) Organics desorbed
from raw materials, even absent any
influence from burning hazardous
waste, contain HAPs; (2) it is reasonable
to hypothesize that the chlorine released
from burning hazardous waste can react
with the organics desorbed from the raw
material to form generally more toxic
chlorinated HAPs; and (3) some
preheater and precalciner kilns feed
containers of hazardous waste at the
preheater or precalciner end of the kiln
near the by-pass duct entrance such that
hazardous waste PICs may not have
time to combust efficiently. We are
concerned that these hazardous waste
PICs may be emitted from the main
stack, and that monitoring of the by-pass
duct may not be adequate to determine
if inefficient combustion occurs. This is
because the by-pass duct gas may not be
representative of kiln off-gas when
containers of hazardous waste are fed at
the off-gas end of the kiln.
However, the Agency does not now
have sufficient data to quantify the
contribution of hazardous waste (if there
is one) to HC emissions in the main
stack, and therefore to develop a MACT
BTF standard for main stack HC for this
class of CKs. We are thus unable to
propose controls for HC from main
stacks of cement kilns with by-pass
stacks. We invite data to remedy this
situation as well as comment on this
issue. We also invite comment on an
alternative of the same 20 ppmv main
stack HC standard for this class of
cement kilns as for the others.
iii. Emissions Standards for By-pass
Ducts.102 The Agency is proposing that
cement kilns with by-pass ducts
monitor and comply with either a CO or
HC concentration limit in the by-pass
duct because levels .of CO and HC in the
by-pass gas are more representative of
combustion efficiency than levels in the
main stack.103 The BIF rule currently
limits HC (in the by-pass duct) to 20
ppmv.104 MACT floor control is
operating under good combustion
conditions, including conditions that
provide adequate oxygen, temperature,
turbulence, and residence time. These
controls will ensure that kilns with low
organic-containing raw materials are
operating under good combustion
conditions to control PICs formed by the
combustion of hazardous waste fuel.105
EPA's MACT analysis of the existing
by-pass duct data of the best performing
sources resulted in a HC MACT floor
level of 6.7 ppmv. The Agency's
database for CO in the by-pass is
incomplete for the purposes of
calculating a statistically-derived
emission limit, but we believe that it is
reasonable and appropriate to establish
the by-pass CO floor level at the same
level allowed in the BIF rule—100
ppmv. Under this standard the facility
would have the option of complying
with either the CO or HC standard in the
by-pass duct.
102 Most precalciner and some preheater kilns are
equipped with by-pass ducts where a portion (e.g.,
5 to 30 percent) of the kiln exhaust is diverted to
a separate APGD, and, sometimes, a separate stack.
These gases are typically diverted to avoid a build-
up of metal salts that can adversely affect the
calcination process.
103 provided that: (1) hazardous waste is fired
only into the kiln (i.e., not at any location
downstream from the kiln exit relative to the
direction of gas flow); and (2) the by-pass duct gas
is representative of kiln gas. To ensure by-pass gas
is representative of kiln gas, the by-pass duct must
divert a minimum of 10 percent of kiln off-gas as
currently required in the BIF rule. See 266.104(g).
104 The BIF rule provides for an alternative
emissions standard for CO of 100 ppmv. See
§ 104(f).
105 when the by-pass duct is vented through a
separate stack, compliance with limits on CO or HC
would ensure application of MACT regarding fuel-
related organic HAPs. When the by-pass is routed
back into the main (only) stack, compliance with
limits on CO or HC will likewise ensure application
of MACT regarding fuel-related organic HAPs.
Absent these controls on the by-pass duct, fuel-
related organic HAPs could be either: (l) masked by
raw material-related HAPs, if the raw material
contains substantial organics; or (2) if the raw
material contains low levels of organics, the kiln
could comply with the main stack standard (if one
were proposed) while operating under poor fuel
combustion conditions.
The Agency also invites comment on
requiring cement kilns with by-pass
ducts to comply with both the CO and
HC standard (measured in the by-pass
duct). Given that CO in the by-pass duct
should be related only to fuel
combustion efficiency, monitoring of
CO in addition to HC may be
appropriate to ensure complete
combustion of organics in the kiln;
however, the Agency is concerned that
some CO may be generated from the CO2
by dissociation at high sintering
temperatures and thus requests
information and data on this option.
Cement kiln sources would not incur
costs to comply with the proposed floor
level since all cement kilns with by-pass
ducts (for which EPA has data)
currently meet the floor level for either
HC or CO. EPA also notes that
approximately half of cement kilns that
measured both HC and CO in the by-
pass achieved the floor level.
As mentioned above, the Agency is
aware of a long wet process cement kiln
that is unable to comply with either the
CO limit of 100 ppmv or the HC limit
of 20 ppmv in the main stack. This kiln
cannot achieve either of these levels due
to the relatively high organic matter
content in the limestone. Since the
majority of the raw material fed to the
kiln is limestone, substitution with an
alternative source of limestone with
lower organic content is not readily
feasible (e.g., prohibitively expensive
transportation costs of a substitute raw
material). The facility attempted to
retrofit the kiln with a by-pass duct thus
allowing monitoring of CO or HC in the
by-pass duct as permitted by current BIF
regulations. However, efforts to
construct and engineer this kiln with a
by-pass duct were not successful due to
the length of the kiln.106
In coordination with state and
regional officials, the cement kiln was
retrofitted with a mid-kiln sampling
port that continuously draws off a
portion of the kiln combustion gas for
analysis of HC or CO. Since this
sampling port does not divert a
minimum of 10 percent of the kiln off-
gas from the kiln, it does not meet the
Agency's current definition of a by-pass
duct defined in § 266.104(g). The kiln's
mid-kiln sampling port diverts
approximately 7 to 8 percent of the kiln
off-gas. The Agency specifically invites
comment on allowing sources with a
mid-kiln sampling port, or other kiln gas
extraction mechanism, that is capable of
continuously extracting a representative
sample of kiln off-gas to comply with
106For example, the kiln experiences a substantial
increase in length due to expansion during start-up
as the kiln heats up to operating levels.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17399
the same HC and CO standards
proposed for kilns with hy-pass ducts.
Commenters should specifically address
how the gas extraction system ensures
that a representable sample of the kiln's
fuel combustion gas would be
monitored for HC or CO.
b. Beyond-the-Floor Considerations.
EPA has considered BTF control for
organic HAP emissions from the main
stack of all CKs (including those with
by-pass ducts) based on use of a
combustion gas afterburner. We believe
that a BTF level for CO of 50 ppmv and
for HC of 6 ppmv are readily achievable
with an afterburner, but not appropriate.
Therefore, we are not proposing such a
BTF standard. EPA has no data
indicating that any cement kilns are
currently meeting these BTF levels with
existing controls. The annualized
engineering costs for the cement kilns to
meet these BTF levels is estimated to be
$280 million, and would provide an
incremental reduction in HC emissions
nationally beyond the floor controls of
approximately 1500 tons per year and
65,000 tons per year for CO.
8. MACT Floor Cost Impacts
The total national annualized
compliance costs107 for existing cement
kilns to meet all the MACT floor levels
are estimated to be $34 million with the
cost per cement kiln averaging
$777,000. On a cost per ton of
hazardous waste burned, these total
compliance costs equate to $40 per ton
of waste. We estimate that up to 2
cement facilities will likely cease
burning hazardous waste due to the
compliance costs associated at the floor.
The Agency is proposing to go
beyond-the-floor for two pollutants for
existing cement kilns: dioxins/furans
and mercury. The total national
annualized compliance costs (i.e., total
costs not incremental costs from the
floor levels) to meet the dioxin/furan
and mercury BTF levels in addition to
the MACT floor levels for the remaining
HAPs are estimated to be $44 million
with the cost per cement kiln averaging
$1,04 million. On a cost per ton of
hazardous waste burned, these total
compliance costs increase to $50 per ton
of waste. Again, we estimate that up to
2 cement facilities \vill likely cease
burning hazardous waste due to the
compliance costs associated with the
proposed standards.
""Compliance costs represent pre-tax
compliance costs. Because compliance costs are tax-
deductible, the portion of pre-tax costs borne by the
firm would bo between 70 and 80 percent of the
values shown above, depending on the specific
firm's margin tax bracket. See "Regulatory Impact
Assessment for Proposed Hazardous Waste
Combustion MACT Standards", November 13,
1905, for details.
B. MACT for New Hazardous Waste-
Burning Cement Kilns
This section summarizes EPA's
rationale for establishing MACT for new
cement kilns for each HAP, HAP
surrogate, or HAP group. Table IV.4.B.1.
summarizes the proposed emissions
limits for new cement kilns, which were
determined using the analytical process
described in Part Three, Section VII and
in the technical background document.
TABLE IV.4.B.1.—PROPOSED MACT
STANDARDS FOR NEW CEMENT KILNS
HAP or HAP surro-
gate
Dioxin/furans (TEQ)
Paniculate Matter ....
Mercury
SVM (Cd, Pb)
LVM (As, Be, Cf, Sb)
HCI + CI2 (total
chlorides).
Hydrocarbons:
Main Stackc
By-pass Stackd
Carbon Monoxide:
Main Stack
By-pass Stackd
Proposed standarda
0.20 ng/dscm (TEQ).
69 mg/dscm (0.030
gr/dscf)-
50 u.g/dscm.
55 u.g/dscm.
44 ng/dscm.b
67 ppmv.
20 ppmv.
6.7 ppmv.
N/A.
100 ppmv.
"All emission levels are corrected to 7 per-
cent O2.
bAn alternative standard of 80 |xg/dscm
would apply if the source elects to document
compliance using a multi-metals GEM.
c Applicable only to long wet and dry proc-
ess cement kilns (i.e., not applicable to pre-
heater and/or precalciner kilns).
d Emissions standard applicable only for ce-
ment kilns configured with a by-pass duct
(typically preheater and/or precalciner kilns).
Source must comply with either the HC or CO
standard in the by-pass stack. A long wet or
long dry process cement kiln that has a by-
pass duct has the option of meeting either the
HC level in the main stack or the HC or CO
limit in the by-pass duct.
1. MACT New for Dioxins/Furans
a. MACT New Floor. As for existing
cement kilns, the Agency is identifying
MACT new floor for D/F based on
temperature control at the inlet to the
ESP or FF. EPA characterized the single
best performing source with the lowest
TEQ dioxin/furan emissions and
determined that the best performing
source had an inlet temperature of
409°F or less.
The Agency then evaluated D/F
emissions from all kilns that operated
the ESP or FF at 409°F or less and
determined that 75 percent had D/F
emissions less than 0.2 ng/dscm (TEQ).
The other 25 percent of kilns generally
had TEQs less than 0.8 ng/dscm (TEQ),
although one kiln emitted 4.7 ng/dscm
(TEQ). The Agency notes that the MACT
new expanded pool was virtually
identical (with the exception of two test
conditions) to the expanded pool of
existing sources.
The Agency is, therefore, identifying
temperature control at the inlet to the
ESP or FF at 409°F as the MACT floor
control. Given that 75 percent of sources
achieve D/F emissions of 0.20 ng/dscm
(TEQ) at that temperature, the Agency
believes that it is appropriate to express
the floor as "0.20 ng/dscm (TEQ), or
(temperature at the inlet to the ESP or
FF not to exceed) 409°F". This would
allow sources that operate at
temperatures above 409°F but that
achieve the same D/F emissions as the
majority of sources that operate below
409°F (i.e., 0.20 ng/dscm (TEQ)) to meet
the standard without incurring the
expense of lowering the temperature at
the ESP or FF.
b. Beyond-The-Floor Considerations.
The Agency has identified activated
carbon injection (CI) at less than 400°F
as a BTF control for D/F for cement
kilns because CI is currently used in
similar applications such as hazardous
waste incinerators, municipal waste
combustors, and medical waste
incinerators. The Agency is not aware of
any CK flue gas conditions that would
preclude the applicability of CI or
inhibit the performance of CI that has
been demonstrated for other waste
combustion applications.
Carbon injection has been
demonstrated to be routinely effective at
removing greater than 95 percent of D/
F and some tests have demonstrated a
removal efficiency exceeding 99 percent
at gas temperatures of 400°F or less. To
determine a BTF emission level, the
Agency considered the emission levels
that could result from gas temperature
control to less than 400°F combined
with CI.
As discussed for existing sources,
when CI is used in conjunction with
temperature control, an additional 95
percent reduction in emissions could be
expected. Accordingly, emissions with
BTF controls could be expected to be
less than a range of 0.04 to 0.24 ng/dscm
(TEQ) (i.e., 95 percent reduction from
0.8 ng and 4.7 ng, respectively). Given
that CI reductions greater than 95
percent are readily feasible, the Agency
believes that it is appropriate to identify
0.20 ng/dscm (TEQJ as a reasonable BTF
level that could be routinely achieved.
The Agency notes that, because we
have assumed a fairly conservative
carbon injection removal efficiency of
95 percent to identify the 0.20 ng/dscm
(TEQ) level, we believe that this
approach adequately accounts for
emissions variability at an individual
kiln because CI removal efficiency is
likely to be up to or greater than 99
percent. EPA thus believes that it is not
-------
17400 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
necessary to add a statistically-derived
variability factor to the 0.20 ng/dscm
(TEQ) level to account for emissions
variability at an individual kiln. Thus,
the 0.20 ng/dscm (TEQ) BTF level
represents the proposed D/F emission
standard for new cement kilns.
EPA solicits comment on this
approach, and notes that if a
statistically-derived variability factor
were deemed appropriate, the BTF level
of 0.20 ng/dscm (TEQ) would be
expressed as a standard of 0.31 ng/dscm
(TEQ). We note, however, that under
this approach, it may be more
appropriate to use a less conservative CI
removal efficiency (i.e., because
emissions variability would be
accounted for using statistics rather than
in the engineering decision to use a
conservative CI removal efficiency),
thus lowering the 0.20 ng/dscm (TEQ)
level to approximately 0.04 ng/dscm
(TEQ) (i.e., 99 percent reduction from
0.8 ng and 4.7 ng results in levels of
0.008 ng to 0.047 ng/dscm (TEQ),
respectively, and 0.04 ng is a reasonable
value within this range). If so, the D/F
standard would be about 0.15 ng/dscm
(TEQ) (i.e., 0.04 ng/dscm TEQ plus the
variability factor of 0.11 ng/dscm TEQ).
For similar reasons as discussed for
existing cement kilns, the Agency is
proposing a BTF standard for D/F of
0.20 ng/dscm (TEQ) for new hazardous
waste-burning cement kilns. Costs for
new sources are discussed in
"Regulatory Impact Assessment for
Proposed Hazardous Waste Combustion
MACT Standards".
2. MACT New for Particulate Matter
a. MACT New Floor. The Agency
analyzed all available PM emissions
data and determined that the control
used by the single best performing
source used a fabric filter with an air-
to-cloth (A/C) ratio of 1.8 acfm/ft2 or
less. Analysis of emissions data from all
CKs using FFs with the 1.8 acfm/ft2 A/
C ratio or less resulted in a level of
0.065 gr/dscf.
For. similar reasons discussed for
existing cement kilns, the Agency has
chosen the existing NSPS standard (an
established regulatory benchmark for
PM), not the statistically-derived limit,
as the MACT for existing hazardous
waste-burning cement kilns. Thus, the
Agency is identifying a MACT floor for
PM and is identifying the floor level as
the NSPS limit of 69 mg/dscm (0.03 gr/
dscf) because it is the lowest federally
enforceable emission standard.
b. Beyond-the-Floor Considerations.
EPA considered but is not proposing a
more stringent BTF level (e.g., 35 mg/
dscm (0.0105 gr/dscf)) for new cement
kilns. For the same reasons discussed
for existing sources, the Agency believes
that a more stringent level than the floor
is not warranted.
3. MACT New for Mercury
a. MACT New Floor. As discussed
earlier, hazardous waste-burning cement
kilns control their mercury input (and
therefore much of their emissions)
through control of the mercury content
in the hazardous waste. The Agency is
defining the MACT floor technology as
feedrate control with a hazardous waste
MTEC less than 28 ng/dscm based on
performance of the best performing
source. Analysis of all existing cement
kiln sources using this hazardous waste
feedrate control resulted in a MACT
new floor level of 82 ng/dscm. EPA
estimates that a source with average
emissions variability must be designed
and operated to routinely achieve an :
emission level of 58 ng/dscm to meet
this standard 99 percent of the time.
Expanded MACT pools are identical.
The MACT new floor analysis results in
the same floor as existing sources
because their respective expanded
MACT pools are identical.
EPA solicits comment on an
alternative method to establishing the
MACT new floor. Under this alternative,
the floor analysis would be similar to
the approach proposed today except
that the variability factor would be
added to the average emissions from the
single best performing source. By
contrast, under the approach proposed
today, the variability factor is added to
the emissions of the highest emitting
source in the expanded MACT pool.
Thus, under this alternative the only
purpose that expanding the MACT pool
would serve is to identify the variability
factor. EPA notes that this approach
results in a MACT new floor of 53 ng/
dscm (4.4 ng/dscm (average emissions
from the best performing source) plus
the statistically-derived variability
factor of 49 ng/dscm).
b. Beyond-the-Floor Considerations.
The Agency has considered the same
BTF control alternatives for improved
Hg control for new cement kilns:
hazardous waste feedrate control of Hg
in conjunction with flue gas
temperature reduction to 400°F or less
followed by either carbon injection (CI)
or carbon bed (CB). The BTF design
emission level under the Cl-controlled
option is 30 ng/dscm (assuming a source
has controlled its Hg emissions to 300
ng/dscm controlling Hg feed in the
hazardous waste). The BTF emission
standard corresponding to a design level
of 30 ng/dscm would be 50 ng/dscm108.
The Agency is proposing 50 ng/dscm as
the MACT standard for new cement
kilns. The Agency specifically requests
comment on establishing BTF emission
standards based on the alternative .
approaches discussed for existing ,
cement kilns.
4. MACT New for Semi volatile Metals
a. MACT New Floor. MACT new
control is based on hazardous waste
feedrate control and PM control. EPA
characterized the single best performing
source with the lowest SVM emissions
and determined that the best performing
source used a fabric filter with an air-
to-cloth ratio of 2.1 acfm/ft2 or less for
a kiln system with a hazardous waste
(HW) MTEC of 36,000 ng/dscm or less.
Analysis of all sources (i.e., expanded
MACT pool of facilities) using this
technology or better resulted in a floor
level of 55 ng/dscm for new cement
kilns.
EPA solicits comment on an
alternative method to establishing the
MACT new floor. Under this alternative,
the floor analysis would be similar to
approach proposed today except that
the variability factor would be added to
the average emissions from the single
best performing source. Thus, the
expanded MACT pool serves only to
identify the variability factor of the floor
technology. EPA notes that this
approach results in a MACT new floor
of 39 ng/dscm (4 ng/dscm (average
emissions from the best performing
source) plus the statistically-derived
variability factor of 35 ng/dscm).
b. Beyond-the-Floor Considerations.
The Agency considered a more stringent
level than the floor level of 55 ng/dscm
based on improved collection efficiency
of the MACT floor FF. Since this level
is virtually identical to the floor level
for existing sources and considering that
EPA is not proposing standards more
stringent than the floor for existing
sources, the Agency believes for the
same reasons that a more stringent floor
level is not warranted for new sources
as well. Finally, we note that
establishing the MACT standard at the
floor would not trigger the need for a
more stringent standard under RCRA.
5. MACT New for Low-Volatile Metals
a. MACT New Floor. MACT new
control is based on hazardous waste
feedrate control and PM control. EPA
characterized the best particulate
control device, and identified the floor
technology as abaghouse (i.e., fabric
filter) with an air-to-cloth ratio of 2.3
acfm/ft2 or less with a hazardous waste
108 To achieve a standard of 50 ng/dscm 99
percent of the time, a source with average emissions
variability must be designed and operated to
achieve an emission level of 30 ng/dscm.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17401
(HW) MTEC less than 25,000 ug/dscm.
Analysis of the expanded MACT pool
resulted in a floor emissions level of 44
Hg/dscm for new cement kilns.
EPA solicits comment on an
alternative method to establishing the
MACT new floor. Under this alternative,
the floor analysis would be similar to
the approach proposed today except
that the variability factor would be
added to the average emissions from the
single best performing source. Thus, the
expanded MACT pool only serves to
identify the variability factor of the floor
technology. EPA notes that this
approach results in a MACT new floor
of 30 ug/dscm (4 ug/dscm (average
emissions from the best performing
source) plus the statistically-derived
variability factor of 26 ug/dscm).
b. Beyond-the-Floor Considerations.
The Agency considered a more stringent
level than the floor of 44 pg/dscm based
on improved collection efficiency of the
MACT floor FF. We initially determined
that selecting such a BTF level is not
warranted for several reasons: (1) It
would not likely be cost effective
considering the small increment of
LVMs removed; (2) LVM are not of
particular concern because they are not
bioaccumulative; (3) establishing the
MACT standard at the MACT new floor
would not trigger the need for a more
stringent RCRA standard.
The Agency is proposing an
alternative compliance option for LVMs
for new cement kilns. Because the
Agency anticipates the likelihood of
development of a multi-metals
continuous emissions monitor (GEM) in
the near future and considering that the
estimated detection limit for the CEM to
be approximately 80 ug/dscm for the
LVM metals combined, the Agency is
proposing an alternative standard of 80
ug/dscm should the source elect to
document compliance using a multi-
metals CEM. Thus, the LVM standard is
•different depending on the compliance
method selected.
6. MACT New for Hydrochloric Acid
and Chlorine
a. MACT New Floor. Cement kilns use
the natural alkalinity of the limestone
used as raw material and hazardous
waste feedrate control to control HC1
and Ch emissions. Thus, the MACT
floor is based on hazardous waste
feedrate control.
EPA characterized the single best
performing source with the lowest HC1/
Cl2 emissions and determined that the
best performing source used feedrate
control with a hazardous waste (HW)
MTEC of 1.6 g/dscm or less. (Combined
emissions of HC1 and Clz were
expressed as HC1 equivalents.) Analysis
of the expanded MACT pool of facilities
resulted in a floor level of 630 ug/dscm
for new cement kilns, which is the same
result as for existing cement kiln
sources because the expanded MACT
pools are identical for both existing and
new cement kilns. ;
Again, as discussed for existing
cement kilns, this determination is
confounding given that the highest
average emissions from any test,
condition in the entire database,
irrespective of hazardous waste MTEC
for total chlorine, was 220 ppmv. This
anomalous finding is apparently
attributable to: (1) The data set having
very high average within-test-condition
variability; and (2) adding the average
variability factor to the log mean rather
than the arithmetic mean of the test
condition within the expanded MACT
pool (those sources using MACT floor
control) with the highest arithmetic
mean. If that source had unusually high
emissions variability, then the log mean
could be substantially higher than the
arithmetic mean, resulting in an
unusually high emission level to which
the variability factor was added.
Because of these concerns, the Agency
invites comment on alternative
approaches that may identify a mor,e
reasonable floor level. One approach
could be to add the average variability
factor for the data set to the arithmetic
mean, rather than the log mean, of the
highest test condition in the expanded
MACT pool. In addition, if this still
resulted in a calculated floor level
greater than any emission level in the
database, irrespective of hazardous
waste MTEC for total chlorine, the floor
level could be capped at the highest
emission level in the database—220
ppmv.
b. Beyond-the-Floor Considerations.
BTF control is being defined as a wet
scrubber in conjunction with the floor
control for hazardous waste chlorine
feedrate. As discussed earlier for
existing systems, more stringent HC1
and C12 control based on use of wet
scrubbers is readily achievable. The
Agency is aware of two cement kilns
(not burning hazardous waste) that
employ a wet and dry scrubber,
respectively, capable of HCl/Ck capture.
Wet scrubber use within the hazardous
waste incineration industry is well
established also, often achieving capture
efficiencies exceeding 99 percent.
Considering that average HC1/C12
emissions from existing cement kilns
range from less than 1 ppmv to 220
ppmv and that a well-designed and
operated wet scrubber •would be
expected to achieve removal efficiencies
greater than 90 percent, if not higher,
the Agency believes that HCl/Ck control
to a standard of 67 ppmv (corresponding
to a design level of 25 ppmv109) is
readily achievable.110 Thus the Agency
is proposing a HCl/Ck standard of 67
ppmv for new cement kilns. See
"Regulatory Impact Assessment for
Proposed Hazardous Waste Combustion
MACT Standards" for further details on
the costs.
7. MACT New for Carbon Monoxide and
Hydrocarbons
a. MACT Floor. The Agency believes
that control of non-dioxin organic HAP
emissions (i.e., non-dioxin PICs that are
also HAPs) can be achieved by
establishing emissions limits on
hydrocarbons and carbon monoxide. As
discussed earlier for existing cement
kilns, the Agency is proposing a MACT
standard of 20 ppmv for HCs in the
main stack (not applicable for preheater
and precalciner kilns), and either a CO
limit of 100 in the by-pass duct or HC
standard of 6.7 ppmv in the by-pass
duct. Thus, the proposed standards for
new cement kilns are identical to those
for existing kilns.
b. Beyond-the-Floor Considerations.
As for existing sources the Agency
requests comment on a main stack
hydrocarbon standard of 6 ppmv and a
carbon monoxide standard of 50 ppmv
for all new cement kilns (including
those with by-pass ducts) based on
performance of a combustion gas
afterburner to burn-out incompletely
combusted organics that escape the
primary combustion zone.
8. MACT New Cost Impacts
A discussion of the costs and
economic impacts for new cement kilns
is presented in Part Seven of today's
proposal.
C. Evaluation ofProtectiveness
In order to satisfy the Agency's
mandate under the RCRA to establish
standards for facilities that manage
hazardous wastes and issue permits that
are protective of human health and the
environment, the Agency conducted an
analysis to assess the extent to which
109 Considering the highest total chlorine data
point of 220 ppmv with a 90 percent removal
efficiency yields a design level of approximately 25
ppmv.
"°The Agency notes that assuming a 99 percent
capture efficiency would result in a design level of
approximately 2.2 ppmv (corresponding to an
emission level of 6.7 ppmv). Since the application
of wet scrubbers is still limited in the cement
industry, EPA believes that a total chlorine standard
of 6.7 ppmv is unnecessarily low and is thus
assuming a more conservative total chlorine
removal efficiency of 90 percent. In addition, the
Agency notes that further controls under RCRA
would not be necessary at a level of 67'ppmv
(corresponding to a design level of 25 ppmv) for
new cement kilns.
-------
17402
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
potential risks from current emissions
would be reduced through
implementation of MACT standards.
The analysis conducted for hazardous
waste-burning cement kilns is similar to
the one described above for hazardous
waste incinerators. The procedures used
in the Agency's risk analyses are
described in detail in the background
document for today's proposal.111 In
evaluating the MACT standards, the
Agency used the design value which is
the value the Agency expects a source
would have to design to in order to be
assured of meeting the standard on a
daily basis and hence is always a lower
value than the actual standard for all
HAPs controlled by a variable control
technology.112
The risk results for hazardous waste-
burning cement kilns are summarized in
Table IV.4.G.1 for cancer effects and
Table IV.4.C.2 for non-cancer effects for
the populations of greatest interest,
namely subsistence farmers, subsistence
fishers, recreational anglers, and home
gardeners. The results are expressed as
a range where the range represents the
variation in exposures across the
example facilities (and example
waterbodies for surface water pathways)
for the high-end and central tendency •
exposure characterizations across the
exposure scenarios of concern. For' '•-'< ..
example, because dioxins • " ,-"
bioaccumulate in both meatFand fish,
the subsistence farmer and subsistence
fisher scenarios are used to determine
the range.113
TABLE IV.4.C.1—INDIVIDUAL CANCER RISK ESTIMATES FOR CEMENT KILNS'
Dioxins
Semi-volatile met-
als2
Low volatile met-
als3
Existing Sources
Floor • •
BTF •
1E-8to9E-5
4E-9 to2E-54 ...
4E-9to2E-65.
1E-9to4E-7 .....
3E-9to1E-7
5E-1 1 to 5E-7
9E-9 to 4E<-6
New Sources
BTF .
CEM Option6
4E-9 to 2E-54 ...
4E-9 to2E-65.
3E-9to 1E-7
3E-9to1E-7
3E-9 to 1 E-6
1 E-8 to 4E-6
1 Lifetime excess cancer risk.
2 Carcinogenic metal: cadmium. ' - -
3 Carcinogenic metals: arsenic, beryllium, and chromium (VI).
"Based on 0.2 ng/dscm TEQ as a central tendency estimate and 1.4 ng/dscm TEQ as a high-end estimate.
s Based on 0.20 ng/dscm TEQ.
6 Based on SVM standard of 60 ug/dscm and LVM standard of 80 ng/dscm (applicable only if the source elects to document compliance using
a multi-metals CEM).
TABLE IVAC.2—INDIVIDUAL NON-CANCER RISK ESTIMATES FOR CEMENT KILNS '
Semi-volatile met-
als2
Low volatile met-
als3
Hydrogen chlo-
ride
Chlorine ,
Existing Sources
Ploor •
<0.001 to 0.06 ....
<0.001 to 0.004
<0.001 to 0.004
<0.001 to 0.01 ....
<0.001 to 0.04 ....
0.01 to 0.1 4
<0.001 to 0.06
0.05 to 0.8s
New Sources
BTF ......
CEM Option6 1 ....
<0.001 to 0.004
<0.001 to 0.004
<0.001 to 0.005
O.001 to 0.01 ....
0.01 to 0.1 4
0.001 to 0.01 * ....
0.05 to 0.8 s
o.oo5 to o.oa?
1 Hazard quotient. . .
2 Cadmium and lead. . • . . •
3 Antimony, arsenic, beryllium, and chromium.
4 HCI + CI2 assuming 100 percent HCL,
5 HCI + CI2 assuming 10 percent Cl2. •
6 Based on SVM standard of 60 ng/dscm and LVM standard of 80 ng/dscm (applicable only if the source elects to document compliance using
a multi-metals CEM). ,
The risk analysis indicates that for the
semi-volatile and low-volatile metals
categories, the MACT standards for
cement kilns are protective at the floor
for both existing and new sources. The
analysis indicates that the CEM
compliance option for new sources is
also protective. For hydrogen chloride
and chlorine (C12), the MACT standards
for cement kilns are also protective at
the floor for both existing and new
sources. However, the analysis indicates
1'' "Risk Assessment Support to the Development
of Technical Standards for Emissions from
Combustion Units Burning Hazardous Wastes:
Background Information Document," February 20,
1995.
112 For the semi-volatile and low volatility metals
categories, the Agency assumed the source could
emit up to the design value for each metal in the
category for the purpose of assessing protectiveness.
>13For the semi-volatile and low volatility metals
categories, the inhalation MET scenarios are also
used. For hydrogen chloride and chlorine (C12) only
the inhalation MEI scenarios are used.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17403
that for dioxins the proposed beyond
the floor standards, rather than the floor
levels, are protective.
V. Lightweight Aggregate Kilns: Basis
and Level for the Proposed NESHAP
Standards for New and Existing
Sources
Today's proposal would establish
maximum achievable control
technology (MACT) emissions standards
for dioxin/furans, mercury, semivolatile
metals (cadmium and lead), low volatile
metals (arsenic, beryllium, chromium,
and antimony), particulate matter (PM),
acid gas emissions (hydrochloric acid
plus chlorine), hydrocarbons, and
carbon monoxide from existing and new
hazardous waste-burning lightweight
aggregate kilns (LWAKs). See proposed
§ 63.1205. The following discussion
addresses how MACT floor and beyond-
the-floor (BTF) levels were established
for each HAP and EPA's rationale for
the proposed standard. The Agency's
overall procedural approach for MACT
determinations has been discussed in
Part Three, Sections V and VI for
existing sources and in Section VII for
new sources.
Again, the Agency wishes to
emphasize that these standards were
developed using a database that
contains primarily short-term
certification of compliance data that
may not adequately reflect more normal,
day-to-day operations and emissions. As
noted earlier, EPA believes it preferable
to use long-term, more normal operating
emissions data for MACT standard-
setting purposes and specifically invites
commenters to submit this type of data.
A. Summary of MACT Standards for
Existing LWAKs
This section summarizes EPA's
rationale for establishing the MACT
floor emission level and choosing
MACT for existing LWAKs for each
HAP, HAP surrogate, or HAP group.
Table IV.5.A.1 summarizes the MACT
standards for existing LWAKs. The basis
for the floor level and BTF
considerations for each HAP or HAP
surrogate is then discussed.
Table IV.5.A.1 .—PROPOSED MACT
STANDARDS FOR EXISTING LWAKs
Table IV.5.A.1.—PROPOSED MACT
STANDARDS FOR EXISTING
LWAKs—Continued
HAP or HAP surro-
gate
Dioxin/furans
Particulate Matter
Mercury ,
SVM [Cd, Pb]
LVM [As, Be, Cr, Sb]
HC! + CI2 ,
CO
Proposed standards1
0.20 ng/dscm TEQ.
0.030 gr/dscf (69 mg/
dscm)
72 ng/dscm.
12 ng/dscm.2
340 ng/dscm.
450 ppmv.
100 ppmv.
HAP or HAP surro-
gate
HC
Proposed standards 1
1 4 ppmv.
1 All emission levels are corrected to 7 per-
cent O2.
2 An alternative standard of 60 jig/dscm
would apply if the source elects to document
compliance using a multi-metals CEM.
1. Dioxin/Furans
a. MACT Floor. EPA has obtained
dioxin/furan (D/F) emissions data for
only one LWAK. The data indicated an
average test condition D/F emission of
0.04 ng/dscm (TEQ). Based on the
Agency's data on the performance of D/
F control technology, the Agency is
identifying the MACT floor for D/F
based on temperature control at the inlet
to the fabric filter. EPA is therefore
identifying the MACT floor level for D/
F emissions from LWAKs as 0.20 ng/
dscm (TEQ) or (temperature at the PM
control device not to exceed) 418° F.
Given that EPA is not aware of any
LWAKs that exceed the floor level, the
rule would not require these sources to
incur costs to achieve compliance.
The Agency recognizes that its data
on dioxin/furan emissions from LWAKs
is limited. Therefore, the Agency is
inviting commenters to submit
additional performance data on LWAK
D/F emissions.
b. Beyond-The-Floor Considerations.
The BTF considerations for LWAKs
were the same as for CKs. Therefore,
EPA is proposing a BTF standard of 0.20
ng/dscm (TEQ) for the same reasons
applicable to CKs. As noted above,
given that EPA is not aware of any
LWAKs that exceed the proposed BTF
standard, LWAKs should not have to
incur costs to achieve compliance. EPA
notes, however, that LWAKs would
nonetheless be required to comply with
operating limits established during
performance testing and conduct
periodic D/F testing to document
compliance with the rule. These costs
are relatively low when compared to the
cost of complying with other provisions
of today's rule.
2. Particulate Matter
a. MACT Floor. LWAKs, like cement
kilns, have high particulate inlet
loadings to the particulate control
device due to the nature of the
lightweight aggregate manufacturing
process; that is, a significant portion of
the finely pulverized raw material fed to
the kiln is entrained in the flue gas
entering the control device. LWAKs are
equipped with fabric filters, although
one facility is equipped with a spray
dryer, venturi scrubber and wet
scrubber, in addition to the fabric filter,
to control PM to a 0.08 gr/dscf standard
under the BIF rule. The PM data for
LWAKs include results from 15 test
conditions collected from 6 facilities,
with a total of 12 units being tested. The
Agency's database shows that the
average controlled PM emissions ranged
from 0.0005 gr/dscf to 0.02 gr/dscf,
corrected to 7 percent oxygen, dry basis.
The Agency analyzed all available PM
emissions data and determined that
sources with emission levels at or below
the level emitted by the median of the
best performing 12 percent of sources
used a fabric filter with an air-to-cloth
ratio of 2.8 acfm/ft2 or less. EPA's
analysis of all LWAKs employing this
floor technology resulted in a MACT
floor emissions level of 110 mg/dscm
(0.049 gr/dscf). EPA estimates that 100
percent of LWAKs are currently meeting
the floor level. The national annualized
compliance cost for LWAKs to meet the
floor level is estimated to be $290,000
for the entire LWAK industry.
b. Beyond-The-Floor Considerations.
EPA is proposing a more stringent
beyond-the-floor (BTF) level of 69 mg/
dscm (0.03 gr/dscf) for LWAKs. As
mentioned above, since 1971, some
cement kilns have been subject to the
more stringent NSPS (see 40 CFR 60.60,
Subpart F) of 0.3 Ib/ton of raw material
feed (dry basis) to the kiln, which is
generally equivalent to 69 mg/dscm
(0.03 gr/dscf). Because of design and
process similarities between LWAKs
and cement kilns, such as high inlet
grain loading and similar APCDs, the
Agency believes that 69 mg/dscm is
achievable for LWAKs.
EPA estimates that 80 percent of
LWAKs are currently meeting this BTF
level. The Agency estimates that there
would be no national incremental
annualized compliance cost for the
remaining LWAKs to meet the BTF level
rather than comply with the floor
controls. This is because sources are
already meeting the BTF level, or they
would be able to meet it with the
upgrades or retrofits needed to meet the
floor level. The BTF level would
provide an incremental reduction of 4
tons per year, or 9 percent, in PM
emissions nationally beyond that
achieved with floor controls. (Note that
emissions reductions estimates are
based on the design level, not the
standard.) Therefore, the Agency is
proposing a MACT standard of 69 mg/
dscm (0.030 gr/dscf) for existing
LWAKs.
EPA considered but is not proposing
an alternative more stringent beyond-
-------
17404
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
the-floor level (e.g., 35 mg/dscm (0.015
gr/dscf)) for LWAKs. EPA notes that, to
ensure compliance with a 35 mg/dscm
standard 99 percent of the time, a source
with average emissions variability must
be designed and operated to achieve an
emission level of approximately 18 mg/
dscm. EPA estimates that 60 percent of
LWAKs currently have average PM
emissions below 18 mg/dscm.
All of the remaining LWAKs may
require the installation of new fabric
filters to comply with the proposed
standards for all HAPs discussed in
today's rule. The average emissions
level for the 40 percent of LWAKs that
do not meet a PM emission level of 18
mg/dscm is 28 mg/dscm. All of these
LWAKs would require an upgrade from
fiberglass bags to improved performance
filter media on the newly installed
fabric filters. Although the engineering
costs to comply with a PM design level
of 18 mg/dscm is modest for LWAKs,
the resulting reduction in PM emissions
is minimal because 40 percent of the
kilns are emitting at an average emission
level slightly above the BTF level.
Lowering the PM design level to 18 mg/
dscm may not be appropriate based on
this minimal impact on overall PM
emissions.
Thus, EPA specifically invites
comment on whether the final rule
should establish BTF standard for PM of
35 mg/dscm (or 0.15 Ib/ton of raw
material (dry basis) feed into the kiln).
3. MACT for Mercury
a. MACT Floor. Mercury emissions
from LWAKs are currently controlled by
the BIF rule, and LWAKs have elected
to comply with the BIF standard by
limiting the feedrate of Hg in the
hazardous waste.114 Thus, the MACT
floor is based on hazardous waste feed
control.
The LWAK mercury emissions data
reflect results from 13 test conditions
collected from 6 facilities, with a total
of 10 kilns being tested. The average
mercury emissions for the test
conditions ranged from 0.4 ug/dscm to
560 ug/dscm.
To identify the floor level for
hazardous waste feed control, the
Agency determined that sources with
Hg emissions at or below the level
emitted by the median of the best
performing 12 percent of sources had
normalized hazardous waste feedrates
(i.e., MTECs)115 of Hg of 17 ug/dscm or
114 EPA notes that one LWAK is equipped with
a venturi scrubber that can provide control of Hg.
That kiln, however, is the highest Hg-emitting kiln
in our database because, EPA believes, it burns
waste with high levels of Hg.
115MTEC, or maximum theoretical emission
concentration, is calculated as the feedrate of (Hg)
less. Analysis of all LWAKs using this
level of hazardous waste feedrate of Hg,
or less (i.e., sources having a MTEC of
17 ug/dscm or less), resulted in a MACT
floor level of 72 ug/dscm. To meet this
standard 99 percent of the time, EPA
estimates that a source with average
emissions variability among runs of a
test condition would need to design and
operate the kiln to meet a level of 36 ug/
dscm.
EPA estimates that approximately 70
percent of LWAKs can meet this floor
level. The national annualized
compliance cost of the remaining
LWAKs to reduce mercury emissions to
the floor level is estimated to be $1.6
million for the entire hazardous waste-
burning LWAK industry, and would
reduce mercury emissions by 540
pounds per year or by 86 percent from
current baseline emissions.
EPA notes that it considered whether
all LWAKs would be likely to be able to
meet the floor level of 72 ug/dscm using
control of hazardous waste feed for Hg
at an MTEC of 17 ug/dscm, given that
Hg emissions also result from Hg in the
raw material feed. EPA has determined
that all LWAKs should be able to meet
the floor level using the floor control
without substituting raw material.
b. Beyond-The-Floor Considerations.
The Agency has considered beyond-the-
floor (BTF) control for Hg using carbon
injection (CI) in combustion gas at
temperatures below 400°F, coupled with
the MACT floor level control of Hg in
the hazardous waste feed. As discussed
for CKs, EPA believes that CI can
control Hg emissions at or above 90
percent removal efficiency.
To identify a BTF level, EPA
considered two approaches that would
result in virtually the same BTF
standard—6 ug/dscm. Under one
approach, EPA would apply a 90
percent removal efficiency for CI to the
floor design level of 36 ug/dscm to
identify a BTF standard of 6 ug/dscm,
which includes a statistically-derived
variability factor.
Under a second approach, EPA could
account for emissions variability by
using a conservative CI removal
efficiency of 80 percent to identify a
BTF emission standard of 7.2 ug/dscm
(based on a design floor level of 36 ug/
dscm). Under this approach, a
statistically-derived variability factor
would not be added.
EPA invites comment on which
approach would be more appropriate for
identifying a BTF level. EPA, however,
is not proposing a BTF standard.
divided by the gas flow rate. It is used to normalize
feedrates of Hg (and other metals and chlorine)
across sources with different waste (or fuel) burning
capacities.
In conjunction with earlier
evaluations, the Agency has evaluated
the cost and emissions reductions
associated with an emission standard of
8 ug/dscm. Although the BTF levels
presented above are somewhat different,
EPA does not believe that the difference
is large enough to significantly affect the
information presented below. /••
One of 11 LWAKs in the database
would be able to meet a BTF level of 8
ug/dscm currently. The national
annualized compliance cost for the
remaining LWAKs to meet the BTF level
is estimated to be $4.4 million for the
entire hazardous waste-burning LWAK
industry. The BTF level would provide
an incremental reduction of 60 pounds
per year (72 percent) in Hg emissions
nationally beyond that achieved with
floor controls.
EPA has considered the costs in
relation to emissions reductions and the
special bioaccumulation potential that
Hg poses and has decided that the floor
level of 72 ug/dscm best balances those
factors. Mercury is one of the more toxic
metals known due to its
bioaccumulation potential and the
neurological health effects at low
concentrations. For further discussion
see the mercury benefits discussion in
Section VII of today's preamble. EPA
invites comment, however, on whether
there are cost-effectiveness or other
factors that would lead the Agency to
promulgate a final rule based on the
BTF level.
4. Semivolatile Metals
a. MACT Floor. Emissions of SVM
from LWAKs are currently controlled
under the BIF rule. LWAKs use a
combination of hazardous waste
feedrate control and PM control to
comply with those standards.
Accordingly, MACT floor control is
based on hazardous waste feedrate
control and PM control.
The LWAK semivolatile metals (SVM)
(consisting of cadmium and lead) data
reflect results from 13 test conditions
collected from 6 facilities, with a total
of 10 units being tested. Average
emissions of the SVM group ranged
from 1 ug/dscm to 1670 ug/dscm.
Control of semivolatile emissions is
associated with PM control (see
discussion of SVM control for existing
cement kilns). All LWAKs are equipped
with a fabric filter as the air pollution
control device, although one facility is
equipped with a spray dryer, venturi
scrubber and wet scrubber in addition to
the fabric filter.
The Agency analyzed all available
lead and cadmium emissions data and
determined that sources with emission
levels at or below the level emitted by
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17405
the median of the best 12 percent of
sources employed either: (1) A fabric
filter with an air-to-cloth ratio of 1.5
acfm/flz or less with a hazardous waste
MTEC less than 270,000 ug/dscm; or (2)
a fabric filter and venturi scrubber with
on air-to-cloth ratio of 4.2 acfm/ft 2 or
less with a hazardous xvaste MTEC less
than 54,000 ug/dscm. Analysis of
emissions data from all LWAKs using
these MACT technologies resulted in a
floor level of 12 ug/dscm.
EPA notes that raw materials and
fossil fuels also contribute to LWAK
SVM feedrates and emissions. Given
that all sources must be able to meet the
floor level using the floor control, EPA
investigated whether all LWAKs could
meet the floor level employing the
MACT floor technologies without being
forced to substitute raw material. EPA
preliminary evaluation determined that
25 percent of sources in the SVM
emissions database had raw material
containing Cd and Pb in greater
concentrations than sources in the
expanded MACT pool; thus, these
sources may not be able to achieve the
floor with MACT alone.116 However, the
Agency believes that the data on which
this preliminary finding is based may
not reflect the normal, day-to-day Pb
and Cd levels in raw material feed.
As noted in the earlier section on
cement kilns, one approach to address
this issue (of sources with higher levels
of SVM metals in their raw materials
than sources in the expanded MACT
pool and that, therefore, cannot meet the
floor level using floor control) is to: (1)
Identify the source with the highest
normalized (by MTEC) feedrate of
metals in raw material; (2) assume the
source is also feeding hazardous waste
with the floor control MTEC level of the
metals; and (3) project SVM emissions
from the source based on combined raw
material and hazardous waste MTECs
using a representative system removal
efficiency (SRE) from the expanded
MACT pool considering an appropriate
variability factor (e.g., variability of
emissions among runs within a test
condition in the expanded MACT pool).
The Agency has not yet conducted this
type of analysis, but intends to do so in
the near future. EPA also believes that
data reflecting normal, day-to-day levels
of Pb and Cd in raw materials would be
important for this type of analysis, and
specifically invites commenters to
submit such data as well as their views
on the approach suggested above.
i m USEPA, "Draft Technical Support Document
forHWC MACT Standards, Volume IH: Selection of
Proposed MACT Standards and Technologies",
February 1096.
EPA estimates that 38 percent of
LWAKs are currently meeting the floor
level. The national annualized
compliance cost of the remaining
LWAKs to reduce SVM emissions to the
floor level is estimated to be $2.1
million for the entire LWAK industry,
and would reduce lead and cadmium
emissions nationally by 0.66 tons per
year, or by 97 percent from current
baseline emissions.
The Agency is proposing an
alternative compliance option for SVMs.
Since the Agency anticipates the
likelihood of development of a multi-
metals continuous emissions monitor
(CEM) in the near future, the Agency is
proposing establishing a higher standard
for sources using a properly designed
and operated multi-metals CEM. This
alternative compliance option would be
based on the minimum detection limit
of the device, which is estimated to be
60 ug/dscm for SVMs combined.
b. Beyond-The-Floor Considerations.
The Agency considered whether to
propose a more stringent level than the
floor of 12 ug/dscm. EPA has
determined that a BTF standard would
not be appropriate. Since control of
semivolatile emissions is associated
with PM control, a more stringent SVM
BTF level would require LWAKs to
upgrade to more expensive fiberglass
bags (e.g., bags backed with teflon
membranes) or the addition of newly
installed FFs with improved
performance media. Although the
engineering costs to comply with a BTF
SVM level are moderate, the resulting
incremental reduction in SVM
emissions from the floor level is
minimal because the floor level already
provides substantial control by reducing
baseline emissions by 97 percent. Thus,
the Agency believes a SVM BTF
standard is not appropriate and is
proposing a SVM MACT standard of 12
ug/dscm for existing LWAKs.
5. Low-Volatility Metals
a. MACT Floor. Emissions of LVM
from LWAKs are also currently
controlled under the BIF rule. LWAKs
use a combination of hazardous waste
feedrate control and PM control to
comply with those standards.
Accordingly, MACT floor control is
based on hazardous waste feedrate
control and PM control.
The low volatility metals (LVM)
(consisting of arsenic, antimony,
beryllium, and chromium) data reflect
results from 13 test conditions collected
from 6 facilities, with a total of 10 units
being tested. Average emissions of the
LVM group ranged from 10 ug/dscm to
289 ug/dscm. Due to the relatively low
volatility of these metals, performance
of the APCD is the most important factor
in controlling LVM emissions.
The Agency analyzed all available
LVM emissions data and determined
that sources with emission levels at or
below the level emitted by the median
of the best 12 percent of sources used a
fabric filter with an air-to-cloth ratio of
1.8 acfm/ft2 or less with a hazardous
waste MTEC less than 46,000 ug/dscm.
Analysis of available emissions data for
all LWAKs employing these controls
resulted in a floor emission level of 340
Hg/dscm.
EPA notes that raw materials and
fossil fuels also contribute to LWAK
LVM feedrates and emissions. Given
that all sources must be able to meet the
floor level using the floor control, EPA
investigated whether all LWAKs could
meet the floor level employing the
MACT floor technologies without being
forced to substitute raw material. EPA's
preliminary evaluation determined that
one of the sources in the LVM emissions
database had raw material containing
LVM in greater concentrations than
sources in the expanded MACT pool;
thus, this sources may not be able to
achieve the floor with MACT alone.117
EPA requests comments on addressing
this issue.
One approach to address this issue (of
sources with higher levels of LVM
metals in their raw materials than
sources in the expanded MACT pool
and that, therefore, cannot meet the
floor level using floor control) is to: (1)
Identify the source with the highest
normalized (by MTEC) feedrate of
metals in raw material; (2) assume the
source is also feeding hazardous waste
with the floor control MTEC level of the
metals; and (3) project LVM emissions
from the source based on combined raw
material and hazardous waste MTECs
using a representative system removal
efficiency (SRE) from the expanded
MACT pool considering an appropriate
variability factor (e.g., variability of
emissions among runs within a test
condition in the expanded MACT pool).
The Agency has not yet conducted this
type of analysis but intends to do so in
the near future. EPA also believes that
data reflecting normal, day-to-day levels
of LVM in raw materials would be
important for this type of analysis and
specifically invites commenters to
submit such data as well as their views
on the approach suggested above.
EPA estimates that 92 percent of
LWAKs are currently meeting the floor
level. The national annualized cost of
117 USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume III: Selection of
Proposed MACT Standards and Technologies",
February 1996.
-------
17406
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules'
the remaining LWAKs to reduce LVM
emissions to the floor level is estimated
to be $380,000 for the entire hazardous
waste-burning LWAK industry;'this
would reduce LVM emissions nationally
by 0.011 ton per year or by 5 percent
from current baseline emissions.
b. Beyond-The-Floor Considerations.
The Agency considered whether to
propose a more stringent level than the
floor of 340 ug/dscm. Since control of
low-volatile emissions is associated
with PM control, a more stringent LVM
BTF level would require LWAKs to
upgrade to more expensive fiberglass
bags (e.g., bags backed with teflon
membranes) or the addition of newly
installed FFs with improved
performance media. Although the
engineering costs to comply with a BTF
LVM level are moderate, the resulting
reduction in LVM emissions is minimal
since LWAK LVM national emissions
are estimated to be 0.2 tons per year for
the entire industry at the floor level.
Thus, the Agency believes a LVM BTF
standard is not appropriate and is
proposing a LVM MACT standard of 340
ug/dscm for existing LWAKs.
6. Hydrochloric Acid and Chlorine
a. MACT Floor. HC1 and C12 emissions
from LWAKs are currently regulated by
the BIF rule. Only one LWAK facility
currently utilizes a venturi scrubber,
which is a dedicated control device,
designed specifically to remove HCl/Ck
(referred to as total chlorine where
combined HC1 and Ck levels are
expressed as HC1 equivalents) from the
flue gas.
The total chlorine emission database
reflects results from 13 test conditions
collected from 6 facilities, with a total
of 10 units being tested. Average total
chlorine emissions range from 13 ppmv
to 2080 ppmv. The Agency analyzed all
available total chlorine emissions data
and determined that sources with
emission levels at or below the level
emitted by the median of the best 12
percent of sources used either: (1)
Hazardous waste feedrate control of
total chlorine with a MTEC less than 1.5
g/dscm; or (2) venturi scrubber with
hazardous waste MTEC less than 14 g/
dscm. The analysis of all available
emissions data for LWAKs using these
technologies resulted in a floor
emissions level of 2100 ppmv, which
the Agency has identified as the MACT
floor level. To meet this standard 99
percent of the time, a source with
average within test condition emission
variability would need to be designed
and operated to achieve an emission
level of 1400 ppmv.
EPA notes mat raw materials and
fossil fuels also contribute to LWAK
chlorine feedrates and emissions. Given
that all sources must be able to meet the
floor level using the floor control, EPA
investigated whether all LWAKs could
meet the floor level employing the
MACT floor technologies without being
forced to substitute raw material. EPA
determined that all LWAKs in the total ••
chlorine emissions database would be
able to meet the floor level using floor
control11B without switching raw
material.
EPA estimates that 85 percent of
LWAKs are currently meeting the floor
level. The national annualized
compliance cost of the remaining
LWAKs to reduce total chlorine
emissions to the floor level is estimated
to be $890,000 for the entire hazardous
waste-burning LWAK industry; this
would reduce total chlorine emissions
nationally by 190 tons per year or 6
percent from current baseline emissions.
b. Beyond-The-Floor Considerations.
The Agency has considered BTF
controls for improved total chlorine
control using a dry scrubber or spray
tower scrubber. A dry scrubber should
achieve a total chlorine removal
efficiency of 90 percent, and a spray
tower scrubber should achieve a
removal efficiency of 99 percent.
Applying the 90 percent removal factor
(the more conservative of the two
removal efficiencies)119 to the highest
test condition in the database resulted
in a BTF standard of 450 ppmv. To meet
this standard 99 percent of the time,
EPA estimates that a source with
average emissions variability (among
runs within a test condition) would
need to meet a design level of 210
ppmv.
EPA believes that dry scrubbers or
spray tower scrubbers are appropriate
controls and is proposing a 450 ppmv
total chlorine emission standard based
on these controls. EPA estimates that 38
percent of LWAKs are currently meeting
this BTF level. The national annualized
compliance cost for the remaining
LWAKs to meet this BTF level rather
than comply with the floor controls is
estimated to be $5.0 million for the
entire hazardous waste-burning LWAK
industry. This BTF level would provide
an incremental reduction of 2200 tons
per year (80 percent) in total chlorine
emissions nationally beyond that
achieved with the floor controls.
"8USEPA, "Draft Technical Support Document
for HWC MACT Standards, Volume ID: Selection of
Proposed MACT Standards and Technologies",
February 1996.
?19The Agency believes that many, but not all,
LWAKs could use a dry scrubber without adversely
affecting the quality of the LWAK dust (which is
primarily raw material) for incorporation into
products or recycling back into the kiln. See
discussion in the text below.
The Agency believes that both wet
and dry scrubbing control techniques
are applicable to LWAKs for chlorine
control. Dry scrubbing is being used at
some hazardous waste-burning LWAKs.
Control efficiency and outlet chlorine
emissions levels are unclear due to
conflicting trial bum results, however.
• One potential problem with the
application of dry scrubbing to LWAKs
is contamination of the captured LWAK
dust with dry sorbent. This may affect
whether captured dust can be recycled
back into the kiln or incorporated into
the final light weight aggregate product.
The addition of dry scrubbing could
force some kilns either to add a
separate, additional FF dedicated to
capturing the dry sorbent or dispose of
the mixed sorbent and LWAK dust. The
Agency invites comment on the
effectiveness (and implications on dust
management) of dry scrubbing for
control of chlorine in hazardous waste-
burning LWAKs.
The Agency also considered an
additional BTF level of 25 ppmv for
LWAKs based on wet scrubbing alone.
A further reduction from the proposed
BTF design level of 210 ppmv (based on
dry scrubbing or spray tower scrubbing)
to 25 ppmv would require all thirteen
LWAK sources to either install new
control equipment, or modify existing
control equipment. The incremental
cost of this enhanced control would be
moderate to high for each of the
individual LWAK sources. Although the
engineering cost for each facility is
moderate to high, the overall cost for
LWAKs as a group is high since
upgrades are required by every facility.
The Agency believes that the resulting
moderate decrease in total chlorine
emissions may not justify this relatively
high engineering cost.
Based on cost-effectiveness
considerations, EPA has determined
that proposing a BTF standard of 450
ppmv is warranted. As discussed
elsewhere in today's preamble, EPA's
risk analysis developed for purposes of
RCRA shows that the emissions of total
chlorine from hazardous waste-burning
LWAKs could pose significant risks by
direct inhalation, and these risks would
be reduced by BTF controls.120 Thus,
the BTF controls would make separate
RCRA standards unnecessary.
Additionally, the Agency requests
comments on an alternative option to
identify the BTF level. Under this
120 EPA notes that under the BIF regulations,
LWAKs are currently subject to site-specific, risk-
based emissions standards for HC1/C12. EPA is
uncertain why our risk assessment to consider
RCRA concerns under today's proposed rule shows
that baseline emissions for some LWAKs can pose
significant risk.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17407
option the 90 percent reduction in
emissions provided by a dry scrubber or
spray tower scrubber would be applied
to the floor level resulting from
hazardous xvaste feedrate control of total
chlorine—2100 ppmv. Thus, at 90
percent control efficiency, the BTF
emission standard would be 210 ppmv.
To comply with this standard 99
percent of the time, a source with
average within test condition emissions
variability would need to be designed
and operated to meet an emission level
of approximately 140 ppmv. EPA invites
comment on whether this option is
more appropriate to establish the BTF
level than applying the BTF percent
reduction to the test condition in the
database with the highest emissions.
As discussed above, EPA believes that
a dry scrubber or spray tower scrubber
(in conjunction with the levels achieved
using MACT floor controls) are
appropriate alternative controls. EPA
estimates that 38 percent of LWAKs are
currently meeting this alternative BTF
level of 210 ppmv. EPA estimates that
this BTF level would provide a further
incremental reduction in total chlorine
emissions nationally beyond that
achieved with the proposed BTF
standard of 450 ppmv. EPA invites
comment on this alternative approach to
identify the BTF level.
7. Carbon Monoxide and Hydrocarbons
The Agency is proposing to use
carbon monoxide (CO) and
hydrocarbons (HC) as surrogates for
non-D/F organic HAPs.121
a. MACT Floor.
i. Carbon Monoxide. The BIF rule
currently limits CO emissions from
LWAKs to 100 ppmv on an hourly
rolling average (HRA). See § 266.104(b).
However, the BIF rule provides an
alternative standard that allows higher
CO levels if HC levels are less than 20
ppmv.
LWAKs generally have low CO levels
(i.e., less than 100 ppmv HRA) achieved
by operating under good combustion
practices. Good combustion practices
include techniques such as thorough
fuel, air, and waste mixing; adequate
excess oxygen; maintenance of adequate
combustion temperature; and blending
of waste fuels to minimize combustion
perturbations. Accordingly, operating
under good combustion practices is
identified as the floor control.
Given that 10 of 12 LWAKs for which
EPA has CO emissions data have
maximum hourly rolling averages for
the test condition of less than 100
ppmv, EPA believes it is reasonable and
appropriate to identify the floor level as
the BIF limit of 100 ppmv. Two LWAKs
have CO levels exceeding the 100 ppmv
level, however, and these higher levels
(i.e., 190 ppmv and 1900 ppmv) are
allowed under the BIF rule. EPA is not
sure whether these elevated CO levels
were caused by operating under poor
combustion conditions, or by trace
levels of organics desorbing from the
raw materials.
If the CO were caused by organics
desorbing from raw material, EPA
would consider this situation analogous
to CKs that do not have a by-pass duct
(and thus stack emissions are affected
by organics desorbed from raw
material). Accordingly, such LWAKs
would be exempt from the CO limit (and
would be subject to a HC limit of 20
ppmv). (In this situation, floor control
(i.e., good combustion practices) could
not be used to meet the floor level.) EPA
invites comment on how to distinguish
between LWAKs that have elevated CO
levels because of poor combustion (and
that should be subject to the 100 ppmv
floor level) and LWAKs that have
elevated CO levels because of
desorption of organics from raw
material (and that should be exempt
from the 100 ppmv floor level). If an
effective approach to distinguish
between these situations is developed,
the final rule could distinguish among
LWAKs based on those high levels of
organics in raw material versus those
with low levels.
EPA estimates that over 80 percent of
LWAKs are currently meeting the
proposed standard. The national
annualized compliance cost of the
remaining LWAKs to reduce carbon
monoxide emissions to the floor
level122 is estimated to be $1.4 million
for the entire LWAK industry; this
would reduce carbon monoxide
emissions nationally by 600 tons per
year, or 81 percent from current baseline
emissions.
ii. Hydrocarbons. As discussed above,
the BIF rule limits HC levels to 20 ppmv
HRA when CO exceeds 100 ppmv HRA.
As with CO, floor control is operating
under good combustion practices. EPA
believes it is appropriate to establish the
floor level at the lower of the BIF
emission limit or the levels that sources
actually achieved. An analysis of the
available HC data determined that
sources with emission levels at or below
the level emitted by the median of the
best 12 percent of sources used good
combustion practices as the control
technology. The analysis of all available
emissions data for LWAKs believed to
be using good combustion practices
resulted in a floor emissions level of 14
ppmv.123
EPA estimates that 86 percent of
LWAKs are currently meeting the floor
HC level. The national annualized
compliance cost of the remaining
LWAKs to reduce hydrocarbon
emissions to the floor level is estimated
to be $760,000 for the entire LWAK
industry; this would reduce
hydrocarbon emissions nationally by 14
tons per year, or 31 percent from current
baseline emissions.
b. Beyond-The-Floor Considerations.
EPA considered BTF levels for CO of 50
ppmv and for HC of 6 ppmv. Control of
organic HAP emissions would require
the use of a combustion gas afterburner.
Addition of an afterburner to a LWAK
would be expensive due to the
requirement of a large amount of
auxiliary fuel to reheat the kiln exit flue
gas to temperatures required for
organics burnout. Preliminary estimates
suggest that going beyond-the-floor for
CO and HC would more than double the
national costs of complying with the
proposed rule. EPA believes that a BTF
standard is not appropriate.
EPA estimates that 29 percent of,
LWAKs are currently meeting the BTF
level of 6 ppmv for HC and that 461
percent of LWAKs are currently meeting
the BTF levels of 50 ppmv for CO. The
Agency has determined that selecting
these BTF levels is not appropriate.
Therefore, the Agency is proposing a
MACT standard for hydrocarbons of 14
ppmv HRA and for carbon monoxide of
100 ppmv HRA.
8. MACT Floor Cost Impacts
The total national annualized
compliance costs for existing LWAKs to
meet all the MACT floor levels are
estimated to be $3 million with the cost
per kiln averaging $390,000. These total
compliance costs equate to $39 per ton
of hazardous waste burned. EPA
estimates that one LWAK facility may
cease burning hazardous waste due to
the compliance costs associated at the
floor.
1Z> This is in addition to controlling PM as a
surrogate for (condensed) senrivolatile HAPs.
>22 EPA assumed that the LWAK with CO levels
of 1900 ppmv would need to install an afterburner
to meet the floor level. EPA acknowledges that this
is inappropriate because all sources must be able to
meet the floor level using floor control—good
combustion practices. As discussed in the text, EPA
invites comment on how to identify appropriate
MACT floor levels for sources that may have
elevated CO levels due to desorption of organics
from raw material.
123 EPA notes that oneofseven LWAKs in the HC
database had substantially higher test condition
maximum HC levels (i.e., 13 ppmv HRA) than the
other sources (i.e., 6 to 8 ppmv HRA). As discussed
in the text above for CO, it is not clear whether the
elevated HC levels were caused by operating under
poor combustion conditions or desorption of
organics from raw material. EPA invites comment
on how to address this situation.
-------
37408 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
The Agency is proposing to go
beyond-the-floor for three pollutants for
existing LWAKs: dioxin/furans,
mercury, and total chlorine. The total
national annualized compliance costs to
meet the .dioxin/furan, mercury and
total chlorine BTF standards in addition
to the MACT floor standards for the
remaining HAPs are estimated to be $4
million with the cost per kiln averaging
$670,000. These total compliance costs
increase the cost per ton of hazardous
waste burned to $56. EPA estimated that
one LWAK facility may cease burning
hazardous waste due to the compliance
costs associated with this suite of floor
and BTF standards.
B. MACT for New Sources
This section summarizes EPA's
rationale for establishing MACT for new
LWAKs for each HAP, HAP surrogate, or
HAP group. Table V.5.B.1 summarizes
the proposed MACT standards for new
LWAKs, which were determined using
the analytical process described in Part
Three, Section VII and in "Draft
Technical Support Document for HWC
MACT Standards, Volume III: Selection
of MACT Standards and Technologies".
TABLE IV.5.B.1.—PROPOSED
EMISSION LEVELS FOR NEW LWAKs
HAP or HAP Surro-
gate
Dioxin/furans
Particulate Matter
Mercury
SVM [Cd, Pb]
LVM [As, Be, Cr, Sb]
HCI + CI2
CO
HC
Proposed Standards1
0.20 ng/dscm TEQ.
0.030 gr/dscf (69 mg/
dscm).
72 ng/dscm.
5.2 ng/dscm2.
55 ng/dscm3.
62 ppmv.
100 ppmv.
14 ppmv.
1 All emission levels are corrected to 7 per-
cent O2.
2 An alternative standard of 60 ng/dscm
would apply if the source elects to document
compliance using a multi-metals GEM.
3 An alternative standard of 80 ng/dscm
would apply if the source elects to document
compliance using a multi-metals GEM.
1. MACT New for Dioxin/Furan
a. MACT NEW Floor. EPA used the
Agency's data on the performance of D/
F control technology to identify MACT
floor controls and the floor level for new
facilities. The MACT floor level for D/
F emissions from LWAKs is 0.20 ng/
dscm (TEQ) or (temperature at the PM
control device not to exceed) 418 °F.
b. Beyond-The-Floor Considerations.
The BTF considerations for new LWAKs
were the same as for CKs. Therefore,
EPA is proposing a BTF standard for
new LWAKs of 0.20 ng/dscm (TEQ) for
the same reasons applicable to CKs.
2. MACT New for Particulate Matter
a. MACT New Floor. EPA's analysis of
available PM data shows that the single
best APCD for controlling particulate
emissions is a fabric filter with an air-
to-cloth ratio less than 1.5 acfm/ft2
which represents MACT technology for
new sources. An evaluation of all
sources employing this technology
shows that this technology can
consistently achieve a PM emission of
0.054 gr/dscf.
b. Beyond-The-Floor Considerations.
For the same reasons as discussed for
existing LWAKs, the Agency is
proposing a lower BTF standard for new
LWAKs. Therefore, the Agency is
proposing the MACT standard of 69 mg/
dscm (0.03 gr/dscf) for new LWAKs.
As discussed above for existing
LWAKs, EPA specifically invites
comment on whether the final rule
should establish an alternative BTF
standard for PM of 35.mg/dscm (or 0.15
Ib/ton of raw material (dry basis) feed
into the kiln).
3. MACT New for Mercury
a. MACT New Floor. The MACT hew
floor analysis is the same as existing
sources because the expanded pools for
each, based on the single best
performing source, are identical. As
discussed earlier, LWAKs control their
mercury input (and therefore much of
their emissions) through the control of
the mercury content in the hazardous
waste. The Agency is defining the
MACT floor technology as feedrate
control with a hazardous waste MTEC
less than 17 ng/dscm based on
performance of the single best
performing source. Analysis of all
existing LWAK sources using this
hazardous feedrate control resulted in a
MACT floor level of 72 ng/dscm.
b. Beyond-the-Floor Consideration.
The Agency is considering the same two
BTF options for new LWAKs as
discussed for existing sources—Option
1 is 6 ng/dscm, and Option 2 is 7.2 u,g/
dscm. The Option 1 mercury BTF level
of 6 ng/dscm is achievable based on the
use of some degree of hazardous waste
feedrate control and/or add-on mercury
control with injection of activated
carbon, assuming a 90 percent
reduction. The Option 2 level of 7.2 ug/
dscm represents an achievable level
based on both achievement of floor
levels and use of carbon injection,
assuming conservative 80 percent
reduction.
Therefore, EPA is proposing a
mercury MACT standard of 72 ng/dscm
for existing LWAKs and requesting
comments on possible BTF standard of
6 ng/dscm and 7.2 ng/dscm.
4. MACT New for Semivolatile Metals
a. MACT New Floor. EPA
characterized the single best performing
source with the lowest SVM emissions
and determined that the best performing
source used a fabric filter with an air-
to-cloth ratio of 1.5 acfm/ft2 or less for
a kiln system with a hazardous waste
(HW) MTEC of 270,000 ng/dscm or less.
Analysis of all sources using this
technology or better (i.e., expanded
MACT pool of facilities) resulted in a
floor level of 5.2 ng/dscm for new
LWAKs.
The Agency recognizes that 5.2 ng/
dscm is a low floor level and is
concerned about potential problems in
its approach to setting the MACT floor
level. The expanded MACT pool
included only one other test condition
besides the single best source, and EPA
is concerned that this low data set
resulted in a low floor level. In addition,
EPA is concerned that the single best
performing source may have low SVM
feedrates in the raw material, which
could result in a floor level that is
unachievable. EPA invites comment on
how to address these potential issues.
The Agency is proposing an
alternative compliance option for SVMs.
Since the Agency anticipates the
likelihood of development of a multi-
metals continuous emissions monitor
(GEM) in the near future, the Agency is
proposing establishing a higher standard
for sources using a properly designed
and operated multi-metals CEM. This
alternative compliance option would be
based on the minimum detection limit
of the device which is estimated to be
60 ug/dscm for SVMs combined.
b. Beyond-the-Floor Considerations.
EPA has determined that proposing a
BTF standard is not warranted for the
same reasons that a more stringent level
was not proposed for existing sources.
Therefore, the Agency is proposing a
semivolatile metals MACT standard of
5.2 ng/dscm for new LWAKs.
5. MACT New for Low-Volatile Metals
a. MACT New Floor. EPA
characterized the best particulate
control device and identified the floor
technology as a fabric filter with an air-
to-cloth ratio of 1.3 acfm/ft2 or less with
a hazardous waste (HW) MTEC less than
37,000 ng/dscm. Analysis of all existing
LWAK sources employing either of
these technologies resulted in a floor
emissions level of 55 ng/dscm for new
LWAKs.
The Agency is proposing an
alternative compliance option for LVMs.
Since the Agency anticipates the
likelihood of development of a multi-
metals continuous emissions monitor
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17409
(GEM) in the near future, the Agency is
proposing establishing a higher standard
for new sources using a properly
designed and operated multi-metals
CEM. This alternative compliance
option would be based on the minimum
detection limit of the device which is
estimated to be 80 ug/dscm for these
LVM metals combined.
b. Beyond-the-Floor Considerations.
EPA has determined that proposing a
BTF standard is not warranted for the
same reasons that a more stringent level
was not proposed for existing sources.
Therefore, the Agency is proposing a
low-volatile metals MACT standard of
55 ug/dscm for new LVVAKs.
6. MACT New for Hydrochloric Acid
and Chlorine
a. MACT New Floor. EPA
characterized the single best performing
source with the lowest HCl/Ck (total
chlorine) emissions and determined that
the best performing source used a
venturi scrubber with a hazardous waste
(HW) MTEC of 14 g/dscm or less.
Analysis of all sources using this
technology or better (i.e., expanded
MACT pool of facilities) resulted in a
floor level of 62 ppmv for new LWAKs.
b. Beyond-the-Floor Considerations.
The MACT floor is characterized by a
technology that is able to achieve a 99
percent removal efficiency. A BTF level
is not warranted because the floor level
is based on a technology that is able to
achieve the highest removal efficiency
for HC1/C12. Therefore, the Agency is
proposing a HCl/Clz MACT standard of
62 ppmv for new LWAKs.
7. MACT New for Carbon Monoxide and
Hydrocarbons
a. MACT New Floor. The Agency
believes that control of non-dioxin
organic emissions can be achieved by
establishing emissions limits on
hydrocarbons and carbon monoxide. As
discussed earlier for existing LWAKs,
the Agency is proposing a MACT
standard of 14 ppmv for HC and of 100
ppmv for CO, based on floor levels
b. Beyond-the-Floor Considerations.
EPA considered control for organic HAP
emissions based on the use of a
combustion gas afterburner. Even
though EPA believes that BTF levels for
CO of 50 ppmv and for HC of 6 ppmv
are achievable with an afterburner,
using these values for a BTF standard is
not appropriate and is not warranted at
this time (see discussion for existing
LWAKs). Therefore, EPA is proposing a
MACT standard of 14 ppmv for HC and
of 100 ppmv for CO for new LWAKs.
8. MACT New Cost Impacts
A detailed discussion of the costs and
economic impacts for new LWAKs is
presented in Part Seven of today's
proposal and "Regulatory Impact
Assessment for Proposed Hazardous
Waste Combustion MACT Standards".
C. Evaluation of Protectiveness
In order to satisfy the Agency's
mandate under the Resource
Conservation and Recovery Act to
establish standards for facilities that
manage hazardous wastes and issue
permits that are protective of human
health and the environment, the Agency
conducted an analysis to assess the
extent to which potential risks from
current emissions would be reduced
through implementation of MACT
standards. The analysis conducted for
hazardous waste-burning LWAKs is
similar to the one described above for
hazardous waste incinerators and
cement kilns. The procedures used in
the Agency's risk analyses are discussed
in detail in the background document
for today's proposal.124 In evaluating the
MACT standards, the Agency used the
design value which is the value the
Agency expects a source would have to
design to in order to be assured of
meeting the standard on a daily basis
and hence is always a lower value than
the actual standard for all HAPs
controlled by a variable control
technology.12S
The risk results for hazardous waste-
burning lightweight aggregate kilns are
summarized in Table V.5.C.1 for cancer
effects and Table V.5.C.2 for non-cancer
effects for the populations of greatest
interest, namely subsistence farmers,
subsistence fishers, recreational anglers,
and home gardeners. The results are
expressed as a range representing the
variation in exposures across the
example facilities (and example
waterbodies for surface water pathways)
for the high-end and central tendency
exposure characterizations across the
exposure scenarios of concern. For
example, because dioxins
bioaccumulate in both meat and fish,
the subsistence farmer and subsistence
fisher scenarios are used to determine
the range.126
TABLE V.5.C.1.—INDIVIDUAL CANCER RISK ESTIMATES FOR LIGHTWEIGHT AGGREGATE KILNS 1
Dioxins
Semi-volatile met-
als2
Low volatile met-
als3
Existing Sources
Baseline
Floor
BTF
2E 9 to 4E 7
1E-8 to 2E-64
1E-8 to2E-65
1 E-8 to 5E 7
1 E-8 to 6 E-8
9E 10 to 4E 7
5E 7 to 1 E 5
New Sources
Floor
BTF
CEMOotion6
1 E-8 to 2E-64
1 E-8 to 2E-65 . .
6E-9 to 3E-8
6E-8 to 3E-7
7E-8 to 2E-6
2E-7 to 5E-6.
1 Lifetime excess cancer risk.
* Carcinogenic metal: cadmium.
'Carcinogenic metals: arsenic, beryllium, and chromium (VI).
'Based on 0.2 ng/dscm TEQ as both a central tendency and high-end estimate.
'Based on 0.20 ng/dsem TEQ.
•Based on SVM standard of 60 ug/dscm and LVM standard of 80 ug/dscm (applicable only if the source elects to document compliance using a multimetals CEM).)
124 "Risk Assessment Support to the Development
of Technical Standards for Emissions from
Combustion Units Burning Hazardous Wastes:
Background Information Document", February 20,
1996.
1MFor tho semi-volatile and low volatility metals
categories, tho Agency assumed the source could
emit up to the design value for each metal in the
category for the purpose of assessing protectiveness.
126 For the semi-volatile and low volatility metals
categories, the inhalation MEI scenarios are also
-------
27420
Federal Register / Vol. 61, No. 77 /Friday, April 19, 1996 / Proposed Rules
TABLE V.5.C.2—INDIVIDUAL NON-CANCER RISK ESTIMATES FOR LIGHTWEIGHT AGGREGATE KILNS 1
Semi-volatile
metals 2
Low volatile
metals3
Hydrogen
chloride
Chlorine
Existing Sources
Baseline
Floor
BTF
<0 001 to 0 006
<0 001
<0 001 to 0 007
<0 001 to 0 08
0 1 to 4
0 8 to 1 4
01 to 024
0 03 to 0 3
4 to 75
0 6 to 1 5
New Sources
Floor
BTF
GEM Option 6
<0.001
<0.001 to 0 001
<0 001 to 0 01
<0 001 to 0 03
0 02 to 0 044
0 01 to 0 024
01 to 0 25
0 07 to 0 1 5
1 Hazard quotient.
2 Cadmium and lead.
3 Antimony, arsenic, beryllium, and chromium.
* HCI + Cb assuming 100 percent HCI.
5 HCI + Clz assuming 10 percent CI2.
6 Based on SVM standard of 60 ng/dscm and LVM standard of 80 |ig/dscm (applicable only if the source elects to document compliance using a multi-metals GEM).
The risk analysis indicates that for the
semi-volatile and low volatility metals
categories, the MACT standards for
lightweight aggregate kilns are
protective at the floor for both existing
and new sources. The analysis indicates
that the GEM compliance option for new
sources is also protective. The analysis
also indicates that for dioxins, both the
floor levels and the proposed beyond
the floor standards are protective. The
analysis also indicates that for hydrogen
chloride and chlorine (C12), the
proposed beyond-the-floor standards for
existing sources, rather than the floor
levels, are protective.
VI. Achievability of the Floor Levels
As discussed in sections III, IV, and
V above, the MACT floor levels were
selected for each source category by
identifying the best performing sources
for each individual HAP or HAP
surrogate. This is the approach typically
used by the Agency in establishing
MACT standards.
Nonetheless, the Agency recognizes
that this approach raises the question of
whether the selected floor levels will be
achievable simultaneously.
An alternative approach that would
ensure simultaneous achievability of the
floor levels would be to identify the best
performing sources for a particular HAP
or HAP surrogate (e.g., D/F or PM) and
to consider emissions only from those
sources127 to establish floor levels for
the other HAPs or HAP surrogates. EPA
To address concerns relating to the
simultaneous achievability of the
proposed standards, which are a
combination of floor and BTF emissions
levels, the Agency investigated whether
sources could achieve the proposed
standards without making any upgrades
to existing equipment. It is important to
note that, under the current approach
used by the agency in establishing
MACT standards (i.e. the HAP by HAP
approach—utilizing the highest emitting
source in the expanded MACT pool),
approximately 5 to 8 percent of the
facilities currently operating will meet
all of the proposed standards.
Furthermore, subject to the data caveats
noted for certain HAPs and source
categories (which the Agency believes
can be resolved properly), it is the
opinion of the Agency that 100 percent
of the facilities who use MACT floor
and beyond-the-floor technologies can
meet all of the proposed standards
simultaneously.
Specific information and data
pertaining to the analysis of
simultaneous achievability can be found
in "Regulatory Impact Assessment for
Proposed Hazardous Waste Combustion
MACT Standards".
VII. Comparison of the Proposed
Emission Standards With Emission
Standards for Other Combustion
Devices
Although not explicitly part of the
MACT standard setting process, EPA
believes, for perspective, it is
appropriate to compare the proposed
emissions standards to those of other
waste-burning devices and similar
devices, (hi some cases, such a
comparison may show that a particular
technology or level of performance is
demonstrated as well.) The standards
used for comparison have either been
proposed by EPA or are guidelines
promulgated by the European Union
(EU). The standards for these various
type of devices will be different for
reasons including: (1) Different statutory
authorities and requirements; (2)
different levels of emission control for
existing sources; and (3) different
potential to emit high levels of specific
HAPs. Nonetheless, EPA believes a
comparison of standards is instructive.
Tables VII.l and VH.2 contain the
standards for municipal waste
combustors (MWCs), medical waste
incinerators (MWIs), EU hazardous
waste combustors, and the standards
proposed here for existing and new
facilities, respectively.
127 Another option would be to consider
emissions from other sources that employ
equivalent or better control for the other HAPs or
HAP surrogates, has not used this approach because
it would result in establishing unreasonably high
floor levels for most HAPs or HAP surrogates that
arbitrarily reflect the control devices (and emission
levels) that happen to be used by sources that are
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17411
TABLE VII.1.—COMPARISON OF STANDARDS FOR EXISTING SOURCES
dscm TEQ and/
or Total
congeners.
PM mo/dscm
Ho iio/dscm
CO ppmv
as HCI equiva-
lents (5).
Large MWCs
30 Total (or 15 if
testing less fre-
quent).
27
80 or 85% Reduct
Cd- 40
Pb- 49
50 to 250 4 to 24
hr avg.
None
31 or 95% Reduct
Proposed MWls
1 9 TEQ or 80
Total.
30
470 or 85%
Reduct..
Cd- 50
Pb: 100
none
50 12-hr avg
None
42 or 97% Reduct
EU HWCs (1)
019 TEQ
13 24-hr avg
1 3-39 30-min avg
(2).
130
Cd" 65
Tl: 65
Pb- 130 (3)
1 1 70 (3)
52, 24 hr avg
1 04, 30 min avg
(4).
156, 10 min avg
(4).
8, 24 hr avg
8-16, 30 min avg
(2).
8 24-hr avg
8-48, 30 min avg
(2).
Proposed HW in-
cinerators
50 10-
270
210
100 1 hr avg. .......
12 1 hr avg
280
Proposed HW ce-
ment kilns
0.20 TEQ.
69 2-hr avg
hr avg
57
130
Wet and Long,
Dry Kilns None.
Kilns with By-pass
1 00 in by-pass
duct (or HC
cannot exceed
6.7) 1 hr avg.
Wet and Long,
Dry Kilns 20 in
main stack 1 hr
avg.
Kilns with By-^pass
6.7 in by-pass
(or CO cannot
exceed 100) 1
hr avg.
630
Proposed HW
LWAKs
72 10-hr avg.
12.
340.
1 00 1 hr avg.
14 1 hr avg.
450.
Notes:' The EU HWC guidelines have been corrected from the European basis of 11% O2 and 0°C to the US basis of 7% O2 and 20°C. Both
are expressed on dry emissions.
2The EU HWC PM, HC, and HCI guidelines are based either 97 % compliance with the lower number or 100% compliance with the higher
number on a 30-minute average over a year.
3The EU LVM guideline is 1300 ng/dscm and includes Sb, As, Pb, Cr, Co, Cu, Mn, Ni, V, Sn. If all metals are emitted equally, their contribu-
tion is 130 jg/dscm. Pb, a SVM, was subtracted from this group, resulting in the 1170 n.g/dscm level.
* The EU HWC CO guideline is based on either 95% compliance with the 156 ppm level on a 10 minute average or 100% compliance with the
104 ppm level on a 30-minute average in any day.
sThe proposed MWC and MWI and the EU MWC guideline are for HCI only.
TABLE VII.2— COMPARISON OF STANDARDS FOR NEW SOURCES
dscm TEQ, and/
or Total
congeners
PM, mg/dscm
SVM, ng/dscm
LVM uo/dscm
CO, ppmv
Large MWCs
13 Total (or 7 if
testing less fre-
quent)
24
80 or 85% Reduct
Cd: 20
Pb: 20
None
50 to 150 4 to 24
hr avg.
MWls
1 9 TEQ or 80
Total
30
470 or 85%
Reduct.
Cd: 50
Pb: 100
None
50 12-hr avg
EU HWCs1
0 19 TEQ
13 24-hr avg
13-39 30-min
avg2.
65
Cd: 3.25
Tl: 3.25
Pb-653 .
585 3
52, 24-hr avg
104, 30 min avg4
156, 10 min avg4
Proposed HW in-
cinerators
50 10-
62 ;
60
100 1 hr avg
Proposed HW ce-
ment kilns
0.20
69 2-hr avg
hr avg
55
44
Wet and Long,
Dry Kilns None
Kilns with By-pass
1 00 in by-pass
duct (or HC
cannot exceed
6.7) 1 hr avg.
Proposed HW
LWAKs -
72 1 0-hr avg.
5.2.
55.
100 1 hr avg.
-------
27422
Federal Register / Vol. 61, No. 77 /Friday, April 19, 1996 / Proposed Rules
TABLE VII.2.—COMPARISON OF STANDARDS FOR NEW SOURCES—Continued
HC
HCI and CI2, ppmv
as HCI equiva-
lents5
Large MWCs
None
25 or 95% Reduct
MWIs
None
42 or 97% Reduct
EUHWCs1
8 24 hr avg
8-16, SOmin
avg2.
8, 24-hr avg
8-^48, 30 min
avg 2
Proposed HW in-
cinerators
1 2 1 hr avg
6
Proposed HW ce-
ment kilns
Wet and Long
Dry Kilns
20 in main stack 1
hravg
Kilns with By-pass
6.7 in by-pass
(or CO cannot
exceed 100) 1
hr avg
7
Proposed HW
LWAKs
14 1 hr avQ
62
Notes:
1The EU HWC guidelines have been corrected from the European basis of 11% O2 and 0°C to the US basis of 7% O2 and 20°C. Both are ex-
pressed on dry emissions.
2The EU HWC PM, HC, and HCI guidelines are based either 97 % compliance with the lower number or 100% compliance with the higher
number on a 30-minute average over a year.
3The EU LVM guideline is 650 (ig/dscm and includes Sb, As, Pb, Cr, Co, Cu, Mn, Ni, V, Sn. If all metals are emitted equally, their contribution
to the guideline is 65.ng/dscm. Pb, a SVM, was subtracted from this group, resulting in the 585 ng/dscm level.
4 The EU HWC CO guideline is based on either 95% compliance with the 156 ppm level on a 10 minute average or 100% compliance with the
104 ppm level on a 30-minute average in any day.
5 The proposed MWC and MWI standards and the EU HWC guideline are for HCI only.
VIII. Alternative Floor (12 Percent)
Option Results and Option to Address
Variability
As described in Part 3, Section 5, EPA
considered another approach (termed
the "12 percent approach") to
establishing the MACT floor. In this
approach, tie Agency selected an
emissions floor level based on the
average emissions of the 12 percent
MACT pool and the average variability
within the pool. As in the other
approaches, the standards are based on
HW MTEC where appropriate, 3-run
averages, and a 99th percentile
confidence interval.
Through the evaluation of the
emissions database using this 12 percent
approach, it was determined that
various sources equipped with floor
controls would be unable to meet the
floor emission limits. EPA believes that,
if this approach is used to determine
emission standards, a situation would
be created that is arguably inconsistent
with the spirit of the Act. Furthermore,
it could subject the regulated
community to an undue burden—one in
which some facilities in the MACT floor
pool must add control equipment in
addition to the recognized floor controls
in order to meet the floor levels. It could
also place EPA in a position of
defending a floor-based standard in
which the identified floor control
technology does not clearly achieve the
specified floor emissions levels for all of
the facilities in the MACT floor pool.
Although we are inclined not to use this
evaluation method due to these
concerns, we invite comment on this
approach versus other MACT floor
approaches.
Additionally, information regarding
the level of protection these standards
provide can be found in U.S. EPA, "Risk
Assessment Support to the Development
of Technical Standards for Emissions
from Combustion Units Burning
Hazardous Wastes: Background
Information Document", February 20,
1996.
A. Summary of Results of 12 Percent
Analysis
Table VIII. 1 shows the results of the
12 percent floor analysis for existing
sources:
TABLE VIII.1.—12 PERCENT APPROACH MACT FLOOR RESULTSI
HAP
D/F
Hg
HCI/CI2
SVM
LVM
PM
CO
HC
1 Inite
ug TEQ
jig/dscm
nomv
|ig/dscm
ng/dscm
gr/dscf
DDmv
Domv .
Incinerators
Stnd
025
13
23
53
61
0024
100
12
Cement kilns
Stnd
023
32
25
240
46
0 03
n/a
Main2-20 by
pass3:6.7 (or
CO 100).
LWA kilns
Stnd.
023
32
1800
61
57
0012
100
14
1 All emissions levels are corrected to 7 percent O2.
2 Applicable only to long wet and dry process cement kilns (i.e., not applicable to preheater and/or precalciner kilns).
3 Emissions standards applicable only for cement kilns configured with a by-pass duct (typically preheater and/or precalciner kilns). Sources
must comply with either the HC or CO standard in the by-pass stack.
Table VHI.2 shows the results of the 12 percent approach considering BTF analyses for select HAPs for existing
sources:
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17413
TABLE VIII.2.—12 PERCENT APPROACH BTF OPTIONI
HAP
n/F
Hg.
Ha
ny
HCI/CU
^VM
LA/M
PM
CO
HC
units
fig TEQ :
p.g/dscm . •
DDmV
DDrrw
Incinerators
Stnd
0.25
13 ..:
23
53
61 „
0.024
100
12
Cement kilns
Stnd
0.23
8
25
240
46
0.03
n/a
Main2:20 bypass
3:6.7(orCO
100).
LWA kilns
Stnd
0.23.
8.
67.
61. .
57.
0.012.
100.
14.
1 All emissions are corrected to 7 percent O. ,
2 Applicable only to long wet and dry kilns (i.e., not applicable to preheater and/or precalciner kilns).
3 Emissions standard applicable only for cement kilns configured with a by-pass duct (typically preheater and/or precajciner kilns). Source must
comply with either the HC or CO standard in the by-pass stack.
Information pertaining to the calculation of these floor emission levels can be found in U.S. EPA, "Draft Technical
Support Document for HWC MACT Standards, Volume HI: Selection of Proposed MACT Standards and Technologies".
B. Summary of MACT Floor Cost Impacts and Emissions Reductions.
Under the 12 percent approach, the total national annualized compliance costs for existing sources to meet the
MACT floor levels are estimated to be: (1) for incinerators, $28 million, with the cost per facility averaging $971,000;
(2) for cement kilns, $59 million, with the cost per facility averaging $879,000; and (3) for LWAKs, $3 million, with
the cost per facility averaging $860,000. These total compliance costs equate to $49 per ton of hazardous waste burned
for incinerators, $65 per ton of hazardous waste burned for cement kilns, and $52 per ton of hazardous waste burned
for LWAKs. EPA estimates that up to four commercial incinerators will cease burning hazardous waste due to the
compliance costs associated at the floor, in addition to three cement kilns and one lightweight aggregate kiln. However,
we also believe that the these estimates are exaggerated because, they are based on emissions levels determined during
trial burns and compliance performance tests, which produce emissions far in excess of the emission levels most facilities
achieve in day-to-day operation.
There would be substantial emissions reductions at the MACT floor level, compared to baseline emissions. Table
VIII.3 summarizes the estimated national emissions for incinerators if tie facilities were operating at a level to meet
the 12 percent MACT floor level. Also, the estimated percent reduction of HAP emissions from baseline are shown.
Tables Vffl.4 and VIII.5 show similar results for cement and lightweight aggregate kilns.
TABLE VIII.3.—NATIONAL EMISSIONS ESTIMATES FOR INCINERATORS 12 PERCENT MACT APPROACH
HAP
SVM (Cd Pb)
LVM (As Cr Sb Be)
HC1/CU
Particulate Matter
Annual emissions at MACT floor level
3 0 grams TEQ/yr
0.2 tons/year
1 .0 tons/year ,
0.8 tons/year
293 tons/year ..".....
650 tons/year
Percent reduc-
tion from
baseline emis-
sions (percent)
96
96
98
97
83
67
TABLE VIII.4.—NATIONAL EMISSIONS ESTIMATES FOR CEMENT KILNS 12 PERCENT MACT APPROACH
HAP
SVM (Cd Pb) .
LVM fAs Cr Sb B&\
HCI/CIz . -
Annual emissions at MACT floor level
7 0 grams TEQ/yr
1 .7 tons/year
4.0 tons/year
0 9 tons/year
761 tons/year '.
1 877 tons/year ;...
Percent reduc-
tion from
baseline emis-
sions (percent)
99
71
87
73
71
56
-------
17414
Federal Register / Vol. 61, No, 77 / Friday, April 19, 1996 / Proposed Rules
TABLE VIII.5.—NATIONAL EMISSIONS ESTIMATES FOR LWAKS 12 PERCENT MACT APPROACH
HAP
Dioxin/Furans (TEQ) „.
Mercury
SVM (Cd, Pb) ;
LVM (As, Cr, Sb, Be) .:..
HCI/Cb , .,
Particulate Matter
Annual emissions at MACT floor level
(not determined) 12S .... - -J
0.03 tons/year
0.04 tons/year Ti
0.07 tons/year , , '. -
2760 tons/year • •••->••••'
26 tons/year -
Percent reduction
from baseline emis-
sions
(not determined)
91%
94%.
67%
9%
45%
C. Alternative Floor Option: Percent
Reduction Refinement
The Agency is also considering
whether to use a refinement technique,
in establishing the MACT floor that
would modify either the 6 percent
approach, used as the hasis of today's,
proposal, or the 12 percent option .
discussed previously. This refinement
attempts to address the unfavorable .
conditions (i.e. worst-case trial burn or
COG testing) under which the emissions
data was generated.
As discussed elsewhere, EPA is
concerned that our hazardous waste
emissions database is biased high due to
the operating conditions that generated
the data (e.g., metals and chlorine
spiking, non-optimal APCD
performance). Therefore, the analysis of
this database results in floor levels that
are artificially inflated and not
adequately representative of day-to-day
emissions levels. One simplified option
to address this concern is to apply a
"percent reduction" to the calculated ;
floor levels derived from either the 6
percent or 12 percent approach. We
invite comment on this approach
particularly with respect to the
appropriate percent reductipn(s) to be .. '..
applied. We also solicit information and
data based on routine facility operations
and emissions' levels that could be used
to calculate MACT floors that better
reflect day-to-day operations and that
would avoid the potential difficulties in
attempting to determine the appropriate
percent reduction(s) to be used.
IX. Additional Data for Comment
The Agency has received submissions
from various stakeholders detailing
alternative approaches to establish
MACT floor and beyond-the-floor levels.
The Agency has placed these
submissions into the docket129 for this
rulemaking and specifically requests
comment on the approaches used and ;
the emission levels identified. This
section provides some information on
analyses conducted by the Cement Kiln
Recycling'Coalition and Waste
Technologies Industries to determine
MACT and MACT floor levels.
A. Data from Cement Kiln Recycling
Coalition '•
The Cement Kiln Recycling Coalition
(CKRC) is a trade association with a
membership comprised of cement
companies that burn hazardous waste
fuel arid related companies engaged in
the processing and marketing of these
fuels. CKRC conducted a technical
analysis of the hazardous waste-burning
cement kiln's emissions database,
identified the best performing sources
and MACT control technology, and
determined MACT floor emission levels
for dioxin and furans and six metal
HAPs. CKRC's initial analysis specified
separate MACT floor levels based on
cement kiln process type (i.e., separate
floors were developed for cement kilns
employing dry production processes
and wet production processes).130 The
MACT floor results are provided in
Table IX.A.l below.
TABLE IX.A.L—CKRC's PROPOSED MACT FLOOR.EMISSION LEVELS FOR EXISTING CEMENT KILNS (BASED ON DRY AND
WET PROCESS SUB-CATEGORIES)
HAP
Arsenic :
Beryllium '.
Cadmium
Chromium
Chromium (VI)
Lead , '
Mercury ,
Dioxins/Furans
Dry process CKs
3 ng/dscm
0 3 (ig/dscm
30 jig/dscm
485 (ig/dscm
8 ug/dscm
143 fiQ/dscm
NA .
1.7 ng/dscm (TEQ)
Wet Process CKs
32 ug/dscm
2.0 ng/dscm (TEQ).
While CKRC states that sub-
categorization is appropriate, they have
analyzed recent data based on no sub-
categorization and arrived at the floor
levels and (generally) achievable
128 The database is insufficient to make a realistic
determination of the emissions at the baseline or for
the 12 percent option.
• 1M In addition to the submission discussed in this
section, the petitions in the docket for this
rulemaking include: (1) Hazardous Waste Treatment
Council (now Environmental Technology Council),
"Petition for Rulemaking under the Resource
beyond-the-floor (BTF) levels presented
in Table IX.A.2.13'Note that this, , ,
subsequent re-analysis does not
differentiate cement kilns by process
type (Le., wet.and dry process). CKRC
Conservation and Recovery Act to Establish
Uniform National Performance Standards for all
Combustion Facilities based on the Best Available
Technology", May 18,1994; and (2) Cement Kiln
Recycling Coalition, "Petition for Rulemaking
under the Resource Conservation .and Recovery Act
to Modify the Rules for the Burning of Hazardous
•Waste", January 18; 1994.
also emphasizes that the levels
identified in Table IX.A.2 were derived
assuming testing under normal facility
operating conditions using hazardous
waste as a fuel and does not reflect use
tan Environmental Risk Sciences Incorporated
(prepared .for CKRC), "An Analysis of Technical
Issues Pertaining to the Determination of MACT
Standards for the Waste Recycling Segment of the
Cement Industry" (Volumes I-in), May 3,1995.
. 131 Letter from Craig Campbell, CKRC, to James
Berlow, U.S. EPA, undated but received February
20,1996. '. .
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17415
of continuous emissions monitors for
PM or individual HAPs. In addition,
CKRC emphasizes that, because of
natural variations found in the cement
industry (e.g., high levels of metals in
some raw materials), not all kilns may
be able to achieve these levels. CKRC
believes this reinforces the need for the
ability to make site-specific adjustments
to the limits.
TABLE IX.A.2.—CKRC's ALTERNATE MACT FLOOR AND BEYOND-THE-FLOOR LEVELS FOR EXISTING CEMENT KILNS (No
SUB-CATEGORIZATION) ,
HAP
Low-volatile metals
MACT floor level
0 030 gr/dscf
118 jig/dscm
261 ng/dscm
229 ng/dscm
BTF levels
0.025 gr/dscf:"
80 ng/dscm.
1 50 |ig/dscm.
130 ug/dscm.
We invite comment on CKRC's
approach to identify MACT floor and
BTF levels.
CKRC presented this re-analysis of
MACT emissions levels in tandem with
a recommendation that monitoring
metals levels in collected cement kiln
dust (CKD) is a more effective approach
to ensure compliance with metals
emission standards than monitoring the
feedrate of metals in all feedstreams.
CKRC suggested that CKD monitoring
for metals should be used until GEM
technologies become a workable
alternative. Although CKD monitoring
for metals is currently allowed under
the BIF rule in lieu of feedstream
monitoring and the same methodology
is incorporated into today's proposal
(see proposed § 63.1210(n)(2)), CKRC
has suggested revisions to the
methodology to make it more workable.
See Part Five, Section H.C.4.c.v of this
preamble for a discussion of CKRC's
recommendations.
B. Data from Waste Technologies
Industries
Waste Technologies Industries (WTI)
has submitted data and information to
the Agency pertaining to identification
of MACT floor levels for incinerators.132
WTI raises the following issues: (1) in
determining MACT floor, the Agency
has not considered all of WTI's
emissions data that have been submitted
to the Agency; and (2) the Agency
should subdivide the incinerator source
category to develop separate MACT
standards for commercial versus on-site
incinerators.
We have investigated WTI's concern
about not considering its emissions data
and, based on a preliminary analysis,
determined that WTI's data would not
affect the MACT floor levels that the
Agency has identified for existing or
new incinerators.133
WTI is recommending that the
Agency subdivide incinerators to
develop separate standards for
commercial and on-site sources. WTI
notes that its emissions levels are
substantially lower than the standards
that (it believes) EPA is considering for
proposal. In addition, WTI presents
what it believes are appropriate MACT
limitations for existing commercial, off-
site incinerators.134 The table below
compares WTI's suggested MACT
limitations for commercial incinerators
to the Agency's proposed standards:
Pollutant
<^\/M fim/rfermt
LVM (jtg/dscm)
WTI's recommended standard
33 (0.01 gr/dscf)
167
72
EPA's proposed standard
69 (0.03 gr/dscf).
270.
210.
We invite comment on whether
incinerators should be subdivided by
commercial, off-site units versus on-site
units. Commenters should consider the
criteria EPA uses to determine whether
to subdivide a source category as
discussed above in Section I of Part
Four of this preamble. We also invite
comment on WTI's approach to identify
MACT limitations for commercial, off-
site incinerators.
PART FIVE: IMPLEMENTATION
I. Selection of Compliance Dates
Sections A and B below explain when
existing and new facilities, respectively,
would have to document compliance
with the proposed MACT standards.
132Letter from Barry Direnfeld, Swidler & Berlin,
to Michael Shapiro, dated January 23,1996, with
an attached letter from Fred Sigg, Von Roll/WTI, to
Sally Katzen, Office of Management and Budget,
dated January 19,1996.
Section C presents a proposal for a one
year compliance extension in order to
institute pollution prevention/waste
minimization measures.
EPA is proposing a different
definition of compliance date for HWCs
than is provided by existing 40 CFR
§ 63.2. Although that section defines
compliance date as the date when a
source must be in compliance with the
standards, 40 CFR § 63.7 requires
performance testing to document
compliance with the emission standards
(and performance evaluations to
document compliance with
requirements for continuous monitoring
systems) after the compliance date. This
use of the term "compliance date" is not
consistent with the current RCRA
definition and regulatory requirements
for HWCs.
To achieve more consistency and to
avoid potential duplication and conflict,
the Agency is proposing to define
compliance date for HWCs in § 63.1201
as the date when a HWC must submit
the initial notification of compliance. In
addition, notification of compliance
would be defined as a notification in
which the owner and operator certify,
after completion of performance
evaluations and tests, that the HWC
meets the emissions standards, CMS,
and other requirements of Subpart EEE,
Part 63, including establishing operating
limits to meet standards for which
compliance is not based on a CEM.
133 See memorandum from Bruce Springsteen,
EER, to Shiva Garg, EPA, dated February 26,1996,
entitled "Determination of the effects of the
inclusion of new WTI test burn data on the MACT
floors."
134 See letter from Gary Liberson, Environmental
Risk Sciences, to Michael Shapiro, EPA, dated
February 21,1996.
-------
27416
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
For HWCs, initial compliance would
thus mean that a facility has: (1)
completed all modifications necessary
to meet the standards; (2) conducted all
emissions tests to verify compliance and
set operating limits; (3) installed and
satisfactorily performance tested all
continuous monitoring systems (CMS)
including continuous emissions
monitors (GEMS); and (4) postmarked a
letter to the director that transmits the
(successful) emission results of the
initial comprehensive performance test,
performance test results for CMS, and
all operating limits, and that states the
facility is in compliance. Requirements
to ensure compliance after the initial
compliance notification are discussed in
the preamble in Section II of Part Five.
A. Existing Sources
EPA proposes that a facility he in
compliance with these standards within
three years after the date of publication
of the final rule in the Federal Register
(which is also the effective date of the
rule). See proposed § 63.1206(a). EPA
believes that the vast majority of sources
(approximately 90 to 95 percent) would
require substantial modifications to
operating and/or emission control
equipment to comply with the proposed
standards. Three years is a reasonable
estimate of the time it will take for a
facility to: read and analyze the final
rule; conduct tests to identify cost-
effective approaches to comply with the
standards; complete the engineering
analysis and design; fabricate, install,
start up and shake down the modified
facility; conduct preliminary emissions
tests; conduct formal compliance
testing; analyze samples and evaluate
test results; prepare the notification of
compliance; and obtain management
certification of the results.
Nonetheless, the Agency believes that
some sources would be able to comply
with the rule (i.e., submit a notification
of compliance) before three years after
the date of publication of the final rule.
For example, some sources may require
only minor modifications to emission
control equipment and could comply
substantially sooner than sources that
need a major retrofit. Accordingly, we
invite comment on how such sources
could be identified and strategies that
could be used to encourage or require'
them to comply at the earliest possible
date.
We note that the CAAA allows a
maximum compliance period of three
years (see § 112(I)(3)(A)), unless a
waiver is granted on a case-specific
basis. Section 63.6(i)(4)(i)(A) provides
for a one year time extension "if such
additional time period is necessary for
the installation of controls." If an owner
or operator needs to modify the RCRA
permit in order to allow modifications
to the facility necessary to comply with
the MACT standards, we believe
inability to comply with the MACT
standards within three years because of
the need to modify the RCRA .permit
could constitute a valid reason for
granting a time extension under
§ 63.6(i). See discussion below. That is,
the modification to the RCRA permit
would be needed "for the installation of
controls."
Sources with RCRA permits can ,
modify their facilities only after
complying with the permit modification
procedures of 40 CFR 270.42. If an
owner and operator make a good faith
effort to obtain the permit modification
in time to submit a notification of
compliance under today's proposed rule
within three years of the effective date
but cannot do so for reasons beyond
their control (for example, the state in
which the facility is located is in the
process of receiving oversight authority,
or the Agency is unable to respond in
a timely manner to all permit
modification requests), the
Administrator may grant a one-year time
extension.
Note also that, as discussed above, the
one-year time extension provided by
§ 63.6(i) applies to a different definition
of compliance than that proposed by
today's rule for HWCs. By the date of
compliance under this proposal, a HWC
must have submitted a notification of .
compliance as defined above. Thus,
although we are proposing a one-year
time extension for initial compliance for
HWCs using the procedures established
in existing § 63.6(i), a HWC must submit
a notification of compliance by the end
of the time extension, if granted, while
other MACT sources would continue
under the current rules unamended (i.e.,
they would conduct their performance
test after the end of the time extension).
See existing § 63.7(a).
A special case for HWCs exists for an
existing unit that would not be subject
to regulation on the effective date of this
rule because it does not burn a
hazardous waste but which
subsequently becomes subject to
regulation under today's proposed
MACT standards because one of its
waste streams later becomes a newly
identified or listed hazardous waste. In
this case, we propose that the facility be
considered an "existing source", since it
would be inappropriate to apply new
source MACT to a facility which has not
altered its conduct, and which only
becomes subject to this rule because of
additional regulatory action taken by
EPA (or an authorized state). Such a
facility would have three years after the
date of publication in the Federal
Register of the final rule listing the
waste as hazardous to come into
compliance with these regulations.135
Finally, EPA wants to ensure that
only those facilities that plan to comply
with the new regulations are allowed to
burn hazardous waste during the
compliance period. Accordingly, the
rule would provide that, if the owner or
operator of an existing source did not
submit a notification of compliance by
the applicable date, the source .must
immediately stop burning hazardous
waste when the owner or operator first
determines that the notification will not
be submitted by the applicable date (i.e.,
following the effective date, but well
before the compliance deadline) and
could not resume burning hazardous
waste except under the requirements for
new MACT sources. To comply with the
deadline for the initial notification of
compliance, a source will have had to
begin making preparations well in
advance of the deadline. We invite
comment on strategies that could be
used to determine when a source could
realistically determine whether or not it
will meet the notification deadline and
comply with the new standards.
We note that there would also be
substantial RCRA implications for a
facility that does not comply with the
applicable deadlines in a timely fashion.
In particular, the source could not
resume burning hazardous waste
without being issued a RCRA operating
permit. Further, if the source had
already been issued a RCRA operating
permit, hazardous waste could only be
burned (after missing the deadline for
submitting an initial notification of
compliance) for a total of 720 hours and
only for the purpose of pretesting or
comprehensive performance testing.
Finally, if a source with a RCRA
operating permit failed to submit an
initial notification of compliance by the
deadline, the source must, within 90
days of missing the initial notification of
compliance, either submit a notification
of compliance with MACT new
standards or begin RCRA closure
procedures unless the Administrator
grants an extension of time in writing
prior to the 90-day deadline for good
cause. Examples of good cause that the
Agency would be willing to evaluate
135 Note that in other cases, an existing source that
begins to burn hazardous waste after the effective
date of this rule (and therefore changes its conduct)
is classified as a new source and would have to
comply with today's rules when the hazardous
waste is first burned. The source would also have
to obtain a RCRA operating permit before
commencing hazardous waste management
activities since it would be ineligible for interim
status (assuming it is conducting no other
hazardous waste management activities).
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17417
are: the facility now must undergo
significant modifications in order to
comply with the more stringent MACT
new standards that will take longer to
complete than the deadline allows, or
the facility must contract for substantial
new services in order to show
compliance with the new standards.
EPA believes that these requirements
are necessary to ensure that owners and
operators that elect not to comply with
the standards do not continue to burn
hazardous waste beyond the date on
which the source determines that they
will not comply with the promulgated
standards.
B. New Sources
Section 63.6 states that new or
reconstructed sources "shall comply
with such standard[s] upon startup of
the source." See also proposed
§ 63.1206(b). One exception, available
only to facilities which commence
construction between proposal and
promulgation, is in the instance where
a standard more stringent than the one
proposed is promulgated. In this
instance, three years can be granted for
the new source to be in compliance with
the standard which is more stringent.
The new source shall be in compliance
upon startup with all standards which
are not more stringent than those
proposed. Section 63.2 defines new
source as"* * * any affected source the
construction or reconstruction of which
is commenced after the Administrator
first proposes a relevant emission
standard * * * ." For discussion on
reconstruction, see section vn.C. of this
part of this preamble.
C. One Year Extensions for Pollution
Prevention/Waste Minimization
EPA is also seeking comment on a
proposal to consider extension of
compliance deadlines for up to one year
beyond the three year deadline from the
date of promulgation of this rule, on a
case-by-case basis, for facilities which
request an extension to implement
pollution prevention/waste
minimization measures that will enable
the facility to meet MACT standards and
that cannot practically be implemented
within the three year compliance
deadline.
During development of the Hazardous
Waste Minimization National Plan
(released in 1994), some companies
pointed out that short compliance
deadlines after the promulgation of
some rules have precluded them from
completing necessary pollution
prevention planning and
implementation that would facilitate
meeting compliance requirements
through source reduction and
environmentally sound recycling. As a
result, companies opt for installing often
expensive "end-of-pipe" pollution
controls in order to meet compliance
deadlines. In addition, once capital has
been sunk into end-of-pipe pollution
controls which are large enough to
handle current and future waste
volumes, there is little incentive for
companies to then spend money
exploring pollution prevention/waste
minimization options.
EPA believes that the three year
compliance deadline for meeting the
MACT standards in this rulemaking
should in most cases be sufficient for a
facility to complete the pollution
prevention planning and
implementation that might be necessary
to meet MACT standards. In cases
where facilities can provide information
that shows that additional time is
necessary to complete this process, EPA
is proposing to grant up to a one year
extension for facilities to complete
pollution prevention planning and
implementation, and to satisfy all of the
procedures in this rule for
demonstrating compliance. This
proposed extension is consistent with
other portions of today's proposal,
including the section on permitting
procedures which describes pollution
prevention/waste minimization options
during the permitting process.
II. Selection of Proposed Monitoring
Requirements
Section 114(a) of the CAA requires
monitoring to ensure compliance with
the standards and the submission of
periodic compliance certifications for
all major stationary sources. Given that
all HWCs are subject to regulation as
major sources, the proposed compliance
monitoring requirements discussed
below would apply to all HWCs.
In this section we discuss the
following: (a) the compliance
monitoring hierarchy; (b) how
operations during comprehensive
performance testing would be used to
establish limits for operating
parameters; (c) for each emission
standard, requirements for continuous
emissions monitors (if any) and limits
on operating parameters to ensure
compliance; (d) compliance with
controls on fugitive combustion
emissions; (e) requirements for
automatic waste feed cutoffs and
emergency safety vent openings; (f)
quality assurance requirements for
continuous monitoring systems (CMS);
and (g) protocols to ensure and
document compliance.
A. Monitoring Hierarchy
The proposed compliance monitoring
requirements were developed by
examining the hierarchy of monitoring
options available for specific processes,
pollutants, and control equipment. The
approach involves describing, on an
emission standard specific basis, what
monitoring is required for a source to be
in compliance. This approach was also
used for the secondary lead smelter
MACT (59 FR at 29772, June 9,1994),
another rule where the sources process
hazardous waste.
The monitoring hierarchy is three-
tiered. The top tier of the monitoring
hierarchy is the use of a continuous
emissions monitor system (GEMS, also
known as "GEM") for that HAP or
standard. In the absence of a GEMS for
that HAP or standard, the second tier is
the use of a GEMS for a surrogate of that
HAP or standard and, when necessary,
setting some operating limits to account
for the limitations of using surrogates.
Lacking a GEMS for either, EPA sets
appropriate feedstream and operating
parameter limits to ensure compliance
and requires periodic testing of the
source. In developing this proposal each
tier of the hierarchy was evaluated
relative to its technical feasibility, cost,
ease of implementation, and relevance
to its underlying process emission limit
or control device.
The proposed standards for hazardous
waste combustors contain monitoring
requirements for process stack
emissions and combustion fugitive
emissions. The proposed standards
require either pollutant monitoring
directly through the use of a GEMS,
surrogate monitoring through the use of
a GEMS, and/or parameter monitoring
that indicates proper operation and
maintenance of a control device.
Recordkeeping is also required to ensure
that specific work practices are being
followed. Section VI of this part
discusses recordkeeping.
B. Use of Comprehensive Performance
Test Data to Establish Operating Limits
Limits on operating parameters (e.g.,
feedrate limits, temperature limits)
would be based on levels that are
achieved during the comprehensive
performance test. See section III of this
part for the discussion on
comprehensive performance tests.
1. Averaging Periods for Limits on
Operating Parameters
The Agency is proposing various
averaging periods for the limits on
operating parameters: a ten-minute
rolling average; a one-hour rolling
average; and a 12-hour rolling
-------
17418
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
average.136 To show compliance with
any of these rolling averages with
respect to operating parameters that are
established based on levels achieved
during the comprehensive performance
test (rather than on manufacturer
specifications), the monitor must make
a measurement of the parameter at least
once each 15 seconds, and four 15-
second measurements must be averaged
each minute to determine a one-minute
average. Then, each one-minute average
is considered along with the previous
one-minute averages over the averaging
period to calculate a new rolling average
level each minute. Thus, irrespective of
the averaging period, a new rolling
average level is calculated each minute.
The duration of the averaging period
affects the number of one-minute
averages used to calculate the level. For
example, if a limit is based on a 12-hour
rolling average, each new one-minute
average is added to the previous 719
one-minute average values to calculate a
new 12-hour rolling average value each
minute.
A ten-minute average is proposed
when the Agency is concerned that
short-term perturbations above the limit
will result in high emissions that cannot
be offset by lower emissions during
periods of more appropriate
operation.137 Since the ten-minute
average is used to control short-term
perturbations and does not control
average emissions, it will always be
used with a one hour average designed
to control average emissions. (An
exception is when the 10-minute
average is used to control a design
specification of the APCD manufacturer.
In this event, a ten-minute average may
be used alone.) It could be argued that
a short term averaging period other than
ten minutes could be used. However,
the Agency is concerned about setting
the averaging period shorter than 10
minutes. Shorter averaging periods
would result in more extreme (i.e.,
absolute maximum or minimum) limits
and could lead to higher emissions.
Conversely, EPA could set a short-term
averaging period longer than ten
minutes, but believes that ten minutes is
an appropriate, achievable,
conservative, and reasonable duration
for the short averaging period.
A one-hour averaging period is
proposed in instances where the Agency
136 We note that today's rale would establish an
instantaneous limit; i.e., a limit where no averaging
is allowed, to ensure that less than ambient
pressure is maintained in the combustion system at
all times to control fugitive combustion emissions.
137 An example is for inlet temperature to dry PM
APCDs to control dioxin. Dioxin increases
exponentially with increasing temperature, so a
short-term increase in temperature will not be offset
by short-term decreases in dioxin emissions.
is less concerned about perturbations
and/or wants to limit average
emissions.138 Hourly rolling averages are
currently required under the BIF rule
and are required for some incinerators.
The value of one-hour averages will
tend to be less extreme than 10-minute
averages since perturbations are
averaged out over more normal data
and, thus, are better at controlling
average emissions than 10-minute
averages. It could be argued that an
averaging period shorter than one hour
would be appropriate, but EPA is
selecting a ten-minute average to control
perturbations and believes this is
sufficient. It could be argued that
averaging periods longer than one hour
could also be appropriate, but setting
limits on operating parameters is at the
bottom of the monitoring hierarchy and,
as such, a conservative approach is
preferable.
The twelve-hour averages are being
proposed in instances when the Agency
wants to control average emissions and
is concerned that the one-hour average
may not be achievable or may be overly
restrictive. Twelve-hour averages are
proposed only for feedrates: metals and
chlorine. For each of these, feedstream
analysis is necessary to determine the
concentration in each of the feedstreams
and this makes using an averaging
period shorter than twelve hours
problematic. EPA could use an
averaging period longer than twelve
hours, but believes that twelve hours is
achievable. EPA is concerned about this
12-hour average in that it may be
inconsistent with averaging periods for
GEMS; namely, it is longer than the
metals, HC1, C12, or PM averaging
periods. A 12-hour average is
inconsistent because, at the top of the
monitoring hierarchy, GEMS averaging
periods should be longer, i.e., less
conservative, than feedstream
monitoring, at the bottom of the
hierarchy. EPA invites comment on this
issue. Alternate averaging periods for
chlorine and metals feedrates are
discussed below in the appropriate
sections.
As noted earlier, for compliance with
these averaging periods, EPA proposes
that averages be calculated every minute
on a rolling-average basis. It is also
proposed that the one-minute average be
the average of the previous four
measurements taken at 15-second
intervals. This is the approach required
by the BIF rule. All 15-second
measurements would be used without
smoothing, rounding, or data checks. No
15-second observations may be "thrown
out" for any reason.
2. How Limits Would Be Established
from Comprehensive Test Data
This section explains how operating
limits for the averaging periods
discussed above are established from
the comprehensive test data. Note that
all averages are rolling averages, based
on a one-minute average.
Ten-minute rolling averages would be
established as the average over all
comprehensive test runs of the highest
or lowest (as specified) ten-minute
rolling average for each run.
One of two approaches would be
specified to establish limits on an
hourly rolling average basis: an average
level or an average of the highest or
lowest (as specified) hourly rolling
average. In most cases, it is derived by
averaging all of the one-minute averages
during all the runs of the
comprehensive performance test. In the
few cases when an average of the
maximum hourly rolling averages is
specified, the limit is derived by taking
the average of the highest hourly
average for each run of the
comprehensive performance test.
Twelve-hour rolling averages for
feedstreams would be derived by
averaging all of the one-minute averages
during all the runs of the-
comprehensive performance test
irrespective of the total duration of the
test.139 Separate twelve-hour averages
would apply to all feed locations.
3. Example of How Limits Would Be
Established
For example, if a facility were to have
a fabric filter (FF), it might have a limit
on maximum FF inlet temperature on a
ten-minute average to ensure
compliance with the dioxin and furan
standard. If this is the case, during the
comprehensive performance test, the
facility would monitor FF inlet
temperature. The facility would then
take the highest single ten-minute
rolling averages of FF inlet temperature
from each of the three comprehensive
test runs and average them together. If
these single largest ten minute rolling
averages from each of the three runs
were 140,150, and 160°C, then the
maximum ten-minute rolling average for
FF inlet temperature would be 150°C.
If the same parameter were also to
have an hourly rolling average based on
all data from all runs, the facility would
138 An example is flue gas flowrate. This
parameter is important, but slight increases in flow
rate can be offset by proportionate decreases in
flowrate. Therefore, average flowrate is important
without regard to perturbations..
139 Or, if the source elects to define different
operating modes and conduct performance testing
under each mode, the one-minute averages would
be averaged for all runs for each test condition
(representing each mode of operation).
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17419
sum up all one-minute averages
occurring during the comprehensive
performance test and average them
together. This would become the hourly
rolling average for this parameter.
Twelve-hour feedrate limits are
calculated similarly. For SVM, the
facility would sum the total feed from
all runs of the comprehensive
performance test and divide that sum hy
the number of minutes of all three runs
of the comprehensive test. For this
example, assume that both Cd and Pb
are fed during the comprehensive
performance test, that the feedrate for
Cd was 5,30, and 25 and for Pb was
100, 70, and 85 for each of the three
runs of the comprehensive performance
test and that the time duration of each
run was 205, 230, and 195 minutes. The
total amount of SVM fed would be 315
and the time duration of the test would
be 630 minutes. Therefore, the SVM
limit would be 315, divided by 630
minutes, or 0.50. During normal
operation the SVM feedrate would be
calculated every minute to ensure it
does not exceed the 0.50 SVM limit by
averaging the current and previous 719
one-minute averages.
C. Compliance Monitoring
Requirements
Monitoring requirements are
proposed to ensure compliance with the
following emission standards: dioxin
and furan (D/F), mercury (Hg),
semivolatile metals (SVM), low-volatile
metals (LVM), carbon monoxide (CO),
hydrocarbons (HC), hydrochloric acid
(HC1) and chlorine gas (C12) (combined
and reported as HC1), and particulate
matter (PM). See proposed § 63.1210.
Monitoring requirements for
combustion fugitive emissions are
proposed as well.
Table V.2.1 summarizes today's
proposed compliance monitoring
requirements.
1. Continued Applicability of RCRA
Omnibus Authority
When a RCRA operating permit is
issued under Part 270 after a source has
submitted its initial notification of
compliance with the proposed MACT
standards, a permit writer would
continue to have the discretion
currently provided by § 264.345(b)(6) of
the incinerator standards and
§§ 266.102(e) subparagraphs (2)(i)(G),
(3)(i)(E), (4)(ii)(J), (4)(iii)(J), and (5)(i)(G)
of the BIF standards to supplement
these operating parameter limits as
necessary to protect human health and
the environment on a site-specific basis
to ensure that today's proposed
emission standards are being met. This
means the RCRA permit writer's
authority to use instantaneous limits or
averaging periods other than those
specified here, or require operating
parameters in addition to those
specified here, is maintained during the
RCRA permitting process. See proposed
§§ 264.340(b)(2)(iii) and
266.102(a)(2)(ii).
TABLE V.2.1 .—SUMMARY TABLE OF PROPOSED MONITORING REQUIREMENTS
Device
Continu-
ous
Mon-
itor.
Carbon
Injec-
tion.
Carbon
Bed.
Dioxin
Inhibi-
tor.
Parameter
Stack GEMS .
Max Inlet Temp
to Dry PM
APOD.
Min Carbon In-
jection
Feedrate
(Carbon Feed
through Injec-
tor).
Min Carrier
Fluid
Flowrate or
Nozzle Pres-
sure Drop.
Carbon Specs
Max Age of
Carbon (Time
in-use).
Carbon Specs
Min Inhibitor
Feedrate.
Inhibitor Speci-
fications.
D/F
•
•
•
•
•
•
•
•
Hg
•
(2)
(2)
P)
(2)
(2)
(2)
PM
•
SVM
(')
•
LVM
C)
•
CO&
HC
•
HCI&
CI2
0)
Limits
from
GEMS
Stnds.
Comp
'Test.
Comp
Test.
Manuf
Spec.
Comp
Test.
Initial
Comp
Test.
Conf
Tests.
Sub.
Comp.
Tests.
Comp
Test.
Comp
Test.
Comp
Test.
Avg pe-
riod
varies
10 min
1 hour
10 min
1 hour
10 min
n/a
n/a
n/a
n/a
n/a
10 min
1 hour.
n/a :
Limits set as
Units of Stand-
ard.
Avg of Max 10
min RA.
Avg over all
runs.
Avg of Min 10
min RA.
Avg over all
runs.
Same brand
and type.
Manuf specs
(no C aging).
Normal C
Change-out
Schedule.
Max C Age is
the age dur-
ing subse-
quent Comp
Tests.
Same brand
and type.
Avg of Min 1 0
min RA.
Avg over all
runs.
Same brand
and type.
-------
17420
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
TABLE V.2.1.—SUMMARY TABLE OF PROPOSED MONITORING REQUIREMENTS—Continued
Device
Catalytic
Oxi-
dizer.
Good
Com-
bus-
tion.
Good
Com-
bus-
tion
and
APCD
Effi-
ciency.
Feed
Con-
trol.
Wet
Scrub-
ber.
Parameter
Min Fine Gas
Temp at En-
trance.
Max Age (Time
in-use).
Catalyst Re-
placement
Specs:.
— Catalytic
Metal Load-
ing (each
metal).
— Space Time
— Substrate
Construction
(mat'ls, pore
size).
Max Flue Gas
Temp at En-
trance.
Maximun Batch
Size, Feeding
Frequency,
and Minimum
Oxygen Con-
centration.
Max Waste
Feedrate.
Min Comb
Chamber
Temp (Exit of
Each Cham-
ber).
Max Flue Gas
Flowrate or
Production
Rage.
Max Total Met-
als Feedrate
(all streams).
Max Pumpable
Liquid Metals
Feedrate.
Max Total Ash
Feedrate (all
streams).
Max Total Chlo-
rine Feedrate
(all streams).
Min Press Drop
Across
Scrubber.
Min Liquid Feed
Press.
Min Liquid pH
Min Slowdown
(Liq Flowrate)
or Max Solid
Content in
Liq.
D/F
•
•
•
•
•
•
•
•
(2)
(2)
(2)
Hg
(2)
(2)
(2)
(2)
(2)
(2)
PM
(2)
(2)
(2)
(2)
(2)
SVM
(2)
•
•
(2)
(2)
(2)
LVM
(2)
•
•
•
(2)
(2)
(2)
CO&
HC
HCI&
CI2
•
!/
•
•
(/
Limits
from
Comp
Test.
Manuf
Spec.
Comp
Test.
Manuf
Spec.
Comp
Test.
Comp
Test
Test.
Comp
Test.
Comp
Test.
Comp
Test.
Test.
Comp
Test.
Manuf
Spec.
Test.
Comp
Test.
Avg pe-
riod
10 min
1 hour
As
spec-
ified..
10 min
n/a
10 min
1 hour
1 hour
12 hour
12 hour
12 hour
10 min
1 hour
1 0 min
1 hour
10 min
1 hour
Limits set as
min RA. ,
Avg over all
runs.
during pre-
vious Comp
Test.
As specified
Lightest batch
fed. Least
frequent
feeding High-
est O2 level.
hour RA.
min. RA
Avg over all
runs.
Avg- of Max 1
hour RA.
Avg over all
runs.
Avg over all
runs.
runs.
Avg of Min 1 0
min RA
Avg over all
runs.
n/a
Avg of Min 1 0
min RA
Avg over all
runs.
Avg of Min/Max
10 min RA
Avg over all
runs.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17421
TABLE V.2.1.—SUMMARY TABLE OF PROPOSED MONITORING REQUIREMENTS—Continued
Device
Ionizing
Wet
Scrub-
ber.
Dry
Scrub-
ber.
pc
ESPs ....
Parameter
Min Liq Flow to
Gas Flow
Ratio.
Min Press Drop
Across
Scrubber.
Min Liquid Feed
Pressure.
Min Slowdown
(Liq Flowrate)
or Max Solid
Content in
Liq.
Min Liq Flow to
Gas Flow
Ratio.
Min Power
Input (kVA:
current and
voltage).
Min Sorbent
Feedrate.
Min Carrier
Fluid
Flowrate or
Nozzle Pres-
sure Drop.
Sorbent Speci-
fications.
Min Press Drop
Across De-
vice.
Min Power
Input (kVA:
current and
voltage).
D/F
(2)
(2)
(2)
(2)
(2)
(2)
(2)
(2)
Hg
(2)
(2)
(2)
(2)
(2)
(2)
(2)
(2)
PM
(2)
(2)
(2)
(2)
(2)
(2)
(2)
(2)
SVM
(2)
(2)
(2)
(2)
(2)
(2)
(2)
(2)
LVM
(2)
(2)
(2)
(2)
(2)
(2)
(2)
(2)
CO&
HC
HCI&
CI2
•
•
•
•
•
•
•
Limits
from
Comp
Test.
Comp
Test.
Manuf
Spec.
Comp
Test.
Comp
Test.
Comp
Test.
Comp
Test.
Manuf
Spec.
Comp
Test.
Comp
Test.
Comp
Test.
Avg pe-
riod
10 min
1 hour
10 min
1 hour
10 min
10 min
1 hour
10 min
1 hour
10 min
1 hour
10 min
1 hour
10 min
n/a
10 min
1 hour
10 min
1 hour
Limits set as
Avg of Min 1 0
min RA
Avg over all
runs.
Avg of Min 1 0
min RA
Avg over all
runs.
n/a
Avg of Min/Max
1 0 min RA
Avg over all
runs.
Avg of Min 10
min RA
Avg over all
runs.
Avg of Min 10
min RA
Avg over all
runs.
Avg of Min 1 0
min RA.
Avg over all
runs.
n/a
Same brand
and type.
Avg of Min 1 0
min RA.
Avg over all
runs.
Avg of Min 1 0
min RA.
Avg over all
runs.
'.Stack GEMS is optional for the SVM, LVM, and HCI and CI2 standards. If a GEMS is used for compliance, none of the feedstream and oper-
ating parameters for that HAP would apply.
(J)»lf GEMS are not required in the final rule for PM and/or Hg, the operating limits for these parameters would apply.
Definitions:
"Comp Test"=Comprehensive Performance Test.
"Conf Test"=Confirmatory Performance Test.
2. Dioxin and Furan (D/F) and complying with limits on operating on-site facilities). Table V.2.2
EPA is proposing that sources comply Parameters and performing D/F test summarizes these limits. See also
with the D/F standard by establishing every 18 months (or 30 months for small proposed § 63.1210(j).
TABLE V.2.2.—SUMMARY OF PROPOSED DIOXIN AND FURAN MONITORING REQUIREMENTS
Parti/^i ilnto Matter /PNvH
Control.
Compliance using
PM GEMS
CO and HC GEMS
CMS at exit of each
chamber.
Max waste feedrate CMS
Limits from
Comp Test
MACT Std
Comp Test
Como Test
Avg. period
10 min
1 hour
1 hour
10 min
11 hour
1 hour
How limit is established
from comp performance
test
Avg of Max 1 0-min RAs.
Avg over all runs.
N/A.
Avg of Max 1 0-min RAs.
Avg over all runs.
Avg of Max 1 hour RAs.
-------
17422
Federal Register / Vol. 61, No. 77. / Friday, April 19, 1996 / Proposed Rules
TABLE V.2.2.—SUMMARY OF PROPOSED DIOXIN AND FURAN MONITORING REQUIREMENTS—Continued
Max Inlet Temp to Dry PM
APCD.
Max Flue Gas Flow/rate or
Production Rate.
Min Carbon Injection Feed
Min Carrier Fluid Flowrate
or Nozzle Pressure Drop.
Carbon Specs
Max Carbon Age, Carbon
Bed.
Min Flue Gas Temp, Cata-
lytic Oxidizer.
Max Age, Catalytic Oxidizer
Catalyst Replacement
Specs.
Max Flue Gas Tempera-
ture, Catalytic Oxidizer.
Min Inhibitor Feedrate
Inhibitor Specs
- Compliance using
For batch fed sources:
-•-limit on batch size, feed-
•ing frequency, and mini-
mum oxygen.
Temp CMS
Flowrate CMS or Produc-
tion Rate.
Feedrate CMS
same
Brand and Type
Max Carbon Lifetime
Inlet to Catalyst
Time in use
Catalytic Metal Loading ....
Space Time
Substrate Construct:
mat'ls, pore size.
Inlet to Catalyst
Feedrate CMS
None
Limits from
Comp Test
Comp Test
Comp Test
Comp Test
Manuf Spec
Comp Test
Initial Comp Test
Conf Tests
Sub. Comp Tests
Comp Test
Manuf Spec .
Comp Test
Manuf Spec
Comp Test
Corno Test
Avg. period
None
10 min
1 hour
1 hour
10 min
1 hour
10 min
N/A
N/A .
N/A
N/A
10 min
1 hour
As specified
N/A
10 min
10 min
1 hour
N/A
How limit is established
from comp performance
test
N/A
Avg of Max 1 0 min RAs
Avg over all runs
Avg of Max 1 hour RAs
Avg of Min 1 0 min RAs
Avg over all runs
N/A
Same brand and type
Manuf Specs (no C aging)
Normal C Change-out
Schedule.
Max C Age is the age dur-
ing sub. Comp Tests.
Avg of Min 1 0 min RAs
Avg over all runs
Same as used during
comp test.
As specified
Avg of Min 1 0 min RAs
Avg over all runs
Same brand and tvoe.
a. Evaluation of Monitoring Options.
D/F partitions into two phases in stack
emissions: a portion is adsorbed onto
particulate and a portion is emitted as
a vapor (gas). Given that there is no
GEMS for D/F, the Agency is proposing
to require a combination of approaches
to control D/F emissions: (1) compliance
with a site-specific PM limit to control
adsorbed D/F; (2) operation under good
combustion conditions to minimize D/F
precursors; (3) temperature control at
the PM control device to limit D/F
formation in the control device; and (4)
compliance with operating limits on
D/F control equipment (e.g., carbon
injection) that a source may elect to use.
b. Operating Parameter Limits.
Today's proposed rule would limit the
following operating parameters to
satisfy the combination of approaches
discussed in the previous paragraph.
i. Control of PM Emissions: To control
D/F and other PICs that are adsorbed to
PM, the rule would require that sources
limit PM emissions to the site-specific
level that occurs when demonstrating
compliance with the D/F (and SVM and
LVM) emission standards. The site
specific operating limit forPM would be
capped at (i.e., could not exceed) the
proposed national MACT standard of 69
mg/dscm. See section 7 of this section
for a discussion on the control of PM
emissions.
ii. Good Combustion: CO and HC
Limits. EPA is proposing CO and HC
standards to ensure good combustion to
help minimize D/F precursors. See
discussion below (section 5 of this
section) for the explanation of the CO
and HC emission standards.
iii. Good Combustion: Maximum
Waste Feedrate. An increase in waste
feedrate without a corresponding
increase in combustion air can cause
inefficient combustion that may
produce (or incompletely destroy) D/F
precursors. Therefore EPA proposes to
limit waste feedrate. For incinerators,
waste feedrate limits would be
established for each combustion
chamber to minimize combustion
perturbations. For CKs and LWAKs
waste feedrate limits would be
established for each location where
waste is fed (e.g., the hot end where
product is discharged, mid-kiln, and at
the cold end where raw material is
fed.140
Feedrate limits would be established
on an hourly rolling average basis as the
average of the highest hourly rolling
average for each run. We specifically
invite comment on whether it would be
more appropriate to establish the limit
based on the average hourly rolling
average over all runs. EPA is not
proposing this more stringent approach
because we consider waste feedrate to
be a secondary control parameter that
may not require such strict control.
See also the discussion in section
II.F.2 below for other requirements to
document compliance with feedrate
limits.
iv. Good Combustion: Combustion
Zone Temperature. As combustion zone
temperatures decrease, combustion
efficiency can decrease resulting in an
increase in formation of (or incomplete
destruction of) D/F precursors. For this
reason, the Agency proposes limiting
combustion zone temperature in each
i4owaste feedrate limits would also be
established for waste fed into a preheater or
precalciner system of a cement kiln facility.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17423
chamber to the minimum level
occurring during the comprehensive
performance test documenting
compliance with the D/F standard.
BIFs and incinerators are already
required to monitor combustion zone
temperature for compliance with metals
emissions standards and destruction
and removal efficiency (DRE).
Monitoring of combustion zone
temperature has been problematic,
however, because the actual burning
zone temperature cannot be measured at
many units (e.g., cement kilns). For this
reason, the BIF rule requires
measurement of the "combustion
chamber temperature where the
temperature measurement is as close to
the combustion zone as possible." See
§266.103(c)(l)(vii).
In some cases, temperature is
measured at a location quite removed
from the combustion zone due to
extreme temperatures and the harsh
conditions at the combustion zone. We
are concerned that monitoring at such
remote locations may not accurately
reflect changes in combustion zone
temperatures. For example, a reduction
in heat transfer chain in a wet cement
kiln due to wear over time or decreasing
raw material feedrate (at a fixed heat
input) in a cement or lightweight
aggregate kiln may increase temperature
at the kiln outlet even if combustion
conditions actually caused a decrease in
combustion zone temperature.
We specifically invite comment on
how to address this issue. For example,
the final rule could require the owner or
operator to identify a parameter that
correlates with combustion zone
temperature and to provide data or
information to support the use of that
parameter in the operating record. The
final rule could also enable the Director
on a case-specific basis to require the
use of alternate parameters as deemed
appropriate, or to determine that there
is no practicable approach to ensure that
minimum combustion chamber
temperature is maintained. In that case,
the Director may determine that the
source could not comply with the
regulations and, thus, could not burn
hazardous waste.
Note also that, in the final rule, we
would revise the existing BIF and
incinerator rules to conform with the
approach used in the final MACT rule.
Those conforming revisions would
become effective six months from the
date of publication of the final rule in
the Federal Register and would remain
in effect until the MACT standards take
effect.
The temperature limit(s) would apply
to each combustion zone into which
hazardous waste is fired. As examples,
for incinerators with a primary and
secondary chamber, separate limits
would be established for the combustion
zone in each chamber. For kilns,
separate temperature limits would apply
at each location where hazardous waste
may be fired (e.g., the hot end where
clinker is discharged; the mid-point of
the kiln; and the cold end of the kiln
where raw material is fed).
EPA proposes that a ten-minute
average be used to control perturbations
in combustion chamber temperature and
that an hourly rolling average be used to
control average combustion chamber
temperature. The ten-minute average
would be established as the average of
the minimum ten-minute rolling average
for each run of the comprehensive
performance test. The hourly average
would be established as the average over
all runs.
v. Good Combustion: Maximum Flue
Gas Rate or Production Rate. Flue gas
flowrates in excess pf those that occur
during performance testing reduce the
time that combustion gases are exposed
to combustion chamber temperatures.
Thus, combustion efficiency can
decrease causing an increase in D/F
precursors.141 Accordingly, today's rule
would limit flue gas flowrate based on
levels that occur during the
comprehensive performance test.
For CKs and LWAKs, the rule would
allow the use of production rate as a
surrogate for flue gas flowrate. This is
the approach currently used for the BIF
rule for these devices, given that flue gas
flowrate correlates with production rate
(e.g., feeflrate of raw materials or rate of
production of clinker or aggregate).
However, production rate may not relate
well to flue gas flowrate in situations
where the moisture content of the feed
to the combustor changes dramatically.
Therefore, EPA invites comment on how
to address moisture content in feeds.
The gas flowrate or production rate
limit would be established as the
average of the maximum hourly rolling
average for each run of the
comprehensive performance test.
vi. Good Combustion: Batch Size,
Feeding Frequency, and Minimum
Oxygen. Some HWCs burn waste or
non-waste fuel in batches, such as metal
drums or plastic containers. Some
containerized waste can volatilize
rapidly, causing a momentary oxygen-
deficient condition that can result in an
increase in D/F precursors.142 To ensure
that D/F precursors are not increased
over levels that occur during the
comprehensive performance test, the
rule would establish site-specific limits
on maximum batch size, batch feeding
frequency, and minimum oxygen
concentration at the end of the
combustion chamber into which the
batch is fed, at the time the batch is
fed.143
This requirement would apply to all
HWCs that burn any waste or non-waste
fuel in batches (i.e., ram or equivalent
feed systems) or containers. For
example, incinerators that use a ram to
charge batches of hazardous or
nonhazardous waste would be subject to
these requirements. Cement kilns that
feed containers of fuel at mid-kiln or at
the "cold", raw material feed end would
also be subject to these requirements, as
would hazardous waste-burning cement
kilns that feed tires in batches.
The rule would provide a conditioned
exemption from the (site-specific)
oxygen limit, however, for cement kilns
that feed up to 1-gallon containers into
the "hot", clinker discharge of the kiln.
We do not believe that it is necessary to
control the oxygen content of
combustion gases when these containers
are fed into the hot end of the kiln given
that the oxygen demand from waste in
the containers would be insignificant
compared to the oxygen demand from
other (non-containerized) fuel burned at
this location. The frequency of firing the
containers would, however, be limited
to the rate occurring during the
performance test.
There would be no averaging period
associated with the limits on these
operating parameters. The maximum
batch size a facility could burn during
normal operations would be limited by
mass and would be established based on
the container or batch fired during the
test having the lowest mass. The
minimum batch feeding interval (i.e.,
the minimum period of time between
batch feedings) a facility could burn
141 We note that an increase in gas flow rate can
also adversely affect the performance of a D/F
emission control device (e.g., carbon injection,
catalytic oxidizer). Thus, gas flow rate is controlled
for this reason as well.
M2The requirements would apply when either
hazardous or non-hazardous waste fuels are batch
fed because the potential for oxygen-deficient
conditions and an increase in D/F precursors is
present irrespective of whether the material fed is
classified as a hazardous waste.
'•"EPA considered whether it would be practical
to establish a national minimum oxygen level for
all HWCs in this proposed rule and believes it is
not practical. A limit on minimum oxygen content
would have to be established on a case-specific
basis given that the minimum oxygen level
necessary for good combustion will vary from
source to source within a given source category, and
will vary within a given source over time as the
type or volume of waste or fuel varies. The Agency
invites comment on whether the final rule should
require a case-specific limit on minimum oxygen
content for all HWCs rather than as proposed for
only batch-fired HWCs. If so, the limits would be
established on a ten-minute and an hourly rolling
average as proposed for combustion chamber
temperature.
-------
17424
Federal Register / Vol. 61, No, 77 /Friday, April 19, 1996 / Proposed Rules
during normal operations would be
established as the longest interval of
time between batch feedings during the
comprehensive performance test. The
minimum oxygen content at which a
facility would charge a containerized
waste into the burner during normal
operations would be the highest
instantaneous oxygen level observed
when any batch was fed during the
comprehensive performance test.
EPA specifically invites comment on
whether the bases of these three
parameters are overly conservative.
Rather than basing maximum batch size
on the smallest container fed during the
comprehensive test, EPA could establish
maximum batch size based on the
average container mass. Feeding
frequency could be based on the average
time interval between batches during
the comprehensive test. Oxygen
concentration could be the average
oxygen level occurring during the test.
To address this issue, EPA needs to
know whether the proposed
requirements are overly conservative
and why, or conversely, whether the
options described in this paragraph are
not restrictive enough.
EPA specifically invites comment on
other approaches to establish limits for
these parameters, and whether (and
how) it would he necessary to limit
maximum volatility of the batch-fired
material.
vii. Dry PM Collection Device Inlet
Temperature. Formation of D/F
emissions on particulate matter
increases with increasing temperature.
Above 350°F and up to approximately
700°F, emissions of D/F can increase a
factor of 10 for every 125°F increase in
temperature.144 Consequently, today's
rule would limit temperature at the inlet
to a dry PM control device to the
maximum levels that occurred during
the comprehensive performance test.
It is proposed that a ten-minute
rolling average be used to control
perturbations in temperatures and that a
one-hour rolling average be used to
control the average temperature. The
ten-minute rolling average limit would
be established as the average of the
highest ten-minute average for each run.
The hourly average would be
established as the average of over all
runs.
viii. Carbon Injection. Facilities may '
use carbon injection to meet the D/F
standard. Today's rule would limit the
following carbon injection parameters:
minimum carbon injection rate;
""See Chapter 7.2 of "Draft Technical Support
Document for HWC MACT Standards, Volume IV:
Compliance with the Proposed MACT Standards",
February 1996.
minimum carrier fluid flowrate or
nozzle pressure drop, and adsorption
characteristics of the carbon.
A minimum carbon feedrate limit is
necessary to.ensure that facilities
maintain the same D/F removal
efficiency as was demonstrated during
the comprehensive performance test. It
is proposed that minimum carbon
injection rate be maintained on a ten-
minute and one-hour average. The ten-
minute average would be established as
the average of the minimum 10-minute
rolling average for each run, and the
one-hour average would be established
as the average over all runs.
A carrier fluid, gas or liquid, is
necessary to transport and inject the
carbon into the gas stream. EPA
proposes that either minimum carrier
gas flowrate or pressure drop across the
nozzle be maintained to ensure good
flow of the injected carbon into the flue
gas stream. It is proposed that either
limit be established on a 10-minute
rolling average and that the limit be
based on the carbon injection
manufacturers specifications.
Finally, to ensure that D/F removal
efficiency is maintained after the
performance test, carbon used after the
test must have the same or better
adsorption properties as carbon used
during the test. Thus, the rule would
require that facilities continue to use the
same brand and type of carbon that was
used during the comprehensive test.
The rule would allow a source to obtain
a waiver from this requirement from the
Director, however,-if the owner or
operator: (1) documents by data or
information key characteristics of
carbon which affect removal of D/F from
combustion gas; (2) documents by data
or information specification levels
corresponding to those characteristics;
and (3) complies with the specification.
ix. Carbon Bed. Some sources may
elect to use a carbon bed to control D/
F. Today's rule would limit the age of
the carbon and the adsorption
characteristics of the carbon to ensure
that D/F control is maintained.
Since carbon beds work by adsorbing
certain chemicals, e.g., dioxin and
mercury, and the carbon in the bed
becomes less effective as the active sites
for adsorption become occupied, an
appropriate control parameter for
carbon beds is the amount of time the
carbon in use. EPA is particularly
concerned about a facility's ability to
know when a carbon bed is spent, i.e.,
when enough active sites get occupied
to make the device inadequate for
removing dioxin or mercury, and
knowing how often carbon must be
replaced from the bed to ensure this
does not occur. This cannot be
determined during the initial
comprehensive performance test. For
that reason, the Agency proposes that
facilities follow the carbon bed
manufacturer's specifications for the
initial comprehensive performance test.
No carbon aging would be required
for this initial test. For confirmatory
tests, facilities would be required to
follow the normal change-out schedule
specified by the manufacturer. For
subsequent comprehensive tests, the
Agency proposes that the D/F test be
conducted at maximum carbon age, i.e.,
at the least frequent carbon change-out,
and that this age be maximum age
allowable under normal operation.
Alternately, the Agency could use
some form of a breakthrough calculation
and use this to assure compliance with
the D/F standard. A breakthrough
calculation would give a theoretical
minimum carbon change-out schedule
which the facility could use to ensure
that breakthrough, i.e., the dramatic
reduction in efficiency of the carbon bed
due to too make active sites being
occupied, does not happen. However a
breakthrough calculation can only be
done after experimentation determines
the relationship between incoming
adsorbed chemicals and the adsorption
rate of the carbon. The adsorption rate
of carbon can be determined
experimentally, but the speciation of
adsorbed chemicals in a flue gas stream
is site-specific and may vary greatly
within a given site over time. Therefore,
EPA proposes using this alternative only
for the initial comprehensive test, when
site data is not available and the carbon
bed is not aged. EPA believes that, for
subsequent comprehensive tests, the
proposed option is preferable, since it
provides for the setting of the minimum
carbon change-out on subsequent D/F
tests. EPA does not believe it is
appropriate to use breakthrough
calculations for the second and
subsequent comprehensive test(s) since
they do not take into account facility
specific characteristics, like the
concentration of adsorbed chemicals in
the flue gas. EPA invites comment on an
approach which would use
breakthrough calculations alone, to see
if it can become workable in another
form than the Agency has envisioned.
An issue that is difficult to address is
that carbon age is dependant not only
on time in service, but also the carbon
bed inlet concentration of substances
(e.g., metals, PM) which adsorb or
absorb onto the carbon. There may be
other factors that affect D/F removal
efficiency of the bed. The Agency
invites comment on how to address
these issues.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 /Proposed Rules
17425
Another issue is whether it is
necessary to control temperature at the
inlet to the carbon bed. EPA does not
believe this is necessary since facilities
will need a PM control device upstream
of a carbon bed and temperature at the
inlet to dry PM APCDs is proposed to
be, controlled. However, the
consequences of a temperature spike at
the carbon bed can be severe: a
temperature spike may cause adsorbed
D/F and Hg to de-adsorb and re-enter
the gas stream, resulting in a significant
amount of D/F and Hg being emitted at
the stack at once. For this reason, the
Agency invites comment on whether
controlling temperature at the inlet to a
carbon bed is necessary.
Finally, as the case with carbon
injection,, to ensure that D/F removal
efficiency is maintained after the
performance test, carbon used post-test
must have the same or better adsorption
properties as carbon used during the
test. Thus, the rule would require that
facilities continue to use the same brand
and type of carbon as was used during
the comprehensive test. The rule would
allow a source to obtain a waiver from
this requirement, however, as discussed
above.
x. Catalytic Oxidizer. Some facilities
may use a catalytic oxidizer to meet the
D/F standard. Catalytic oxidizers used
to control stack emissions are similar to
those used in automotive and industrial
applications. The flue gas passes over a
catalytic metals, such as palladium and
platinum, supported by an alumina
washcoat on some metal or ceramic
substrate. When the flue gas passes
through the catalyst, a reaction takes
place similar to combustion, converting
hydrocarbons to carbon monoxide, then
carbon dioxide. Catalytic oxidizers can
also be "poisoned" by lead and other
metals just as automotive and industrial
catalysts are.
The rule would require sources to
establish site-specific limits on the
following operating parameters for
catalytic oxidizers: minimum flue gas
temperature at the inlet of the catalyst,
maximum age in use, catalyst
replacement specifications, and
maximum flue gas temperature at the
inlet of the catalyst. The rule would
allow a waiver from these provisions if
the owner documents to the Director
that establishing limits on other
operating parameters would be more
appropriate to ensure that the D/F
destruction efficiency of the oxidizer is
maintained after the performance test.
The owner or operator would provide
such documentation, including how
limits on the alternative operating
parameters would be established and
appropriate averaging periods, and a
request for a waiver as part of the
notification to conduct the
comprehensive performance test and
draft test protocol. The Director would
grant the waiver in writing, if
warranted. .
Minimum flue gas temperature at the
inlet of the catalyst is necessary to
ensure that the catalyst is above light-off
temperature. Light-off temperature is
that minimum temperature at which the
catalyst is hot enough to catalyze the
reactions of hydrocarbons and carbon
monoxide. EPA proposes that minimum
flue gas temperature be maintained on
both a ten-minute and one-hour average.
The ten^minute average limit would be
established as the average of the
minimum ten-minute rolling average for
each run during the comprehensive
performance test. The hourly average
limit would be established as the
average hourly average over all runs.
Due to poisoning and general
degradation of the catalyst,
manufacturers often establish a
maximum time in-use for the catalyst.
EPA proposes that the manufacturer's
specification for maximum age be used
as maximum age of the catalyst.
When a catalyst is replaced, it must be
of the same design of the previous
catalyst to ensure that the replacement
catalyst will work as efficiently as the
previous one. Therefore, EPA proposes
that the following design parameters be
used in specifying replacement
catalysts: loading of catalytic metals;
space time; and monolith substrate
construction.
Catalytic metal loading is important
because, without sufficient catalytic
metal on the catalyst, it would not,
properly function. Also, some catalytic
metals are more efficient than others.
Therefore, EPA proposes that
replacement catalysts have at least the
same catalytic metal loading for each
catalytic metal as the catalyst used
during the comprehensive performance
test.
Space time, expressed in inverse
seconds (s~.1), is defined as the
maximum rated volumetric flow
through the catalyst divided by the
volume of the catalyst. This is important
because it is a measure of the gas flow
residence time and, hence, the amount
of time the flue gas is in the catalyst.
The longer the gas is in the catalyst, the
more time the catalyst has to cause
hydrocarbons and carbon monoxide to
react. It is proposed that'replacement
catalysts have at the same or lower
space time as the one used during'the
comprehensive performance test.
Substrate construction is also an
important parameter. Substrates for
industrial applications are typically
monoliths, made of rippled metal plates
banded together around the
circumference of the catalyst. Ceramic
monoliths and pellets can also be used. ;
Because of the many types of substrates,,
EPA proposes that the same materials of
construction, monolith or pellets and
metal or ceramic, be used as was used .
during the comprehensive performance
test. Monoliths also form a honeycomb
like structure when viewed from one
end. The pore density, i.e., number of
pores per square inch, is critical because
they must be small enough to ensure
intimate contact between the flue gas
and the catalyst, but large enough to
allow unrestricted flow through the
catalyst. Therefore, if a monolith
substrate is used, EPA proposes that the
same pore density as the one used
during the comprehensive performance
test. Finally, catalysts are supported by
a washcoat, typically alumina. EPA
proposes that replacement catalysts
have the same type and loading of
washcoat as was on the catalyst used
during the comprehensive performance
test.
Finally, EPA believes it is also
important to control maximum flue gas
temperature into the catalyst. This is
because sustained high flue gas
temperature can result in sintering of
the catalyst, degrading its performance.
The Agency proposes that maximum
flue gas temperature into the catalyst be
controlled and that it be a ten-minute
rolling average, based on manufacturer
specifications.
xi. D/F Inhibitor. Some facilities may
use a D/F inhibitor fe.g., sulfur) to meet
the D/F standard. In such cases, th^ rule
would establish a minimum inhibitor
feedrate. Limits would be established on
both a ten-minute and one-hour average.
The ten-minute average limit would be
established as the average of the
minimum ten-minute rolling average for
each run, and the one-hour average limit
would be established as the average over
all runs. See also the discussion in
section II.F.2 below for other
requirements to document compliance
with feedrate limits.
This minimum inhibitor feedrate ,
pertains to additives to feedstreams, not
naturally occurring inhibitors that may
be found in fossil fuels or hazardous .
waste. It is conceivable that a1 facility
would choose to burn high sulfur fuel
or waste specially during the
comprehensive test and switch back to
low sulfur fuels or waste after the test,
thus reducing D/F emissions during the
comprehensive test to levels that would
not be maintained after the test. EPA
invites comment on whether and how to
address this concern, including whether
it would be appropriate to establish
-------
17426 Federal Register /Vol. 61, No, 77 / Friday* April 19, 1996 / Proposed Rules
limits on the amount of naturally
occurring inhibitor, either during
performance testing or as an operating
limit., Comments and documentation are
also requested to help identify such
inhibitors.
As was the case with carbon used in
carbon injection and carbon beds, EPA
is concerned that facilities may use a
less effective, and presumably less'
expensive, D/F inhibitor during normal
operation than was used during the
comprehensive performance test. For
this reason, the rule would require that
facilities continue to use the same type
and brand of inhibitor as was used
during the comprehensive test. The rule
would,allow a source to obtain a waiver
from this requirement from the Director,
however, if the owner or operator: (1)
documents by data or information key
characteristics of the inhibitor which
inhibit formation of D/F; (2) documents
by data or information specification
levels corresponding to those
characteristics; and (3) complies with
the specification.
xii. Rapid Quench. Some facilities
may elect to use a rapid quench to lower
flue gas temperature to meet the D/F
standard. The rule would not establish
limits on operating parameters for rapid
quench systems because we believe that
a maximum dry PM control device
temperature is sufficient to ensure that
the quench was adequate. We note,
however, that a facility may use a rapid
quench for control of D/F emissions yet
hot have a dry PM control device. One
way to address this situation is to
require that a maximum flue gas
temperature be established at the stack.
EPA doubts, however, that there will
be any facilities which use a rapid
quench without a dry PM control
device. Consequently, we invite
comment on whether the final rule
should establish a maximum flue gas
temperature limit that would address
such apparently'hypothetical situations.
xiii. Consideration of Feed
Restrictions on Metals, Halogens, and
Dioxih Precursors. The rule would not
establish feedrate limits on metals,
halogens, or D/F precursors to ensure
compliance with the D/F standard.
Some research indicates that certain
metals, copper for instance, in the feed
may catalyze the formation of D/F.
However, this research is inconclusive
and there is not yet a consensus among
the research community that catalytic
metal in the feed necessarily causes
increased D/F emissions.145 Therefore,
EPA proposes not limiting the feed of
catalytic metals in the feed.
Research and common sense has also
indicated that the presence of halogens,
such as chlorine, in the feed may
contribute to the production of
halogenated D/F. While the presence of
chlorine in the feed is necessary for the
formation of chlorinated D/F, current
science seems to support the-view that
there is not a clear correlation between
the level of chlorine in the feed and the
level of dioxin in the flue gas. In other
words, increasing halogen feedrate
above de minimis levels does not appear
to cause increased emissions of
chlorinated D/F.146 Therefore, the rule
would not limit the amount of chlorine
fed to ensure compliance with the D/F
standard, particularly in light of the
suite of other compliance assurance
measures.
Nonetheless, we believe that it is
prudent to require that chlorine be fed'
at normal levels (or greater) during the
D/F comprehensive performance test.
This is because, while more chlorine '•
does not necessarily form more dioxin,
some chlorine is needed to form
chlorinated D/F. We invite comment on
how to ensure that normal levels of ,
chlorine are fed during the
comprehensive performance test. For
sources that do not elect to use a GEMS
for SVM, LVM, HC1 and C12 and, thus,
must maximize chlorine feedrate during
the test, this is not an issue. We believe -
that the vast majority of sources will be
in this situation. For sources that elect
to use such GEMS (assuming that multi-
metal and Cl2 GEMS become
commercially available),'defining
normal chlorine feedrates is an issue'.
Some arguments have been made that
the presence of organic dioxin
precursors in the feed would result in,
an increased level of D/F in the flue gas.
EPA has briefly examined certain
facilities which feed dioxin or known
organic dioxin precursors (e.g.,
chlorophenol and chlorobenzene) to
those which are known not' to feed
organic dioxin precursors. Although our
limited study suggests that no strong
correlation exists between the level of .,
dioxins or organic dioxin precursors in
the feed and D/F emissions, we do not
believe the issue has been sufficiently
examined in detail (indeed, other
evidence suggests that a correlation
might exist). EPA invites comment on ,
whether feed restrictions on D/F and
organic dioxin precursors are warranted
and, if so, whether this should be an .
operating parameter or a feed
requirement during the comprehensive
test (such as proposed for chlorine).
3. Mercury (Hg)
Table V.2.3 Summarizes the proposed
compliance monitoring requirements
and other options being considered for
Hg. See also proposed §63.1210(k).
TABLE V.2.3.—PROPOSED HG MONITORING REQUIREMENTS AND OTHER OPTIONS BEING CONSIDERED
Proposed Requirement
Option 1 • Elemental Hg
GEMS. ,
' " !
GEMS . ...
Surrogate GEMS
Max Flue Gas
Flowrate or Produc-
tion Rate.
Min Press Drop Wet
Scrubber.
Min Liq Feed Press
Wet Scrubber.
Min Liq pH
Compliance using
Total Hg or Multi-
metal GEMS.
Elemental Hg GEMS
Same
Pressure Drop Across
Scrubber.
Pressure
oH
Limits from
GEMS Std
Comp Test
Comp Test
Comp Test
Manuf Spec
Comp Test
Avg. period
10 hour.
10 hour .....
1 hour
10 min
1 hour
10 min.
10 min
Operating limit avg pd
basis
Avg over all runs.
Avg of Max 1 hour
RAs.
Avg of Min 10 min
RAs.
Avg over all runs.
Avg of Min 1 0 min
RAs.
14? See Chapter 7.2 of USEPA, "Draft technical
Support Document for HWC MACT Standards,
Volume IV: Compliance with the Proposed MACT
Standards", February 1996. •
"^See Chapter 7.3 of USEPA, "Draft Technical
Support Document for HWC MACT Standards,
Volume IV: Compliance with the Proposed MACT
Standards", February 1996. '. •' •
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17427
TABLE V.2.3.—PROPOSED HG MONITORING REQUIREMENTS AND OTHER OPTIONS BEING CONSIDERED—Continued
Option 2: No GEMS ....
Min Liq/Gas Ratio
Wet Scrubber.
Max Total Hg
Feedrate, all
streams.
Max Inlet Temp to
Dry PM APCD.
Min Carbon Injection
Rate.
Carbon Specs
Min Carrier Fluid
Flowrate or Nozzle.
Max Carbon Age
Max Flue Gas
Flowrate of Produc-
tion Rate.
Min Press Drop Wet
Scrubber.
Min Liq Feed Press
Wet Scrubber.
Min Liq pH, Wet
Scrubber.
Min Liq/Gas Ratio,
Wet Scrubber.
Compliance using
Scrubber Liquid and
Flue Gas Flowrate.
Feedstream Analysis
Temp
Feedrate CMS
Brand and Type
Same
Max Carbon
Flowrate CMS or Pro-
duction Rate.
Pressure Drop Across
Scrubber.
Pressure
pH
Scrubber Liquid and-
Flue Gas Flowrate.
Limits from
Comp Test
Comp Test
Comp Test
Comp Test
Comp Test
Manuf Spec
Initia
Conf Tests
Subsequent Comp
Tests.
Comp Test
Comp Test
Manuf Spec
Comp Test
Comp Test
Avg. period
1 hour
10 min
1 hour
12 hour
t
10 min
1 hour
10 min
1 hour
N/A
10 min
N/A
N/A
N/A
1 hour
10 min
1 hour
1 0 min
10 min
1 hour
10 min
1 hour
Operating limit avg pd
basis
Avg over all runs.
Avg of Min 1 0 min
RAs.
Avg over all runs.
Avg over all runs.
Avg of Max 1 0 min
RAs.
Avg over all runs.
Avg of Min 1 0 min
RAs.
Avg over all runs.
N/A.
N/A
Manuf Specs
Normal C Change-out
Schedule.
Max C Age is the age
during subsequent
Comp Tests.
Avg of Max 1 hour
RAs.
Avg of Min 1 0 min
RAs.
Avg over all runs.
Avg of Min 10 min
RAs.
Avg over all runs
Avg of Min 1 0 min
RAs.
Avg over all runs.
a. Evaluation of Monitoring Options.
Several types of GEMS exist or are
under development which measure Hg.
Therefore, the rule proposes use of a Hg
GEMS to document compliance with the
Hg standard.147
The rule would allow two alternative
GEMS approaches: the use of a multi-
metal GEMS or the use of a total Hg
GEMS. (In addition, we discuss below
our concerns with allowing the use of
an elemental Hg GEMS.) If a facility
elects to use a multi-metal (MM) GEMS
for compliance with the SVM and LVM
standards, the MM GEMS can be used
for compliance with the Hg standard as
well. See the discussion below on SVMs
and LVMs for discussion on MM GEMS.
If a facility elects not to use a MM
GEMS, the source may use a total Hg
GEMS.
'•"In February 1996, the Agency initiated a
demonstration program to determine whether Hg
(and PM) GEMS can comply with the performance
specifications proposed today. The demonstration
will also evaluate lone-term durability (e.g., 6
months or longer) of the GEMS. Results of the
demonstration will be made available for review
and comment prior to promulgation of the final
rule.
In case the final rule does not require
compliance with the Hg standard using
a GEMS, we also invite comment on
ensuring compliance by establishing
limits on operating parameters.
b. Total Mercury GEMS. The rale
would require use of a GEMS to monitor
Hg emissions (see below, small-on site
sources could obtain a waiver from the
GEMS requirement.) If a facility elects
not to use a MM GEMS for compliance
with all of the metals standards, EPA
recommends that facilities use a total Hg
GEMS.
An example of such a unit is a total
Hg GEMS made by the German company
Verewa and marketed in the US by
Euramark. The device has recently been
certified by TUV, a quasi-governmental
German agency charged with approving
compliance devices and methods. The
GEMS uses wet chemistry techniques
prior to an elemental Hg UV absorption
analyzer to convert all species of Hg into
elemental Hg. The analyzer then
determines the total Hg in the flue gas.
The performance specification for a
total Hg GEMS are proposed here as Part '
60, Appendix B, Performance
Specification 12. In addition, the
appendix to Part 63, Subpart EEE,
Quality Assurance for GEMS would
require quarterly testing of the analyzer
and relative accuracy testing of the total
system every 3 years (or 5 years for
small on-site facilities).
Also, EPA invites comments on
allowing small on-site sources (defined
in § 63.1208(b)(l)(ii) in the proposed
regulations) to obtain a waiver from the
requirement of installing Hg GEMS. If
the waiver is promulgated and granted
by the permitting authority, the facility
would demonstrate compliance with the
Hg standard by establishing operating
parameter limits described in subsection
d, "Alternative to a GEMS," below.
c. Elemental Mercury GEMS. EPA
invites comment on another approach to
continuously monitor Hg emissions, the
use of an elemental Hg GEMS. Although
the elemental Hg GEMS may be less
expensive than a total Hg GEMS, EPA
has several concerns with allowing the
use of an elemental Hg GEMS.
An elemental Hg GEMS does not
measure species other than elemental,
or metallic Hg. It does not measure Hg
-------
27428
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
salts such as mercuric chloride
Therefore, it would be necessary for the
facility to measure elemental Hg using
the GEMS and elemental and Hg salts
separately using manual methods
during the comprehensive performance
test.
Data from the comprehensive test
would be used to identify the elemental
Hg emission level at which the facility
is considered to be hi compliance with
the total Hg standard. However,
following the comprehensive test a
facility could have higher levels of
undetectable Hg salt emissions than
occurred during the comprehensive test.
This could happen in one of two ways:
the scrubber may not be working as
effectively; or the Hg and halogen feed
may have increased such that, at a fixed
scrubber efficiency, more Hg salts are
emitted as a result. Ensuring that the
scrubber efficiency is maintained at
performance test levels can be
accomplished using the parameters
described above. However, it is difficult
to determine whether the same amount
of Hg salts, relative to the amount of
total Hg, is being emitted. One could
, correlate Hg and halogen feed with
scrubber efficiency at various scrubber
conditions, but this would require many
data points and seems infeasible from a
monetary and technical standpoint.
Even if an approach can be developed,
the Agency is inclined to believe it
would require a lot of oversight to
ensure it is done properly.
If the issue of correlating total Hg
emissions to an elemental Hg GEMS can
be successfully addressed, establishing
the site-specific limit and the averaging
period for the elemental Hg standard
would then have to be addressed.
Facilities would be able to use the mean
of the results during the test, along with
a variability factor, as their site-specific
elemental Hg level. The averaging
period could be the time duration of
three runs of the comprehensive
performance test, but manual methods
tests do not end on the exact hour and
there may be more than one
comprehensive test with, likely,
different sampling periods. So, a
problem would arise as to what
averaging period to use.
For these reasons, EPA believes the
use of an elemental Hg GEMS is
infeasible to implement under self-
implemented MACT standards.
Nonetheless, if these issues can be
resolved, the final rule may allow some
use of an elemental Hg GEMS.
d. Alternative to a GEMS. If the final
rule does not require that Hg emissions
be continuously monitored, the rule
would ensure compliance with the Hg
standard by establishing limits on
operating parameters. Also if the
provision allowing small on-site
facilities (defined in § 63.1208(b)(l)(ii)
of the proposed regulations) to waive
the Hg GEMS requirement is
promulgated and such a facility elects
not to use an Hg GEMS, the facility
would have to establish these operating
parameter limits to document
compliance with the Hg standard. The
proposed operating limits are:
maximum Hg feedrate, Hg scrubber
operating parameters, maximum flue gas
feedrate, minimum carbon injection
rate, and carbon bed operating
parameters.
i. Maximum Hg Feedrates. Absent a
requirement to monitor Hg emissions
with a GEMS, the final rule would
establish a maximum Hg feedrate limit.
This is because the amount of Hg fed
into the combustor directly affects
emissions and the ability of control
equipment to remove Hg. This
maximum feedrate pertains to all feeds
into the combustor: hazardous waste,
raw materials, additives, and fossil
fuels. Feedrate sampling and analysis
protocols would be described in the
facility's waste analysis plan. The limit
would be based on a twelve-hour
average and established as twelve times
the hourly average feedrate during all
runs of the comprehensive performance
test. See also the discussion in section
II.F.2. below for other requirements to
document compliance with feedrate
limits.
As mentioned above in Subsection B,
this twelve-hour average is inconsistent
with the ten hour averaging period for
metals GEMS. GEMS should have longer
averaging periods than operating
parameters such as feedrates. Therefore,
EPA invites comment on whether the
averaging period for Hg feedrate should
be promulgated at six, instead of 12,
hours. EPA believes a six-hour
averaging period for Hg feedrate is
sufficiently conservative, relative to the
GEMS averaging period and achievable.
ii. Max Inlet Temp to Dry PM APCD.
High inlet temperatures to dry PM
APCDs can cause low recovery of Hg in
the APCD. This is because Hg volatility
increases with increasing temperature.
Therefore, absent a requirement to
monitor Hg emissions with a GEMS, the
final rule would control inlet
temperature to a dry PM APCD. Limits
would be based on both a 10-minute
and a one-hour average. The 10-minute
average would be the average of the
maximum PM APCD inlet temperatures
experienced during each compliance
test run and the one-hour average would
be the average over all runs.
iii. Carbon Injection. Some facilities
may need to use carbon injection as an
aftertreatment to comply with the Hg
standard. Absent a Hg GEMS
requirement, the final rule would
establish controls on the following
carbon injection operating parameters:
minimum carbon injection rate, carbon
specifications, and minimum carrier
flowrate or nozzle pressure drop. The
controls would be established under the
same approach as proposed for carbon
injection used for D/F control. See the
previous discussion.
iv. Carbon Bed. Rather than carbon
injection, some facilities may elect to
use a carbon bed to control Hg
emissions. Absent a requirement to
monitor Hg emissions with a GEMS, the
final rule would establish controls on
carbon bed operating parameters under
the same approach as proposed for
carbon beds used for D/F control. See
the previous discussion.
v. Maximum Flue Gas Flowrate or
Production Rate. As discussed above for
compliance with the D/F standard, an
increase in flue gas flowrate can
decrease collection efficiency of the
emission control device. Accordingly,
absent a requirement to monitor Hg
emissions continuously, the final rule
would limit flue gas flowrate or
production rate under the same
approach as proposed for D/F
compliance. See the previous
discussion.
vi. Wet Scrubber Parameters. The
efficiency of wet scrubbers directly
affects the removal of Hg salts from flue
gas. Key operating parameters would
include: maximum flue gas flowrate or
production rate, minimum pressure
drop across the wet scrubber, minimum
liquid feed pressure, minimum liquid
pH, and minimum liquid to gas ratio.
Refer to the section below on
compliance requirements for the HC1
and Cla standard for discussion on these
parameters. Absent a requirement to
monitor Hg emissions continuously, the
final rule would establish limits on
these parameters under the same
approach as proposed for compliance
with the HC1 and C12 standard.
4. Semivolatile Metals (SVM) and Low
Volatile Metals (LVM)
Table V.2.4 Summarizes the proposed
compliance monitoring requirements
and other options being considered. See
also proposed § 63.1210 (1) and (m).
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17429
TABLE V.2.4.—SUMMARY OF PROPOSED SVM AND LVM COMPLIANCE MONITORING REQUIREMENTS AND OTHER OPTIONS
BEING CONSIDERED
Proposed Option 1
(Facility Choice).
Proposed Option 2
(Facility Choice).
GEMS
Good PM Control
Max Inlet Temp to
Dry PM APCD.
Max Total SVM and
LVM Feedrates.
Max Pumpable LVM
Feedrate.
Max Chlorine
Feedrate.
Compliance using
Multi-metal GEMS
PM GEMS (see PM
for Others).
Same
Feedstream Analysis
Feedstream Analysis
Feedstream Analysis
Limit from
GEMS Std
Comp Test
Comp Test
Comp Test
Comp Test
Comp Test ....
Avg period
10 hour
10 min
1 hour
10 min
1 hour
12 hour
12 hour
12 hour
Operating limit avg pd
basis
Avg of Max 10 min
RAs.
Avg over all runs.
Avg of Max 10 min
RAs.
Avg over all runs
Avg over all runs.
Avg over all runs.
Avg over all runs
a. Evaluation of Monitoring Options.
EPA proposes two compliance options
for the SVM and LVM standards: use of
a multi-metal GEMS (MM GEMS) or
compliance with limits on operating
parameters. A facility would he allowed
to use either of these options to
demonstrate compliance. We are not
proposing to require the use of a GEMS
because a GEMS is not commercially
available for LVMs and SVMs at this
time, and the Agency is uncertain
whether a GEMS that could meet the
proposed performance specifications
discussed below would be available at
promulgation of the final rule.
b. Option 1: Use of a Multi-metal
GEMS to Document Compliance. EPA is
proposing to allow the use of a MM
GEMS for compliance with the Hg,
SVM, and LVM standards. If a facility
elects to use a MM GEMS, limits on
operating parameters would not be
required.M8
EPA is proposing to allow the use of
a MM GEMS (and may require the use
of MM GEMS if they would be
commercially available by the
promulgation date of the final rule)
because it is difficult to ensure
compliance with the emission standards
by limiting operating parameters.
Sampling and analysis of feedstreams to
monitor metals feedrate has drawbacks
in that representative sampling is
sometimes difficult and expensive to
achieve,149 and the available analytical
methods may not extract all metals from
some feedstreams (and thus metal
feedrates may be higher than indicated
by analysis). In addition, it is often
'*< Although a site-specific limit on PM would
also not bo required for compliance with the SVM
and LVM omission standards, it would be needed
to comply with the D/F standard.
'•"We noto that several cement and light-weight
aggregate kilns have been fined because of
inadequate fcedstream analysis plans.
difficult to use limits on operating
parameters of the metal emission
control device to ensure that collection
efficiency is maintained. It is also
difficult to ensure that the other major
factors that can affect metals emissions
are adequately addressed by operating
limits. For example, factors that affect
metal volatility and subsequently metals
emissions may include chlorine
feedrates, combustion chamber
temperature, and temperature at the
inlet of the emission control device.
Finally, the common process of spiking
metals during compliance testing to
ensure an adequate operating envelope
is expensive, potentially dangerous to
the testing crew that must handle the
toxic metals, and causes higher than
normal emission rates during
compliance testing. If a MM GEMS were
available, there would not be a need to
spike metals during compliance testing.
i. How to Address Metals that a GEMS
May Not Be Able to Measure. Several
MM GEMS are currently under
development, and not all of them will
be able to measure all metals in the
SVM (Pb and Cd) and LVM (As, Be, Cr,
and Sb) groupings. Clearly, a MM GEMS
cannot be used to document compliance
for a metal it cannot measure. For
metals a MM GEMS cannot measure, it
is proposed that facilities assume that
all of that metal fed is emitted at the
stack and that this metal feedrate be
used in calculating the emissions for the
metal group. Alternately, EPA could
decide that a MM GEMS which does not
measure all the metals could not be
used as GEMS for compliance with the
SVM and LVM standards. EPA invites
comment on this issue.
For example, x-ray fluorescence
analyzers do not measure Be. If a facility
chooses to use a MM GEMS which
employs an x-ray fluorescence analyzer,
it would take the MM GEMS results for
As, Cr, and Sb, and the mass feedrate for
Be (corrected to effluent concentrations
by dividing by the average gas flowrate)
and sum the four together. This would
constitute the LVM emissions for the
averaging period that would be used to
determine compliance.
ii. Performance Specifications for a
MM GEMS. The performance
specification for a MM GEMS is
proposed here as Part 60, Appendix B,
Performance Specification (PS) 10.
Lacking a commercially available MM
GEMS to test prior to developing the
performance specification created
unique challenges to developing a MM
GEMS PS. The Agency's approach to
developing the PS was to base
performance criteria as much as
possible on existing performance
specifications. The Agency also worked
closely with MM GEMS developers,
through the American Society of
Mechanical Engineers, to ensure that the
MM GEMS PS would be representative
of the performance of commercially
available devices. EPA specifically
invites comment on the performance
specification.
It is also proposed that special quality
assurance (QA) requirements also
pertain to MM GEMS. (See subsection
F.I. of this section for more information
on GEMS QA requirements.) We
propose that the owner/operator
perform a relative accuracy test audit
(RATA) on the MM GEMS at least once
every three years (five years for small
on-site facilities). The RATA compares
the output of the MM GEMS to the
reference method. For the purposes of
these source categories, the reference
method for stack metals determinations
is the current BIF Method 0012 (SW-
846 Method 0060). The QA
requirements also propose that an
absolute calibration audit (ACA) be
conducted in years the RATA is not
-------
27430 Federal Register / Vol. 61, No. 77 A Friday, April 19, 1996 / Proposed Rules
conducted. The AC A would involve
making nine measurements using an
NIST traceable calibration standard at
three levels for each metal the GEMS
measures. NIST traceable solutions of
metals are currently available which
challenge the analyzer device only. EPA
is currently developing the NIST
traceable metal standard which will
challenge the entire system, not just the
analyzer.
c. Option 2: Use of Limits on
Operating Parameters to Document
Compliance. If a source elects not to use
a MM GEMS (or a GEMS is not
commercially available), the rule would
require the source to establish a site-
specific PM limit and comply with
limits on metals feedrate, chlorine
feedrate, and maximum temperature at
the inlet to the PM control device. These
limits would be established during the
comprehensive performance test when
the source demonstrates compliance
with the emission limits by manual
stack sampling.
i. PM Limit. SVM and LVM (and
adsorbed D/F) are controlled by the PM
control device. To ensure that the
collection efficiency of the PM device is
maintained after the comprehensive
performance test, EPA is proposing to
require that a PM limit be established as
the lower of the level occurring during
the SVM, LVM, and D/F performance
testing or the MACT standard. For PM
monitoring requirements see section 7,
below.
ii. Maximum Inlet Temperature to Dry
PM APCDs. High inlet temperatures to
dry PM APCDs can cause low recovery
of metals in the APCD because at higher
temperatures a larger portion of some
metals will be in the vapor phase. (Dry
PM control devices do not control vapor
phase metals.) This happens because
metal volatility increases with
increasing temperature. Therefore, EPA
proposes that the inlet temperature to a
dry PM APCD be maintained at a level
no higher than that during the
comprehensive performance test.
The Agency proposes that maximum
inlet temperature to a dry PM APCD be
maintained on both a 10-minute and a
one-hour average. The 10-minute
average would be the average of the
maximum inlet temperatures
experienced during each compliance
test run and the 'one-hour average would
be the average over all runs.
iii. Maximum SVM and LVM Feedrate
Limits. Given the correlation between
feedrate and emission rate, the rule
would limit feedrate of SVM and LVM
to levels fed during the comprehensive
performance test. For LVM, feedrate
limits would be set on both pumpable
liquids and total feedstreams separately.
A separate limit is proposed for
pumpable feedstreams because metals
present in pumpable feedstreams may
partition between the combustion gas
and bottom ash (or kiln product) at a
higher rate than metals in nonpumpable
feedstreams.
For SVM, the feedrate limit would
apply to all feedstreams. Separate limits
would not be established for pumpable
versus total feedstreams. This is because
partitioning between the combustion gas
and bottom ash or product does not
appear to be affected by the physical
state of the feedstrearn. 15°
Sources would be required to perform
sampling and analysis of all feedstreams
(including hazardous waste, raw
materials, and other fuels and additives)
for SVM and LVM content to document
compliance with the feedrate limits. See
also the discussion in section II.F.2.
below for other requirements to
document compliance with feedrate
limits.
The rule would base the feedrate limit
for SVM and LVM on a twelve-hour
average basis. The limit would be
established as twelve times the average
hourly feedrate during the
comprehensive performance test. Also,
facilities would be required to record
not only the total feed at each
individual feed location for SVM and
LVM, but the total sum of the SVM feed
and the LVM feed at the various
locations.
As mentioned above in Subsection B,
this twelve-hour average is inconsistent
with the ten-hour averaging period for
metals GEMS. GEMS should have longer
averaging periods than operating
parameters such as feedrates. Therefore,
EPA invites comment on whether the
averaging period for all SVM and LVM
feedrates should be promulgated at six,
instead of 12, hours. EPA believes a six-
hour averaging period for all SVM and
LVM feedrates is sufficiently
conservative, relative to the GEMS
averaging period and achievable.
The grouping of metals by volatility
means that it is possible for one metal
within the volatility group to be used
during performance testing as a
surrogate for other metals in that
volatility group. For instance, As may be
used as a surrogate during the
comprehensive performance test for all
LVMs. Similarly, lead could be used as
a surrogate for Cd, the other SVM. In
addition, either SVM could be used as
a surrogate for any LVM. This will help
alleviate concerns facilities have voiced
150 See USEPA, "Draft Technical Support
Document for HWC MACT Standards, Volume IV:
Compliance with the Proposed MACT Standards",
February 1996.
regarding the need to spike each metal
during BIF certification of compliance
testing. Facilities would not need to
spike each metal to comply with today's
rule, but only one metal within the
group (or potentially one SVM for both
categories).
iv. Maximum Chlorine Feedrate. The
rule would establish a maximum
feedrate for total chlorine and chloride
based on the level fed during the
comprehensive performance test. A
limit on maximum chlorine feed is
necessary because most metals are more
volatile in the chlorinated form.
Although most of the volatilized SVM
and LVM will condense to particulate
form before entering the PM control
device, the metals condense in a fine
particulate fume that is more difficult
for most PM control devices to collect
than larger particulate.
The rule would require sampling and
analysis of each feedstream for total
chlorine and chloride to document
compliance with the feedrate limit for
total feedstreams. The maximum
feedrate would be based on a twelve-
hour average, and would be established
as twelve times the hourly average
feedrate during the comprehensive
performance test. Note also the
requirements for documenting
compliance with feedrate limits
discussed in section II.F.2.
Again, this twelve-hour average is
inconsistent with the one-hour
averaging period for HC1 and Cla GEMS.
GEMS should have longer averaging
periods than operating parameters such
as feedrates. Therefore, EPA invites
comment on whether the averaging
period for chlorine feedrate should be
promulgated at one, instead of 12,
hours. EPA believes a twelve-hour
averaging period for chlorine feedrate is
not be sufficiently conservative, relative
to the one-hour GEMS averaging period.
However, EPA also believes that a
shorter averaging period for feedrates
may be difficult for some facilities,
particularly those with diverse
feedstreams, to achieve routinely. For
this reason, the twelve-hour average is
proposed and comment is sought on the
one hour-average.
We note that if a facility uses a GEMS
for compliance with the Hg, SVM, LVM,
and HC1 and C12 standards, there would
be no need for the facility to establish
a total chlorine and chloride feedrate
limit.
v. Special Requirements for Cement
and Lightweight Aggregate Kilns that
Recycle Collected Particulate Matter.
Cement kilns and lightweight aggregate
kilns that recycle collected particulate
matter (which is primarily raw material
that is entrained in kiln gas) pose a
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17431
special problem to ensure compliance
with metals emission standards. These
sources (particularly cement kilns) feed
a variety of feedstocks which makes
feedstream analysis problematic. Also,
when these sources spike metals in
feedstreams for purposes of performance
testing, it may take several hours or days
to reach steady-state emissions.
Under the BIF rule, these sources
must comply with one of three
requirements: (1) Daily monitoring of
collected PM to ensure that metals
levels do not exceed limits that relate
concentration of the metal in the
collected PM to emitted PM; (2) daily
stack sampling for metals; or (3)
conditioning of the furnace system prior
to performance testing to ensure that
metals emissions are at equilibrium
with metals feedrates. See 56 FR 7176-
78 (February 21,1991), existing
§ 266.103(c)(6), and proposed
§ 63.1210(n). We propose to continue to
require that these sources comply with
one of the three BIF alternative
approaches for compliance with the
MACT metals standards.
We understand, however, that the
approach of daily monitoring collected
PM to document compliance with the
BIF metal standards (see Section 10 of
Appendix IX to Part 266, "Alternative
Methodology for Implementing Metals
Controls") is not currently being used
by any facility because it is too
complicated and burdensome. (The
methodology involves empirically
relating the concentration of each metal
in the emitted PM to the concentration
of the metal in collected PM (i.e., the
enrichment factor).) The Cement Kiln
Recycling Coalition (CKRC) has
suggested several revisions to the
methodologyISI including: (1) Reduced
testing frequency to establish and
periodically confirm the enrichment
factor; (2) assuming PM emissionsI52 are
at normal levels rather than maximum
allowable levels; (3) a less conservative
approach to estimate the enrichment
factor for nondetect metals in collected
PM (based on new sampling and
analysis techniques and improved
understanding of metals behavior); and
(4) allowing all kilns to comply with a
revised methodology, not just kilns that
recycle collected PM. (The Agency
believes the approach may, in fact, be
appropriate for any HWC and invites
comment on this matter.) hi addition,
CKRC raises several questions regarding
m Soo lottor from Craig Campbell, CKRC, to
Jnmcs Bcrlow, EPA, undated but received on
February 20,1996.
'"Note that PM emissions from CKs are
comprised primarily of raw material entrained in
the kiln off-gas. The material is known as cement
kiln dust (CKD).
the statistical foundations of the
methodology.
The Agency invites comment on
CKRC's recommendations to improve
the collected PM monitoring
methodology and on other approaches
to make the methodology a more
workable but effective compliance
approach in lieu of monitoring feedrates
of metals in feedstreams.
5. Carbon Monoxide (CO),
Hydrocarbons (HC), and Oxygen (Oa)
EPA is proposing that facilities
demonstrate compliance with the CO
and HC standards by using CEMS. See
proposed § 63.1210(p) and (q). EPA is
not proposing a standard for O2,153 but
all of the standards are based on
correction to 7 percent O2. Therefore,
EPA proposes facilities monitor Oz by
using a CEMS. Many HWCs are already
equipped with these monitors to comply
with the existing incinerator or BIF
regulations.
EPA proposes performance
specifications for CO and O2 CEMS in
Performance Specification 4B of
Appendix B, Part 60. EPA proposes a
total hydrocarbon (THC) CEMS
performance specifications based on the
use of a heated flame ionization detector
(i.e., heated FID). The HC PS will be
Performance Specification 8A contained
in Appendix B, Part 60. Both PSs are
similar to those currently used for BIFs.
The minor proposed changes are
discussed below.
a. Averaging Period for CO and HC
CEMS. The averaging period for CO and
HC CEMS is proposed to be a one-hour
rolling average. This is because this a
one-hour rolling average is the same
averaging period currently used in the
BIF rule. Changing the averaging period
would necessitate changing the
emission standard (see Part Four,
Section II) to maintain the same
stringency for the different averaging
period. EPA does not believe this is
warranted, so the one-hour rolling
average is proposed.
b. CO and HC CEMS Performance
Specifications. Performance
specifications for CO and Oz CEMS are
proposed here as Performance
Specification 4B. This performance
specification is essentially the same as
the specification for BIFs provided in
Appendix IX of Part 266. This
performance specification is the very
similar to existing Appendix B
Performance Specifications 3 (for 02)
and 4A (for CO). It references many of
153 Except that batch-fired HWCs would be
required to comply with a minimum combustion
chamber oxygen level prior to feeding a batch to
maintain compliance with the D/F standard.
the provisions of the two other
specifications. What the proposed
specification does do is describe how
the current BIF CEMS performance
specifications differ from performance
specifications 3 and 4A and prescribes
the BIF specifications in instances when
differences occur. EPA is proposing
specification 4B because it believes it is
important to "grandfather" in the
current performance specifications for
administrative and cost reasons.
Performance specification 4B does not
differ substantially from the current Part
60 specifications. Therefore, EPA invites
comment on whether to not propose
performance specification 4B and
instead rely on the existing
specifications 3 and 4A.
Also, performance specifications 3
and 4A (which performance
specification 4B refers to) requires a
Relative Accuracy Test Audit (RATA) be
performed on the CEMS. It also allows
for a waiver of the RATA requirement
if an acceptable substitute is used. The
Agency is currently moving away from
requiring RATAs for CEMS for which
cylinder gases are available. Cylinder
gases are available for both CO and O2,
so we invite comment on whether the
RATA requirements not be included in
performance specification 4B. EPA
would still require facilities to perform
quarterly absolute calibration audits
(ACAs) using calibration error (CE) test
procedures for these CEMS. EPA invites
comment on whether the RATA
requirement should not be promulgated
and whether just a quarterly ACA is
adequate without a RATA.
HC CEMS performance specifications
are proposed here as Performance
Specification 8A. It is identical to the
performance specification contained in
section 2.2 of Appendix IX of Part 266,
except the quality assurance section has
been deleted and placed in the
appendix to Subpart EEE, Part 63, to be
consistent with the Agency's approach
to Part 60 performance specifications.
There is an existing performance
specification, number 8, for a volatile
organic compound (VOC) CEMS.
Performance specification 8 does not
rely on heated sampling lines and
detector. A cold VOC monitor does not
measure less volatile hydrocarbons
which, due to heating, are measured by
a heated FID but not a cold VOC
monitor. (Heavy hydrocarbons would
condense out in the sampling line and
in the analyzer in a VOC CEMS and not
be measured as hydrocarbon emissions.
Therefore, a VOC CEMS measures a
subset of what a heated FID measures.)
Using the VOC performance
specification would be problematic
because the emission standard was
-------
17432 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
established using the results from
heated FIDs, not cold VOC GEMS. EPA
believes allowing compliance with a
GEMS that measures only a subset of the
pollutants represented by the standard
is inappropriate. For this reason, we
decided against proposing the use of
performance specification 8. EPA
believes it is appropriate to propose
performance specification 8A to
"grandfather" in the current
specifications and keep compliance
monitoring in agreement with how the
standard was derived.
One issue that has arisen during the
implementation of the BIF rule is that
the stated span values for the CO GEMS
may lead to high error in the facility's
calculated emission value. For instance,
a CK may analyze for CO emissions in
the bypass duct, and analyses in bypass
ducts can have very high oxygen
correction factors, on the order of 10. At
the low range CO span of 200 ppm with
an acceptable calibration drift of 3
percent, or 6 ppm, this means that error
in the standard due to calibration drift
would be 60 ppm if the oxygen
correction factor is ten. An absolute
calibration drift of 60 ppm is more than
half the CO standard of 100 ppm and
many believe this is unacceptable.
Therefore, EPA wishes to clarify the
ranges for GEMS, stating that the spans
for low and high ranges are expressed at
an oxygen correction factor of 1.
Facilities which normally operate at
oxygen correction factors more than 2
would have to use GEMS with spans
proportionately lower than the stated
values, relative to the oxygen correction
factor at the sampling point.
In the example above, where the
oxygen correction factor is 10, the
suggested value of the low range span
for the CO GEMS would be 200 divided
by 10, or 20 ppm. If the low CO range
is 20, the oxygen correction factor is 10,
and the calibration drift is 3 percent of
the span of the range, then the absolute
calibration drift would be 6 ppm.
Because the span value is a suggested
value, the facility could use a 25 ppm
span value to satisfy this requirement.
This modification is contained in the
GEMS Quality Assurance section of the
proposed rules and would apply to-the
other GEMS except the oxygen GEMS,
where the oxygen correction factor does
not apply. It is proposed that
corresponding changes be made to the
BIF rule as well.
An issue which also relates to the
oxygen correction factor is that it grows
exponentially as oxygen levels increase,
particularly at oxygen concentrations
above 15 to 17 percent. Some facilities
experience high oxygen correction
factors at times of start-up or shut-down
because combustion has just
commenced or is just completing and,
as a result, there is very high levels of
excess oxygen in the combustor. For this
reason, EPA invites comment on
whether it would be appropriate to cap
the oxygen correction factor at some
multiplier above the facility's normal
operating correction factor for a
specified period of time, on the order of
minutes, after a start-up or prior to a
shut-down.
6. Hydrochloric Acid (HC1) and
Chlorine Gas (C12)
Table V.2.5 summarizes the proposed
HCl/Cla compliance monitoring
requirements and other options being
considered. See also proposed
§63.1210(o).
TABLE V.2.5.—PROPOSED HCI/CI2 COMPLIANCE MONITORING REQUIREMENTS AND OTHER OPTIONS BEING CONSIDERED
(Facility Choice).
(Facility Choice).
Flowrate or Produc-
tion Rate.
Feedrate.
Scrubber.
sure, Wet Scrubber.
Min Liq pH Wet
Scrubber.
Wet Scrubber.
Feedrate, Dry
Scrubber.
Min Carrier Fluid
Flowrate or Nozzle
Pressure Drop, Dry
Scrubber.
Scrubber.
GEMS
Surrogate CEMS
Formation.
Compliance using
scrubber.
Pressure
DH
Scrubber liquid and
gas flowrates.
Carrier fluid flowrate
or pressure drop.
HCI and CI2 CEMS
HCI CEMS
TBD
Limits from
Comp Test
Comp Test
Comp Test
Manuf Spec
Comp Test . . ..
Comp Test ...
Comp Test
Manuf Spec
Comp Test
CEMS Std
Comp Test
Comp Test
Avg period
1 hour
12 hour
10 min
1 hour
10 min.
10 min
1 hour
10 min
1 hour
10 min
1 hour .
10 min.
N/A
2 hours.
2 hours
TBD
Operating limit avg pd
basis
Avg of Max 1 hour
RAs.
Avg over all runs.
Avg of Min 10 min
RAs.
Avg over all runs.
Avg Min 10 min RAs.
Avg over all runs.
Avg Min 10 min RAs.
Avg over all runs.
Avg of Min 1 0 min
RAs.
Avg over all runs.
Same brand and
type.
Avg over all runs.
TBD.
a. Evaluation of Monitoring Options.
The rule would allow sources the option
of using separate CEMS to monitor HCI
and Cl2 emissions or to comply with
limits on operating parameters.
HCI CEMS are commercially available
and have been used at permitted
municipal waste combustor sources and
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17433
some HWCs for many years. C12 GEMS
are currently being marketed by a
European manufacturer. Although the
Agency prefers the use of GEMS
whenever they are available for
compliance monitoring, we are
concerned that the use of GEMS to
monitor HC1 and C12 emissions may not
be cost-effective. This is because
facilities are likely to be required to
monitor chlorine feed to demonstrate
compliance with the SVM and LVM
standards anyway, given that a multi-
metal GEMS may not be commercially
available for some time.154 Accordingly,
the rule would allow, but not require,
the use of GEMS for HC1 and C12.
We note that we considered the
feasibility of allowing the use of an HC1
GEMS only, whereby the HC1 GEMS
would be used as a surrogate for the
HC1/C12 standard. As discussed below,
we determined, however, that this
approach would be more complicated,
more costly, have technical problems,
and/or provide less assurance of
compliance. We nonetheless invite
comment on whether the use of an HC1
GEMS as a compliance parameter for the
HC1 and C12 standard could be a
workable approach.
b. Compliance Using Limits on
Operating Parameters. If a source elects
not to use separate HC1 and C12 GEMS
to demonstrate compliance with the
HCl/Ck standard, the rule would
require the source to establish limits on
the following operating parameters
based on operations during the
comprehensive performance test to
ensure it maintains compliance with the
standard: maximum feedrate of total
chlorine and chloride from all
feedstreams, and limits on the acid gas
APCD operating parameters discussed
below.
i. Maximum Flue Gas Flowrate or
Production Rate. If flue gas flowrates
exceed those during the comprehensive
performance test, the HC1/C12 collection
efficiency of the control device may not
be maintained which may result in
emissions that exceed the standard.
Therefore, EPA proposes that maximum
flue gas flowrate be controlled to levels
that are no higher than those during the
performance test. Alternatively, CKs and
LWAKs may establish a maximum
production rate (e.g., raw material
feedrate or clinker or aggregate
production rate) in lieu of a maximum
gas flowrate given that production rate
directly relates to flue gas flowrate. The
limit would be based on a one-hour
'"If wa determine that multi-metal GEMS are
commercially available at promulgation and require
their USD in the final rule, we may also require the
uso of GEMS to monitor HC1 and C12 emissions.
average and be established as the
average of the maximum hourly rolling
average for each run of the
comprehensive performance test.
ii. Maximum Total Chlorine or
Chloride Feedrate. The rule would limit
the amount of total chlorine or chloride
fed in all feedstreams to levels that were
fed during the comprehensive
performance test demonstrating
compliance with the HC1/C12 standard.
Sources would be required to perform
sampling and analysis of each
feedstream for total chlorine and
chloride content to document
compliance with the feedrate limit for
total feedstreams. See also the
discussion in section II.F.2 for other
requirements to document compliance
with feedstream limits.
The total chlorine and chloride
feedrate limit would be averaged over a
twelve-hour period and would be
established as twelve times the hourly
feedrate during the comprehensive
performance test.
We again note that there is an
inconsistency between this twelve-hour
feedrate average and the proposed one-
hour averaging period for HC1 and C12
GEMS. EPA invites comment on
whether the averaging period for
chlorine feed should be promulgated at
one, instead of twelve, hours.
Note that if a facility uses a CEMS for
compliance with the HC1 and C12, Hg,
SVM, and LVM standards, no chlorine
feed monitoring would be required.
iii. Wet Scrubber Parameters. Wet
scrubbers can be used to control HC1
and C12 emissions. To ensure that the
control efficiency of a wet scrubber is
maintained at levels achieved during,
the comprehensive performance test, the
rule would require sources to establish
limits on the following operating
parameters: pressure drop across the •
scrubber; liquid feed pressure; liquid
(blowdown) pH; and liquid to gas flow
ratio.
Pressure drop across a wet scrubber is
an important parameter because it is an
indicator of good mixing of the two
fluids, the scrubber liquid and the flue
gas. A low pressure drop would indicate
poor mixing and, hence, poor efficiency.
A high pressure drop would indicate
good removal efficiency. Therefore, EPA
proposes that the pressure drop across
the scrubber be limited to the minimum
level during the comprehensive
performance test. Limits would be based
on both a ten-minute and a one-hour
average. The ten-minute average limit
would be established as the average of
the lowest ten-minute rolling average for
each run, and the hourly average limit
would be established as the average over
all runs. . •..',.
Scrubber liquid feed pressure is
important because it directly relates to
the amount of scrubber liquid pumped
into the scrubber and is easier to
measure than scrubber liquid flow
directly. The more scrubber liquid
pumped into the scrubber, the better the
removal efficiency. If liquid flow were
to decrease, the removal efficiency
would also decrease. EPA proposes that,
minimum liquid feed pressure be
maintained on a ten-minute average and
that the limit be the minimum value
established by the scrubber
manufacturer.
The pH of the scrubber liquid is also
important because, at low pH, the .
scrubber solution is more acidic and
removal efficiency of HC1 decreases. We
propose that the pH be determined from
the blowdown liquid. This is because it
is the best indicator of scrubber
efficiency by measuring pH of scrubber
liquid. EPA proposes that minimum pH
of the scrubber water be controlled on
both a ten-minute and a one-hour
average. The ten-minute average limit
would be established as the average of
the lowest ten-minute rolling average for
each run, and the hourly average limit
would be the average over all runs.
EPA solicits comment on whether the
alkaline reagent (such as lime)
concentration in the scrubber should be
a control parameter for alkaline wet-
scrubbers. This parameter is closely
related to the just mentioned pH since'
the concentration of alkaline reagent in
the scrubber will keep the scrubber
liquid pH high. EPA believes this
parameter is important because the
alkaline reagent is what removes Cl2
and, to a lesser extent, HC1 from the flue
gas. pH is a secondary indicator of this
parameter. EPA's concern is alkaline
reagent concentrations can be low,
enough to lower the efficiency of Wet
scrubbers yet buffer the scrubber liquid
enough to maintain pH. However, the
concentration of alkaline reagent in the
scrubber liquid can not be continuously
monitored as easily as pH. We invite
comment on whether the concentration
of alkaline reagent in the scrubber liquid
should be a control parameter for wet
scrubbers, whether this parameter
should be in addition to or in lieu of the
pH parameter, and what averaging '
period(s) such a parameter should have.
In addition, EPA invites comment on
whether a ten-minute average is
appropriate for pH (and/or alkaline
reagent concentration). Some facilities
may not automate their wet scrubbers to
add scrubbing solutions as needed to
maintain scrubber efficiency. Such
facilities make up batches of virgin
scrubber solution and add it to the
scrubber liquid. In this case, it might be
-------
37434 Federal Register / Vol. 61, No. 77 /Friday, April 19, 1996 / Proposed Rules
more appropriate to establish a
parameter ensuring that hatches of new
scrubber solution is added to the wet ,
scrubber prior to the scrubber liquid pH
(and/or possibly alkaline reagent)
reaching a certain level.
Liquid to gas flow ratio is another
important wet scrubber parameter. A
high liquid to gas flow ratio indicates
good scrubber removal, while a low -
liquid to gas flow ratio indicates less
efficient removal. EPA proposes that the
minimum scrubber liquid to flue gas
flow ratio be controlled on both a ten-
minute and a one-hour average. The ten-
minute average limit would be
established as the average of the lowest
ten-minute rolling average for each run,
and the hourly average limit would be
established as the average over all runs.
iv. Dry Scrubber Parameters. A dry
scrubber removes HC1 from the flue gas
by adsorbing the HC1 onto some sorbent,
normally an alkaline substance like
limestone. To ensure that the collection
efficiency of the scrubber is maintained
at comprehensive performance test
levels, the rule would require sources to
establish limits,on the following
operating parameters: sorbent feedrate;
carrier fluid flowrate or nozzle pressure
drop; and sorbent specifications.
Sorbent feedrate is important because,
when more sorbent is fed into the dry
scrubber, removal efficiency for HC1 and
C12 will increase.155 Conversely, lower
sorbent feedrates tend to cause removal
efficiency to decrease. Therefore, EPA
proposes that the minimum sorbent
feedrate into the .dry scrubber be
controlled,on both a ten-minute and a
one-hour rolling average. The tenT
minute average limit would be ,
established-as the average of the lowest
ten-minute rolling average for each run,
and the hourly average limit, would be
established as the average over all runs.
Carrier fluid is some liquid or gas
(normally air or water) which transports
the sorbent into the dry scrubber.
Without proper carrier flow to the dry
scrubber the s.orbent flow into the dry
scrubber will decrease, and efficiency
will also decrease. Nozzle pressure drop
is also an indicator of carrier gas flow
into the scrubber. At a relatively high
pressure drop, more sorbent is carried to
the dry scrubber. At lower pressure
drop, less sorbent is carried to the
scrubber. Therefore, the rule would
require that carrier fluid flowrate or
nozzle pressure: drop be maintained to
the minimum levels occurring during
J55EPA notes that sorbent to a dry scrubber
should be fed in excess of the stoichiometric
requirements for neutralizing the anion component
in the flue gas. Lower concentration of sorbent,
even above stoichiometric requirements, would
limit the removal of acid gasses.
the comprehensive performance test.
Limits would be established on both a
ten-minute and a one-hour rolling
average. The ten-minute average limit
would be established as the average of
the lowest ten-minute rolling average for
each run, and the hourly average limit
would be established as the average over
allruns. -'.;••"
As was the case with maintaining the
quality of carbon used in carbon
injection and carbon bed systems for
control of D/F and Hg, the rule would
require that the quality of sorbent be
maintained after the comprehensive
performance test. Therefore, the rule
would require sources to continue to
use the same sorbent brand and type as
they used during the comprehensive
performance test. The rule would allow
a source to obtain a waiver from this
requirement from the Director, however,
if the owner or operator: (1) documents
by data or information key
characteristics of the sorbent which
controls HC1 and Cl2>r(2) documents by
data or information specification levels
corresponding to those characteristics;
and (3) complies with the specification.
As was the case for pH in wet
scrubbers, EPA invites comment on
whether a ten-minute average is
appropriate for sorbent feedrate. Some
facilities may not automate their dry
scrubbers to add sorbent solutions as
needed to maintain scrubber efficiency.
Such facilities make up batches of virgin
sorbent solution and add it to a dry
scrubber feed tank containing the
sorbent. In this case, it might be more
appropriate to establish a parameter
ensuring that* batches of new scrubber
sorbent is added to the dry scrubber
prior to the Sorbent concentration in the
dry scrubber reaching a certain level.
c. Compliance Using Separate HC1
and C12 GEMS. The rule would allow
sources to use separate HC1 and C12
GEMS to demonstrate compliance with
the HC1/C12 standard. This option would
allow for the direct measurement of the
standard, at the top of the monitoring
hierarchy, but does so at a higher cost
relative to the previous option of
compliance with limits on operating
parameters. EPA' seeks comment on
whether the use of separate HC1 and Cla
GEMS is in fact cost-effective and
should be required in the final rule in
lieu of allowing compliance with
operating limits.
Under this option, compliance would
be demonstrated by measuring HC1
emissions (in ppmv) with the HC1 GEMS
and measuring C\2 emissions (in ppmv)
with a Cla monitor! Since the HC1 and
C12 standard is based on equivalents of
HC1, the ppmv emissions of C12 must be
multiplied by two and added to the HC1
emissions to determine the combined ••
emission level. If this result is lower
than the emission standard, then the , ;..
facility is in compliance with the HG1/
Ck standard.
i. HC1 GEMS. HC1 GEMS are proven,
technologies, available worldwide, and
are currently required in the permits of-
many MWCs. Several HWCs also use ;
HC1 GEMS. HC1 GEMS are not ... .
expensive; the purchase cost are
$12,000 to $55,000.156 ." :
Performance specifications for a HC1
GEMS are proposed today as •
Performance Specification 13 of
Appendix B, Part 60. The proposed
appendix to Part 63, Subpart EEE, also .,
proposes certain RATA and ACA
requirements.
ii: C12 GEMS. Cl2-specific GEMS are
currently being marketed by Opsis, a
European GEMS manufacturer. These;
devices have been certified for use in
Germany and can also be used to .
monitor for HC1, CO, NOX, SOX, and
NH.3. This device would likely be a cost-
effective option for new facilities or -
existing facilities purchasing a,Suite of
new GEMS.
Performance specifications for C12, ,
analyzers are proposed here as
Performance Specification 14 of Part 60,
Appendix B. The proposed appendix to
Part 63, Subpart EEE, also proposes
certain RATA and ACA requirements.
d. Consideration of Using an HC1 •
GEMS Only. EPA requests comment on
whether the use solely of an HC1
monitor for compliance with the HC1/
Cl2 standard could be workable. If so, .
this approach could be allowed as an;
option in the final rule. , •
This approach would provide direct
monitoring of the HC1 portion of the
standard and act as a surrogate monitor
for the C12 portion. However, EPA is
concerned that.poor correlation between
HC1 and C12 emissions may result in HC1
being a poor surrogate for C12. For an
HC1 GEMS alone to be a feasible
Surrogate monitor for the HC1/C12 .
standard, this and other issues
discussed below must be addressed. •
C12 and HC1 form a post-combustion
equilibrium. At temperatures above
1000°F the equilibrium is quite stable >
and correlation is good. At lower
temperatures, though, formation of Cl2
is favored over HC1 and the equilibrium
no longer holds. All HWCs experience
temperatures lower than 1000I?F, so the
HC1/C12 equilibrium does not hold. The
formation of C12 under these , •;
circumstances is dependent on a
156See Chapter 2.6 of USEPA, "Draft Technical
Support Document for HWC MACT Standards, v •,
Volume IV: Compliance with the Proposed MACT
Standards", February 1996. ' ••.. >,. ,--,' •';':i;.'!;. .••,,'.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17435
number of site-specific conditions, such
as the post-combustion temperature
profile and hence the rate of conversion
to Cl2, and residence time from the
point where CU formation is favored to
the stack. In fact, these conditions may
vary at any given facility depending on
the circumstances at any time after
combustion. Given that HC1 appears to
be a poor indicator of CU emissions,
direct measurement of Clz is desired.
If this issue can be adequately
addressed, the use of only a HC1 GEMS
to demonstrate compliance with the
standard would involve determining a
site-specific HC1 limit representative of
the combined HCl/Cla emissions. This
would involve a comprehensive
performance test at maximum chlorine
feed and under conditions which are
worst-case for CLz formation and
emissions and optimal for HC1 removal.
The resulting HC1 level would become
the site-specific limit to demonstrate
compliance with the HC1/C12 standard.
Limits on operating conditions would
also be necessary to ensure that the ratio
of Cla to HC1 emissions is not higher
than experienced during the
comprehensive performance test, and
that HC1 control equipment is not
operated more efficiently (note
emphasis) after the performance test.
Otherwise, the HC1 emissions during
normal operations may under-predict
combined HC1 and Ck emissions.
7. Particulate Matter (PM)
As discussed above in the sections on
operating limits for compliance with the
D/F, SVM, and LVM standards, a PM
limit would be established as the lower
of either the levels that occurred during
the comprehensive performance test to
demonstrate compliance with the D/F,
SVM, and LVM emission standards (as
a compliance parameter for those
standards) or the national PM standard.
Table V.2.6 below summarizes the
proposed monitoring requirements and
options being considered.
TABLE V.2.6.— PROPOSED PM MONITORING REQUIREMENTS AND OTHER OPTIONS BEING CONSIDERED
Proposed Requirement
Option: Feedstream
and Operating Pa-
rameter Limits.
GEMS
Max Flue Gas
Flowrate or Produc-
tion Rate.
Max Ash Feedrate ....
Min Press Drop, Wet
Scrubber including
Ionizing Wet Scrub-
ber.
Min Scrubber Feed
Press, Wet Scrub-
ber including Ioniz-
ing Wet Scrubber.
Min Slowdown or
Max Solid Content
in Liq, Wet Scrub-
ber including Ioniz-
ing Wet Scrubber.
Min Liq/Gas Ratio,
Wet Scrubber in-
cluding Ionizing
Wet Scrubber.
Min Pressure Drop,
Fabric Filter.
Min Power Input
Compliance using
PM GEMS
Same
Feedstream Analysis
Press drop across
scrubber.
Pressure
Liquid Flowrate or
Solid Content.
Scrubber Liquid and
Gas Flowrates.
Pressure Drop Across
Fabric Filter.
Voltage
Limits from
GEMS Std
D/F or SVM/LVM
Comp Test.
Comp Test
Comp Test
Comp Test
Manuf Specs
Comp Test
Comp Test
Comp Test
Comp
Avg. period
2 hours
10 Min
1 hour
1 hour
12 hour
10 min
1 hour
10 min
10 min
1 hour
10 min
1 hour
10 min
1 hour
10 min
1 hour
Operating limit avg.
pd basis
Lowest Avg Min 1 0
min RAs.
Lowest Avg over all
runs.
Avg of Max 1 hour
RAs.
Avg over all runs.
Avg of Min 1 0 min
RAs.
Avg over all runs
N/A.
Avg of Min/Max 1 0
min RAs.
Avg over all runs
Avg Min 10 min RAs
Avg over all runs
Avg Min 10 min RAs
Avg over all runs.
Avg Min 10
Avg over all runs.
a. Evaluation of Monitoring Options.
Continuous PM GEMS are commercially
available and installed on stacks
worldwide. EPA proposes that facilities
maintain continuous compliance with
the PM standard through the use of a
PM GEMS. PM GEMS are installed for
compliance purposes in the European
Union (EU) with the EU hazardous
waste combustor PM standard of 13 mg/
dscm. Germany has been in the forefront
in the development, certification, and
application of PM GEMS.
i. Evaluation of PM GEMS feasibility
and use. EPA in the past has relied on
opacity monitors to indicate compliance
with a PM standard. Opacity GEMS
used in accordance with performance
specification 1 have been a valid tool to
indicate PM APCD failures and the
necessity for corrective action as a
result. However, opacity monitors are
not, relatively speaking, very sensitive.
They are typically useful down to about
45 mg/dscm. Today's proposed
regulation will limit PM emissions to 69
mg/dscm. Opacity monitors would not
be sufficient because to maintain
compliance with 69 mg/dscm, facilities
would generally need to operate around
35 mg/dscm. Thus, emissions will
typically be below the detection limit of
opacity monitors most of the time.
While normal emission levels below the
detection limits of GEMS are acceptable,
facilities often desire the detection limit
to be below one-tenth of the emission
limit, or 7 mg/dscm for the proposed
standard. This gives one sufficient
-------
17436
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
warning of how emissions are changing
before the emission limit is approached,
and allows the facility, based on GEMS
readings, to change operations as
necessary to be in compliance with the
applicable standard. EPA has relied on
opacity GEMS because there has not
been available an acceptable
quantitative monitor for continuous
mass PM emissions. Opacity GEMS
standards are established at a given
percent opacity limit (generally 5-10
percent) over a 6-mimite averaging
period and, as stated, cannot distinguish
particulate concentrations below 45 mg/
dscm. In other words, opacity GEMS as
they are currently used can be used to
ensure PM APCD efficiency but not to
determine mass emissions in real time.
If possible, EPA desires a quantitative,
continuous measure of PM mass
concentration rather than opacity. EPA
has recently determined that GEMS do
exist that do this: beta gauges and light
scattering based GEMS. These GEMS
rely on calibration of the device to
manual gravimetric measurements.
Therefore, EPA is proposing use of
GEMS based on the availability of these
newer technology PM GEMS and a
related PM GEMS Performance
Specification for monitoring PM mass
concentration. This PS does not specify
the type of GEMS used and allows the
use of opacity monitors, which can also
be calibrated to relate opacity to mass
concentration. However, opacity is more
sensitive to PM size distribution and
physical properties, and has high
detection limitations relative to the
newer PM GEMS. As a result the
calibration will be less stable for an
opacity GEMS calibrated according to
the proposed performance specification
than one of the newer technology
instruments.
EPA believes that mass emission
monitoring is feasible, and opacity
monitoring has borderline sensitivity
relative to today's proposed PM
emission limit. The newer technology
PM GEMS can give a real-time
quantitative measure of PM mass
emissions while opacity GEMS cannot.
From a cost standpoint opacity
monitoring is no less expensive than the
alternative proposed here. As a result,
EPA proposes to require mass emission
monitoring rather than opacity
monitoring.
The German approach to using GEMS
for PM compliance monitoring is based
on the application of a practical
engineering philosophy. PM GEMS are
used despite the known sensitivities to
various factors such as particle
composition and size distribution since
these devices are designed to minimize
the impacts of these changes on the
accurate measure of PM mass ,
concentrations. The German experience
on PM GEMS is that at controlled
sources, i.e., those with low loading or
equipped with PM control devices such
as baghouses or ESPs, these sensitivities
are not as important as they are at
facilities with no control or high and/or
highly varying grain loadings. The
Germans have found that PM GEMS can
be calibrated to manual methods to
achieve a statistically reliable and
enforceable calibration curve at
controlled sources.157
At periods when the particle
composition and size changes
dramatically, the PM GEMS calibration
is not valid. However, this occurs when
fuel is changed or the PM control device
fails and causes very high grain loadings
to occur. To account for the PM GEMS'
sensitivity to fuel type, the Germans
mandate a new calibration be made
whenever the fuel is changed. During
times of high grain loading the PM
GEMS cannot accurately determine how
high the PM emissions were. But at
controlled devices, this only occurs
when the PM control device fails and/
or otherwise exceeds the PM standard.
Therefore, PM GEMS remain a reliable
indicator of compliance with a PM
standard.
In Germany, calibration of the PM
GEMS defines a statistically derived
site-specific calibration of the PM
GEMS' response to various PM loadings.
This is done by installing a plate in lieu
of a bag in the baghouse or by varying
the ESP voltage to allow various grain
loadings to flow through the control
device to the stack. The PM GEMS and
manual methods are run simultaneously
at various PM loadings to determine
emissions. These PM GEMS outputs and
manual methods results are used to
statistically define the calibration curve
for the PM GEMS.
EPA has tested several of these
devices at a hazardous waste incinerator
and a cement kiln and has found that
PM GEMS maintain calibration, even in
a water saturated flue gas.
ii. Types of PM GEMS available. The
many types of PM GEMS fall into three
broad categories: accumulated mass,
impaction, and light scattering.
For accumulated mass PM GEMS,
stack gas is extracted isokinetically and
particles are deposited on a sensing
surface for mass measurement. Two
types of accumulated mass devices are
p-radiation attenuators, commonly
referred to as "P-gauge" devices, and
loaded oscillators. EPA has tested a
stack-type p-gauge but testing was
inconclusive.158 EPA knows of no
available stack-type loaded oscillator
device.
For impaction devices, particles
impact upon a sensor surface due to the
inertia imparted by the approaching gas
stream. Two types of impaction PM
GEMS are contact electrification,
commonly referred to as "triboelectric",
and acoustic energy. Stack-type
triboelectric devices are commercially
available and in widespread use in
France. However, EPA has concern
about triboelectric PM GEMS since the
physical property of PM which they
work on, contact electrification, can
vary the most from particle to particle
even at controlled sources. For this
reason, facilities should be aware that
triboelectric PM GEMS may not be
quantitative enough to be used for
compliance with the PM standard.
Acoustic energy PM GEMS are not in
widespread use.
Light scattering GEMS are preferred in
Germany and are believed to be the PM
GEMS most suitable for making
measurements at low particulate levels
typical of a well controlled source. Light
scattering PM GEMS operate by sending
a light beam across a path and
measuring the light reflected back to a
sensor at some angle from the source
light. Several hundred of these devices
have been certified for stack-use in the
EU. EPA has also tested a time-
dependant optical transmission device.
Under certain circumstances, it can give
results comparable to those of the light
scattering device.
To be in compliance with the PM
limit, facilities would comply with the
performance specifications and
operating practices for the GEMS
proposed here. If a PM GEMS is used at
a facility, no feedstream or operating
parameter limits will be necessary to
document compliance with the PM
limit. If a PM GEMS is not used,
compliance with limits on feedstream
and operating parameters will be
necessary.
iii. Control of PM Emissions. We are
proposing to use a PM GEMS as a
compliance parameter to ensure: (1)
compliance with the national MACT PM
standard; and (2) that the collection
efficiency of the PM control device is
maintained at performance test levels
achieved when documenting
compliance with the SVM, LVM, and D/
F standards. Thus, it is necessary to
157 See Chapter 2.1 of USEPA, "Draft Technical
Support Document for HWC MACT Standards,
Volume IV: Compliance with the Proposed MACT
Standards", February 1996.
is8 See Chapter 2.1 of USEPA, "Draft Technical
Support Document for HWC MACT Standards,
Volume IV: Compliance with the Proposed MACT
Standards", February 1996.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17437
establish the PM limit as the lower of
the level that occurs during the SVM,
LVM, and D/F performance tests or the
MACT standard. This is because a
source could be operating well below
the national PM standard during the
performance test and, after the test,
operate the PM control device at lower
collection efficiency (e.g., to reduce
operating costs, or because of reduced
efficiency from "wear and tear"). In this
case, the source could continue to be in
compliance with the national PM
standard, yet exceed the D/F, LVM, and
SVM emission limits because of
increased emissions of adsorbed D/F,
LVM, and SVM.
To ensure that the collection
efficiency is maintained while meeting
the site-specific PM limit, the rule
would require that feedstocks with
normal levels of ash, i.e., those levels
which the facility routinely experiences
during normal operations, be fed during
the performance test. This would
preclude a source from artificially
increasing the PM loading during the
performance test using high ash
feedstocks to obtain a high site-specific
PM limit. If this were the case, the
source could meet the PM limit during
normal operations when feeding
feedstocks with normal ash content
while operating the PM control device
under less efficient conditions. This
could result in an increase in emissions
of metals and D/F adsorbed onto PM.
We invite comments on how to ensure
that feedstocks with normal ash content
are fed during the comprehensive
performance test.
The comprehensive performance tests
would be conducted as follows. During
the D/F, SVM, and LVM comprehensive
performance tests, the facility would
make manual measurements of D/F and
metals and GEMS measurements of PM.
Emissions of PM would be limited to
the national standard of 69 mg/dscm
during the tests. Following the tests the
facility would establish two site-specific
limits for PM: a ten-minute limit to
control perturbations and a one-hour
limit to control average emissions. The
ten-minute average would be based on
the highest ten-minute rolling averages
occurring during each comprehensive
test. The hourly average would be the
average of all one-minute averages
occurring during each comprehensive
test. (Note that, if the facility were to
perform separate D/F and metals tests,
the lowest of the two PM averages
would be the applicable PM limit.)
The facility need not determine or
record two-hour averages to document
compliance with the MACT PM
standard during normal operation, only
during the comprehensive test. Since
the one-hour average is the average of
all one-minute averages during the
comprehensive performance test and the
time duration of the test is longer than
two hours, the one-hour average would
have a numerical value lower than the
two hour national standard.
Demonstration of compliance with a
lower numerical limit over a shorter
averaging period proves compliance
with a higher number over a longer
averaging period.
In lieu of a site-specific PM limit, EPA
could limit key operating parameters for
the PM control device to ensure that the
device's collection efficiency is
maintained at performance test level.
We are concerned, however, that
limiting key operating parameters (e.g.,
pressure drop across a fabric filter) may
not be adequate because there are many
complex operating and maintenance
factors that affect collection efficiency of
a PM control device. We believe that
continuous monitoring of a surrogate
emission (i.e., PM) is far preferable to
continuous monitoring of operating
parameters that less effectively relate to
collection efficiency. (We note,
however, that if the use of a PM GEMS
is not required in the final rule, the rule
would establish limits on the PM
control device operating parameters as
the next preferable approach.)
Also, EPA invites comment on
allowing small on-site sources (defined
in § 63.1208(b)(l)(ii) in the proposed
regulations) to obtain a waiver from the
requirement of installing a PM GEMS. If
the waiver is promulgated and allowed
by the permitting authority, the facility
would demonstrate compliance with
PM by establishing operating parameter
limits described in subsection b,
"Operating Parameter Limits," below.
iv. Proposed PM GEMS Performance
and Calibration Specifications. There
are existing performance specifications
(PS) developed by the International
Standards Organization (ISO) for PM
GEMS. The ISO specifications have been
modified slightly to account for the US
regulatory environment. This PM GEMS
PS is proposed here as Part 60,
Appendix B, Performance Specification
11. EPA invites comment on this
specification.
It is proposed that HWCs follow the
German approach to using PM GEMS.
This approach involves deriving a site-
specific statistically derived calibration
curve of PM GEMS response to manual
methods results for each fuel type.
When the facility changes fuel type or
supplier, a new PM GEMS calibration
would be performed.
It is proposed that PM GEMS be
calibrated to the reference method, 40
CFR 60, Appendix A, Method 5.
Performance specification 11 requires
that at least 15 measurements be made
at least three grain loadings. During
calibration, Method 5 and the GEMS
will be run simultaneously during each
of the 15 measurements. The average
output response from the GEMS is then
compared to the results of each of the
15 measurements. Two calibration
procedures are possible for PM GEMS:
linear and quadratic. The performance
specification proposes that facilities first
calculate the calibration using the linear
relationship, then the quadratic. If the
quadratic relationship proves to be a
better fit to the data, it is used.
Otherwise the linear relationship is
used.
The quality assurance (QA)
requirements for HWC GEMS propose
that an absolute calibrations audit
(ACA) be performed quarterly (every
three months) and a relative calibration
audit (RCA) be performed every 18
months (30 months for small on-site
facilities). If the calibration has drifted,
a new calibration shall be performed.
An absolute calibration audit would not
be required during quarters when a
response calibration audit is conducted.
Also, there is a concern that the
suitability of a calibration curve for a
PM GEMS is dependant on the type of
fuel used. For the purposes of this
source category it is proposed that fuel
type be defined by the physical state of
the fuel: gas, liquid, or solid. Therefore,
a facility that burns only gas, liquid, or
solid fuel would need to generate only
one calibration curve. Facilities which
wish to burn a combination of fuel types
would need to establish a single or
multiple calibration curves which
encompasses all combinations of fuel
mix. Facilities which use multiple
curves must describe in their quality
assurance plan their methodology for
deriving the curves and how the proper
curves will be used during normal
operation. See the TBD for more
information on calibration due to fuel
changes.
b. Operating Parameter Limits. If the
final rule does not require the use of a
PM GEMS, we would rely on limits on
ash feedrate and key PM APCD
operating parameters to ensure
continued compliance with the PM
emission standard. In addition, if the
provision allowing small on-site
facilities (defined in § 63.1208(b)(l)(ii)
of the proposed regulations) to waiver
the PM GEMS requirement is
promulgated and the facility elects not
to use a PM GEMS, the facility would
have to establish these operating
parameter limits to document
compliance with the PM emission limit.
-------
17438
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
i. Maximum Flue Gas Flowrate or
Production Rate. EPA is concerned that
flue gas flowrates exceeding those of the
performance test could decrease the
collection efficiency of the PM control
device. For that reason, EPA proposes
limiting flue gas flowrate. Alternately,
CKs and LWAKs could limit production
rate (e.g., production rate of clinker or
aggregate, or raw material feedrate)
since production rate is proportional to
flue gas flowrate. Either flue gas
flowrate or production rate would be
established as a one hour average. The
one-hour average would be the average
of the maximum hourly rolling averages
occurring during the comprehensive
performance tests.
ii. Maximum Ash Feedrate. A portion
of the.ash fed into a HWC is emitted as
PM. To limit the amount of PM emitted
at the stack, maximum ash feedrate
would be used as a compliance
parameter. As set out in the BIF rule,
however, EPA does not believe that an
ash feedrate limit is necessary for CKs
or LWAKs because entrained raw
materials comprise virtually all of their
PM emissions. See 266.103(c)(l)(iv) and
56 FR at 7146. Thus, for a cement or
lightweight aggregate kiln, variation in
ash content of the hazardous waste is
not likely to have a significant effect on
PM loading at the inlet to the PM
control device or PM emissions.
Conceptually, however, the feedrate of
ash in liquid feeds and the rate at which
air pollution control dust (e.g., cement
kiln dust) is returned to the kiln may
have significant effect on the loading of
small particles. Absent a GEMS, EPA ,
seeks comment on addressing this issue.
It is proposed that the limit on ash
feedrate be established on a one-hour
average to coincide with the other
control parameters for PM. This one-
hour average for ash feed is also
consistent with and conservative
relative to the two-hour (national)
averaging period for a PM GEMS.
iii. Wet Scrubber Parameters,
including Venturi and Ionizing Wet
Scrubbers. Venturi and other wet
scrubbers remove PM by capturing
particles in liquid droplets and
separating the droplets from the gas
stream. The wet scrubber parameters
pertinent to PM control are minimum
pressure drop across the wet scrubber,
minimum liquid feed pressure to the
wet scrubber, minimum blowdown or
solids content of the scrubber liquid,
and minimum liquid to gas ratio.
Ionizing wet scrubbers have the
additional parameter of minimum
power input. Parameters for pressure
drop, liquid feed pressure, and liquid to
gas ratio are described, below, in the
section dealing with HC1 and Clz .
standard. Parameters for blowdown or
solids content and power input to an
IWS are described in the next
paragraphs.
Slowdown is the amount of scrubber
liquid removed from the process and
not recycled back into the wet scrubber.
Blowdown is an important wet scrubber
parameter because, as scrubber liquid is
removed and not recycled, solids are
removed as well and not recycled.
Alternately, solids content can be used
as a direct indicator of solids content in
the scrubber liquid. When the scrubber
liquid contains high solids, there is a
lack of a driving force for more solids
to go into solution. Conversely, when
little or no solids are in the scrubber
liquid, there is a strong driving force for
liquids to go into solution. Therefore,
establishing a maximum solids content
for a wet scrubber is desirable.
If a PM GEMS is not required in the
final rule, we propose that either a
minimum blowdown or a maximum
solids content limit be established. Both
would be established on both a ten-
minute and a one-hour average. The ten-
minute average would be the average of
the minimum, for blowdown, or
maximum, for solids content, ten-
minute averages occurring during each
run of the comprehensive performance
test. The one-hour average would be the
average over all runs.
Power input to an IWS is important
because IWSs charge the particulate
prior to it entering a packed bed wet
scrubber. The charging aids in the
collection of the particulate onto the
packing surface in the bed. The
particulate is then washed off of the
packing by the scrubber liquid.
Therefore, power input to an IWS is a
key parameter to the proper operation of
an IWS and EPA proposes that facilities
establish a limit on minimum power
input to an IWS. This limit would be
established on both a ten-minute and
one-hour average. The ten-minute .
average would be the average of the
minimum 10 minute averages occurring
during each run of the comprehensive
performance test and the one-hour
average would be the average across all
runs.
Facilities may obtain a waiver from
these requirements for wet scrubbers
from the Director if they can identify
other key parameters which affect good
control of PM through their use and use
these parameter limits during normal
operation.
iv. Fabric Filters. Fabric filters (FFs),
also known as baghouses, are used to
filter PM from stack flue gas prior to the
stack. Performance of a fabric filter
directly affects PM emissions. Filter
failure is typically due to filter holes,
bleed-through migration of particulate
through the filter and cake, and small
"pin holes" in the filter and cake. Since
low pressure drop is an indicator of one
of these types of failure, pressure drop
across the fabric filter is the best
indicator that the fabric filter has not
failed.
If the final rule does not require the
use of a PM GEMS, EPA proposes that
a limit on minimum pressure drop
across the fabric filter be established to
ensure that collection efficiency is
maintained. EPA proposes that this
limit be established on both a ten-
minute and a one-hour average. The ten-
minute average would be the average of
the single lowest 10-minute rolling
averages occurring during each run of
the comprehensive performance test.
The one-hour average would be the
average over all runs.
EPA believes it would also be useful
to establish other, potentially better
parameters as measures of collection
efficiency for the fabric filter. Collection
efficiency from fabric filters is a
function of filter type, face velocity
(which in turn is a function of flue gas
flowrate and filter material area), cake
build-up on the filter, and particulate
matter characteristics (primarily
particulate size distribution).
Unfortunately, the Agency is not aware
of a way to establish parameters for
these indicators of collection efficiency.
Therefore, EPA invites comment on
what type of parameters could be used
as better indicators of collection
efficiency and on what averaging period
they should be established.
Facilities may obtain a waiver from
these requirements for PM APCDs from
the Director if they can identify key
parameters which affect good control of
PM through their use and use these
parameter limits during normal
operation.
v. Electrostatic Precipitators.
Electrostatic precipitators (ESPs)
capture PM by charging particulate in
an electric field and collecting the
charged particulate on an inversely
charged collection plate. Electrical
power is the product of the electrical
voltage and the current. High voltage'
leads to high magnetic field strength
which results in an increase in the
saturation charge level the particle can
obtain, which in turn causes an increase
in charged particle migration to the
collection plate. High current leads to
an increased particle charging rate and
increased electric field strength near the
collection electrode due to a phenomena
called "ionic space charge" and, thus,
increased collection at the plate. High
voltage is also important on the
collection plates, since this will increase
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17439
collection of the inversely charged
particles on the plates. Therefore,.
maximizing both voltage and current is
desirable for good collection. Therefore,
power input to the ESP is a direct
function of ESP efficiency since, the
lower the power input, the lower the
collection efficiency.
For these reasons, EPA proposes that
facilities establish a limit on minimum
power input to the ESP to ensure that
collection efficiency is maintained at
performance test levels if the final rule
does not require the use of a PM GEMS.
This limit would be established on both
a ten-minute and one-hour average. The
ten-minute average would be the
average of the minimum 10-minute
averages for power input which occurs
during each run of the comprehensive
performance test. The one-hour average
would be the average over all runs.
Since very high power can be
supplied to either the charging or
collection parts of an ESP, EPA also
invites comment on whether power
input to each part of the ESP should be
controlled.
Facilities may obtain a waiver from
these requirements for ESPs from the
Director if they can identify more
appropriate parameters that would
ensure that collection efficiency is
maintained at performance test levels.
8. Waiver of Operating Limits
We believe that a provision to waive
any or all of the operating limits
discussed in this section is appropriate
given that many sources will employ
unique and innovative combinations of
emission control devices. Fixed,
national monitoring and compliance
requirements may not be applicable or
reasonable in some situations.
Accordingly, the proposed rule would
allow the Director to grant a waiver from
any or all of the operating limits
discussed in this section if a source
documents in writing that other, more
appropriate operating limits would
ensure compliance with the pertinent
emission standard. See proposed
§ 63.1210(s). The documentation must
include recommended averaging
periods for the alternative operating
limits, and the basis for establishing the
limits based on operations during the
comprehensive performance test.
9. Request for Comment on Waiver of
GEMS Requirements for Small, On-Site
Sources
We specifically invite comment on
whether the final rule should allow
small, on-site sources the option of not
having to use a mercury and PM GEMS.
Under a waiver, the source would be
required to comply with the operating
limits discussed above in lieu of using
a GEMS. As a separate issue, EPA is
proposing less stringent RATA and RCA
frequencies for the mercury and PM
GEMS (and testing in general, see
section El of this part) for these sources.
Sources with a gas flowrate less than
23,127 acfm would be considered small.
See discussion in Part Four, Section I,
for the rationale for that demarcation
between small and large units. See also
§ 63.1208(b)(l)(ii) of the proposed rule.
We believe that this waiver could be
warranted because small, on-site
sources may be better able to effectively
sample and analyze feedstreams to
ensure compliance with feedrate limits,
and because their emission rates (i.e.,
environmental loading) would be less
than from large sources.
We also invite comment on basing the
definition of what is small on a gas
flowrate and the value proposed for
defining what is a small source.
D. Combustion Fugitive Emissions
Operating parameters on combustion
fugitive emissions are necessary to
ensure that these emissions do not leak
from the combustion device, APCDs, or
any ducting connecting them. The
current BIF and incinerator rules
establish provisions for controlling
combustion fugitive emissions (see
§§ 266.102(e)(7)(I) and 264.345(d)).
Today's proposed rule would require
sources to comply with those
requirements, with minor clarifications.
See proposed § 63.1207(b). Specifically,
it is proposed that sources shall:
—keep the combustion chamber and all
ducting and devices from the
combustion chamber to the stack
totally sealed against fugitive
emissions; or
—maintain the maximum pressure on
an instantaneous basis in the
combustion chamber and in all
ducting and devices from the
combustion chamber to the stack at
lower than ambient pressure at all
times;159 or
—use some other means of control
demonstrated to provide equivalent
control. Support for such
demonstration shall be included in
the operating record with prior
written approval obtained from the
Director.
In addition, the rule would require the
owner or operator to specify in the
operating record the method used for
fugitive emission control.
EPA continues to believe this
approach (already in effect for
159 That is, on an instantaneous basis, without an
averaging period. The recording system must record
the instantaneous values continuously.
incinerators and BIFs) is appropriate
and is proposing to retain it here. There
are cases, however, particularly at
munitions incinerators, where
combustion fugitive emissions are a
problem even when less than ambient
pressure is apparently being
maintained. In these cases, the Director
may require in the RCRA operating
permit continual video surveillance of
the equipment to ensure there are no
leaks. If leaks occur, each occurrence is
a violation, and would require an
automatic waste feed cut-off (AWFCO).
In addition, as with all AWFCOs, the
owner or operator must identify the
cause of the leak and identify remedial
action taken to minimize future
occurrences.
We are also proposing to make
conforming changes to the existing BIF
and incinerator requirements for
combustion fugitive emissions. See
proposed §§ 264.347(e), 265.347(c), and
266.102(e). The effective date of these
conforming requirements would be 6
months after publication of the final
rule in the Federal Register, and so
would take effect before the MACT
standard compliance date.
E, Automatic Waste Feed Cutoff
(AWFCO) Requirements and Emergency
Safety Vent (ESV) Openings
We explain in this section that the
source must be in compliance with the
CEMS-monitored emission standards
and the operating limits at all times.
This would be ensured by requiring that
all operating parameters for which
limits would be established (as
discussed above) must be interactive
with an automatic waste feed cutoff
(AWFCO) system. Further, we also
describe the periodic reporting
requirements that would apply if 10
AWFCOs that result in an exceedance of
a CEMS-monitored emission standard or
operating limit occur during any 60-day
period. Finally we explain the
consequences of, and reporting
requirements for, emergency safety vent
openings.
1. Automatic Waste Feed Cutoff System
Sources must be in compliance with
the CEMS-monitored emission
standards and operating limits at all
times. See proposed §63.1207 (a)(l) and
(a)(2). If a facility exceeds a standard or
operating limit, today's rule proposes
that the hazardous waste feed be
instantaneously and automatically cut
off. This requirement now exists under
current incinerator permits and the
Agency's BIF rules (see
§ 266.102(e)(7)(ii)). After an AWFCO,
the source must continue to monitor all
AWFCO operating parameters (and
-------
17440
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
CEMS-monitored emissions) and cannot
begin feeding hazardous waste again
until all parameters come within
allowable levels. Further, to minimize
emissions of regulated pollutants,
including products of incomplete
combustion that could result from the
perturbation caused by the waste feed
cutoff, combustion gases must continue
to be routed through the air pollution
control system after a cutoff, and
minimum combustion temperature must
be maintained for as long as hazardous
waste remains in the combustion
chamber.
As currently required under the BIF
rule, all AWFCO parameters must
continue to be monitored after an
AWFCO, and hazardous waste firing
cannot resume until all parameters are
within allowable levels. Thus, all rolling
averages must continue to be calculated
even when hazardous waste is not being
burned.160
Today's proposed rule would require
the following parameters to be AWFCO
parameters:161
—CEMS-monitored emission standards
—All applicable feedrate limits (e.g.,
hazardous waste, pumpable LVM
metals, total SVM and LVM metals)
—Minimum combustion chamber
temperature (each chamber)
—Maximum combustion chamber
temperature
—Maximum temperature at the inlet to
the initial dry PM control device
—Maximum combustion chamber
pressure (if used to control
combustion fugitive emissions)
—Maximum flue gas flowrate (or
production rate)
—Minimum flue gas flowrate (where
required (e.g., under §63.1208(h)(l))
(or production rate)
—Limits on operating parameters of the
emission control equipment (e.g.,
carbon injection rate)
160 This requirement that all parameters must
continue to be monitored after a AWFCO assumes
that the operator intends to begin burning
hazardous waste as soon as the operating
parameters return to allowable levels. If not,
however, it may not be practicable to require
monitoring of AWFCO parameters when hazardous
waste is not burned. We specifically request
comment on a reasonable interval of time after a
AWFCO and before hazardous waste firing could be
resumed during which the operator would not be
required to monitor the AWFCO parameters. For
example, if the operator did not intend to begin
burning hazardous waste for 8 hours after the
AWFCO, it may not be appropriate to require
monitoring of AWFCO parameters during that
period.
161 We note that during the RCRA permitting
process, permit writers may identify additional
operating parameters they determine to be
necessary on a case-specific basis in order for the
source to comply with the standards. See
subsection C.I. of this part, "Continued
Applicability of RCRA Omnibus Authority," for
more information on this.
—Failure of the Automatic Waste Feed
Cut-off system.
—Whenever continuous monitoring
systems (CMS) or the measurement
component of the CMS registers a
value beyond its rated scale.
We note that the current requirements
for BIFs and incinerators do not require
a AWFCO whenever a measurement
component of the CMS registers a value
beyond its rated scale or when the
AWFCO system fails. To ensure that
those standards conform with today's
proposal, we are proposing to add this
requirement to those rules. The effective
date of these conforming requirements
would be six months after publication of
the final rule in the Federal Register,
and thus would precede the MACT
standard compliance date.
If an operating limit or CEMS-
monitored emission standard is
exceeded after the hazardous waste feed
has ceased but while hazardous waste
remains in the combustion chamber, it
is a violation of the relevant emission
standard.162
As currently required for BIFs, the
AWFCO system and associated alarms
must be tested at least once every seven
days when hazardous waste is burned to
verify operability, unless the owner or
operator documents in the operating
record that weekly inspections will
unduly restrict or upset operations and
that less frequent inspections will be
adequate. At a minimum, operational
testing must be conducted at least once
every 30 days.
Under today's proposed rule, owners
and operators would be required to
document in the operating log the cause
of each AWFCO that is associated with
an exceedance of an operating limit or
CEMS-monitored emission standard163
and document the preventive measures
taken to minimize future AWFCOs.
Also, we are proposing a reporting
requirement for excessive AWFCOs
caused by violations to alert regulatory
officials that a source is having
operational problems. Thus, regulatory
officials can increase frequency of
inspections and review the sources
operating plan. In addition, the Director
may specify requirements through the
RCRA permit beyond recordkeeping and
reporting for addressing AWFCOs (i.e.,
I62If an operating limit is exceeded (when
hazardous waste is in the combustion chamber), the
source has violated the emission standard for which
the operating limit is used to ensure compliance.
IS3Not all AWFCOs are the result of an
exceedance of an emission standard or operating
limit. AWFCOs which are not associated with a
violation must be recorded in the operating log but
need not be reported.
approval to restart hazardous waste
feed, etc.)
Owners or operators would be
required to submit an "Excessive
AWFCO Report" to the Administrator if
more than 10 AWFCOs associated with
an exceedance of an operating limit or
CEMS-monitored emission standard
occur during any 60 calendar-day
period. After 10 such cutoffs occur, the
60 calendar-day clock would begin
anew. The report would have to be
postmarked within five calendar days of
the tenth AWFCO associated with an
exceedance, and would have to
document the cause of each such cutoff
and preventive measures taken to
minimize future cutoffs.
We invite comments on alternative
exceedance frequencies that would
trigger the need to submit an Excessive
AWFCO Report, such as incurring 5
cutoffs in any 30 calendar-day period. A
shorter accounting period would enable
enforcement officials to better identify
problem facilities.
2. Emergency Safety Vent (ESV)
Openings
Today's rule would require that
combustion gases always pass through
the emission control system in place
during the comprehensive performance
test. Thus, opening an emergency safety
vent (ESV) (including emergency vent
stacks, bypass stacks, thermal relief
valves, and pressure relief valves) to
bypass any part of the emission control
system would be a violation of that
requirement and the emission standard
the by-passed control device is designed
to control. See proposed § 63.1207(a)(3).
We are also proposing to make
conforming changes to the RCRA
incinerator standards of Part 264,
Subpart O, to provide consistency.
While this section specifically addresses
ESVs, the requirements apply to any
type of air pollution control bypass
stack while hazardous waste remains in
the combustion chamber.
ESVs are safety devices which are
designed to allow combustion gases to
bypass the air pollution control
equipment in order to: (1) Prevent
ground-level releases which could
endanger workers, in the event of an
overpressure, or (2) prevent damage to
the air pollution control equipment in
the event of excessively high
temperatures. An ESV opening allows
uncontrolled emissions to directly enter
the atmosphere. Some ESVs are situated
prior to the secondary combustion
chamber. This chamber is important for
organics destruction in an incinerator.
Further, since incinerators normally
demonstrate compliance with the
regulatory performance standards while
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17441
using their secondary combustion
chambers and air pollution control
devices, emissions from ESVs are
expected to be in excess of levels set by
the performance standards for the
respective devices.
There are situations where the
alternative to opening an ESV (e.g.,
fugitive emissions at ground level, or
even an explosion) are worse from a
health and environmental standpoint.
Thus, EPA would like to emphasize that
simply eliminating an ESV itself is one •
solution, but not appropriate in some
cases. Rather, EPA believes that
emergency (or other) situations which
would cause either an ESV opening or
fugitive emissions from the combustor
can, and should be, prevented to the
greatest extent possible.
EPA believes that most facilities can
readily make changes in their operations
which can reduce ESV openings. To
minimize ESV openings, facilities may
need to repair or replace unreliable
equipment, better control the feeding of
waste, or add redundant systems where
necessary.
In the preamble to the proposed
amendments for hazardous waste
incinerators (55 FR17890, April 27,
1990), EPA proposed to clarify the
regulatory status of ESV openings. The
Agency proposed that no ESV openings
be allowed while hazardous waste is in
the unit. In this case any ESV opening
while hazardous waste remains in the
unit would be a permit violation and
subject to enforcement action. This is
being reproposed today.
Also in the proposed rule for
hazardous waste incinerators (55 FR at
17891), EPA proposed to amend
§ 264.345(a) to clarify that an incinerator
must operate in accordance with the
operating requirements specified in
their permit whenever there is
hazardous waste in the incinerator.
Today's rule is again proposing to
amend § 264.345(a) to clarify that an
incinerator must be operated in
accordance with the conditions
specified in the permit and meet the
applicable emission standards at all
times that hazardous waste or hazardous
waste residues remain in the chamber.
(This is a conforming change.)
For BIFs, the regulations state that
they must be operated in accordance
with the operating limits and the
applicable emission standards at all
times when there is waste in the unit.
§266.103(c)(l). Further,
§266.102(e)(7)(ii)(B) requires that
combustion gases must be routed
through the air pollution control system
as long as waste remains in the unit.
The BIF final rule discusses that a BIF
must be in compliance at all times that
there is hazardous waste in the unit,
regardless of whether an automatic
waste feed cutoff has occurred. See 56
FR at 7160. The activation of the
automatic waste feed cutoff system does
not relieve the facility from its
obligation to comply with the permit
conditions while waste remains in the
unit. Today's rule does not propose any
changes to this regime.
Finally, today's proposed rule would
require the owner or operator to record
in the operating log the ESV opening,
the reason for the opening, and
corrective measures taken to minimize
the frequency of openings. Further, the
owner or operator would have to submit
a written report to the Administrator
within 5 calendar days of each ESV
opening documenting the information
provided in the operating log.
While it is understood that there can
be mitigating circumstances which
require the use of ESVs, these instances
should be minimized. Therefore, it is
proposed that the owner or operator
prepare an ESV Operating Plan in which
the owner or operator shall address
what they will do to prevent the use Of
the ESV and release uncontrolled
emissions into the air and what they
will do to minimize the hazard from
such releases (such as back-up systems,
maintaining flame temperature, and
combustion air to combustion organics.)
This plan is analogous to the
"Preparedness and Prevention and
Contingency Plan" discussed in the
1990 proposed revisions to the
hazardous waste incinerator rule (55 FR
at 17890). A corresponding change to
the current hazardous waste incinerator
rules are proposed as well. ~
F. Quality Assurance for Continuous ,.
Monitoring Systems
EPA proposes specific quality -.•
assurance (QA) requirements for
continuous monitoring systems (CMS).
These systems can be classified as:
continuous emissions monitoring
systems (GEMS); analysis of
feedstreams; and continuous monitoring
systems to comply with limits on other
operating parameters.
1. Continuous Emissions Monitoring
Systems (GEMS)
The rule would require HWCs to
comply with the general monitoring
requirements under § 63.8 for all MACT
sources except as discussed below. In
addition, the rule would establish in the
appendix to Part 63, Subpart EEE,
specific quality assurance (QA) and
quality control (QC) requirements for
GEMS used by HWCs. These
requirements would supersede the
requirements in Appendix F of Part 60
for these sources. We are proposing an
appendix to Subpart EEE in lieu of the
requirements of Appendix F because the
proposed appendix to Subpart EEE
would incorporate various issues
particularly relating to HWCs (e.g.,
requirements for specific GEMS not
addressed by Appendix F; out-of-control
periods and data reporting are not
relevant to HWCs because HWCs cannot
burn hazardous waste if the GEMS is not
meeting performance specifications).
a. Applicability of § 63.8 .
Requirements. Most of the §63.8
monitoring requirements for MACT
sources would apply to HWCs including
requirements for the owner and operator
to develop and implement a quality
control program (§ 63.8(d)(2)) and
conduct a performance evaluation test
in conjunction with the performance
test to demonstration compliance with
the emission standards (§ 63.8(d)(2) and
(e)(4)). Section 63.8(f) also provides for
approval of an alternative monitoring
method.
Several provisions of § 63.8, however,
would not apply to HWCs. They are as
follows:
i. §63.8 (c)(l)(I)-(iii), (c)(4), (c)(7),
(c)(8), and (g)(5) would not apply
because these paragraphs address
requirements relating to operations
when the GEMS is out of compliance
with the relevant performance
specifications. Hazardous waste .cannot
be fed (or remain in the combustion
chamber) if the GEMS is not in
compliance with performance
specifications.
ii. § 63.8 (c)(4)(ii) and (g)(2) would not
apply because these paragraphs define
continuous operation and data
reduction inconsistently with today's
proposed rule. Under today's rule, the
performance specifications in Appendix
B to Part 60 and the data quality
objectives in the appendix to Part 63,
Subpart EEE, define continuous
operation specific to the GEMS.
b. Quality Assurance Procedures. The
proposed appendix to Part 63, Subpart
EEE, defines quality assurance
procedures for GEMS at HWCs. If a
GEMS component is not in compliance
with applicable quality assurance •
procedures or performance
specifications (provided in Appendix B,
Part 60), hazardous waste burning must
cease immediately and cannot be
resumed until the owner or operator
documents that the GEMS meets the
performance specifications.
The appendix would require owners
and operators to develop and implement
a quality assurance and quality control
(QA/QC) program. It would define
requirements for determining
compliance with calibration and zero
-------
27442
Federal Register / Vol. 61, No. 77 I Friday, April 19, 1996 / Proposed Rules
drift specifications provided in
Appendix B. It would also define ,
requirements for performance .'.'.•
evaluations, that is, performance audits
including relative accuracy tests and
absolute calibration audits.
The appendix also deals with issues
specific to these source categories. It
establishes specific testing intervals for
GEMS for HWCs. It defines the one
minute and rolling averages, the oxygen
correction factor, GEMS span values^
and provides a provision to allow the
use of alternative span values. It
provides procedures for reestablishing a
rolling average after short term
interruptions such as calibration and
maintenance and long-term
interruptions such as periodic
downtime for kiln maintenance or for
weekends and holidays when the
facility is not being operated. It also
allows up to 20 minutes of GEMS
downtime for calibration purposes.
c. Conforming changes to the BIF and
incinerator rules. Conforming changes
are also proposed to the BIF and
incinerator rules: deleting the current
Part 266, Appendix IX, GEMS
requirements; and, instead, requiring
the use of the Part 60, Appendix B,
performance specifications and the data
quality specifications in the appendix to
SubpartEEE.
d. Zero Drift and Zero Gas
Requirements. The Agency specifically
invites comment on two other issues
which affect all GEMS: whether the zero
drift requirements contained in the
appendix to Subpart EEE (and the
various performance specifications)
should be promulgated, or whether the
zero gas requirements should be
changed from the current 0—20 percent
levels to a 0—0.1 ppm level.
Many of the performance
specifications require that zero gas, or
zero level gas, contain between 0 to 20
per cent of the measured constituent. :
However, facilities often use just one
zero grade gas for all their GEMS, one
of "zero-grade nitrogen." Therefore,
EPA invites comment on whether this
requirement should be changed from 0
to 20 percent to 0 to 0.1 ppm of the
measured constituent.
: e. EPA certification of GEMS. EPA
invites comment on whether a process
should be established whereby GEMS
manufacturers could certify that their
GEMS meet the established performance
specifications. If this were promulgated,
a GEMS would not be allowed for use
on a hazardous waste combustor unless
it has been certified by EPA. The GEMS
certification would be similar to the
certifications used for TUV approval in
Germany and for GEMS used for.
compliance with EPA's acid,rain
program.
Issues EPA .needs to address in order
to promulgate such a process include:
what benefits the regulated community
and industry would incur as a result of
such S certification; how the program
would work; and whether a
nongovernment agency could dp this
task. ; '
vi. Correcting GEMS Readings for
Moisture Content. One quality
assurance issue that must be considered
is how often facilities need to measure
the moisture content of their flue gas.
All the standards proposed today are on
a dry basis, so knowing the flue gas
moisture content to correct GEMS
outputs to a dry basis is necessary, EPA
is considering two alternative
approaches to obtain the moisture
content of the flue gas. One involves
making periodic measurements of the
moisture content of the flue gas using
Method 4, found in Part 60, Appendix
A. Under this scheme, a facility would
take flue gas moisture measurements
quarterly, while conducting the ACA.
This moisture level would then be used
to correct GEMS outputs for moisture
throughout the next quarter.
Another alternative is that facilities
make instantaneous measurements of
the flue gas temperature at the GEMS
sampling point. The temperature would
then be used to determine the saturation
water concentration of the flue gas. The
saturation water concentration would .
then be used to correct the GEMS output
for moisture. .
EPA favors using the saturation water
concentration as a surrogate for flue gas
moisture because it is continuous,
frequently conservative, and cost-
effective compared to running a manual
method. One issue with this approach is
that facilities with wet APCS may have
a water concentration higher than the ;
saturated water concentration due to
entrained water droplets in the flue gas.
However, we do not have data on the
amount of entrained water droplets in
the flue gas and, thus, cannot determine
at this point how important this issue is.
The Agency requests data and
information from facilities with a wet
APCS regarding the total water
concentration (including water droplets)
in the flue gas compared with the
saturated water concentration. The
Agency will evaluate data and
recommendations of commenters on
these or other approaches in making a
determination on the best approach for
the final rale.
2. Analysis of Feedstreams
In this section, we discuss the
following proposed requirements for
analysis of feedstreams: (1) required
analysis plan; (2) requirement to submit
the plan for review and approval the
Director's request; (3) frequency of
analysis; and (4) information that must
be determined and recorded to
document compliance. (We note that
HWCs are already subject to these
requirements under 40 CFR Parts 261,
264, 265, 266, and 270.) We also request
comment on analysis of gaseous
feedstreams, including natural gas. We
also propose making a conforming
change to the BIF and incinerator rules
to clarify that constituent monitoring is
required for all feedstreams.
a. Feedstream Analysis Plan. The rule
would require (in § 63.1210(c)) an
owner or operator to obtain an analysis
of each feedstream that is sufficient to
document compliance with the
applicable feedrate limits. The owner or
operator must obtain the analyses for
each feedstream prior to feeding into the
combustor; This is done in order to
document compliance with the
applicable feedrate limits at all times.
To ensure that the owner or operator
will obtain an adequate analysis, the
owner or operator would be required to
develop and implement a feedstream
analysis plan and record it in the
operating record. The operating plan
must specify at a minimum: (1) the
parameters for which each feedstream
will be analyzed to ensure compliance
with proposed §63.1210; (2) whether
the owner or operator will obtain the
analysis by performing sampling and
analysis, or by other methods such as
using analytical information obtained
from others164 or using other published
or documented data or information; (3)
how the analysis will be used to
document compliance with applicable
feedrate limits (e.g., if hazardous wastes
are blended and analyses are obtained of
the wastes prior to blending but not of
the blended, as-fired, waste, the plan
must describe how the owner and
operator will determine the pertinent
parameters of the blended waste); (4) the
test methods which will be used to
obtain the analyses;165 (5) the sampling
method which will be used to obtain a
representative sample of each
feedstream to be analyzed using
sampling methods described in
Appendix I, Part 261, or an equivalent
method; and (6) the frequency with
which the initial analysis of the
feedstreanrwill be reviewed or repeated
iM when analytical information is provided by
others, the analysis plan must document how the
owner or operator will ensure it is complete and
accurate. . '
165 The information must be provided whether the
owner or operator conducts the analyses or the •
analyses are obtained from others. - . •
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17443
to ensure that the analysis is accurate
and up to date.166
We note that guidance on developing
a feedstream analysis plan is provided
in Waste Analysis At Facilities That
Generate, Treat, and Dispose of
Hazardous Waste, (OSWER [Office of
Solid Waste and Emergency Response]
#9938.4-03, April 1994). The document
!s available from the National Technical
Information Services (NTIS),
publication # PB94-963-603. In
addition, in April 1995, EPA published
a Notice of Availability for public
comment on Waste Analysis Guidance
for Facilities That Burn Hazardous
Wastes-Draft (Office of Enforcement and
Compliance Assurance # EPA 530-R—
94-019) (see 60 FR18402). This
guidance document provides assistance
in developing waste analysis plans
specifically for HWCs. The comment
period for this document closed on June
2,1995, and EPA is currently reviewing
and evaluating the comments received.
b. Review and Approval of Analysis
Plan. Under today's proposed rule, the
Director could require the owner or
operator to submit the analysis plan for
review and approval at any time. Given
that feedstream analysis is a primary
compliance approach for the SVM,
LVM, and HCl/Cb emission standards,
it is imperative that the source develop
and implement an adequate analysis
plan. Consequently, the Agency would
like to review and approve analysis
plans for each existing source at the
time of initial compliance (i.e., initial
notification of compliance).167
Because of resource constraints,
however, the Agency will review
analysis plans on a priority basis,
considering factors such as whether the
source accepts off-site waste, volume of
waste burned, and compliance
history.168 Therefore, the Agency wishes
to preserve flexibility on whether to
require a source to submit its analysis
plan for review and approval.
c. How to Comply with Feedrate
Limits. To comply with the feedrate
limits, the source must: (1) know the
concentration of the limited parameter
(e.g., SVM) in the feedstream at all
times; (2) know the feedrate of the
feedstream at all times; and (3) record
the feedrate (the product of the
'«*Tha analysis must bo repeated as necessary to
ensure that it is accurate and up to date. At a
minimum, the analysis must be repeated when the
owner or operator is notified or has reason to
believe that the process or operation generating or
producing the feedstream has changed.
167 Analysis plans would be reviewed and
approved for new sources during the RCRA
pormilting process (i.e., prior to commencement of
construction).
IMNote that the analysis plan will be reviewed
during facility inspections as well.
concentration times the feedstream rate)
in the operating record. The source
would know the concentration of the
parameter in the feedstream by
implementing the analysis plan
discussed above.
The source would know the feedrate
of the feedstream by using a continuous
monitor of the volumetric or mass
flowrate.169 If a volumetric flowrate
monitor is used, the source must know
the density of the feedstream at all times
if it is necessary to know the mass per
unit time feedrate.
In order for a facility to know the
concentration of the parameters at all
times, the source must record the
feedrate in the operating record. It
would be preferable to reduce the
burden on regulatory inspectors to
continuously record all of the
parameters used to calculate the
feedrate (e.g., concentration of metal,
volumetric flowrate, density) as well as
the feedrate itself. Other approaches
may be acceptable, however, such as
continuously recording only volumetric
flowrate, but clearly noting in the record
the concentration and density
associated with that volumetric flowrate
so that the inspector could readily
confirm that the feedrate was not
exceeded at the recorded flowrates. If a
source prefers the second approach, we
recommend that it informally notify the
Director for concurrence.
d. Request for Comment on
Monitoring Gaseous Feedstreams. We
request comment here on how to
address the difficulty of continuously
sampling gaseous feedstreams—both
natural gas and process gas—for
nonvapor constituents (metals, chloride
salts).
Natural gas is a primary fuel for
several HWCs. Under today's rule (as
well as the BIF regulations), this
feedstream, like all other feedstreams,
would be subject to the continuous
monitoring and recording provisions,
including feedstream sampling and
analysis for metal and chlorine
constituents.
Facilities have questioned whether it
is necessary to sample and analyze
natural gas for constituents they feel are
not reasonably expected to be present.
Therefore, the Agency is soliciting data
and information on whether (and at
what concentrations) the seven metals
that would be regulated in today's rule
are likely to be present in natural gas.
Based on the information submitted by
commenters, the final rule could
incorporate a number of options
including: (1) determine that natural gas
feedstreams need not be considered in
feedrate determinations because levels
of metals and chlorine and chloride are
not likely to be significant; (2) allow
sources to make a one-time, site-specific
determination of metals and chlorine
levels that could be used for feedrate
determinations provided that the
natural gas supplier does not change; or
(3) establish generic concentration
levels for metals and chlorine and
chloride that could be assumed to be
present. We also invite comment on
these or other approaches to address
this issue.
Process gas feedstreams pose a similar
problem. One approach for these
feedstreams would be to allow sources
to make a one-time determination of
metals and chlorine levels (by sampling
and analysis, process knowledge, or
other information) that could be used
for feedrate determinations until process
changes or other factors occurred that
could change the composition of the
gas. We invite comments on this or
alternative approaches to address this
issue.
3. Quality Assurance for Continuous
Monitoring Systems Other Than CEMS
Continuous monitoring systems
(CMS) other than CEMS include the
systems associated with monitors such
as thermocouples, pressure transducers,
stress/strain gages, flow meters, and pH
meters. In addition to the requirements
discussed below, we are proposing to
require compliance with the general
quality assurance procedures for
continuous monitoring systems (CMS)
provided by existing § 63.8(c)(4). See
proposed § 63.1210(d). That paragraph
requires owners and operators to verify
the operational status of CMS by, at a
minimum, complying with the
manufacturer's written specifications or
recommendations for installation,
operation, and calibration of the system.
To make current rules consistent with
the ones which will be promulgated
here, EPA proposes making conforming
changes to the BIF and incinerator rules
to incorporate quality assurance
requirements for CMS.
a. Sampling and Detection Frequency.
We are proposing to require that CMS
(other than CEMS)170 sample the
regulated parameter without
interruption, and evaluate the detector
response at least once each 15 seconds,
and compute and record the average
values at least every 60 seconds.
b. Exceeding CMS Span Would
Trigger a AWFCO. The rule would also
169 Quality assurance for the flowrate monitor is
discussed below in the text.
170The proposed CEM performance specifications
and data quality objectives define acceptable
sampling and detection frequency.
-------
17444
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
require that the automatic waste feed
cutoff (AWFCO) system be engaged if
the span of any CMS (other than a
GEMS) is exceeded. This is because it is
not practicable to establish span values
for each CMS as we have proposed for
each GEMS.
The issue arises because facilities
have the discretion of purchasing
equipment with any span. For CMS, the
span is defined as lite range between the
highest certifiable reading a CMS can
make (the "upper span") and its
corresponding minimum (the "lower
span.") If a CMS were to have an upper
span which is too low, say a
thermocouple with a upper span of
630°C, there would be no way.to
document accurately a temperature
higher than 630°C. This is a problem if
the facility routinely operates at a
temperature of, say, 750°C. For this
reason, it is important to ensure that
CMS are operated within their certified
span.
III. MACT Performance Testing and
Related Issues
Today's rule would require
performance testing to demonstrate
compliance with the proposed MACT
emission standards. The requirements
and procedures for MACT performance
testing are discussed here. In addition,
HWCs would continue to be subject to
the existing trial burn requirements
during the RCRA permitting process.
The interaction between the RCRA trial
burn and the MACT performance test is
also discussed here. In addition, we
discuss in this section the waiver for
performance testing for Hg, SVM, LVM,
and HCl/Cla that would be provided for
sources that feed de minimis levels of
these metals or chlorine. Finally, we
discuss in this section requirements for
relative accuracy tests for GEMS.
A. MACT Performance Testing
Two types of performance testing
would be required to demonstrate
compliance with the proposed MACT
emission standards: comprehensive
performance testing and confirmatory
performance testing. See proposed
§63.1208.
1. Comprehensive Performance Testing
The purpose of the comprehensive
performance test is to initially and
periodically thereafter: (1) demonstrate
that the source is in compliance with
the CEMS-monitored emission
standards (e.g., PM, Hg, CO, HC); (2)
conduct manual stack sampling to
demonstrate compliance with the
emission standards for pollutants that
are not monitored with a GEMS (e.g., D/
F, SVM, LVM, HC1/C12); (3) establish
limits on the applicable operating
parameters provided by proposed
§ 63.1210 (Monitoring Requirements) to
ensure that compliance is maintained
with those emission standards for which
a GEMS is not used for compliance
monitoring; and (4) demonstrate
performance of CMS is consistent with
the requirements and quality assurance
plan. Thus, the comprehensive
performance test has purposes similar to
the RCRA trial burn and BIF interim
status compliance test. It would be more
like a BIF interim status compliance
test, however, because of the low level
of Agency oversight and high degree of
facility self-implementation, as
discussed below.
a. Operations During Comprehensive
Performance Testing. Given that limits
will be established on operating
parameters during the comprehensive
performance test, sources will likely
want to operate during the test at the
edge of the operating envelope that they
believe is both necessary to operate
efficiently and comply with the
emission standards. Accordingly,
sources may elect to spike feedstreams
with metals or chlorine, for example, to
ensure that the feedrate limits are high
enough to accommodate normal
operations while allowing some
flexibility to feed higher rates at times.
In addition, sources may identify two
or more modes of operation for which
separate performance tests would be
conducted and for which separate limits
on operating conditions would be
established. In this situation, the source
would be required to note in the
operating record under which mode of
operation it was operating at all times.
An example of when two modes of
operation must be identified would be
a cement kiln that routes its kiln off-gas
through the raw meal mill to help dry
the raw meal. When the raw meal mill
is not operating (perhaps one third of
the time), the kiln gas bypasses the raw
meal mill. Emissions of PM and other
HAPs or HAP surrogates may vary
substantially depending on whether the
kiln gas bypasses the raw meal mill.
When conducting the comprehensive
performance test, sources must also
operate under representative conditions
for the following parameters to ensure
that emissions are representative of
normal operating conditions: (1) types
of organic compounds in the waste (e.g.,
aromatics, aliphatics, nitrogen content,
halogen/carbon ratio, oxygen/carbon
ratio) and volatility of wastes, when
demonstrating compliance with the D/F
emission standard; and (2) cleaning
cycle of the PM control device (e.g., ESP
rapping cycle) when demonstrating
compliance with the SVM and LVM
emission standard when using manual
stack sampling and the D/F emission
standard.
b. Frequency of Testing. The rule
would require that the comprehensive
performance test be performed
periodically because the Agency is
concerned that long-term wear-and-tear
on critical components (e.g., firing
systems, emission control equipment)
could adversely affect emissions. Large
sources (i.e., those with a stack gas flow
rate greater than 23,127 acfm.) and
sources that accept waste from off-site
would be required to perform
comprehensive performance testing
every three years.
Small, on-site sources would be
required to perform testing every five
years, unless the Director determines
otherwise on a case-specific basis. The
proposed testing frequency would be
less for small, on-site sources because of
cost-effectiveness concerns. In addition,
we note that, from the RCRA
perspective, small, on-site sources are
more familiar with the wastes they burn,
the waste may be more homogeneous
and less complex, and they burn smaller
volumes of waste. Thus, their emissions
may not pose the same hazard as
emissions from large or commercial
facilities. We invite comment on this
approach.
The Director may determine,
however, that a small, on-site source
may pose the same potential hazard as
a large or off-site source because of the
factors listed above, compliance history,
or other reasons. Accordingly, the rule
would allow discretion for the Director
to require a three-year testing frequency
for such small, on-site sources as
warranted.
c. Agency Oversight. The proposed
rule would require the owner or
operator to submit a "notification of
performance test" to the Administrator
60 days prior to the planned test date.
The notification must be accompanied
by a site-specific test plan for review
and approval by the Administrator. This
is consistent with the general provisions
for MACT sources provided by § 63.7 (b)
and (c). See those paragraphs for
provisions regarding: (1) Agency
approval of the test plan; (2) 30-day
period for the Agency to approve or
disapprove the test plan; m and (3)
notwithstanding Agency approval or
disapproval, or failure to approve or
disapprove, the test plan, the owner or
operator must comply with the
applicable requirements, including the
171 Generally, §63.7(c)(3) provides that the source
can assume the test plan is approved if the Agency
does not take action within 30 days of receiving the
original plan or any supplementary information.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17445
deadline for submitting the initial and
subsequent notifications of compliance.
In addition, the Agency has the option
of observing the performance test.
d. Operating Conditions During
Subsequent Tests. Although the rule
xvould allow the burning of hazardous
waste only under the operating limits
established during the previous
comprehensive performance test (to
ensure compliance with emission
standards not monitored with a GEMS),
two types of waivers from this
requirement would be provided during
subsequent comprehensive performance
tests: (1) an automatic waiver to 'exceed
current operating limits up to 5 percent;
and (2) a waiver that the Director may
grant if warranted to allow the source to
exceed the current operating limits
without restriction. The rationale and
implementation of these waivers is
discussed below.
The rule would provide an automatic
waiver because, without the waiver, the
operating limits would become more
and more stringent with subsequent
comprehensive performance tests. This
is because sources would be required to
operate within the more stringent
conditions to ensure that they did not
exceed a current operating limit. This
would result in a shrinking operating
envelope over time.
Accordingly, EPA is proposing to
allow sources to operate under the
"same" operating conditions as the
previous comprehensive performance
test in order to duplicate the current
operating limits. It is not practicable to
require a source to operate under the
exact same operating conditions as the
previous comprehensive performance
test, however. Therefore, the rule would
allow sources to deviate during
comprehensive performance testing by
up to 5 percent from the current
operating limits provided that the
source accept operating limits based on
the new performance test levels that are
the more stringent of the current
operating limits or levels achieved
during the new performance test. We
invite comment on whether this
provision would meet our objective of
ensuring that the operating envelope
does not shrink over time as subsequent
comprehensive performance tests are
conducted. For example, an additional
approach would be to provide for a site-
specific xvaiver of the 5 percent
deviation limit to allow deviations from
current operating limits as warranted to
ensure that the operating envelope does
not shrink.
The rule also proposes a waiver that
the Administrator may grant if
warranted to allow the source to exceed
the current operating limits without
restriction. This is because the source
may want to operate under less
restrictive limits and believes that it can
still comply with the emission
standards under the less restrictive
limits. For example, a source may want
to burn a waste with higher metal or
chlorine content, and/or the source may
want to install an improved emission
control device.
To accommodate such situations, the
rule would allow the Administrator tp
grant a site-specific waiver of the
operating limits if the source provides
supporting documentation that it is
likely to be able to meet the emission
standards under less restrictive
operating limits. The documentation
must be submitted prior to or at the time
of submittal of the notification of
performance test, and must include
empirical data or other data and
information to support the request. If
the waiver request is submitted with the
notification of performance test (which
must be accompanied by the test plan),
the Director will approve or disapprove
the waiver request under the procedures
for approving or disapproving the test
plan.
e. Testing Schedule and Notification
of Compliance. The owner or operator
must submit to the Administrator a
notification of compliance under
proposed § 63.1211(c) documenting
compliance with the emission standards
and CMS requirements, and identifying
applicable operating limits. (This
provision is similar to § 63.7(g).) The
notification must be postmarked by the
90th day following the completion of
performance testing and CMS
performance evaluation.
The initial notification of compliance
must be postmarked within 36 months
after the date of publication of the final
rule. Subsequent notifications must be
submitted within 90 days after the
completion of subsequent performance
testing. Subsequent comprehensive
performance testing must be initiated 36
months for large and off-site sources or
60 months for small, on-site sources,
respectively, after initiation of the initial
performance test.
Given the complexity of
comprehensive performance testing and
to allow for unforeseen events, however,
the rule would allow the subsequent test
to be initiated within a range of 30 days
before or after the 36 or 60-month
anniversary. The rule would require that
the anniversary date remain based on
the initial comprehensive performance
test. This would simplify recordkeeping
and preclude a source from
intentionally scheduling the test toward
the end of the 30-day grace period and
thus effectively obtaining a 37 or 61-
month testing frequency.
The rule would give a source the,
option of performing a comprehensive
performance test at any time before the
36 or 60-month anniversary. A source
may want to retrofit or add a new
emission control device prior to a test
anniversary date. To do so, the source
would be required to conduct a new
comprehensive performance test to
document compliance with emission
standards and to establish new
operating limits. The rule would require
the source to follow the same
procedures for this comprehensive
performance test as discussed above
(e.g., submittal of notification of
performance testing and test plan;
review and approval of test plan). Note
that conducting a comprehensive
performance test prior to the normal
anniversary date would establish a new
anniversary date.
f. Time Extensions for Subsequent
Performance Tests. The rule would
allow the Administrator to grant up to
a 1 year time extension for any
performance test subsequent to the
initial comprehensive performance
test.172 This would enable the source to
consolidate, into one test, both the
MACT-related performance testing and
the RCRA trial burn testing, which are
both required for issuance and
reissuance of RCRA operating
permits.173 (Trial burn testing
requirements are discussed below.)
For example, if the comprehensive
performance test anniversary were a
date proximate to the date scheduled for
the trial burn, we believe it is reasonable
to allow the source to conduct only one
test to satisfy both requirements (i.e., the
MACT-related performance test and the
RCRA trial burn). To address this
situation, the rule would, allow up to a
one-year time extension for the
performance test.174
When the trial burn and performance
tests are consolidated, the anniversary
dates for subsequent performance tests
would be correspondingly adjusted. For
example, if the anniversary date for a
172 Note that we discuss in Part Five, Section I
(Selection of Compliance Dates) of the preamble
that the rule would provide up to a ;L-year time
extension to submit the initial notification of
compliance.
173 In addition, the source may experience a major
outage whereby the performance test could not be
conducted within the 2-month window around the
anniversary date. This time extension provision
could address this situation as well.
174 Note that, if the trial burn were scheduled
before, rather than after, the performance test
anniversary date, there would not be a problem
because the source qan conduct a comprehensive
performance test at any time prior to the
anniversary date. If so, the anniversary date is
simply moved up.
-------
17446
Federal Register / VoL 61, No. 77 /Friday, April 19, 1996 / Proposed Rules
confirmatory performance test for a
large or off-site source is January 1 and
the trial burn is scheduled for
September 1 of that year, the source may
adjust the anniversary date of the
confirmatory performance test to
September 1. This would also delay the
anniversary date for subsequent
comprehensive performance tests by 9
months. As noted above, under the
proposal a maximum of 12 months
delay could be granted.
The procedure for granting or denying
a time extension would be the same as
those for existing § 63.6(i) which allows
the Administrator to grant MACT
sources up to 1 additional year (in
addition to the 3 years beginning with
publication of applicable standards (e.g.,
MACT standards for HWCs) in the
Federal Register) to comply with the
standard.175 (These are also the same
procedures that would apply to a
request for a time extension for the
initial notification of compliance.)
We invite comment on alternative
maximum time periods for the
extension to allow sources to reasonably
consolidate performance and trial burn
testing, and whether the time.extension
should be automatic or require prior
approval by the Administrator.
vL Failure to Submit a Timely
Notification of Compliance. If the owner
or operator does not submit a
notification of compliance by the
required date, the rule would require
the source to immediately stop burning
hazardous waste (the same manner as
applied to BIFs certifying compliance
under RCRA § 266.103 in 1991). If the
source wanted to burn hazardous waste
in the future, it would be required to
comply with the standards and permit
requirements for new MACT and RCRA
sources. For example, if the source were
operating under RCRA interim status, it
would need to obtain a RCRA operating
permit and meet MACT standards for
new facilities before hazardous waste
burning could resume. Moreover, the
rule would require the source to obtain
written approval from the Administrator
before hazardous waste burning could
resume. (For RCRA interim status
sources, issuance of a RCRA operating
permit would constitute such written
approval.)
g. Failure of a Comprehensive
Performance Test. When a source
determines (e.g., based on GEMS
recordings, results of analysis of
samples taken during manual stack
sampling, or results of the CMS
175 Note, however, that § 63.6(i) applies to an
entirely different situation: extension of time for
initial compliance with the standard whereby
performance testing is conducted after the date of
compliance.
performance evaluation) that it has
failed any emission standard during the
performance test, it would be required
to immediately stop burning hazardous
waste. If, however, a source conducts
the comprehensive performance test
under two or more modes of operation
and meets the emission standards when
operating under one or more modes of
operation, it would be allowed to
continue burning under the modes of
operation for which it has met the
standards.
For sources that fail one or more
emission standards during all modes of
operation tested, the rule would enable
the source to burn hazardous waste only
for a total of 720 hours and only for the
purposes of pretesting (i.e., informal
testing to determine if it could meet the
standards operating under modified
conditions) or comprehensive
performance testing under modified
conditions.
Finally, failure to comply with an
emission standard after initial
notification of compliance would be a
violation of the rule.
We note that HWCs are currently
subject to virtually these same
requirements under RCRA rules.
h. Applicability of Existing Part 63
General Requirements for MACT
Sources. Part 63 establishes
requirements for performance testing in
§ 63.7 and requirements for extension of
compliance dates in § 63.6(i). Some of
these provisions would be directly
applicable to HWCs, some would be
applicable in modified form, some
would be superseded by today's rule,
and others are not applicable.
The following § 63.7 requirements
would be applicable to HWCs:
(1) Paragraph (a)(l) (Applicability)
and (a)(3)
(2) Paragraphs (b) (Notification of
performance test) and (c) (Quality
Assurance Program), except that all
sources would be required to submit the
test plan for review and approval
(3) Paragraph (d) (Performance testing
facilities)
(4) Paragraph (e) (Conduct of
performance tests), except that operating
conditions during comprehensive
performance testing would be as
discussed above (i.e., not normal
operating conditions), and operating
conditions during confirmatory
performance testing discussed below
would be under normal conditions as
defined in that discussion. Also,
emissions during startup and shutdown
would be included in the performance
tests, if the sources wishes to have the
authority to burn hazardous waste
during those periods.
(5) Paragraph (f) (Use of an alternative
test method)
(6) Paragraph (g) (Data analysis,
recordkeeping, and reporting), except
that the test results would have to be
reported 90 days after completion of the
test, rather than 60 days.
The following § 63.7 requirements
would not be applicable to HWCs:
(1) Paragraph (a)(2) (establishing
deadlines for performance testing)
because new HWCs would be required
to obtain a RCRA operating permit
before commencing construction. The
RCRA operating permit would specify
allowable periods of operation and
operating conditions prior to (and
following) performance testing. Existing
HWCs would be required to submit a
notification of compliance within 3-
years of the date of publication of the
final rule in the Federal Register.
(2) Paragraph (h) (Waiver of
performance tests), because the bases for'
the waiver are not relevant to HWCs as
follows: (1) the rule would allow the
Administrator to grant a time extension
to submit a notification of compliance;
and (2) the purpose of periodic testing
is to determine whether sources are
meeting the standards on a continuous
basis.
2. Confirmatory Performance Testing
Confirmatory performance testing for
D/F would be required mid-way
between the cycle required for
comprehensive performance testing to
determine if the source is continuing to
meet the emission standard. The Agency
is proposing such testing only for D/F
given: (1) the health risk posed by D/F;
(2) there is no GEMS for D/F; (3) there
is no feedrate limit of a material that
directly and unambiguously relates to
D/F emissions (as opposed to, for
example, metals feedrates, which
directly relate to metals emissions); and
(4) wear and tear on the equipment,
including any emission control
equipment, which over time could
result in an increase in D/F emissions
even though the source stays in
compliance with applicable operating
limits.
Confirmatory testing differs from
comprehensive testing, however, in that
the source would be required to operate
under normal, representative conditions
during confirmatory testing. This would
reduce the cost of the test while
providing the essential information
because the source would not have to
establish new operating limits based on
the confirmatory test.
a. Definition of Normal Operating
Conditions. Normal operating
conditions would be defined as
operations during which: (1) the GEMS
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17447
that measure parameters that could
relate to D/F emissions—PM, CO, HC—
are recording emission levels within the
range of the average value for each
GEMS (the sum of all one-minute
averages, divided hy the number of one
minute averages) over the previous 12
months to the maximum allowed; and
(2) each operating limit established to
maintain compliance with the D/F
emission standard (see discussion in
Part Five, section II.C.l) is held within
the range of the average values over the
previous 12 months and the maximum
or minimums, as appropriate, that are
alloxved. The Agency believes it is
necessary to define normal operating
conditions in this manner because,
otherwise, sources could elect to limit
levels of the regulated D/F operating
parameters (e.g., hazardous waste
feedrate, combustion chamber
temperature, temperature at the inlet to
the dry PM control device) to ensure
minimum emissions. Thus, without
specifying what constitutes normal
conditions, EPA believes the
confirmatory test could be meaningless.
On the other hand, the proposed
definition of normal conditions is broad
enough to allow the source flexibility in
operations during the test.
When conducting the confirmatory
performance test for D/F, sources must
also operate under representative
conditions for the following parameters
to ensure that emissions are
representative of normal operating
conditions: (1) types of organic
compounds in the waste (e.g., aromatics,
aliphatics, nitrogen content, halogen/
carbon ratio, oxygen/carbon ratio) and
volatility of wastes, when demonstrating
compliance with the D/F emission
standard; and (2) cleaning cycle of the
PM control device (e.g., ESP rapping
cycle).
Finally, when conducting the
confirmatory test for D/F, the source
would also be required to conduct a
performance evaluation of the CMS that
are required to maintain compliance
with the D/F emission standard.
b. Frequency of Testing. Large and off-
site sources would be required to
conduct confirmatory performance
testing 18 months after the previous
comprehensive performance test. Small,
on-site sources would be required to
conduct the testing 30 months after the
previous comprehensive performance
test. The same 2-month testing window
applicable for comprehensive tests
would also apply to confirmatory tests.
c. Agency Oversight, Notification of
Performance Test, Notification of
Compliance, Time Extensions, and
Failure to Submit a Timely Notice of
Compliance. The requirements that
would apply to comprehensive tests
would also apply to confirmatory tests.
d. Failure of a Confirmatory
Performance Test. When a source
determines (e.g., based results of
analysis of samples taken during
manual stack sampling) that it has failed
the D/F emission standard, it would
have violated the rule. The source
would be required to immediately stop
burning hazardous waste. If, however, a
source had conducted the
comprehensive performance test under
two or more modes of operation and met
the D/F emission standards during
confirmatory testing when operating
under one or more modes of operation,
it would be allowed to continue burning
under the modes of operation for which
it has met the standards.
For sources that fail one or more
emission standard during all modes of
operation tested, the rule would require
the source to modify design or operation
of the unit and conduct a new
comprehensive performance test to
demonstrate compliance with the D/F
emission standard and establish new
operating limits. Further, prior to
submitting a notification of compliance
based on the new comprehensive
performance test, the source could burn
hazardous waste only for a total of 720
hours, and only for purposes of informal
pretesting or comprehensive
performance testing.
B. RCRA Trial Burns
HWCs are also subject to the existing
permit requirements under RCRA that
are established at 40 CFR Parts 264, 266,
and 270. Those rules require HWCs
(among other things) to conduct a trial
burn to demonstrate compliance with
applicable emission standards.
Operating conditions are included in
the permit to ensure that compliance is
maintained.
We are proposing to amend those
rules today to refer to the proposed
MACT requirements. Thus, the existing
RCRA emission standards and ancillary
requirements would be superseded by
the proposed MACT standards, with one
exception: destruction and removal
efficiency (ORE).
1. The RCRA DRE Requirement Would
Be Implemented Under RCRA Authority
The destruction and removal
efficiency (DRE) requirement under the
RCRA standards would continue to
apply to all HWCs. Although the DRE
requirement, which is statutory for
incinerators, RCRA § 3004(o)(l)(B),
could be proposed as a MACT surrogate
parameter to minimize organic HAPs by
ensuring good combustion, we are not
doing so. This is because the DRE
standard is complex and impracticable
to self-implement.176 Consequently, the
Agency would continue to apply the
DRE standard under RCRA authority
alone.
2. Coordinating Trial Burns and MACT
Performance Tests
As discussed above, the rule would
allow a source to consolidate a trial
burn test with a comprehensive or
confirmatory test if the trial burn test
were conducted within a year after the
anniversary date for the MACT
performance test.177 If the tests are
consolidated, however, the unified test
must of course satisfy the objectives of
both tests.
We note that the level of Agency
oversight for trial bums is substantially
greater than the oversight that might be
provided for MACT performance tests.
Accordingly, as current practice, the
Agency's implementation procedures
for trial burns will deviate from those
proposed for the MACT performance
tests. As examples, the Agency will
require that the test plan be submitted
more than 60 days in advance of the
planned trial burn test, and extensive
public participation will be provided for
review of the test plan, test results, and
determination of operating limits.
C. Waiver of MACT Performance Testing
for HWCs Feeding De Minimis Levels of
Metals or Chlorine
Today's rule would provide a waiver
of performance testing requirements for
Hg, SVM, LVM, or HC1/C12 for HWCs
that feed de minimis levels of these
metals or chlorine.178. Under the waiver,
a source would be required to assume
that all Hg, SVM, LVM, or chlorine fed
in each feedstream is emitted from the
stack and to document that resulting
emission concentrations do not exceed
the emission standards, considering
stack gas flow rate. Thus, the source
would be required to: (1) establish and
comply with maximum feedrate limits
for total feedstreams for Hg, SVM, LVM,
or chlorine; and (2) establish and
comply with, as a minimum stack gas
flow rate, the flow rate used to
document compliance (by calculation
176 We note that, for this reason, the Agency chose
not to require BIFs operating under interim status
to comply with the DRE standard even though they
were subject to all other emission standards that
would be applicable under a operating permit.
177 If the trial burn were scheduled prior to the
performance test, the source could elect to
consolidate the tests and, thus, move up the
anniversary date for the performance test.
178 Note that the term de minimis means simply
low concentration of metals or chlorine. It does not
denote or imply low risk.
-------
17448
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
rather than emissions testing) with the
emission standard.
To accommodate sources that may
operate under a wide range of gas flow
rates, the rule would allow a source to
establish different modes of operation
with corresponding minimum stack gas
flow rate limits and maximum feedrates
for metals or chlorine. If a source uses
this approach, the operating record must
clearly identify which operating mode is
in effect at all times.
Sources claiming the waiver would be
required to do so in the initial
notification of performance test and
would not be required to establish or
comply with operating limits for the
performance test (i.e., Hg, SVM, LVM, or
HC1/C12) for which the waiver is
claimed. Sources eligible for a waiver
from the Hg standard would not be
required to install a Hg GEMS.
D. Relative Accuracy Tests for OEMS
This section describes the testing
requirements for GEMS proposed today.
Note that GEMS for multi-metals, HC1,
and C12 are proposed to be optional.
Facilities need not perform tests
described below for GEMS they elect not
to use.
A relative accuracy test audit (RATA)
for Hg and multi-metal GEMS would be
required every three years (or five years
for small on-site facilities). RATAs for
CO and O2 GEMS would be required
annually. 17» RATAs for Hg and multi-
metals involve comparing the output of
the GEM to the results of manual
method tests in order to determine the
overall accuracy of the GEM and would
be conducted in conjunction with a
comprehensive test. RATAs for CO and
O2 would be conducted during a
comprehensive test or on the
anniversary date of the previous
comprehensive test.
A relative calibration audit (RCA) for
PM GEMS would be required every 18
months (30 months for small on-site •
facilities). These are similar to a RATA
in that they involve comparing the
output of lie GEM to the results of
manual method tests in order to verify
the validity of the GEM and its
calibration, and would be conducted
whenever a comprehensive or
confirmatory test is performed.
An absolute calibration audit (ACA) is
a test which determines the calibration
error (CE) associated with a GEM. These
audits do so by challenging the analyzer
using gas bottles or solutions of metals
or particulate with known
179 Note that EPA invites comment on waiving the
RATA requirements for CO and 02, instead relying
on quarterly calibration error tests using cylinder
gasses.
concentrations of the compound being
analyzed. ACA's are conducted
quarterly for all GEMS except for multi-
metals, which are conducted annually.
Calibration drift (CD) and zero drift
(ZD) 18° tests are conducted daily using
cylinder gas bottles, filters, or internal
(to the GEMS) calibration standards.
IV. Selection of Manual Stack Sampling
Methods
This section discusses the manual
emission test methods that would be
required for emission tests and
calibration of GEMS and relies heavily
on the BIF methods currently in Part
266, Appendix IX. EPA previously
proposed incorporating many of these
methods in SW-846, Test Methods for
Evaluating Solid Wastes (60 FR 37974,
July 25,1995). Accordingly, both the
BIF and proposed SW-846 numbers are
given.
The emission test method for D/F
would be the proposed SW-846 Method
0023A (60 FR 37974, July 25, 1995). It
is identical to the BIF Method 23 in
Appendix IX of Part 266 except Method
0023A requires that collection
efficiencies be determined for both the
particulate and sorbent. BIF Method 23
is the same as the Air Method 23 in Part
60, Appendix A. Method 23 determines
the efficiency off the sorbent only and
assumes the same recovery off the
particulate as from the sorbent. We are
also proposing today to make a
conforming change to the BIF rule to
require use of Method 0023 A rather
than Method 23.
It is proposed that BIF Method 0012
(SW-846 method 0060) be used as the
manual method test for Hg. The
proposed manual emission test method
for the SVM and LVM standards is BIF
Method 0012 contained in section 3.1 of
Appendix IX, Part 266 (SW-846 method
0060). This method is also commonly
known as Air Method 29.
For compliance with the HC1/C12
standard, the rule would use BIF
Methods 0050, 0051, and 9057
contained in section 3.3 of Appendix IX,
Part 266, as the manual test method
(SW-846 would retain the same
numbering). These methods are
commonly known as Air Method 26A,
found in Appendix A of Part 60.
Existing § 63.7 describes procedures
for allowing the use of alternative test
methods for MACT sources. This
procedure involves using Method 301 of
Part 60, Appendix A, to validate the
proposed method. The data from the
Method 301 validation is submitted to
EPA. EPA then decides if the proposed
180 Note that EPA invites comment on whether the
ZD requirements should be deleted.
method is acceptable. Absent this
approval under § 63.7 procedures,
alternate methods cannot be used.
V. Notification, Recordkeeping,
Reporting, and Operator Certification
Requirements
Today's proposed rule would
establish several notification,
recordkeeping, and reporting
requirements for HWCs. This section
discusses the applicability to HWCs of
existing requirements in §§ 63.9 and
63.10 and Parts 264, 265, 266, and 270.
In addition, we discuss in this section
new requirements that would apply
specifically to HWCs. Finally, we
discuss whether operator certification
requirements should be promulgated.
A. Notification Requirements
HWCs would be required to submit
the following notifications:
• Initial notification. The initial
notification requirements of existing
§63.9(b) would apply. These
notifications are intended to alert
regulatory officials that a source is
subject to the regulations. Even though
all existing HWCs have already notified
the Administrator of their hazardous
waste activities under RCRA
requirements, and new HWCs must
notify the Administrator and obtain an
operating permit before commencing
construction, these RCRA-required
notifications will not always be received
by the same regulatory officials
implementing the MACT standards. For
example, when a state is authorized for
Title V permitting, various state
regulatory authorities, including local
air boards, could be the implementing
authority. In contrast, RCRA regulations
are implemented by Agency and state
officials. Accordingly, to ensure that all
appropriate regulatory officials are
apprised that a HWC is subject to the
MACT and RCRA regulations, we are
proposing to retain the initial
notification requirement under § 63.9(b).
• Notification of performance test and
CMS performance evaluation. This
notification includes the planned test
date, performance test plan (to
demonstrate compliance with
emissions), CMS performance
evaluation plan, and quality assurance
plan. It is required by existing § 63.9(c),
except that all sources must submit their
test plan and CMS performance
evaluation plan for review and
approval.
• Notification of compliance. This
notification includes results of
performance test and CMS performance
evaluation and certification by the
owner and operator that the source is in
compliance with the applicable
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17449
standards. It is similar to that required ,
by existing § 63.9(h) with several
important differences. Under today's
rule, a source must notify that it is
actually in compliance with all
applicable standards, not merely
identify its status with respect to
compliance as allowed by § 63.9(h). In
addition, paragraphs (h)(2) (D) and (E)
requiring the source to identify the type
and quantity of pollutants emitted and
an analysis of whether the source is a
major or area source are not applicable
to HWCs. This is because today's
proposed rule would apply to all HWCs
irrespective of whether it meets the
definition of a major source. Finally,
today's rule would require the
notification to be submitted 90 days
after completion of testing, rather than
60 days as now required by paragraph
.
Request for extension of time to
submit a notification of compliance. A
notification for a time extension for
initial compliance is provided by
§ 63.9(c). Today's rule would require
sources to submit a notification of
compliance after each performance test
(both comprehensive and confirmatory)
and alloxv requests for time extensions .
to submit those notifications.
• Request for a time extension to
consolidate a performance test with a
trial burn. Today's rule would allow a
source to request to consolidate a trial
burn with a performance test if the trial
burn test date is no later than 12 months
after the performance test anniversary
date.
To summarize applicability of
existing § 63.9 notification requirements
and to assist the regulated community
in understanding the applicable
requirements, the following list is
provided as guidance:
• Paragraph (a) (Applicability and
general information) applies.
• Paragraph (b) (Initial notifications)
applies as discussed above.
• Paragraph (c) (Request for extension
of compliance) applies for the purposes
discussed above.
• Paragraph (d) (Notification that
source is subject to special compliance
requirements) applies.
• Paragraph (e) (Notification of
performance test) applies as discussed
above.
• Paragraph (f) (Notification of
opacity and visible emission
observations) is not applicable because
the rule would establish a PM emission
standard and other compliance/
monitoring requirements in lieu of
opacity and visible emission standards.
• Paragraph (g) (Additional
notification requirements for sources
with CMS) applies.
• Paragraph (h) (Notification of
compliance status) applies with the
caveats discussed above.
• Paragraph (i) (Adjustments to time
periods or postmark deadlines for
submittal and review of required
communications) applies.
• Paragraph (j) (Change in
information already provided) applies.
The rule would require the following
additional notification requirements:
• Small quantity on-site burner
exemption. See discussion in Part Six,
Section II.A.l.
• Pre-trial burn period (shakedown).
See discussion in Part Six, Section
II.F.1.
B. Reporting Requirements
HWCs would be required to submit
the following reports:
• Excessive AWFCO report. See
discussion in Part Five, Section II.E.l.
• ESV opening report. See discussion
in Part Five, Section II.E.l.
For guidance to the regulated
community, the applicability of the
existing reporting requirements under
§§ 63.10(d) (General reporting
requirements), 63.10(e) (Additional
reporting requirements for sources with
CMS), and 63.10(f) (Waiver of
recordkeeping or reporting
requirements) would be as follows:
• Paragraph (d)(l) applies. This
paragraph references the reporting
requirements in the specific standards
for a source category, in this case
proposed Subpart EEE.
• Paragraph (d)(2) (Reporting results
of performance tests) applies, except
that the report may be submitted up to
90 days after completion of the test.
• Paragraph (d)(3) (Reporting results
of opacity or visible emission
observations) does not apply because
the rule would not regulate opacity or
visible emissions.
• Paragraph (d)(4) (Progress reports)
applies.
• Paragraph (d)(5) (Periodic startup,
shutdown, and malfunction reports; and
immediate startup, shutdown, and
malfunction reports) does not apply.
Given that HWCs could not burn
hazardous waste under the proposed
rule except in compliance with all
applicable emission standards,
operating limits, and CMS performance
specifications, the rule would not
require a startup, shutdown, and
malfunction plan as required by
§ 63.6(e)(3) for other MACT sources.
There will be no excess hazardous waste
emissions during these periods (unless
the HWC violates the standards) and the
Agency does not need information about
how quickly a HWC is able to correct a
malfunction or come back into
compliance again so that it may resume
hazardous waste burning.181
• Paragraph (e)(l) (General) applies.
• Paragraph (e)(2) (Reporting results
of CMS performance evaluations)
applies.
• Paragraph (e)(3) (Excess emissions
and CMS performance report and
summary report) does not apply because
HWCs cannot burn hazardous waste
except in compliance with all
applicable standards.
• Paragraph (e)(4) (Reporting
continuous opacity monitoring system
data produced during a performance
test) does not apply because COMs are
not required in this proposal.
• Paragraph (f) (Waiver of
recordkeeping or reporting
requirements) would not apply because
the bases for considering the waiver are
not relevant to HWCs as follows: (1)
Recordkeeping and reporting should not
be waived because "the source is
achieving the relevant standards"
because recordkeeping and reporting
would be the primary means of
compliance assurance for the HWC
rules; (2) recordkeeping and reporting
should not be waived during a time
extension because the requirements
would not apply until a HWC submitted
the initial notification of compliance'
irrespective of whether a time extension
were granted; and (3) recordkeeping and
reporting should not be waived if a time
extension is granted for a subsequent
notification of compliance (because the
source will be burning hazardous waste
under the standards).
C. Recordkeeping Requirements
Existing § 63.10(b)(l) requires MACT
sources to keep the records discussed
below for at least five years from the
date of each occurrence, measurement,
maintenance, corrective action, report,
or record. At a minimum, the most
recent two years of data must be
retained off-site. The remaining three
years of data may be retained on site.
Such files may be maintained on:
microfilm, a computer, computer floppy
disks, optical disk, magnetic tape, or
microfiche.
'si One exception to this is the operation of
cement kilns when the hazardous waste feed has
been cut off and there is no hazardous waste
remaining in the combustion chamber. In this
situation, the HWC emission standards, operating
limits, and CMS performance specifications would
not apply. Given that the Agency plans to propose
MACT standards for cement kilns that do not burn
hazardous waste, however, a cement kiln that is
temporarily not subject to today's proposed '
standards because the waste feed has been cutoff
(and there is no hazardous waste remaining in the
combustion chamber) would nonetheless remain (or
become) subject to any MACT standards the Agency
may promulgate.
-------
17450
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
1. Information Required in the
Operating Record
The rule would require HWCs to
record the following in the operating
record:
• Comprehensive test results used to
determine operating limits. See
discussion in Part Five, Section II.B.
• All operating parameter limits
established. See discussion in Part Five,
Section II.C.
• Operating data which substantiates
compliance, including minute-by-
minuf e operating parameter data,
including feedstream; arid minute-by-
minute GEM data. See discussion in Part
Five, Section II.B.
• Documentation for performance test
waiver. See discussion in Part Five,
Section III.C.
• Description of and operating data
substantiating compliance with
provisions to limit combustion fugitive
emissions. See discussion in Part Five,
Section H.D.
• For each occurrence of an
exceedance of a GEM or operating
parameter limit, including'what
operating parameter of GEM limit was
violated: the cause of the violation, and
what corrective action was taken to
ensure the violation will be prevented
in the future. See discussion in Part
Five, Section II.E.l.
• For each ESV opening:
documentation that the ESV opened, the
reason for the opening, and corrective
measures taken to minimize the
frequency of openings. See discussion
Part Five, Section II.E.2.
• ESV operating plan. See discussion
Part Five, Section H.E.2.
• GEM quality assurance document,
including: definition of compliance'
with the calibration and zero drift
specifications, and how relative
accuracy and absolute calibration audits
will be performed. See discussion Part
Five, Section II.F:I.
• Feedstream Analysis Plan,
including: the parameters for which
each feedstream will be analyzed to
ensure compliance; whether the owner
or operator will obtain the analyses by
performing sampling and analysis or by
other methods; how the analysis will be
used to document compliance; the test
methods used; the sampling method
used; and the frequency of testing. See
discussion in Part Five, Section II.F.2.
• Other Continuous Monitoring
Systems (CMS), including:
manufacturer's written specifications for
installation, operation, and calibration
of a CMS; and technical specifications
of CMS, such as spans and percent
error. See discussion in Part Five,
Section II.F.3.
In addition, HWCs would be required
to develop and keep in the operating
record a feedstream management plan
that enables the source to maintain
compliance with CEM-monitored
emission standards. Although a facility
using a GEM for compliance would not
be required to comply with feedrate
limits, the owner and operator would be
required to develop a feedstream
management plan (and include it in the
operating record) that will enable the
source to know the feedrate in all
feedstreams of Hg (as well as other
metals and chlorine if the source elects
to use a GEM for compliance
monitoring) at all times to minimize
automatic waste feed cutoffs and
exceedances of the emission standard.
Knowledge of Hg (and other metals and
chlorine) concentration of feedstreams
can come from the waste generator,
supplier, or other information, and need
not be obtained by sampling and
analysis by the burner. If the source
experiences frequent AWFCOs or
exceedances, enforcement officials will
determine if a feedstream management
plan is in place. If the plan is
determined to be inadequate, the
Director may require that it be
upgraded, taking into account whether a
good faith effort has been made to
develop a plan, even if the plan is
determined to be inadequate.
Note that RCRA/HSWA already ;
requires the facility owner to certify no
less than annually, that the facility has
a waste minimization program in place,
and the certification must be maintained
in the facility's operating record. The
facility owner is encouraged to
coordinate the development of the
feedstream analysis plan and the
feedstream management plan with the
facility's waste minimization program.
EPA published Interim Final "Guidance
to Hazardous Waste Generators on the
Elements of a Waste Minimization
Program in Place," (1993) and the
"Pollution Prevention Facility Planning
Guide" (1993), which provide
information to facility owners on how to
prepare analyses of waste streams and
options for reducing wastestreams using
alternative pollution prevention/waste
minimization measures. Information on
these documents can be requested by
calling the RCRA hotline at 1-800-424-
9346.
Many states provide free pollution
prevention/waste minimization
technical assistance that may aid
facilities in the development of
pollution prevention/waste
minimization plans. At least 20 states
have requirements for certain facilities
to prepare pollution prevention/waste
minimization plans. As noted elsewhere
in today's rule, facilities can get further
information on available technical
assistance by contacting the National
Pollution Prevention Roundtable in
Washington, D.C. at (202) 466-7272, or
from Enviro$ense, an electronic library
of information on pollution prevention,
technical assistance, and environmental
compliance,'that can be accessed by
contacting a system operator at (703)
908-2007, via modem at (703) 908-
2092, or on the Internet at http://
wastenotiinel.gov/enviro-sense.
2. Applicability of § 63.10
Recdrdkeeping Requirements
The applicability of the existing
recordkeeping requirements of § 63.10
would be as follows:
• Paragraph (a) (Applicability and
general information) applies, except for
(a) (2) that exempts sources that are
operating under a compliance
extension. This is because sources that
receive a time extension to submit the
initial notification of compliance would
not be subject to any of the proposed
standards. Further, sources that receive
an extension for a subsequent
notification of compliance need to
comply with recordkeeping and
reporting requirements to provide
compliance assurance given that they
are burning hazardous waste during the
extension.
• Paragraph (b) (General
recordkeeping requirements) applies,
except for (b)(2) (iv)—(vi) that pertain to
actions during malfunctions, and (b)(3)
regarding recordkeeping for
applicability determinations.
• Paragraph (c) (Additional
recordkeeping requirements for sources
with CMS) would apply, except for
(c)(6)-(8), (c)(13), and (c)(15) that
pertain to malfunctions.
3. New Recordkeeping Requirements
The rule will also require
recordkeeping requirements for the
following:
• Comparable fuels. Sampling and
analysis plan, including revisions; and
certifications from burners. Under
§ 261.4 records will be kept for as long
as the generator manages a comparable
fuel, plus five years. See discussion in
Part 6, Section I.E.6.
• Comparable fuels. Results of
sampling and analysis; and records of
off-site shipments for five years. See
discussion in Part 6, Section II.E.6.
• Small quantity on-site burner
exemption. Under § 266.108, records
will be kept for 3 years. See discussion
in Part Six,. Section II.D.
• Regulation of residues. Under
§ 266.112, records will be kept until
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17451
closure. See discussion in Part Six,
Section H.D.
D. Operator Certification
The Agency notes that section 129 of
the Clean Air Act requires EPA to
develop and promulgate a model
program for the training and
certification of municipal waste
conibustor (MWC) and medical waste
combustor (MWI) operators.
Accordingly, the Agency has
promulgated operator certification and
training requirements for MWCs and has
proposed requirements for MWIs. The
Agency is today requesting comment on
whether similar requirements are
necessary and appropriate for operators
ofHWCs.
The MWC and MWI requirements call
for (in part) full operator certification of
all shift supervisors and chief facility
operators by the American Society of
Mechanical Engineers (ASME) or a State
certification program. In addition, a
least one of the following persons is
required to he on duty at all times
during which the unit is combusting
waste: a fully certified chief facility
operator; a fully certified shift
supervisor; or a provisionally certified
control room operator.
We note that the ASME has recently
established a Standard for the
Qualification and Certification of
Hazardous Waste Incinerator Operators
(ASME QHO-1-1994, January 31,1995).
We request comment on whether: (1)
operator certification requirements are
necessary for HWCs; and (2) the ASME
standard, or an equivalent State
certification program) is appropriate and
sufficient
The ASME standard has been
developed specifically for hazardous
waste incinerators. We are not aware of
an equivalent standard for operators of
cement kilns and lightweight aggregate
kilns that burn hazardous waste. We
note, however, that the Cement Kiln
Recycling Coalition has stated that it is
committed to the development of an
operating training and certification
program for its member facilities.182 We
invite comment and information from
owners and operators of waste-burning
kilns regarding the need for a
certification standard and the status of
development of a standard for such
combustors.
VI. Permit Requirements
The rulemaking approach in today's
proposal, to promulgate final standards
under joint RCRA/CAA authority, raises
<" Latter from Craig Campbell, CKRC, to Ronald
Bastlan, Chairman, ASME QHO, dated January 5,
1994.
some challenging implementation
questions. In this section, permitting
strategies are discussed. EPA requests
comment on how these strategies can be
further simplified while retaining basic
environmental protection goals.
A. Coordination ofRCRA and CAA
Permitting Processes
The rulemaking approach chosen for
today's proposal is to promulgate the
final standards for hazardous waste
combustors under joint RCRA/CAA
authority. However, the standards will
only appear under 40 CFR Part 63
(Clean Air Act section). The RCRA
regulations in 40 CFR Parts 264 and 266
will make reference to these Part 63
standards, thereby incorporating them
as RCRA standards as well. Thus,
legally, the new standards will be part
of both the RCRA and CAA regulations
and both regulatory programs (RCRA &
CAA) will have an obligation to address
these standards in permits issued under
their authority.
Although the Agency believes that a
single permit would be ideal to
implement these two programs, today's
proposed approach does not always
eliminate the need for two separate
permits. However, it does provide a
variety of options for State
implementation. By using both the CAA
and RCRA authorities, today's approach
provides maximum flexibility for
permitting authorities at the Regional,
State, and/or local levels to coordinate
the issuance of permits and enforcement
activities in the way which most
effectively addresses their particular
situation.
Currently, combustion facilities are
required to obtain two permits; a RCRA
permit and a CAA permit. Although it
is EPA's long term goal is to have one
permit that would address both RCRA
and CAA requirements, it is difficult
because (1) different pieces of the rule
rely on different authorities, and (2)
significant coordination is needed
between Regional, State, and local
authorities. After careful consideration,
EPA's goal in today's proposal is to
coordinate as much as possible between
the two permitting programs to avoid
duplication of effort, inconsistent
requirements, and redundant
procedures.
EPA explored the possibility of
requiring combustion facilities to have
only one EPA permit issued under
either RCRA authority or CAA
authority. Promulgating these standards
in the CAA regulations and requiring
only a CAA permit looked promising
because RCRA allows EPA to defer
RCRA regulation to other authorities
administered by EPA, if RCRA core
values are covered by the other federal
requirements (RCRA Section
1006(b)(l)), in this case, the CAA.
However, EPA believes that several
RCRA core requirements (e.g., corrective
action, omnibus conditions, DRE, etc.)
cannot be addressed in a CAA permit,
since the CAA does not provide the
legal authority to address them.
Promulgating these requirements
under RCRA authority and issuing only
a RCRA permit is not possible because
the CAA does not allow permits for
major sources to be waived. As
previously discussed, all facilities
covered by this rulemaking will be
considered major sources. Also, CAA
specific concerns (e.g., acid rain, criteria
pollutants, etc.) would not be addressed
in a RCRA permit.
EPA considered placing the revised
air emission standards in the CAA
regulations and including a RCRA
permit-by-rule provision that would
defer to the CAA permit. Under this
option, the CAA regulations would
contain the air emission requirements
and the CAA permit would contain the
emission standards. In addition, a
separate RCRA permit would address
RCRA-specific concerns (e.g., corrective
action, omnibus conditions, DRE,
storage, etc.). This approach would
avoid duplicating air emission
requirements in both permits. EPA is
not proposing regulatory language that
would require this approach because
there is concern that it might limit the
permitting flexibility of the.
implementing agencies by specifying
which program would be required to
address air emissions. Some states have
expressed concerns about this approach.
Many states—for example, those that
regulate air emission standards under
their hazardous waste program—may
find it difficult to implement this
option; also, some states were
concerned about the ability of local
permitting programs being solely
responsible for the air emissions
. permitting for these facilities. On the
other hand, the flexibility EPA is
suggesting in today's proposal would
not preclude states from using this
permitting approach.
More broadly, EPA has not specified
any one permitting approach in today's
proposal. The flexibility the Agency is
proposing would allow states to decide
which permitting approach to take. The
important things are that all substantive
requirements are met and that a timely
and full opportunity for public
involvement is provided during the
permitting process.
EPA has identified a range of possible
permitting scenarios under today's
proposed approach. Some examples of
-------
27452
Federal Register /Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
coordinated efforts between the RCRA
and CAA programs include: (1) issuing
a single permit using both (or either)
RCRA and CAA authority, and (2)
issuing two separate permits with close
coordination between the two programs.
In the first example, the two
permitting programs would work
together to issue one permit that meets
all the requirements of both programs.
This joint permit would include CAA-
specific items (e.g., acid rain, criteria
pollutants, etc.), RCRA-specific items
(e.g., corrective action, omnibus
conditions, DRE, etc.), and items
common to both programs (e.g., air
emission standards, etc.). The permit
would be issued under joint authority
and signed by the Director(s) of both
programs. This scenario is likely to be
most appropriate where a State has
authority for both programs and the two
programs have experience working
together. This approach could also be
implemented by using the CAA in
combination with the RCRA permit-by-
rule provision as discussed above.
In the second example, the two
permitting programs (one responsible
for RCRA, and one responsible for CAA)
would coordinate their permitting
efforts. Each program would issue a
permit. The requirements common to
both programs (e.g., stack emission
standards, etc.) would be included in
one permit and the other permit would
incorporate the common requirements
by reference. This approach would
avoid duplicative and conflicting
requirements. In this example, each
permit would go through the applicable
procedures for issuance. To coordinate
permit issuance, all public participation
requirements (notices, comments,
hearings, etc.) could be combined.
Under this approach permits would be
subject to applicable appeal procedures
and enforcement provisions under each
program; however, EPA would not
expect to enforce under both permits.
The appropriate enforcement response
will be determined on a case-by-case
basis. We invite comment on this point
in particular.
EPA will work with the States to
identify issues relating to streamlining
the permitting programs and to develop
any needed guidance materials or model
processes. Additionally, EPA will
continue to pursue a mechanism to
issue one permit that would address
both RCRA and CAA requirements.
An Agency-wide initiative led by the
Permits Improvement Team (PIT) has
recommended ways to improve
permitting activities for all
environmental programs. Under this
initiative EPA continues to seek the best
ways to permit facilities throughout its
various media programs. The approach
in today's proposal is consistent with
the current direction of the PIT, which
suggests avoiding duplication of effort
by incorporating the air emission
standards into one permitting program.
EPA is committed to harmonizing these
two permitting processes as much as
possible for the implementation of
today's proposal.
B. Permit Application Requirements
EPA reviewed information required
for permit applications under both the
CAA (§ 70.5) and RCRA (Part 270) to
identify any duplication that could be
eliminated and to determine whether
any CAA or RCRA permit application
requirements for hazardous waste
combustors could be combined.
Historically, determinations for permit
approval for facilities regulated under
the CAA generally focused solely on the
efficiency of the air pollution control
device (APCD). Conversely, the basis for
permit approval under RCRA has
traditionally been more specific and
related to details of the combustion unit
and process (for example, design
characteristics of the unit, variability of
the waste burned, information on the
type of waste to determine the effect it
may have on the quality of the operation
of the unit over time, etc.). Specific
information requirements are listed in
§§ 270.15-270.26 (see specific technical
information requirements in § 270.19 for
incinerators and § 270.22 for BIFs). For
these reasons, EPA has concluded that
the current Part B information
requirements and the information
requirements in the CAA regulations are
not duplicative and is proposing that
both be retained under the existing
regulations to assure that all RCRA and
CAA concerns are addressed.
Although some of the general
information required under § 270.13,
Contents of Part A of the RCRA permit
application, is also requested in § 70.5
of the CAA permit application
requirements, EPA believes that because
this information is so minimal, it would
not be a burden for the applicant to
duplicate it on two separate
applications. Section 270.13 requires
further information under the Part A,
such as a scale drawing of the facility
showing the location of all past, present,
and future TSD areas, specifications of
the hazardous waste listed or designated
under 40 CFR Part 261 to be handled at
the facility and a list of all permits or
construction approvals received or
applied for under other programs, to list
a few. In addition, standards relating to
the overall operation of the facility are
listed under Part B (§270.14). These
standards include, but are not limited
to, chemical and physical analyses of
the hazardous waste and hazardous
debris to be handled at the facility,
description of the security procedures,
contingency plans, closure and post-
closure plans (including cost estimates)
and a description of the continuing
training programs. Such standards are
not required in the application for a
CAA permit. EPA has therefore
concluded that it would be reasonable
to keep the application requirements
where they now exist and cross-
reference them where appropriate.
C. Clarifications on Definitions and
Permit Process Issues
Because of the incorporation of the
technical standards into both the RCRA
and CAA regulations, as described
previously, both RCRA and CAA
permitting procedures are applicable.
For issues such as the meaning of the
term "construction", there could be
confusion since the definitions and
interpretations under one Act differ
from those under the other. Our intent
is not to reconcile these issues on a
national basis but to continue to let both
apply. As in the past, sources regulated
under both Acts will need to coordinate
with both RCRA and CAA permitting
authorities to see how these procedures
apply to them. We note in passing that
this approach means that the most
restrictive limitations or processes will
generally govern.
The Agency requests comment on
whether these issues should be
addressed at the national level. EPA's
current preference is not to do so, but
to leave flexibility for the states and
EPA Regions to address these issues.
1. Prior Approval
RCRA and CAA are similar in that
both require EPA approval before
construction or reconstruction of a
facility (generally) (Sections 61.07, 63.5,
270.10(f)). Both programs use
hypothetical emissions data to make the
construction approval decision. If a
facility is existing before the effective
date of the final regulation, both RCRA
and CAA require notification of
operation but do not require approval of
the construction that has already
occurred (Sections 60.7,
266.103(a)(l)(ii)). (Modification of a
permitted facility also requires prior
approval.)
2. 50 Percent Benchmark
. RCRA and CAA both classify a
modification of a facility that costs more
than 50 percent of the replacement cost
of the facility as "reconstruction".
However, the significance of this term is
different under the two statutes. Under
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17453
RCRA, the issue of reconstruction is
relevant to interim status facilities. An
interim status facility planning
modifications which constitute
reconstruction must receive a RCRA
permit prior to construction of the
modifications and operation
(§ 270.72(b)). Under the CAA,
reconstruction subjects the facility to
standards applicable to new facilities
(§§60.15, 60.488, and 63.5).
3. Facility Definition
RCRA and CAA define "facility"
differently. This definition has bearing
in determining the value of the facility
with respect to the 50 percent rule on
modifications just discussed. CAA
defines facility as the entire industrial
process at the site (profit making
productive process and pollution
control devices), while RCRA for
purposes of reconstruction refers to a
"comparable entirely new hazardous
waste facility" (Section 270.72)
excluding other industrial processes at
the site from consideration in the cost
of the existing facility. For a site where
the only activities are RCRA hazardous
waste activities, the two definitions are
identical. However, sites with non-
RCRA industrial activities will have
differing cost figures for each rule.
Therefore, the two programs have
differing determinations of how much
reconstruction can occur before the 50
percent benchmark is exceeded.
However, EPA believes this difference
should not constitute a problem, since
the reconstruction determination has
different applications under each Act.
The RCRA definition should be used for
the RCRA application to changes during
interim status, and the CAA definition
should be used when determining
applicability of new versus existing
MACT standards.
4. No New Eligibility for Interim Status
This joint CAA/RCRA proposed
rulemaking revises emission standards
for incinerators and BIFs and hence
amends the original incinerator and
industrial furnace rules that were
finalized in 1981 and 1991, respectively.
Because these rules established the date
on which incinerators and BIFs were
first subject to a permit requirement, the
effective dates of those rules created the
only opportunity for interim status
eligibility. § 270.10(e)(l)(A)(ii). The
interim status windows that occurred in
1981 and 1991 thus will not and legally
cannot be modified by this rule. Of
course, facilities currently burning
wastes that become newly listed under
other, future rules would still be able
under existing law to qualify for interim
status (§ 270.42(g)).
To avoid the possibility that readers
of Part 63 might be unaware of their
obligations under RCRA, EPA has
inserted a note into Section written
Section 63.1206 to alert them to this
point. This note states: "an owner or
operator wishing to commence
construction of a HWI or hazardous
waste-burning equipment for a cement
kiln or lightweight aggregate kiln must
first obtain some type of RCRA
authorization, whether it be a RCRA
permit, a modification to an existing
RCRA permit, or a change under already
existing interim status. Please see 40
CFR Part 270."
5. What Constitutes Construction
Requiring Approval
RCRA and CAA both have restrictions
requiring approval prior to construction.
The definition of construction under the
RCRA regulations and associated
interpretations differ from the CAA
approach to defining construction (case-
specific call, see Sections 60.5, 61.06)
Facilities need to comply with both and
should be consulting with applicable
permitting authorities to assure
appropriate site-specific interpretations.
We believe the RCRA construction
definition is generally broader (more
restrictive) and thus will govern in most
cases. The Agency believes retaining the
two differing definitions will not cause
problems since they are already being
applied concurrently. Also, the Agency
feels that creating a third construction
definition for this small subset of the
RCRA and CAA facilities would create
more confusion than it would eliminate.
D. Pollution Prevention/Waste
Minimization Options
EPA believes pollution prevention/
waste minimization measures may
provide facilities additional flexibility
in meeting MACT standards. Pollution
prevention/waste minimization
measures have been used by many
companies to modify processes and
install new or improved technologies
which reduce or eliminate the volume
and/or toxicity of hazardous wastes
generation that would otherwise enter
combustion unit feedstreams, or be
treated or disposed of in some other
fashion. EPA is soliciting comment on
two pollution prevention/waste
minimization options for reducing or
eliminating hazardous constituents that
enter on-site as well as commercial
combustor feedstreams, and that can be
considered in the definitions of changes
in facility operating parameters and/or
new or improved control technologies
for meeting MACT standards.
The first option would require all
facilities to provide adequate
information on alternative pollution
prevention/waste minimization
measures that reduce hazardous
constituents entering the feedstream,
particularly the most persistent,
bipaccumulative, and toxic constituents,
in all permit applications. EPA believes
this approach is consistent with the
national policies of the Pollution
Prevention Act of 1990, CAA, RCRA,
and over 20 states who encourage or
require pollution prevention plans.
Facilities are encouraged to reference
existing EPA documents, such as the
Interim Final "Guidance to Hazardous
Waste Generators on the Elements of a
Waste Minimization Program in Place,"
(May 1993), which provides a guide for
developing pollution prevention/waste
minimization programs. Facilities are
also encouraged to reference EPA's
"Pollution Prevention Facility Planning
Guide" (May 1992), "An Introduction to
Environmental Accounting As A
Business Management Tool" (June
1995), and "Setting Priorities for
Minimization of Combusted Hazardous
Waste" (November 1995), and to contact
the National Pollution Prevention
Roundtable, and state pollution
prevention technical assistance
programs for additional pollution
prevention resources. These documents
were published as aides to facility
owners in preparing analyses of
pollution prevention/waste
minimization measures. EPA believes
this approach provides maximum
flexibility to facilities for identifying
controls through the application of
processes, or systems (including
pollution prevention/waste
minimization measures) for reducing
emissions.183
EPA believes in many cases, facilities
may already be required or encouraged
to prepare this information in the more
than 20 States which have pollution
prevention facility planning
requirements already in place. EPA
believes this approach will promote
consistency in States which are
requiring facilities to develop pollution
prevention/waste minimization plans as
a basis for developing multi-media
permits. This approach will enhance,
without duplicating, the requirements
in this proposal for facilities to prepare
a feedstream analysis plan and a
feedstream management plan. In cases
where this information has been already
developed by the facility in accord with
State requirements within 18 months
prior to the date of application, no
IBS Under the Clean Air Act Section 112[d)(2),
MACT standards include, among other things,/
process changes, substitution of materials or other
modifications.
-------
27454
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
additional pollution prevention/waste
minimization information will be • "
required as part of the permit
application.
• In the second option, EPA proposes to
give EPA Regions and States discretion
to make case by case determinations
regarding whether a facility must
provide adequate information for
reducing measures, including pollution
prevention/waste minimization
measures, that will minimize hazardous
constituents entering the feedstream.
EPA believes this determination should
be made based on the facility's ability to
verify that they have a waste
minimization program in place as
required under RCRA, the extent to
which the facility has reported pollution
prevention information in annual Toxic
Release Inventory reports (for facilities
subject to TRI reporting requirements),
and the extent to which information has
already been prepared under existing
state pollution prevention planning
requirements, or in conjunction with
State or local pollution prevention
technical assistance programs.
EPA believes this option provides the
regulated community and States broad
flexibility to integrate existing pollution
prevention/waste minimization
programs into the objectives of this
rulemaking. States, universities and
local governments operate over 200
technical assistance programs that work
cooperatively with companies to
identify waste minimization options to
reduce waste generation and
management. Some states combine this
approach with compliance assistance,
and a few have in place enforceable
waste minimization requirements
ranging from mandatory waste
minimization plans to incorporating
waste minimization opportunities into
permitting, inspection and/or
enforcement activities. As noted
elsewhere, facilities can contact the
National Pollution Prevention
Roundtable in Washington, B.C. at (202)
466—7272 for further information on
technical assistance opportunities, or
Enviro$ense, an electronic library of
information on pollution prevention,
technical assistance, and environmental
compliance. Enviro$ense can be
accessed by contacting a system
operator at (703) 908-2007, via modem
at (703) 908-2092, or on the Internet at
http://wastenot.inel.gov/enviro-sense.
E. Permit Modifications Necessary To
Come Into Compliance With MACT
Standards
This Notice of Proposed Rulemaking
would require facilities to come into
compliance with a number of new
MACT emission standards within three
years following final promulgation of
this rule. Some facilities would need to
perform facility modifications to come
into compliance with the MACT
standards through changing operating
parameters or adding new or improved
control technology(ies) to reduce
emissions. For example, incinerators
that currently operate above the MACT
PM emissions standards would
potentially need to add or modify
electrostatic precipitators (ESP) or
baghouses to reduce emissions.
Incinerators with a need to reduce
dioxin emissions may need to look into
establishing better controls on
temperature or the use of carbon
injection. LWAKs with potential
exceedances in acid gas emissions may
need to add control technology such as
wet scrubbers. These facility changes
may need to be added to a facility's
existing RCRA permit through a permit
modification. The facility, in this case,
would need to apply for and receive
approval for a permit modification
(unless it is a class 1 modification)
before commencing with its proposed
change (s).
This rule is being proposed under
both RCRA and the Clean Air Act
Amendments. With regard to coming
into compliance with these proposed
standards, the Clean Air Act creates a
mandatory compliance deadline of three
years for facilities subject to these
regulations (with a one year allowance
for an extension granted on a case-by-
case basis). The MACT standards are
self-implementing in that they take
effect in the absence of a CAA permit.
As mentioned earlier in this notice, the
Agency is also taking comment on
whether it would be appropriate to
move up the compliance date of this
rulemaking from the proposed three
year timeframe following promulgation
to a timeframe closer to many RCRA-
based regulations, that of six months to
a year. The Agency is taking comment,
as well, on any other timeframes which
can be considered both technically and
legally feasible.
However, these sources also hold
RCRA permits (or operate under interim
status) which likely would have to be
modified as a result of efforts to comply
with the MACT emission standards.
With respect to facilities with RCRA
permits, EPA is concerned that these
facilities could submit a high number of
Class 2 or Class 3 permit modification
requests within the three year window
before MACT compliance begins. This
large influx could potentially lead to
difficulties in timely processing of
modification requests by EPA or State
agencies. As a result, facilities
potentially would not have conformed
their RCRA permits to reflect the
changes needed to meet the MACT
standards. The Agency anticipates that
many of the permit modification
requests will contain either identical or
similar proposed changes, given the
similarities in incinerator, cement kiln,
and LWAK design and operation. Given
the large number and the potential jbr
duplication of modification requests,
and the desire to achieve timely
emissions reductions, the Agency is
considering options that will streamline
the RCRA permit modification process
to ensure that necessary modifications
are made expeditiously, particularly in
light of the fact that these standards
could potentially become effective in a
shorter period of time, depending on
comments received from the public on
this proposed rulemaking.
In today's proposal, we are seeking
comment on five main options (referred
to as modification options 1-5) which
propose various mechanisms to
expeditiously authorize changes made
to comply with this rule. Also, the
Agency is seeking comment on three
approaches to address whether EPA or
a state would process necessary permit
modifications (referred to as
implementation approaches 1-3) where
a state is authorized to issue RCRA
incineration and BIF permits but is not
authorized to implement the new
combustion rule. This situation should
arise only where a state does not adopt
the necessary provisions of the new rule
within the time required by 40 CFR Part
271.21. EPA strongly urges states to
adopt this rule, once finalized,
expeditiously in order to streamline the
processing of necessary modifications.
This notice seeks comment on which
modification option or combination of
modification options would be the most
viable. The Agency is also taking
comment on any combination of the
above implementation approaches and
options if an intermediate option and
implementation approach combination
seems more appropriate. Under the
current RCRA permit modification
scheme, a permitted facility would refer
to Appendix I of 40 CFR 270.42 to
determine if its proposed modification
is classified in the modifications table.
A modification may rank as Class 1,2,
or 3 (see 53 FR 37912 (Sept. 28,1988)).
A higher modification class signifies an
increased significance of the facility
change which is accompanied with a
commensurate increase in the level of
public participation. Facilities can
proceed with most Class 1 changes
without notifying the Agency, though
some Class 1 modifications require prior
Agency approval. Owners and operators
must, in all cases, notify the public and
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17455
the authorized Agency once they have
made a Class 1 modification. For cause,
the Agency may reject any Class 1
modification.
Class 2 modifications provide for
considerably more participation by both
the facility and the public including an
informational meeting between the
owner and the public regarding the
owner's request prior to the Agency
decision. Class 3 modifications
substantially alter the facility or its
operations. As a result, they require the
most Agency review and are subject to
more public participation requirements
than a Class 1 or 2 modification,
including the full part 124 procedures
for processing draft permit decisions.
1. Proposed Options Regarding
Modifications
To provide a procedural framework
that allows these facilities to make the
necessary changes in RCRA permits, the
Agency proposes to amend the interim
status and permit modification
requirements.
a. Modifications During Interim
Status. Interim status facilities can make
certain facility alterations with fewer
procedural hurdles than apply to
permitted facilities. However, many
changes do require Agency approval. In
addition, interim status facilities must
adhere to all reconstruction
requirements found in 40 CFR Part
270.72 and must revise their Part A
permit applications. To ensure that
facilities making changes to come into
compliance with today's proposed
MACT standards are not constrained by
the reconstruction limits under § 270.72,
the Agency is proposing to add a new
sub-section as (b)(8) that would exempt
those facilities from the reconstruction
limitation. The Agency does not expect
that the costs to come into compliance
would exceed the 50 percent limit for
reconstruction—defined as 50 percent of
the cost of a new, comparable hazardous
waste management facility. However,
since the limit is cumulative for all
changes at the interim status facility,
there could be cases where this
provision could pose problems (e.g.,
where the facility had invested in a
number of prior changes).
b. Permit Modifications. For
permitted facilities, EPA's goal is to
implement a procedural system which
is as streamlined as possible, but still
allows for a satisfactory level of public
input. The Agency believes that a
streamlined process can result in earlier
achievement of the more stringent
MACT requirements by facilities,
leading to more environmentally
protective operations. The approach is
consistent with general efforts within
the Agency to improve environmental
permits by focusing on performance
standards, rather than on a detailed
review of the technology requirements.
The Agency's first, most streamlined
option is that the facility would be given
overall self-implementing authority (as
it has under the CAA) to perform all
necessary facility modifications to
comply with the new standards without
having to obtain a permit modification
from either the state or the Agency. This
option provides the facility with the
greatest latitude and authority since it
would allow the facility the opportunity
to make changes to its waste
management process and to operate
under conditions which are different
than those which are specified in either
the HSWA or base portion of its existing
RCRA permit. Under this option, there
would be no immediate need for the
facility to request a permit modification
to incorporate these operating changes
into the existing permit. These changes,
provided they enable the facility to meet
the new CAA standards, would be
incorporated into the permit at some
later date (e.g. during the permit
renewal process). It should be noted that
this option does not provide for public
participation at the time the facility is
altering its process to comply with the
new standards. Public involvement
would instead occur as part of a later
permit action, such as permit
reissuance. It would also not provide for
State or Federal agency oversight prior
to design or operating changes. This
option is based on the theory that, so
long as the facility is meeting the
applicable performance standards, there
may be no need to review how it comes
into compliance.
The Agency's second modification
option would consider all modification
requests due to the MACT standards to
be Class 1 modifications requiring no
prior approval. The basis for this option
would be to ensure that facilities are
capable of meeting the new standards
within the three year compliance
window because like Option 1, it
relieves the facility of possible delays
associated with obtaining prior approval
for modifications needed to come into
compliance. It also puts substantial
compliance responsibility on the facility
to make the correct changes within the
allotted time.
The Agency's third option, for which
rule language has been proposed, would
revise Appendix I of 40 CFR 270.42 to
designate as Class 1 modifications with
prior Agency approval all initial
requests for permit modifications made
by facilities in order to comply with
today's MACT standards. Appendix I of
40 CFR 270.42 would be revised to
reflect this classification by adding item
L(9) entitled "Initial Technology
Changes Needed to Meet MACT
Standards under 40 CFR Part 63
(National Emission Standards for
Hazardous Air Pollutants From
Hazardous Waste Combustors)". The
prior approval under this option would
provide for an Agency review of the
proposed physical and operational
changes to the facility before they are
implemented in order to ensure that
these changes do not lead to other
undesirable consequences.
Experience suggests that steps
intended to reduce emissions may not,
in all cases, lead to enhanced
environmental protection. On.the other
hand, it could be argued that it should
be the responsibility of the facility, not
the permitting Agency, to assure that the
regulated unit meets the required
performance standards. EPA requests
comment on the need for Agency
oversight.
The abbreviated procedures in
options 1 through 3 would be limited to
facilities making initial changes to
existing permits in order to come into
compliance with § 112 standards The
procedures would not apply to general
retrofitting changes outside the
framework of meeting MACT related
technology changes or to subsequent
changes relating to maintaining
compliance with § 112 standards. The
Agency is aware that the criteria for
deciding on the classification of a
modification request deviate from past
decision making criteria used to
differentiate among modification
classifications in Appendix I of Part
270. Many of the changes facilities
might make to conform to the new
standards would likely be Class 2 or 3
modifications under the current scheme.
However, the Agency believes that a
streamlined approach may be justified
because EPA did not consider newer,
more stringent standards becoming
effective under shorter timeframes when
it developed the current permit
modification table. Also, these changes
are mandated under a different
regulatory scheme for which the
modification tables were not designed to
account. This streamlining of the
modifications process has been
addressed in the past by the Agency to
ensure that changes made at facilities
needed to meet LDR levels for newly
listed or newly identified hazardous
waste could be met (see 54 FR 9596,
March 7,1989). These previous
modifications needed to meet the LDR
levels for newly identified wastes were
redesignated as Class 1 modifications.
These MACT standards impose more
stringent operating standards than
-------
17456
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
current requirements; the Agency
anticipates that the public will be
receptive to these improvements and
upgrades. Also, the Agency would still
have control over the modification
process under option 3 since it would
still be reviewing the details of
proposed new equipment or fixes to
existing equipment.
The Agency's fourth modification
option, like modification option 3,
would consider all initial modification
requests to existing permits to be Class
1 modifications requiring prior approval
by the Director, but would give the
Director the authority to elevate this
modification to a Class 2 modification if
the Director believes that additional
public participation is warranted. This
option to elevate a Class 1 modification
requiring prior approval to a Class 2
modification would apply only to
facilities requesting modifications to
comply with today's proposed MACT
standards. It would not apply to other
class 1 modifications.
The fifth modification option
represents a "no change" option. Most
modifications requested would likely be
handled as Class 2 or 3 modifications
given the types of facility changes we
expect in response to the MACT
standards. Under this option, facilities
would be urged to submit their permit
modification requests as soon as
possible in order to maximize the
chances of completing the modification
procedures, including administrative
appeals, prior to the compliance
deadline. EPA believes this alternative
could thwart the Agency's chief
objective of minimizing RCRA/CAA
interface problems, and would be
difficult to implement within the CAA
compliance deadlines. Therefore, EPA
does not favor this alternative.
Finally, the Agency realizes that many
states have not yet adopted the
modification table in Appendix I of 40
CFR 270.42. It hopes that states will, at
a minimum, adopt the modification
scheme that is promulgated in the final
rule to ensure expeditious
implementation of the new MACT
standards. Alternatively, if option 2 or
3 is selected in the final rule, States that
rely on a two-tiered system of major and
minor modifications could classify these
changes as "minor modifications".
In light of these proposed options for
facilities attempting to comply with the
MACT standards proposed in this
notice, the Agency is, under a separate
process, investigating ways to
streamline the entire RCRA permit
modification and renewal process for all
industry categories to further reduce
redundancies and inefficiencies in the
process, while making sure that the
public has adequate notice and
involvement in the process. The Agency
is in the early stages of this effort and .
wishes to solicit comment from the
public on ways to achieve a more
effective and efficient overall RCRA
permit modification and renewal
system.
2. Proposed Approaches To Address
Potential Implementation Conflict
As mentioned earlier, the Agency is
also taking comment on three
companion approaches to deal with
possible permit implementation
conflicts which may occur in the event
that a state does not become authorized
to carry out the provisions of this
rulemaking in time to handle necessary
modifications. These approaches are
relevant to modification options 2
through 5; if option 1 is chosen, no
permit modification will be necessary,
so the issues discussed in this section
would not arise. It is important to
remember that the standards in this rule
would take effect automatically under
the CAA. Therefore, the facility would
be obligated under that statute to make
the necessary changes to achieve
compliance. The issue discussed herein
relates to the respective roles of EPA
and authorized states in processing
RCRA permit modification requests.
The Agency's first approach provides
a narrow interpretation of the scope of
this rulemaking. Under this approach,
only the numerical standards imposed
by this rulemaking would be viewed as
within the scope of this rule, and so,
within the scope of HSWA. The manner
in which facility changes are performed
would be interpreted to be beyond the
scope of the rule. Therefore, for those
facilities needing a RCRA permit
modification to reflect changes in
permit conditions, the facility would be
required to request the modification
through the agency(ies) that implement
the portion(s) of the permit to be
modified.
Under the Agency's second approach,
both the proposed MACT standards as
well as the modification(s) needed to
come into compliance with these
standards would be interpreted to fall
within the scope of today's HSWA
rulemaking. Accordingly, the Agency
would make the modifications under
HSWA for facilities in states that have
not yet become authorized for this rule.
Although this approach would facilitate
changes, the Agency does recognize that
it could potentially create a possibility
for conflict between state and federal
permit portions. In areas where these
modifications would be inconsistent
with currently existing state-issued
portions of the facility's permit, the
State would need to perform parallel
modification procedures to correct the
inconsistencies. In the event that a State
could not do this (e.g. there is no "cause
for modification" under the State
regulations to cover the type of change
that would be necessary), EPA would
attempt to secure agreement from the
state that the new HSWA conditions are
more stringent than any inconsistent
state permit conditions and take
precedence over such conditions. The
state might memorialize this agreement
through memorandum or letter to the
facility or to the rulemaking record. This
approach might require an extensive
amount of communication between the
State and the Agency, e.g. to come to
agreement that the HSWA change is an
improvement over any conflicting
conditions in the state portion of the
permit.
Under the Agency's third approach, in
states that have not yet become
authorized under RGRA for this rule, the
Agency would not only modify the
permit by adding conditions necessary
for facilities to come into compliance
with these MACT standards, but would
also delete or modify conditions of the
state portion of a permit if conflicts exist
between the state- administered base
program portion of a permit and the
federally-administered HSWA portion.
This approach is similar to the second
approach, except that all modifications
to any portion of a RCRA permit would
be viewed as an integral part of EPA's
role in carrying out the new HSWA
requirements.
VII. State Authorization
A. Authority for Today's Rule
Today's rule is being proposed under
the joint authority of the Clean Air Act
(42 U.S.C. 7401 et seq.) and RCRA (42
U.S.C. 6924(o) and 6924(q)). The
proposed approach would apply the
new standards to both regulatory
programs. Although the proposed
standards would be located in 40 CFR
Part 63, which addresses Clean Air Act
requirements, the RCRA regulations in
40 CFR Parts 264 and 266 would
incorporate these standards by
reference. States may also promulgate
these standards under their CAA
program, and then incorporate them by
reference into their RCRA regulations.
Alternatively, States may promulgate
these standards in both the RCRA and
CAA sections of their State code for
several reasons. Also, States without an
approved CAA Title V permit program
may promulgate these standards under
their RCRA program only. Note
however, that EPA strongly encourages
States to adopt and apply for
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17457
authorization or delegation under both
regulatory programs for today's
proposed standards when finalized. (In
the implementation of RCRA and the
CAA by States, there is no functional
distinction between the authorization of
a State to implement RCRA in lieu of
EPA, and the delegation to a State to
administer the CAA. See the discussion
below.) EPA believes that State
implementation of this rule will
facilitate the coordination between the
RCRA and CAA regulatory programs.
B. Program Delegation Under the Clean
Air Act
Section 112(1) of the Clean Air Act
allows EPA to approve State rules or
programs for the implementation and
enforcement of emission standards and
other requirements for air pollutants
subject to section 112. Under this
authority, EPA has developed
delegation procedures and requirements
located at 40 CFR Part 63, Subpart E, for
NESHAPS under Title HI of the CAA
(See 57 FR 32250, July 21,1992).
Related requirements for permit
programs under Title V are located at 40
CFR Part 70 (See 58 FR 62262,
November 26,1993).
Under 40 CFR 70.4(a) and § 502(d) of
the CAA, States were required to submit
to EPA a proposed Part 70 (Title V)
permitting program by November 15,
1993. If a State CAA Title V program
does not receive EPA approval by
November 15,1995, the Title V program
must be implemented by EPA for that
State.
Submission of rules or programs by
States under 40 CFR Part 63 is
voluntary. Once a State receives
approval from EPA for a standard under
section 112(1) of the CAA, the State is
delegated the authority to implement
and enforce the approved State rules or
programs in lieu of the otherwise
applicable federal rules (the approved
State standard would be federally
enforceable). States may also apply for
a partial Title III program, such that the
State is not required to adopt all rules
promulgated in 40 CFR Part 63. EPA
will administer any rules federally
promulgated under section 112 of the
CAA that have not been delegated to the
State.
The section 112(1) rule for delegation
under Title HI (see 58 FR 62262,
November 26,1993), is currently the
subject of litigation. (See Louisiana
Environmental Network v.
Environmental Protection Agency, No.
94-1042 (D.C. Cir., filed January 21,
1994).) The outcome of this case could
severely limit the ability of States to
receive delegation for air toxics
standards that differ from the
comparable federal standards. A
decision is expected in early 1996.
C. RCRA State Authorization
1. Applicability of Rules in Authorized
States
Under section 3006 of RCRA, EPA
may authorize qualified States to
administer and enforce the RCRA.
program within the State. Following
authorization, EPA retains enforcement
authority under sections 3008, 3013,
and 7003 of RCRA, although authorized
States have primary enforcement
responsibility. The standards and
requirements for authorization are
found in 40 CFR Part 271.
Prior to HSWA, a State with final
authorization administered its
hazardous waste program in lieu of EPA
administering the Federal program in
that State. The Federal requirements no
longer applied in the authorized State,
and EPA could not issue permits for any
facilities that the State was authorized
to permit. When new, more stringent
Federal requirements were promulgated
or enacted, the State was obliged to
enact equivalent authority within
specified time frames. New Federal
requirements did not take effect in an
authorized State until the State adopted
the requirements as State law.
In contrast, under RCRA section
3006(g) (42 U.S.C. 6926(g)), new
requirements and prohibitions imposed
by HSWA take effect in authorized
States at the same time that they take
effect in unauthorized States. EPA is
directed to carry out these requirements
and prohibitions in authorized States,
including the issuance of permits, until
the State is granted authorization to do
so.
Today's rule is being proposed
pursuant to sections 3004(o) and
3004(q), of RCRA (42 U.S.C. 6924(o) and
6924(q)), which are HSWA provisions.
The rule would be added to Table 1 in
40 CFR 271.l(j), which identifies the
Federal program requirements that are
promulgated pursuant to HSWA. States
may apply for final authorization for the
HSWA provisions in Table 1, as
discussed in the following section of
this preamble.
2. Effect on State Authorization
Today's proposed rule is considered
to be more stringent than the existing
standards in 40 CFR Parts 264 and 266.
Thus, because today's revised technical
standards for hazardous waste
combustors are being proposed under
HSWA authority, when finalized, this
rule would be implemented by EPA in
authorized States until their programs
are modified to adopt this rule and the
modification is approved by EPA. Note
that these standards would also apply to
all covered facilities under CAA
authority, regardless of whether a State
has been delegated the provisions of the
final rule because these standards
would be largely self-implementing.
Because today's rule is proposed
pursuant to HSWA, a State submitting a
program modification may apply to
receive interim or final authorization
under RCRA section 3006(g)(2) or
3006(b), respectively, on the basis of
requirements that are substantially
equivalent or equivalent to EPA's. The
procedures and schedule for State
program modifications for final
authorization are described in 40 CFR
271.21. It should be noted that all
HSWA interim authorizations will
expire January 1, 2003. (See § 271.24(c)
and 57 FR 60132, December 18, 1992.)
In addition, note that 40 CFR Part 63,
Subpart E provides for interim
approvals under the CAA only in
limited circumstances.
Section 271.21(e)(2) requires that
States with final authorization must
modify their programs to reflect Federal
program changes and to subsequently
submit the modification to EPA for
approval. The deadline by which the
State would have to modify its program
to adopt these regulations is specified in
section 271.21(e). This deadline can be
extended in certain cases (see section
271.21(e)(3)). Once EPA approves the
modification, the State requirements
become Subtitle C RCRA. requirements.
States with authorized RCRA
programs may already have
requirements similar to those in today's
proposed rule. These State regulations
have not been assessed against the
Federal regulations being proposed
today to determine whether they meet .
the tests for authorization. Thus, a State
is not authorized to implement these
requirements in lieu of EPA until the
State program modifications are
approved. Of course, states with existing
standards could continue to administer
and enforce their standards as a matter
of State law pending authorization for
revised standards. In implementing the
Federal program, EPA will work with
States under agreements to minimize
duplication of efforts. In most cases,
EPA expects that it will be able to defer
to the States in their efforts to
implement their programs rather than
take separate actions under Federal
authority.
States that submit official applications
for final RCRA authorization less than
12 months after the effective date of
these regulations are not required to
include standards equivalent to these
regulations in their application. ;
-------
17458
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
However, the State must modify its
RCRA program by the deadline set forth
in § 271.21(e). States that submit official
applications for final authorization 12
months after the effective date of these
regulations must include standards
equivalent to these regulations in their
application. The requirements a State
must meet when submitting its final
authorization application are set forth in
40CFR271.5.
3. Streamlined Authorization Under
RCRA
Recently, EPA has initiated a series of
rulemakings intended to streamline and
speed the State authorization of RCRA
rules. On August 22,1995, EPA
proposed abbreviated authorization
procedures for certain routine Land
Disposal Restrictions (LDR) provisions
as part of the Phase IV LDR rule (see 60
FR 43654 and 43686). This proposal
would implement streamlined
authorization procedures for certain
minor and routine rulemakings for those
States which certify that they have
authority equivalent to and no less
stringent than the federal rule. EPA
believes that the abbreviated
authorization procedures proposed in
the August 22,1995, proposal would be
appropriate for RCRA Subtitle C
authorization for those States that are
approved to implement this rule
pursuant to 40 CFR Part 63, Subpart E,
and are simply incorporating this rule
into their RCRA regulations. EPA
requests comment regarding the use of
this proposed procedure for this
authorization scenario. Note however,
that EPA is not proposing to use RCRA
authorization as a substitute for CAA
section 112(1) approvals.
The primary reason that EPA is
proposing to use an abbreviated
authorization procedure when States are
approved to implement this rule under
the CAA, is that the delegation process
and requirements in Part 63 are similar
to authorization under 40 CFR 271.21.
For example, section 112(1)(1) of the
CAA requires that a program submitted
by a State "shall not include authority
to set standards less stringent than those
promulgated by the Administrator."
Further, section 116 of the CAA
precludes a State from adopting or
enforcing less stringent standards than
those under section 112. See 40 CFR
§§ 63.12(a)(l), 271.l(h), and section
3009 of RCRA. States may also establish
more stringent requirements as long as
they are not inconsistent with the CAA.
Further, section 112(1)(5)(A) of the CAA
requires States to have adequate
authorities to ensure compliance,
similar to the requirement in section
3006(b) of RCRA. Thus, for EPA to
approve a State rule or program, the
procedures and criteria in 40 CFR
63.91(b) must be met, as well as any
applicable requirements of §§ 63.92
through 63.94. These requirements are
equivalent to those under RCRA.
Therefore, using an abbreviated RCRA
authorization procedure would prevent
States from going through substantial
authorization procedures under both the
CAA program and the RCRA program.
EPA is also committed to streamlining
the authorization process for States that
would not be incorporating delegated
CAA standards stemming from the final
rule. EPA believes that authorized States
have experience implementing
sophisticated combustion regulatory
programs and would have the ability to
effectively implement today's proposed
standards. Thus, EPA requests comment
on whether all States that are authorized
for the incinerator regulations under 40
CFR Part 264 and the Boiler and
Industrial Furnace (BIF) regulations
should use the authorization procedure
proposed on August 22,1995. EPA is
also developing a second authorization
procedure for those RCRA rules which
have more significant impacts on State
hazardous waste programs that is
slightly more extensive than the
procedure proposed on August 22,1995.
This second procedure is also intended
to significantly streamline the
authorization process, and will be
described in detail in the upcoming
Hazardous Waste Identification Rule
(HWIR) proposal for contaminated
media. EPA believes that this second
procedure may be more appropriate for
today's proposal, given its significance .
and complexity. In the upcoming HWIR-
Media proposal, EPA will request
comment whether this procedure
should be used for RCRA authorization
in this case.
VIII. Definitions
Many of the terms used in today's
proposal have been defined either in the
Clean Air Act or in existing § 63.2. For
terms that are not already defined, we
are proposing definitions in §63.1201.
In addition, we are proposing
conforming definitions to the existing
RCRA regulations in §§ 260.10 and
270.2.
A. Definitions Proposed in §63.1201
We are proposing definitions for the
following terms in § 63.1201: Air
Pollution Control System, Automatic
Waste Feed Cutoff System, Cement Kiln,
Combustion Chamber, Compliance Date,
Comprehensive Performance Test,
Confirmatory Performance Test,
Continuous Monitor, Dioxins and
Furans, Feedstream, Flowrate, Fugitive
Combustion Emissions, Hazardous
Waste, Hazardous Waste Combustor,
Hazardous Waste Incinerator, Initial
Comprehensive Performance Test,
Instantaneous Monitoring, Lightweight
Aggregate Kiln, Low Volatility Metals,
New Source, Notification of
Compliance, One-Minute Average,
Operating Record, Reconstruction,
Rolling Average, Run, Semivolatile
Metals, and TEQ.
We believe that the definitions of
these terms is self-explanatory as
proposed.
B. Conforming Definitions Proposed in
§§ 260.10 and 270.2
To avoid confusion and ambiguity, we
are proposing conforming definitions in
§§ 260.10 and 270.2 for the following
terms that pertain to implementation of
the current RCRA requirements and
RCRA requirements that would not be
superseded by the proposed MACT
standards: RCRA operating permit, DRE
performance standard, closure and
financial responsibility requirements,
addition of permit conditions as
warranted on a site-specific basis to
protect human health and the
environment.
Because these definitions pertain to
existing RCRA requirements, the
effective date for the definitions would
be six months after the date of
publication in the Federal Register.
C. Clarification of RCRA Definition of
Industrial Furnace
Today's proposed rule applies to
combustion units that are already
subject to regulation under RCRA. These
devices are presently classified as
hazardous waste incinerators or
hazardous waste-burning industrial
furnaces, depending on their mode of
operation. As discussed below, the
distinctions between these
classifications (i.e., incinerator and
industrial furnace) are important in
determining the level for Clean Air Act
technology-based standards and also in
applying a variety of RCRA regulatory
provisions.
From the RCRA perspective, the
distinction between incinerators and
industrial furnaces (and boilers, for that
matter) is important, among other
things, for determining facility
eligibility for interim status, the
regulatory regime for classification of
. combustion residue (i.e., for example,
product or non-product), and eligibility
for Bevill status for combustion residue.
EPA defines industrial furnaces as those
designated devices that are an integral
part of a manufacturing process and that
use thermal treatment to recover
materials or energy. 40 CFR 260.10.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17459
Other criteria in the rule indicate what
it means to be an "integral part of a
manufacturing process." The RCRA
rules thus set out "aspects of industrial
furnaces that distinguish them from
hazardous waste incinerators", 48 FR
14472,14483 (April 4,1983); 50 FR 614,
626-27 0anuary 4,1985). These include
whether the device is designed and used
"primarily to accomplish recovery of
material products", the "use of the
device to burn or reduce raw materials
to make a material product", "the use of
the device to burn or reduce secondary
materials as effective substitutes for raw
materials, in processes using raw
materials as principal feedstocks", "the
use of the device to burn or reduce
secondary materials as ingredients in an
industrial process to make a material
product", and "the use of the device in
common industrial practice to produce
a material product. 40 CFR 260.10.
EPA interprets the regulatory
definition of industrial furnace as
applying only to devices that are
enumerated in the rule and that also
satisfy the narrative portion of the
definition, that is, functions as an
integral part of a manufacturing process,
taking into account the narrative criteria
in the rule. Thus, for example, if a
device which is otherwise a cement kiln
is not used as an integral component of
a manufacturing process, it is not an
industrial furnace. See 56 FR at 7140,
7141 (February 21,1991) (Device-by-
device application of industrial furnace
regulatory definition); 48 FR at 14485
(April 4,1983) (same). A cement kiln
used primarily to burn contaminated
soil from Times Beach so as to destroy
dioxins thus is not an industrial furnace
because it would not be an integral
component of a manufacturing process
but essentially a waste treatment unit.
Among other things, it would not be
used "primarily for recovery of material
products." 40 CFR 260.10(13)(I); See
also Background Document for the
Regulatory Definition of Boiler,
Incinerator, and Industrial Furnace
(October 1984), at page 6. Conversely, a
cement kiln making cement from raw
materials but burning some hazardous
waste for destruction as an adjunct to its
normal activities could be classified as
an industrial furnace.
Industrial furnaces burning hazardous
wastes for any purpose—energy
recovery, material recovery, or
destruction—are currently subject to the
rules for BIFs in Part 266 subpart H. 56
FR at 7138; 40 CFR 266.100. In this
regard, the BIF rule changed the
previous regulatory regime whereby if a
combustion device burned hazardous
waste for destruction, it was regulated
as an incinerator no matter what the
proportion of burning for destruction to
other activities. 40 CFR 264.340(a) and
265.340(a) as promulgated at 50 FR at
665-66 (January 4,1985); 48 FR at
14484 and n. 15 (April 4, 1983).
However, a device must still satisfy the
regulatory definition of industrial
furnace, and thus must in the first
instance be an integral component of a
manufacturing process. This means,
among other things, that enclosed
combustion devices that burn hazardous
wastes for destruction may not be
industrial furnaces. See 1984
Background Document for Definition of
Boiler, Incinerator, and Industrial
Furnace (cited above), page 6. This is
because hazardous waste destruction
devices may not be designing and using
the device primarily to accomplish
recovery of material products, may not
be using the device to combust
secondary materials as effective
substitutes for raw materials, etc.184
PART SIX: MISCELLANEOUS
PROVISIONS AND ISSUES
I. Comparable Fuel Exclusion
EPA is proposing to exclude from the
definition of solid and hazardous waste
materials that meet specification levels
for concentrations of toxic constituents
and physical properties that affect
burning. Generators that comply with
sampling and analysis, notification and
certification, and recordkeeping
requirements would be eligible for the
exclusion.185 See proposed
§261.4(a)(13).
Hazardous waste is burned for energy
recovery in boilers and industrial
furnaces in lieu of fossil fuels. There are
benefits to this energy recovery in the
form of diminished use of petroleum-
based fossil fuels. Industry sources
contend that in some cases, hazardous
waste fuels can be "as clean or cleaner"
(meaning they present less risk) than the
fossil fuels they displace. This claim has
not been documented with full
emissions and risk analysis. Industry
further contends that currently
regulating these materials under normal
184 The Administrator specifically rejects the
contrary suggestion of the Agency's Environmental
Appeals Board that "the purpose for which
hazardous waste is burned at the facility has little
or no bearing on whether the facility meets the
industrial furnace definition." In re Marine Shale
Processors, Inc., RCRA Appeal No. 94-12 (March
17,1995) p. 25 n. 32.
IBS We note that DOW Chemical Company (Dow)
in a petition to the Administrator, dated August 10,
1995, specifically requested that the Agency
develop a generic exclusion for "materials that are
burned for energy recovery in on-site boilers which
do not exceed the levels of fossil fuel constituents.
..." (Petition, at p. 3). This proposal also
responds to that petition.
hazardous waste regulations acts as a
disincentive to using them as fuels.
EPA's goal is to develop a comparable
fuel specification which is of use to the
regulated community but assures that an
excluded waste is similar in
composition to commercially available
fuel and poses no greater risk than
burning fossil fuel. Accordingly, EPA is
using a "benchmark approach" to
identify a specification that would
ensure that constituent concentrations
and physical properties of excluded
waste are comparable to those of fossil
fuels. We note that this is consistent
with the main approach discussed in
the Dow Chemical Company petition of
August 10,1995, which also points out
a number of benefits that would result
from promulgating this type of
exemption: (1) support for the Agency's
goal of promoting beneficial energy
recovery and resource conservation; (2)
reduction of unnecessary regulatory
burden and allowing all parties to focus
resources on higher permitting and
regulatory priorities; and (3)
demonstration of a common-sense
approach to regulation.186
The rationale for the Agency's
approach is that if a secondary material-
based fuel is comparable to a fossil fuel
in terms of hazardous and other key
constituents and has a heating value
indicative of a fuel, EPA has ample
authority to classify such material as a
fuel product, not a waste. Indeed,
existing rules already embody this
approach to some degree. Under
§ 261.33, commercial chemical products
such as benzene, toluene, and xylene
are not considered to be wastes when
burned as fuels because normal fossil
fuels can contain significant fractions of
these chemicals and these chemicals
have a fuel value. Given that'a
comparable fuel would have legitimate
energy value and the same hazardous
constituents in comparable
concentrations to those in fossil fuel,
classifying such material a non-waste
would promote RCRA's resource recover
goals without creating any risk greater
than those posed by the commonly used
commercial fuels. Under these
circumstances, EPA can permissibly
classify a comparable fuel as a non-
waste. See also 46 FR at 44971 (August
8,1981) exempting from Subtitle C
regulation spent pickle liquor used as a
wastewater treatment agent in part
because of its similarity in composition
to the commercial acids that would be
used in its place.
IBO vve also note there are other details in the
DOW petition that are congruent with aspects of
today's proposal. The Agency specifically invites
comment on the DOW petition as part of this
rulemaking.
-------
17460
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
As discussed below, EPA seeks
comment on a number of options
including what fossil fuel or fuels
should be used as a benchmark, and
how to select appropriate specification
limits given the range of values both
within and across fuel types. EPA also
requests additional data on hazardous
constituents naturally occurring in
commercially available fuels. (The
Agency's current data on fossil fuel
composition are provided in the docket
to this rulemaking.)
Also, the exclusion would operate
from the point of fuel generation to the
point of burning. Thus, the fuel's
generator would be eligible for the
exclusion and could either burn the
excluded comparable fuel on site or
ship it off-site directly to a burner. Thus,
the Agency must ensure that storage and
transportation of excluded comparable
fuel poses no greater hazard than fossil
fuel. The Agency invites comment on
whether the applicable Department of
Transportation (DOT) and Office of
Occupational Safety and Health- (OSHA)
requirements are adequate to address
this concern so that separate, potentially
duplicative RCRA regulation would not
be needed.
Note also that, because EPA is
proposing to eliminate or amend other
combustion-related exemptions in this
rulemaking (i.e., the exemption for
incinerators for wastes that are
hazardous solely because they are
ignitable, corrosive, or reactive and
contain no or insignificant levels of
Appendix VIII, Part 261, toxic
constituents; and the low-risk waste
exemption under BIF), the inclusion of
a comparable fuels exemption may
offset the effects of these changes at a
number of affected facilities.
EPA also invites comment on whether
acutely hazardous wastes should be
ineligible for the exemption. See the
section called "CMA Clean Fuel
Proposal", below, for what is considered
an acutely hazardous waste.
A. EPA's Approach to Establishing
Benchmark Constituent Levels
1. The Benchmark Approach
EPA considered using risk to human
health and the environment as the way
to determine the scope and levels of a
"clean fuels" specification. However,
the Agency encountered several
technical and implementation problems
using a purely risk-based approach.
Specifically, we have insufficient data
relating to the types of waste burned
. and the risks they pose. To pursue a
risk-based "clean fuels" approach, EPA
needs to examine emissions from a
number of example facilities at which
"clean fuel" would be burned. The
Agency could then analyze risks while
the facility is burning the "clean fuel".
EPA also does not have sufficient data
to determine the relationship between
the amount of "clean fuel" burned and
emissions, especially dioxins and other
non-dioxin PICs. EPA also does not
know how emissions relate to real
individual facilities as compared to
example facilities used to derive the
"clean fuel" specification. (Emissions
and/or risks at a given facility could be
higher than those of the example
facilities given site-specific
considerations.) Without this, it is not
clear how the Agency can use risk to
establish a "clean fuel" specification.
The Agency requests data and invites
comment on deriving a risk based
specification.
The Agency is instead proposing to
develop a comparable fuel specification,
based on the level of hazardous and
other constituents normally found in
fossil fuels. EPA calls this the
"benchmark approach". For this
approach, EPA would set a comparable
fuel specification such that
concentrations of hazardous
constituents in the comparable fuel
could be no greater than the
concentration of hazardous constituents
naturally occurring in commercial fossil
fuels. Thus, EPA would expect that the
comparable fuel would pose no greater
risk when burned than a fossil fuel and
would at the same time be physically
comparable to a fossil fuel.
2. The Comparable Fuel Specification
EPA is proposing to use this
benchmark approach to develop a series
of technical specifications addressing:
(1) physical specifications:
—Kinematic viscosity (cST at 100° F),
—Flash point (°F or °C), and
—Heating value (BTU/lb);
(2) general constituent specifications
for:
—Nitrogen, total (ppmw), and
—Total Halogens (ppmw, expressed as
Cl'~), including chlorine, bromine,
and iodine;187 and
(3) individual hazardous constituent
specifications, for:
—Individual Metals (ppmw), including
antimony, arsenic, barium, beryllium,
cadmium, chromium, cobalt, lead,
manganese, mercury, nickel,
selenium, silver, and thallium, and
—Individual Appendix VIII, Part 261,
Toxic Organics and Fluorine (ppmw).
(Note that ppmw is an alternate way of
expressing the units mg/kg.) The
constituent specifications and heating
value would apply to both gases and
liquids. The flash point and kinematic
viscosity would not apply to gases. EPA
invites comment on whether this list of
specifications should be expanded to
include other parameters, specifically
ash and solids content, to ensure that
excluded comparable fuels have the
same handling and combustion
properties as fossil fuels.
There are existing specifications for
fossil fuels that are developed and
routinely updated by the American
Society for Testing and Materials
(ASTM). (See ASTM Designation D 396
for fuel oils and D 4814 for gasoline.)
These requirements specify limits for
physical properties of fossil fuels, such
as flash point, water and sediment,
distillation temperatures,188 viscosity,
ash, sulfur, corrosion, density, and pour
point. The ASTM requirements do not
limit specific constituents in fuel. As a
result, fossil fuels are quite diverse in
their hydrocarbon constituent make-up.
Specific levels of hydrocarbon
constituents are a function of the crude
oil, the processes used to generate the
fuels, and the blending that occurs. This
makes ASTM requirements for fuels of
no use for deriving individual
hazardous constituent specifications,
but useful for deriving physical
specifications. EPA invites comment on
whether ASTM's physical specifications
for flash point and viscosity should be
used instead of the results of EPA's
analysis.18919°
a. Standards for CAA Metal HAPs.
EPA is proposing limits for two metals
that are not found on Part 261,
Appendix VIII: cobalt and manganese.
EPA included these metals in the
analysis because they are listed in the
Clean Air Act (CAA) as hazardous air
pollutants (HAPs). See CAA, section
112(b). These metals are included
because burning does not destroy
metals, and will cause the release of
metals into the air. Therefore, if a
comparable fuel contained more of a
metal than a fossil fuel, the result would
be more air emissions of that metal than
would be the case if the facility burned
only fossil fuels. From a CAA
perspective, it would not be acceptable
to increase emissions of CAA HAP
metals, relative to what would be
emitted if fossil fuels were burned.
187 See discussion below concerning another
halogen, fluorine.
188 The temperature at which a certain volumetric
fraction of the fuel has distilled.
189 The issue is that all analytical results should
meet ASTM's specifications. Thus, basing a
specification limit on analysis of samples will result
in limits more restrictive than the ASTM
specification defining an acceptable fuel.
190 ASTM does not specify a heating value
requirement.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19,, 1996 / Proposed Rules
17461
Therefore, constituent levels (or
detection limits) for the two CAA HAPs
are proposed as well.
b. Heating Value. With respect to
heating value, the Agency is concerned
with the issues of overall environmental
loading and acceptability of the waste as
a fuel. Comparable fuels may have a
lower heating value than the fossil fuels
they would displace. In these situations,
more comparable fuels would be burned
to achieve the same net heating loads,
with the result that more of the
hazardous constituents in the
comparable fuel would be emitted (e.g.,
halogenated organic compounds and
metals) than if fossil fuel were to be
burned. This would lead to greater
environmental loading of potentially
toxic substances, which is not in
keeping with the intent of the
comparable fuels exclusion.
To address environmental loading,
the Agency could establish a minimum
heating value specification comparable
to the BTU content of the benchmark
fossil fuel(s). Fossil fuels have a higher
heating value than most hazardous
waste fuels, however; so this approach
might exclude many otherwise suitable
fuels. Therefore the Agency chose to
establish the specification(s) for
comparable fuels at a heating value of
10,000 BTU/lb.191 EPA chose 10,000
BTU/lb because it is typical of current
hazardous waste burned for energy
recovery.192 However, hazardous waste
fuels have a xvide range of heating
values. Therefore, EPA is proposing
that, when determining whether a waste
meets the comparable fuel constituent
specifications, a generator must first
correct the constituent levels in the
candidate waste to a 10,000 BTU/lb
heating value basis prior to comparing
them to the comparable fuel
specification tables. In this way, a
facility that burns a comparable fuel
would not be feeding more total mass of
hazardous constituents than if it burned
fossil fuels.193
Also, EPA wants to ensure that
currently defined wastes which meet
the comparable fuels exclusion have a
legitimate use as a fuel. Historically, the
Agency has relied on a heating value of
11,500 J/g (5,000 BTU/lbm) as a
minimum heating value specification
for determining if a waste is being
i»i Constituent levels presented in today's
proposed rule have been corrected from the fuel's
heating valuo (approximately 20,000 BTU/lb) to
10,000 BTU/lb.
""Consult USEPA, "Draft Technical Support
Document for HWC MACT Standards, Volume IT:
HWC Emissions Database", February 1996.
183 Note that the heating value correction would
apply only to allowable constituent levels in fuels,
not to detection limits. Detection limits would not
bo corrected for heating value.
burned for energy recovery. (See
§ 266.103(c)(2)(ii).) EPA proposes this
limit today as a minimum heating value
for a comparable fuel to ensure that
comparable fuels are legitimate fuels.
c. Applicability of the specifications:
A separate issue is the applicability of
these specifications. EPA is proposing
that these specifications apply to all
gases and liquids currently defined as
hazardous wastes. (However as noted
elsewhere, used oil, and used crude oil
that is also a hazardous waste, would
remain subject to regulation as used oil
under 40 CFR Part 279, even if it meets
the comparable fuel specifications.) The
specifications for viscosity and flash
point would only pertain to liquid fuels.
This is because gases are inherently less
viscous than liquids and flash point
does not apply to gases. Therefore, EPA
proposes that the specifications for
viscosity and flash point not apply to
gaseous comparable fuels.
d. Organic Constituent Specifications.
With respect to Appendix VIII organic
toxic constituents and other toxic
synthetic chemicals, such as pesticides
and pharmaceuticals, the Agency needs
to ensure that only waste fuels
comparable to fossil fuels are excluded.
Therefore, the Agency proposes to limit
the Appendix VIII constituents in
comparable fuels to those found in the
benchmark fossil fuel. These limits were
calculated using a statistical analysis of
individual samples EPA obtained.
If the benchmark fossil fuel has no
detectable level of a particular
Appendix VIII constituent, then the
comparable fuel specification would be
"non-detect" with an associated,
specified maximum allowable detection
limit for each compound. (Note
exception in the following section.) The
detection limit is a statistically derived
level based on the quantification limit
determined for each sample.
There are also compounds found on
Appendix VIII which were not analyzed
for, either because an analytical method
is not available or could not be
identified in time for this analysis.
These compounds are not listed in
today's specifications. If EPA is able to
identify methods for analyzing these
compounds and is able to analyze for
these compounds prior to promulgation,
an appropriate specification level or
detection limit will be promulgated for
Appendix VIII compounds missing from
today's specification. If EPA is not able
to analyze for compounds on Appendix
Vni, we propose that the standard for
these remaining Appendix VIII
constituents be "nondetect" without a
maximum detection limit proposed.
e. Specification Levels for Undetected
Pure Hydrocarbons. A corollary issue is
that, since fossil fuels are comprised
almost entirely of pure hydrocarbons 194
in varying concentrations, it is possible
that many pure hydrocarbons on
Appendix VIII, Part 261, could be
present in fossil fuel but below
detection limits. Therefore, EPA
proposes allowing pure hydrocarbons
on Appendix VIII to be present up to the
detection limits in EPA's analysis.
Compounds on Appendix VIII which
contain atoms other than hydrogen and
carbon would be limited to "non-
detect" levels as described in the
previous paragraph.
/. Specification Levels for Other Fuel-
like Compounds. In addition there are
classes of fuel-like compounds that are
not found in fossil fuels. These include
oxygenates, an organic compound
comprised solely of hydrogen, carbon,
and oxygen above a minimum oxygen-
to-carbon ratio. Examples of oxygenates
which are used as fuels or fuel additives
include alcohols such as methanol and
ethanol, and ethers such as Methyl tert-
butyl ether (MTBE).195 However,
Appendix VIII oxygenates are not
routinely found in fossil fuels and were
not detected in EPA's sampling and
analysis program.196 Since oxygenates
can serve as fuels and are believed to
burn well (i.e., may not produce
significant PICs), EPA invites comment
on: (1) whether these compounds
should also be allowed up to the
detection limits in EPA's analysis; and
(2) an appropriate minimum oxygen-to-
carbon ratio to identify an oxygenate.
g. Total Halogen Specification and
Fluorine. Another issue is that the
methods for determining total halogens
do not measure fluorine, the lightest of
the halogen compounds. Fluorine is,
however, listed as an Appendix VIII
constituent and methods are available
for measuring fluorine directly.
Therefore, EPA proposes that the total
halogen limit pertain only to halogens
other than fluorine, i.e., chlorine,
bromine, and iodine. EPA also proposes
that a fluorine limit be established
separately from the total halogen limit.
Specification values for fluorine are
included in the specifications described
below.
h. Specification Levels for
Halogenated Compounds. EPA invites
comment on whether it is necessary to
""Excluding sulfur, carbon and hydrogen
comprise 99.6 to 100 percent of liquid fossil fuels.
195 A compound such as 2,3,7,8-TCDD is not an
oxygenate since it contains atoms other than
hydrogen, carbon, and oxygen. Compounds such as
Dibenzo-p-dioxin and Dibenzofuran are not
oxygenates even though they are comprised solely
of hydrogen, carbon, and oxygen because the
oxygen-to-carbon ratio is too low.
196 See the appendix for this notice for the results
of EPA's analysis.
-------
17462
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
specify limits for halogenated
compounds found on Appendix Vin.
Nondetect levels of halogens were found
in EPA's fossil fuel analysis and the
nondetect levels for total halogens were
much less than those of the individual
halogenated compounds. Therefore, a
waste that meets the total halogen limit
should, by default, meet the non-detect
levels specified for halogenated
compounds. EPA prefers this approach
since it will simplify the comparable
fuels specification and mean fewer and
less costly sampling and analysis of
comparable fuel streams for generators.
We invite comment on this approach.
EPA also invites comment on whether
this approach could be expanded to
other Appendix VIII constituents as well
(e.g., whether the total nitrogen
specification level would ensure
compliance with specification levels for
individual compounds containing
nitrogen).
3. Selection of the Benchmark Fuel
Another issue is selecting the
appropriate fossil fuel(s) for the
benchmark, and therefore the basis of
the comparable fuel specification.
Commercially available fossil fuels are
diverse. They range from gases, such as
natural gas and propane, to liquids,
such as gasoline and fuel oils, to solids,
such as coal, coke, and peat.
EPA does not believe, from an
environmental standpoint, that the
comparable fuel specification, which
would exclude a hazardous waste fuel
from RCRA subtitle C regulation, should
be based on fossil fuels that have high
levels of toxic constituents that may (or
will) not be destroyed or detoxified by
burning (e.g., metals and halogens). One
would expect that solid fuels, such as
coal, would have relatively high metal
and possibly halogen levels. Metals and
halogens are not destroyed in the
combustion process and as a result can
lead to increases in HAP emissions,
unlike organic Appendix VIII
constituents which (ideally) are
destroyed or detoxified through
combustion. Therefore, EPA is not
inclined to include a solid fuel as a
benchmark fuel. Also, we believe that
basing the comparable fuel specification
on a gas fuel would be overly
conservative and have no utility to the
regulated industry. Liquid fuels, on the
other hand, are widely used by industry
and do not have disadvantages of solid
or gaseous fuels. Liquid fuels seem a
good compromise among the fuel types.
The Agency is therefore proposing to
base the comparable fuel specification
on benchmark liquid fuels.
However, even liquid fossil fuels are
diverse and add to the complexity of
selecting a benchmark fuel. For
instance, gasoline has relatively higher
levels of toxic organics, such as benzene
and toluene but lower concentrations of
metals. Conversely, we have also found
and would continue to expect that
typical'fuel oils have lower
concentrations of toxic organics and
higher concentrations of metals than
gasoline. We also have found that
heavier fuel oils (e.g., No. 6) contain
more metals than lighter fuel oils (e.g.,
No. 2).i9?
In addition, EPA could choose a
vegetable oil-based fuel, such as "tall
oil", rather than a fossil fuel. EPA has
no data on concentrations of hazardous
constituents in these fuels, however.
Also, these fuels are not as widely used
as commercial fuels. In keeping with the
benchmark approach, EPA believes it is
appropriate to base the comparable fuel
specification on an appropriate and
widely used type of commercial fuel,
i.e., fossil fuels.
We specifically request constituent
data for gasoline, automotive diesel, and
No. 1 (kerosene/Jet fuel), No. 2 (different
from automotive diesel}, No. 4, and No.
6 fuel oils. These data should be
complete and include analyses for all
Appendix VTfl constituents including
nondetect values. When supplying data
during the comment period,
commenters should follow the same
analytical and quality procedures EPA
used. It would assist the Agency greatly
if the data were supplied in electronic
(1.44-MB PC or Macintosh floppy disk)
as well as hard-copy form. Electronic
versions should be in a spreadsheet
form (for instance, Lotus 1,2,3, or
Microsoft Excel) or an ASCII file with a
description of how the records are
classified/organized into which fields.
Consult the Technical Background
Document for a complete list of
constituents and additional information
concerning EPA's sampling and analysis
and quality assurance protocols used.
B. Sampling, Analysis, and Statistical
Protocols Used
This section describes the sampling,
analysis, and statistical protocols used
to derive the comparable fuels
specifications described below. For
more detailed discussion, refer to the
Technical Background Document.
1. Sampling
EPA obtained a total of 27 fossil fuel
samples. They were comprised of eight
gasoline and eleven No. 2, one No. 4,
and seven No. 6 fuel oil samples. The
samples were collected at random from
sources across the country: Irvine, CA;
north west New Jersey; north east
Connecticut; Coffeyville, KS; Fredonia,
KS; Norco, LA; Hopewell, VA; and
Research Triangle Park, NC.
Only one No. 4 fuel oil sample was
obtained. Very little "No. 4" fuel oil198
is sold in the United States. Rather,
what is used as No. 4 is essentially a
blend of No. 2 and 6 fuel oils/These
blends vary, are contract specific, and
are not No. 4 fuel oil, per se. EPA
specifically requests data on (genuine)
No. 4 fuel oil constituent levels.
2. Analysis of the Fuel Samples
Analytical methods have not been
defined for all compounds on Part 261,
Appendix VIII. Where analytical
methods have not been defined, analysis
of those constituent levels in fossil fuels
are not possible. However, EPA is
working on identifying methods for
compounds on Appendix VIII which
were not analyzed for during this initial
analysis. If EPA is able to identify
analysis methods for these compounds,
constituent specifications for these
compounds will be included in the final
rule using the same methodology for
constituent specifications described in
today's notice.
After the samples were obtained, they
were analyzed at a laboratory
accustomed to analyzing fossil fuels.
SW-846 methods were used whenever
possible. Where SW—846 methods were
not available, established ASTM
procedures or other EPA methods for
fuel analyses were used. Table VI. 1.1
summarizes the analytical methods
used.
TABLE Vl.1.1: ANALYTICAL METHODS
USED FOR COMPARABLE FUELS
ANALYSIS
Property of interest
Heating Value
Kinematic Viscosity ...
Flash Point
Total Nitrogen
Total Halogens
Antimony
Arsenic
Barium
Beryllium
Cadmium . .
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Method
EPA 325.3/PARR.
ASTM D240.
SW-846 1010.
ASTM D4629.
EPA 325.3/PARR.
SW-846 7040.
SW-846 7060.
SW-846 7080.
SW-846 7090.
SW-846 71 30.
SW-846 71 90.
SW-846 7200.
SW-846 7420.
SW-846 7460.
SW-846 7470.
SW-846 7520.
SW-846 7740.
SW-846 7760.
197 See the appendix to this notice for the results
of EPA's analysis.
198 No. 4 fuel oil is defined as fuel that meets the
physical specifications established by the American
Society of Testing and Materials.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17463
TABLE Vl.1.1: ANALYTICAL METHODS
USED FOR COMPARABLE FUELS
ANALYSIS—Continued
Property of Interest
Thallium
Appendix IX Volatile
Organics.
Appendix IX
Semivolatile
Organics.
Method
SW-846 7840
SW-846 8240.
SW-846 8270.
In addition, the analysis was
conducted in such a way as to ensure
the lowest detection limits, also called
"quantification limits," possible.
Detection limits were determined by
calculating the "method detection
limit" (MDL) for each analysis. To do
this, EPA used a modified version of the
procedures defined by EPA in 40 CFR
136, Appendix B, Definition and
Procedure for Determination of Method
Detection Limits, Revision 1.1. The
modification involved spiking for each
of the samples being analyzed instead of
spiking once for all the samples, as
stated by the method.
One issue concerning the analysis is
that, even when attempts are made to
minimize detection limits, detection
limits can still be extremely high. This
is particularly so for volatile organic
compounds in the gasoline samples.
There is no feasible analytical way to
address this issue, so it is addressed
when deriving the comparable fuel
specification.
3. Statistical Procedures Used
Due to the small sample sizes of each
fuel type, EPA used a nonparametric
"order statistics" approach to analyze
the fuel data. If enough data are received
to determine the distribution of the
enlarged data set, statistical procedures
appropriate to the distribution, i.e.,
different than those described here, may
be used for the promulgated
specification.
"Order statistics" involves ranking
the data for each constituent from
lowest to highest concentration,
assigning each data point a percentile
value from lowest to highest percentile,
respectively. Result percentiles were
then calculated from the data
percentiles. Consult the Technical
Background document for more
information regarding the statistical
approach.
EPA is considering using either the
90th or 50th percentile values to
determine the comparable fuel
specification. If the exclusion were to be
based on specifications from one or
more individual benchmark fuels (e.g.,
separate gasoline or fuel oil based
specifications), EPA believes it is more
appropriate to establish the
specification(s) based on the 90th
percentile rather than the 50th
percentile values. The 90th percentile
represents an estimate of an upper limit
of what is in a particular fuel while the
50th percentile values would exclude
up to 50 percent of the fossil fuel
samples. For composite specifications
(discussed in detail below), EPA is
considering using either the 50th or
90th percentile, but the considerations
differ. A 50th percentile analysis was
conducted because it represents what,
"on average", is found in all potential
benchmark fuels that were studied. A
90th percentile was also conducted
because it represents the upper bound of
what is found in all fuels. EPA invites
comment on which percentile(s) is
appropriate for both the individual
specifications as well as the composite
specification.
C. Options for the Benchmark Approach
As just described, EPA has several
options for deciding what fossil fuel(s)
to use as the benchmark. The following
options range from developing a suite of
comparable fuel specifications based on
individual benchmark fuels (i.e.,
gasoline, No. 2, No. 6) to basing the
specification on composite values
derived from the analysis of all
benchmark fuels.
The Agency invites comment on
which of the following options should
be selected. Again, EPA desires to
provide constructive relief to the
regulated community by having a
comparable fuel specification that can
be used in practice. On the other hand,
EPA needs to ensure that the release of
toxic compounds is not increased
significantly by burning comparable
fuels in lieu of fossil fuels. For this
reason, we are offering several options
for comment. Commenters should also
address in their comments the
justification needed to support their
preferred option.
The options discussed below are not
the only possible options. If commenters
have other options they wish the
Agency to consider, they should
recommend them and explain how they
meet the objectives of a benchmark
approach to comparability.
1. Individual Benchmark Fuel
Specifications
Under this option, EPA invites
comment on establishing individual
specifications based on the benchmark
fuels for which EPA has obtained data:
gasoline, and No. 2 and No. 6 fuel
oils.199 20° Each would have a unique set
of constituent and physical
specifications, based on the individual
benchmark fossil fuel. A generator
would use one of these specifications
(after correcting for heating value) to
determine if a waste qualifies for the
exclusion. As mentioned in subsection
A.2.B., above, heating value of a
comparable fuel would have to exceed
11,500 J/g (5,000 BTU/lbm).
EPA envisions that individual fuel
specification(s) could be implemented
in one of two ways under this approach.
First, a facility could use any of the
individual benchmark specifications,
without regard to what fuel it currently
burns. This approach would provide
flexibility for the facility in choosing
which specification to use. Although
this approach could allow higher
emissions of certain toxic compounds at
the particular site than would be the
case if they burned their normal fuel(s),
overall (total) emissions of hazardous
constituents may be lower since a
comparable fuel is unlikely to have high
levels of all constituents. In addition,
the amounts of excluded waste may
well be small relative to the quantity of
fossil fuels burned annually.
The second approach is to link the
comparable fuel specification to the
type of fuel burned at the facility and
being displaced by the comparable fuel.
In this case, if a facility burns only No.
2 fuel oil, it could only use the No. 2
fuel oil comparable fuel specification to
establish whether its current waste
stream is a comparable fuel.
Implementation issues include the
following: what specification would
apply if a facility uses a gas or solid fuel,
and what is the degree of inflexibility
introduced?
EPA prefers the first implementation
approach, but invites comment on
whether a single fuel should be used to
base a comparable fuel specification and
if so, which implementation should be
adopted.
2. A Composite Fuel as the Benchmark
One issue associated with the single
fuel specification approach is that
'"This list could be expanded, depending on the
amount and quality of data received during the
comment period.
200 EPA is reluctant to propose a No. 4 oil
specification at this time. As noted, EPA has been
able to obtain only one sample of No. 4 oil. EPA
desires more data on genuine samples of this fuel
before establishing a comparable fuel specification
based on No. 4 fuel oil. As is the case with other
types of fuel, if a sufficient number of samples are
obtained, a No. 4 fuel oil comparable fuel
specification may be promulgated.
-------
17464
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
gasoline has relatively high levels of
volatile organic compounds while No. 6
fuel oil has higher levels of semivolatile
organic compounds and metals. If a
potential comparable fuel were to have
a volatile organic constituent
concentration below the gasoline
specification but higher than the others,
and a particular metal concentration
lower than the No. 6 fuel oil
specification but higher than gasoline, it
would not be a comparable fuel since it
meets no single specification entirely.
Therefore, EPA is concerned that
establishing specifications under this
option would limit the utility of the
exclusion.
To address this issue, one option is to
use a composite approach to setting the
comparable fuel specification. In this
option, EPA would use a variety of
liquid fuels from which certain
compounds would be selected to derive
the complete specification.
EPA determined composite fuel
specifications for this proposal by
compositing the data from all fuels
analyzed (gasoline and the three fuel
oils individually). Compositing all the
fuels has the advantage that it may
better reflect the range of fuel choices
and potential for fuel-switching
available nationally to burners. A
facility would be allowed to use the
composite fuel specification regardless
of which fuel(s) it burns.
One technical issue is that EPA has
different number of samples for each
fuel type. Therefore, the fuel with the
largest number of samples would
dominate the composite database. To
address this issue, EPA's statistical
analysis "normalizes" the number of
samples, i.e., treat each fuel type in the
composite equally without regard to the
number of samples taken.
The Agency has evaluated
establishing a composite specification
using: (1) the 90th percentile aggregate
values for the benchmark fuels; and (2)
the 50th percentile aggregate values for
the benchmark fuels. Under either
approach, high gasoline volatile organic
nondetects would be omitted from the
analysis.
The 90th percentile approach has the
virtue of being representative of a range
of fuels that are burned nationally in
combustion devices. It also provides
maximum flexibility for the regulated
community. However, the 90th
percentile composite approach does
allow for higher amounts of toxic
constituents than other approaches EPA
is considering. As a practical matter,
though, no excluded fuel is likely to
contain constituent levels at or near all
of the 90th percentile composite
specification level. EPA invites
comment on this issue.
The 50th percentile approach ensures
the comparable fuel specification is
representative of a range of benchmark
fuels commonly burned at combustion
devices, perhaps even more so than the
90th percentile approach since it better
represents an "average" level for fuels
in general. It also provides flexibility for
the regulated community, though the
specification levels (and potentially the
usefulness) would be lower than those
resulting from the 90th percentile
approach. If facilities indeed are likely
to have at least several constituents near
the 90th percentile composite levels, a
50th percentile composite would be
more restrictive and less useful than the
90th percentile composite approach.
EPA seeks comments on whether a
composite of fuels should be used to
base a comparable fuel specification
and, if so, whether a 90th or 50th
percentile approach would be more
appropriate. Further, the Agency seeks
comment on whether the exclusion
should be based on a suite of
specifications comprised of the
individual benchmark fuel-based
specifications plus a composite
specification. Under this approach the
generator could select any specification
in the suite as the basis for the
exclusion.
3. Waste Minimization Approaches
By proposing this comparable fuels
exemption the Agency does not wish to
discourage pollution prevention/waste
minimization opportunities to reduce or
eliminate the generation of wastes in
favor of burning wastes as comparable
fuels. EPA solicits comments on the
effect of today's comparable fuels
proposal on facilities' efforts to promote
source reduction and environmentally
sound recycling (which does not
include burning for energy recovery as
a form of recycling in the RCRA waste
management hierarchy.)
D. Comparable Fuel Specification
In this section, EPA will outline the
five specifications discussed above:
gasoline, No. 2 fuel oil, No. 6 fuel oil,
composite 50th percentile values, and
composite 90th percentile values. For
reasons stated above, the individual fuel
specifications were based on the 90th
percentile values. EPA is not proposing
any particular approach at this time, but
invites comments on which
approach(es) should be promulgated in
a final rule. EPA is also presenting the
results of the No. 4 fuel oil sample for
comparison.
1. Hazardous Constituent Specifications
a. Gasoline Specification. The
gasoline-based specification is
presented in Table 1 of the appendix to
this preamble. As stated above, gasoline
contains more volatile organic
compounds (such as benzene and
toluene) than the other fuels. This
results in detection limits for volatile
organic compounds an order of
magnitude higher than the other fuel
specifications. EPA believes analysis of
comparable fuels will more likely result
in detection limits much lower than
gasoline and similar to those associated
with analysis of fuel oils. To address
this issue, EPA has performed an
analysis of a fuel oil-only composite
(one which does not include gasoline in
the composite) at the 90th percentile to
use as a surrogate for the volatile
organic gasoline non-detect values.
Those values from the fuel oil-only
composite are presented as the volatile
organic nondetect values in Table 1.
EPA invites comment on whether the
approach of substituting fuel oil-only
volatile organic nondetect values in lieu
of those values for gasoline is
appropriate.
b. Number 2 Fuel Oil Specification.
The No. 2 fuel oil-based specification is
presented in Table 2 of the appendix to
this preamble. As suggested above, No.
2 fuel oil contains more volatile organic
compounds than the other fuel oils, but
less than gasoline. In addition, its metal
concentrations are lower than the other
fuel oils, but more than gasoline.
c. Number 4 Fuel Oil Specification.
The No. 4 fuel oil-based specification is
presented in Table 3 of the appendix. It
follows a similar trend, having fewer
organic constituents than those previous
described, but more metals.
However, this specification is based
on only one sample. The Agency is
concerned that one sample may not be
representative of true No. 4 fuel oil. As
a result, EPA believes that we will not
be able to promulgate a No. 4 fuel oil
specification unless more data is
received during the comment period.
d. Number 6 Fuel Oil Specification.
The No. 6 fuel oil-based specification is
presented in Table 4 of the appendix.
e. Composite Fuel Specifications. Two
alternative composite fuel specifications
are presented in Tables 5 and 6 of the
appendix. Table 5 presents a
specification based on the aggregate
50th percentile values for the
benchmark fuels, and Table 6 presents
a specification based on the aggregate
90th percentile values of the benchmark
fuels.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 /Proposed Rules
17465
As was the case with the gasoline
specification, volatile organic detection
limits for gasoline are quite large. For
this reason, EPA is relying on surrogate
values for volatile organic detection
limits, one based on the detection limits
from a fuel oil-only composite. For the
50th percentile composite fuel
specification, the 50th percentile fuel
oil-only volatile organic nondetect
values were used. The 90th percentile
composite fuel specification was
handled similarly, using the 90th
percentile volatile organic nondetect
values from the fuel oil-only composite.
See the discussion for the gasoline
sample for EPA's concerns regarding
gasoline's high detection limits.
2. Physical Specifications (Flash Point
and Kinematic Viscosity)
Alternative physical specifications for
the options evaluated are presented
collectively in Tables 7 and 8 of the
appendix. Table 7 presents the results of
the analyses EPA conducted. Table 8
presents an alternate approach, using
the requirements for viscosity and flash
point for fuel oil specified by ASTM.
Physical specifications for viscosity and
flash point for gasoline are not required
by ASTM, but their upper and lower
limits, respectively, are available from
other reference sources.
When considering a composite
physical specifications using the
reference values presented in Table 8,
EPA believes it is appropriate to use the
second highest viscosity and second
lowest flash point as the specifications.
This would have the effect of not
considering the extremes, No. 6 fuel oil
viscosity (50.0 cSt at 100°C) and
gasoline flash point (-42°C), and using
as the specification the viscosity of No.
4 fuel oil (24.0 cSt at 40°C) and the flash
point of No. 2 fuel oil (38°C). EPA
believes this approach will result in
specifications which are representative
of comparable fuels and the fossil fuels
they displace, and ensure adequate
safety during transportation and storage.
Subsection A.2.b. discusses the
proposed minimum heating value of
11.500 J/g (5,000 BTU/lbm).
E, Exclusion of Synthesis Gas Fuel
EPA is also proposing to exclude from
the definition of solid waste (and,
therefore regulation as hazardous waste)
a particular type of hazardous waste-
derived fuel, namely a type of synthesis
gas ("syngas") meeting particular,
stringent specifications. The Agency
believes that many fuels produced from
hazardous wastes are more waste-like
than fuel- or product-like, and must be
regulated as such. We are aware,
however, of certain fuels and products
produced from hazardous waste that are
more appropriately classified and
managed as products rather than wastes.
EPA believes that syngas meeting the
requirements of the proposed exclusion
is such a material. Syngas is a
commercial product which has
important uses in industry as both a
feedstock and commercial fuel, and it
may be used as both a feedstock and
commercial fuel at a manufacturing
facility. The Agency is therefore
proposing this exclusion to clarify the
distinction between syngas products
meeting these stringent specifications
and hazardous wastes and other waste-
derived fuels. The Agency believes it is
useful to provide a conditional
exclusion for these particular fuels,
possibly before promulgating the
broader rule being proposed today. This
is because, although there may be much
debate about the generic comparable
fuel specification levels discussed
above, the syngas at issue here appears
to be well within the bounds of what
would be excluded, whatever the final
rule levels may actually be for other
comparable fuels.
The proposal applies to syngas that
results from thermal reaction of
hazardous wastes which is optimized to
both break organic bonds and
reformulate the organics into hydrogen
gas (H2) and carbon monoxide (CO).
This process is more similar to a
chemical reaction, rather than to
combustion. The process is optimized to
produce an end-product, rather than
merely to destroy organic matter.
EPA is aware of one such process,
proposed to be operated by Molten
Metals Technology (MMT). MMT
intends to operate a catalytic extraction
process (CEP) unit that generates certain
gas streams from the thermal reaction of
various hazardous wastes, including
chlorinated hazardous wastes. See letter
of July 21,1995, from Molten Metal
Technology to EPA. This letter and
other information on the MMT process
are in the docket for today's proposed
rule. MMT claims that the syngas
generated by the processes has
legitimate fuel value (i.e., 6,000 to 7,000
Btu/lb), has a chlorine level of 1 ppmv
or less, and does not contain hazardous
compounds at higher than parts per
billion levels. Thus, this syngas
possesses standard product indicia in
the form of fuel value plus being the
output of a process designed to optimize
these properties, and the syngas product
does not contain hazardous constituents
at levels higher than those present in
fossil fuel.
To ensure that any excluded syngas
meets these low levels of hazardous
compounds relative to levels in fossil
fuels in order to be excluded from the
definition as a solid waste, the Agency
is proposing the following syngas
specifications:
—Minimum Btu value of 5,000 Btu/lb;
—Less than 1 ppmv 202 of each
hazardous constituent listed in
Appendix VIII of Part 261 (that could
reasonably be expected to be in the
gas), except the limit for hydrogen
sulfide is 10 ppmv;
—Less than 1 ppmv of total chlorine;
and
—Less than 1 ppmv of total nitrogen,
other than diatomic nitrogen (N2).
EPA seeks comment on whether there
are other hazardous waste-derived
synthesis gas fuels (i.e., other than
MMT's) that meet the criteria for this
proposed exclusion.
We also note that conditions imposed
for exclusion of syngas fuels in no way
precludes the use of syngas as an
ingredient in manufacturing, which is
evaluated under a different set of
criteria, when the syngas is produced
from hazardous waste. In other words,
if the syngas were to be used as either
a product in manufacturing or burned as
a fuel, it would be excluded as a
product when it met the criteria for use
as a product and was used for that
purpose and excluded as a fuel when
burned.
If EPA adopts this exclusion for
syngas fuel, we believe that the
implementation procedures for the
generic comparable fuel exclusion
discussed subsequently in Section F
would also be appropriate for syngas.
This includes requirements for the
syngas producer to notify the Regional
Administrator that an excluded fuel is
produced, a certification that the syngas
meets the exclusion specification levels,
and sampling and analysis
requirements. EPA invites comment on
these implementation procedures for
syngases and whether any of these
procedures should be modified to
address any unique characteristics of
syngases.
Finally, we note that in Section F
below we discuss whether the burning
of hazardous waste excluded under the
generic comparable fuel exclusion
should be restricted only to stationary
sources either with air permits or that
otherwise have their air emissions
regulated by a federal, state, or local
entity. We specifically request comment
on whether this restriction would also
be appropriate for excluded syngas.
Given that the Agency may undertake
final rulemaking to provide an
202 All specification levels would be documented
at normal temperature and pressure of the gas at the
point that the exclusion is claimed.
-------
17466
Federal Register / Vol. 61, No, 77 I Friday, April 19, 1996 / Proposed Rules
exclusion for syngas before
promulgating a generic exclusion for
comparable fuels, however, we request
comment on whether more restrictive
requirements on burning excluded
syngas would be appropriate to
minimize concern about burning a
hazardous waste-derived gas. For
example, the exclusion could be limited
to syngas which is burned in an
industrial boiler, industrial furnace (as
defined in 40 CFR 260.10) or
incinerator. We note that these units
would not necessarily have to be RCRA
Subtitle C units.
F. Implementation of the Exclusion
The implementation scheme
described here is adapted from the
current used oil management system
and is tailored to the particular
characteristics of the comparable fuel
universe.203 It provides for one-time
notification and certification, sampling
and analysis, and recordkeeping
requirements. Other issues addressed
include blending, ensuring that the
comparable fuel is burned, and
treatment to meet the specification.
1. Notification and Certification
EPA proposes that a generator (or
syngas producer204) who claims that a
(currently defined) hazardous waste
meets the specification for exclusion
must submit a one-time notification and
certification to the Regional
Administrator. The notification would
state that the generator manages a
comparable fuel and certifies (through a
responsible company official) that the
generator is in compliance with the
conditions of the exclusion regarding
sampling and analysis, recordkeeping,
blending, and ultimate use of the waste
as a fuel. EPA understands that a
"generator" may be a company with
multiple facilities. For this reason, a
single company would be allowed to
submit one notification, but must
specify at what facilities the comparable
fuels notification applies. All other
provisions apply to each stream at the
point of generation.
203 Note that used oil has its own separate
management system, as allowed under RCRA,
tailored to the unique characteristics of used oil
recycling practices. The comparable fuel exclusion
proposed today would not apply to used oil because
it is adequately and appropriately managed under
its own tailored system. Used oil will still be
managed under 40 CFR Part 279. This proposal in
no way reopens the used oil specification or
management structure in 40 CFR Part 279.
204 Requirements applicable to the generator of an
excluded fuel would also apply to producers of
excluded syngas.
2. Sampling and Analysis
EPA believes it is appropriate that the
generator document by sampling and
analysis that the hazardous waste meets
the specification. Until such
documentation is obtained, the waste
would not be excluded. Waste analysis
rules for TSDFs would apply to
comparable fuel generators.
Consequently, generators would
implement a comparable fuels analysis
plan.
The sampling and analytical
procedures for determining that the
waste meets the specification must be
documented in a comparable fuels
analysis plan. The comparable fuel
analysis plan would involve sampling
and analyzing for all Appendix VIII
constituents initially and at least every
year thereafter for constituents that the
generator could have reason to believe
are present in the comparable fuel. EPA
specifically invites comment on
whether to allow a generator to use
process knowledge to determine what
compounds to sample and analyze for
during the first analysis, as well.
The generator would use current EPA
guidance for developing waste analysis
plans to derive their comparable fuel
analyze plan. This will ensure that
generators sample and analysis as often
as necessary, i.e., more frequently than
every year, for constituents present in
the fuel to ensure that excluded •waste
meets the specification.
Analytical methods provided by SW—
846 must be used, unless written
approval is obtained from the Regional
Administrator to use an equivalent
method. EPA invites comment on
establishing a procedure similar to Part
63, Appendix A, Method 301 to validate
alternate analytical methods. EPA also
invites comment on whether to limit the
Agency's time to approve an equivalent
method. In this case, the Regional
Administrator would have a set period
of time, such as 60 days, to respond to
the request. If an approval is not
received within 60 days, the alternative
method is considered approved. If the
Regional Administrator later rejects the
method, the rejection would only
pertain to analyses conducted after the
rejection of the method.
3. Use as a Fuel
An integral part of the comparable
fuel exclusion is that the fuel must be
burned. To ensure that the comparable
fuel is burned, the person who claims
the exclusion must either:
—Burn the comparable fuel on-site; or
—Ship the waste off-site to a person
who in turn burns the comparable
fuel.
This provision would not allow any
party to manage the fuel other than
those who generate or burn the fuel (and
other than transportation related
handling). EPA is reluctant to allow
persons other than the generator and the
burner to manage the comparable fuel
because it would likely be too difficult
to ensure that the excluded fuel meets
the specification and is burned. We
invite comment on how to allow third
party intermediaries, such as fuel
blenders, to handle an excluded
comparable fuel without precipitating
serious enforcement and
implementation difficulties.
Additionally, EPA is concerned that
comparable fuel shipped directly to an
off-site burner may not in fact be
burned. Therefore, EPA invites
comment on whether, for off-site
shipments to a burner, the following
information should be retained in the
record for each shipment:
—Name and address of the receiving
facility;
—Cross-reference to a certification from
the facility certifying that the
comparable fuel will be burned;
—Quantity of excluded waste shipped;
—Date of shipment; and
—A cross-reference to the analyses ,
performed to determine that the waste
meets the specification.
A comparable fuel which is not burned
remains a hazardous waste and is
subject to regulation cradle-to-grave.205
This documentation would provide a
paper trail to ensure that the comparable
fuel is burned.
EPA invites comment on •whether the
burning of a comparable fuel should be
restricted to only stationary sources
either with air permits or that otherwise
have their air emissions regulated by a
federal, state, or local entity. EPA's
primary concern is that excluded fuel
may be burned in unregulated
combustion devices. EPA believes that
unregulated burners may be unaware of
or unprepared to handle many unique
issues related to fuels other than fossil
fuels. In addition, EPA invites comment
on whether comparable fuels should be
allowed for use in sources other than
stationary sources, i.e., mobile sources
(on- and off-road automobiles, trucks,
and engines) and small engines.
4. Blending To Meet the Specification
The issue of whether to allow
blending to meet the comparable fuel
specification also needs to be addressed.
205 Note that the only disposal method for a
comparable fuel is burning. Any disposal method
other than burning is a RCRA violation, unless the
comparable fuel is properly managed as a
hazardous waste. >.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17467
One alternative is to exclude only those
comparable fuels that meet the
specification as generated and which are
destined for burning. The facilities
would be required to demonstrate, for
compliance purposes, that the waste as
generated meets the specification and to
certify that the waste is destined for
burning.
If blending to lower the
concentrations of hazardous
constituents in a waste were allowed to
meet the specification, EPA believes
that a very extensive compliance and
enforcement system would have to be
instituted to ensure that blending was
done properly (with any necessary
storage and treatment permits) and that
the resultant mixture meets the
specification continually. This
alternative appears to warrant a degree
of oversight that may be infeasible from
the industry viewpoint and unworkable
from the Agency's viewpoint. EPA is
also investigating whether blending
removes the incentive for facilities to
engage in source reduction and
recycling of waste. Finally, this
alternative raises the issue of whether
blending is simply a form of prohibited
or objectionable dilution that could
result in an overall increase in
environmental loading of toxic,
persistent, or bioaccumulative
substances.
Complicating this issue is the fact that
blending to lower hazardous constituent
concentrations in used oil is allowed.
(40 CFR 279.50(a).) However, EPA
believes it is appropriate to deviate from
the approach for used oil in this case.
Used oil is better defined and
understood in its origins and use than
currently defined hazardous wastes.
Used crankcase oil is a petroleum
product analogous to a thick fuel with
enriched metal concentrations due to its
use for lubricating metal-bearing parts
in situations of tight tolerance. In the
case of used oil, blending a thick fuel
enriched with metals with a thinner fuel
with low concentrations of metals is
appropriate since the resulting mixture
would be wholly a petroleum product
with similar levels of metals as other
petroleum fuels.
Comparable fuels, however, differ
substantially from used oil in both the
nature of materials to which the
exclusion pertains and the scope of the
exclusion. A comparable fuel is
presently defined as a hazardous waste
and is unlikely to be a petroleum
distillate. The issue of toxic organic
constituents is important for comparable
fuels due to the diversity of processes
and process ingredients from which
potential comparable fuels may result.
This is not relevant for the used oil rules
since they deal with the post-use
material stemming from a highly
consistent and well known petroleum
distillate. Therefore, blending used oil
would result in a more predictable
mixture, one which would be expected
to contain the same organic compounds
in varying concentrations. The same
cannot be said for the large variety of
potential comparable fuels, which can
vary significantly in the constituents
present.
The issue of metals in a comparable
fuel is similarly different from the case
of used oil. While used oil does contain
enriched levels of metals relative to
virgin oil or petroleum fuels, those
levels are greatly understood (relative to
hazardous waste in general) due to their
use in only one process, the lubrication
of metal-bearing parts. Therefore, there
is essentially a real-world limit to the
amount and type of metal that could be
entrained in a used oil, so blending to
meet metal specifications is more
appropriate. In the case of comparable
fuels if there were no prohibition on
blending to meet constituent
specifications, a generator would be
allowed to take a predominantly metal
waste, blend it into a fuel to levels lower
than the constituent specification levels,
and (through pure dilution) meet the
exclusion. For these reasons, EPA
believes the specially tailored used oil
program does not provide a satisfactory
model to use for addressing the issue of
blending potential comparable fuels.
We also note that the LDR program
specifically prohibits dilution as a form
of treatment. (40 CFR 268.3.) Allowing
blending to meet the specification may,
in effect, allow dilution as a form of
treatment contrary to the LDR
prohibition for these hazardous wastes.
For these reasons, EPA desires to stay
consistent with other rules and policies
and not allow blending to meet the
comparable fuels specification.
Similarly, EPA proposes that the
specification for heating value be met on
an as-generated basis as well. In other
words, blending would not be allowed
to meet the heating value specification.
If the Agency were to allow blending to
meet the heating value specification,
wastes with no heating value could be
blended with high heating value fossil
fuels and meet the comparable fuel
heating value specification. EPA does
not believe this approach can be
justified, allowing a waste which as
generated has little or no heating value
to be a comparable fuel. Therefore, we
propose that heating value be met on an
as generated basis.
For these reasons, EPA is proposing
that the comparable fuel constituent and
heating value specifications be met on
an "as generated" basis, and that
blending to meet the constituent and
heating value specifications not be
allowed. However, if the constituent
and heating value specifications have
been met as generated, EPA believes it
may be appropriate for a comparable
fuel to be treated like any other fuel and
allow it to be blended after the
constituent and'heating value
specifications have been met. This
includes blending for the purposes of
meeting other physical specifications
(flash point and viscosity), pH
neutralization, etc.
After blending, generators would have
to retest the prospective comparable fuel
to ensure that blending did not increase
the levels of constituents to above the
specification levels or decrease it to
below the heating value requirement. If
the waste were blended with a clean
fossil fuel, such as No. 2 fuel oil, it
would be sufficient to document that
the substance the prospective
comparable fuel is being blended with
has lower constituent levels and a
higher heating value than the
comparable fuel specification. If the
waste is above constituent specifications
or below the heating value requirement
after blending, the waste would not be
a comparable fuel.
EPA invites comment on the issue of
blending only to meet the physical
specifications, flash point and kinematic
viscosity.
5. Treatment To Meet the Specification
It is possible, as a technical matter, for
hazardous wastes to undergo treatment
that destroys or removes hazardous
constituents and thereby produce a
comparable fuel. Likewise, it is possible
to treat a waste such that the heating
value of the waste is increased. For
example, distillation could remove
certain organic constituents from the
waste matrix, thereby allowing the
treated waste to meet the comparable
fuel specification. Similarly, decanting
to decrease the water concentration of
the waste stream would increase the
heating value of the waste by
concentrating those compounds which
are burned. The issue discussed here is
whether such processes should be
allowed under a comparable fuel
regime, and if so, under what
circumstances. The Agency is proposing
to allow treatment under limited
circumstances.
The Agency's concern about allowing
such treatment is that it could increase
the incentive and opportunity for
impermissible blending or otherwise
fraudulent treatment. Thus, at the least,
EPA would seek to set up controls to
reduce the possibility of such practices
-------
17468
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
if treatment were allowed. This might be
done by requiring treaters to document
that the comparable fuel specification is
being satisfied through treatment that
destroys or removes hazardous
constituents and/or increases heating
value by removing constituents from the
waste, not through blending or other
dilution-type activities. Second, where
the treater has a RCRA permit for the
storage/treatment activity (i.e., treatment
of hazardous waste conducted in any
unit except a 90-day generator unit not
subject to permitting requirements
under § 262.34), the rule could
authorize permit writers to add
conditions to the permit to assure the
integrity of the permitted process. Such
conditions could take the form of extra
conditions on the treatment process,
conditions on the wastes which could
be treated to produce comparable fuels,
and additional sampling and analysis of
both incoming wastes and outgoing
comparable fuels. The Agency solicits
comment on what limitations or
conditions should be imposed on
treatment activities and whether and
how to adapt such limitations or
conditions to the non-permitted context
of 90-day generator units.
Finally, it should be noted that if
hazardous wastes are treated to produce
comparable fuels, only the comparable
fuel would be excluded from RCRA
subtitle C regulation. The hazardous
wastes would be regulated from point of
generation until a comparable fuel is
produced, so that generation, transport,
storage, and treatment of the waste until
' production of the comparable fuel
would remain subject to the applicable
subtitle C rules. Also any residuals
resulting from treatment remain
hazardous wastes as a result of the
derived-from rule.
6. Recordkeeping
It is proposed that documentation
pertaining to verification that the waste
meets the comparable fuel specification
and the information on shipments be
retained for three years. The sampling
and analysis plan and all revisions to
the plan since its inception would be
retained for as long as the person claims
to manage excluded waste, plus three
years. Certifications from burners (if
required in the final rule) would be
retained for as long as the burner is
shipped comparable fuels, plus three
years.
The generator would retain the
records supporting its claim for the
exemption. For comparable fuels which
are not blended, the records that must
be retained are the as generated results.
For comparable fuels which are blended
to meet the flash point and/or kinematic
viscosity specifications, the records
which must be retained are those after
blending.
7. Small Business Considerations:
Inherently Comparable Fuel
Small businesses may, hypothetically,
generate wastes (such as mineral spirits
used to clean automotive parts) that
could meet a comparable fuel
specification. However, the Agency is
concerned that the proposed
implementation scheme for the
comparable fuel exclusion may be
overly burdensome to small businesses
because of the small volume of waste
each business may generate. EPA
requests data on whether categories of
high volume inherently comparable fuel
from a large number of small generators
exist. If so, EPA would consider
providing an exclusion for these fuels in
the final rule. For these fuels to be
excluded, the Agency would need
constituent data from various small
generators indicating that these wastes
would meet the comparable fuel
exclusion levels on a routine basis.
If an inherently comparable fuel
exclusion were promulgated in the final
rule, the Agency would promulgate a
petitioning process whereby classes of
generators could document that a
specific type of waste is virtually always
likely to meet the comparable fuel
specification. If the Agency granted the
petition through rulemaking, such waste
would be classified as inherently
comparable fuel. As such, the generator
would not be subject to the proposed
implementation requirements for the
comparable fuel exclusion: notification,
sampling and analysis, and
recordkeeping. In addition, such
inherently comparable fuel could be
blended, treated, and shipped off-site
without restriction given that it would
be excluded from regulation as
generated.
EPA invites comment on whether
high volumes of comparable fuel is
generated from a large number of small
generators. If so, the Agency requires
data on whether this approach provides
relief to small businesses while ensuring
protection of human health and the
environment. In addition, EPA invites
analytical data supporting classification
of particular wastes as inherently
comparable fuel. The Agency would
provide notice and request comment on
such data prior to making a final
determination that the waste is
inherently comparable fuel.
G. Transportation and Storage
Waste derived fuels can pose risks
during transportation and storage, not
just when burned. For instance,
comparable fuels could be reactive and
corrosive (virgin fossil fuels are neither),
more volatile than fossil fuels, or have
other special properties affecting
handling and storage. The Agency
believes we can exempt comparable
fuels from RCRA storage and
transportation requirements and
therefore rely on the storage and
transportation regulations of other
federal and state agencies. However, the
affected industries may have more
direct knowledge of how these
requirements actually affect shipments
and storage of the potential fuels,
particularly with respect to the extent of
state regulatory controls. We are
therefore asking commenters to give
EPA information on the adequacy of
DOT and OSHA requirements related to
storage and transportation, particularly
with respect to whether a combustion
facility (including an industrial boiler)
will be on proper notice about the
nature and behavior of the comparable
fuel to allow for safe handling and
burning.
In this regard, EPA believes it is
appropriate to set a minimum flash
point for comparable fuels. (See section
A.2. for a general discussion concerning
the Comparable Fuels Specification.)
The flash point is defined as the
minimum temperature at which a
substance gives off enough flammable
vapors which in contact with a spark or
flame will ignite. Setting a minimum
flash point would ensure that under
ambient conditions the comparable fuel
would not ignite during transportation
and storage.
A shortcoming of this approach is that
a purchaser or other off-site facility may
desire a comparable fuel with a flash
point lower than the comparable fuel
specified flash point. EPA does not wish
to preclude low flash point comparable
fuels from the exemption. Therefore, the
Agency is inclined to allow some waiver
of the minimum flash point
specification under certain
circumstances.
EPA is proposing to allow low flash
point comparable fuels if there is some
notice to intermediate carriers and the
ultimate user of what the flash point of
this comparable fuel is. To do this, EPA
needs to be assured that these low flash
point comparable fuels can be stored,
handled, and transported safely. EPA is
inclined to believe current DOT and
OSHA requirements for transportation
and storage of hazardous or combustible
liquids are adequate for this purpose,
but we specifically seek comment on
this issue.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 7 Proposed Rules
17469
H. Speculative Accumulation
EPA is also proposing that
comparable fuels remain subject to the
speculative accumulation test found in
§ 261.2(c)(4). This means that persons
generating or burning comparable fuels
must actually put a given volume of the
fuel to its intended use during a one-
year period, namely 75 per cent of what
is on nand at the beginning of each
calendar year commencing on January 1.
See the definition of "accumulated
speculatively" in § 261.1(c)(8). (The
rules also provide for variances to
accommodate circumstances where
such turnover is not legitimately
practical. § 260.31(a).) EPA applies this
test to other similar exclusions of
recycled secondary materials in the
rules (see § 261.2(e)(2)(iii).) This is
because over accumulation of hazardous
waste-derived recyclables has led to
many of the most severe hazardous
waste damage incidents. See 50 FR at
658-61 and 634-37 (January 4,1985).
There is no formal recordkeeping
requirement associated with the
speculative accumulation test, but the
burden of proof is on the person
claiming the exclusion to show that the
test has been satisfied. § 261.2(f) and 50
FR at 636-37.
/. Regulatory Impacts
EPA also requests data from the
regulatory community concerning the
regulatory impacts of this proposed
comparable fuel exclusion. Impact data
includes the quantity of waste which
would be excluded (by weight) and the
cost savings as a result of the exclusion.
Based on the data submitted, EPA will
develop a full regulatory impact
assessment during the final rulemaking.
/. CMA Clean Fuel Proposal
The Chemical Manufacturers
Association (CMA) submitted a proposal
to exempt certain "clean" liquid wastes
from RCRA regulation206. Unlike EPA's
benchmark-based comparable fuel
proposal, the CMA approach would
establish clean fuel specifications for
mercury, LVM, and SVM metals based
on the technology-based MACT
emission standards proposed today. For
mercury, CMA calculated the maximum
feed rate the facility would be allowed
if it had a given gas flowrate, no
mercury control, and yet complied with
today's proposed standards. This would
establish the maximum mercury
concentration of the CMA "clean fuel"
specification. Limits would be
established for LVM and SVM metals in
a similar fashion. For chlorine, CMA
presented a specification level based on
the concentration of chlorine found in
coal. Limits for ash content would be
derived from No. 4 fuel oil.
The CMA proposal also appears to
rely solely on adequate thermal
destruction of the organics to control
potential organic contamination and
risks therefrom. Combustion would be
limited to on-site boilers or boilers
owned and operated by the clean fuel
generator, where these boilers meet a
100 ppmv hourly rolling average CO
limit.
CMA's clean fuel proposal would also
establish limits on physical
specifications. The heating value of a
CMA clean fuel would have to be at
least 5,000 BTU/lb, viscosity would
have to be less that 26.4, and the clean
fuel must be a liquid.
Acutely hazardous wastes207 would
not be eligible for CMA's proposed
clean fuel exemption, nor would dioxin-
listed wastes (hazardous waste numbers
F020, F021, F022, F023, F026, F028.)
EPA invites comment on CMA's
proposed "clean fuels" specification.
Specifically, EPA requests commentors
address the following issues and
questions:
—Is reliance on the technology-based
MACT emission standards approach
appropriate for establishing a clean
fuel exemption under RCRA, either
with or without restrictions on the
type of device that can be used to
burn the clean fuel? How does EPA
justify not establishing specific
constituent limits for the other five
RCRA metals?
—Does a CO limit alone ensure
adequate destruction of toxic organics
in a clean fuel scenario? Would
additional controls, such as an HC
limit, limits on inlet temperature to a
dry PM APCD, DRE testing, and site-
specific risk assessment also be
appropriate?
—Does CMA's proposal adequately
address new facilities? Would it be
appropriate to allow off-site shipment
to a facility not owned by the
generator if the generator owns no
combustion device in the vicinity? If
so, how would EPA be able to ensure
compliance regarding the CO
emissions (and possibly other testing
and operational conditions) of a
combustion device not owned by the
generator?
504 SOB Revised CMA Proposal for Clean Waste
Fuels Exemption to BCRA dated March 1,1996.
207 That is, discarded commercial chemical
products listed in § 261.33 ("P" listed wastes), and
acutely hazardous [those with "H" hazard codes)
wastes listed in §§ 261.31 and 261.32 (hazardous
wastes from non-specific and specific sources, "F"
and "K" listed wastes, respectively.)
—Should CMA's clean fuel approach be
expanded to include gaseous as well
as liquid fuels?
—Are there wastes other than those
identified by CMA (acutely toxic and
dioxin-listed wastes) which should
not be eligible for a "clean fuel"
exemption? If so, what would be the
practical impacts of such expanded
ineligibility?
—Are data available documenting that
emissions from burning a "clean fuel"
would not pose a significant risk for
the potential combustion and
management scenarios in which the
clean fuel exclusion from RCRA might
be used?
II. Miscellaneous Revisions to the
Existing Rules
This section provides several
miscellaneous revisions to the RCRA
hazardous waste combustion rules
provided by 40 CFR Parts 260-270. We
note that we are also proposing other
revisions to Parts 260-270 that would be
conforming revisions to ensure that the
RCRA rules are consistent with similar
provisions of the proposed Part 63 rules.
Those proposed conforming revisions
are discussed elsewhere in the
preamble.
A. Revisions to the Small Quantity
Burner Exemption Under the BIFRule
The Agency is proposing to revise the
small quantity burner (SQB) exemption
provided by § 266.108 of the BIF rule
because the current exemption may not
be protective of human health and the
environment. Under the exemption,
BIFs could burn up to the exempt
quantities absent regulation other than
notification and recordkeeping
requirements. Under a settlement
agreement, the environmental
petitioners in Horsehead Resource
Development Company, Inc., v. EPA
(No. 91-1221 and Consolidated Cases),
the Agency must reevaluate whether the
small quantity burner exemption is
sufficiently protective given that the
Agency did not consider indirect
exposure pathways in calculating the
exemption levels. In addition, the
petitioners argued that the exemption is
inconsistent with the intent of RCRA
§ 3004(q)(2)(B) which specifically
allows the Administrator to exempt
facilities which burn de minimis
quantities of hazardous waste because
the exemption as promulgated would
allow sources to burn up to 2,000
gallons of hazardous waste per month
absent substantive emissions controls.
Petitioners believe that 2,000 gallons per
month is not a de minimis quantity.
EPA attempted to reevaluate exempt
quantities considering indirect exposure
-------
17470
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
pathways for, in particular, emissions of
dioxins and furans (D/F). Unfortunately,
we were not able to adequately predict
emission levels of D/F for purposes of
conducting a generic, national risk
assessment to back-calculate exempt
quantities. We could not effectively
predict D/F emissions because: (1)
There may be little relationship between
quantity of hazardous waste burned and
D/F emissions (i.e., other factors may
result in high or low D/F emissions);
and (2) there are several site-specific
factors that can affect D/F emissions,
including combustion efficiency (that is
affected by factors such as combustion
zone temperature, oxygen levels, and
residence time in the combustion zone),
gas temperature at the particulate matter
control device, and presence of
precursors such as PCBs.
In addition, we found it difficult to
identify an appropriate indirect
exposure scenario for purposes of
assessing risk to support a generic
exemption. We note that to evaluate
whether the proposed MACT standards
met RCRA protectiveness requirements,
we analyzed 11 example facilities
assuming the example facilities emitted
HAPs at the regulatory option levels. We
did not have site-specific stack gas
properties (e.g., gas flow rate, gas
temperature, stack height) and exposure
information to conduct similar indirect
exposure assessments for example SQB
facilities.
Given these difficulties, the Agency is
proposing to revise the SQB exemption
to limit exempt quantities to 100 kg/mo
(27 gal/mo), which is the current
exemption level for small quantity
generators (SQG) provided by § 261.5.
We believe that this is appropriate given
that SQG hazardous waste is already
exempt from regulation and thus, may
be burned absent emission controls. We
note, however, that the SQB exemption
can apply to facilities owned or
operated by large quantity generators.
Thus, under today's proposal, wastes
not eligible for the SQG exemption
could be eligible for the SQB exemption.
Nonetheless, we believe that 27 gal/mo
is a reasonable level for the exemption
because it is truly a de minimis quantity
and such quantities can be burned
absent emission controls under existing
SQG regulations.
We believe that approximately 200
boilers are currently operating under the
SQB exemption. Many of these boilers
are likely burning quantities in excess of
27 gallons/mo, and so would be subject
to full regulation as a BIF under today's
proposal. We note, however, that we are
also proposing today a comparable fuels
exclusion that would exclude from the
definition of solid and hazardous waste
any material that meets the proposed
comparable fuels specification.
Although we currently have no
information on how many SQBs could
use the comparable fuels exclusion,
some heretofore SQBs are expected to be
eligible for this proposed exclusion.
Sources that burn hazardous waste
that do not meet the comparable fuels
specification may determine that it is
less expensive to send their waste to a
commercial burner than comply with
the BIF regulations. Those sources that
choose to continue burning hazardous
waste would be required to comply with
the substantive requirements of the BIF
rule. Since the BIF rule would subject
some of these facilities to RCRA
regulation for the first time (assuming
no other permitted units are at the
facility), these SQB facilities would be
eligible for interim status. See 56 FR at
7186 (February 21,1991) for
requirements regarding permit
modifications, section 3010
notifications, and Part A permit
applications. Such sources would also
be required to submit a certification of
precompliance (required by
§ 266.103(b)) within 6 months of the
date of publication of the final rule in
the Federal Register, and a certification
of compliance (required by § 266.103(c))
within 18 months of the date of
publication of the final rule.
B. The Waiver of the PM Standard
Under the Low Risk Waste Exemption of
the BIF Rule Would Not Be Applicable
to HWCs
Section 266.109 of the BIF rule
provides a conditional exemption from
the destruction and removal efficiency
(DRE) standard and the particulate
matter (PM) emission standard. The
DRE standard is waived if the owner or
operator complies with prescribed
procedures to show that emissions of
toxic organics are not likely to pose a
potential hazard to human health
considering the direct inhalation
pathway. The PM standard is waived if
the DRE standard is waived and the
source complies with the Tier I or
adjusted Tier I feedrate limits for metals.
We are proposing today to restrict
eligibility for the waiver of the PM
standard to BIFs other than cement and
lightweight aggregate kilns. This is
because: (1) Compliance assurance with
the proposed MACT standards for D/F,
SVM, and LVM is based on compliance
with a CEM-monitored, site-specific PM
emission limit;208 and (2) the proposed
MACT PM standard would be used to
help minimize emissions of adsorbed
2os NOf to exceed the proposed national MACT
standard.
non-D/F organic HAPs. Given that this
restriction for cement and lightweight
aggregate kilns is needed to ensure
compliance with the proposed MACT
standards, the restriction would be
effective at the time that the kiln begins
to comply with the MACT standard (i.e.,
when the source submits the initial
notification of compliance).
Finally, we note that, as a practical
matter, we believe that this proposed
restriction of eligibility for the PM
waiver for kilns will have little or no
effect on the regulated community. We
are not aware of any cement or
lightweight aggregate kilns that both
meet the conditions for the exemption
and have elected or intend to elect to
request the waiver.
The Agency solicits comment on the
application of waste minimization to
lower the volume of waste streams fed
to combustors so that the combustor can
meet the proposed revised SQB feed
limitations. Such reductions might be
achieved by meeting the proposed
HWIR standards and thus removing
entire streams from Subtitle C
requirements. The Agency is
particularly interested in technical and
economic information about commercial
or experimental processes to reduce
stream volume.
C. The "Low Risk Waste" Exemption
from the Emission Standards Provided
by the Existing Incinerator Standards
Would Be Superseded by the MACT
Rules
Section 264.340(c) exempts certain
incinerators from the emission
standards if the hazardous waste burned
contains insignificant concentrations of
Appendix VIII, Part 261, hazardous
constituents which would reasonably be
expected to be in the waste. In
implementing this provision, the
Agency has used various measures of
risk potential to define "insignificant"
concentrations. We believe that a risk-
based waiver is inconsistent with
today's proposed technology-based
MACT standards for incinerators, and in
any case could not supersede those
standards. Thus, we are proposing that
this provision no longer be applicable to
an incinerator at the time it begins
complying with the MACT standards
(i.e., when the initial notification of
compliance is submitted).
We also note that § 264.340(b)
provides the same exemption from
emission standards if the hazardous
waste burned does not contain any (i.e.,
nondetect levels) of the Appendix VIII
constituents. We are proposing that this
provision also be superseded by the
proposed MACT standards because: (1)
Detection limits may be high for some
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17471
waste matrices; and (2) nontoxic
organics in the waste can result in
emissions of toxic organics under poor
combustion conditions or conditions
favorable to formation of D/F in the
post-combustion zone (e.g., aPM
control device operating at temperatures
above 400°F).
D. Bevill Besidues
1. Required Testing Frequency for Bevill
Residues
The Agency is proposing to set a
minimum sampling and analysis
frequency for residues derived from the
burning or processing of hazardous
waste in units that may qualify for the
Bevill exemption by satisfying the
requirements of § 266.112 (a) and (b).
The Agency believes a minimum testing
frequency is necessary to prevent large
quantities of hazardous residues from
being managed in an environmentally
unsound manner.
Current regulations require that waste
derived residue be sampled and
analyzed "as often as necessary to
determine whether the residue
generated during each 24-hour period"
meets requirements to qualify for the
Bevill exemption. Because large
volumes of residue are generated in any
24-hour period, it is possible that a
facility may have disposed of the
residue after a sample had been taken,
but before the analysis results are
received. The Agency stated in the
preamble to the BIF regulations (56 FR
42504 (August 27,1991)) that "if the
waste derived residue is sampled and
analyzed less often than on a daily
basis, and subsequent analysis
determines that the residue fails the test
and is fully regulated hazardous waste,
the Agency considers all residue
generated since the previous successful
analysis to be fully regulated hazardous
waste absent documentation otherwise."
Residue generated after the failed test
may also be considered hazardous waste
until the next passing test. The residue
disposal area or unit would also become
subject to Subtitle C requirements.
In the interest of protecting human
health and the environment and
avoiding the scenarios mentioned
above, the Agency is today proposing
that if a facility elects to sample and
analyze less frequently than every day,
approval must be granted by the
Regional Administrator and the
sampling and analysis frequency used
must be based on and justified by
statistical analysis. The Agency is also
proposing that, in the event the Regional
Administrator approves less than daily
sampling at a facility, the facility must,
at a minimum, sample and analyze its
residues at least once every month for
metals and once every six months for
other compounds. A more frequent
minimum sampling frequency has been
proposed for metals because of the
variability of metal content in feed
materials and because metals cannot be
destroyed in the furnace. The proposed
sampling frequency will minimize the
possibility of large volumes of
hazardous residues being placed on the
land or otherwise being stored or
disposed of contrary to Subtitle C
requirements. The Agency does not
believe these proposed requirements
will unduly burden the regulated
community and requests comments on
this issue.
The following factors must be
considered when determining an
appropriate sampling frequency:
—Selection of a statistical method and
distribution of data (normal or log
normal distribution)
—Feedrates of wastes and all other feed
streams
—Volatility of metals in all feed streams
—Physical form of various feed streams
(solid versus liquid)
—Type of feed system
—Levels and types of organic
constituents in all feedstreams (for
example, difficulty of destruction or
formation of by-products)
—Levels and types of metals regulated
under RCRA, other than those
regulated by the BIF regulations (for
example, selenium)
—Changes in feed streams
—Changes in operating conditions or
equipment
—Operating conditions when sampling
compared with those when not
sampling
—Trends in partitioning of metals in fly
as compared with bottom ash
Facilities with a high variability of
hazardous constituents in their residues
should closely examine these factors in
deciding upon a sampling frequency.
Facilities with residues that exhibit
little or no constituent variability may
be able to sample at the minimum
frequency, pending approval of less
than daily sampling by the Regional
Administrator.
2. Dioxin Testing of Bevill Residues
a. Regulatory History. Under 40 CFR
§266.112 of the boiler/industrial
furnace (BIF) rule, EPA codified
procedures for owners and operators of
Bevill devices to determine whether
their residues retain the Bevill
exemption when the facilities co-fire or
co-process hazardous waste fuels along
with fossil fuels or normal raw
materials. These procedures were
deemed necessary to ensure that the
burning of hazardous waste does not
alter the residues so that they are no
longer the "high volume, low hazard"
materials exempted by the Bevill
amendment. This test was upheld by the
B.C. Circuit in Horsehead Resource
Development Co. v. Browner, 16 F. 3d
1246 (B.C. Cir. 1994).
Specifically, 40 CFR § 266.112
requires facilities that claim the Bevill
exemption for residues from co-burning
hazardous waste along with Bevill raw
materials to conduct sampling, and
analysis of their residues to document
that either: (1) Levels of toxic
constituents in the waste-derived
residue are not significantly higher than
normal (i.e., when not burning
hazardous waste) residues; or (2) levels
of toxic constituents in waste-derived
residue do not exceed health-based
levels specified in the rule. This is
commonly referred to as the two-part
Bevill test. The constituents for which
analysis must be conducted include: (1)
Appendix VIII, Part 261, hazardous
constituents that could reasonably be
expected to be in the hazardous waste
burned, and that are listed in § 268.40
for F039 non-wastewaters (see 59 FR
4982 of September 19,1994); and (2)
compounds that the Agency has
determined are common products of
incomplete combustion (i.e., they may
be formed during combustion of the
waste) and have been listed in
Appendix VIII of Part 266.
b. Addition of Dioxin/Furan
Compounds to the Appendix VIII, Part
266 Product of Incomplete Combustion
List. The Appendix VIII, Part 266
product of incomplete combustion (PIC)
list does not currently include
polychlorinated dibenzo-p-dioxin
(PCBD) and polychlorinated dibenzo-
furan (PCBF) compounds. In addition,
most BIF facilities do not burn wastes
which could reasonably be expected to
contain PCBB/PCBF compounds. Thus,
few § 266.112 facilities have been
analyzing their residues on a routine
basis for PCBB/PCBF compounds to
determine whether burning hazardous
waste has affected the character of the
residue.
EPA believes that it is important to
add PCBB/PCBF compounds to the PIC
list in order to make residue analysis for
PCBB/PCBFs a mandatory component
of the two-part Bevill test. First, dioxin/
furan compounds are likely to be PICs
and, as such, should rightfully be
included on the PIC list. As described
in Chapter 4 of the May 1994 Braft
Combustion Emissions Technical
Resource Document (CETREB), there is
a considerable body of evidence to show
that PCBB/PCBF compounds can be
-------
17472
Federal Register / Vol. 61, No, 77 / Friday, April 19, 1996 / Proposed Rules
formed in the post-combustion regions
of boilers, industrial furnaces and
incinerators, even if no PCDD/PCDF
compounds are fed to the combustion
device. Secondly, the level of dioxins in
residues can be influenced by hazardous
waste burning activities. The October
1994 Cement Kiln Dust Notice of Data
Availability, which augmented the
December 1993 Report to Congress on
Cement Kiln Dust, provided a regression
analysis to determine the impact of
hazardous waste fuel use on dioxin and
furan concentrations. Every one of the
dioxins and furans evaluated appeared
in significantly higher concentrations in
cement kiln dust generated by plants
that burned hazardous waste fuel in
comparison with plants that did not
burn any hazardous waste fuels. The
Report concluded that the strength and
consistency of this relationship for
cement kiln dust was striking, and that
it provides very strong evidence that
dioxin and furan concentrations in the
dust are systematically higher at plants
that burn hazardous waste fuel.
Finally, it is important to note that,
where the potential for excess risks were
identified in the Report, the constituents
of concern included metals and dioxin/
furan compounds. Metals are already
covered by the two-part test of
§ 266.112. However, it is equally
important to include PCDDs/PCDFs in
the two-part test to make sure that
residues from hazardous waste-burning
devices continue to meet the high
volume, low hazard criteria presumed
by the Bevill exemption.
c. Use of Land Disposal Restriction
Standards as Interim Limits for PCDD/
PCDFs. On November 9,1993, EPA
published an interim final rule
establishing alternate concentration
limits for nonmetals to be used for the
health-based comparison portion of the
two-part Bevill test (i.e., 40 CFR
§ 266.112(b)(2)). The alternate levels
were based on the land disposal
restriction (LDR) limits for F039 non-
wastewaters pending further
administrative action to determine
whether more-appropriate health-based
levels should be developed. Although
the LDR limits are not health-based
levels, the Agency noted in the
preamble (58 FR at 59598 (Nov. 9,
1994)) that the technology-based LDR
treatment limits should serve to identify
residues that have the "low toxicity"
attribute that is one of the key bases for
the temporary exemption of Bevill
residues from the definition of
hazardous waste. See Horsehead
Resource Development Co. v. Browner,
16 F. 3d. The Agency also noted that the
LDR levels are promulgated limits and
so have been scrutinized and subject to
public comment in previous
rulemakings.
As part of today's proposal to add
PCDD/PCDF constituents to the
Appendix VIII, Part 266 PIC list, the
Agency would continue the interim
practice of basing the concentration
limits for the health-based portion of the
two-part Bevill test on the LDR F039
nonwastewater levels. The LDR
regulation establishes concentration
limits of 1 part-per-billion (ppb) for total
HxPCDDs, total HxPCDFs, total
PePCDDs, total PePCDFs, total TCDDs
and total TCDFs. The Agency believes
that these levels for dioxin/furan
compounds will serve as adequate
screening levels on an interim basis to
ensure that residues from hazardous
waste-burning devices continue to meet
the "low toxicity" attribute presumed
by the Bevill exemption.
The Report to Congress on Cement
Kiln Dust provides some support for the
1 ppb PCDD/PCDF screening criteria. In
baseline risk modeling for fifteen case
study facilities managing CKD on-site,
dioxin/furan compounds were not
identified as contributors to adverse
health effects for either direct or indirect
exposure pathways (see Report, Exhibit
6-14). Risk from PCDD/PCDFs only
reached levels of concern when the
Agency performed a sensitivity analysis
to examine the change in risks that
would occur at five baseline facilities
based on the hypothetical management
of CKD containing the highest measured
PCDD/PCDF concentrations found in
EPA's sampling at 11 cement plants.
The highest concentrations were
observed in samples from a cement
facility, and were at least 2V2 times
higher than concentrations observed at
any other facility. All of the samples
from that facility exceeded 1 ppb for at
least one homolog listed as part of the
LDR F039 criteria (i.e., total HxPCDDs,
total HxPCDFs, total PePCDDs, total
PePCDFs, total TCDDs or total TCDFs).
Thus, the levels which showed potential
for adverse health effects in the site-
specific modeling would be screened by
application of the 1 ppb criteria listed
in the F039 LDR. By comparison, none
of the samples from facilities other than
the above facility had any PCDD/PCDF
homologs exceeding 1 ppb.
The Agency is proposing continued
use of the LDR levels because it does not
believe that it is appropriate to establish
a more specific health-based level for
dioxin/furan compounds at this time.209
209 EPA notes that, by establishing LDR exemption
levels for Bevill residue, the Agency is not
suggesting that: (l) the technology-based treatment
standards are equivalent to, or appropriate to use
as, health-based limits; or (2) Bevill excluded
residues should necessarily be subject to the LDR
A separate regulatory process is
underway which will establish controls
on management of cement kiln dust (60
FR 7366). Any health-based level
established in advance of these
controlled CKD management standards
would quickly become obsolete because,
at a minimum, the fate and transport
assumptions would be different. The
Agency specifically requests comment
regarding whether the interim LDR F039
limits for PCDD/PCDF constituents are
appropriate. Alternatively, the Agency
requests information regarding an
appropriate methodology for
establishing more specific health-based
limits.
d. Clarification of Appendix VIII, Part
266 PIC List Applicability. There has
historically been some confusion
regarding whether each of the
constituents listed on the Appendix
VIII, Part 266 list must be a mandatory
component of the residue testing at
every facility, or whether a facility
could exclude some of the constituents
on the list. Today, the Agency clarifies
that the Appendix VIII, Part 266 list is
applicable to every facility in its
entirety, without exclusion.
3. Application of Derived From Rule to
Residues From Hazardous Waste
Combustion in non-Bevill Boilers and
Industrial Furnaces
As part of a settlement agreement of
the lawsuit over the 1991 BIF
regulations, EPA agreed to reconsider
the appropriateness of applying the
derived from rule to residues from co-
processing listed hazardous waste fuels
and raw materials in non-Bevill boilers
and industrial furnaces. An example
would be an oil-fired boiler burning
listed hazardous waste fuel and
generating emission control dusts or
scrubber effluents, which dusts or
effluents would not be considered to be
Bevill excluded. If this type of burning
occurs in a boiler or furnace whose
residues are otherwise within the scope
of the Bevill amendment, the residues
remain exempted from subtitle C (i.e.
remain exempted by virtue of the Bevill
amendment) so long as they are not
"significantly affected" by burning
hazardous waste. § 266.112. A residue is
not significantly affected if there is no
statistically significant increase between
baseline, non-hazardous waste-derived
residues, or if hazardous constituents in
the residue do not exceed health-based
(or health-based surrogate) levels. Id.
Consistent with the settlement
agreement mentioned above, EPA
solicits comment as to whether this
rules. See 58 FR at 59603 (November 9,1994).
These issues are the subject of other rulemakings.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17473
same type of test could be applied to
burning of hazardous waste in non-
Bevill boilers and furnaces. The logic
could be that if hazardous properties are
not contributed by the hazardous waste,
the derived from rule should not apply.
EPA's inclination is not to apply any
type of significantly affected test to
residues at this time. The recently-
proposed exit levels, and methodology,
in the Hazardous Waste Identification
Rule (HWIR) provide a means of
automatic exit from the subtitle C
system when wastes (including derived-
from wastes) are no longer hazardous.
Furthermore, the "significantly
affected" test is closely linked to the
Bevill amendment, and in fact defines
the scope of that amendment in co-
processing situations. EPA sees no
persuasive reason to apply the test to
non-Bevill residues, particularly when
the Agency has proposed a means
whereby such residues can
automatically exit the system. It appears
to EPA to be the better approach to
make subtitle C exit determinations on
the basis of hazards actually posed by
the waste rather than by comparisons
\vith a non-waste baseline. (Indeed, this
is one component of the significantly
affected test already. See
§ 266.112(b)(2).) The Agency solicits
comment on this matter, however.
E. Applicability of Regulations to
Cyanide Wastes
The Agency has received several
inquiries regarding the applicability of
§ 266.100(c)(2)(i) criteria for processing
cyanide wastes solely for metal
recovery. Specifically, cyanide wastes
do not meet the common dictionary
meaning of being an organic, but can be
destroyed by industrial furnaces. The
Agency's intent of this exemption was
to preclude burning of waste streams
that contain greater than 500 ppm
nonmetal compounds listed in
Appendix VIII of Part 61, that are
provided a level of destruction by the
furnace. The Agency inappropriately
chose the word 'organic' instead of
'nonmetal' in the above regulation. An
amendment is being proposed to
provide the needed clarification that
wastes containing cyanides are eligible
to be included in this exemption. We
are also proposing similar amendments
(i.e., revisions to use the term
"nonmetal" rather than "organic") to
subparagraphs (c)(2)(ii), (c)(3)(i)(B), and
F. Shakedown Concerns
There is a concern within the Agency
that some new units do not effectively
use their allotted 720 hour pre-trial burn
period (commonly referred to as
"shakedown") or extensions thereof to
correct operational problems prior to the
trial burn period. This ineffective use of
the pretrial burn period can potentially
lead to emission exceedances which
pose unnecessary risks to human health
and the environment. In addition,
failure(s) during trial burn testing at one
or more test conditions reduce a
facility's flexibility to burn hazardous
waste in a subsequent permit developed
from the trial burn or may even lead to
a need to perform other trial burns or a
termination of the permit. A failure to
perform adequate shakedown may also
lead to difficulties in making an
interpretation of trial burn data and in
setting of permit conditions due to
excessive variability in trial burn
operation.
The Agency believes that an approach
using system start-up and system
problem solving with the use of a non-
hazardous waste feed followed by a
gradual, carefully planned introduction
of hazardous waste feed is essential to
avoid the potential problems which
could result from the burning of
hazardous waste in an undiagnosed
system which may not yet be operating
at steady state conditions. The absence
of this type of approach has caused
many previous trial burns not to be
carried through to completion or has
caused them to occur in a very different
fashion from that prescribed in the trial
burn plan. Other efforts during the trial
burn have resulted in diminished
operating allowances or in the need for
additional trial burn testing. As a result
of these occurrences, the Agency is
proposing three options which center
around the pretrial burn period in an
attempt to enhance regulatory control
over trial burn testing. The Agency is
also requesting comment on the
applicability of these options to interim
status facilities. The shakedown period
has, in the past, been applied
exclusively to new facilities and has not
addressed existing facilities operating
under interim status. The Agency
believes that these options could apply
to interim status facilities if the newly
proposed waste to be burned
represented a very different waste than
that which had been burned.
As its primary option, the Agency
would require that facilities be required
to show the Director prior to trial burn
dates .being scheduled that the facility
has provided a minimum showing of
operational readiness. This showing of
operational readiness would be one
which has been established by the
Director and would be incorporated as
part of the permit application process
for both interim status and new devices.
The manner in which this notification
of readiness would occur would be
determined by the Director. A trial burn
could not be scheduled until this
minimum showing to the Director has
occurred. Criteria for trial burn
readiness would include, but would not
be limited to the following examples: (1)
The ability of a facility to show that it
has operated the device to be permitted
under its planned trial burn conditions
(e.g. temperature, feedrate) for a
specified time period set by the
Director, or (2) the ability of a facility to
operate for a designated period of time
(to be.established by the Director)
without an Automatic Waste Feed Cut-
Off (AWFCO) occurring. To show
readiness to the Director, the
composition of the feed stream to the
device during this showing would need
to be nearly identical (if not identical)
to the waste intended to be burned
during the operational lifetime of the
facility. This similarity should be
consistent with respect to the physical,
thermal, and fluid characteristics of the
waste not only being burned during the
trial burn tests, but also during the
lifetime of the facility. It is the Agency's
belief that facilities which fail their trial
burn tests often fail because facilities
tend to stress their devices for the first
time only during trial burn testing. The
system has to that point never
undergone "break point" testing with an
increased feedrate or maximum capacity
feedrate. A trial burn should not be
scheduled until a facility has shown the
Director that it can operate without
constant shutdowns at feedrates
consistent with that of the trial burn.
A second option which the Agency
offers for comment is a more restrictive
option. This option proposes
requirements on both the operations
prior to and following the shakedown
period. It incorporates the notification
requirements found in the primary
option along with an additional
notification requirement which would
occur prior to the beginning of
shakedown. This option would require
a facility to notify the Director that it
has achieved steady state operation with
non-hazardous waste during this period
leading up to shakedown at operational
levels set by Director (e.g. flowrates)
which are comparable to that to be
tested at trial burn and to certify that the
device is ready to begin shakedown
operations. As before, this option would
also require a facility to notify the
Director following shakedown that
operational readiness with hazardous
waste has been achieved and to certify
that the device is ready for trial burn.
tests. Although this option would
impose two more operational
-------
17474
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 7 Proposed Rules
requirements for a facility, it would
ensure that the facility has brought the
device up to operational standards
whereby the addition of hazardous
waste would not represent an excessive
risk to human health or the
environment. The Agency believes that
this option would also provide for a
more efficient trial burn since it has
required a facility to become operational
without constant shutdowns prior to the
trial burn prior to shakedown and after
shakedown. Portions of this option may
not be directly applicable to interim
status facilities since they have been
burning hazardous waste to date and
may have most of their operational
problems worked out.
A third option upon which the
Agency is requesting comment is a
"guidance only" option. Although this
option would not impose any specific
regulatory requirements for a showing of
operational readiness prior to or after a
shakedown period, it would provide
guidance to industry and permit writers
on how to effectively achieve
preparedness prior to a trial burn
without the need of formalizing it
within the constraints of the regulations.
Permit writers would have the ability, as
they do now, to set readiness
demonstration requirements if they
deem it necessary for a specific site.
G. Extensions of Time Under
Certification of Compliance
The Boiler and Industrial Furnace
Rule, at 40 CFR § 266.103(c)(7), allows
a facility to obtain a case-by-case
extension under certain circumstances
when events were outside of the control
of the facility. There have been
questions as to whether this provision
meant that after August 21,1992, a
facility could no longer apply for a case-
by-case extension. The Agency wants to
clarify that it never intended this
restrictive interpretation and so is
proposing to amend this section to
provide the clarification. EPA intended
the case-by-case extension to apply at
any time during the certification of
compliance cycle, including during
Revised Certification of Compliance
under § 266.103(c)(8j, and during
Periodic Recertifications under
§ 266.103(d). See 56 FR at 7182
(February 21,1991). The basis of
granting the case-by-case extension is
proposed to remain unchanged by
today's rule. Additionally, EPA is
clarifying that the automatic one year
extension is not valid for facilities
which were not in existence on August
21,1991.
H. Technical Amendments to the BJF
Rule
1. Facility Requirements at Closure
•EPA is today proposing to amend
§ 266.103(1) to stipulate that at closure,
the owner or operator must remove all
hazardous waste and hazardous waste
residues not only from the boiler or
industrial furnace, but also from its air
pollution control system (APCS).
Although the APCS is an integral part of
the facility, this minor amendment will
make it explicitly clear that no
hazardous waste or residues can remain
in the APCS after closure.
2. Definitions under the BIF Rule
We are adding several definitions
under § 260.10 for frequently used terms
in combustion regulations like fugitive
emissions, automatic waste feed cutoff
system, run, air pollution control system
and operating record. The purpose is to
clarify these technical terms of thermal
treatment, expedite permit writing as
well as increase the enforceability of
obvious technical violations. Some of
these definitions already exist in the air
regulations.
/. Clarification of Regulatory Status of
Fuel Blenders
EPA is proposing to revise 40 CFR
266.101 ("Management prior to
burning") to clarify that fuel blending
activities, including those which
constitute treatment, are regulated
under RCRA. Section 266.101 (formerly
266.34) was written with the
understanding that hazardous waste
fuel-blending activities were
traditionally performed in containers or
tank systems where the storage
standards of Part 264 could be applied.
The Agency believes that protection of
human health and the environment is
accomplished when the permit
addresses the containment of the waste
being treated. Therefore, no direct
reference to "treatment" was included
in Section 266.34; treatment was
understood to be implicit in the
regulation, as shown by the reference in
section 261.6 to the "* * * applicable
provisions of Part 270." EPA has in fact
explicitly interpreted § 266.101
(formerly § 266.34) to require tank
storage standards to apply to tanks in
which hazardous waste fuels are
blended. See 52 FR 11820 (April 13,
1987).
More recently, it has come to the
Agency's attention that fuel blenders
may be using devices such as
microwave units and distillation
columns in their hazardous waste
handling operations that differ from the
traditional fuel-blending practices.
These practices are, in fact, hazardous
waste treatment activities requiring a
RCRA permit, without which the unit
cannot operate. For many such
operations, the "miscellaneous unit"
requirements of Part 264, Subpart X,
would apply. Due to various inquiries
regarding this issue, EPA has written
several policy memoranda confirming
that treatment, as well as storage,
conducted by fuel blenders requires a
RCRA permit. These memoranda are
part of the Agency's RCRA Permit
Policy Compendium and are available
from the RCRA Hotline. They are also
included in this rulemaking docket.
EPA is taking this opportunity to clarify
this issue in the regulations by revising
the language in § 266.101.
/. Change in Reporting Requirements for
Secondary Lead Smelters Subject to
MACT
EPA recently promulgated MACT
standards for the secondary lead smelter
source category. 60 FR 29750 (June 23,
1995). In that rule, the Agency found,
with unanimous support from
commenters, that RCRA emission
standards were unnecessary at the
present time for these sources since the
MACT standards provide significant
health protection, area secondary lead
sources will be regulated by these
MACT standards, and the ultimate issue
of the protectiveness of the standard
•will be evaluated during the section
112(f) residual risk determination.
EPA is proposing here to modify
existing § 266.100(c), which provides an
exemption from RCRA air emission
standards for (among other sources)
industrial furnaces burning hazardous
waste solely for material recovery.
Secondary lead smelters complying
with conditions enumerated in
§ 266.100(c)(l) and (3) are among this
type of industrial furnace. The Agency
is proposing to amend § 266.100(c)and
is proposing to add a new § 266.100(g)
to state that RCRA provisions for air
emissions do not apply to secondary
lead smelters when the MACT rule takes
effect (in June, 1997), provided the
smelters do not burn hazardous wastes
containing greater than 500 ppm
nonmetal hazardous constituents (or
burn wastes enumerated in 40 CFR Part
266 Appendix XI), submit a one-time
notice to EPA or an authorized state,
sample and analyze as necessary to
document the basis for their claim, and
keep appropriate records. These
amendments also could take the form of
an exemption (subject to the same
conditions) for such secondary lead
smelters from the present proposed rule.
:. • This proposed amendment is similar
to the exemption found in the existing
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17475
RCRA BIF rules but does eliminate
certain recordkeeping and reporting
requirements for secondary lead
smelters presently required as a
condition of the RCRA exemption. The
Agency tentatively does not believe
these extra reporting requirements are
needed once the MACT standards take
effect. At the same time, secondary lead
smelters choosing to burn hazardous
wastes different from those evaluated in
the secondary lead NESHAP (i.e.
hazardous wastes with greater than 500
ppm toxic nonmetals or those hazardous
waste not listed in Appendix XI to Part
266) would have to meet applicable
standards for hazardous waste
combustion units (i.e. either the existing
BIF standards or revised standards
based on MACT), as well as those for
secondary lead smelters. EPA would
administer this proposal by not
requiring a secondary lead smelter that
has already submitted a notification to
EPA or an authorized state under
existing 266.100(c)(l) or (3), to renotify
under proposed 266.100 (g).
PART SEVEN: ANALYTICAL AND
REGULATORY REQUIREMENTS
I. Executive Order 12866
Under Executive Order 12866, (58 FR
51735 (October 4,1993)) the Agency
must determine whether this regulatory
action is "significant." A determination
of significance will subject this action to
full OMB review and compliance under
Executive Order 12866 requirements.
The order defines "significant
regulatory action" as one that is likely
to result in a rule that may:
(1) Have an annual effect on the
economy of $100 million or more,
adversely affect in a material way the
economy, a sector of the economy,
productivity, competition, jobs, the
environment, public health or safety, or
state, local, or tribal governments or
communities;
(2) Create a serious inconsistency or
otherwise interfere with an action taken
or planned by another agency;
(3) Materially alter the budgetary
impact of entitlement, grants, user fees,
or loan programs, or the rights and
obligations of recipients thereof; or
(4) Raise novel legal or policy issues
arising out of legal mandates, the
President's priorities, or the principles
set forth in the terms of the Executive
Order.
The Agency believes that today's
proposal, represents a significant action.
If adopted, the proposed rule would
most likely result in a cost greater than
$100 million. As a result, this
rulemaking action, and supporting
analyses, are subject to full OMB review
under the requirements of the Executive
Order. The Agency has prepared
"Regulatory Impact Assessment for
Proposed Hazardous Waste Combustion
MACT Standards" and "Addendum to
the Regulatory Impact Assessment for
Proposed Hazardous Waste Combustion
MACT Standards" in support of today's
action; this report is available in the
public docket for today's rule. A
summary of this analysis and findings is
presented below.
II. Regulatory Options
During the regulatory developmental
phases, EPA considered seven different
regulatory MACT options for existing
sources. Refer to the RIA for a detailed
discussion of the seven options. This
preamble discusses and assesses the
floor option and the Agency preferred
option. For more detail on the specific
methodology used in developing floor
and "beyond-the-floor" control levels,
the reader should refer to the preamble
Options section, Part Four of this
preamble. Below is a summary of the
impact of floor levels and the preferred
option 1 on the combustion industry.
TOTAL ANNUAL COMPLIANCE COSTS
[Millions]
III. Assessment of Potential Costs and
Benefits
A. Introduction
The Agency has prepared a regulatory
impact assessment to accompany
today's proposed rulemaking. The
Agency has evaluated cost, economic
impacts, and other impacts such as
environmental justice, unfunded
mandates, regulatory takings, and waste
minimization incentives. The focus of
the economic impact assessment was on
how the MACT standards may affect the
hazardous waste-burning industry. The
Agency would like to note that although
the cement kiln industry profits are
generated by two components: cement
production and hazardous waste
burning, the RIA only estimated the
impact the MACT standards will have
on hazardous waste burning. The
Agency is in the process of beginning an
analysis that will study the impact of
today's rule on cement production,
cement prices, and competition in the
cement industry. The Agency would
like to solicit comments and request
information in this area as we begin our
research.
To develop cost estimates, EPA
categorized the combustion units by
size, and estimated engineering costs for
the air pollution control devices
(APCDs) needed to achieve the
standards in the regulatory options.
Based on information regarding current
emissions and APCD trains EPA
developed assumptions regarding the
type of upgrades that units would
require. Because EPA's data was
limited, this analysis is meant to
develop estimates of national economic
impacts, and not site specific impacts.
B. Analysis and Findings
Total annual compliance costs for the
floor option and the Agency's proposed
standards range in costs from an
estimated $93 million to $136 million.
Options
6 percent Floor
6 percent BTF
Cement
kilns
$27
44
LWA
kilns
$2
4
Commer-
cial incin-
erators
$13
20
On-site
inciner-
ators
$50
67
Total
$93
136
This rule will result in a significant
impact to the combustion industry. The
regulatory impact assessment used a
number screening indicators to assess
the impact of this rule. One Indicator
the analysis used was the average total
annual compliance cost per unit. This
indicator assesses the relative impact
the rule has on each facility type in the
combustion universe. According to this
indicator, cement kilns incur the
greatest average incremental cost per
unit totaling $770,000 annually for the
floor and $1.1 million annually for the
proposed standards, which include
beyond the floor standards. The cost per
unit for LWAKs range from $490,000 to
$825,000 and for on-site incinerators
from $340,000 to $486,000. Commercial
incinerators annual average cost per
unit total $493,000 for the floor and
-------
17476
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
$730,000 for the proposed standards.
One should note however, that the per
unit costs are presented assuming no
market exit. Once market exit occurs,
per unit should be significantly lower
particularly for on-site incinerators.
Looking at the price per ton, in the
baseline, cement kilns have the lowest
cost ($104 per ton) to burn hazardous
waste today with commercial
incinerators have $800 per ton costs and
on-site incinerators have $28,460 per
ton costs. For compliance costs, cement
kilns have the smallest impact ($40 to
$50 per ton) with on-site incinerators
experiencing a high compliance cost of
$47 to $57 per ton.
EPA also looked at baseline cost of
burning hazardous waste as a
percentage of compliance cost. This
indicator assesses the relative impact of
facilities within the sector but it also
can be a predictor for how prices might
increase for burning hazardous waste.
According to the table below, the floor
compliance costs are 40 percent of the
current baseline cost of burning
hazardous waste for cement kilns and
over 20 percent for LWAKs. Many on-
site incinerators and commercial
incinerators have existing APCDs and
have larger volumes of waste to
distribute compliance costs across, thus
compliance costs tend to be a smaller
addition to baseline costs.
AVERAGE TOTAL ANNUAL BASELINE—INCREMENTAL COMPLIANCE
[Cost per Ton]
Options
Baseline
6 percent Floor
6 percent BTF
Cement
kilns
$104
$40
50
LWA
kilns
$194
$39
56
Commer-
cial incin-
erators
$806
$23
31
On-site
inciner-
ators
$28 500
$47
57
Note: Baseline costs were calculated by identifying all costs associated with hazardous waste burning. Thus, for commercial incinerators and
on-site incinerators, all costs associated with unit construction, operation and maintenance are included. This also includes RCRA permits and
existing APCDs. The costs for on-site burners are extremely high because total costs for incineration is distributed across the small amount of
hazardous waste burned. For cement kilns and LWAKs, only those incremental costs associated with burning hazardous waste are included
such as, permits. The cost of the actual units (which have a primary purpose of producing cement or aggregate) are not included in the baseline.
Also these costs are after consolidation occurs.
Although cement kilns incur a
significant impact, they still have the
lowest average waste burning cost after
the regulation. As the table above
illustrates in the post-regulatory
scenario, cement kilns cost per ton for
burning waste would total $154
compared to a cost per ton for
commercial incinerators of $837. EPA
expects that this advantage for cement
kilns in the market will allow them to
continue to set the market price for
waste burning.
Not all facilities however, will be able
to absorb the compliance cost to this
rule and remain competitive. The
economic impact assessment estimates
that of the facilities which are currently
burning hazardous waste 3 cement
kilns, 2 LWAK, 6 commercial
incinerators and 85 on-site incinerators
will likely stop burning waste in the
long term. Most of these units are ones
which burn smaller amount of
hazardous waste.
C. Total Incremental Cost per
Incremental Reduction in HAP
Emissions
Cost effectiveness is calculated by
first estimating the compliance
expenditures associated with the
specific hazardous air pollutant (HAP).
The estimation of costs per HAP is often
difficult to ascertain because the air
pollution control devices usually
control more than one HAP. Therefore,
estimation of precise cost per HAP was
not feasible. Once the compliance
expenditures has been estimated, the
total mass emission reduction achieved
when combustion facilities comply with
the standards for a given option must be
estimated. With the total compliance
costs and the total mass emissions, the
total incremental cost per incremental
reduction in HAP emissions can be
estimated. For a more detailed
discussion of how the cost per HAP was
calculated, please see chapter 5 of
"Regulatory Impact Assessment for
Proposed Hazardous Waste Combustion
MACT Standards".
Results of the cost-effectiveness
calculations for each HAP for all
facilities are found below. For results on
a facility-type level, please see chapter
5 of the RIA. Considering all facilities as
a group, the results indicate that dioxin,
mercury, and metals cost per unit
reduction are quite high. This is the case
because small amounts of the dioxin
and metals are released into the
environment. For other pollutants,
expenditures per ton are much lower.
COST EFFECTIVENESS FOR ALL
FACILITIES
COST EFFECTIVENESS FOR ALL
FACILITIES—Continued
HAP
D/F
Mercury
LVM
SVM
Chlorine
*
Unit
$/g
$/lb
$/Mton ...
$/Mton ...
$/Mton ...
Baseline
to 6 per-
cent floor
$12000
2,600
407,000
315,000
7,000
6 percent
floor to 6
percent
BTF
$560 000
5,400
NA
NA
2,240
HAP
Particu-
late.
CO
THC
Unit
$/Mton ...
$/Mton
$/Mton ...
Baseline
to 6 per-
cent floor
4,400
1 360
2,800
6 percent
floor to 6
percent
BTF
3,200
NA
NA
Note: NA = Zero incremental reduction in
HAP emissions (Dollars divided by zero =
NA).
D. Human Health Benefits
1. Dioxin benefits
Fob/chlorinated dibenzo-p-dioxins
and polychlorinated dibenzofurans,
hereafter referred to collectively as
dioxins, are ubiquitous in the
environment. The more highly
chlorinated dioxins, which are
extremely stable under environmental
conditions, persist in the environment
for decades and are found particularly
in soils, sediments, and foods. It has
been hypothesized that the primary
mechanism by which dioxins enter the
terrestrial food chain is through
atmospheric deposition.210 Dioxins may
be emitted directly to the atmosphere by
a variety of anthropogenic sources or
indirectly through volatilization or
particle resuspension from reservoir
^10USEPA, "Estimating Exposure to Dioxin-Like
Compounds", Volume I, June 1994. ,
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17477
sources such as soils, sediments, and
vegetation.
The most well known incident of
environmental contamination with
dioxins occurred in Seveso, Italy in an
industrial accident. Symptoms of acute
exposures such as chloracne occurred
immediately following the incident.
Since then, significant increases in
certain types of cancers have also been
observed.21' After evaluating a variety of
carcinogenicity studies in human
populations and laboratory animals,
EPA has concluded that 2,3,7,8-
tetrachlorodibenzo-p-dioxin and related
compounds are probable human
carcinogens.212 EPA estimates that a
dose of 0.01 picograms on a toxicity
equivalent (TEQ) basis per kilogram
body weight per day is associated with
a plausible upper bound lifetime excess
cancer risk of one in one million
(1x10 — 6).213 Toxicity equivalence is
based on the premise that a series of
common biological steps are necessary
for most if not all of the observed
effects, including cancer, from
exposures to 2,3,7,8 chlorine-substituted
dibenzo-p-dioxin and dibenzofuran
compounds in vertebrates, including
humans. Given the levels of background
TEQ exposures discussed below, as
many as 600 cancer cases may be
attributable to dioxin exposures each
year in the United States.
EPA has also concluded that there is
adequate evidence from both human
populations and laboratory animals, as
well as other experimental data, to
support the inference that humans are
likely to respond with a broad spectrum
of non-cancer effects from exposure to
dioxins if exposures are high enough.
Although it is not possible given
existing information to state exactly
how or at what levels exposed humans
will respond, the margin of exposure
between background TEQ levels and
levels where effects are detectable in
humans is considerably smaller than
previously thought.214
Dioxins are commonly found in food
produced for human consumption.
Consumption of dioxin contaminated
food is considered the primary route of
exposure in the general population. EPA
evaluated data collected in four U.S.
studies, three of which included
analyses of all 2,3,7,8 chlorine-
substituted congeners of dibenzo-p-
dioxin and dibenzofuran. EPA's
evaluation concluded that
"background" levels in beef, milk, pork,
chicken, and eggs are approximately 0.5,
0.07, 0.3, 0.2, and 0.1 parts per trillion
fresh weight, respectively, on a toxicity
equivalent (TEQ) basis.215 EPA then
used these background levels, together
with information on food consumption,
to estimate dietary intake in the general
population. That estimate is 120
picograms TEQ per day.216
EPA has also collected data on
dioxins in fish taken from 388 locations
nationwide and found that at 89 percent
of the locations, fish contained
detectable levels of at least two of the
dioxin and furan compounds for which
analyses were conducted.217 (Of the
2,3,7,8 chlorine-substituted congeners,
only octachlorodibenzo-p-dioxin and
octachlorodibenzofuran were not
analyzed.) Seven of the compounds,
including 2,3,7,8-TCDD, were detected
at over half the locations. Detection
limits were generally at or below 1 part
per trillion on a toxicity equivalent
basis. The median (50th percentile)
concentration in fish on a toxicity
equivalent basis (TEQ) was 3 parts per
trillion (ppt) while the 90th percentile
was approximately 30 ppt TEQ. Five
percent of the sites exceeded 50 ppt
TEQ. At most sites, both a composite
sample of bottom feeders and a
composite sample of game fish were
collected. At sites considered
representative of background levels, the
median concentration was 0.5 ppt TEQ.
EPA has estimated that hazardous
waste incinerators and hazardous waste-
burning cement and lightweight
aggregate kilns currently emit 0.08, 0.86,
and less than 0.01 kg TEQ of dioxins per
year, respectively, or a total of 0.94 kg
TEQ per year. Excluding non-hazardous
waste-burning cement kilns, an
emission rate of approximately 9 kg
TEQ per year is estimated for all other
U.S. sources.218 Therefore, hazardous
waste-burning sources represent about 9
percent of total anthropogenic emissions
of dioxins in the U.S. The following
table shows hazardous waste-burning
sources relative to other major emitters
of dioxins:
Ji'USEPA, "Health Assessment Document for
2,3,7,8-Totrachlorodibenzo-p-Dioxin (TCDD) and
Related Compounds, Volume II, June 1994.
J"USEPA, "Health Assessment Document for
2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and
Rotated Compounds, Volume III," August 1994.
Source category
Medical Waste Incinerators
Dioxin
emissions
(kg TEQ/
year)
5.1
215TJSEPA, "Estimating Exposure to Dioxin-Like
Compounds," Volume II, June 1994.
216ffild.
zi'USEPA, "National Study of Chemical Residues
in Fish," Office of Science and Technology,
September 1992. . . •-
2I8USEPA, "Estimating Exposure to Dioxin-Like
Compounds", Volume II, June 1994.
Source category
Municipal Waste Incinerators
Hazardous Waste-burning Inciner-
ators, Cement Kilns, and Light-
weight Aggregate Kilns ..-
Dioxin
emissions
(kg TEQ/
year)
3.0
0.9
There is information to suggest,
however, that dioxin emissions
nationwide from all sources are higher
than have been estimated. Public
comments on EPA's dioxin
reassessment have identified a number
of possible additional sources of
dioxins, including decomposition of
materials containing chlorophenols (i.e.
wood treated with PCP), metals
processing industries, diesel fuel and
unleaded gasoline, PCB manufacturing,
and re-ehtrainment of reservoir sources.
Reservoir sources may be a significant
source of vapor phase dioxins. On the
other hand, emissions from at least one
of the sources, medical waste
incinerators, is probably significantly
overestimated. Supporting the view that
dioxin emissions may be higher than
previously estimated are indications
that deposition may be considerably
greater than can be accounted for by
presently identified emissions.
The impact of emissions on exposure
and risk depends on the relative
geographic locations of the emission
sources and receptors which contribute
to exposure and risk, primarily farm
animals. This applies to both near field
dispersion and long-range transport and
it affects exposure and risk both in
determining whether the trajectory of an
air parcel impacts receptors of concern
and in determining the chemical fate of
the emissions. The fate of dioxins
depends on degradation processes that
can occur in the atmosphere. These
processes can increase or decrease the
toxicity of the original emissions
through dechlorination. This process
can have different effects on different
emission sources, depending on the
congener distributions, residence time
in the atmosphere, and climatic
conditions.
Considering all these factors, it is
apparent that hazardous waste-burning
sources contribute significantly to the
overall loading of dioxins to the
environment, although the relative
magnitude of the contribution remains
to be determined. While there is not a
one-to-one relationship between
emissions and risk, it may be inferred
that hazardous waste-burning sources
likely do contribute significantly to
dioxin levels in foods used for human
consumption and, to an extent as yet
unknown, the estimated 600 cancer
-------
17478
Federal Register / Vol. 61, No. 77 /Friday, April 19, 1996 / Proposed Rules
cases attributable to dioxin exposures
annually.
EPA estimates that dioxin emissions
from hazardous waste-burning sources
will be reduced to 0.07 kg TEQ per year
at the floor levels and to 0.01 kg TEQ
per year at the proposed beyond the
floor standard. These reductions would
result in decreases of approximately 8
and 9 percent, respectively in .total
estimated anthropogenic U.S. emissions.
EPA expects that reductions in dioxin
emissions from hazardous waste-
burning sources, in conjunction with
reductions in emissions from other
dioxin-emitting sources, will help
reduce dioxin levels over time in foods
used for human consumption and,
therefore, reduce the likelihood of
adverse health effects, including cancer,
occurring in the general population.
2. Mercury Benefits
Mercury has long been a concern in
both occupational and environmental
settings. The most bioavailable form of
mercury and, therefore, the form most
likely to have an adverse effect, is
methyl mercury. Human exposures to
methyl mercury occur primarily from
ingestion of fish. As a result of mercury
contamination, there are currently fish
consumption bans or advisories in effect
for at least one waterbody in over two
thirds of the States.
Nationally, about 60 percent of all fish
consumption bans and advisories are
due to mercury. In several States the
mercury advisories are statewide, with
the most widespread concerns being in
the northern Great Lakes states and
Florida. The bans.and advisories vary
from State to State with respect to the
levels of concern, the recommended
limits on consumption, and other
factors. Therefore, it is difficult to
develop a national estimate of potential
risk based on this information.
Nevertheless, these bans and advisories
provide one indication of the extent and
severity of mercury contamination.
Even low levels of mercury in surface
waters can lead to high levels of
mercury in fish. EPA has estimated that
bioaccumulation factors, which
represent the ratio of the total mercury
concentration in fish tissue to the total
concentration in filtered water, range
from 5,000 to 10,000,000 depending on ,
the species of fish, the age of the fish,
and the waterbody the fish inhabit.
The most well known example of
mercury poisoning from ingestion of
fish occurred in the vicinity of
Minamata Bay, Japan. Severe
neurological effects resembling cerebral
palsy occurred in the offspring of
exposed pregnant women. EPA has
estimated what it considers a safe level
of exposure to methyl mercury. This
level, referred to as the reference dose,
is 1E-4 mg/kg-day. The reference dose
is based on an evaluation of 81
maternal-infant pairs exposed to methyl
mercury in an incident in Iraq in which
methyl mercury treated seed grain was
diverted,for use in making bread.
Sources of uncertainty in the reference
dose are the relatively small number of
maternal-infant pairs in the Iraqi study,
the short duration of maternal exposure
(approximately three months), latency
in the appearance of effects (from as
little as a month to as long as a year),
possible misclassification of maternal
exposures, differences in the vehicle of ,
exposure (i.e., grain versus fish), and the
selection of the neurologic and
behavioral endpoints used in the
analysis. EPA intends to further
evaluate the reference dose for methyl
mercury when the results from studies
of fish-eating populations become
available.
EPA collected data on chemical
residues in fish taken from 388 locations
nationwide and found that at 92 percent
of the locations, fish contained
detectable levels of mercury.219
(Detection limits varied between 0.001
and 0.05 parts per million.) The median
(50th percentile) mercury concentration
in fish was 0.2 ppm while the 90th
percentile was 0.6 ppm. Two percent of
the sites exceeded 1 ppm. At most sites,
both a composite sample of bottom
feeders and a composite sample of game
fish were collected. The highest
concentration, 1.8 ppm, was measured
at a remote site considered to represent
background conditions.
Similar results have been obtained in
other studies, strongly suggesting that
long-range atmospheric transport and
deposition of anthropogenic emissions
is occurring. Air emissions of mercury
Contribute, then, to both regional and
global deposition, as well as deposition
locally. Congress, in fact, explicitly
found this to be the case and required
EPA to prioritize MACT controls for
mercury for this reason. (See S. Rep. No.
228,101st Cong. 1st Sess. at 153-54.)
An indication of the significance of
mercury contamination in fish is
illustrated by combining data on the
levels of mercury in fish with data on
fish consumption and comparing it to
the reference dose for methyl mercury.
For example, a fish consumption rate of
140 g/day (a 90th percentile rate
associated with recreational fishing) in
conjunction with a mercury
concentration of 0.6 ng/g (a 90th
21' USEPA,/'National Study of Chemical Residues
in Fish," Office of Science and Technology,
September 1992.
percentile concentration) translates into
an average daily dose of 1E-3 mg/kg-
day, or 10 times the reference dose.
Using the same fish concentration with
a mean fish consumption rate for
recreational anglers of 30 g/day gives a
dose that is three times the reference
dose. At the median fish concentration
of 0.2 ug/g and a fish consumption rate
of 30 g/day, the dose is nearly 90
percent of the reference dose. These
results indicate that for persons who eat
significant amounts of freshwater fish,
exposures to mercury are significant
when compared with EPA's estimate of
the threshold at which effects may occur
in susceptible individuals. However, it
must be recognized that EPA's threshold
estimate represents a lower bound; the
true threshold may be higher than EPA's
estimate.
EPA has estimated that hazardous
waste incinerators and hazardous waste-
burning cement and lightweight
aggregate kilns currently emit 4.2, 5.6,
and 0.3 Mg of mercury per year,
respectively, or a total of 10.1 Mg per
year. In addition, EPA estimates that
approximately 230 Mg per year are
emitted by all other U.S. sources. Based
on these estimates, hazardous waste-
burning sources represent about 4
percent of total anthropogenic emissions
of mercury in the U.S. Therefore,
hazardous waste-burning sources do
contribute to the overall loading of
mercury to the environment and, it may
be inferred, to mercury levels in fish.
EPA estimates that mercury emissions
from hazardous waste-burning sources
will be reduced to 3.3 Mg per year at the
proposed floor levels and to 2.0 Mg per
year at the proposed beyond the floor
standard. These reductions would result
in reductions of total anthropogenic
U.S. emissions of approximately 3
percent. EPA expects that reductions in
mercury emissions from hazardous
waste-burning sources, in conjunction
with reductions in emissions from other
mercury-emitting sources, will help
reduce mercury levels in fish over time
and, therefore, reduce the likelihood of
adverse health effects occurring in fish-
consuming populations.
E. Other Benefits
Other benefits that EPA investigated
included ecological benefits, property
value benefits, soiling and material
damage, aesthetic damages and
recreational and commercial fishing
impacts. Overall, the analysis of the
ecological risk suggest that only when
assuming very high emissions water
quality criteria is exceeded in the
watersheds small in size and located
near waste combustion facilities. These
watersheds are typically located near
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17479
cement kilns appear to exceed the water
quality criteria. According to the
property value analysis, there may be
property value benefits associated with
reduction in emission from combustion
facilities. The property value work is
on-going and is undergoing refinements.
In addition, EPA investigated other
benefits such as benefits received from
avoided clean-up as result of reduced
particulate matter releases. For further
detail, please see chapter 5 of the RIA.
IV. Other Regulatory Issues
A. Environmental Justice
The U.S. EPA completed analyses that
identified demographic characteristics
of populations near cement plants and
commercial hazardous waste
incinerators and compared them to the
populations of county and state. The
analysis focuses on the spatial
relationship between cement plants and
incinerators and minority and low
income populations. The study does not
describe the actual health status of these
populations, and how their health might
be affected proximity to facilities.
EPA used a sample of 41 cement
plants was analyzed from a universe of
113 plants and a sample of 21
commercial incinerators was analyzed
from a universe of 35. The complete
methodology results of the analyses are
found in two reports filed in the docket
titled, "Race , Ethnicity, and Poverty
Status of the Populations Living Near
Cement Plants in the United States and
Race," "Ethnicity, and Poverty Status of
the Populations Living Near
Commercial Incinerators." Below is a
summary of the key results found in the
studies.
The Agency looked at whether
minority percentages within a one mile
radius are significantly different than
the minority percentages at the county
for all cement plants and sample of
incinerators, the results are as follows:
^•27 percent of the universe of all
cement plants (29 plants) and 37
percent of sample of incinerators (21
plants) have minority percentages
within a one mile radius which exceed
the corresponding county minority
percentages by more than five
percentage points.
^ 36 percent of the universe of all
cement plants (41 plants) and 44
percent of sample of incinerators have
minority percentages within a one mile
radius which fall below the
corresponding county minority
percentages by more than five
percentage points.
>• 38 percent of the universe of all
cement plants (43 plants) and 20
percent of sample of incinerators
minority percentages within a one mile
radius which fall within five percentage
points (above or below) of the
corresponding county minority
percentages.
With regard to the question of
whether poverty percentages within a
one mile radius significantly different
from the poverty percentages for the
county for all cement plants. The results
are as follows:
^" 18 percent of the universe of all
cement plants (20 plants) and 36
percent of the sample of incinerators (21
plants) have poverty percentages at a
one mile radius which exceed the
corresponding county poverty
percentages by more than five
percentage points.
^" 22 percent of the universe of all
cement plants (25 plants) and 37
percent of the sample of incinerators (21
plants) have poverty percentages at a
one mile radius which fall below the
corresponding county poverty
percentages by more than five
percentage points.
^- 60 percent of the universe of all
cement plants (68 plants) and 28
percent of sample of incinerators (21
plants) have poverty percentages at a
one mile radius which fall within five
percentage points (above or below) of
the corresponding county poverty
percentages.
B. Unfunded Federal Mandates
The Agency also evaluated the
proposed MACT standards for
compliance with the Unfunded
Mandates Reform Act (UMRA) of 1995.
Today's rule contains no Federal
mandates under the regulatory
provisions of Title II of the UMBRA for
State, local or tribal governments or the
private sector. The Agency concluded
that the rule implements requirement
specifically set forth by Congress, as
stated in the Clean Air Act and the
Resource Conservation Recovery Act. In
addition, promulgation of these MACT
standards is not expected to result in
mandated costs of $100 million or more
to any state, local, or tribal governments,
in any one year. Finally, the MACT
standards will not significantly or
uniquely affect small governments.
C. Regulatory Takings
EPA found no indication that the
MACT standards would be considered a
"taking," as defined by legislation
currently being considered by Congress.
Property would not be physically
invaded or taken for public use without
the consent of the owner. Also, the
MACT standards will not deprive
property owners of economically
beneficial or productive use of their
property, or reduce the property's value.
D. Incentives for Waste Minimization
and Pollution Prevention
The RIA results do not incorporate
waste minimization at this time.
However, the Agency did analyze the
potential for waste minimization and
the preliminary results suggest that
generators have a number of options for
reducing or eliminating waste at a much
lower cost. To evaluate whether
facilities would adopt applicable waste
minimization measures, a simplified
pay back analysis was used. Using
information on per-facility capital costs
for each technology, EPA estimated the
period of time required for the cost of
the waste minimization measure to be
returned in reduced combustion
expenditures. The assessment of waste
minimization yields estimates of the
tonnage of combusted waste that might
be eliminated. Comprehensive data to
evaluate waste minimization were not
available. Improved information on the
capital investment and operating costs
associated with waste minimization are
needed.
Overall, EPA was able to estimate that
630,000 tons of waste, a significant
portion of all combusted waste, may be
amenable to waste minimization. Three
waste generating processes account for
the reduction. These processes include
solvent and product recovery, product
processing waste, and process waste
removal and cleaning. EPA is
continuing analysis of waste
minimization options and requests
comments and information in this area.
For a complete description of the
analysis, see the regulatory impact
assessment.
E. Evaluation of Impacts on Certain
Generators
EPA is aware of the potential impact
today's proposal may have on small
business hazardous waste generators.
The emission standards proposed today
will require many combustion facilities
to install new emission control
equipment, undertake expanded
monitoring, and comply with additional
recordkeeping and reporting
requirements. Combustion facilities will
incur higher capital and operating costs
as a result of today's rule. Some
facilities are predicted to leave the
waste management business altogether.
As capacity decreases and costs
increase, facilities are likely to increase
the waste management prices they
charge generators.
EPA believes many larger generators
will respond to waste management cost
increases by accelerating their waste
-------
17480
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
minimization efforts. By undertaking
cost-effective waste minimization
initiatives, companies can reduce the
amount of waste requiring combustion,
thereby deflecting some of the impacts
of increases in waste management costs.
The same waste minimization options
may not be so readily available to
smaller businesses. Small businesses
often do not have the financial resources
to make the capital or process
improvements necessary to minimize
hazardous waste generation, even if
such improvements will have a net cost
benefit in the long run. In addition,
small businesses often lack the technical
expertise necessary for effective waste
minimization.
Those small businesses that are
unable to minimize waste generation
will either incur higher costs to operate
their businesses or, if allowed under
federal and state regulations, manage
their hazardous wastes using
unregulated disposal options. Many
small businesses, because they generate
less than 100 kg per month or less than
10 kg of acutely hazardous waste per
month, are classified as conditionally
exempt small quantity generators
(CESQGs). CESQGs are exempt from
many of the generator requirements
under 40 CFR 262 and are not required
under the federal RCRA regulations to
manage their wastes in TSDFs. Many
CESQGs, however, send their wastes to
third-party collection companies who
mix CESQG waste with waste from
larger generators and manage it as a
fully regulated hazardous waste.
Increases in waste management costs
due to today's proposal could encourage
some number of third-party collection
companies to segregate CESQG wastes
and manage them using less expensive,
yet legal, alternatives, such as
unpermitted boilers, space heaters, and
non-TSDF cement kilns.
EPA plans to revise the Regulatory
Impact Assessment (RIA) issued with
today's rule to include additional
analysis, as appropriate and feasible,
focusing on these issues. EPA is seeking
comments on any of the issues raised
here.
V. Regulatory Flexibility Analysis
The Regulatory Flexibility Act (RFA)
of 1980 requires Federal agencies to
consider impact on "small entities"
throughout the regulatory process.
Section 603 of the RFA calls for an
initial screening analysis to be
preformed to determine whether small
entities will be adversely affected by the
regulation. If affected small entities are
identified, regulatory alternatives must
be considered to mitigate the potential
impacts. Small entities as described in
the Act are only those "businesses,
organizations and governmental
jurisdictions subject to regulation."
EPA used information from Dunn &
Bradstreet, the American Business
Directory and other sources to identify
small businesses. Based on the number
of employees and annual sales
information, EPA identified 11 firms
which may be small entities. The
proposed rule is unlikely to adversely
affect many small businesses for two
important reasons. First, few
combustion units are owned by
businesses that meet the SBA definition
as a small business. Furthermore, over
one-third of those that are considered
small have a relatively small number of
employees, but have an annual sales in
excess of $50 million per year.
Second, small entities most impacted
by the rule are those that burn very little
waste and hence face very high cost per
ton burned-. Those that burn very little
waste in their existing units will
discontinue burning hazardous waste
rather than comply with the proposed
rule and dispose of waste off-site. EPA
looked at the costs of alternative
disposal and concludes the costs of
discontinuing burning wastes will not
be so high as to result in a significant
impact. Therefore, EPA believes that
today's proposed rule will have a minor
impact on small businesses.
VI. Paperwork Reduction Act
The information collection
requirements in this proposed rule have
been submitted for approval to the
Office of Management and Budget
(OMBJ under the Paperwork Reduction
Act, 44 U.S.C. 3501 et seq. Two
Information Collection Request (ICR)
documents have been prepared by EPA.
One ICR document covers the reporting
and recordkeeping requirements for
NESHAPs from hazardous waste
combustors and the other ICR document
covers the new and amended reporting
and recordkeeping requirements for
boilers and industrial furnaces burning
hazardous waste. Copies may be
obtained from Sandy Farmer, OPPE
Regulatory Information Division; U.S.
Environmental Protection Agency
(2136); 401 M St., SW; Washington, DC
20460 or by calling (202) 260-2740.
The annual public reporting and
recordkeeping burden for the NESHAP
collection of information is estimated to
average 36 hours per response. The
annual public reporting and
recordkeeping burden for the BIF
collection of information is estimated to
average 2 hours per response. These
estimates include the time needed to
review instructions; develop, acquire,
install, and utilize technology and
systems for the purposes of collecting,
validating, and verifying information,
processing and maintaining
information, and disclosing and
providing information; adjust the
existing ways to comply with any
previously applicable instructions and
requirements; train personnel to
respond to a collection of information;
search existing data sources; complete
and review the collection of
information; and transmit or otherwise
disclose the information.
An Agency may not conduct or
sponsor, and a person is not required to
respond to a collection of information
unless it displays a currently valid OMB
control number. The OMB control
numbers for EPA's regulations are
displayed in 40 CFR Part 9.
Send comments regarding the burden
estimate or any other aspect of this
collection of information, including
suggestions for reducing this burden to
Chief, OPPE Regulatory Information
Division; U.S. Environmental Protection
Agency (2136); 401 M St., SW;
Washington, DC 20460; and to the
Office of Information and Regulatory
Affairs, Office of Management and
Budget, Washington, DC 20503, marked
"Attention: Desk Officer for EPA."
Include the ICR number in any
correspondence. The final rule will
respond to any OMB or public
comments on the information collection
requirements contained in this proposal.
VII. Request for Data
EPA requests the following data to
help refine the RIA:
(1) Waste Quantity Burned: data on
hazardous and non-hazardous waste
burned at on-site facilities (by
combustion unit) broken down by
quantity of liquids, sludges, and solids.
(2) Price Data: Aggregate prices by
waste type and how they vary by
geographic region and waste
contamination level.
(3) Combustion Alternatives:
—Information on likelihood of on-site
incinerators shipping waste to on-site
boilers as an alternative.
—Realistic waste minimization
practices. Information on how
combustion and waste minimization
prices become attractive.
—Information on the type of
commercial incinerator most likely to
receive waste from on-site facilities to
ship waste off-site.
(4) Capacity: practical capacity levels
for each combustion unit.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17481
Appendix—Comparable Fuel Constituent and Physical Specifications
Note: All numbers ia the tables of this appendix are expressed to two significant figures.
TABLE 1 .—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE GASOLINE SPECIFICATION
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion limit (mg/kg)
Total Nitrogen as N
Total Halogens as Cl
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Thallium
a-Naphthylamine
a,a-Dlmethylphenethylamine
fi-Naphthylamine
1,1-Dichloroethylene
1,1,2-TrichIoroethane
1,1,2,2-Tetrachloroethane
1,2-Dibromo-3-chloropropane
1,2-Dtchloroethylene (cis-or trans-) ..
1,2,3-Trichloropropane
1,2,4-Trichlorobenzene
1,2,4,5-Tetrachlorobenzene
1,3,5-Trin'rtrobenzene
1,4-DtohIoro-2-butene (cis- or trans-).
1,4-Naphthoquinone
2-Acetylaminofluorene
2-Chloroethy) vinyl ether
2-Chloronaphthalene
2-Chlorophenol
2-Piocoline
2,3,4,6-Tetrachlorophenol
2,4-Dtehlorophenol
2,4-Dimethylpheno!
2,4-Dinitrophenol
2,4-Dinitrotoluene
2,4,5-Trichlorophenol
2,4,6-Trichlorophenol
2,6-Dichlorophenol
2,6-Dinitrotoluene
3-3'-Dimethylbenzidine
3-Methylcholanthrene
3,3'-Dichlorobenzidine
4-Aminobiphenyl
4-Bromophenyl phenyl ether
4,6-Dinitro-o-cresol
5-Nitro-o-toluidine
7,12-Dimethylbenz[a]anthracene
Acetonitrile
Acetophenone
Acrolein
Acrylonitrile
Ally! chloride
Aniline
Aramite
Benzene
Benzidine
Benzo [a] anthracene
Benzo [a] pyrene
Benzo [b] fluoranthene :.
Benzo [k] fluoranthene
Bromoform
Butyl benzyl phthalate
9.2
25
3500
(1)
340
340
7.0
0.14
14
0.70
0.70
1.4
2.8
7.0,
0.70
0.10
' 2.8
0.14
1.4
14
670
670
670
34
34 ,
34
34
34
34
670
670
670
34
670
670
34
670
670
670
670
670
670
670
670
670
670
670
670
670
670
670
670
670
670
670
670
34
670
34
34
34
670
670
670
670
670
34
670
-------
17482
Federal. Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
TABLE 1 .—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE GASOLINE SPECIFICATION—Continued
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion limit (mg/kg)
Carbon disulfide
Carbon tetrachloride
Chlorobenzene ,
Chlorobenzilate .....:
Chloroform
Chloroprene
Chrysene
cis-1,3-Dichloropropene
Cresol (o-, m-, or p-)
Di-n-butyl phthalate
Di-n-octyl phthalate
Diallate
Dibenzo [a,h] anthracene
Dibenz [a,j] acridine
Dichlorodifluoromethane
Diethyl phthalate
Dimethoate
Dimethyl phthalate
Dinoseb
Diphenylamine
Disulfoton
Ethyl methacrylate
Ethyl methanesulfonate
Famphur
Fluoranthene
Fluorene
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclopentadiene
Hexachloroethane
Hexachlorophene
Hexachloropropene
lndeno(1,2,3-cd) pyrene
Isobutyl alcohol
Isodrin
Isosafrole
Keppne
m-Dichlorobenzene
Methacrylonitrile
Methapyrilene
Methyl bromide .'..
Methyl chloride
Methyl ethyl ketone
Methyl iodide
Methyl methacrylate
Methyl methanesulfonate
Methyl parathion
Methylene chloride
N-Nitrosodi-n-butylamine
N-Nitrosodiethylamine
N-Nitrosomethylethylamine
N-Nitrosomorpholine
N-Nitrosopiperidine
N-Nitrosopyrrolidine
Naphthalene
Nitrobenzene r
o-Dichlorobenzene
o-ToIuidine
O,O-Diethyl O-pyrazinyl phospho- thioate
O,O,O-Triethyl phosphorothionate
p-(Dimethylamino) azobenzene
p-Chloro-m-cresol
p-Chloroaniline
p-Dichlorobenzene
p-Nitroaniline ;.,
p-Nitrophenol
p-Phenylenediamine
Parathion
Pentachlorobenzene
Pentachloroethane
34
34
34
670
34
34
340
0)
340
(1)
340
34
670
670
670
670
34
670
670
670
670
670
670
34
670
670
670
670
670
670
670
670
17000
670
670
34
670
670
1300
670
34
670
34
34
34
34
34
670
670
34
670
670
670
670
670
670
2800
670
670
670
670
670
670
670
670
670
670
670
670
670
670
34
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17483
TABLE 1 .—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE GASOLINE SPECIFICATION—Continued
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion limit (mg/kg)
Pentachloronitrobenzene ,
Pentachlorophenol
Phenacetin
Phenol
Phorate
Pronamide
Pyrkiine
Safrole
Tetrachloroethylene
Tetraethyldithiopyrophosphate
Toluene
Trichloroethylene
Trichlorofluoromethane
Vinyl Chloride
670
670
670
670
670
670
670
670
34
670
35000
34
34
34
1 Non-detect.
TABLE 2.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 2 FUEL OIL SPECIFICATION
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion limits
(mg/kg)
Total Nitrogen as N
Total Halogens as Cl
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Thallium
a-Naphthylamine
ct,a-Dimethylphenethylamine
p-Naphthylamine
1,1-Dichloroethylene
1,1,2-Trichloroethane
1,1,2,2-Tetrachloroethane
1,2-Dibromo-3-chloropropane
1,2-Dichloroethylene (cis-or trans-) ..
1,2,3-Trichloropropane
1,2,4-Trichlorobenzene
1,2,4,5-Tetrachlorobenzene
1,3,5-Trinitrobenzene ,
1,4-Dichloro-2-butene (cis- or trans-),
1,4-Naphthoquinone
2-Acetylamlnofluorene
2-Chloroethyl vinyl ether
2-Chloronaphthalene
2-Chlorophenol
2-Piccoline
2,3,4,6-TetrachIorophenol
2,4-Dichlorophenol
2,4-Dimethylphenol
2,4-Dinitrophenol
2,4-Dinitrotoluene
2,4,5-Trichlorophenol
2,4,6-Trichlorophenol
2,6-Dichlorophenol
2,6-Dinitrotoluene
3-3'-Dimethylbenzidine
3-Methytcholanthrene
3,3'-Dichlorobenzidine
110
25
6.0
0.12
12
0.60
0.60
1.2
2.4
6.6
0.070
0.60
0.11
2.4
1.2
12
1200
1200
1200
34
34
34
34
34
34
1200
1200
1200
34
1200
1200
34
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
-------
27484
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
TABLE 2.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 2 FUEL OIL SPECIFICATION—Continued
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maxjmum detec-
tion limits
(mg/kg)
4-Aminobiphenyl
4-Bromophenyl phenyl ether
4,6-Dinitro-o-cresol
5-Nitro-o-toluidine
7,12-Dimethylbenz[a]anthracene
Acetonitrile
Acetophenone
Acrolein
Acrylonitrile
AHyl chloride
Aniline
Aramite
Benzene
Benzidine
Benzo[a]anthracene
Benzofajpyrene
Benzo[b]fluoranthene
Benzo[k]fluoranthene
Bromoform
Butyl benzyl phthalate
Carbon disulfide
Carbon tetrachloride
Chlorobenzene
Chlorobenzilate
Chloroform
Chloroprene
Chrysene
cis-1,3-Dichloropropene
Cresol (o-, n-, or p-)
Di-n-butyl phthalate
Di-n-octyl phthalate
Diallate
Dibenzo[a,h]anthracene
Dibenz[a,j]acridine
Dichlorodifluoromethane
Diethyl phthalate
Dimethoate
Dimethyl phthalate
Dinoseb
Diphenylamine
Disulfoton
Ethyl methacrylate
Ethyl methanesulfonate
Famphur
Fluoranthene
Fluorene
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclopentadiene
Hexachloroethane
Hexachlorophene
Hexachloropropene
lndeno(1,2,3-cd)pyrene
Isobutyl alcohol
Isodrin
Isosafrole
Kepone
m-Dichlorobenzene
Methacrylonitrile
Methapyrilene
Methyl bromide
Methyl chloride
Methyl ethyl ketone
Methyl iodide
Methyl methacrylate
Methyl methanesulfonate
Methyl parathion
Methylene chloride
N-Nitrosodi-n-butylamine
N-Nitrosomorpholine
21
(1)
610
610
610
(1)
(1)
(1)
610
(1)
610
O
0
(1)
0)
1200
1200
1200
1200
1200
34
1200
34
34
34
1200
1200
1200
1200
1200
34
1200
34
34
34
1200
34
34
34
1200
1200
1200
D
1200
34
1200
1200
1200
1200
1200
1200
34
1200
1200
1200
1200
1200
1200
1200
1200
29000
1200
1200
34
1200
1200
2300
1200
34
1200
34
34
34
34
34
1200
1200
34
1200
1200
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17485
TABLE 2.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 2 FUEL OIL SPECIFICATION—Continued
Chemical name
Vinyl Chloride
Concentration
limit (md/kg at
10,000 BTU/lb)
(1)
(1)
(1)
(1)
1200
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
150
(1)
(1)
(1)
Maximum detec-
tion limits
(mg/kg)
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
34
1200
1200
1200
1200
1200
1200
1200
1200
34
1200
34
34
34
1 Non-detect.
TABLE 3.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 4 FUEL OIL SPECIFICATION
Chemical name
Hnhalf
i oari
MirL-al
Thallium • •
1.2.4.5-Tetrachlorobenzene
Concentration
limit (mg/kg at
10,000 BTU/lb)
1500
10
(1)
(1)
(1)
(1)
(1)
(1)
(1)
9.9
(1)
(1)
16
0.13
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
D
(1)
0)
(1)
Maximum detec-
tion limits
(mg/kg)
11
0.23
23
1.1
1.1
2.3
4.6
1.1
0.18
2.3
23
200
200
200
17
17
17
17
17
17
200
200
-------
17486
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
TABLE 3.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 4 FUEL OIL SPECIFICATION—Continued
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion limits
(mg/kg)
1,3,5-Trinitrobenzene
1,4-Dichlorc-2-butene (cis- or trans-)
1,4-Naphthpquinone
2-Acetylaminofluorene
2-Chloroethyl vinyl ether
2-Chloronaphthalene
2-Chlorophenol
2-Piccoline
2,3,4,6-Tetrachlorophenol
2,4-Dichlorophenol :.
2,4-Dimethylphenol
2,4-Dinitropheno! ,
2,4-Dinitrotoluene
2,4,5-Trichlorophenol
2,4,6-Trichlorophenol
2,6-Dichlorophenol
2,6-Dinitrotoluene
3-3'-Dimethylbenzidine
3-Methylcholanthrene
3,3'-Dichlorobenzidine
4-AminobiphenyI
4-Bromophenyl phenyl ether
4,6-Dinitro-o-cresol
5-Nitro-o-toluidine
7,12-Dimethylbenz[a]anthracene
Acetonitrile
Acetophenone
Acrolein
Acrylonitrile
Allyl chloride
Aniline
Aramite
Benzene
Benzidine
Benzo[a]anthracene
Benzofajpyrene
Benzo[b]fluoranthene
Benzo[k]fluoranthene
Bromoform
Butyl benzyl phthalate
Carbon disulfide
Carbon tetrachloride
Chlorobenzene
Chlorobenzilate
Chloroform
Chloroprene
Chrysene
cis-1,3-DichIoropropene
Cresol (o-, m-, or p-)
Di-n-butyl phthalate
Di-n-octyl phthalate
Diallate
Dibenzo[a,h]anthracene
Dibenz[a,j]acridine
Dichlorodifluoromethane
Diethyl phthalate
Dimethoate
Dimethyl phthalate
Dinoseb ,
Diphenylamine
Disulfoton
Ethyl methacrylate
Ethyl methanesulfonate
Famphur
Fluoranthene
Fluorene
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclopentadiene
Hexachloroethane
22
(1)
100
100
100
(1)
(1)
(1)
100
(1)
100
110
200
17
200
200
17
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
17
200
17
17
17
200
200
200
200
200
17
200
17
17
17
200
17
17
17
200
200
200
200
17
200
200
200
200
200
200
17
200
200
200
200
200
200
200
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17487
TABLE 3.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 4 FUEL OIL SPECIFICATION—Continued
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion limits
(mg/kg)
Hexachlorophene
Hexachloropropene
lndeno{1,2,3-cd)pyrene
Isobutyl alcohol
tsodrin
Isosafrole
Kepone
m-Dtehlorobenzene
Methacrylonitrile
Methapyrilene
Methyl bromide
Methyl chloride
Methyl ethyl ketone
Methyl iodide
Methyl methacrylate
Methyl methanesulfonate
Methyl parathion
Methylene chloride
N-Nitrosodi-n-butylamine
N-Nitrosomethylethylamine
N-Nitrosomorpholine
N-Nitrosopiperidine
N-Nitrosopyrrolidine
N-Nitrosodiethylamine
Naphthalene
Nitrobenzene
o-Dichlorobenzene
o-TolukJine
O.O Diethyl O-pyrazinyl phosphoro- thioate
O.O.O-Triethyl phosphorothionate
p-(Dimethylamino)azobenzene
p-Chloro-m-cresol
p-Chloroaniline
p-Dichlorobenzene
p-Nitroaniline
p-Nitrophenol
p-Phenylenediamine
Parathion
Pentachlorobenzene
Pentachloroethane :.
Pentachloronitrobenzene
Pentachlorophenol
Phenacetin
Phenol
Phorale
Pronamide
Pyridine
Safrole
Tetrachloroethylene
Tetraethyldithiopyrophosphate
Toluene
Trichloroethylene
Trichlorofluoromethane
Vinyl Chloride
1 Non-detect.
C)
340
C)
110
5000
200
200
17
200
200
400
200
17
200
17
17
17
17
17
200
200
17
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
17
200
200
200
200
200
200
200
200
17
200
17
17
17
TABLE 4.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 6 FUEL OIL SPECIFICATION
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion level (mg/kg)
Total Nitrogen as N ...
Total Halogens as Cl
Antimony
Arsenic
Barium
Beryllium
Cadmium
3500
10
6.5
0.20
20
1.0
1.0
-------
17488 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
TABLE 4.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 6 FUEL OIL SPECIFICATION—Continued
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
:ion level (mg/kg)
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Thallium
a-Naphthylamine
a,a-Dimethylphenethylamine
p-Naphthylamine
1,1-Dichloroethylene
1,1,2-Trichloroethane
1,1,2,2-Tetrachloroethane
1,2-Dibromo-3-chloropropane
1,2-Dichloroethylene (cis- or trans-) .,
1,2,3-Trichloropropane
1,2,4-Trichlorobenzene
1,2,4,5-Tetrachlorobenzene
1,3,5-Trinitrobenzene
1,4-Dichloro-2-butene (cis- or trans-)
1,4-Naphthoquinone
2-Acetylaminofluorene
2-Chloroethyl vinyl ether
2-ChloronaphthaIene
2-Chlorophenol
2-Piccoline
2,3,4,6-Tetrachlorophenol
2,4-Dichlorophenol
2,4-Dimethylphenol
2,4-Dinitrophenol
2,4-Dinitrotoluene
2,4,5-TrichIorophenol
2,4,6-Trichlorophenol
2,6-DichIorophenol
2,6-Dinitrotoluene
3-3'-Dimethylbenzidine
3-Methylcholanthrene
3,3'-Dichlorobenzidine
4-Aminobiphenyl
4-Bromophenyl phenyl ether
4,6-Dinitro-o-cresol
5-Nitro-o-toluidine
7,12-Dimethylbenz[a]anthracene
Acetonitrile
Acetophenone
Acrolein
Acrylonitrile
Allyl chloride
Aniline
Aramite
Benzene
Benzidine
Benzo[a]anthracene
Benzo[a]pyrene
Benzo[b]fluoranthene
Benzo[k]fluoranthene
Bromoform
Butyl benzyl phthalate
Carbon disulfide
Carbon tetrachloride
Chlorobenzene
Chlorobenzilate
Chloroform
Chloroprene
Chrysene
cis-1,3-Dichloropropene
Cresol (o-, m-, p-)
Di-n-butylphthalate
30
36
0.12
D
0)
11
(1)
930
530
420
1300
(1)
n
2.0
4.1
1.0
0.22
2.0
20
640
640
640
20
20
20
20
20
20
640
640
640
20
640
640
20
640
640
640
640
640
640
640
640
640
640
640
640
640
640
640
640
640
640
640
640
20
640
20
20
20
640
640
640
640
20
640
20
20
20
640
20
20
20
640
640
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17489
TABLE 4.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 6 FUEL OIL SPECIFICATION—Continued
Chemical name
Di-n-octyl phthalate
Diallate
Dibenzo[a,h]anttiracene
Dibenz[a,j]acridine
Dichlorodifluoromethane :
Diethyl phthalate
Dimethoate
Dimethyl phthalate .-
Dinoseb , . .
Diphenylamine
Disulfoton ... .
Ethyl methacrylate
Ethyl methanesulfonate
Famphur
Fluoranthene
Fluorene
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclopentadiene
Hexachloroethane
Hexachlorophene
Hexachloropropene :
lndeno(1 ,2,3-cd)pyrene
Isobutyl alcohol
Isodrin
Isosafrole
Kepone
m-Dlch!orobenzene
Methacrylonitrile
Methapyrilene
Methyl bromide
Methyl chloride
Methyl ethyl ketone
Methyl iodide
Methyl methacrylate
Methyl methanesulfonate
Methyl parathion
Methylene chloride
N-Nitrosodi-n-butylamlne
N-Nitrosomethylethylamine
N-Nitrosomorphol!ne
N-Nitrosopiperidine
N-Nitrosopyrrolidine
N-Nitrosodiethylamine
Naphthalene
Nitrobenzene
o-Dichlorobenzene
o-TolukJine
O,O Diethyl O-pyrazinyl phosphothioate
O,O,O-Triethyl phosphorothionate
p-(Dimethylamino)azobenzene
p-Chloro-m-cresol
p-Chloroaniline
p-Dichlorobenzene
p-Nitroaniline
p-Nitrophenol
p-Phenylenediamine
Parathion
Pentachlorobenzene
Pentachloroethane
Pentachloronitrobenzene
Pentachlorophenol
Phenacetin
Phenol
Phorate
Pronamide
PyrWine
Safrole
Tetrachloroethylene
Tetraethyldithiopyropnosphate
Concentration
limit (mo/kg at
10,000 BTU/lb)
350
(1)
350
(1)
n\
n\
n\
<1)
m
o)
n)
n
n)
m
n)
350
(1)
(1)
(1)
H)
(1)
(1)
350
(1)
(1)
(1)
(1)
C1)
1
(1)
n
<1)
P)
(1)
p)
(1)
p)
(1)
p)
(1)
P)
(1)
py
p)
570
P)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
n\
n\
n)
D
n)
/n
n
n
D
h\
m
D
n\
n\
(1)
D
Maximum detec-
tion level (mg/kg)
640
640
20
640
640
640
640
640
640
20
640
640
640
640
640
640
640
16000
640
20
640
640
1300
640
20
640
20
20
20
20
20
640
640
20
640
640
640
640
640
640
640
640
1300
640
640
640
640
640
640
640
640
640
640
640
20
640
640
640
640
640
640
640
640
20
640
-------
17490
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
TABLE 4.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE NUMBER 6 FUEL OIL SPECIFICATION—Continued
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion level (mg/kg)
Toluene
Trichloroethylene
Trichlorofluoromethane
Vinyl Chloride
41
20
20
20
1 Nqn-detect.
TABLE 5.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE COMPOSITE FUEL SPECIFICATION—50TH
PERCENTILE ANALYSIS
Chemical name
Concentration
limit (mg/kg at
10,000 BTU/lb)
Maximum detec-
tion limits
(mg/kg)
Total Nitrogen as N
Total Halogens as Cl
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Thallium
ec-Naphthylamine
a.cc-Dimethylphenethylamine ,
p-Naphthylamine ,
1,1-Dichloroethylene ,
1,1,2-TrichIoroethane ,
1,1,2,2-Tetrachloroethane ,
1,2-Dibromo-3-chloropropane
1,2-Dichloroethylene (cis- or trans-) .
1,2,3-Trichloropropane
1,2,4-Trichlorobenzene
1,2,4,5-Tetrachlorobenzene
1,3,5-Trinitrobenzene
1,4-Dichloro-2-butene (cis- or trans-)
1,4-Naphthoquinone
2-Acetylaminofluorene
2-Chloroethyl vinyl ether
2-Chloronaphthalene
2-Chlorophenol
2-Piccoline-
2,3,4,6-Tetrachlorophenol
2,4-Dichlorophenol
2,4-Dimethylphenol
2,4-Dinitrophenol
2,4-Dinitrotoluene
2,4,5-Trichlorophenol
2,4,6-Trichlorophenol
2,6-Dichlorophenol
2,6-Dinitrotoluene
3-3'-DimethyIbenzidine
3-MethyIcholanthrene
3,3'-Dichlorobenzidine
4-Aminobiphenyl
4-Bromophenyl phenyl ether
4,6-Dinitro-o-cresol .'
5-Nitro-o-toluidine
7,12-Dimethylbenz[a]anthracene
Acetonitrile
Acetophenone
Acrolein
Acrylonitrile
170
10
4.7
7.0
2.4
0.090
O
0.14
18
0.90
0.90
1.8
3.6
0.90
0.11
1.8
18
220
220
220
17
17
17
17
17
17
220
220
220
17
220
220
17
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
220
17
220
17
17
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17491
TABLE 5.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE COMPOSITE FUEL SPECIFICATION—SOTH
PERCENTILE ANALYSIS—Continued
Chemical name
Allyl chloride
Aniline
Aramite
Benzene „ .
Benzidine
Benzo[a]anthracene
Benzofajpyrsne
Bonzojbjfluoranthene
Benzo[k]fluoranthene
Bromoform
Butyl benzyl phthalate
Carbon disulfide
Carbon tetrachloride . . .
Chlorobenzene •.
Chlorobenzilate
Chloroform
Chloroprene
Chrysene
cis-1 ,3-DichIoropropene
Cresol (o-, n-, or p-)
Di-n-butyl phlhalate
Di-n-octyl phthalate
Diallate
Dibenzofa h]anthracene
Dibonz[a,j]acridine
Dichlorodifluoromethane
Diethyi phthalate
Dimethoate
Dimethyl phthalate
Dinoscb : ....
Diphenylamine
Disulfoton
Ethyl methacrylate
Ethyl methanesulfonate
Famphur
Fluoranthene
Fluorene .
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclopentadiene
Hexachloroethane
Hexachlorophene
Hexachloropropene
lndeno(1,2,3-cd)pyrene
Isobutyl alcohol
Isodrin
Isosafrole ;
Kepone
m-Dichlorobenzene
Methacrylonitrile
Methapyrilene
Methyl bromide
Methyl chloride
Methyl ethyl ketone ,
Methyl iodide
Methyl methacrylate
Methyl methanesulfonate
Methyl parathion
Methylene chloride
N-Nitrosodi-n-butylamine
N-Nitrosomethylethylamine
N-Nitrosomorpholine . . ..
N-Nitrosopiperidine
N-Nitrosopyrrolidine
N-Nitrosodiethylamine
Naphthalene
Nitrobenzene
o-Dichlorobenzene
o-Toluidine
Concentration
limit (mg/kg at
10,000 BTU/lb)
(1)
(i)
(i)
21
P)
140
140
140
(1)
(1)
(1)
(1)
P)
(i)
(1)
(1)
(i)
140
P)
(1)
P)
120
(1)
140
P)
(1)
(1)
P)
(1)
P)
P)
P)
(1)
(1)
(1)
P)
120
(1)
(1)
(1)
(1)
(1)
(1)
140
(1)
(1)
(')
P)
(n
P)
(1)
P)
(1)
(1)
(1)
(1)
P)
(1)
(1)
P)
P)
P)
P)
P)
P)
360
P)
P)
P)
Maximum detec-
tion limits
(mg/kg)
17
220
220
220
220
17
220
17
17
17
220
17
17
17
220
220
220
220
17
220
220
220
220
220
220
17
220
220
220
220
220
220
220
5500
220
17
220
220
440
220
17
220
17
17
17
17
17
220
220
17
220
220
220
220
220
220
220
220
270
-------
17492
Federal Register 7 Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
TABLE 5.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE COMPOSITE FUEL SPECIFICATION—50TH
PERCENTILE ANALYSIS—Continued
Chemical name
p-Chloroaniline ••••
Pyridine • — • *
Trichlorofluoromethane •
Vinyl Chloride •
Concentration
limit (mg/kg at
10,000 BTU/lb)
(1)
(i)
(i)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(i)
(1)
(1)
(1)
D
(1)
(1)
(1)
(1)
(1)
110
(1)
(1)
(1)
Maximum detec-
tion limits
(mg/kg)
220
220
220
220
220
220
220
220
220
220
220
17
220
220
220
220
220
220
220
220
17
220
17
17
17
1 Non-detect.
TABLE 6.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE COMPOSITE FUEL SPECIFICATION—90TH
PERCENTILE ANALYSIS
Chemical name
Beryllium
Cobalt
Manganese • •
Mercury — • •
Nickel
Silver
Thallium
1 1 -Dichloroethylene
•j 2-Dibromo-3-chioropropane
•j 2-Dichloroethylene (cis- or trans-)
•) 2 3-Trichloropropane -—
•\ 2 4-Trichlorobenzene
•\ 3 5-Trinitrobenzene • •
1 4-Naphthoouinone
2-Chloroethyl vinyl ether •••• • •
2-Chloronaohthalene
Concentration
limit (mg/kg at
10,000 BTU/lb)
1800
25
5.8
(1)
(1)
(1)
(1)
(1)
(1)
22
(1)
(1)
18
0.12
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
Maximum detec-
tion limit (mg/kg)
0.22
22
1.1
1.1
2.2
4.4
1.1
0.18
2.2
22
700
700
700
34
34
34
34
34
34
700
700
900
34
700
700
34
700
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17493
TABLE 6.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE COMPOSITE FUEL SPECIFICATION—90TH
PERCENTILE ANALYSIS—Continued
Chemical name „
2-Chlorophenol
2-PtecolIne
2,3,4,6-Tetrachlorophenol
2,4-DichlorophenoI
2,4-Dimethylphenol
2,4-Dinitrophenol
2,4-D!nitrotoluene
2,4,5-Trichlorophenol
2,4,6-Trichlorophenol
2,6-Dtehlorophenol
2,6-Dlnitrotoluene ;
3-3'-DimethyIbenzidine
3-Methyteholanthrene
3,3'-DIchlorobenzidine
4-Arninobiphenyl 1
4-Bromophenyl phenyl ether
4,6-DInilro-o-cresol
5-Nitro-o-toluidine
7,12-Dimelhylbenz[a]anthracene ,
Acetonitrile
Acetophenone
Acrolein
Acrylonitrile
Ally! chloride
Aniline
Aramite
Benzene
Benzidine
Benzo[a]anthracene
Benzo{a]pyrene
Benzo[b]fluoranthene
Benzo[k]fluoranthene
Bromoform
Butyl benzyl phthalate
Carbon disulfide
Carbon tetrachloride
Chlorobenzene
Chlorobenzilate
Chloroform
Chloroprene
Chrysene
cls-1 ,3-Dichloropropene
Cresol (o-, n-, or p-) ,
Di-n-butyl phthalate
Di-n-octyl phthalate
Diallate
Dibenzo[a,h]anthracene
Dibenz[a,j]acridine
Dichlorodifluoromethane
Diethyl phthalate
Dimethoate '.
Dimethyl phthalate
Dinoseb
Diphenylamine
Disulfoton
Ethyl methacrylate
Ethyl methanesulfonate
Famphur
Fluoranthene
Fluorene
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclopentadiene .
Hexachloroethane
Hexachlorophene
Hexachloropropene ...
lndeno(1,2,3-cd)pyrene
fsobutyl alcohol
Isodrin »...'. ;.
Concentration
limit (mg/kg at
10,000 BTU/lb)
n)
(1)
m
n\
n\
n\
m
n\
m
n\
(1)
n)
n)
n\
n\
(1)
(1)
m
m
n\
n)
n)
n)
n\
n\
n\
3300
m
610
530
390
m
n)
n\
h\
n\
n\
n\
(1)
n\
610
(1)
n)
m
360
n\
360
n\
n\
n\
it)
ii\
n\
it)
(t\
n\
m
n)
n)
360
n)
ii)
n)
n\
ii)
ii)
360
/n
n
Maximum detec-
tion limit (mg/kg)
700
700
700
700
700
700
700
700
700
700
700
700
700
700
700
700
700
700
700
34
700
34
34
34
700
700
700
700
34
700
34
34
34
700
34
34
34
700
700
700
700
34
700
700
700
700
700
700
34
700
700
700
700
700
700
700
18000
700
34
700
-------
17494
Federal Register /Vol. 61, No. 77 /.Friday, April 19, 1996 / Proposed Rules
TABLE 6.—DETECTION AND DETECTION LIMIT VALUES FOR A POSSIBLE COMPOSITE FUEL SPECIFICATION—90TH
PERCENTILE ANALYSIS—Continued
Chemical name
Safrolo
Vinyl Chloride .- • •
Concentration
limit (mg/kg at
10,000 BTU/lb)
(1)
(i)
. ,(ij-,-,
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
1300
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
25,000
(1)
(1)
(1)
Maximum detec-
tion limit (mg/kg)
700
1400
700
34
700
34
34
34
34
34
700
700
34
700
700
700
700
700
700
700
700
1000
700
700
700
700
700
700
700
700
700
700
700
34
700
700
700
700
700
700
700
700
34
700
34
34
34
1 Non-detect.
TABLE 7.—POSSIBLE PHYSICAL SPECIFICATIONS—FROM EPA's DATA
Fuel type (physical param)
Flash Point (°C) •
Kinematic viscosity (cSt @ 40°C) '.
Gasoline
<0
No. 2
44
3.7
No. 4
66
6.4
No. 6
69
660
Comp. 50th
63
6.4
Comp 90th
<0
Note: Kinematic viscosity for gasoline is less than measureable levels.
TABLE 8.—POSSIBLE PHYSICAL SPECIFICATIONS—FROM ASTM AND OTHER PUBLISHED LITERATURE
Fuel type220 (parameter)
Flashpoint (°C) '•'•"••
Kinematic viscosity (cSt@40 °C) .'....'...:
Gasoline
221 _ 42
222 0.6
No. 2
' 38
3.4
•Me
No. 4
55
24
No. 6
60
50 (at 100°C)
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17495
221 FeWer, M.F., and R.W. Rousseau, Elementary Principles of Chemical Processes, John Wiley and Sons, New York 1978 420
222 Perry, Robert H., Don W. Green, and James O. Moloney, Perry's Chemical Engineers' Handbook: Sixth Edition, McGraw-Hill Book Co.,
New York, 1984, 9-13.
List of Subjects
40 CFR Part 60
Environmental protection
Administrative practice and procedure
Air pollution control
Aluminum
Ammonium sulfate plants
Batteries
Beverages
Carbon monoxide
Cement industry
Coal
Copper
Dry cleaners
Electric power plants
Fertilizers
Fluoride
Gasoline
Glass and glass products
Grains
Graphic arts industry
Heaters
Household appliances
Insulation
Intergovernmental relations
Iron
Labeling
Lead
Lime
Metallic and nonmetallic mineral
processing plants
Metals
Motor vehicles
Natural gas
Nitric acid plants
Nitrogen dioxide
Paper and paper products industry
Particulate matter
Paving and roofing materials
Petroleum
Phosphate
Plastics materials and synthetics
Polymers
Reporting and recordkeeping
requirements
Sewage disposal
Steel
Sulfur oxides
Sulfuric acid plants
Tires
Ure thane
Vinyl
Volatile organic compounds
Waste treatment and disposal
Zinc
40 CFR Part 63
Air pollution control
Hazardous substances
Reporting and recordkeeping
requirements
40 CFR Part 260
Administrative practice and procedure
Confidential business information
Environmental Protection Agency
Hazardous waste
40 CFR Part 261
Environmental Protection Agency
Hazardous waste
Recycling
Reporting and recordkeeping
requirements
40 CFR Part 264
Air pollution control
Environmental Protection Agency
Hazardous waste
Insurance
Packaging and containers
Reporting and recordkeeping
requirements
Security measures
Surety bonds
40 CFR Part 265
Air pollution control
Environmental Protection Agency
Hazardous waste
Insurance
Packaging and containers
Reporting and recordkeeping
requirements
Security measures
Surety bonds
Water supply
40 CFR Part 266
Energy
Environmental Protection Agency
Hazardous waste
Recycling
Reporting and recordkeeping
requirements
40 CFR Part 270
Administrative practice and procedure
Confidential business information
Environmental Protection Agency
Hazardous materials transportation
Hazardous waste
Reporting and recordkeeping
requirements
Water pollution control
Water supply
40 CFR Part 271
Administrative practice and procedure
Confidential business information
Environmental Protection Agency
Hazardous materials transportation
Hazardous waste
Indians-lands
Intergovernmental relations
Penalties
Reporting and recordkeeping
requirements
Water pollution control
Water supply
Dated: March 20,1996.
Carol M. Browner,
Administrator.
For the reasons set out in the
preamble, it is proposed to amend Title
40 of the Code of Federal Regulations as
follows:
PART 60—STANDARDS OF
PERFORMANCE FOR NEW
STATIONARY SOURCES
I. In part 60:
1. The authority citation for part 60
continues to read as follows:
Authority: 42 USC 7401, 7411, 7414, 7416,
7429, and 7601.
2. Appendix B in Part 60 is amended
by adding four entries to the table of
contents, and by adding new
performance specifications 4B, 8A, 10,
11, and 12:
Appendix B—Performance
Specifications
*****
Performance Specification 4B—
Specifications and test procedures for carbon
monoxide and oxygen continuous monitoring
systems in stationary sources.
*****
Performance Specification 8A—
Specifications and test procedures for total
hydrocarbon continuous monitoring systems
in hazardous waste-burning stationary
Performance Specification 10—
Specifications and test procedures for multi-
metals continuous monitoring sytems in
stationary sources.
Performance Specification 11—
Specifications and test procedures for
particulate matter continuous monitoring
systems in stationary sources.
Performance Specification 12—
Specifications and test procedures for total
mercury monitoring systems in stationary
Performance Specification 4B—
Specifications and test procedures for carbon
monoxide and oxygen continuous monitoring
systems in stationary sources.
1. Applicability and Principle
1.1 Applicability. This specification is to
be used for evaluating the acceptability of
carbon monoxide (CO) and oxygen (02)
continuous emission monitoring systems
(GEMS) at the time of or soon after
installation and whenever specified in the
regulations. The GEMS may include, for
certain stationary sources, (a) flow
monitoring equipment to allow measurement
-------
17496
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
of the dry volume of stack effluent sampled,
and M an automatic sampling system.
This specification is not designed to
evaluate the installed GEMS' performance
over an extended period of time nor does it
identify specific calibration techniques and
auxiliary procedures to assess the GEMS'
performance. The source owner or operator,
however, is responsible to properly calibrate,
maintain, and operate the GEMS. To evaluate
the GEMS' performance, the Administrator
may require, under Section 114 of the Act,
the operator to conduct GEMS performance
evaluations at other times besides the initial
test.
The definitions, installation and
measurement location specifications, test
procedures, data reduction procedures,
reporting requirements, and bibliography are
the same as in PS 3 (for O2) and PS 4A (for
CO) except as otherwise noted below.
1.2 Principle. Installation and
measurement location specifications,
performance specifications, test procedures,
and data reduction procedures are included
in this specification. Reference method tests,
calibration error tests, and calibration drift
tests, and interferant tests are conducted to
determine conformance of the GEMS with the
specification.
2. Definitions
2.1 Continuous Emission Monitoring
System (GEMS). This definition is the same
as PS 2 Section 2.1 with the following
addition. A continuous monitor is one in
which the sample to be analyzed passes the
measurement section of the analyzer without
interruption.
2.2 Response Time. The time interval
between the start of a step change in the
system input and the time when the
pollutant analyzer output reaches 95 percent
of the final value.
2.3 Calibration Error (CE). The difference
between the concentration indicated by the
GEMS and the known concentration
generated by a calibration source when the
entire GEMS, including the sampling
interface) is challenged. A CE test procedure
is performed to document the accuracy and
linearity of the GEMS over the entire
measurement range.
3. Installation and Measurement Location
Specifications
3.1 The GEMS Installation and
Measurement Location. This specification is
the same as PS 2 Section 3.1 with the
following additions. Both the CO and O2
monitors should be installed at the same
general location. If this is not possible, they
may be installed at different locations if the
effluent gases at both sample locations are
not stratified and there is no in-leakage of air
between sampling locations.
3.1.1 Measurement Location. Same as PS
2 Section 3.1.1.
3.1.2 Point GEMS. The measurement
point should be within or centrally located
over the centroidal area of the stack or duct
cross section.
3.1.3 Path GEMS. The effective
measurement path should be (1) have at least
70 percent of the path within the inner 50
percent of the stack or duct cross sectional
area, or (2) be centrally located over any part
of the centroidal area.
3.2 Reference Method (RM) Measurement
Location and Traverse Points. This
specification is the same as PS 2 Section 3.2
with the following additions. When pollutant
concentrations changes are due solely to
diluent leakage and CO and O2 are
simultaneously measured at the same
location, one half diameter may be used in
place of two equivalent diameters.
3.3 Stratification Test Procedure.
Stratification is defined as the difference in
excess of 10 percent between the average
concentration in the duct or stack and the
concentration at any point more than 1.0
meter from the duct or stack wall. To
dete'rmine whether effluent stratification
exists, a dual probe system should be used
to determine the average effluent
concentration while measurements at each
traverse point are being made. One probe,
located at the stack or duct centroid, is used
as a stationary reference point to indicate
change in the effluent concentration over
time. The second probe is used for sampling
at the traverse points specified in method 1,
appendix A, 40 CFR part 60. The monitoring
system samples sequentially at the reference
and traverse points throughout the testing
period for five minutes at each point.
4. Performance and Equipment
Specifications
4.1 Data Recorder Scale. For O2, same as
specified in PS 3, except that the span shall
be 25 percent. The span of the O2 may be
higher if the O2 concentration at the sampling
point can be greater than 25 percent For CO,
same as specified in PS 4A, except that the
low-range span shall be 200 ppm and the
high range span shall be 3000 ppm. In
addition, the scale for both GEMS must
record all readings within a measurement
range with a resolution of 0.5 percent.
4.2 Calibration Drift. For O2, same as
specified in PS 3. For CO, the same as
specified in PS 4A except that the GEMS
calibration must not drift from the reference
value of the calibration standard by more
than 3 percent of the span value on either the
high or low range.
4.3 Relative Accuracy (RA). For O2, same
as specified in PS 3. For CO, the same as
specified in PS 4A.
4.4 Calibration Error (CE). The mean
difference between the GEMS and reference
values at all three test points (see Table I)
must be no greater than 5 percent of span
value for CO monitors and 0.5 percent for O2
monitors.
4.5 Response Time. The response time for
the CO or O2 monitor shall not exceed 2
minutes.
5. Performance Specification Test Procedure
5.1 Calibration Error Test and Response
Time Test Periods. Conduct the CE and
response time tests during the CD test period.
6.0 The OEMS Calibration Drift and
Response Time Test Procedures
The response time test procedure is given
in PS 4A, and must be carried out for both
the CO and 02 monitors.
7. Relative Accuracy and Calibration Error
Test Procedures
7.1 Calibration Error Test Procedure.
Challenge each monitor (both low and high
range CO and O2) with zero gas and EPA
Protocol 1 cylinder gases at three
measurement points within the ranges
specified in Table I.
TABLE 1.—CALIBRATION ERROR
CONCENTRATION RANGES
Measure-
ment point
1
2
3
CO low
range
(ppm)
0-40
60-80
140-160
CO high
range
(ppm)
0-600
900-1200
2100-2400
02
(per-
cent)
0-2
8-10
14-16
Operate each monitor in its normal
sampling mode as nearly as possible. The
calibration gas shall be injected into the
sample system as close to the sampling probe
outlet as practical and should pass through
all GEMS components used during normal
sampling. Challenge the CEMS three non-
consecutive times at each measurement point
and record the responses. The duration of
each gas injection should be sufficient to
ensure that the CEMS surfaces are
conditioned.
7.1.1 Calculations. Summarize the results
on a data sheet. Average the differences
between the instrument response and the
certified cylinder gas value for each gas.
Calculate the CE results according to:
CE = |d/FS| x 100 (1)
Where d is the mean difference between the
CEMS response and the known reference
concentration and FS is the span value.
7.2 Relative Accuracy Test Procedure.
Follow the RA test procedures in PS 3 (for
O2) section 3 and PS 4A (for CO) section 4.
7.3 Alternative RA Procedure. Under
some operating conditions, it may not be
possible to obtain meaningful results using
the RA test procedure. This includes
conditions where consistent, very low CO
emission or low CO emissions interrupted
periodically by short duration, high level
spikes are observed. It may be appropriate in
these circumstances to waive the RA test and
substitute the following procedure.
Conduct a complete CEMS status check
following the manufacturer's written
instructions. The check should include
operation of the light source, signal receiver,
timing mechanism functions, data
acquisition and data reduction functions,
data recorders, mechanically operated
functions, sample filters, sample line heaters,
moisture traps, and other related functions of
the CEMS, as applicable. All parts of the
CEMS must be functioning properly before
the RA requirement can be waived. The
instrument must also successfully passed the
CE and CD specifications. Substitution of the
alternate procedure requires approval of the
Regional Administrator.
8. Bibliography
1. 40 CFR Part 266, Appendix IX, Section
2, "Performance Specifications for
Continuous Emission Monitoring Systems."
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17497
Performance Specification 8A—
Specifications and test procedures for total
hydrocarbon continuous monitoring systems
in hazardous waste-burning stationary
sources.
1. Applicability and Principle
1.1 Applicability. These performance
specifications apply to hydrocarbon (HC)
continuous emission monitoring systems
(GEMS) installed on hazardous waste-
burning stationary sources. The
specifications include procedures which are
intended to be used to evaluate the
acceptability of the CEMS at the time of its
installation or whenever specified in
regulations or permits. The procedures are
not designed to evaluate CEMS performance
over an extended period of time. The source
owner or operator is responsible for the
proper calibration, maintenance, and
operation of the CEMS at all times.
1.2 Principle. A gas sample is extracted
from the source through a heated sample line
and heated filter to a flame ionization
detector (FID). Results are reported as volume
concentration equivalents of propane.
Installation and measurement location
specifications, performance and equipment
specifications, test and data reduction
procedures, and brief quality assurance
guidelines are Included in the specifications.
Calibration drift, calibration error, and
response time tests are conducted to
determine conformance of the CEMS with the
specifications.
2. Definitions
2.1 Continuous Emission Monitoring
System (CEMS). The total equipment used to
acquire data, which includes sample
extraction and transport hardware, analyzer,
data recording and processing hardware, and
software. The system consists of the
following major subsystems:
2.1.1 Sample Interface. That portion of
the system that is used for one or more of the
following: Sample acquisition, sample
transportation, sample conditioning, or
protection of the analyzer from the effects of
the stack effluent.
2.1.2 Organic Analyzer. That portion of
the system that senses organic concentration
and generates an output proportional to the
gas concentration.
2.1.3 Data Recorder. That portion of the
system that records a permanent record of the
measurement values. The data recorder may
include automatic data reduction
capabilities.
2.2 Instrument Measurement Range. The
difference between the minimum and
maximum concentration that can be
measured by a specific instrument. The
minimum is often stated or assumed to be
zero and the range expressed only as the
maximum.
2.3 Span or Span Value. Full scale
instrument measurement range. The span
value shall be documented by the CEMS
manufacturer with laboratory data.
2.4 Calibration Gas. A known
concentration of a gas in an appropriate
diluent gas.
2.5 Calibration Drift (CD). The difference
in the CEMS output readings from the
established reference value after a stated
period of operation during which no
unscheduled maintenance, repair, or
adjustment takes place. A CD test is
performed to demonstrate the stability of the
OEMS calibration over time.
2.6 Response Time. The time interval
between the start of a step change in the
system input (e.g., change of calibration gas)
and the time when the data recorder displays
95 percent of the final value.
2.7 Accuracy. A measurement of
agreement between a measured value and an
accepted or true value, expressed as the
percentage difference between the true and
measured values relative to the true value.
For these performance specifications,
accuracy is checked by conducting a
calibration error (CE) test.
2.8 Calibration Error (CE). The difference
between the concentration indicated by the
CEMS and the known concentration of the
cylinder gas. A CE test procedure is
performed to document the accuracy and
linearity of the monitoring equipment over
the entire measurement range.
2.9 Performance Specification Test (PST)
Period. The period during which CD, CE, and
response time tests are conducted.
2.10 Centroidal Area. A concentric area
that is geometrically similar to the stack or
duct cross section and is no greater than 1
percent of the stack or duct cross-sectional
area.
3. Installation and Measurement Location
Specifications
3.1 CEMS Installation and Measurement
Locations. The CEMS shall be installed in a
location in which measurements
representative of the source's emissions can
be obtained. The optimum location of the
sample interface for the CEMS is determined
by a number of factors, including ease of
access for calibration and maintenance, the
degree to which sample conditioning will be
required, the degree to which it represents
total emissions, and the degree to which it
represents the combustion situation in the
firebox. The location should be as free from
in-leakage influences as possible and
reasonably free from severe flow
disturbances. The sample location should be
at least two equivalent duct diameters
downstream from the nearest control device,
point of pollutant generation, or other point
at which a change in the pollutant
concentration or emission rate occurs and at
least 0.5 diameter upstream from the exhaust
or control device. The equivalent duct
diameter is calculated as per 40 CFR part 60,
appendix A, method 1, section 2.1. If these
criteria are not achievable or if the location
is otherwise less than optimum, the
possibility of stratification should be
investigated as described in section 3.2. The
measurement point shall be within the
centroidal area of the stack or duct cross
section.
3.2 Stratification Test Procedure.
Stratification is defined as a difference in
excess of 10 percent between the average
concentration in the duct or stack and the
concentration at any point more than 1.0
meter from the duct or stack wall. To • f- •"
determine whether effluent stratification
exists, a dual probe system should be used
to determine the average effluent
concentration while measurements at each
traverse point are being made. One probe,
located at the stack or duct centroid, is used
as a stationary reference point to indicate the
change in effluent concentration over time..
The second probe is used for sampling at the
traverse points specified in 40 CFR part 60
appendix A, method 1. The monitoring
system samples sequentially at the reference
and traverse points throughout the testing
period for five minutes at each point.
4. CEMS Performance and Equipment
Specifications
If this method is applied in highly
explosive areas, caution and care shall be
exercised in choice of equipment and
installation.
4.1 Flame Ionization Detector (FID)
Analyzer. A heated FID analyzer capable of
meeting or exceeding the requirements of
these specifications. Heated systems shall
maintain the temperature of the sample gas
between 150 °C (300 °F) and 175 °C (350 °F)
throughout the system. This requires all
system components such as the probe,
calibration valve, filter, sample lines, pump,
and the FID to be kept heated at all times
such that no moisture is condensed out of the
system. The essential components of the
measurement system are described below:
4.1.1 Sample Probe. Stainless steel, or
equivalent, to collect a gas sample from the
centroidal area of the stack cross-section.
4.1.2 Sample Line. Stainless steel or
Teflon tubing to transport the sample to the
analyzer.
Note: Mention of trade names or specific
products does not Constitute endorsement by
the Environmental Protection Agency.
4.1.3 Calibration Valve Assembly. A
heated three-way valve assembly to direct the
zero and calibration gases to the analyzer is
recommended. Other methods, such as •
quick-connect lines, to route calibration gas
to the analyzers are applicable.
4.1.4 Particulate Filter. An in-stack or
out-of-stack sintered stainless steel filter is
recommended if exhaust-gas particulate
loading is significant. An out-of-stack filter
must be heated.
4.1.5 Fuel. The fuel specified by the
manufacturer (e.g., 40 percent hydrogen/60
percent helium, 40 percent hydrogen/60
percent nitrogen gas mixtures, or pure
hydrogen) should be used.
4.1.6 Zero Gas. High purity air with less
than 0.1 parts per million by volume (ppm)
HC as methane or carbon equivalent or less
than 0.1 percent of the span value, whichever
is greater.
4.1.7 Calibration Gases. Appropriate
concentrations of propane gas (in air or
nitrogen). Preparation of the calibration gases
should be done according to the procedures
in EPA Protocol 1. In addition, the
manufacturer of the cylinder gas should
provide a recommended shelf life for each
calibration gas cylinder over which the
concentration does not change by more than
±2 percent from the certified value.
4.2 CEMS Span Value. 100 ppm propane.
The span value shall be documented by the
CEMS manufacturer with laboratory data.
-------
17498
Federal Register / Vol. 61, No. 77 / .Friday, April 19, 1996 7 Proposed Rules
4.3 Daily Calibration Gas Values. The
owner or operator must choose calibration
gas concentrations that include zero and
high-level calibration values.
4.3.1 The zero level may be between zero
and 0.1 ppm (zero and 0.1 percent of the
span value).
4.3.2 The high-level concentration shall
be between 50 and 90 ppm (50 and 90
percent of the span value).
4.4 Data Recorder Scale. The strip chart
recorder, computer, or digital recorder must
be capable of recording all readings within
the GEMS' measurement range and shall have
a resolution of 0.5 ppm (0.5 percent of span
value).
4.5 Response Time. The response time for
the GEMS must not exceed 2 minutes to
achieve 95 percent of the final stable value.
4.6 Calibration Drift. The GEMS must
allow the determination of CD at the zero and
high-level values. The GEMS calibration
response must not differ by more than ±3
ppm (+3 percent of the span value) after each
24-hour period of the 7-day test at both zero
and high levels.
4.7 Calibration Error. The mean
difference between the GEMS and reference
values at all three test points listed below
shall be no greater than 5 ppm (±5 percent
of the span value).
4.7.1 Zero Level. Zero to 0.1 ppm (0 to 0.1
percent of span value).
4,7.2 Mid-Level. 30 to 40 ppm (30 to 40
percent of span value).
4.7.3 High-Level. 70 to 80 ppm (70 to 80
percent of span value).
4.8 Measurement and Recording
Frequency. The sample to be analyzed shall
pass through the measurement section of the
analyzer without interruption. The detector
shall measure the sample concentration at
least once every 15 seconds. An average
emission rate shall be computed and
recorded at least once every 60 seconds.
4.10 Retest. If the GEMS produces results
within the specified criteria, the test is
successful. If the GEMS does not meet one or
more of the criteria, necessary corrections
must be made and the performance tests
repeated.
5. Performance Specification Test (PST)
Periods
5.1 . Pretest Preparation Period. Install the
GEMS, prepare the PTM test site according to
the specifications in section 3, and prepare
the GEMS for operation and calibration
according to the manufacturer's written
instructions. A pretest conditioning period
similar to that of the 7-day CD test is
recommended to verify the operational status
of the GEMS.
5.2 Calibration Drift Test Period. While
the facility is operating under normal
conditions, determine the magnitude of the
CD at 24-hour intervals for seven consecutive
days according to the procedure given in
section 6.1. All CD determinations must be
made following a 24-hour period during
which no unscheduled maintenance, repair,
or adjustment takes place. If the combustion
unit is taken out of service during the test
period, record the onset and duration of the
downtime and continue the CD test when the
unit resumes operation.
5.3 Calibration Error Test and Response
Time Test Periods. Conduct the CE and
response time tests during the CD test period.
6. Performance Specification Test
Procedures
6.1 Relative Accuracy Test Audit (RATA)
and Absolute Calibration Audits (ACA). The
test procedures described in this section are
in lieu of a RATA and ACA.
6.2 Calibration Drift Test.
6.2.1 Sampling Strategy. Conduct the CD
test at 24-hour intervals for seven
consecutive days using calibration gases at
the two daily concentration levels specified
automatic or manual adjustments are made to
the GEMS zero and calibration settings,
conduct the CD test immediately before these
adjustments, or conduct it in such a way that
the CD can be determined. Record the GEMS
response and subtract this value from the
reference (calibration gas) value. To meet the
specification, none of the differences shall
exceed 3 percent of the span of the GEM.
6.2.2 Calculations. Summarize the results
on a data sheet. An example is shown in
Figure 1. Calculate the differences between
the GEMS responses and the reference
values.
6.3 Response Time. The entire system
including'sample extraction and transport,
sample conditioning, gas analyses, and the
data recording is checked with this
procedure.
6.3.1 Introduce the calibration gases at
the probe as near to the sample location as
possible. Introduce the zero gas into the
system. When the system output.has
stabilized (no change greater than 1 percent
of full scale for 30 sec), switch to monitor
stack effluent and wait for a stable value.
Record the time (upscale response .time)
required to reach 95 percent of the final
stable value.
6.3.2 Next, introduce a high-level
calibration gas and repeat the above
procedure. Repeat the entire procedure three
times and determine the mean upscale and
downscale response times. The longer of the
two means is the system response time.
6.4 Calibration Error Test Procedure.
6.4.1 Sampling Strategy. Challenge the
GEMS with zero gas and EPA Protocol 1
cylinder gases at measurement points within
the ranges specified in section 4.7.
6.4.1.1 The daily calibration gases, if
Protocol 1, may be used for this test.
Source: •
Monitor:
4.9 Hourly Rolling Average Calculation. in section 4.3. Introduce the two calibration .
The GEMS shall calculate every minute an gases into the sampling system as close to the f6"
hourly rolling average, which is the sampling probe outlet as practical. The gas Date:
arithmetic mean of the 60 most recent 1- shall pass through all GEM components used Locat
minute average values Curing normal sampling. Tf periodic Span
Number:
ion:
Day
Zero/low level:
1
2
3
4
5
6
7
High level:
1
2
3
4
5
6
7
Date
Time
Calibration value
Monitor response
Difference
Percent of span
(1)
1=Acceptance Criteria: < 3% of span each day for seven days.
Figure 1: Calibration Drift Determination
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17499
6.4.1.2 Operate the GEMS as nearly as
possible in its normal sampling mode. The
calibration gas should be injected into the
sampling system as close to the sampling
probo outlet as practical and shall pass
through all filters, scrubbers, conditioners,
and other monitor components used during
normal sampling. Challenge the GEMS three
non-consecutive times at each measurement
point and record the responses. The duration
of each gas injection should be for a
sufficient period of time to ensure that the
OEMS surfaces are conditioned.
6.4.2 Calculations. Summarize the results
on a data sheet. An example data sheet is
shown in Figure 2. Average the differences
between the instrument response and the
certified cylinder gas value for each gas.
Calculate three CE results according to
Equation 1. No confidence coefficient is used
in CE calculations.
7. Equations
7.1 Calibration Error. Calculate CE using
Equation 1.
CE = |d/FS|xlOO (Eq.l)
Where:
d = Mean difference between GEMS response
and the known reference concentration,
determined using Equation 2.
(Eq.2)
n=
i=i
di = Individual difference between GEMS
response and the known reference
concentration.
8. Reporting
At a minimum, summarize in tabular form
the results of the CD, response time, and CE
test, as appropriate. Include all data sheets,
calculations, GEMS data records, and
cylinder gas or reference material
certifications.
Source:
Monitor:
Serial Number:
Date:
Location:
Span:
Run No.
1-Zero
2-Mid
3-High
4-MkJ
5-Zero
6-High
7-Zero
8-Mid
9-H!gh
Calibration
value
Mean Difl
Calibratio
Monitor
response
ference =
n Error =
Difference
Zero/Low
Mid
High
Figure 2: Calibration Error Determination
9. References
1. Measurement of Volatile Organic
Compounds-Guideline Series. U.S.
Environmental Protection Agency, Research
Triangle Park, North Carolina, 27711, EPA-
450/2-78-041, June 1978.
2. Traceability Protocol for Establishing
True Concentrations of Gases Used for
Calibration and Audits of Continuous Source
Emission Monitors (Protocol No. 1). U.S.
Environmental Protection Agency ORD/
EMSL, Research Triangle Park, North
Carolina, 27711, June 1978.
3. Gasoline Vapor Emission Laboratory
Evaluation-Part 2. U.S. Environmental
Protection Agency, OAQPS, Research
Triangle Park, North Carolina, 27711, EMB
Report No. 76-GAS-6, August 1975.
*****
Performance Specification 10—
Specifications and test procedures for multi-
metals continuous monitoring systems in
stationary sources.
1, Applicability and Principle
1.1 Applicability. This specification is to
bo used for evaluating the acceptability of
multi-metals continuous emission
monitoring systems (GEMS) at the time of or
soon after installation and whenever
specified in the regulations. The GEMS may
include, for certain stationary sources, (a) a
diluent (Oz) monitor (which must meet its
own performance specifications: 40 CFR part
60, Appendix B, Performance Specification
3), (b) flow monitoring equipment to allow
measurement of the dry volume of stack
effluent sampled, and (c) an automatic
sampling system.
A multi-metals GEMS must be capable of
measuring the total concentrations
(regardless of speciation) of two or more of
the following metals in both their vapor and
solid forms: Antimony (Sb), Arsenic (As),
Barium (Ba), Beryllium (Be), Cadmium (Cd),
Chromium (Cr), Lead (Pb), Mercury (Hg),
Silver (Ag), Thallium (Tl), Manganese (Mn),
Cobalt (Co), Nickel (Ni), and Selenium (Se).
Additional metals may be added to this list
at a later date by addition of appendices to
this performance specification. If a GEMS
does not measure a particular metal or fails
to meet the performance specifications for a
particular metal, then the GEMS may not be
used to determine emission compliance with
the applicable regulation for that metal.
This specification is not designed to
evaluate the installed GEMS' performance
over an extended period of time nor does it
identify specific calibration techniques and
auxiliary procedures to assess the GEMS'
performance. The source owner or operator,
however, is responsible to properly calibrate,
maintain, and operate the GEMS. To evaluate
the GEMS' performance, the Administrator
may require, under Section 114 of the Act,
the operator to conduct GEMS performance
evaluations at other times besides the initial
test. See Sec. 60.13 (c) and "Quality
Assurance Requirements For Multi-Metals
Continuous Emission Monitoring Systems
Used For Compliance Determination."
1.2 Principle. Installation and
measurement location specifications,
performance specifications, test procedures,
and data reduction procedures are included
in this specification. Reference method tests
and calibration drift tests are conducted to
determine conformance of the GEMS with the
specification.
2. Definitions
2.1 Continuous Emission Monitoring
System (GEMS). The total equipment
required for the determination of a metal
concentration. The system consists of the
following major subsystems:
2.1.1 Sample Interface. That portion of
the GEMS used for one or more of the
following: sample acquisition, sample
transport, and sample conditioning, or
protection of the monitor from the effects of
the stack effluent.
2.1.2 Pollutant Analyzer. That portion of
the GEMS that senses the metals
concentrations and generates a proportional
output.
2.1.3 Diluent Analyzer (if applicable).
That portion of the GEMS that senses the
diluent gas (O2) and generates an output
proportional to the gas concentration.
-------
27500 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
2.1.4. Data Recorder. That portion of the
GEMS that provides a permanent record of
the analyzer output. The data recorder may
provide automatic data reduction and GEMS
control capabilities.
2.2 Point GEMS. A GEMS that measures
the metals concentrations either at a single
point or along a path equal to or less than
10 percent of the equivalent diameter of the
stack or duct cross section.
2.3 Path GEMS. A GEMS that measures
the metals concentrations along a path
greater than 10 percent of the equivalent
diameter of the stack or duct cross section.
2.4 Span Value. The upper limit of a
metals concentration measurement range
defined as twenty times the applicable
emission limit for each metal. The span value
shall be documented by the GEMS
manufacturer with laboratory data.
2.5 Relative Accuracy (RA). The absolute
mean difference between the metals
concentrations determined by the GEMS and
the value determined by the reference
method (KM) plus the 2.5 percent error
confidence coefficient of a series of tests
divided by the mean of the RM tests or the
applicable emission limit.
2.6 Calibration Drift (CD). The difference
in the GEMS output readings from the
established reference value after a stated
period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.7 Zero Drift (ZD). The difference in the
GEMS output readings for zero input after a
stated period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.8 Representative Results. Defined by
the RA test procedure defined in this
specification.
2.9 Response Time. The time interval
between the start of a step change in the
system input and the time when the
pollutant analyzer output reaches 95 percent
of the final value.
2.10 Centroidal Area. A concentric area
that is geometrically similar to the stack or
duct cross section and is no greater than \
percent of the stack or duct cross sectional
area.
2.11 Batch Sampling. Batch sampling
refers to the technique of sampling the stack
effluent continuously and concentrating the
pollutant in some capture medium. Analysis
is performed periodically after sufficient time
has elapsed to concentrate the pollutant to
levels detectable by the analyzer.
2.12 Calibration Standard. Calibration
standards consist of a known amount of
metal(s) that are presented to the pollutant
analyzer portion of the GEMS in order to
calibrate the drift or response of the analyzer.
The calibration standard may be, for
example, a solution containing a known
metal concentration, or a filter with a known
mass loading or composition.
3. Installation and Measurement Location
Specifications
3.1 The GEMS Installation and
measurement location. Install the GEMS at an
accessible location downstream of all
pollution control equipment where the
metals concentrations measurements are
directly representative or can be corrected so
as to be representative of the total emissions
from the affected facility. Then select
representative measurement points or paths
for monitoring in locations that the GEMS
will pass the RA test (see Section 7). If the
cause of failure to meet the RA test is
determined to be the measurement location
and a satisfactory correction technique
cannot be established, the Administrator may
require the GEMS to be relocated.
Measurement locations and points or paths
that are most likely to provide data that will
meet the RA requirements are listed below.
3.1.1 Measurement Location. The
measurement location should be (1) at least
eight equivalent diameters downstream of the
nearest control device, point of pollutant
generation, bend, or other point at which a
change of pollutant concentration or flow
disturbance may occur, and (2) at least two
equivalent diameters upstream from the
effluent exhaust. The equivalent duct
diameter is calculated as per 40 CFR part 60,
Appendix A, Method 1, Section 2.1.
3.1.2 Point GEMS. The measurement
point should be (1) no less than 1.0 meter
from the stack or duct wall or (2) within or
centrally located over the centroidal area of
the stack or duct cross section. Selection of
traverse points to determine the
representativeness of the measurement
location should be made according to 40 CFR
part 60, Appendix A, Method 1, Sections 2.2
and 2.3.
3.1.3 Path GEMS. The effective
measurement path should be (1) totally
within the inner area bounded by a line 1.0
meter from the stack or duct wall, or (2) have
at least 70 percent of the path within the
inner 50 percent of the stack or duct cross
sectional area, or (3) be centrally located over
any part of the centroidal area.
3.2 Reference Method (RM) Measurement
Location and Traverse Points. The RM
measurement location should be (1) at least
eight equivalent diameters downstream of the
nearest control device, point of pollutant
generation, bend, or other point at which a
change of pollutant concentration or flow
disturbance may occur, and (2) at least two
equivalent diameters upstream from the
effluent exhaust. The RM and GEMS
locations need not be the same, however the
difference may contribute to failure of the
GEMS to pass the RA test, thus they should
be as close as possible without causing
interference with one another. The
equivalent duct diameter is calculated as per
40 CFR part 60, Appendix A, Method 1,
Section 2.1. Selection of traverse
measurement point locations should be made
according to 40 CFR part 60, Appendix A,
Method 1, Sections 2.2 and 2.3. If the RM
traverse line interferes with or is interfered
by the GEMS measurements, the line may be
displaced up to 30 cm (or 5 percent of the
equivalent diameter of the cross section,
whichever is less) from the centroidal area.
4. Performance and Equipment
Specifications
4.1 Data Recorder Scale. The GEMS data
recorder response range must include zero
and a high level value. The high level value
must be equal to the span value. If a lower
high level value is used, the GEMS must have
the capability of providing multiple outputs
with different high level values (one of which
is equal to the span value) or be capable of
automatically changing the high level value
as required (up to the span value) such that
the measured value does not exceed 95
percent of the high level value.
4.2 Relative Accuracy (RA). The RA of
the GEMS must be no greater than 20 percent
of the mean value of the RM test data in
terms of units of the emission standard for
each metal, or 10 percent of the applicable
standard, whichever is greater.
4.3 Calibration Drift, The GEMS design
must allow the determination of calibration
drift at concentration levels commensurate
with the applicable emission standard for
each metal monitored. The GEMS calibration
-may not drift or deviate from the reference
value (RV) of the calibration standard used
for each metal by more than 5 percent of the
emission standard for each metal. The
calibration shall be performed at a point
equal to 80 to 120 percent of the applicable
emission standard for each metal.
4.4 Zero Drift. The GEMS design must
allow the determination of calibration drift at
the zero level (zero drift) for each metal. If
this is not possible or practicable, the design
must allow the zero drift determination to be
made at a low level value (zero to 20 percent
of the emission limit value). The GEMS zero
point for each metal shall not drift by more
than 5 percent of the emission standard for
that metal.
4.5 Sampling and Response Time. The
GEMS shall sample the stack effluent
continuously. Averaging time, the number of
measurements in an average, and the
averaging procedure for reporting and
determining compliance shall conform with
that specified in the applicable emission
regulation.
4.5.1 Response Time for Instantaneous,
Continuous GEMS. The response time for the
GEMS must not exceed 2 minutes to achieve
95 percent of the final stable value.
4.5.2 Waiver from Response Time
Requirement. A source owner or operator
may receive a waiver from the response time
requirement for instantaneous, continuous
GEMS in section 4.5.1 from the Agency if no
GEM is available which can meet this
specification at the time of purchase of the
GEMS.
4.5.3 Response Time for Batch CEMS.
The response time requirement of Sections
4.5.1 and 4.5.2 do not apply to batch CEMS.
Instead it is required that the sampling time
be no longer than one third of the averaging
period for the applicable standard. In
addition, the delay between the end of the
sampling period and reporting of the sample
analysis shall be no greater than one hour.
Sampling is also required to be continuous
except in that the pause in sampling when
the sample collection media are changed
should be no greater than five percent of the
averaging period or five minutes, whichever
is less.
5. Performance Specification Test Procedure
5.1 Pretest Preparation. Install the CEMS
and prepare the RM test site according to the
specifications in Section 3, and prepare the
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17501
GEMS for operation according to the
manufacturer's written instructions.
5.2 Calibration and Zero Drift Test
Period. While the affected facility is
operating at more than 50 percent of normal
load, or as specified in an applicable subpart,
determine the magnitude of die calibration
drift (CD) and zero drift (ZD) once each day
(at 24-hour intervals) for 7 consecutive days
according to the procedure given in Section
6. To meet the requirements of Sections 4.3
and 4.4 none of the CD's or ZD's may exceed
the specification. All CD determinations
must be made following a 24-hour period
during which no unscheduled maintenance,
repair, or manual adjustment of the CEMS
took place.
5.3 RA Test Period. Conduct a RA test
following the CD test period. Conduct the RA
tost according to the procedure given in
Section 7 while the affected facility is
operating at more than 50 percent of normal
load, or as specified in the applicable
subpart.
6.0 The CEMS Calibration and Zero Drift
Test Procedure
This performance specification is designed
to allow calibration of the CEMS by use of
standard solutions, filters, etc. that challenge
the pollutant analyzer part of the CEMS (and
as much of the whole system as possible), but
which do not challenge the entire CEMS,
including the sampling interface. Satisfactory
response of the entire system is covered by
the RA requirements.
The CD measurement is to verify the ability
of the CEMS to conform to the established
CEMS calibration used for determining the
omission concentration. Therefore, if
periodic automatic or manual adjustments
are made to the CEMS zero and calibration
settings, conduct the CD test immediately
before the adjustments, or conduct it in such
a way that the CD and ZD can be determined.
Conduct the CD and ZD tests at the points
specified in Sections 4.3 and 4.4. Record the
CEMS response and calculate the CD
according to:
CD = -
Rv
-xlOO,
(1)
Where CD denotes the calibration drift of the
CEMS in percent, RCEM is the CEMS
response, and Rv is the reference value of the
high level calibration standard. Calculate the
ZD according to:
ZD = -
R
xlOO,
(2)
EM
Where ZD denotes the zero drift of the CEMS
in percent, RCEM is the CEMS response, Rv
is the reference value of the low level
calibration standard, and REM is the
emission limit value.
7. Relative Accuracy Test Procedure
7.1 Sampling Strategy for RA Tests. The
RA tests are to verify the initial performance
of the entire CEMS system, including the
sampling interface, by comparison to RM
measurements. Conduct the RM
measurements in such a way that they will
yield results representative of the emissions
from the source and can be correlated to the
CEMS data. Although it is preferable to
conduct the diluent (if applicable), moisture
(if needed), and pollutant measurements
simultaneously, the diluent and moisture
measurements that are taken within a 30 to
60-minute period, which includes the
pollutant measurements, may be used to
calculate dry pollutant concentration.
A measure of relative accuracy at a single
level is required for each metal measured for
compliance purposes by the CEMS. Thus the
concentration of each metal must be
detectable by both the CEMS and the RM. In
addition, the RA must be determined at three
levels (0 to 20, 40 to 60, and 80 to 120
percent of the emission limit) for one of the
metals which will be monitored, or for iron.
If iron is chosen, the three levels should be
chosen to correspond to those for one of the
metals that will be monitored using known
sensitivities (documented by the
manufacturer) of the CEMS to both metals.
In order to correlate the CEMS and RM
data properly, note the beginning and end of
each RM test period of each run (including
the exact time of day) in the CEMS data log.
Use the following strategy for the RM
measurements:
7.2 Correlation of RM and CEMS Data.
Correlate the CEMS and RM test data as to
the time and duration by first determining
from the CEMS final output (the one used for
reporting) the integrated average pollutant
concentration for each RM test period.
Consider system response time, if important,
and confirm that the pah- of results are on a
consistent moisture, temperature, and diluent
concentration basis. Then compare each
integrated CEMS value against the
corresponding average RM value.
7.3 Number of tests. Obtain a minimum
of three pairs of CEMS and RM
measurements for each metal required and at
each level required (see Section 7.1). If more
than nine pairs of measurements are
obtained, then up to three pairs of
measurements may be rejected so long as the
total number of measurement pairs used to
determine the RA is greater than or equal to
nine. However, all data, including the
rejected data, must be reported.
7.4 Reference Methods. Unless otherwise
specified in an applicable subpart of the
regulations, Method 3B, or its approved
alternative, is the reference method for
diluent (O2) concentration. Unless otherwise
specified in an applicable subpart of the
regulations, the manual method for multi-
metals in 40 CFR part 266, Appendix IX,
Section 3.1 (until superseded by SW-846), or
its approved alternative, is the reference
method for multi-metals.
As of March 22,1995 there is no approved
alternative RM to Method 29 (for example, a
second metals CEMS, calibrated absolutely
according to the alternate procedure to be
specified in an appendix to this performance
specification to be added when an absolute
system calibration procedure becomes
available and is approved).
7.5 Calculations. Summarize the results
on a data sheet. An example is shown in
Figure 2-2 of 40 CFR part 60, Appendix B,
Performance Specification 2. Calculate the
mean of the RM values. Calculate the
arithmetic differences between the RM and
CEMS output sets, and then calculate the
mean of the differences. Calculate the
standard deviation of each data set and
CEMS RA using the equations in Section 8.
7.6 Undetectable Emission Levels. In the
event of metals emissions concentrations
from the source being so low as to be
undetectable by the CEMS operating in its
normal mode (i.e., measurement times and
frequencies within the bounds of the
performance specifications), then spiking of
the appropriate metals in the feed or other
operation of the facility in such a way as to
raise the metal concentration to a level
detectable by both the CEMS and the RM is
required in order to perform the RA test.
8. Equations
8.1 Arithmetic Mean. Calculate the
arithmetic mean of a data set as follows:
Where n is equal to the number of data
points.
8.1.1 Calculate the arithmetic mean of the
difference, d, of a data set, using Equation 3
and substituting d for x. Then
di=xi-yi, (4)
Where x and y are paired data points from
the CEMS and RM, respectively.
8.2 Standard Deviation. Calculate the
standard deviation (SD) of a data set as
follows:
SD = "
n-1
(5)
8.3 Relative Accuracy (RA). Calculate the
RA as follows:
RA = -
R
(6)
RM
Where d is equal to the arithmetic mean of
the difference, d, of the paired CEMS and RM
data set, calculated according to Equations 3
and 4, SD is the standard deviation
calculated according to Equation 5, RRM is
equal to either the average of the RM data set,
calculated according to Equation 3, or the
value of the emission standard, as applicable
(see Section 4.2), and to.975 is the t-value at
2.5 percent error confidence, see Table 1.
-------
17502
TABLE 1
[t-Values]
na
,
0
6 • ,-•
to.975
12.706
4.303
3.182
2.776
2.571
na
7
8
9
10
11
to.975
2.447
2.365
2.306
2.262
2.228
na
12
13
14
15
16
to.975
2.201
2.179
2.160
2.145
2.131
"The values in this table are already corrected for n-1 degrees of freedom. Use n equal to the number of individual values
9. Reporting
At a minimum (check with the appropriate
regional office, or State, or local agency for
additional requirements, if any) summarize
in tabular form the results of the CD tests and
the RA tests or alternate RA procedure as
appropriate. Include all data sheets,
calculations, and records of GEMS response
necessary to substantiate that the
performance of the GEMS met the
performance specifications.
The GEMS measurements shall be reported
to the agency in units of ug/m3 on a dry basis,
corrected to 20°C and 7 percent 02.
10. Alternative Procedures
A procedure for a total system calibration,
when developed, will be acceptable as a
procedure for determining RA. Such a
procedure will involve challenging the entire
GEMS, including the sampling interface, with
a known metals concentration. This
procedure will be added as an appendix to
this performance specification when it has
been developed and approved. The RA
requirement of Section 4.2 will remain
unchanged.
11. Bibliography
1. 40 CFR part 60, Appendix B,
"Performance Specification 2—Specifications
and Test Procedures for SOa and NOX
Continuous Emission Monitoring Systems in
Stationary Sources."
2. 40 CFR part 60, Appendix B,
"Performance Specification 1—Specification
and Test Procedures for Opacity Continuous
Emission Monitoring Systems in Stationary
Sources."
3. 40 CFR part 60, Appendix A, "Method
1—Sample and Velocity Traverses for
Stationary Sources."
4. 40 CFR part 266, Appendix IX, Section
2, "Performance Specifications for
Continuous Emission Monitoring Systems."
5. Draft Method 29, "Determination of
Metals Emissions from Stationary Sources,"
Docket A-90-45, Item II-B-12, and EMTIC
CTM-012.WPF.
6. "Continuous Emission Monitoring
Technology Survey for Incinerators, Boilers,
and Industrial Furnaces: Final Report for
Metals CEM's," prepared for the Office of
Solid Waste, U.S. EPA, Contract No. 68-D2-
0164 (4/25/94).
Performance Specification 11—
Specifications and test procedures for
particulate matter continuous monitoring
systems in stationary sources.
1. Applicability and Principle
1.1 Applicability. This specification is to
be used for evaluating the acceptability of
particulate matter continuous emission
monitoring systems (GEMS) at the time of or
soon after installation and whenever
specified in the regulations. The GEMS may
include, for certain stationary sources, a) a
diluent (O2) monitor (which must meet its
own performance specifications: 40 CFR part
60, Appendix B, Performance Specification
3), b) flow monitoring equipment to allow
measurement of the dry volume of stack
effluent sampled, and c) an automatic
sampling system.
This performance specification requires
site specific calibration of the PM GEMS'
response against manual gravimetric method
measurements. The range of validity of the
response calibration is restricted to the range
of particulate mass loadings used to develop
the calibration relation. Further, if conditions
at the facility change (i.e., changes in
emission control system or fuel type), then a
new response calibration is required. Since
the validity of the response calibration may
be affected by changes in the properties of
the particulate, such as density, index of
refraction, and size distribution, the
limitations of the GEMS used should be
evaluated with respect to these possible
changes on a site specific basis.
This specification is not designed to
evaluate the installed GEMS' performance
over an extended period of time nor does it
identify specific calibration techniques and
auxiliary procedures to assess the GEMS'
performance. The source owner or operator,
however, is responsible to properly calibrate,
maintain, and operate the GEMS. To evaluate
the GEMS' performance, the Administrator
may require, under Section 114 of the Act,
the operator to conduct GEMS performance
evaluations at other times besides the initial
test. See Sec. 60.13 (c) and "Quality
Assurance Requirements For Particulate
Matter Continuous Emission Monitoring
Systems Used For Compliance
Determination."
1.2 Principle. Installation and
measurement location specifications,
performance specifications, test procedures,
and data reduction procedures are included
in this specification. Reference method tests
and calibration drift tests are conducted to '
determine conformance of the GEMS with the
specification.
2. Definitions
2.1 Continuous Emission Monitoring
System (GEMS). The total equipment
required for the determination of particulate
matter mass concentration. The system
consists of the following major subsystems:
2.1.1 Sample Interface. That portion of
the GEMS used for one or more of the
following: sample acquisition, sample
transport, and sample conditioning, or
protection of the monitor from the effects of
the stack effluent.
2.1.2 Pollutant Analyzer. That portion of
the GEMS that senses the particulate matter
concentration and generates a proportional
output.
2.1.3 Diluent Analyzer (if applicable).
That portion of the GEMS that senses the
diluent gas (O2) and generates an output
proportional to the gas concentration.
2.1.4 Data Recorder. That portion of the
GEMS that provides a permanent record of
the analyzer output. The, data recorder may
provide automatic data reduction and GEMS
control capabilities.
2.2 Point GEMS. A GEMS that measures
particulate matter mass concentrations either
at a single point or along a path equal to or
less than 10 percent of the equivalent
diameter of the stack or duct cross section.
2.3 Path GEMS. A GEMS that measures
particulate matter mass concentrations along
a path greater than 10 percent of the
equivalent diameter of the stack or duct cross
section.
2.4 Span Value. The upper limit of the
GEMS measurement range. The span value
shall be documented by the GEMS
manufacturer with laboratory data.
2.5 Confidence Interval. The interval with
upper and lower limits within which the
GEMS response calibration relation lies with
a given level of confidence.
2.6 Tolerance Interval. The interval with
upper and lower limits within which are
contained a specified percentage of the
population with a given level of confidence.
2.7 . Calibration Drift (CD). The difference
in the GEMS output readings from the
established reference value after a stated
period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.8 Zero Drift (ZD). The difference in the
GEMS output readings for zero input after a
stated period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.9 Representative Results. Defined by
the reference method test procedure defined
in this specification.
2.10 Response Time. The time interval
between the start of a step change in the
system input and the time when the
pollutant analyzer output reaches 95 percent
of the final value.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17503
2.11 Centroidal Area. A concentric area
that is geometrically similar to the stack or
duct cross section and is no greater than 1
percent of the stack or duct cross sectional
area.
2.12 Batch Sampling. Batch sampling
refers to the technique of sampling the stack
effluent continuously and concentrating the
pollutant in some capture medium. Analysis
is performed periodically after sufficient time
has elapsed to concentrate the pollutant to
levels detectable by the analyzer.
2.13 Calibration Standard. Calibration
standards produce a known and unchanging
response when presented to the pollutant
analyzer portion of the GEMS, and are used
to calibrate the drift or response of the
analyzer.
3. Installation and Measurement Location
Specifications
3.1 The GEMS Installation and
measurement location. Install the CEMS at an
accessible location downstream of all
pollution control equipment where the
particulate matter mass concentrations
measurements are directly representative or
can bo corrected so as to be representative of
the total emissions from the affected facility.
Then select representative measurement
points or paths for monitoring in locations
that the CEMS will meet the calibration
requirements (see Section 7). If the cause of
failure to meet the calibration requirements
is determined to be the measurement location
and a satisfactory correction technique
cannot be established, the Administrator may
require the CEMS to be relocated.
Measurement locations and points or paths
that are most likely to provide data that will
moot the calibration requirements are listed
below.
3,1.1 Measurement Location. The
measurement location should be (1) at least
eight equivalent diameters downstream of the
nearest control device, point of pollutant
generation, bend, or other point at which a
change of pollutant concentration or flow
disturbance may occur and (2) at least two
equivalent diameters upstream from the
effluent exhaust. The equivalent duct
diameter is calculated as per 40 CFR part 60,
Appendix A, Method 1, Section 2.1.
3.1.2 Point CEMS. The measurement
point should be (1) no less than 1.0 meter
from the stack or duct wall or (2) within or
centrally located over the centroidal area of
the stack or duct cross section. Selection of
traverse points to determine the
representativeness of the measurement
location should be made according to 40 CFR
part 60, Appendix A, Method 1, Section 2.2
and 2.3.
3.1.3 Path CEMS. The effective
measurement path should be (1) totally
within the inner area bounded by a line 1.0
meter from the stack or duct wall, or (2) have
at least 70 percent of the path within the
inner 50 percent of the stack or duct cross
sectional area, or (3) be centrally located over
any part of the centroidal area.
3.1.4 Sampling Requirement for Saturated
Flue Gas. If the CEMS is to be installed
downstream of a wet air pollution control
system such that the flue gases are saturated
with water, then the CEMS must
isokinetically extract and heat a sample of
the flue gas for measurement so that the
pollutant analyzer portion of the CEMS
measures only dry particulate. Heating shall
be to a temperature above the water
condensation temperature of the extracted
gas and shall be maintained at all points in
the sample line, from where the flue gas is
extracted to and including the pollutant
analyzer. Performance of a CEMS design
configured in this manner must be
documented by the CEMS manufacturer.
3.2 Reference Method (RM) Measurement
Location and Traverse Points. The RM
measurement location should be (1) at least
eight equivalent diameters downstream of the
nearest control device, point of pollutant
generation, bend, or other point at which a
change of pollutant concentration or flow
disturbance may occur and (2) at least two
equivalent diameters upstream from the
effluent exhaust. The RM and CEMS
locations need not be the same, however the
difference may contribute to failure of the
CEMS to pass the RA test, thus they should
be as close as possible without causing
interference with one another. The
equivalent duct diameter is calculated as per
40 CFR part 60, Appendix A, Method 1,
Section 2.1. Selection of traverse
measurement point locations should be made
according to 40 CFR part 60, Appendix A,
Method 1, Sections 2.2 and 2.3. If the RM
traverse line interferes with or is interfered
by the CEMS measurements, the line may be
displaced up to 30 cm (or 5 percent of the
equivalent diameter of the cross section,
whichever is less) from the centroidal area.
4. Performance and Equipment
Specifications
4.1 Span and Data Recorder Scale.
4.1.1 Span. The span of the instrument
shall be three times the applicable emission
limit. The span value shall be documented by
the CEMS manufacturer with laboratory data.
4.1.2 Data Recorder Scale. The CEMS
data recorder response range must include
zero and a high level value. The high level
value must be equal to the span value. If a
lower high level value is used, the CEMS
must have the capability of providing
multiple outputs with different high level
values (one of which is equal to the span
value) or be capable of automatically
changing the high level value as required (up
to the span value) such that the measured
value does not exceed 95 percent of the high
level value.
4.2 CEMS Response Calibration
Specifications. The CEMS response
calibration relation must meet the following
specifications.
4.2.1 Correlation Coefficient. The
correlation coefficient shall be > 0.90.
4.2.2 Confidence Interval. The confidence
interval (95 percent) at the emission limit
shall be within ±20 percent of the emission
limit value.
4.2.3 Tolerance Interval. The tolerance
interval at the emission limit shall have 95
percent confidence that 75 percent of all
possible values are within ±35 percent of the
emission limit value.
4.3 Calibration Drift. The CEMS design
must allow the determination of calibration
drift at concentration levels commensurate
with the applicable emission standard. The
CEMS calibration may not drift or deviate
from the reference value (RV) of the
calibration standard by more than 2 percent
of the reference value. The calibration shall
be performed at a point equal to 80 to 120
percent of the applicable emission standard.
4.4 Zero Drift. The CEMS design must
allow the determination of calibration drift at
the zero level (zero drift). If this is not
possible or practicable, the design must allow
the zero drift determination to be made at a
low level value (zero to 20 percent of the
emission limit value). The CEMS zero point
shall not drift by more than 2 percent of the
emission standard.
4.5 Sampling and Response Time. The
CEMS shall sample the stack effluent
continuously. Averaging time, the number of
measurements in an average, and the
averaging procedure for reporting and
determining compliance shall conform with
that specified in the applicable emission
regulation.
4.5.1 Response Time. The response time
of the CEMS should not exceed 2 minutes to
achieve 95 percent of the final stable value.
The response time shall be documented by
the CEMS manufacturer.
4.5.2 Response Time for Batch CEMS.
The response time requirement of Section
4.5.1 does not apply to batch CEMS. Instead
it is required that the sampling time be no
longer than one third of the averaging period
for the applicable standard. In addition, the
delay between the end of the sampling time
and reporting of the sample analysis shall be
no greater than one hour. Sampling is also
required to be continuous except in that the
pause in sampling when the sample
collection media are changed should be no
greater than five percent of the averaging
period or five minutes, whichever is less.
5. Performance Specification Test Procedure
5.1 Pretest Preparation. Install the CEMS
and prepare the RM test site according to the
specifications in Section 3, and prepare the
CEMS for operation according to the
manufacturer's written instructions.
5.2 Calibration and Zero Drift Test
Period. While the affected facility is
operating at more than 50 percent of normal
load, or as specified in an applicable subpart,
determine the magnitude of the calibration
drift (CD) and zero drift (ZD) once each day
(at 24-hour intervals) for 7 consecutive days
according to the procedure given in Section
6. To meet the requirements of Sections 4.3
and 4.4 none of the CD's or ZD's may exceed
the specification. All CD determinations
must be made following a 24-hour period
during which no unscheduled maintenance,
repair, or manual adjustment of the CEMS
took place.
5.3 CEMS Response Calibration Period.
Calibrate the CEMS response following the
CD test period. Conduct the calibration
according to the procedure given in Section
7 while the affected facility is operating at
more than 50 percent of normal load, or as
specified in the applicable subpart.
-------
17504 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
6.0 The OEMS Calibration and Zero Drift
Test Procedure
This performance specification is designed
to allow calibration of the GEMS by use of
calibration standard that challenges the
pollutant analyzer part of the GEMS (and as
much of the whole system as possible), but
which does not challenge the entire GEMS,
including the sampling interface. Satisfactory
response of the entire system is covered by
the GEMS response calibration requirements.
The CD measurement is to verify the ability
of the GEMS to conform to the established
GEMS response calibration used for
determining the emission concentration.
Therefore, if periodic automatic or manual
adjustments are made to the GEMS zero and
calibration settings, conduct the CD test
immediately before the adjustments, or
conduct it in such a way that the CD and ZD
can be determined.
Conduct the CD and ZD tests at the points
specified in Sections 4.3 and 4.4. Record the
GEMS response and calculate the CD
according to:
(1)
R
Where CD denotes the calibration drift of the
GEMS in percent, RCEM is the GEMS
response, and Rv is the reference value of the
high level calibration standard. Calculate the
ZD according to:
ZD =
_ (RCEM~RVJ
R
xlOO,
(2)
EM
Where ZD denotes the zero drift of the GEMS
in percent, RCEM is the GEMS response, Rv
is the reference value of the low level
calibration standard, and REM is the emission
limit value.
7. CEMS Response Calibration Procedure
7.1 Sampling Strategy for Response
Calibration. The GEMS response calibration
is carried out in order to verify and calibrate
the performance of the entire GEMS system,
including the sampling interface, by
comparison to RM measurements. Conduct
the RM measurements in such a way that
they will yield results representative of the
emissions from the source and can be
correlated to the CEMS data. Although it is
preferable to conduct the diluent (if
applicable), moisture (if needed), and
pollutant measurements simultaneously, the
diluent and moisture measurements that are
taken within a 30 to 60-minute period, which
includes the pollutant measurements, may be
used to calculate dry pollutant concentration.
7.2 Correlation of RM and CEMS Data. In
order to correlate the CEMS and RM data
properly, note the beginning and end of each
RM test period of each ran (including the
exact time of day) in the CEMS data log.
Correlate the CEMS and RM test data as to
the time and duration by first determining
from the CEMS final output (the one used for
reporting) the integrated average pollutant
concentration for each RM test period.
Consider system response time, if important,
and confirm that the pair of results are on a
consistent moisture, temperature, and diluent
concentration basis. Then compare each
integrated GEMS value against the
corresponding average RM value.
7.3 Number of tests. The CEMS response
calibration shall be carried out by making
simultaneous CEMS and RM measurements
at three (or more) different levels of
particulate mass concentrations. Three (or
more) sets of measurements shall be obtained
at each level. A total of at least 15
measurements shall be obtained. The
different levels of particulate mass
concentration should be obtained by varying
the process conditions as much as the
process allows within the range of normal
operation. Alternatively, emission levels may
be varied by adjusting the particulate control
system. It is recommended that the CEMS be
calibrated over PM levels ranging from a
minimum normal level to a level roughly
twice the emission limit, as this will provide
the smallest confidence interval bounds on
the calibration relation at the emission limit
level.
7.4 Reference Methods. Unless otherwise
specified in an applicable subpart of the
regulations, Method 3B, or its approved
alternative, is the reference method for
diluent (O2) concentration. Unless otherwise
specified in an applicable subpart of the
regulations, Method 5 (40 CFR Part 60,
Appendix A), or its approved alternative, is
the reference method for particulate matter
mass concentration.
7.5 Calculations. Summarize the results
on a data sheet. An example is shown is
shown in Figure 2-2 of 40 CFR part 60,
Appendix B, Performance Specification 2.
Calculate the calibration relation, correlation
coefficient, and confidence and tolerance
intervals using the equations in Section 8.
8. Equations
8.1 Linear Calibration Relation. A linear
calibration relation may be calculated from
the calibration data by performing a linear
least squares regression. The CEMS data are
taken as the x values, and the reference
method data as the y values. The calibration
relation, which gives the predicted mass
emission, y, based on the CEMS response x,
is given by
y = a-
where:
(y,-J) m
From which the scatter of y values about the
regression line (calibration relation) SL can be
determined:
a =
and
'xy
by
ni=1
ni=,
Si =
yy
n-2
1—
xv
(8)
The two-sided confidence interval yc for the
predicted concentration y at point x is given
by
yc = y±
, with f = n-2 (9)
The two-sided tolerance interval yt for the
regression line is given by
yT=y±kT-sL
At the point x with kT=un' vf and f=n -,
where
n-(x-x)2
(10)
(ID
1 +
The tolerance factor un' for 75 percent of the
population is given in Table I as a function
of n'. The factor vr as a function of f is also
given in Table I as well as the t-factor at the
95 percent confidence level.
The correlation coefficient r may be
calculated from
(12)
TABLE I.—FACTORS FOR CALCULATION
OF CONFIDENCE AND TOLERANCE IN-
TERVALS
(3)
(4)
= y-a-x (5)
The mean values of the data sets are given
(6)
f
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
tf
2.365
2.306
2.262
2.228
2.201
2.179
2.160
2.145
2.131
2.120
2.110
2.101
2.093
2.086
2.080
2.074
2.069
2.064
2.060
Vf
1.7972
1.7110
1.6452
1 .5931
1.5506
1.5153
1.4854
1.4597
1.4373
1.4176
1.4001
1 .3845
1.3704
1 .3576
1.3460
1.3353
1.3255
1.3165
1.3081
n'
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
un'
(75)
1.233
1.223
1.214
1.208
1.203
1.199
1.195
1.192
1.189
1.187
1.185
1.183
1.181
1.179
1.178
1.177
1.175
1.174
1.173
Where xs and yi are the absolute values of the
individual measurements and n is the
number of data points. The values Sxx, Syy,
and SXy are given by
8.2 Quadratic Calibration Relation. In
some cases a quadratic regression will
provide a better fit to the calibration data
than a linear regression. If a quadratic
regression is used to determine a calibration
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17505
relation, a test to determine if the quadratic 8.2.1 Quadratic Regression. A least- ~ _ , , ,2
regression gives a better fit to the data than squares quadratic regression gives the best fit y = « + bix + b2x 03)
a linear regression must be performed, and coefficients b0, bi, and b2 for the calibration The coefficients bo, bi, and b2 are detennined
the relation with the best fit must be used. relation: from the solution to the matrix equation
Ab=B where:
A =
n S] S2
S, S2 S3
S2 S3 S4_
b =
V
b.
t>2
B =
ss"
S6
S7
and
(14)
The solutions to bo, bi, and b2 are:
Where:
bo = (Ss -S2 -S4 +S, .S3 -S7 +S2 -S6 -S3 -S7 -S2 -S2 -S3 -S3 -S5 -S4 -S6 -S,)/ del A (15)
bj = (n-S6 -S4 +S5 -S3 -S2 +S2 -Sj -S7 -S2 -S6 -S2 -S7 -S3 -n-S4 -S, -S5)/detA (16)
b2 = (n-S2 -S7 +S, -S6 -S2 +S5 -S, -S3 -S2 -S2 -S5 -S3 -S6 -n-S7 -S, -Sj/detA (17)
det A = n-S2 -S4 +8, -S3 -S2 +S2 -Sj -S3 -S2 -S2 -S2 -S3 -S3 -S2 -S4 -S, •S1 (18)
8.2.2 Confidence Interval. For any f = n - 3,
positive value of x, the confidence interval is
given by: tf is given in Table I,
yc, =y±trsQVA
Where:
(19)
SQ=.
n-3 i=
i-yi) -and (20)
A = C0 +2C,x + (2C2 +C3)x2 +2C4x3 +C5x4. (21)
The C coefficients are given below:
_ S2-S4-S3 _S3-S2-Sj-S4 S!-S3-S2
^0 ' Ul ' » *-2 =
D D D
_nS4-S^ _S,.S2-nS3 _nS2-Sf
\_*Q — j ^^"4 ~~ " • ^-"" ~~ —
O
D
D
(22)
Where:
-------
17506
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
D = n(s2.S4-S^)+S1(S3.S2-S1.S4)+S2(si.S3-S^). (23)
8.2.3 Tolerance Interval. For any positive
value of x, the tolerance interval is given by:
yTI=y±kT.sQ, (24)
Where:
kT = u , -vf withf = n-3, and (25)
i n *
n' = l/Awithn'>2.
(26)
The Vf and un, factors can also be found in
Table I.
8.3 Test to Determine Best Regression Fit.
The test to determine if the fit using a
quadratic regression is better than the fit
using a linear regression is based on the
values of s calculated in the two
formulations. If SL denotes the value of s from
the linear regression and SQ the value of s
from the quadratic regression, then the
quadratic regression gives a better fit at the
95 percent confidence level if the following
relationship is fulfilled:
(n-2).s2L-(n-3)4
>Ff
(27)
With f = n-3 and the value of Ff at the 95
percent confidence level as a function of f
taken from Table II below.
TABLE II.—VALUES FOR Ff
f
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Ft
161.4
18.51
10.13
7.71
6.61
5.99
5.59
5.32
5.12
4.96
4.84
4.75
4.67
4.60
4.54
/
16
17
18
19
20
22
24
26
28
30
40
50
60
80
100
F*
4.49
4.45
4.41
4.38
4.35
4.30
4.26
4.23
4.20
4.17
4.08
4.03
4.00
3.96
3.94
9. Reporting
At a minimum (check with the appropriate
regional office, or State, or local agency for
additional requirements, if any) summarize
in tabular form the results of the CD tests and
the GEMS response calibration. Include all
data sheets, calculations, and records of
GEMS response necessary to substantiate that
the performance of the GEMS met the
performance specifications.
The GEMS measurements shall be
reported to the agency in units of mg/
m3 on a dry basis, corrected to 20°C and
7 percent Oa-
10. Bibliography
1. 40 CFR part 60, Appendix B,
"Performance Specification 2—Specifications
and Test Procedures for SOz and NOX
Continuous Emission Monitoring Systems in
Stationary Sources."
2. 40 CFR part 60, Appendix B,
"Performance Specification 1—Specification
and Test Procedures for Opacity Continuous
Emission Monitoring Systems in Stationary
Sources."
3. 40 CFR part 60, Appendix A, "Method
1—Sample and Velocity Traverses for
Stationary Sources."
4. 40 CFR part 266, Appendix IX, Section
2, "Performance Specifications for
Continuous Emission Monitoring Systems."
5. ISO 10155, "Stationary Source
Emissions—Automated Monitoring of Mass
Concentrations of Particles: Performance
Characteristics, Test Procedures, and
Specifications," available from ANSI.
6. G. Box, W. Hunter, J. Hunter, Statistics
for Experimenters (Wiley, New York, 1978).
7. M. Spiegel, Mathematical Handbook of
Formulas and Tables (McGraw-Hill, New
York, 1968).
Performance Specification 12—
Specifications and test procedures for total
mercury continuous monitoring systems in
stationary sources.
1. Applicability and Principle
1.1 Applicability. This specification is to
be used for evaluating the acceptability of
total mercury continuous emission
monitoring systems (GEMS) at the time of or
soon after installation and whenever
specified in the regulations. The GEMS must
be capable of measuring the total
concentration (regardless of speciation) of
both vapor and solid phase mercury. The
GEMS may include, for certain stationary
sources, (a) a diluent (Oa) monitor (which
must meet its own performance
specifications: 40 CFR part 60, Appendix B,
Performance Specification 3), (b) flow
monitoring equipment to allow measurement
of the dry volume of stack effluent sampled,
and (c) an automatic sampling system.
This specification is not designed to
evaluate the installed GEMS' performance
over an extended period of time nor does it
identify specific calibration techniques and
auxiliary procedures to assess the GEMS'
performance. The source owner or operator,
however, is responsible to properly calibrate,
maintain, and operate the GEMS. To evaluate
the GEMS' performance, the Administrator
may require, under Section 114 of the Act,
the operator to conduct GEMS performance
evaluations at other times besides the initial
test.
1.2 Principle. Installation and
measurement location specifications,
performance specifications, test procedures,
and data reduction procedures are included
in this specification. Reference method tests,
calibration error tests, and calibration drift
tests, and interferant tests are conducted to
determine conformance of the GEMS with the
specification. Calibration error is assessed
with standards for elemental mercury (Hg(0))
and mercuric chloride (HgCk). The ability of
the GEMS to provide a measure of total
mercury (regardless of speciation and phase)
at the facility at which it is installed is
demonstrated by comparison to manual
reference method measurements.
2. Definitions
2.1 Continuous Emission Monitoring
System (CEMS). The total equipment
required for the determination of a pollutant
concentration. The system consists of the
following major subsystems:
2.1.1 Sample Interface. That portion of
the CEMS used for one or more of the
following: sample acquisition, sample
transport, and sample conditioning, or
protection of the monitor from the effects of
the stack effluent.
2.1.2 Pollutant Analyzer. That portion of
the CEMS that senses the pollutant
concentration(s) and generates a proportional
output.
2.1.3 Diluent Analyzer (if applicable).
That portion of the CEMS that senses the
diluent gas (O2) and generates an output
proportional to the gas concentration.
2.1.4 Data Recorder. That portion of the
CEMS that provides a permanent record of
the analyzer output. The data recorder may
provide automatic data reduction and CEMS
control capabilities.
2.2 Point CEMS. A CEMS that measures
the pollutant concentrations either at a single
point or along a path equal to or less than
10 percent of the equivalent diameter of the
stack or duct cross section.
2.3 Path CEMS. A CEMS that measures
the pollutant concentrations along a path
greater than 10 percent of the equivalent
diameter of the stack or duct cross section.
2.4 Span Value. The upper limit of a
pollutant concentration measurement range
defined as twenty times the applicable
emission limit. The span value shall be
documented by the CEMS manufacturer with
laboratory data.
2.5 Relative Accuracy (RA). The absolute
mean difference between the pollutant
concentration(s) determined by the CEMS
and the value determined by the reference
method (RM) plus the 2.5 percent error
confidence coefficient of a series of tests
divided by the mean of the RM tests or the
applicable emission limit.
2.6 Calibration Drift (CD). The difference
in the CEMS output readings from the
established reference value after a stated
period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.7 Zero Drift (ZD). The difference in the
CEMS output readings for zero input after a
stated period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.8 Representative Results. Defined by
the RA test procedure defined in this
specification.
2.9 Response Time. The time interval
between the start of a step change in the
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17507
system input and the time when the
pollutant analyzer output reaches 95 percent
of the final value.
2.10 Controidal Area. A concentric area
that is geometrically similar to the stack or
duct cross section and is no greater than 1
percent of the stack or duct cross sectional
area.
2.11 Batch Sampling. Batch sampling
refers to the technique of sampling the stack
effluent continuously and concentrating the
pollutant in some capture medium. Analysis
is performed periodically after sufficient time
has elapsed to concentrate the pollutant to
levels detectable by the analyzer.
2.12 Calibration Standard. Calibration
standards consist of a known amount of
pollutant that is presented to the pollutant
analyzer portion of the GEMS in order to
calibrate the drift or response of the analyzer.
The calibration standard may be, for
example, a solution containing a known
concentration, or a filter with a known mass
loading or composition.
2.13 Calibration Error (CE). The
difference between the concentration
indicated by the GEMS and the known
concentration generated by a calibration
source when the entire GEMS, including the
sampling interface) is challenged. A CE test
procedure is performed to document the
accuracy and linearity of the GEMS over the
entire measurement range.
3. Installation and Measurement Location
Specifications
3.1 The GEMS Installation and
measurement location. Install the GEMS at an
accessible location downstream of all
pollution control equipment where the
mercury concentration measurements are
directly representative or can be corrected so
as to be representative of the total emissions
from the affected facility. Then select
representative measurement points or paths
for monitoring in locations that the GEMS
will pass the RA test (see Section 7). If the
cause of failure to meet the RA test is
determined to be the measurement location
and a satisfactory correction technique
cannot be established, the Administrator may
require the CEMS to be relocated.
Measurement locations and points or paths
that are most likely to provide data that will
meet the RA requirements are listed below.
3.1.1 Measurement Location. The
measurement location should be (1) at least
eight equivalent diameters downstream of the
nearest control device, point of pollutant
generation, bend, or other point at which a
change of pollutant concentration or flow
disturbance may occur and (2) at least two
equivalent diameters upstream from the
effluent exhaust. The equivalent duct
diameter is calculated as per 40 CFR part 60,
Appendix A, Method 1, Section 2.1.
3.1.2 Point CEMS. The measurement
point should be (1) no less than 1.0 meter
from the stack or duct wall or (2) within or
centrally located over the centroidal area of
the stack or duct cross section. Selection of
traverse points to determine the
representativeness of the measurement
location should be made according to 40 CFR
part 60, Appendix A, Method 1, Section 2.2
and 2.3.
3.1.3 Path CEMS. The effective
measurement path should be (1) totally
within the inner area bounded by a line 1.0
meter from the stack or duct wall, or (2) have
at least 70 percent of the path within the
inner 50 percent of the stack or duct cross
sectional area, or (3) be centrally located over
any part of the centroidal area.
3.2 Reference Method (RM) Measurement
Location and Traverse Points. The RM
measurement location should be (1) at least
eight equivalent diameters downstream of the
nearest control device, point of pollutant
generation, bend, or other point at which a
change of pollutant concentration or flow
disturbance may occur and (2) at least two
equivalent diameters upstream from the
effluent exhaust. The RM and CEMS
locations need not be the same, however the
difference may contribute to failure of the
CEMS to pass the RA test, thus they should
be as close as possible without causing
interference with one another. The
equivalent duct diameter is calculated as per
40 CFR part 60, Appendix A, Method 1,
Section 2.1. Selection of traverse
measurement point locations should be made
according to 40 CFR part 60, Appendix A,
Method 1, Sections 2.2 and 2.3. If the RM
traverse line interferes with or is interfered
by the CEMS measurements, the line may be
displaced up to 30 cm (or 5 percent of the
equivalent diameter of the cross section,
whichever is less) from the centroidal area.
4. Performance and Equipment
Specifications
4.1 Data Recorder Scale. The CEMS data
recorder response range must include zero
and a high level value. The high level value
must be equal to the span value. If a lower
high level value is used, the CEMS must have
the capability of providing multiple outputs
with different high level values (one of which
is equal to the span value) or be capable of
automatically changing the high level value
as required (up to the span value) such that
the measured value does not exceed 95
percent of the high level value.
4.2 Relative Accuracy (RA). The RA of
the CEMS must be no greater than 20 percent
of the mean value of the RM test data in
terms of units of the emission standard, or 10
percent of the applicable standard,
whichever is greater.
4.3 Calibration Error. Calibration error is
assessed using standards for Hg(0) and HgCl2.
The mean difference between the indicated
CEMS concentration and the reference
concentration value for each standard at all
three test levels listed below shall be no
greater than ±15 percent of the reference
concentration at each level.
4.3.1 Zero Level. Zero to twenty (0-20)
percent of the emission limit.
4.3.2 Mid-Level. Forty to sixty (40-60)
percent of the emission limit.
4.3.3 High-Level. Eighty to one-hundred
and twenty (80-120) percent of the emission
limit.
4.4 Calibration Drift. The CEMS design
must allow the determination of calibration
drift of the pollutant analyzer at
concentration levels commensurate with the
applicable emission standard. The CEMS
calibration may not drift or deviate from the
reference value (RV) of the calibration
standard by more than 10 percent of the
emission limit. The calibration shall be
performed at a level equal to 80 to 120
percent of the applicable emission standard.
Calibration drift shall be evaluated for
elemental mercury only.
4.5 Zero Drift. The CEMS design must
allow the determination of calibration drift at
the zero level (zero drift). The CEMS zero
point shall not drift by more than 5 percent
of the emission standard.
4.6 Sampling and Response Time. The
CEMS shall sample the stack effluent
continuously. Averaging time, the number of
measurements in an average, and the
averaging procedure for reporting and
determining compliance shall conform with
that specified in the applicable emission
regulation.
4.6.1 Response Time. The response time
of the CEMS should not exceed 2 minutes to
achieve 95 percent of the final stable value.
The response time shall be documented by
the CEMS manufacturer.
4.6.2 Waiver from Response Time
Requirement. A source owner or operator
may receive a waiver from the response time
requirement for instantaneous, continuous
CEMS in section 4.5.1 from the Agency if no
GEM is available which can meet this
specification at the time of purchase of the
CEMS.
4.6.3 Response Time for Batch CEMS.
The response time requirement of Section
4.5.1 does not apply to batch CEMS. Instead
it is required that the sampling time be no
longer than one third of the averaging period
for the applicable standard. In addition, the
delay between the end of the sampling time
and reporting of the sample analysis shall be
no greater than one hour. Sampling is also
required to be continuous except in that the
pause in sampling when the sample
collection media are changed should be no
greater than five percent of the averaging
period or five minutes, whichever is less.
4.7 CEMS Interference Response. While
the CEMS is measuring the concentration of
mercury in the high-level calibration sources
used to conduct the CE test the gaseous
components (in nitrogen) listed in Table I
shall be introduced into the measurement
system either separately or in combination.
The interference test gases must be
introduced in such a way as to cause no
change in the mercury or mercuric chloride
calibration concentration being delivered to
the CEMS. The concentrations listed in the
table are the target levels at the sampling
interface of the CEMS based on the known
cylinder gas concentrations and the extent of
dilution (see Section 9). Interference is
defined as the difference between the CEMS
response with these components present and
absent. The sum of the interferences must be
less than 10 percent of the emission limit
value. If this level of interference is
exceeded, then corrective action to eliminate
the interference(s) must be taken.
-------
17508
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
TABLE I.—INTERFERENCE TEST GAS
CONCENTRATIONS IN NITROGEN
Gas
Carbon Monoxide
Carbon Dioxide
Oxygen
Sulfur Dioxide
Nitrogen Dioxide
Water Vapor
Hydrogen Chloride (HCI)
Chlorine (CI2)
Concentration
500±50 ppm.
10±1 percent.
20.9±1 percent.
500±50 ppm.
250±25 ppm.
25±5 percent.
50±5 ppm.
1 Cttl ppm.
4.8 Calibration Source Requirements for
Assessment of Calibration Error. The
calibration source must permit the
introduction of known (NIST traceable) and
repeatable concentrations of elemental
mercury (Hg(0)) and mercuric chloride
(HgCl2) into the sampling system of the
GEMS. The GEMS manufacturer shall
document the performance of the calibration
source, and submit this documentation and
a calibration protocol to the administrator for
approval. Determination of GEMS calibration
error must then be made in using the
approved calibration source and in
accordance with the approved protocol.
4.8.1 Design Considerations. The
calibration source must be designed so that
the flowrate of calibration gas introduced to
the CEMS is the same at all three calibration
levels specified in Section 4.3 and at all
times exceeds the flow requirements of the
CEMS.
4.8.2 Calibration Precision. A series of
three injections of the same calibration gas,
at any dilution, shall produce results which
do not vary by more than ±5 percent from the
mean of the three injections. Failure to attain
this level of precision is an indication of a
problem in the calibration system or the
CEMS. Any such problem must be identified
and corrected before proceeding.
5. Performance Specification Test Procedure
5.1 Pretest Preparation. Install the CEMS
and prepare the RM test site according to the
specifications in Section 3, and prepare the
CEMS for operation according to the
manufacturer's written instructions.
5.2 Calibration and Zero Drift Test
Period. While the affected facility is
operating at more than 50 percent of normal
load, or as specified in an applicable subpart,
determine the magnitude of the calibration
drift (CD) and zero drift (ZD) once each day
(at 24-hour intervals) for 7 consecutive days
according to the procedure given in Section
6. To meet the requirements of Sections 4.4
and 4.5 none of the CD's or ZD's may exceed
the specification. All CD determinations
must be made following a 24-hour period
during which no unscheduled maintenance,
repair, or manual adjustment of the CEMS
took place.
5.3 CE Test Period. Conduct a CE test
prior to the CD test period. Conduct the CE
test according to the procedure given in
Section 8.
5.4 CEMS Interference Response Test
Period. Conduct an interference response test
in conjunction with the CE test according to
the procedure given in Section 9.
5.5 RA Test Period. Conduct a RA test
following the CD test period. Conduct the RA
test according to the procedure given in
Section 7 while the affected facility is
operating at more than 50 percent of normal
load, or as specified in the applicable
subpart.
6.0 The CEMS Calibration and Zero Drift
Test Procedure
This performance specification is designed
to allow calibration of the CEMS by use of
standard solutions, filters, etc. that challenge
the pollutant analyzer part of the CEMS (and
as much of the whole system as possible), but
which do not challenge the entire CEMS,
including the sampling interface. Satisfactory
response of the entire system is covered by
the RA and CE requirements.
The CD measurement is to verify the ability
of the CEMS to conform to the established
CEMS calibration used for determining the
emission concentration. Therefore, if
periodic automatic or manual adjustments
are made to the CEMS zero and calibration
settings, conduct the CD test immediately
before the adjustments, or conduct it in such
a way that the CD and ZD can be determined.
Conduct the CD and ZD tests at the points
specified in Sections 4.4 and 4.5. Record the
CEMS response and calculate the CD
according to:
Rv
Where CD denotes the calibration drift of the
CEMS in percent, RCEM is the CEMS
response, and Rv is the reference value of the
high level calibration standard. Calculate the
ZD according to:
(2)
R
EM
Where ZD denotes the zero drift of the CEMS
in percent, RCEM is the CEMS response, Rv
is the reference value of the low level
calibration standard, and REM is the emission
limit value.
7. Relative Accuracy Test Procedure
7.1 Sampling Strategy for RA Tests. The
RA tests are to verify the initial performance
of the entire CEMS system, including the
sampling interface, by comparison to RM
measurements. Conduct the RM
measurements in such a way that they will
yield results representative of the emissions
from the source and can be correlated to the
CEMS data. Although it is preferable to
conduct the diluent (if applicable), moisture
(if needed), and pollutant measurements
simultaneously, the diluent and moisture
measurements that are taken within a 30 to
60-minute period, which includes the
pollutant measurements, may be used to
calculate dry pollutant concentration.
A measure of relative accuracy at a single
level that is detectable by both the CEMS and
the RM is required.
In order to correlate the CEMS and RM
data properly, note the beginning and end of
each RM test period of each run (including
the exact time of day) in the CEMS data log.
7.2 Correlation of RM and CEMS Data.
Correlate the CEMS and RM test data as to
the time and duration by first determining
from the CEMS final output (the one used for
reporting) the integrated average pollutant
concentration for each RM test period.
Consider system response time, if important,
and confirm that the pair of results are on a
consistent moisture, temperature, and diluent
concentration basis. Then compare each
integrated CEMS value against the
corresponding average RM value.
7.3 Number of tests. Obtain a minimum
of three pairs of CEMS and RM
measurements. If more than nine pairs of
measurements are obtained, then up to three
pairs of measurements may be rejected so
long as the total number of measurement
pairs used to determine the RA is greater
than or equal to nine. However, all data,
including the rejected data, must be reported.
7.4 Reference Methods. Unless otherwise
specified in an applicable subpart of the
regulations, Method 3B, or its approved
alternative, is the reference method for
diluent (O2) concentration. Unless otherwise
specified in an applicable subpart of the
regulations, the manual method for multi-
metals in 40 CFR part 266, Appendix IX,
Section 3.1 (until superseded by SW-846), or
its approved alternative, is the reference
method for mercury.
7.5 Calculations. Summarize the results
on a data sheet. An example is shown in
Figure 2-2 of 40 CFR part 60, Appendix B,
Performance Specification 2. Calculate the
mean of the RM values. Calculate the
arithmetic differences between the RM and
CEMS output sets, and then calculate the
mean of the differences. Calculate the
standard deviation of each data set and
CEMS RA using the equations in Section 10.
8. Calibration Error Test Procedure
8.1 Sampling Strategy. The CEMS
calibration error shall be assessed using
calibration sources of elemental mercury and
mercuric chloride in turn (see Section 4.8 for
calibration source requirements). Challenge
the CEMS at the measurement levels
specified in Section 4.3. During the test,
operate the CEMS as nearly as possible in its
normal operating mode. The calibration gases
should be injected into the sampling system
as close to the sampling probe outlet as
practical and shall pass through all filters,
scrubbers, conditioners, and other monitor
components used during normal sampling.
8.2 Number of tests. Challenge the CEMS
three non-consecutive times at each
measurement point and record the responses.
The duration of each challenge should be for
a sufficient period of time to ensure that the
CEMS surfaces are conditioned and a stable
output obtained.
8.3 Calculations. Summarize the results
on a data sheet. Calculate the mean
difference between the CEMS response and
the known reference concentration at each
measurement point according to equations 5
and 6 of Section 10. The calibration error
(CE) at each measurement point is then given
by
CE=|d/Rv|xlOO,
(3)
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17509
Whore Ry is the reference concentration
value.
9. Interference Response Test Procedure
9.1 Tost Strategy. Perform the
interference response test while the GEMS is
being challenged by the high level calibration
source for mercury (after the CE
determination has been made), and again
while the GEMS is being challenged by the
high level calibration source for mercuric
chloride (after the CE determination has been
made). The interference test gases should be
injected into the sampling system as close to
the sampling probe outlet as practical and
shall pass through all filters, scrubbers,
conditioners, and other monitor components
used during normal sampling.
9.2 Number of tests. Introduce the
interference test gas three times alternately
with the high-level calibration gas and record
the responses both with and without the
interference test gas. The duration of each
tost should be for a sufficient period of time
to ensure that the GEMS surfaces are
conditioned and a stable output obtained.
9.3 Calculations. Summarize the results
on a data sheet. Calculate the mean
difference between the GEMS response with
and without the interference test gas by
taking the average of the GEMS responses
with and without the interference test gas
(see equation 5) and then taking the
difference (d). The percent interference (I) is
then given by:
I = |d/RHL|xlOO, (4)
Where RHL is the value of the high-level
calibration standard. If the gaseous
components of the interference test gas are
introduced separately, then the total
interference is the sum of the individual
interferences.
10. Equations
10.1 Arithmetic Mean. Calculate the
arithmetic mean of a data set as follows:
*--£*,.
n M
(5)
Where n is equal to the number of data
points.
10.1.1 Calculate the arithmetic mean of
the difference, d, of a data set, using Equation
5 and substituting d for x. Then
(6)
Where x and y are paired data points from
the GEMS and RM, respectively.
10.2 Standard Deviation. Calculate the
standard deviation (SD) of a data set as
follows:
n-l
(7)
10.3 Relative Accuracy (RA). Calculate
the RA as follows:
RA = -
R
(8)
RM
Where d is equal to the arithmetic mean of
the difference, d, of the paired GEMS and RM
data set, calculated according to Equations 5
and 6, SD is the standard deviation
calculated according to Equation 7, RRM is
equal to either the average of the RM data set,
calculated according to Equation 5, or the
value of the emission standard, as applicable
(see Section 4.2), and to.9?s is the t-value at
2.5 percent error confidence, see Table II.
TABLE II
[t-Values]
n»
2
3
4
5
6
tb.975
12.706
4.303
3.182
2.776
2.571
na
7
8
9
10
11
to.975
2.447
2.365
2.306
2.262
2.228
na
12
13
14
15
16
to.975
2.201
2.179
2.160
2.145
2.131
"The values in this table are already corrected for n-1 degrees of freedom. Use n equal to the number of individual values.
11. Reporting
At a minimum (check with the appropriate
regional office, or State, or local agency for
additional requirements, if any) summarize
in tabular form the results of the CE,
interference response, CD and RA tests.
Include all data sheets, calculations, and
records of GEMS response necessary to
substantiate that the performance of the
GEMS met the performance specifications.
The GEMS measurements shall be reported
to the agency in units of ng/m3 on a dry basis,
corrected to 20 °C and 7 percent 62.
12. Bibliography
1.40 CFR Part 60, Appendix B,
"Performance Specification 2—Specifications
and Test Procedures for SO2 and NOX
Continuous Emission Monitoring Systems in
Stationary Sources."
2.40 CFR Part 60, Appendix B,
"Performance Specification 1—Specification
and Test Procedures for Opacity Continuous
Emission Monitoring Systems in Stationary
Sources."
3. 40 CFR Part 60, Appendix A, "Method
1—Sample and Velocity Traverses for
Stationary Sources."
4. 40 CFR Part 266, Appendix IX, Section
2, "Performance Specifications for
Continuous Emission Monitoring Systems."
5. Draft Method 29, "Determination of
Metals Emissions from Stationary Sources,"
Docket A-90-45, Item II-B-12, and EMTIC
CTM-012.WPF.
6. "Continuous Emission Monitoring
Technology Survey for Incinerators, Boilers,
and Industrial Furnaces: Final Report for
Metals CEM's," prepared for the Office of
Solid Waste, U.S. EPA, Contract No. 68-D2-
0164 (4/25/94).
7. 40 CFR Part 60, Appendix A, Method 16,
"Semicontinuous Determination of Sulfur
Emissions from Stationary Sources."
8. 40 CFR Part 266, Appendix IX,
Performance Specification 2.2, "Performance
Specifications for Continuous Emission
Monitoring of Hydrocarbons for Incinerators,
Boilers, and Industrial Furnaces Burning
Hazardous Waste."
Performance Specification 13—
Specifications and test procedures for
hydrochloric acid continuous monitoring
systems in stationary sources
1. Applicability and Principle
1.1 Applicability. This specification is to
be used for evaluating the acceptability of
hydrogen chloride (HC1) continuous emission
monitoring systems (GEMS) at the time of or
soon after installation and whenever
specified in the pertinent regulations. Some
source specific regulations require the
simultaneous operation of diluent monitors.
These may be Oi or CO2 monitors.
This specification does not evaluate the
performance of installed GEMS over
extended periods of time. The specification
does not identify specific calibration
techniques or other auxiliary procedures that
will assess the GEMS performance. Section
114 of the Act authorizes the administrator
to require the operator of the GEMS to
conduct performance evaluations at times
other than immediately following the initial
installation.
This specification is only applicable to
monitors that unequivocally measure the
concentration of HC1 in the gas phase. It is
not applicable to GEMS that do not measure
gas phase HC1, per se, or GEMS that may
have significant interferences. The
Administrator believes that HC1 GEMS must
measure the concentration of gaseous HC1
thereby eliminating interferences from
volatile inorganic and/or organic chlorinated
compounds. GEMS that are based upon
infrared measurement techniques, non-
dispersive infrared (NDIR), gas filter
correlation infrared (GFC-IR) and Fourier
Transform infrared (FTIR) are examples of
acceptable measurement techniques. Other
measurement techniques that unequivocally
-------
17510
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
measure the concentration of HC1 in the gas
phase may also be acceptable.
1.2 Principle. This specification includes
installation and measurement location
specifications, performance and equipment
specifications, test procedures, and data
reduction procedures. This specification also
provides definitions of acceptable
performance.
This specification stipulates that audit gas
tests and calibration drift tests be used to
assess the performance of the GEMS. The
determination of the accuracy with which the
GEMS measures HC1 is measured by
challenging the GEMS with audit gas of
known concentration. There is no absolute
determination of interference with the
measurement of gas phase HC1 with other
constituents in the stack gases.
2. Definitions
2.1 Continuous Emission Monitoring
System. The total equipment required for the
determination of the concentration of a gas or
its emission rate. The GEMS consist of the
following subsystems:
2.1.1 Sample Interface. That portion of
the GEMS used for one or more of the
following: sample acquisition, sample
transportation, sample conditioning, and
protection of the monitor from the effects of
the stack effluent.
2.1.2 Pollutant Analyzer. That portion of
the GEMS that senses the pollutant gas and
generates an output that is proportional to
the gas concentration.
2.1.3 Diluent Analyzer. That portion of
the GEMS that senses the concentration of
the diluent gas (e.g., CO2 or 02) and generates
an output that is proportional to the
concentration of the diluent.
2.1.4 Data Recorder. That portion of the
GEMS that provides a permanent record of
the analyzer output. The data recorder may
include automatic data reduction
capabilities.
2.2 Point GEMS. A GEMS that measures
the gas concentration either at a single point
or along a path equal to or less than 10
percent of the equivalent diameter of the
stack or duct cross section. The equivalent
diameter must be determined as specified in
Appendix A, Method 1 of this Part.
2.3 Path GEMS. A GEMS that measures
the gas concentration along a path greater
than 10 percent of the equivalent diameter
(Appendix A, Method 1) of the stack of duct
cross section.
2.4 Span Value. The upper limit of a gas
concentration measurement range specified
for affected source categories in the
applicable subpart of the regulations. The
span value shall be documented by the GEMS
manufacturer with laboratory data.
2.5 Accuracy. A measurement of
agreement between a measured value and an
accepted or true value, expressed as the
percentage difference between the true and
measured values relative to the true value.
For these performance specifications,
accuracy is checked by conducting a
calibration error (CE) test.
2.6 Calibration Error (CE). The difference
between the concentration indicated by the
GEMS and the known concentration of, the
cylinder gas, A CE test procedure is .
performed to document the accuracy and
linearity of the monitoring equipment over
the entire measurement range.
2.7 Calibration Drift. (CD). The difference
between the GEMS output and the
concentration of the calibration gas after a
stated period of operation during which no
unscheduled maintenance, repair, or
adjustment-took place.
2.8 Centroidal Area. A concentric area
that is geometrically similar to the stack or
duct cross section is no greater than 1 percent
of the stack or duct cross-sectional area.
2.9 Representative Results. Defined by
the RM test procedure outlined in this
specification.
3. Installation and Measurement Location
Specifications
3.1 GEMS Installation and Measurement
Locations. The GEMS shall be installed in a
location in which measurements
representative of the source's emissions can
be obtained. The optimum location of the
sample interface for the GEMS is determined
by a number of factors, including ease of
access for calibration and maintenance, the
degree to which sample conditioning will be
required, the degree to which it represents
total emissions, and the degree to which it
represents the combustion situation in the
firebox. The location should be as free from
in-leakage influences as possible and
reasonably free from severe flow
disturbances. The sample location should be
at least two equivalent duct diameters
downstream from the nearest control device,
point of pollutant generation, or other point
at which a change in the pollutant
concentration or emission rate occurs and at
least 0.5 diameter upstream from the exhaust
or control device. The equivalent duct
diameter is calculated as per 40 CFR part 60,
appendix A, method 1, section 2.1. If these
criteria are not achievable or if the location
is otherwise less than optimum, the
possibility of stratification should be
investigated as described in section 3.2. The
measurement point shall be within the
centroidal area of the stack or duct cross
section.
3.1.1 Point GEMS. It is suggested that the
measurement point be (1) no less than 1.0
meter from the stack or duct wall or (2)
within or centrally located over the
centroidal area of the stack or duct cross
section. •'
3.1.2 Path GEMS. It is suggested that the
effective measurement path (1) be totally
within the inner area bounded by a line 1.0
meter from the stack or duct wall, or (2) have
at least 70 percent of the path within the
inner 50 percent of the stack or duct cross-
sectional area.
3.2 Stratification Test Procedure.
Stratification is defined as a difference in
excess of 10 percent between the average
concentration in the duct or stack and the
concentration at any point more than 1.0
meter from the duct or stack wall. To
determine whether effluent stratification
exists, a dual probe system should be used
to determine the average effluent
concentration while measurements at each
traverse point are being made. One probe,
located, at the stack or duct centroid, is used
as a stationary reference point to indicate the
change in effluent concentration over time.
The second probe is used for sampling at the
traverse points specified in 40 CFR part 60
appendix A, method 1. The monitoring
system samples sequentially at the reference
and traverse points throughout the testing
period for five minutes at each point.
4. Performance and Equipment
Specifications
4.1 Data Recorder Scale. The GEMS data
recorder response range must include zero
and a high-level value. The high-level value
is chosen by the source owner or operator
and is defined as follows:
For a GEMS intended to measure an
uncontrolled emission (e.g., at the inlet of a
scrubber) the high-level value must be
between 1.25 and 2.0 times the average
potential emission concentration, unless
another value is specified in an applicable
subpart of the regulations. For a GEMS
installed to measure controlled emissions or
emissions that are in compliance with an
applicable regulation, the high-level value
must be between 1.5 times the HC1
concentration corresponding to the emission
standard level and the span value. If a lower
high-level value is used, the operator must
have the capability of requirements of the
applicable regulations.
The data recorder output must be
established so that the high-level value is
read between 90 and 100 percent of the data
recorder full scale. (This scale requirement
may not be applicable to digital data
recorders.) The calibration gas, optical filter
or cell values used to establish the data
recorder scale should produce the zero and
high-level values. Alternatively, a calibration
gas, optical filter, or cell value between 50
and 100 percent of the high-level value may
be used in place of the high-level value,
provided that the data recorder full-scale
requirements as described above are met.
The GEMS design must also allow the
determination of calibration drift at the zero
and high-level values. If this is not possible
or practicable, the design must allow these
determinations to be conducted at a low-level
value (zero to 20 percent of the high-level
value) and at a value between 50 and 100
percent of the high-level value.
4.2 Calibration Drift. The GEMS
calibration must not drift or deviate from the
reference value of the gas cylinder, gas cell,
or optical filter by more than 2.5 percent of
the span value. If the span value of the GEMS
is 20 ppm or less then the calibration drift
must be less than 0.5 parts per million, for
6 out of 7 test days.
If the GEMS includes both HC1 and diluent
monitors, the calibration drift must be
determined separately for each in terms of
concentrations (see PS 3 for the diluent
specifications).
4.3 Calibration Error (CE). Calibration
error is assessed using EPA protocol 1
cyinder gases for HC1. The mean difference
between the indicated GEMS concentration
and the reference concentration value for
each standard at all three test levels indicated
below shall be no greater than 15 percent of
the reference concentration at each level.
4.3.1 Zero Level. Zero to twenty (0-20)
percent of the emission limit.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17511
4.3.2 Mid Level. Forty to sixty (40-60)
percent of the emission limit.
4.3.3 High Level. Eighty to one-hundred
and twenty (80-120) percent of the emission
limit.
4.4 GEMS Interference Response Test.
Introduce the gaseous components listed in
Table PS HC1—1 into the measurement system
of the GEMS, while the measurement system
is measuring the concentration of HC1 in a
calibration gas. These components may be
introduced separately or as gas mixtures.
Adjust the HC1 calibration gas and gaseous
component flow rates so as to maintain a
constant concentration of HC1 in the gas
mixture being introduced into the
measurement system. Record the change in
the measurement system response to the HC1
on a form similar to Figure PS HC1-1. If the
sum of the interferences is greater than 2
percent of the applicable span concentration,
take corrective action to eliminate the
interference.
TABLE PS HCL-1.—INTERFERENCE TEST GASES CONCENTRATIONS
Gas
Concentration
Carbon Monoxide
Carbon Dioxide
Oxygen
Sulfur Dioxide
Water Vapor
Nitrogen Dioxide
500+50 ppm
1 0+1 percent
20 9+1 percent
500+50 ppm
25+5 percent
250±25 ppm
Figure PS HC1-1—Interference Response
Date of Test _
Analyzer Type
Serial Number
HCI—CALIBRATION GAS CONCENTRATION
Test gas
Concentra-
tion
Analyzer re-
sponse
Analyzer
error
Percent of
span
Conduct an interference response test of
•each analyzer prior to its initial use in the
Hold. Thereafter, re-check the measurement
system if changes are made in the
instrumentation that could alter the
interference response, e.g., changes in the
type of gas detector.
4.5 Sampling and Response Time. The
GEMS shall sample the stack effluent
continuously. Averaging time, the number of
measurements in an average, and the
averaging procedure for reporting and
determining compliance shall conform with
that specified in the applicable emission
regulation.
4.5.1 Response Time. The response time
of the GEMS should not exceed 2 minutes to
achieve 95 percent of the final stable value.
The response time shall be documented by
the GEMS manufacturer.
4.5.2 Waiver from Response Time
Requirement. A source owner or operator
may receive a waiver from the response time
requirement for instantaneous, continuous
GEMS in section 4.5.1 from the Agency if no
GEM is available which can meet this
specification at the time of purchase of the
GEMS.
4.5.3 Response Time for Batch GEMS.
The response time requirement of Section
4.5.1 does not apply to batch GEMS. Instead
it is required that the sampling time be no
longer than one third of the averaging period
for the applicable standard. In addition, the
dolay between the end of the sampling time
and reporting of the sample analysis shall be
no greater than one hour. Sampling is also
required to be continuous except in that the
pause in sampling when the sample
collection media are changed should be no
greater than five percent of the averaging
period or five minutes, whichever is less.
5. Performance Specification Test Procedure
5.1 Pretest Preparation. Install the GEMS,
prepare the RM test site according to the
specifications in Section 3, and prepare the
GEMS for operation according to the
manufacturer's written instructions.
5.2 Calibration Drift Test Period. While
the affected facility is operating at more than
50 percent of normal load, or as specified in
an applicable subpart, determine the
magnitude of the calibration drift (CD) once
each day (at 24-hour intervals) for 7
consecutive days, according to the procedure
given in Section 6. The CD may not exceed
the specification given in Section 4.2.
5.3 CE Test Period. Conduct a CE test
prior to the CD test period. Conduct the CE
test according to the procedure given in
section 7.
6. The CEMS Calibration Drift Test Procedure
The CD measurement is to verify the ability
of the CEMS to conform to the established
CEMS calibration used for determining the
emission concentration or emission rate.
Therefore, if periodic automated or manual
adjustments are made to the CEMS zero and
calibration settings, conduct the CD test
immediately before these adjustments, or
conduct it in such a way that the CD can be
determined.
Conduct the CD test at the two points
specified in Section 4.1. Introduce the
reference gases, gas cells or optical filters
(these need not be certified) to the CEMS.
Record the CEMS response and subtract this
value from the reference value (see the
example data sheet in Figure 2-1).
7. Calibration Error Test Procedure
7.1 Sampling Strategy. The CEMS
calibration error shall be assessed using the
calibration source specified in Section 4.3.
Challenge the CEMS at the measurement
levels specified in Section 4.3. During the
test, operate the CEMS as nearly as possible
in its normal operating mode. The calibration
gases should be injected into the sampling
system as close to the sampling probe outlet
as practical and shall pass through all filters,
scrubbers, conditioners, and other monitor
components used during normal sampling.
7.2 Number of tests. Challenge the CEMS
three non-consecutive times at each
measurement point and record the responses.
The duration of each challenge should be for
a sufficient period of time to ensure that the
CEMS surfaces are conditioned and a stable
output obtained.
7.3 Calculations. Summarize the results
on a data sheet. Calculate the mean
difference between the CEMS response and
the known reference concentration at each
measurement point according to equations 1
and 2 of Section 8. The calibration error (CE)
at each measurement point is then given by:
-------
27512 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
= |d/rv|xlOO,
(1)
Where Rv is the reference concentration
value.
8. Equations
8.1 Arithmetic Mean. Calculate the
arithmetic mean of the difference, d, of a data
set as follows:
(2)
Where:
n=number of data points.
Sn _ Algebraic sum of the individual
i ~ differences dj
i=l
When the mean of the differences of pairs of
data is calculated, be sure to correct the data
for moisture, if applicable.
9. Reporting
At a minimum (check with the appropriate
regional office, or State, or local agency for
additional requirements, if any) summarize
in tabular form the results of the CD tests and
the relative accuracy tests or alternative RA
procedure as appropriate. Include all data
sheets, calculations, charts (records of GEMS
responses), cylinder gas concentration
certifications (if applicable), necessary to
substantiate that the performance of the
GEMS met the performance specifications.
Performance Specifications 14—
Specifications and test procedures for
chlorine continuous monitoring systems in
stationary sources.
1. Applicability and Principle
1.1 Applicability. This specification is to
be used for evaluating the acceptability of
chlorine (Ck) continuous emission
monitoring systems (GEMS) at the time of or
soon after installation and whenever
specified in the regulations. This
performance specification applies only to
those GEMS capable of directly measuring
the gas phase concentration of the chlorine
(Ck) molecule. The GEMS may include, for
certain stationary sources, a) a diluent (O2)
monitor (which must meet its own
performance specifications: 40 CFR part 60,
Appendix B, Performance Specification 3), b)
flow monitoring equipment to allow
measurement of the dry volume of stack
effluent sampled, and c) an automatic
sampling system.
This specification is not designed to
evaluate the installed GEMS' performance
over an extended period of time nor does it
identify specific calibration techniques and
auxiliary procedures to assess the GEMS'
performance. The source owner or operator,
however, is responsible to properly calibrate,
maintain, and operate the GEMS. To evaluate
the GEMS' performance, the Administrator
may require, under Section 114 of the Act,
the operator to conduct GEMS performance
evaluations at other times besides the initial
test.
1.2 Principle. Installation and
measurement location specifications,
performance specifications, test procedures,
and data reduction procedures are included
in this specification. Calibration error tests,
and calibration drift tests, and interferant
tests are conducted to determine
conformance of the GEMS with the
specification. Calibration error is assessed
with cylinder gas standards for chlorine. The
ability of the GEMS to provide an accurate
measure of chlorine concentration in the flue
gas of the facility at which it is installed is
demonstrated by comparison to manual
reference method measurements.
2. Definitions
2.1 Continuous Emission Monitoring
System (GEMS). The total equipment
required for the determination of a pollutant
concentration. The system consists of the
following major subsystems:
2.1.1 Sample Interface. That portion of
the GEMS used for one or more of the
following: sample acquisition, sample
transport, and sample conditioning, or
protection of the monitor from the effects of
the stack effluent.
2.1.2 Pollutant Analyzer. That portion of
the GEMS that senses the pollutant
concentration(s) and generates a proportional
output.
2.1.3 Diluent Analyzer (if applicable).
That portion of the GEMS that senses the
diluent gas (02) and generates an output
proportional to the gas concentration.
2.1.4 Data Recorder. That portion of the
GEMS that provides a permanent record of
the analyzer output. The data recorder may
provide automatic data reduction and GEMS
control capabilities.
2.2 Point GEMS. A GEMS that measures
the pollutant concentrations either at a single
point or along a path equal to or less than
10 percent of the equivalent diameter of the
stack or duct cross section.
2.3 Path GEMS. A GEMS that measures
the pollutant concentrations along a path
greater than 10 percent of the equivalent
diameter of the stack or duct cross section.
2.4 Span Value. The upper limit of a
pollutant concentration measurement range '
defined as twenty times the applicable
emission limit. The span value shall be
documented by the GEMS manufacturer with
laboratory data.
2.5 Accuracy. A measurement of
agreement between a measured value and an
accepted or true value, expressed as the
percentage difference between the true and
measured values relative to the true value.
For these performance specifications,
accuracy is checked by conducting a
calibration error (CE) test.
2.6 Calibration Drift (CD). The difference
in the GEMS output readings from the
established reference value after a stated
period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.7 Zero Drift (ZD). The difference in the
GEMS output readings for zero input after a
stated period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.8 Representative Results. Defined by
the RA test procedure defined in this
specification.
2.9 Response Time. The time interval
between the start of a step change in the
system input and the time when the
pollutant analyzer output reaches 95 percent
of the final value.
2.10 Centroidal Area. A concentric area
that is geometrically similar to the stack or
duct cross section and is no greater than 1
percent of the stack or duct cross sectional
area.
2.11 Calibration Standard. Calibration
standards consist of a known amount of
pollutant that is presented to the pollutant
analyzer portion of the GEMS in order to
calibrate the drift or response of the analyzer.
The calibration standard may be, for
example, a gas sample containing known
concentration.
2.12 Calibration Error (CE). The
difference between the concentration
indicated by the GEMS and the known
concentration generated by a calibration
source when the entire GEMS, including the
sampling interface) is challenged. A CE test
procedure is performed to document the
accuracy and linearity of the GEMS over the
entire measurement range.
3. Installation and Measurement Location
Specifications
3.1 GEMS Installation and Measurement
Locations. The GEMS shall be installed in a
location in which measurements
representative of the source's emissions can
be obtained. The optimum location of the
sample interface for the GEMS is determined
by a number of factors, including ease of
access for calibration and maintenance, the
degree to which sample conditioning will be
required, the degree to which it represents
total emissions, and the degree to which it
represents the combustion situation in the
firebox. The location should be as free from
in-leakage influences as possible and
reasonably free from severe flow
disturbances. The sample location should be
at least two equivalent duct diameters
downstream from the nearest control device,
point of pollutant generation, or other point
at which a change in the pollutant
concentration or emission rate occurs and at
least 0.5 diameter upstream from the exhaust
or control device. The equivalent duct
diameter is calculated as per 40 CFR part 60,
appendix A, method 1, section 2.1. If these
criteria are not achievable or if the location
is otherwise less than optimum, the
possibility of stratification should be
investigated as described in section 3.2. The
measurement point shall be within the
centroidal area of the stack or duct cross
section.
3.1.1 Point GEMS. It is suggested that the
measurement point be (1) no less than 1.0
meter from the stack or duct wall or (2)
within or centrally located over the
centroidal area of the stack or duct cross
section.
3.1.2 Path GEMS. It is suggested that the
effective measurement path (1) be totally
within the inner area bounded by a line 1.0
meter from the stack or duct wall, or (2) have
at least 70 percent of the path within the
inner 50 percent of the stack or duct cross-
sectional area.
3.2 Stratification Test Procedure.
Stratification is defined as a difference in
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17513
excess of 10 percent between the average
concentration in the duct or stack and the
concentration at any point more than 1.0
motor from the duct or stack wall. To
determine whether effluent stratification
exists, a dual probe system should be used
to determine the average effluent
concentration while measurements at each
traverse point are being made. One probe,
located at the stack or duct centroid, is used
as a stationary reference point to indicate the
change in effluent concentration over time.
The second probe is used for sampling at the
traverse points specified in 40 CFR part 60
appendix A, method 1. The monitoring
system samples sequentially at the reference
and traverse points throughout the testing
period for five minutes at each point.
4. Performance and Equipment
Specifications
4.1 Data Recorder Scale. The GEMS data
recorder response range must include zero
and a high level value. The high level value
must be equal to the span value. If a lower
high level value is used, the GEMS must have
the capability of providing multiple outputs
with different high level values (one of which
is equal to the span value) or be capable of
automatically changing the high level value
as required (up to the span value) such that
the measured value does not exceed 95
percent of the high level value.
4.2 Relative Accuracy (RA). The RA of
the GEMS must be no greater than 20 percent
of the mean value of the RM test data in
terms of units of the emission standard, or 10
percent of the applicable standard,
whichever is greater.
4.3 Calibration Error. Calibration error is
assessed using certified NIST traceable
cylinder gas standards for chlorine. The
mean difference between the indicated GEMS
concentration and the reference
concentration shall be no greater than ±15
percent of the reference concentration. The
reference concentration shall be the greater of
80 to 120 percent of the applicable emission
standard or 50 ppm CU, in nitrogen.
4.4 Calibration Drift. The GEMS design
must allow the determination of calibration
drift at concentration levels commensurate
with the applicable emission standard. The
CEMS calibration may not drift or deviate
from the reference value (RV) of the
calibration standard by more than 2 percent
of tho reference value. The calibration shall
be performed at a level equal to 80 to 120
percent of the applicable emission standard.
4.5 Zero Drift. The GEMS design must
allow the determination of calibration drift at
tho zero level (zero drift). The GEMS zero
point shall not drift by more than 2 percent
of the emission standard.
4.6 Sampling and Response Time. The
CEMS shall sample the stack effluent
continuously. Averaging time, the number of
measurements in an average, and the
averaging procedure for reporting and
determining compliance shall conform with
that specified in the applicable emission
regulation.
4.6.1 Response Time. The response time
of the CEMS should not exceed 2 minutes to
achieve 95 percent of the final stable value.
The response time shall be documented by
tho CEMS manufacturer.
4.7 CEMS Interference Response. While
the CEMS is measuring the concentration of
chlorine in the high-level calibration source
used to conduct the CE test, the gaseous
components (in nitrogen) listed in Table I
shall be introduced into the measurement
system either separately or in combination.
The interference test gases must be
introduced in such a way as to cause no
change in the calibration concentration of
chlorine being delivered to the CEMS. The
concentrations listed in the table are the
target levels at the sampling interface of the
CEMS based on the known cylinder gas
concentrations and the extent of dilution (see
Section 9). Interference is defined as the
difference between the CEMS response with
these components present and absent. The
sum of the interferences must be less than 2
percent of the emission limit value. If this
level of interference is exceeded, then
corrective action to eliminate the
interference(s) must be taken.
TABLE I.—INTERFERENCE TEST GAS
CONCENTRATIONS IN NITROGEN
Gas
Carbon Monoxide
Carbon Dioxide
Oxygen
Sulfur Dioxide
Nitrogen Dioxide .
Water Vapor ..
Hydrogen Chloride (HCI) ...
Concentration
500 + 50 ppm
10 + 1 percent
20 9 + 1 per-
cent.
500 + 50 ppm
250 ± 25 ppm
25 + 5 percent
50 ± 5 ppm.
5. Performance Specification Test Procedure
5.1 Pretest Preparation. Install the CEMS
and prepare the RM test site according to the
specifications in Section 3, and prepare the
CEMS for operation according to the
manufacturer's written instructions.
5.2 Calibration and Zero Drift Test
Period. While the affected facility is
operating at more than 50 percent of normal
load, or as specified in an applicable subpart,
determine the magnitude of the calibration
drift (CD) and zero drift (ZD) once each day
(at 24-hour intervals) for 7 consecutive days
according to the procedure given in Section
6. To meet the requirements of Sections 4.4
and 4.5 none of the CD's or ZD's may exceed
the specification. All CD determinations
must be made following a 24-hour period
during which no unscheduled maintenance,
repair, or manual adjustment of the CEMS
took place.
5.3 CE Test Period. Conduct a CE test
prior to the CD test period. Conduct the CE
test according to the procedure given in
Section 8.
5.4 CEMS Interference Response Test
Period. Conduct an interference response test
in conjunction with the CE test according to
the procedure given in Section 9.
6.0 The CEMS Calibration and Zero Drift
Test Procedure
This performance specification is designed
to allow calibration of the CEMS by use of
gas samples, filters, etc, that challenge the
pollutant analyzer part of the CEMS (and as
much of the whole system as possible), but
which do not challenge the entire GEMS,
including the sampling interface. Satisfactory
response of the entire system is covered by
the RA and CE requirements.
The CD measurement is to verify the ability
of the CEMS to conform to the established
GEMS calibration used for determining the
emission concentration. Therefore, if
periodic automatic or manual adjustments
are made to the CEMS zero and calibration
settings, conduct the CD test immediately
before the adjustments, or conduct it in such
a way that the CD and ZD can be determined.
Conduct the CD and ZD tests at the points
specified in Sections 4.4 and 4.5. Record the
CEMS response and calculate the CD
according to:
(1)
R
Where CD denotes the calibration drift of the
CEMS in percent, RCEM is the CEMS
response, and Rv is the reference value of the
high level calibration standard. Calculate the
ZD according to:
ZD = -
R
-xlOO,
(2)
EM
Where ZD denotes the zero drift of the CEMS
in percent, RCEM is the CEMS response, Rv
is the reference value of the low level
calibration standard, and REM is the emission
limit value.
7. Calibration Error Test Procedure
7.1 Sampling Strategy. The CEMS
calibration error shall be assessed using the
calibration source specified in Section 4.3.
Challenge the CEMS at the measurement
levels specified in Section 4.3. During the
test, operate the CEMS as nearly as possible
in its normal operating mode. The calibration
gases should be injected into the sampling
system as close to the sampling probe outlet
as practical and shall pass through all filters,
scrubbers, conditioners, and other monitor
components used during normal sampling.
7.2 Number of tests. Challenge the CEMS
three non-consecutive times at each
measurement point and record the responses.
The duration of each challenge should be for
a sufficient period of time to ensure that the
CEMS surfaces are conditioned and a stable
output obtained.
7.3 Calculations. Summarize the results
on a data sheet. Calculate the mean
difference between the CEMS response and
the known reference concentration at each
measurement point according to equations 5
and 6 of Section 10. The calibration error
(CE) at each measurement point is then given
by:
CE = |d/Rv|xlOO, . I/' (3)
Where Rv is the reference concentration
value. :
8. Interference Response Test Procedure
8.1 Test Strategy. Perform the
interference response test while the CEMS is
being challenged by the high level calibration
source (after the CE determination has been
-------
17514
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
made). The interference test gases should be
injected into the sampling system as close to
the sampling probe outlet as practical and
shall pass through all filters, scrubbers,
conditioners, and other monitor components
used during normal sampling.
8.2 Number of tests. Introduce the
interference test gas three times alternately
with the high-level calibration gas and record
the responses both with and without the
interference test gas. The duration of each
test should be for a sufficient period of time
to ensure that the GEMS surfaces are
conditioned and a stable output obtained.
8.3 Calculations. Summarize the results
on a data sheet. Calculate the mean
difference between the GEMS response with
and without the interference test gas by
taking the average of the. GEMS responses
with and without the interference test gas
(see equation 5) and then taking the
difference (d). The percent interference (I) is
then given by:
I = |d/RHL|xlOO, (4)
Where RHL is the value of the high-level
calibration standard. If the gaseous
components of the interference test gas are
introduced separately, then the total
interference is the sum of the individual
interferences.
9. Equations
9.1 Arithmetic Mean. Calculate the
arithmetic mean of a data set as follows:
ni=l
Where n is equal to the number of data
points.
9.1.1 Calculate the arithmetic mean of the
difference, d, of a data set, using Equation 5
and substituting d for x. Then
d,=x,-yif (6)
Where x and y are paired data points from
the CEMS and RM, respectively.
10. Reporting
At a minimum (check with the appropriate
regional office, or State, or local agency for
additional requirements, if any) summarize
in tabular form the results of the CE,
interference response, CD and RA tests.
Include all data sheets, calculations, and
records of CEMS response necessary to
substantiate that the performance of the
CEMS met the performance specifications.
• The CEMS measurements shall be reported
to the agency in units of ng/m3 on a dry basis,
corrected to 20 °C and 7 percent Oi.
11. Bibliography
1. 40 CFR Part 60, Appendix B,
"Performance Specification 2—Specifications
and Test Procedures for SO2 and NOx
Continuous Emission Monitoring Systems in
Stationary Sources."
2. 40 CFR Part 60, Appendix B,
"Performance Specification 1—Specification
and Test Procedures for Opacity Continuous
Emission Monitoring Systems in Stationary
Sources." •':-
3. 40 CFR Part 60, Appendix A, "Method
1—Sample and Velocity Traverses for
Stationary Sources."
4. 40 CFR Part 266, Appendix IX, Section
2, "Performance Specifications for
Continuous Emission Monitoring Systems."
5. "Continuous Emission Monitoring
Technology Survey for Incinerators, Boilers,
and Industrial Furnaces: Final Report for
Metals CEM's," prepared for the Office of
Solid Waste, U.S. EPA, Contract No. 68-D2-
0164 (4/25/94).
PART 63—NATIONAL EMISSION
STANDARDS FOR HAZARDOUS AIR
POLLUTANTS FOR SOURCE
CATEGORIES
II. In part 63:
1. The authority citation for part 63
continues to read as follows:
Authority: 42 U.S.C. 7401 et seq.
2. Part 63 is revised by adding subpart
EEE, to read as follows:
Subpart EEE—National Emission
Standards for Hazardous Air Pollutants
From Hazardous Waste Combustors
Sec.
63.1200 Applicability.
63.1201 Definitions.
63.1202 Construction and reconstruction.
63.1203 Standards for hazardous waste
incinerators (HWIs).
63.1204 Standards for cement kilns (CKs)
that burn hazardous waste.
63.1205 Standards for lightweight aggregate
kilns (LWAKs) that burn hazardous
waste.
63.1206 Initial compliance dates.
63.1207 Compliance with standards and
general requirements.
63.1208 Performance testing requirements.
63.1209 Test methods.
63.1210 Monitoring requirements.
63.1211 Notification requirements.
63,1212 Recordkeeping and reporting
requirements.
Appendix to Subpart EEE—Quality
Assurance Procedures for Continuous
Emissions Monitors Used for Hazardous
Waste Combustors
§63.1200 Applicability.
(a) The provisions of this subpart
apply to all hazardous waste combustors
(HWCs): hazardous waste incinerators,
cement kilns that burn hazardous waste,
and lightweight aggregate kilns that
burn hazardous waste.
(b) HWCs are subject to the provisions
of part 63 as major sources irrespective
of the quantity of hazardous air
pollutants emitted.
(c) When a HWC continues to operate
when hazardous waste is neither being
fed nor remains in the combustion
chamber, the source remains subject to
this subpart until hazardous waste
burning is terminated.
(l) A source has terminated hazardous
waste burning if:
(i) It has stopped feeding hazardous
waste and hazardous waste does not
remain in the combustion chamber;
(ii) The owner or operator notifies the
Administrator in writing within 5
calendar days after hazardous waste
burning has ceased that hazardous
waste burning has terminated.
(2) A source that has terminated
hazardous waste burning may resume
hazardous waste burning provided that:
(i) It complies with requirements in
this subpart for new sources; and
(ii) The owner and operator submits a
notification of compliance based on
comprehensive performance testing
after burning has been resumed.
Hazardous waste cannot be burned for
more than 720 hours prior to submittal
of the notification of compliance, and
may be burned only for purposes of
emissions testing in preparation for
performance testing or performance
testing.
(d) HWCs are also subject to
applicable requirements under parts
260-270 of this chapter.
(e) The more stringent of requirements
of an operating permit issued under part
270 of this chapter or the requirements
of this subpart (and part) apply. If
requirements of the operating permit
issued under part 270 of this chapter
conflict with any requirements of this
subpart (and part 63), the requirements
of this subpart (and part 63) take
precedence.
(f) If the only hazardous wastes that
a HWC burns are those exempt from
regulation under § 266.100(b) of'this
chapter, the HWC is not subject to the
requirements of this subpart.
fg) Waiver of emission standards. (1)
Nondetect levels of Hg, SVM, or LVM in
feedstreams. If no feedstream to a HWC
contains detectable levels of Hg, SVM,
or LVM, the HWC is not subject to the
emission standards and ancillary
performance testing, monitoring,
notification, and recordkeeping and
reporting requirements for those
standards provided in this subpart. To
be eligible for this waiver, the owner
and operator must also develop and
implement a feedstream sampling and
analysis plan to document that no
feedstream contains detectable levels of
the metals.
(2) Nondetect levels of chlorine in
feedstreams. If no feedstream to a HWC
contains detectable levels of chlorine,
the HWC is not subject to the HC1/C12
emission standard and ancillary
performance testing, monitoring,
notification, and recordkeeping and
reporting requirements for that standard
in this subpart. To be eligible for this
waiver, the owner and operator must
also develop and implement a
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17515
feedstream sampling and analysis plan
to document that no feedstream
contains detectable levels of the
chlorine.
§63.1201 Definitions.
The terms used in this part are
defined in the Act, in subpart A of this
part, or in this section as follows:
Air pollution control system means
the equipment used to reduce the
release of particulate matter and other
pollutants to the atmosphere.
Automatic waste feed cutoff (AWFCO)
system means a system comprised of
cutoff valves, actuator, sensor, data
manager, and other necessary
components and electrical circuitry
designed, operated and maintained to
stop the flow of hazardous waste to the
combustion unit automatically and
immediately when any of the
parameters to which the system is
interlocked exceed the limits
established in compliance with
applicable standards, the operating
permit, or safety considerations.
By-pass duct means a device which
diverts a minimum of 10 percent of a
cement kiln's off gas.
Cement kiln means a rotary kiln and
any associated preheater or precalciner
devices that produces clinker by heating
limestone and other materials for
subsequent production of cement for
use in commerce, and that burns
hazardous waste.
Combustion chamber means the area
in which controlled flame combustion
of hazardous waste occurs.
Compliance date means the date by
which a hazardous waste combustor
must submit a notification of
compliance under this subpart.
Comprehensive performance test
means the performance test during
which a HWC demonstrates compliance
with emission standard and establishes
or re-establishes operating limits.
Confirmatory performance test means
the performance test conducted under
normal operating conditions to
demonstrate compliance with the D/F
emission standard.
Continuous monitor means a device
which continuously samples the
regulated parameter without
interruption except during allowable
periods of calibration, and except as
defined otherwise by the CEM
Performance Specifications in appendix
B, part 60.
Dioxins andfurans (D/F) means tetra-,
penta-, hexa-, hepta-, and octa-
chlorinated dibenzo dioxins and furans.
Feedstream means any material fed
into a HWC, including, but not limited
to, any pumpable or nonpumpable solid
or gas.
Flowrate means the rate at which a
feedstream is fed into a HWC.
Fugitive combustion emissions means
particulate or gaseous matter generated
by or resulting from the burning of
hazardous waste that is not collected by
a capture system and is released to the
atmosphere prior to the exit of the stack.
Hazardous waste is defined in § 261.3
of this chapter.
Hazardous waste combustor (HWC)
means a hazardous waste incinerator, or
a cement kiln, or a lightweight aggregate
kiln.
Hazardous waste incinerator means a
device defined in 260.10 of this chapter
that burns hazardous waste.
Initial comprehensive performance
test means the comprehensive
performance test that is used as the
basis for initially demonstrating
compliance with the standards.
Instantaneous monitoring means
continuously sampling, detecting, and
recording the regulated parameter
without use of an averaging period.
Lightweight aggregate kiln means a
rotary kiln that produces for commerce
(or for manufacture of products for
commerce) an aggregate with a density
less than 2.5 g/cc by slowly heating
organic-containing geologic materials
such as shale and clay, and that burns
hazardous waste.
Low volatility metals means arsenic,
beryllium, chromium, and antimony,
and their compounds.
New source means a HWC that first
begins to burn hazardous waste, or the
construction or reconstruction of which
is commenced, after April 19,1996.
Notification of compliance means a
notification in which the owner and
operator certify, after completion of
performance evaluations and tests, that
the HWC meets the emission standards,
CMS, and other requirements of this
subpart, and that the source is in
compliance with operating limits.
One-minute average means the
average of detector responses calculated
at least every 60 seconds from responses
obtained at least each 15 seconds.
Operating record means a
documentation of all information
required by the standards to document
and maintain compliance with the
applicable regulations, including data
and information, reports, notifications,
and communications with regulatory
officials.
Reconstruction means the
replacement or addition of components
of a hazardous waste combustor to such
an extent that:
(1) The fixed capital cost of the new
components exceeds 50 percent of the
fixed capital cost that would be required
to construct a comparable new source.
(2) Upon reconstruction, the
combustor becomes subject to the
standards for new sources, including
compliance dates, irrespective of any
change in emissions of hazardous air
pollutants from that source.
Rolling average means the average of
all one-minute averages over the
averaging period.
Run means the net period of time
during which an air emission sample is
collected under a given set of operating
conditions. Three or more runs
constitutes an emissions test. Unless
otherwise specified, a run may be either
intermittent or continuous.
Semivolatile metals means cadmium
and lead, and their compounds.
TEQ means the international method
of expressing toxicity equivalents for
dioxins and furans as defined in U.S.
EPA, Interim Procedures for Estimating
Risks Associated with Exposures to
Mixtures of Chlorinated Dibenzo-p-
Dioxins and -Dibenzofurans (CDDs and
CDFs) and 1989 Update, March 1989.
§63.1202 Construction and
reconstruction.
The requirements of § 63.5 apply,
except the following apply in lieu of
§§63.5(d)(3)(v) and (vi) and (e)(l)(ii)(D),
as follows:
(a) A discussion of any technical
limitations the source may have in
complying with relevant standards or
other requirements after the proposed
replacements. The discussion shall be
sufficiently detailed to demonstrate to
the Administrator's satisfaction that the
technical limitations affect the source's
ability to comply with the relevant
standard and how they do so.
(b) If in the application for approval
of reconstruction the owner or operator
designates the affected source as a
reconstructed source and declares that
there are no technical limitations to
prevent the source from complying with
all relevant standards or other
requirements, the owner or operator
need not submit the information
required in paragraphs (d)(3) (iii)
through (v) of this section.
(c) Any technical limitations on
compliance with relevant standards that
are inherent in the proposed
replacements.
§ 63.1203 Standards for hazardous waste
incinerators (HWIs).
(a) Emission limits for existing
sources. No owner or operator of an
existing HWI shall discharge or cause
combustion gases to be emitted into the
atmosphere that contain:
(1) Dioxins and furans in excess of
0.20 ng/dscm (TEQ) corrected to 7
percent oxygen;
-------
17516
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
(2) Mercury in excess of 50 ug/dscm,
over a 10-hour rolling average, and
corrected to 7 percent oxygen;
(3) Lead and cadmium in excess of
270 ug/dscm, combined emissions,
corrected to 7 percent oxygen, and
measured over a 12-hour rolling average
if compliance is based on a GEMS;
(4) Arsenic, beryllium, chromium,
and antimony in excess of 210 ug/dscm,
combined emissions, corrected to 7
percent oxygen and measured over a 10-
hour rolling average if compliance is .
based on a GEMS;
(5) Carbon monoxide in excess of 100
parts per million by volume, over an
hourly rolling average, dry basis and
corrected to 7 percent oxygen;
(6) Hydrocarbons in excess of 12 parts
per million by volume, over an hourly
rolling average, dry basis, corrected to 7
percent oxygen, and reported as
propane;
(7) Hydrochloric acid and chlorine gas
in excess of 280 parts per million by
volume, combined emissions, expressed
as hydrochloric acid equivalents, dry
basis and corrected to 7 percent oxygen,
and measured over a hourly rolling
average if compliance is based on a
GEMS; and
(8) Particulate matter (PM) in excess
of 69 mg/dscm, over a 2-hour rolling
average and corrected to 7 percent
oxygen.
(b) Emission limits for new sources.
No owner or operator that commences
construction or reconstruction of a HWI,
or that first burns hazardous waste in an
existing incinerator, after April 19,
1996, shall discharge or cause
combustion gases to be emitted into the
atmosphere that contain:
(1) Dioxins and furans in excess of
0.20 ng/dscm (TEQ), corrected to 7
percent oxygen;
(2) Mercury in excess of 50 ug/dscm,
over a 10-hour rolling average, corrected
to 7 percent oxygen;
(3) Lead and cadmium in excess of 62
ug/dscm, combined emissions,
corrected to 7 percent oxygen and
measured over a 10-hour rolling
average;
(4) Arsenic, beryllium, chromium,
and antimony in excess of 60 ug/dscm
(or 80 ug/dscm if compliance is based
on a GEMS), combined emissions,
corrected to 7 percent oxygen and
measured over a 10-hour rolling average
if compliance is based on a GEM;
(5) Carbon monoxide in excess of 100
parts per million by volume, over an
hourly rolling average, dry basis and
corrected to 7 percent oxygen;
(6) Hydrocarbons in excess of 12 parts
per million by volume, over an hourly
rolling average, dry basis, corrected to 7
percent oxygen, and reported as
propane;
(7) Hydrochloric acid and chlorine gas
in excess of 67 parts per million by
volume, combined emissions, expressed
as hydrochloric acid equivalents, dry
basis and corrected to 7 percent oxygen,
and measured over a hourly rolling
average if compliance is based on a
GEM; and
(8) Particulate matter (PM) in excess
of 69 mg/dscm, over a 2-hour rolling
average and corrected to 7 percent
oxygen.
(c) Significant figures. The emission
limits provided by paragraphs (a) and
(b) of this section shall be considered to
have two significant figures. Emissions
measurements may be rounded to two
significant figures to demonstrate
compliance.
(d) Air emission standards for
equipment leaks, tanks, surface
impoundments, and containers. Owners
and operators of HWIs are subject to the
air emission standards of Subparts BB
and CC, part 264, of this chapter.
§63.1204 Standards for cement kilns
(CKs) that burn hazardous waste.
(a) Emission limits for existing
sources. No owner or operator of an
existing CK shall discharge or cause
combustion gases (resulting solely or
partially from burning hazardous waste)
to be emitted into the atmosphere that
contain:
(1) Dioxins and furans in excess of
0.20 ng/dscm, TEQ, corrected to 7
percent oxygen;
(2) Mercury in excess of 50 ug/dscm,
over a 10-hour rolling average, and
corrected to 7 percent oxygen;
(3) Lead and cadmium in excess of 57
ug/dscm, combined emissions,
corrected to 7 percent oxygen, and
measured over a 10-hour rolling average
if compliance is based on a GEMS;
(4) Arsenic, beryllium, chromium,
and antimony in excess of 130 ug/dscm,
combined emissions, corrected to 7
percent oxygen and measured over a 10-
hour rolling average if compliance is
based on a GEMS;
(5) Carbon Monoxide. For kilns
equipped with a by-pass duct, either:
(i) Carbon monoxide in the by-pass
duct in excess of 100 parts per million
by volume, over an hourly rolling
average, dry basis and corrected to 7
percent oxygen; or
(ii) Hydrocarbons in the by-pass duct
in excess of 6.7 parts per million by
volume, over an hourly rolling average,
dry basis, corrected to 7 percent oxygen,
and reported as propane.
(6) Hydrocarbons. Hydrocarbons in
the main stack of kilns not equipped
with a by-pass duct in excess of 20 parts
per million by volume, over an hourly
rolling average, dry basis, corrected to 7
percent oxygen, and reported as
propane;
(7) Hydrochloric acid and chlorine gas
in excess of 630 parts per million by
volume, combined emissions, expressed
as hydrochloric acid equivalents, dry
basis, corrected to 7 percent oxygen, and
measured over a hourly rolling average
if compliance is based on a GEMS; and
(8) Particulate matter (PM) in excess
of 69 mg/dscm over a 3-hour rolling
average and corrected to 7 percent
oxygen.
(b) Emission limits for new sources.
No owner or operator that commences
construction or reconstruction of a CK,
or that first burns hazardous waste in an
existing CK, after April 19,1996, shall
discharge or cause combustion gases to
be emitted into the atmosphere that
contain:
(1) Dioxins and furans in excess of
0.20 ng/dscm (TEQ) corrected to 7
percent oxygen;
(2) Mercury in excess of 50 ug/dscm,
over a 10-hour rolling average, corrected
to 7 percent oxygen;
(3) Lead and cadmium in excess of 55
ug/dscm, combined emissions,
corrected to 7 percent oxygen, or if
compliance is based on a GEMS, 60 ug/
dscm, combined emissions, corrected to
7 percent oxygen and measured over a
10-hour rolling average;
(4) Arsenic, beryllium, chromium,
and antimony in excess of 44 ug/dscm,
combined emissions, corrected to 7
percent oxygen, or, if compliance is
based on a GEM, 80 ug/dscm, combined
emissions, corrected to 7 percent oxygen
and measured over a 10-hour rolling
average;
(5) Carbon Monoxide. For kilns
equipped with a by-pass duct, either:
(i) Carbon monoxide in the by-pass
duct in excess of 100 parts per million
by volume, over an hourly rolling
average, dry basis and corrected to 7
percent oxygen; or
(ii) Hydrocarbons in the by-pass duct
in excess of 6.7 parts per million by
volume, over an hourly rolling average,
dry basis, corrected to 7 percent oxygen,
and reported as propane.
(6) Hydrocarbons. Hydrocarbons in
the main stack of kilns not equipped
with a by-pass duct in excess of 20 parts
per million by volume, over an hourly
rolling average, dry basis, corrected to 7
percent oxygen, and reported as
propane;
(7) Hydrochloric acid and chlorine gas
in excess of 67 parts per million,
combined emissions, expressed as
hydrochloric acid equivalents, dry basis
and corrected to 7 percent oxygen, and
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17517
measured over a hourly rolling average
if compliance is based on a GEMS; and
(8) Particulate matter (PM) in excess
of 69 mg/dscm over a 2-hour rolling
average and corrected to 7 percent
oxygen.
(c) Significant figures. The emission
limits provided by paragraphs (a) and
(b) of this section shall be considered to
have two significant figures. Emissions
measurements may be rounded to two
significant figures to demonstrate
compliance.
(d) Air emission standards for
equipment leaks, tanks, surface
impoundments, and containers. Owners
and operators of CKs are subject to the
air emission standards of subparts BB
and CC, part 264, of this chapter.
§ 63.1205 Standards for lightweight
aggregate kilns (LWAKs) that burn
hazardous waste.
(a) Emission limits for existing
sources. No owner or operator of an
existing LWAK shall discharge or cause
combustion gases to be emitted into the
atmosphere mat contain:
(1) Dioxins and furans in excess of
0.20 ng/dscm (TEQ), corrected to 7
percent oxygen;
(2) Mercury in excess of 72 ug/dscm,
over a 10-hour rolling average, and
corrected to 7 percent oxygen;
(3) Lead anoV cadmium in excess of 12
ug/dscm, combined emissions,
corrected to 7 percent oxygen, or, if
compliance is based on a GEMS, 60 ug/
dscm, combined emissions, corrected to
7 percent oxygen and measured over a
10-hour rolling average;
(4) Arsenic, beryllium, chromium,
and antimony in excess of 340 ug/dscm,
combined emissions, corrected to 7
percent oxygen, and measured over a
10-hour rolling average if a GEMS is
used for compliance;
(5) Carbon monoxide in excess of 100
parts per million by volume, over an
hourly rolling average, dry basis and
corrected to 7 percent oxygen;
(6) Hydrocarbons in excess of 14 parts
per million by volume, over an hourly
rolling average, dry basis, corrected to 7
percent oxygen, and reported as
propane;
(7) Hydrochloric acid and chlorine gas
in excess of 450 parts per million by
volume, combined emissions, expressed
as hydrochloric acid equivalents, dry
basis and corrected to 7 percent oxygen,
and measured over a hourly rolling
average if compliance is based on a
GEMS; and
(8) Particulate matter (PM) in excess
of 69 mg/dscm over a 2-hour rolling
average and corrected to 7 percent
oxygen.
(bj Emission limits for new sources.
No owner or operator that commences
construction or reconstruction of a
LWAK, or that first burns hazardous
waste in an existing LWAK, after April
19,1996, shall discharge or cause
combustion gases to be emitted into the
atmosphere that contain:
(1) Dioxins and furans in excess of
0.20 ng/dscm (TEOJ, corrected to 7
percent oxygen;
(2) Mercury in excess of 72 ug/dscm,
over a 10-hour rolling average, corrected
to 7 percent oxygen;
(3j Lead and cadmium in excess of 5.2
ug/dscm, combined emissions,
corrected to 7 percent oxygen, or, if
compliance is based on a GEMS, 60 ug/
dscm, combined emissions, corrected to
7 percent oxygen and measured over a
10-hour rolling average;
(4) Arsenic, beryllium, chromium,
and antimony in excess of 55 ug/dscm,
combined emissions, corrected to 7
percent oxygen, or, if compliance is
based on a GEMS, 80 ng/dscm,
combined emissions, corrected to 7
percent oxygen and measured over a 10-
hour rolling average;
(5) Carbon monoxide in excess of 100
parts per million by volume, over an
hourly rolling average, dry basis and
corrected to 7 percent oxygen;
(6) Hydrocarbons in excess of 14 parts
per million by volume, over an hourly
rolling average, dry basis, corrected to 7
percent oxygen, and reported as
propane;
(7) Hydrochloric acid and chlorine gas
in excess of 62 parts per million by
volume, combined emissions, expressed
as hydrochloric acid equivalents, dry
basis and corrected to 7 percent oxygen,
and measured over a hourly rolling
average if compliance is based on a
GEMS; and
(8) Particulate matter (PM) in excess
of 69 mg/dscm over a 2-hour rolling
average and corrected to 7 percent
oxygen.
(c) Significant figures. The emission
limits provided by paragraphs (a) and
(b) shall be considered to have two •
significant figures. Emissions
measurements may be rounded to two
significant figures to demonstrate
compliance.
(d) Air emission standards for
equipment leaks, tanks, surface
impoundments, and containers. Owners
and operators of LWAKs are subject to
the air emission standards subparts BB
and CC, part 264, of this chapter.
§63.1206 Initial Compliance dates.
(a) Existing sources. (1) Compliance
Date. Each owner or operator of an
existing hazardous waste combustor
(HWC) shall submit to the
Administrator under §63.1211 an initial
notification of compliance certifying
compliance with the requirements of
this subpart no later than [date 36
months after publication of the final
rule], unless an extension of time is
granted under § 63.6(i).
(2) Failure to meet compliance date.
(i) Termination of waste burning. If an
owner or operator fails to submit the
notification of compliance as specified
in paragraph (a)(l) of this section,
hazardous waste burning must
terminate on the date that the owner or
operator determine that the notification
will not be submitted by the deadline,
but not later than the date the
notification should have been
submitted.
(ii) Requirements for resuming waste
burning. (A) If a source that fails to ,
submit a timely initial notification of
compliance has not been issued a RCRA
operating permit under part 270 of this
chapter for the HWC, the source may
not resume burning hazardous waste
until a RCRA permit is issued.
(B) If a source that fails to submit a
timely initial notification of compliance
has already been issued a RCRA
operating permit under part 270 of this
chapter for the HWC, the source may
resume burning hazardous waste only
for a total of 720 hours and only for
purposes of pretesting or comprehensive
performance testing prior to submitting
an initial notification of compliance. If
the owner and operator do not submit
an initial notification of compliance
within 90 days after the date it is due,
they must begin closure procedures
under the RCRA operating permit unless
an extension of time is granted prior to
that date in writing by the
Administrator for good cause.
(C) The source must comply with the
requirements for new sources under this
subpart.
(b) New sources. (1) Sources that
begin burning hazardous waste before
the effective date but after the date of
proposal. Each owner or operator of a
new source that first burns hazardous
waste prior to [date of publication of
final rule] but after April 19,1996 shall:
(i) For any requirements of this
subpart (and part) that are not more
stringent than the proposed
requirement, submit to the
Administrator a notification of
compliance at the time specified in the
operating permit issued under part 270
of this chapter;
(ii) For any requirements of this
subpart (and part) that are more
stringent than the proposed
requirement:
(A) Submit to the Administrator a
notification of compliance not later than
[date 36 months after publication of the
-------
27528
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
final rule], unless an extension of time
is granted under § 63.6(i); and
IB) Comply with the standards as
proposed in the interim until the
notification of compliance is submitted.
(2) Sources that begin burning
hazardous waste after the effective date.
Each owner or operator of a new source
that first bums hazardous waste after
[date of publication of final rule] must
submit the notification of compliance at
the time specified in the operating
permit issued under part 270 of this
chapter.
Note to paragraph (b) of this section: An
owner or operator wishing to commence
construction of a hazardous waste incinerator
or hazardous waste-burning equipment for a
cement kiln or lightweight aggregate kiln
must first obtain some type of RCRA
authorization, whether it be a RCRA permit,
a modification to an existing RCRA permit,
or a change under already existing interim
status. See 40 CFR part 270.
§ 63.1207 Compliance with standards and
general requirements.
(a) Compliance with standards. (1)
Standards are in effect at all times. A
hazardous waste combustor (HWC) shall
not burn hazardous waste (that is,
hazardous waste must not be fed and
hazardous waste must not remain in the
combustion chamber) except in
compliance with the standards of this
subpart, including periods of startup,
shutdown, and malfunction. Therefore,
the owner or operator of a HWC is not
subject to the requirements of §§ 63.6(e)
and (f)(l) (regarding operation and
maintenance in conformance with a
startup, shutdown, and malfunction
plan) when burning hazardous waste.
(2) Automatic waste feed cutoff
(AWFCO). During the initial
comprehensive performance test
required under § 63.1208, and upon
submittal of the initial notification of
compliance under § 63.1211, a HWC
must be operated with a functioning
system that immediately and
automatically cuts off the hazardous
waste feed when any of the following
are exceeded: applicable operating
limits specified under § 63.1210; the
emission levels monitored by GEMS; the
span value of any CMS detector, except
a GEMS; the automatic waste feed cutoff
system fails; or the allowable
combustion chamber pressure.
(i) Ducting of combustion gases.
During a AWFCO, combustion gases
must continue to be ducted to the air
pollution control system while
hazardous waste remains in the
combustion chamber;
(ii) Restarting waste feed. The
operating parameters for which limits
are established under § 63.1210 and the
emissions required under that section to
be monitored by a GEMS must continue
to be monitored during the cutoff, and
the hazardous waste feed shall not be
restarted until the operating parameters
and emission levels are within
allowable levels;
(iii) Violations. If, after a AWFCO, a
parameter required to be interlocked
with the AWFCO system exceeds an
allowable level while hazardous waste
remains in the combustion chamber, the
owner and operator have violated the
emission standards of this subpart.
(iv) Corrective measures. After any
AWFCO that results in a violation as
defined in paragraph (a) (2) (iii) of this
section, the owner or operator must
investigate the cause of the AWFCO,
take appropriate corrective measures to
minimize future AWFCO violations, and
record the findings and corrective
measures in the operating record.
(v) Excessive AWFCO report. If a
HWC experiences more than 10
AWFCOs in any 60-day period that
result in an exceedance of any
parameter required to be interlocked
with the AWFCO system under this
section, the owner or operator must
submit a written report within 5
calendar days of the 10th AWFCO
documenting the results of the
investigation and corrective measures
taken.
(vi) Limit on AWFCOs. The
Administrator may limit the number of
cutoffs per an operating period on a
case-by-case basis.
(vii) Testing. The AWFCO system and
associated alarms must be tested at least
weekly to verify operability, unless the
owner and operator document in the
operating record that weekly
inspections will unduly restrict or upset
operations and that less frequent
inspection will be adequate. At a
minimum, operational testing must be
conducted at least monthly.
(3) ESV Openings, (i) Violation. If an
emergency safety vent opens when
hazardous waste is fed or remains in the
combustion chamber, such that
combustion gases are not treated as
during the most recent comprehensive
performance test (e.g., if the combustion
gas by-passes any emission control
device operating during the
performance test), it is a violation of the
emission standards of this subpart.
(ii) ESV Operating Plan. The ESV
Operating Plan shall explain detailed
procedures for rapidly stopping waste
feed, shutting down the combustor,
maintaining temperature in the
combustion chamber until all waste
exits the combustor, and controlling
emissions in the event of equipment
malfunction or activation of any ESV or
other bypass system including
calculations demonstrating that
emissions will be controlled during
such an event (sufficient oxygen for
combustion and maintaining negative
pressure), and the procedures for
executing the plan whenever the ESV is
used, thus causing an emergency release
of emissions.
(iii) Corrective measures. After any
ESV opening that results in a violation
as defined in paragraph (b)(l) of this
section, the owner or operator must
investigate the cause of the ESV
opening, take appropriate corrective
measures to minimize future ESV
violations, and record the findings and
corrective measures in the operating
record.
(iv) Reporting requirement. The
owner or operator must submit a written
report within 5 days of a ESV opening
violation documenting the result of the
investigation and corrective measures
taken.
(b) Fugitive emissions. (1) Fugitive
emissions must be controlled by:
(i) Keeping the combustion zone
totally sealed against fugitive emissions;
or
(ii) Maintaining the maximum
combustion zone pressure lower than
ambient pressure using an
instantaneous monitor; or
(iii) Upon prior written approval of
the Administrator, an alternative means
of control to provide fugitive emissions
control equivalent to maintenance of
combustion zone pressure lower than
ambient pressure;
(2) The owner or operator must
specify in the operating record the
method used for fugitive emissions
control.
(c) Finding of compliance. The
procedures of determining compliance
and finding of compliance provided by
§ 63.6(f)(2) and (3) are applicable to
HWCs, except that paragraph
(f)(2)(iii)(B) (testing is to be conducted
under representative operating
conditions) is superseded by the
requirements for performance testing
under § 63.1208.
(d) Use of an alternative nonopacity
emission standard. The provisions of
§ 63.6(g) are applicable to HWCs.
(e) Extension of compliance with
emission standards. The provisions of
§ 63.6(i) are applicable to HWCs.
(f) Changes in design, operation, or
maintenance. If the design, operation, or
maintenance of the source is changed in
a manner that may affect compliance
with any emission standard that is not
monitored with a GEMS, the source
shall:
(1) Conduct a comprehensive
performance test to re-establish
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17519
operating limits on the parameters
specified in § 63.1210; and
(2) Burn hazardous waste after such
change for no more than a total of 720
hours and only for purposes of
pretesting or comprehensive
performance testing (including
demonstrating compliance with CMS
requirements).
§ 63.1208 Performance testing
requirements.
(a) Types of performance tests. (1)
Comprehensive performance test. The
purpose of the comprehensive
performance test is to demonstrate
compliance with the emission standards
provided by §§ 63.1203,63.1204, and
63.1205, establish limits for the
applicable operating parameters
provided by § 63.1210, and demonstrate
compliance with the performance
specifications for CMS.
(2) Confirmatory performance test.
The purpose of the confirmatory
performance test is to demonstrate
compliance with the D/F emission
standard when the source operates
under normal operating conditions.
(b) Frequency of testing. Testing shall
be conducted periodically as prescribed
in this paragraph Ob). The date of
commencement of the initial
comprehensive performance test shall
be the basis for establishing the
anniversary date of commencement of
subsequent performance testing. A
source may conduct comprehensive
performance testing at any time prior to
the required date. If so, the anniversary
date for subsequent testing is advanced
accordingly. Except as provided by
paragraph (c) of this section, testing
shall be conducted as follows:
(1) Comprehensive performance
testing, (i) Large or off-site sources.
HVVCs that receive hazardous waste
from off-site and HWCs with a gas flow
rate exceeding 23,127 acfm at any time
that hazardous waste is burned or
remains in the combustion chamber
shall commence testing within 35-37
months of the anniversary date of the
initial comprehensive performance test,
and within every 35-37 months of that
anniversary date thereafter.
(ii) Small, on-site sources. HWCs that
burn hazardous waste generated on site
only and that have a gas flow rate of
23,127 acfm or less shall commence
testing xvithin 59-61 months of the
anniversary date of the initial
comprehensive performance test, and
within every 59-61 months of that
anniversary date thereafter. However,
the Administrator may determine on a
case-specific basis that such a source
may pose the same potential to exceed
the standards of this part as a large or
off-site source. If so, the Administrator
may require such a source to comply
with the testing frequency applicable to
large and off-site sources. Factors that
the Administrator may consider
include: type and volume of hazardous
wastes burned, concentration of toxic
constituents in the hazardous waste,
and compliance history.
(2) Confirmatory performance testing.
(i) Large or off-site sources shall
commence confirmatory performance
testing within 17-19 months after the
anniversary date of each comprehensive
performance test.
(ii) Small, on-site sources shall
conduct confirmatory performance
testing within 29-31 months after the
anniversary date of each comprehensive
performance test.
(3) Duration of testing. Performance
testing shall be completed within 30
days after the date of commencement.
(c) Time extension for subsequent
performance tests. After the initial
performance test, a HWC may request
under procedures provided by § 63.6(i)
up to a 1-year time extension for
conducting a performance test in order
to consolidate performance testing with
trial burn testing required under part
270 of this chapter, or for other reasons
deemed acceptable by the
Administrator. If a time extension is
granted, a new anniversary date for
subsequent testing is established as the
date that the delayed testing
commences.
(d) Operating conditions during
testing. (1) Comprehensive performance
testing, (i) The source must operate
under representative conditions (or
conditions that will result in higher
than normal emissions) for the
following parameters to ensure that
emissions are representative (or higher
than) of normal operating conditions:
(A) When demonstrating compliance
with the D/F emission standard, types of
organic compounds in the waste (e.g.,
aromatics, aliphatics, nitrogen content,
halogen/carbon ratio, oxygen/carbon
ratio), and feedrate of chlorine; and
(B) When demonstrating compliance
with the SVM or LVM emission
standard when using manual stack
sampling (i.e., rather than a GEMS) and
the D/F emission standard, normal
feedrates of ash and normal cleaning
cycle of the PM control device.
(ii) Given that limits will be
established for the applicable operating
parameters specified in § 63.1210, a
source may conduct testing under two
or more operating modes to provide
operating flexibility. If so, the source
must note in the operating record under
which mode it is operating at all times.
(2) Confirmatory performance testing.
Confirmatory performance testing for D/
F shall be conducted under normal
operating conditions defined as follows:
(i) The CO, HC, and PM GEM
emission levels must be within the
range of the average value to the
maximum (or minimum) value allowed.
The average value is defined as the sum
of all one-minute averages, divided by
the number of one-minute averages over
the previous 18 months (30 months for
small, on-site facilities defined in
§63.1208(b)(l)(ii));
(ii) Each operating limit established to
maintain compliance with the D/F
emission standard must be held within
the range of the average value over the
previous 18 months (30 months for
small, on-site facilities defined in
§ 63.1208(b)(l)(ii)) and the maximum or
minimum, as appropriate, that is
allowed; and
(iii) The source must feed
representative types (or types that may
result in higher emissions than normal)
of organic compounds in the waste (e.g.,
aromatics, aliphatics, nitrogen content,
halogen/carbon ratio, oxygen/carbon
ratio), and chlorine must be fed at
normal feedrates or greater.
(e) Notification of performance test
and approval of test plan. The
provisions of § 63.7 (b) and (c) apply.
Notwithstanding the Administrator's
approval or disapproval, or failure to
approve or disapprove the test plan, the
owner or operator must comply with all
applicable requirements of this part,
including deadlines for submitting the
initial and subsequent notifications of
compliance.
(f) Performance testing facilities. The
provisions of § 63.7(d) apply.
(g) Notification of compliance. Within
90 days of completion of the ,
performance test, the owner or operator
must postmark a notification of
compliance documenting compliance
with the emission standards and CMS
requirements, and identifying
applicable operating limits. See § 63.7(g)
for additional requirements.
(h) Failure to submit a timely
notification of compliance. If an owner
or operator determines (based on CEM
recordings, results of analyses of stack
samples, or results of CMS performance
evaluations) that the source has failed
any emission standard during the
performance test for a mode of
operation, it is a violation of the
standard and hazardous waste burning
must cease immediately under that
mode of operation. Hazardous waste
burning could not be resumed under
that mode of operation, except for
purposes of pretesting or comprehensive
performance testing and for a maximum
-------
17520
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
of 720 hours, until a notification of
compliance is submitted subsequent to
a new comprehensive performance test.
(i) Waiver of performance test. The
following waiver provision applies in
lieu of § 63.7(h). Performance tests are
not required to document compliance
with the following standards under the
conditions specified and provided that
the required information is submitted to
the Administrator for review and
approval with the site-specific test plan
as required by paragraph (e) of this
section:
(1) Mercury. The owner or operator is
deemed to be in compliance with the
mercury emission standard (and
monitoring Hg emissions with a CEMS
is not required) if the maximum
possible emission concentration
determined as specified below does not
exceed the emission standard:
(i) Establish a maximum feedrate of
mercury from all feedstreams, and
monitor and record the feedrate
according to § 63.1210(c);
(ii) Establish a minimum stack gas
flow rate, or surrogate for gas flow rate,
monitor the parameter with a CMS and
record the data, and interlock the limit
on the parameter with the automatic
waste feed cutoff system;
(iii) Calculate a maximum possible
emission concentration assuming all
mercury from all feedstreams is emitted.
(2) SVM (semivolatile metals). The
owner or operator is deemed to be in
compliance with the SVM (cadmium
and lead, combined) emission standard
if the maximum possible emission
concentration determined as specified
below does not exceed the emission
standard:
(i) Establish a maximum feedrate of
cadmium and lead, combined, from all
feedstreams, and monitor and record the
feedrate according to § 63.1210(c);
(ii) Establish a minimum stack gas
flow rate, or surrogate for gas flow rate,
monitor the parameter with a CMS and
record the data, and interlock the limit
on the parameter with the automatic
waste feed cutoff system;
(iii) Calculate a maximum possible
emission concentration assuming all
cadmium and lead from all feedstreams
is emitted.
(3) LVM (low volatility metals). The
owner or operator is deemed to be in
compliance with the LVM (arsenic,
beryllium, chromium, and antimony,
combined) emission standard if the
maximum possible emission
concentration determined as specified
below does not exceed the emission
standard:
(i) Establish a maximum feedrate of
arsenic, beryllium, chromium, and
antimony, combined, from all
feedstreams, and monitor and record the
feedrate according to § 63.1210(c);
(ii) Establish a minimum stack gas
flow rate, or surrogate for gas flow rate,
monitor the parameter with a CMS and
record the data, and interlock the limit
on the parameter with the automatic
waste feed cutoff system;
(iii) Calculate a maximum possible
emission concentration assuming all
LVM from all feedstreams is emitted.
(4) HC1/C12. The owner or operator is
deemed to be in compliance with the
HC1/C12 emission standard if the
maximum possible emission
concentration determined as specified
below does not exceed the emission
standard:
(i) Establish a maximum feedrate of
total chlorine and chloride from all
feedstreams, and monitor and record the
feedrate according to § 63.1210(c);
(ii) Establish a minimum stack gas
flow rate, or surrogate for gas flow rate,
monitor the parameter with a CMS and
record the data, and interlock the limit
on the parameter with the automatic
waste feed cutoff system;
(iii) Calculate a maximum possible
emission concentration assuming all
total chlorine and chloride from all
feedstreams is emitted.
§ 63.1209 Test methods.
(a) Dioxins and furans. (1) Method
0023A, provided by SW-846
(incorporated by reference in § 260.11 of
this chapter), shall be used to determine
compliance with the emission standard
for dioxin and furans;
(2) If the sampling period for each run
is six hours or greater, nondetects shall
be assumed to be present at zero
concentration. If the sampling period for
any run is less than six hours,
nondetects shall be assumed to be
present at the level of detection for all
runs.
(b) Mercury. Method 0060, provided
by SW-846 (incorporated by reference
in § 260.11 of this chapter), shall be
used to evaluate the mercury CEMS as
required by § 63.1210.
(c) Cadmium and lead. Method 0060,
provided by SW—846 (incorporated by
reference in § 260.11 of this chapter),
shall be used to determine compliance
with the emission standard for cadmium
and lead or to calibrate and/or evaluate
a CEMS as provided by § 63.1210.
(d) Arsenic, beryllium, chromium,
and antimony. Method 0060, provided
by SW-846 (incorporated by reference
in § 260.11 of this chapter), shall be
used to determine compliance with the
emission standard for arsenic,
beryllium, chromium, and antimony or
to calibrate and/or evaluate a CEMS as
provided by § 63.1210.
(e) HC1 and chlorine gas. Methods
0050, 0051, and 9057, provided by SW-
846 (incorporated by reference in
§ 260.11 of this chapter), shall be used
to determine compliance with the
emission standard for HC1 and C12
(combined) or to calibrate and/or
evaluate the HC1 and chlorine gas CEMS
as provided by § 63.1210.
(f) Particulate Matter. Method 5 in
appendix A of part 60 shall be used to
calibrate and/or evaluate a PM CEMS as
provided by § 63.1210.
(g) Feedstream Analytical methods.
Analytical methods used to determine
feedstream concentrations of metals,
halogens, and other constituents shall
be those provided by SW-846
(incorporated by reference in § 260.11 of
this chapter.)
Alternate methods may be used if
approved in advance by the Director.
§ 63.1210 Monitoring requirements.
(a) Continuous emissions monitors
(CEMS). (1) HWCs shall be equipped
with CEMS for PM, Hg, CO, HC, and O2
for compliance monitoring, except as
provided by paragraph (a)(3). Owners
and operators may elect to use CEMS for
compliance monitoring for SVM, LVM,
HC1, and C12.
(2) At all times that hazardous waste
is fed into the HWC or remains in the
combustion chamber, the CEMS must be
operated in compliance with the
appendix to this subpart.
(3) Waiver of CEMS requirement for
mercury. The following waiver
provision applies in lieu of § 63.7(h). A
mercury CEMS is not required to
document compliance with the mercury
standard under the conditions specified
and provided that the required
information is submitted to the
Administrator for review and approval
with the site-specific test plan as
required by § 63.1209(e). The owner or
operator is deemed to be in compliance
with the mercury emission standard if
the maximum possible emission
concentration determined as specified
below does not exceed the emission
standard:
(i) Establish a maximum feedrate of
mercury, combined, from all
feedstreams, and monitor and record the
feedrate according to § 63.1210(c);
(ii) Establish a minimum stack gas
flow rate, or surrogate for gas flow rate,
monitor the parameter with a CMS and
record the data, and interlock the limit
on the parameter with the automatic
waste feed cutoff system;
(iii) Calculate a maximum possible
emission concentration assuming all
mercury from all feedstreams is emitted.
(b) Other continuous monitoring
systems. (1) CMS other than CEMS (e.g.,
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17521
thermocouples, pressure transducers,
flow meters) must be used to document
compliance with the applicable
operating limits provided by this
section.
(2) Non-CEMS CMS must be installed
and operated in conformance with
§63.8(c)(3) requiring the owner and
operator, at a minimum, to comply with
the manufacturer's written
specifications or recommendations for
installation, operation, and calibration
of the system.
(3) Non-CEMS CMS must sample the
regulated parameter without
interruption, and evaluate the detector
response at least once each 15 seconds,
and compute and record the average
values at least every 60 seconds.
(4) The span of the detector must not
be exceeded. Span limits shall be
interlocked into the automatic waste
feed cutoff system required by
§63.1207(a)(2).
(c) Analysis of feedstreams. (1)
General. The owner or operator must
obtain an analysis of each feedstream
prior to feeding the material that is
sufficient to document compliance with
the applicable feedrate limits provided
by this section.
(2) Feedstream analysis plan. The
owner or operator must develop and
implement a feedstream analysis plan
and record it in the operating record.
The plan must specify at a minimum:
(i) The parameters for which each
feedstream will be analyzed to ensure
compliance with the operating limits of
this section;
(ii) Whether the owner or operator
will obtain the analysis by performing
sampling and analysis, or by other
methods such as using analytical
information obtained from others or
using other published or documented
data or information;
(iii) How the analysis will be used to
document compliance with applicable
feedrate limits (e.g., if hazardous wastes
are blended and analyses are obtained of
the wastes prior to blending but not of
the blended, as-fired, waste, the plan
must describe how the owner and
operator will determine the pertinent
parameters of the blended waste);
(iv) The test methods which will be
used to obtain the analyses;
(v) The sampling method which will
be used to obtain a representative
sample of each feedstream to be
analyzed using sampling methods
described in appendix I, part 261, of this
chapter, or an equivalent method; and
(vi) The frequency with which the
initial analysis of the feedstream will be
reviewed or repeated to ensure that the
analysis is accurate and up to date.
(3) Review and approval of analysis
plan. The owner and operator must
submit the feedstream analysis plan to
the Administrator for review and
approval, if requested.
(4) Compliance with feedrate limits.
To comply with the applicable feedrate
limits of this section, feedrates must be
monitored and recorded as follows:
(i) Determine and record the value of
the parameter for each feedstream by
sampling and analysis or other method;
(ii) Determine and record the mass or
volume flowrate of each feedstream by
a CMS. If flowrate of a feedstream is
determined by volume, the density of
the feedstream shall be determined by
sampling and analysis and shall be
recorded (unless the constituent
concentration is reported in units of
weight per unit volume (e.g., mg/1));
(iii) Calculate and record the mass
feedrate of the parameter per unit time.
(d) Performance evaluations. (1) The
requirements of §63.8(d) (Quality
control program) and (e) (Performance
evaluation of continuous monitoring
systems) apply, except that performance
evaluations of components of the CMS
shall be conducted under the frequency
and procedures (for example, submittal
of performance evaluation test plan for
review and approval) applicable to
performance tests as provided by
§63.1208.
(2) Performance specifications and
evaluations of GEMS are prescribed in
the appendix to this subpart.
(e) Conduct of monitoring. The
provisions of § 63.8(b) apply.
(f) Operation and maintenance of
continuous monitoring systems. The
provisions of § 63.8(c) are superseded by
this section, except that paragraphs
(c)(2), (c)(3), and (c)(6) are applicable.
(g) [Reserved]
(h) Use of an alternative monitoring
method. The provisions of § 63.8(0
apply.
(i) Reduction of monitoring data. The
provisions of § 63.8(g) apply, except for
paragraphs (g)(2) and (g)(5).
(j) Dioxins and furans. To remain in
compliance with the emission standard
for dioxins and furans, the owner or
operator shall establish operating h'mits
for the following parameters and
comply with those limits at all times
that hazardous waste is fed or that
hazardous waste remains in the
combustion chamber:
(1) Maximum temperature at the dry
PM control device. If a source is
equipped with an electrostatic
precipitator, fabric filter, or other dry
emissions control device where
particulate matter is collected and
retained in contact with combustion gas,
the maximum allowable temperature at
the inlet to the first such control device
in the air pollution control system must
be established and complied with as
follows:
(i) A 10-minute rolling average shall
be established as the average over all
runs of the highest 10-minute rolling
average for each run;
(ii) An hourly rolling average shall be
established as the average level over all
runs.
(2) Minimum combustion chamber
temperature, (i) The temperature of each
combustion chamber shall be measured
at a location as close to, and as
representative of, each combustion
chamber as practicable;
(ii) A 10-minute rolling average shall
be established as the average over all
runs of the minimum 10-minute rolling
average for each run; and
(iii) An hourly rolling average shall be
established as the average level over all
runs.
(3) Maximum flue gas flowrate or
production rate. As an indicator of gas
residence time in the combustion
chamber, the maximum flue gas
flowrate, or a parameter that the owner
or operator documents in the site-
specific test plan is an appropriate
surrogate, shall be established as the
average over all runs of the maximum
hourly rolling average for each run, and
complied with on a hourly rolling
average basis.
(4) Maximum hazardous waste
feedrate. The maximum hazardous
waste feedrate shall be established as
the average over all runs of the
maximum hourly rolling average for
each run, and complied with on a
hourly rolling average basis. A
maximum waste feedrate shall be
established for each waste feed point.
(5) Batch size, feeding frequency, and
minimum oxygen, (i) Except as
provided below, HWCs that feed a
feedstream in a batch (e.g., ram fed
systems) or container must comply with
the following:
(A) The maximum batch size shall be
the mass of that batch with the lowest
mass fed during the comprehensive
performance test;
(B) The minimum batch feeding
frequency (i.e., the minimum period of
time between batch or container
feedings) shall be the longest interval of
time between batch or container
feedings during the comprehensive
performance test; and
(C) The minimum combustion zone
oxygen content at the time of firing the
batch or container shall be the highest
instantaneous oxygen level observed at
the time any batch or container was fed
during the comprehensive performance
test.
-------
27522 Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
(ii) Cement kilns that fire containers
of material into the hot, clinker
discharge end of the kiln are exempt
from the requirements of this paragraph
provided the owner or operator
documents in the operating record:-
(A) The volume of each container
does not exceed 1 gallon; and
(B) The frequency of firing the
containers does not exceed the rate
occurring during the comprehensive
performance test.
(6) PM limit, (i) PM shall be limited
to the level achieved during the
comprehensive performance test;
(ii) During the comprehensive
performance test the owner and operator
shall demonstrate compliance with the
PM standards in §§ 63.1203, 63.1204,
and 63.1205, corrected to 7 percent
oxygen, based on a 2-hour rolling
average, and monitored with a GEMS;
(A) The owner or operator shall
install, calibrate, maintain, and
continuously operate a GEMS that
measures particulate matter at all times
that hazardous waste is fed or that
hazardous waste remains in the
combustion chamber.
(B) The PM GEMS shall meet the
requirements provided in the appendix
to this subpart.
(iii) The site-specific PM limit shall be
determined from the performance test as
follows:
(A) A 10-minute rolling average shall
be established as the average over all
runs of the maximum 10-minute rolling
average for each run;
(B) An hourly rolling average shall be
established as the average of all one
minute averages over all runs.
(7) Carbon injection parameters. If
carbon injection is used:
(i) Injection rate. Minimum carbon
injection rates shall be established as:
(A) A 10-minute rolling average
established as the average over all runs
of the minimum 10-minute rolling
average for each run; and
(B) An hourly rolling average
established as the average level over all
runs.
(ii) Carrier fluid. Minimum carrier
fluid (gas or liquid) flowrate or pressure
drop shall be established as a 10-minute
rolling average based on the carbon
injection system manufacturer's
specifications.
(iii) Carbon specification. (A) The
brand (i.e., manufacturer) and type of
carbon used during the comprehensive
performance test must be used until a
subsequent comprehensive performance
test is conducted, unless the owner or
operator document in the site-specific
performance test plan required under
§ 63.1208 key parameters that affect
adsorption and establish limits on those
parameters based on the carbon used in
the performance test.
(B) The owner or operator may
request approval from the Administrator
at any time to substitute a different
brand or type of carbon without having
to conduct a comprehensive
performance test. The Administrator
may grant such approval if he or she
determines that the owner or operator
has sufficiently documented that the
substitute carbon will provide the same
level of dioxin and furan control as the
original carbon.
(8) Carbon bed. If a carbon bed is
used, a carbon replacement rate must be
established as follows:
(i) Testing Requirements. Testing of
carbon beds shall be done in the
following manner:
(A) Initial comprehensive
performance test. For the initial
comprehensive performance test, the
carbon bed shall be used in accordance
with manufacturer's specifications. No
aging of the carbon is required.
(B) Confirmatory tests prior to
subsequent comprehensive tests. For
confirmatory tests after the initial but
prior to subsequent comprehensive
tests, the facility shall follow the normal
change-out schedule specified by the
carbon bed manufacturer.
(C) Subsequent comprehensive tests.
The age of the carbon in the carbon bed
shall be determined as the length of
time since carbon was most recently
added and the amount of time the
carbon that has been in the bed the
longest.
(ii) Determination of maximum
allowable carbon age. (A) Prior to
subsequent comprehensive performance
tests, the manufacturer shall follow the
manufacturer's suggested change-out
interval for replacing used carbon with
unused carbon.
(B) After the second comprehensive
test the maximum allowable age of a
carbon bed shall be the amount of time
since carbon has most recently been
added and the amount of time that the
carbon the has been in the bed the
longest, based on what those two time
intervals were during the
comprehensive performance test.
(iii) Carbon specification. (A) The
brand (i.e., manufacturer) and type of
carbon used during the comprehensive
performance test must be used until a
subsequent comprehensive performance
test is conducted, unless the owner or
operator document in the site-specific
performance test plan required under
§ 63.1208 key parameters that affect
adsorption and establish limits on those
parameters based on the carbon used in
the performance test.
(B) The owner or operator may
request approval from the Administrator
at any time to substitute a different
brand or type of carbon without having
to conduct a comprehensive
performance test. The Administrator
may grant such approval if he or she
determines that the owner or operator
has sufficiently documented that the
substitute carbon will provide the same
level of dioxin and furan control as the
original carbon.
(7) Catalytic oxidizer. If a catalytic
oxidizer is used, the following
parameters shall be established:
(i) Minimum flue gas temperature at
the entrance of the catalyst. A minimum
flue gas temperature at the entrance of
the catalyst shall be established as
follows:
(A) A 10-minute average shall be
established as the average over all runs
of the minimum temperature 10-minute
rolling average for each run;
(B) An hourly average shall be
established as the average level over all
runs.
(ii) Maximum time in-use. A catalytic
oxidizer shall be replaced with a new
catalytic oxidizer when it has reached
the maximum service tune specified by
the manufacturer.
(iii) Catalyst replacement
specifications. When a catalyst is
replaced with a new one, the new
catalyst shall be identical to the one
used during the previous
comprehensive test, including:
(A) Catalytic metal loading for each
metal;
(B) Space time, expressed in the units
s"1, the maximum rated volumetric
flow of the catalyst divided by the
volume of the catalyst;
(C) Substrate construction, including
materials of construction, washcoat
type, and pore density.
(iv) Maximum flue gas temperature.
Maximum flue gas temperature at the
entrance of the catalyst shall be
established as a 10-minute rolling
average, based on manufacturer's
specifications.
(8) Inhibitor feedrate. If a dioxin
inhibitor is fed into the unit, the
following parameters shall be
established:
(i) Minimum inhibitor feedrate.
Minimum inhibitor feedrate shall be
established as:
(A) A 10-minute rolling average shall
be established as the average over all
runs of the minimum 10-minute rolling
average for each run;
(B) An hourly average shall be
established as the average level over all
runs.
(ii) Inhibitor specifications. (A) The
brand (i.e., manufacturer) and type of
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17523
inhibitor used during the
comprehensive performance test must
be used until a subsequent
comprehensive performance test is
conducted, unless the owner or operator
document in the site-specific
performance test plan required under
§ 63.1208 key parameters that 'affect the
effectiveness of a D/F inhibitor and
establish limits on those parameters
based on the inhibitor used in the
performance test.
(B) The owner or operator may
request approval from the Administrator
at any time to substitute a different
brand or type of inhibitor without
having to conduct a comprehensive
performance test. The Administrator
may grant such approval if he or she
determines that the owner or operator
has sufficiently documented that the
substitute inhibitor will provide the
same level of dioxin and furan control
as the original inhibitor.
(k) Mercury GEMS. (1) The owner or
operator shall install, calibrate,
maintain, and continuously operate a
GEMS for mercury at all times that
hazardous waste is fed or that hazardous
xvaste remains in the combustion
chamber.
(2) The mercury GEMS shall meet
Performance Specification 10, if the
GEM measures other metals as well as
mercury, or Performance Specification
12, if the GEM measures only mercury.
Both performance specifications are
provided in the appendix to this
subpart.
(a) The owner and operator shall
comply with the quality assurance
procedures provided in the appendix to
this subpart.
(1) Semivolatile metals (SVM). The
owner or operator shall demonstrate
compliance with the SVM (cadmium
and lead) emission standard by either:
(1) GEMS, (i) Installing, calibrating,
maintaining, and continuously
operating a GEMS that measures
multiple metals at all times that
hazardous waste is fed or remains in the
combustion chamber.
(ii) The multi-metal GEMS shall meet
the requirements provided in the
appendix to this subpart; or
(2) Operating limits. Establishing and
complying with the following operating
limits, except that cement kilns and
lightweight aggregate kilns must comply
with alternative requirements provided
by paragraph (f) of this section:
(0 PM limit. (A) PM shall be limited
to the level achieved during the
comprehensive performance test;
(B) During the comprehensive
performance test the owner and operator
shall demonstrate compliance with the
applicable PM standard in §§ 63.1203,
63.1204, and 63.1205, corrected to 7
percent oxygen, based on a 2-hour
rolling average, and monitored with a
GEMS;
(i) The owner or operator shall
install, calibrate, maintain, and
continuously operate a GEMS that
measures particulate matter at all times
that hazardous waste is fed or that
hazardous waste remains in the
combustion chamber.
(2) The PM GEMS shall meet the
requirements provided in the appendix
to this subpart.
(C) The site-specific PM limit shall be
determined from the performance test as
follows:
(1) A 10-minute rolling average shall
be established as the average over all
runs of the maximum 10-minute rolling
average for each run;
(2) An hourly rolling average shall be
established as the average of all one
minute averages over all runs.
(ii) Maximum feedrate of Cd and Pb.
A 12-hour rolling average limit for the
feedrate of Cd and Pb, combined, in all
feedstreams shall be established as the
average feedrate over all runs.
(iii) Maximum total chlorine and
chloride feedrate. A 12-hour rolling
average limit for the feedrate of total
chlorine and chloride in all feedstreams
shall be established as the average
feedrate over all runs.
(iv) Minimum gas flowrate. An hourly
rolling average limit for gas flowrate, or
a surrogate parameter, shall be
established as the average over all runs
of the lowest hourly rolling average for
each run.
(m) Low volatility metals (LVM). The
owner or operator shall demonstrate
compliance with the LVM (arsenic,
beryllium, chromium, and antimony)
emission standard by either:
(1) GEMS, (i) Installing, calibrating,
maintaining, and continuously
operating a GEMS that measures
multiple metals at all times that
hazardous waste is fed or remains in the
combustion chamber.
(ii) The multi-metals GEMS shall meet
the requirements provided in the
appendix to this subpart; or
(2) Operating limits. Establishing and
complying with the following operating
limits, except that cement kilns and
lightweight aggregate kilns must comply
with alternative requirements provided
by paragraph (f) of this section:
(i) PM limit. (A) PM shall be limited
to the level achieved during the
comprehensive performance test;
(B) During the comprehensive
performance test the owner and operator
shall demonstrate compliance with, the
applicable PM standard in §§ 63.1203,
63.1204, or 63.1205, corrected to 7
percent oxygen, based on a 2-hour
rolling average, and monitored with a
GEMS;
(1) The owner or operator shall
install, calibrate, maintain, and
continuously operate a GEMS that
measures particulate matter at all times
that hazardous waste is fed or that
hazardous waste remains in the
combustion chamber.
(2) The PM GEMS shall meet the
requirements provided in the appendix
to this subpart.
(C) The site-specific PM limit shall be
determined from the performance test as
follows:
(1) A 10-minute rolling average shall
be established as the average over all
runs of the maximum 10-minute rolling
average for each run;
(2) An hourly rolling average shall be
established as the average of all one
minute averages over all runs.
(ii) Maximum feedrate of As, Be, Cr,
and Sb. (A) A 12-hour rolling average
limit for the feedrate of As, Be, Cr, and
Sb, combined, in all feedstreams shall
be established as the average feedrate
over all runs.
(B) A 12-hour rolling average limit for
the feedrate of As, Be, Cr, and Sb,
combined, in all pumpable feedstreams
shall be established as the average
feedrate in pumpable feedstreams over
all runs.
(iii) Maximum chlorine and chloride
feedrate. A 12-hour rolling average limit
for the feedrate of total chlorine and
chloride in all feedstreams shall be
established as the average feedrate over
all runs.
(iv) Minimum gas flowrate. An hourly
rolling average limit for gas flowrate, or
a surrogate parameter, shall be
established as the average over all runs
of the lowest hourly rolling average for
each run.
(n) Special requirements for CKs and
LWAKs for compliance with metals
standards. Owners and operators of
cement kilns and lightweight aggregate
kilns that recycle collected particulate
matter back into the kiln must comply
with one of the following alternative
approaches to demonstrate compliance
with the emission standards for SVM,
combined (cadmium and lead), and for
LVM, combined (arsenic, beryllium,
chromium and antimony):
(1) Feedstream monitoring. The
requirements of paragraphs (d) and (e)
of this section only after the kiln system
has been conditioned to enable it to
reach equilibrium with respect to metals
fed into the system and metals
emissions. During conditioning,
hazardous waste and raw materials
having the same metals content as will
be fed during the performance test must
-------
17524
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
be fed at the feedrates that will be fed
during the performance test; or
(2) Monitor recycled PM. The special
testing requirements prescribed in
"Alternative Method for Implementing
Metals Controls" in appendix IX, part
266, of this chapter; or
(3) Semicontinuous emissions testing.
Stack emissions testing for a minimum
of 6 hours each day while hazardous
waste is burned. The testing must be
conducted when burning normal
hazardous waste for that day at normal
feedrates for that day and when the air
pollution control system is operated
under normal conditions. Although
limits on metals in feedstreams are not
established under this option, the owner
or operator must analyze each
feedstream for metals content
sufficiently to determine if changes in
metals content may affect the ability of
the facility to meet the metal emissions
standards under §§ 63.1204 and
63.1205.
(o) HC1 and chlorine gas. The owner
or operator shall demonstrate
compliance with the HC1/C12 emission
standard by either:
(1) GEMS, (i) Installing, calibrating,
maintaining, and continuously
operating a GEMS for HC1 and C12 at all
times that hazardous waste is fed or that
hazardous waste remains in the
combustion chamber.
(ii) The HC1 and C12 GEMS shall meet
the requirements provided in the
appendix to this subpart; or
(2) Operating limits. Establishing and
complying with the following operating
limits:
(i) Feedrate of total chlorine and
chloride. A 12-hour rolling average limit
for the total feedrate of total chlorine
and chloride in all feedstreams shall be
established as the average feedrate over
all runs.
(ii) Maximum flue gas flowrate or
production rate. As an indicator of gas
residence time in the control device, the
maximum flue gas flowrate, or a
parameter that the owner or operator
documents in the site-specific test plan
is an appropriate surrogate, shall be
established as the average over all runs
of the maximum hourly rolling average
for each run, and complied with on a
hourly rolling average basis.
(iii) Wet Scrubber. If a wet scrubber is
used, the following operating parameter
limits shall be established.
(A) Minimum pressure drop across
the scrubber. Minimum pressure drop
across a wet scrubber shall be
established.
(2) A 10-minute rolling average shall
be established as the average over all
runs of the minimum 10-minute rolling
averages for each run.
(2) An hourly rolling average shall be
established as the average level over all
runs.
(B) Minimum liquid feed pressure.
Minimum liquid feed pressure shall be
established as a ten minute average,
based on manufacturer's specifications.
(C) Minimum liquid pH. Minimum
liquid pH shall be established.
(1) A 10-minute rolling average shall
be established as the average over all
runs of the minimum 10-minute rolling
averages for each run.
(2) An hourly rolling average shall be
established as the average level over all
runs.
(D) Minimum liquid to gas flow ratio.
Minimum liquid to gas flow ratio shall
be established.
(1) A. 10-minute rolling average shall
be established as the average over all
runs of the minimum 10-minute rolling
averages for each run.
(2) An hourly rolling average shall be
established as the average level over all
runs.
(iv) Ionizing Wet Scrubber. If an
ionizing wet scrubber is used, the
following operating parameter limits
shall be established.
(A) Minimum pressure drop across
the scrubber. Minimum pressure drop
across an ionizing wet scrubber shall be
established on both a ten minute and
hourly rolling average.
(I) A 10-minute rolling average shall
be established as the average over all
runs of the minimum 10-minute rolling
averages for each run.
(2) An hourly rolling average shall be
established as the average level over all
runs.
(B) Minimum liquid feed pressure.
Minimum liquid feed pressure shall be
established as a ten minute average,
based on manufacturer's specifications.
(C) Minimum liquid to gas flow ratio.
Minimum liquid to gas flow ratio shall
be established on both a ten minute and
hourly rolling average.
(1) A 10-minute rolling average shall
be established as the average over all
runs of the minimum 10-minute rolling
averages for each run.
(2) An hourly rolling average shall be
established as the average level over all
runs.
(v) Dry scrubber. If a dry scrubber is
used, the following operating parameter
limits shall be established.
(A) Minimum sorbent feedrate.
Minimum sorbent feedrate shall be
established on both a ten minute and
hourly rolling average.
(1) A 10-minute rolling average shall
be established as the average over all
runs of the minimum 10-minute rolling
averages for each run.
(2) An hourly rolling average shall be
established as the average level over all
runs.
(B) Minimum carrier fluid flowrate or
nozzle pressure drop. Minimum carrier
fluid (gas or liquid) flowrate or nozzle
pressure drop shall be established as a
ten minute average, based on
manufacturer's specifications.
(C) Sorbent specifications. (1) The
brand (i.e., manufacturer) and type of
sorbent used during the comprehensive
performance test must be used until a
subsequent comprehensive performance
test is conducted, unless the owner or
operator document in the site-specific
performance test plan required under
§ 63.1208 key parameters that affect the
effectiveness of a sorbent and establish
limits on those parameters based on the
inhibitor used in the performance test.
(2) The owner or operator may request
approval from the Administrator at any
time to substitute a different brand or
type of inhibitor without having to
conduct a comprehensive performance
test. The Administrator may grant such
approval if he or she determines that the
owner or operator has sufficiently
documented that the substitute sorbent
will provide the same level of HC1 and
C12 control as the original sorbent.
(p) Carbon monoxide GEMS. (1) The
owner or operator shall install, calibrate,
maintain, and continuously operate a
GEMS for carbon monoxide at all times
that hazardous waste is fed or that
hazardous waste remains in the
combustion chamber.
(2) The carbon monoxide GEMS shall
meet the requirements provided in the
appendix to this subpart.
(q) Hydrocarbon GEMS. (1) The owner
or operator shall install, calibrate,
maintain, and continuously operate a
GEMS for hydrocarbons at all times that
hazardous waste is fed or that hazardous
waste remains in the combustion
chamber.
(2) The hydrocarbon GEMS shall meet
the requirements provided in the
appendix to this subpart.
(r) Oxygen GEMS. (1) The owner or
operator shall install, calibrate,
maintain, and continuously operate a
GEMS for oxygen at all times that
hazardous waste is fed or remains in the
combustion chamber.
(2) The oxygen GEMS shall meet the
requirements provided in the appendix
to this subpart.
(s) Maximum combustion chamber
pressure. If a source complies with the
fugitive emissions requirements of
§ 63.1207(b) by maintaining the
maximum combustion chamber zone
pressure lower than ambient pressure,
the source must monitor the pressure
instantaneously and the automatic
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules 17525
waste feed cutoff system must be
engaged when negative pressure is not
maintained at any time.
(t) Waiver of operating limits. The
owner or operator may request in
writing a waiver from any of the
operating limits provided by this
section. The waiver must include
documentation that other operating
parameters or methods to establish
operating limits are more appropriate to
ensure compliance with the emission
standards. The waiver must also include
recommended averaging periods and the
basis for establishing operating limits.
§63.1211 Notification requirements.
(a) Notifications. HWCs shall submit
the following notifications as
applicable:
(1) Initial notification. HWCs shall
comply with the initial notification
requirements of § 63.9(b).
(2) Notification of performance test
and CMS evaluation. The notification of
performance test requirements of
§ 63.9(c) apply to all performance tests
and CMS evaluations required by
§ 63.1208, except that all notifications
shall be submitted for review and
approval at the times specified in that
section.
(3) Notification of compliance. The
notification of compliance status
requirements of § 63.9(h) apply, except
that:
(i) The notification is a notification of
compliance (rather than compliance
status), as defined in § 63.1200;
(ii) The notification is required for
each performance test;
(SiiJ The requirements of § 63.9{h)(2)(i)
(D) and (E) pertaining to major source
determinations do not apply; and
(iv) Under § 63.9(h)(2)(ii), the
notification shall be sent before the
close of business on the 90th day
following the completion of relevant
compliance demonstration activity
specified in this subpart.
(4) Request for extension of time to
submit a notification of compliance.
HWCs that elect to request a tune
extension of up to one year to submit an
initial notification of compliance under
§ 63.9(c) or a subsequent notification of
compliance under § 63.1208(c) must
submit a written request and
justification as required by those
sections.
(b) Applicability of § 63.9
(Notification requirements). The
following provisions of § 63.9 are
applicable to HWCs:
CD Paragraphs (a), (b), (c), (d), (e), (g),
(i), and (j); and
(2) Paragraph (h), except as provided
in paragraphs (a)(3) (iii) and (iv) of this
section.
§ 63.1212 Recordkeeping and reporting
requirements.
(a) The following provisions of § 63.10
are applicable to HWCs:
(1) Paragraph (a) (Applicability and
general information), except (a)(2);
(2) Paragraph (b) (General
recordkeeping requirements), except
(b)(2) (iv) through (vi), and (b)(3); and
(3) Paragraph (c) (Additional
recordkeeping requirements for sources
with CMS), except (c)(6) through (8),
(c)(13), and (c)(15).
(4) Paragraph (d) (General reporting
requirements) applies as follows:
(i) Paragraphs fd)(l), (d)(4) apply; and
(ii) Paragraph (d)(2) applies, except
that the report may be submitted up to
90 days after completion of the test; and
(5) In paragraph (e) (Additional
reporting requirements for sources with
CMS), paragraphs (e)(l) (General) and
(e)(2) (Reporting results of CMS
performance evaluations) apply.
(b) Additional reporting requirements.
HWCs are also subject to the reporting
requirements for excessive automatic
waste feed cutoffs under § 63.1207(a)(2)
and emergency safety vent openings
under § 63.1207(a)(3).
(c) Additional recordkeeping
requirements. HWCs must also retain
the feedstream analysis plan required
under § 63.1210(c) in the operating
record.
Appendix to Subpart EEE—Quality
Assurance Procedures for Continuous
Emissions Monitors Used for Hazardous
Waste Combustors
1. Applicability and Principle
1.1 Applicability. These quality
assurance requirements are used to evaluate
the effectiveness of quality control (QC) and
quality assurance (QA) procedures and the
quality of data produced by continuous
emission monitoring systems (GEMS) that are
used for determining compliance with the
emission standards on a continuous basis as
specified in the applicable regulation. The
QA procedures specified by these
requirements represent the minimum
requirements necessary for the control and
assessment of the quality of GEMS data used
to demonstrate compliance with the emission
standards provided under subpart EEE, part
63, of this chapter. Owners and operators
must meet these minimum requirements and
are encouraged to develop and implement a
more extensive QA program. These
requirements supersede those found in Part
60, Appendix F of this chapter. Appendix F
does not apply to hazardous waste-burning
devices.
Data collected as a result of the required
QA and QC measures are to be recorded in
the operating record. In addition, data
collected as a result of GEM performance
evaluations required by Section 5 in
conjunction with an emissions performance
test are to be submitted to the Director as
provided by § 63.8(e)(5) of this chapter.
These data are to be used by both the Agency
and the GEMS operator in assessing the
effectiveness of the GEMS QA and QC
procedures in the maintenance of acceptable
GEMS operation and valid emission data.
1.2 Principle. The QA procedures consist
of two distinct and equally important
functions. One function is the assessment of
the quality of the GEMS data by estimating
accuracy. The other function is the control
and improvement of the quality of the GEMS
data by implementing QC policies and
corrective actions. These two functions form
a control loop. When the assessment function
indicates that the data quality is inadequate,
the source must immediately stop burning
hazardous waste. The GEM data control effort
must be increased until the data quality is
acceptable before hazardous waste burning
can resume.
In order to provide uniformity in the
assessment and reporting of data quality, this
procedure explicitly specifies the assessment
methods for response drift and accuracy. The
methods are based on procedures included in
the applicable performance specifications
provided in Appendix B to Part 60. These
procedures also require the analysis of the
EPA audit samples concurrent with certain
reference method (RM) analyses as specified
in the applicable RM's.
Because the control and corrective action
function encompasses a variety of policies,
specifications, standards, and corrective
measures, this procedure treats QC
requirements in general terms to allow each
source owner or operator to develop a QC
system that is most effective and efficient for
the circumstances.
2. Definitions
2.1 Continuous Emission Monitoring
System (GEMS). The total equipment
required for the determination of a pollutant
concentration. The system consists of the
following major subsystems:
2.1.1 Sample Interface. That portion of
the GEMS used for one or more of the
following: sample acquisition, sample
transport, and sample conditioning, or
protection of the monitor from the effects of
the stack effluent.
2.1.2 Pollutant Analyzer. That portion of
the GEMS that senses the pollutant
concentration and generates a proportional
output.
2.1.3 Diluent Analyzer. That portion of
the GEMS that senses the diluent gas (O2) and
generates an output proportional to the gas
concentration.
2.1.4 Data Recorder. That portion of the
CEMS that provides a permanent record of
the analyzer output. The data recorder may
provide automatic data reduction .and CEMS
control capabilities.
2.2 Relative Accuracy (RA). The absolute
mean difference between the pollutant
concentration determined by the CEMS and
the value determined by the reference
method (RM) plus the 2.5 percent error
confidence coefficient of a series of test
divided by the mean of the RM tests or the
applicable emission limit.
2.3 Calibration Drift (CD). The difference
in the CEMS output readings from the
established reference value after a stated
-------
17526
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
period of operation during which no
unscheduled maintenance, repair, or
adjustment took place.
2.4 Zero Drift (ZD). The difference in
GEMS output readings at the zero pollutant
level after a stated period of operation during
which no unscheduled maintenance, repair,
or adjustment took place.
2.5 Tolerance Interval. The interval with
upper and lower limits within which are
contained a specified percentage of the
population with a given level of confidence.
2.6 Calibration Standard. Calibration
standards produce a known and unchanging
response when presented to the pollutant
analyzer portion of the GEMS, and are used
to calibrate the drift or response of the
analyzer.
2.7 Relative Accuracy Test Audit (RATA)
Comparison of GEMS measurements to
reference method measurements in order to
evaluate relative accuracy following
procedures and specification given in the
appropriate performance specification.
2.8 Absolute Calibration Audit (ACA).
Equivalent to calibration error (CE) test
defined in the appropriate performance
specification using NIST traceable calibration
standards to challenge the GEMS and assess
accuracy.
2.9 Response Calibration Audit (RCA).
For PM GEMS only, a check of stability of the
calibration relationship determined by
comparison of GEMS response to manual
gravimetric measurements.
2.10 Fuel Type. For the purposes of PM
CEMs, fuel type is defined as the physical
state of the fuel: gas, liquid, or solid.
2.11 Rolling Average. The average
emissions, based on some (specified) time
period, calculated every minute from a one-
minute average of four measurements taken
at 15-second intervals.
3. QA/QC Requirements
3.1 QC Requirements. Each owner or
operator must develop and implement a QC
program. At a minimum, each QC program
must include written procedures describing
in detail complete, step-by-step procedures
and operations for the following activities.
1. Checks for component failures, leaks,
and other abnormal conditions.
2. Calibration of CEMS.
3. CD determination and adjustment of
CEMS.
4. Integration of CEMS with the automatic
waste feed cutoff (AWECO) system.
5. Preventive Maintenance of CEMS
(including spare parts inventory).
6. Data recording, calculations, and
reporting.
7. Checks of record keeping.
8. Accuracy audit procedures, including
sampling and analysis methods.
9. Program of corrective action for
malfunctioning CEMS.
10. Operator training and certification.
11. Maintaining and ensuring current
certification or naming of cylinder gasses,
metal solutions, and particulate samples used
for audit and accuracy tests, daily checks,
and calibrations.
Whenever excessive inaccuracies occur for
two consecutive quarters, the current written
procedures must be revised or the CEMS
modified or replaced to correct the deficiency
causing the excessive inaccuracies. These
written procedures must be kept on record
and available for inspection by the
enforcement agency.
3.2 QA Requirements. Each source owner
or operator must develop and implement a
QA plan that includes, at a minimum, the
following.
1. QA responsibilities (including
maintaining records, preparing reports,
reviewing reports).
2. Schedules for the daily checks, periodic
audits, and preventive maintenance.
3. Check lists and data sheets.
4. Preventive maintenance procedures.
5. Description of the media, format, and
location of all records and reports.
6. Provisions for a review of the CEMS data
at least once a year. Based on the results of
the review, the owner or operator shall revise
or update the QA plan, if necessary.
4. CD and ZD Assessment and Daily System
Audit
4.1 CD and ZD Requirement. Owners and
operators must check, record, and quantify
the ZD and the CD at least once daily
(approximately 24 hours) in accordance with
the method prescribed by the manufacturer.
The CEMS calibration must, at a minimum,
be adjusted whenever the daily ZD or CD
exceeds the limits in the Performance
Specifications. If, on any given ZD and/or CD
check the ZD and/or CD exceed(s) two times
the limits in the Performance Specifications,
or if the cumulative adjustment to the ZD
and/or CD (see Section 4.2) exceed(s) three
times the limits in the Performance
Specifications, hazardous waste buring must
immediately cease and the CEMS must be
serviced and recalibrated. Hazardous waste
burning cannot resume until the owner or
operator documents that the CEMS is in
compliance with the Performance
Specifications by carrying out an ACA.
4.2 Recording Requirements for
Automatic ZD and CD Adjusting Monitors.
Monitors that automatically adjust the data to
the corrected calibration values must record
the unadjusted concentration measurement
prior to resetting the calibration, if
performed, or record the amount of the
adjustment.
4.3 Daily System Audit. The audit must
include a review of the calibration check
data, an inspection of the recording system,
an inspection of the control panel warning
lights, and an inspection of the sample
transport and interface system (e.g.,
flowmeters, filters, etc.) as appropriate.
4.4 Data Recording and Reporting. All
measurements from the CEMS must be
retained in the operating record for at least
5 years.
5. Performance Evaluation
5.1 Multi-Metals CEMS. The CEMS must
be audited at least once each calendar year.
In years when a performance test is also
required under §63.1208 of this chapter to
document compliance with emission
standards, the performance evaluation (i.e.,
audit) shall coincide with the performance
test. Successive yearly audits shall be at least
9 months apart. The audits shall be
conducted as follows.
5.1.1 Relative Accuracy Test Audit
(RATA). The RATA must be conducted at
least once every three years (five years for
small on-site facilities defined in
§ 63.1208(b)(l)(ii)). Conduct the RATA as
described in the RA test procedure (or
alternate procedures section) described in the
applicable Performance Specifications. In
addition, analyze the appropriate
performance audit samples received from the
EPA as described in the applicable sampling
methods (i.e., SW-846 method 0060).
5.1.2 Absolute Calibration Audit (ACA).
The ACA must be conducted at least once
each year except when a RATA is conducted
instead. Conduct an ACA using NIST
traceable calibration standards at three levels
for each metal that is being monitored for
compliance purposes. The levels must
correspond to 0 to 20, 40 to 60, and 80 to 120
percent of the applicable emission limit for
each metal. (For the SVM and LVM standards
where the standard applies to combined
emissions of several metals, the average
annual emission concentration for each
individual metal in a group for which a
standard applies should be assumed by
projecting emissions based on feedrate
estimates determined from the waste
management plan required under
§ 63.1210(c)(2) of this chapter. The estimated
average annual emission concentration
should be used as a surrogate metal emission
limit for purposes of the ACA.) At each level
and for each metal, make nine
determinations of the RA as defined in
Section 8 of the applicable Performance
Specifications using the value of the
calibration standard in the denominator of
Equation (6).
5.1.3 Reference method. The reference
method is SW-846 method 0060.
5.1.4 Excessive Audit Inaccuracy. If the
RA using the RATA or ACA exceeds the
criteria in Section 4.2 of the Performance
Specifications, hazardous waste burning
must immediately cease. Before hazardous
waste burning can resume, the owner or
operator must take necessary corrective
action to eliminate the problem, and must
audit the CEMS with a RATA to document
that the CEMS is operating within the
specifications.
5.2 Particulate Matter CEMS. The CEMS
must be audited at least once each quarter
(three calendar months.) A response
calibration audit (RCA) shall be conducted
every 18 months. An absolute calibration
audit (ACA) shall be conducted quarterly,
except when an RCA is conducted instead.
The audits shall be conducted as follows.
5.2.1 Response Calibration Audit (RCA).
The RCA must be conducted at least every 18
months (30 months for small on-site facilities
defined in § 63.1208(b)(l)(ii)). Conduct the
RCA as described in the CEMS Response
Calibration Procedure described in the
applicable Performance Specifications
(Sections 5 and 7). A minimum of nine tests
are required at three particulate levels. The
three particulate levels should be at the high-
end, low-end, and midpoint of the particulate
range spanned-by the current calibration of
the CEMS.
5.2.2 Absolute Calibration Audit (ACA).
The ACA must be conducted at least
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17527
quarterly, except when an RCA is conducted
instead. Conduct an ACA using NIST
traceable calibration standards, making three
measurements at three levels (nine
measurements total). The levels must
correspond to 10 to 50 percent, 80 to 120
percent, and 200 to 300 percent of the
emission limit. At each level make a
determination of the instrument response
and compare it to the nominal response by
calculating the calibration error CE:
Where:
RCEM is the GEMS response;
RN is the nominal response generated by the
calibration standard, and
REM is the emission limit value.
5.2.3 Excessive Audit Inaccuracy.
5.2.3.1 RCA. If less than 75 percent
percent of the test results from the RCA fall
within the tolerance interval established for
the current calibration (see Sections 7 and 8
of the Performance Specifications), then a
now calibration relation is required.
Hazardous waste burning must cease
immediately, and may not be resumed until
a new calibration relation is calculated from
the RCA data according to the procedures
specified in Section 8 of the Performance
Specifications.
5.2.3.2 ACA. If the calibration error is
greater than 2 percent of the emission limit
for any of the calibration levels, hazardous
waste burning must cease immediately. If
adjustments to the instrument reduce the
calibration error to less than 2 percent of the
emission limit at all three levels, then
hazardous waste burning can resume. If not,
the instrument must be repaired and must
pass a complete ACA before hazardous waste
burning can resume.
5.2.4 Calibrating for Fuel Type. The
owner or operator shall derive a sufficient
number of calibration curves to use for all
fuel type and mixtures of fuel type.
5.2.5 Reference Method. The reference
method is Method 5 found in 40 CFR Part 60,
Appendix A.
5.3 Total Mercury GEMS. An Absolute
Calibration Audit (ACA) must be conducted
quarterly, and a Relative Accuracy Test Audit
(RATA) must be conducted every three years
(five years for small on-site facilities defined
in §63.1208(b)(l)(ii)). An Interference
Response Tests shall be performed whenever
an ACA or a RATA is conducted. In years
when a performance test is also required
under § 63.1208 of this chapter to document
compliance with emission standards, the
RATA shall coincide with the performance
tost. The audits shall be conducted as
follows.
5.3.1 Relative Accuracy Test Audit
(RATA). The RATA must be conducted at
least every three years (five years for small
on-slte facilities defined in
§ 63.1208(b)(l)(ii)). Conduct the RATA as
described in the RA test procedure (or
alternate procedures section) described in the
applicable Performance Specifications. In
addition, analyze the appropriate
performance audit samples received from the
EPA as described in the applicable sampling
methods.
5.3.2 Absolute Calibration Audit (ACA).
The ACA must be conducted at least
quarterly except in a quarter when a RATA
is conducted instead. Conduct an ACA as
described in the calibration error (CE) test
procedure described in the applicable
Performance Specifications.
5.3.3 Interference Response Test. The
interference response test shall be conducted
whenever an ACA or RATA is conducted.
Conduct an interference response test as
described in the applicable Performance
Specifications.
5.3.4 Excessive Audit Inaccuracy. If the
RA from the RATA or the CE from the ACA
exceeds the criteria in the applicable
Performance Specifications, hazardous waste
burning must cease immediately. Hazardous
waste burning cannot resume until the owner
or operator take corrective measures and
audit the GEMS with a RATA to document
that the GEMS is operating within the
specifications.
5.3.5 Reference Methods. The reference
method for mercury is SW-846 method 0060.
5.4 Hydrogen Chloride (HC1), Chlorine
(C12), Carbon Monoxide (CO), Oxygen (O2),
and Hydrocarbon (HC) GEMS. An Absolute
Calibration Audit (ACA) must be conducted
quarterly, and a Relative Accuracy Test Audit
(RATA) (if applicable, see sections 5.4.1 and
5.4.2) must be conducted yearly. An
Interference Response Tests shall be
performed whenever an ACA or a RATA is
conducted. In years when a performance test
is also required under § 63.1208 of this
chapter to document compliance with
emission standards, the RATA shall coincide
with the performance test. The audits shall
be conducted as follows.
5.4.1 Relative Accuracy Test Audit
(RATA). This requirement applies to O2 and
CO GEMS. The RATA must be conducted at
least yearly. Conduct the RATA as described
in the RA test procedure (or alternate
procedures section) described in the
applicable Performance Specifications. In
addition, analyze the appropriate
performance audit samples received from the
EPA as described in the applicable sampling
methods.
5.4.2 Absolute Calibration Audit (ACA).
This requirements applies to all GEMS listed
in 5.4. The ACA must be conducted at least
quarterly except in a quarter when a RATA
(if applicable, see section 5.4.1) is conducted
instead. Conduct an ACA as described in the
calibration error (CE) test procedure
described in the applicable Performance
Specifications.
5.4.3 Interference Response Test. The
interference response test shall be conducted
whenever an ACA or RATA is conducted.
Conduct an interference response test as
described in the applicable Performance
Specifications.
5.4.4 Excessive Audit Inaccuracy. If the
RA from the RATA or the CE from the ACA
exceeds the criteria in the applicable
Performance Specifications, hazardous waste
burning must cease immediately. Hazardous
waste burning cannot resume until the owner
or operator take corrective measures and
audit the GEMS with a RATA to document
that the GEMS is operating within the
specifications.
6. Other Requirements
6.1 Performance Specifications. GEMS
used by owners and operators of HWCs must
comply with the following performance
specifications in Appendix B to Part 60:
TABLE I.—PERFORMANCE
SPECIFICATIONS FOR GEMS
GEMS
Carbon monoxide
Oxygen
Total hydrocarbons
Mercury, semivolatile metals,
and low volatile metals.
Particulate matter
Mercury
Hydrochloric acid (hydrogen
chloride).
Chlorine gas (diatomic chlorine)
Perform-
ance speci-
fication
4B
4B
8A
.10
11
12
13
14
6.2 Downtime due to Calibration.
Facilities may continue to burn hazardous
waste for a maximum of 20 minutes while
calibrating the GEMS. If all GEMS are
calibrated at once, the facility shall have
twenty minutes to calibrate all the GEMS. If
GEMS are calibrated individually, the facility
shall have twenty minutes to calibrate each
GEMS. If the GEMS are calibrated
individually, other GEMS shall be .
operational while the individual GEMS is
being calibrated.
6.3 Span of the GEMS.
6.3.1 Multi-metals, Particulate Matter,
Mercury, Hydrochloric Acid, and Chlorine
Gas GEMS. The span shall be at least 20
times the emission limit at an oxygen
correction factor of 1.
6.3.2 CO GEMS. The CO GEM shall have
two ranges, a low range with a span of 200
ppmv and a high range with a span of 3000
ppmv at an oxygen correction factor of 1. A
one-range GEM may be used, but it must
meet the performance specifications for the
low range in the specified span of the low
range.
6.3.3 O2 GEMS. The O2 GEM shall have
a span of 25 percent. The span may be higher
than 25 percent if the O2 concentration at the
sampling point is greater than 25 percent.
6.3.4 HC GEMS. The HC GEM shall have
a span of 100 ppmv, expressed as propane,
at an oxygen correction factor of 1.
6.3.5 GEMS Span Values When the
Oxygen Correction Factor is Greater than 2.
When a owner or operator installs a GEMS
at a location of high ambient air dilution, i.e.,
where the maximum oxygen correction factor
as determined by the permitting agency is
greater than 2, the owner or operator shall
install a GEM with a lower span(s),
proportionate to the larger oxygen correction
factor, than those specified above.
6.3.6 Use of Alternative Spans. Owner or
operators may request approval to use
alternative spans and ranges to those
specified. Alternate spans must be approved
in writing in advance by the Director. In
considering approval of alternative spans and
ranges, the Director will consider that
measurements beyond the span will be
recorded as values at the maximum span for
purposes of calculating rolling averages.
-------
27528
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
6.3.7 Documentation of Span Values. The
span value shall be documented by the GEMS
manufacturer with laboratory data.
6.4.1 Oxygen Correction Factor.
Measured pollutant levels shall be corrected
for the amount of oxygen in the stack
according to the following formula:
Pc=Pmxl4/(E-Y)
where:
Pc=concentration of the pollutant or standard
corrected to 7 percent oxygen;
Pm=measured concentration of the pollutant;
E=volume fraction of oxygen in the
combustion air fed into the device, on a
dry basis (normally 21 percent or 0.21 if
only air is fed);
Y=measured fraction of oxygen on a dry basis
at the sampling point.
The oxygen correction factor is:
OCF=14/(E-Y)
6.4.2 Moisture Correction. Method 4 of
appendix A of this Part shall be used to
determine moisture content of the stack
gasses.
6.4.3 Temperature Correction. Correction
values for temperature are obtainable from
standard reference materials.
6.5 Rolling Average. A rolling average is
the arithmetic average of all one-minute
averages over the averaging period.
6.5.1 One-Minute Average. One-minute
averages are the arithmetic average of the
four most recent 15-second observations and
shall be calculated using the following
equation:
- vc
"£4
Where:
c=the one minute average
Ci=a fifteen-second observation from the CEM
Fifteen second observations shall not be
rounded or smoothed. Fifteen-second
observations may be disregarded only as a
result of a failure in the GEMS and allowed
in the source's quality assurance'plan at the
time of the CMS failure. One-minute averages
shall not be rounded, smoothed, or
disregarded.
6.5.2 Ten Minute Rolling Average
Equation. The ten minute rolling average
shall be calculated using the following
equation:
'RA
10 <;.
~
Where:
CRA=The concentration of the standard,
expressed as a rolling average
£i=a one minute average
6.5.3 n-Hourly Rolling Average Equation.
The rolling average, based on a specific
number integer of hours, shall be calculated
using the following equation:
60*N -z
~[ 60*N
Where:
CRA=The concentration of the standard,
expressed as a rolling average
N=The number of hours of the rolling average
Q=a one minute average
6.5.4 New rolling averages. When a
rolling average begins due to the provisions
of § 6.5.4.2 of this appendix or when no
previous one-minute average have been
recorded, the rolling average shall be the
average all one-minute averages since the
rolling average commenced. Then when
sufficient time has passed such that there are
enough one-minute averages to calculate a
rolling average specified in § 6.5.2 or 6.5.3 of
this appendix, i.e., when the period of time
since the rolling average was started is equal
to or greater than the averaging period, the
average shall be calculated using the
equation specified there.
6.5.4.1 Short term interruption of a
rolling average. When rolling averages which
are interrupted (such as for a calibration or
failure of the GEMS), the rolling average shall
be restarted with the one-minute averages
prior to the interruption being the i=l to
(60*N-1) values and the i=60*N value being
the one minute average immediately after the
interruption. A short term interruption is one
with a duration of less than the averaging
period for the given standard or parameter.
6.5.4.2 Long term interruptions of the
rolling average. When ten minute rolling
averages are interrupted for periods greater
than ten minutes, the rolling average shall be
restarted as provided in § 6.5.4 of this
appendix. When rolling averages with
averaging periods in excess of the averaging
period for the given standard or parameter,
the rolling average shall be restarted as
provided in §6.5.4 of this appendix.
6.6 Units of the Standards for the
Purposes of Recording and Reporting
Emissions. Emissions shall be recorded and
reported expressed after correcting for
oxygen, temperature, and moisture.
Emissions shall be reported in metric, but
may also be reported in the English system
of units, at 7 percent oxygen, 20 °C, and on
a dry basis.
6.7 Rounding and Significant Figures.
Emissions shall be rounded to two significant
figures using ASTM procedure E-29-90 or its
successor. Rounding shall be avoided prior to
rounding for the reported value.
7. Bibliography
1. 40 CFR Part 60, Appendix F, "Quality
Assurance Procedures: Procedure 1. Quality
Assurance Requirements for Gas Continuous
Emission Monitoring Systems Used For
Compliance Determination".
PART 260—HAZARDOUS WASTE
MANAGEMENT SYSTEM: GENERAL
III. In part 260:
1. The authority citation for part 260
continues to read as follows:
Authority: 42 U.S.C. 6905, 6912(a), 6921-
6927, 6930, 6934, 6935, 6937, 6938, 6939,
and 6974.
2. Subpart B of part 260 is amended
by revising the definition of "industrial
furnace" and adding the following
definitions to read as follows:
§260.10 Definitions.
When used in parts 260 through 270
of this chapter, the following terms have
the meanings given below:
*****
Air pollution control system means
the equipment used to reduce the
release of particulate matter and other
pollutants to the atmosphere.
Automatic waste feed cutoff (AWFCO)
system means a system comprised of
cutoff valves, actuator, sensor, data
manager, and other necessary
components and electrical circuitry
designed, operated and maintained to
stop the flow of hazardous waste to the
combustion unit automatically and
immediately when any of the
parameters to which the system is
interlocked exceed the limits
established in compliance with
applicable standards, the operating
permit, or safety considerations.
*****
Cement kiln means a rotary kiln and
any associated preheater or precalciner
devices that produces clinker by heating
limestone and other materials for
subsequent production of cement for
use in commerce.
*****
Combustion chamber means the area
in which controlled flame combustion
of hazardous waste occurs.
* - * * * *
Continuous monitor means a device
which continuously samples the
regulated parameter without
interruption except during allowable
periods of calibration, and, for GEMS,
except as defined otherwise by the
applicable performance specification.
*****
Dioxins andfurans (D/F) means tetra,
penta, hexa, hepta, and octa-chlorinated
dibenzo dioxins and furans.
*****
Feedstream means any material fed
into a HWC, including, but not limited
to, any pumpable or nonpumpable solid
or gas.
*****
Flowrate means the rate at which a
feedstream is fed into a HWC.
*****
Fugitive combustion emissions means
particulate or gaseous matter generated
by or resulting from the burning of
hazardous waste that is not collected by
a capture system and is released to the
atmosphere prior to the exit of the stack.
*****
Industrial furnace means any of the
following enclosed devices that are
integral components of manufacturing
processes and that use thermal
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17529
treatment to accomplish recovery of
materials or energy:
(1) Cement kilns
(2) Lime kilns
(3) Lightweight aggregate kilns
*****
Lightweight aggregate kiln means a
rotary kiln that produces for commerce
(or for manufacture of products for
commerce) an aggregate with a density
less than 2.5 g/cc by slowly heating
organic-containing geologic materials
such as shale and clay.
*****
One-minute average means the
average of detector responses calculated
at least every 60 seconds from responses
obtained at least each 15 seconds.
*****
Operating record means all
information required by the standards to
document and maintain compliance
with the applicable regulations,
including data and information, reports,
notifications, and communications with
regulatory officials.
*****
Rolling average means the average of
all one-minute averages over the
averaging period.
Run means the net period of time
during which an air emission sample is
collected under a given set of operating
conditions. Three or more runs
constitutes an emissions test. Unless
otherwise specified, a run may be either
intermittent or continuous.
*****
Synthesis gas fuel means a gaseous
fuel produced by the thermal treatment
of hazardous waste and which meets the
specification provided by
TEQ means the international method
of expressing toxicity equivalents for
dioxins and furans as defined in U.S.
EPA, Interim Procedures for Estimating
Risks Associated with Exposures to
Mixtures of Chlorinated Dibenzo-p-
Dioxins and -Dibenzofurans (CDDs and
CDFs) and 1989 Update, March 1989.
PART 261—IDENTIFICATION AND
LISTING OF HAZARDOUS WASTE
IV. In part 261:
1. The authority citation for part 261
continues to read as follows:
Authority: 42 U.S.C. 6905, 6912(a), 6921,
6922, and 6938.
2. Section 261.4 is amended by
adding paragraph (a)(13) to read as
follows:
§261.4
(a)*
Exclusions.
(13) Wastes that meet the following
comparable fuel specifications, under
the conditions of paragraph (a)(13)(iv):
(i) Generic comparable fuel
specification. (A) Constituent
specifications. For compounds listed
below, the specification levels and,
where non-detect is the specification,
maximum allowable detection limits
are: [values to be determined].
(B) Physical specifications. (1)
Heating value. The heating value must
exceed 11,500 J/g (5,000 BTU/lbm).
(2) Flash point. The flash point must
not be less than [value to be
determined].
(3) Viscosity. The viscosity must not
exceed [value to be determined]
(ii) Synthesis gas fuel specification.
(A) Synthesis gas (syngas) which is
generated from hazardous waste and
which:
(2) Has a minimum Btu value of
11,500 J/g (5,000 Btu/lb);
(2) Contains less than 1 ppmv of each
hazardous constituent listed in
Appendix VIII of this part that could
reasonably be expected to be in the gas,
except the limit for hydrogen sulfide
(H2S) is 10 ppmv; and
(3) Which contains less than 1 ppmv
each of total chlorine and total nitrogen
other them diatomic nitrogen (N2).
(B) Measurements of concentrations of
constituents specified in paragraph
(a)(13)(ii)(A) are to be taken at the
temperature and pressure of the gas at
the point that the exclusion is first
claimed.
(iii) Implementation. Waste that meets
the comparable fuel specifications
provided by paragraphs (a)(13)(i) or (ii)
of this section is excluded from the
definition of solid waste provided that:
(A) The person who generates the
waste or produces the syngas must
claim the exclusion. For purposes of
this paragraph, that person is called the
waste-derived fuel producer;
(B) (2) The producer must submit a
one-time notice to the Director claiming
the exclusion and certifying compliance
with the conditions of the exclusion.
(2) If the producer is a company
which produces comparable fuel at
more than one facility, the producer
shall specify at which sites the
comparable fuel will be produced and
each specified site must be in
compliance with the conditions of the
exclusion at each point of production;
(C) Sampling and analysis. (1) The
producer must obtain information by
sampling and analysis as often as
necessary to document that fuel claimed
to be excluded meets the comparable
fuel specification provided by
paragraphs (a)(13)(i) or (ii) of this
section. At a minimum, the producer
must sample and analyze the fuel for all
constituents for which specifications are
established when the exclusion is first
claimed, and at least annually
thereafter, for all constituents that, using
the results of the initial test and process
knowledge, the producer reasonably
expects to be found in the comparable
fuel.
(2) The producer must develop and
implement a comparable fuel sampling
and analysis plan, using the same
protocols used to develop waste
analysis plans, to document that the
comparable fuel meets the
specifications.
(3) Analytical methods provided by
SW-846 must be used unless prior
written approval is obtained from the
Director to use an equivalent method;
(4) If a waste-derived fuel is blended
in order to meet the flash point and
kinematic viscosity specifications, the
producer shall analyze the fuel as
produced to ensure that it meets the
constituent and heating value
specifications and then analyze the fuel
again after blending to ensure that it
meets all specifications.
(5) If not blended, the comparable fuel
shall be analyzed as produced.
(D) (2) Comparable fuel shall be
burned on-site or shipped directly to a
person who burns the waste.
(2} No person other than the producer
and the burner shall manage a
comparable fuel other than incidental
transportation related handling.
(E) Treatment to meet the
specification. (2) Bona fide treatment of
hazardous waste to remove or destroy
constituents listed in the specifications
or to raise the heating value by
removing constituents or materials can
be used to meet the specification.
(2) Owners and operators of RCRA
permitted hazardous waste treatment
facilities qualify as producers of waste-
derived fuel eligible for the exclusion
provided that the newly generated waste
results from bona fide treatment to
remove or destroy constituents listed in
the specifications or to increase the
heating value.
(3) Residuals resulting from the
treatment of a hazardous waste listed in
subpart D of this part to generate a
comparable fuel remain a hazardous
waste.
(4) Treatment by incidental settling
during storage or blending operations is
not bona fide treatment for purposes of
this exclusion; and
(F) Blending to meet the specification.
Blending a waste containing, as
generated, higher concentration(s) of
hazardous constituent(s) than allowed
in the comparable fuel specifications
with materials with lower •
-------
27530
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
concentrations of such constituents to
meet the specifications is prohibited.
(An excluded comparable fuel, however,
may be blended with other materials
without restriction.)
(G) Speculative Accumulation.
Producers and burners are subject to the
speculative accumulation test under
§ 261.2(c)(4).
(H) Recordkeeping. Producers
claiming the exclusion must keep
records of:
(i) One-time notification to the
Director required by paragraph
(a)(13)(ii)(B) of this section;
(2) Sampling and analysis or other
information documenting that the fuel
meets the comparable fuel specification;
(3) The comparable fuel sampling and
analysis plan; and
(4) For waste that is treated before
meeting particular constituent limits of
the comparable fuel specification,
documentation that the treatment
resulted in removal or destruction of
those constituents to meet the
specification.
(I) Records Retention. Records must
be retained for three years. The
sampling and analysis plan and all
revisions to the plan shall be retained
for as long as the producer claims the
exclusion, plus three years.
PART 264—STANDARDS FOR
OWNERS AND OPERATORS OF
HAZARDOUS WASTE TREATMENT,
STORAGE, AND DISPOSAL
FACILITIES
V. In part 264:
1. The authority citation for part 264
continues to read as follows:
Authority: 42 U.S.C. 6905, 6912(a), 6924,
and 6925.
2. Section 264.340 is amended by
redesignating paragraphs (b), (c), and (d)
as paragraphs (c), (d), and (e),
respectively, and adding paragraph (b),
to read as follows:
§264.340 Applicability.
*****
(b) Incorporation of MACT standards.
(1) The requirements applicable to
hazardous waste incinerators under
subpart EEE, part 63, of this chapter are
incorporated by reference.
(2) When an owner and operator begin
compliance (i.e., submit a notification of
compliance) with the requirements of
subpart EEE, part 63, of this chapter:
(i) The applicability provisions of
§ 264.340(b) and (c) no longer apply;
(ii) The performance standards
provided by §264.343(b) and (c) are
superseded (i.e., replaced) by the
subpart EEE, part 63, standards such
that an operating permit issued or
reissued under part 270 of this chapter
must ensure compliance with the
subpart EEE, part 63, standards as well
as the DRE performance standard under
§ 264.343;
(iii) The operating requirements of
§ 264.345(b)(l) through (4) and the
monitoring requirements of
§264.347(a)(l) and (2) are superseded
(i.e., replaced) by the operating and
monitoring requirements of § 63.1210 of
this chapter such that an operating
permit issued or reissued under part 270
of this chapter must ensure compliance
with the subpart EEE, part 63, standards
as well as the remaining standards
under §§ 264.345 and 264.347; and
(iv) The operating requirements of
§264.345(d)(l)-(3) and § 264.345(e) are
superseded (i.e., replaced) by the
operating and monitoring requirements
of § 63.1207 of this chapter such that an
operating permit issued or reissued
under part 270 of this chapter must
ensure compliance with the subpart
EEE, part 63, standards as well as the
remaining applicable standards under
§ 264.345.
*****
3. Section 264.345 is amended by
revising paragraph (a) and adding
paragraph (g) to read as follows:
§264.345 Operating Requirements
(a) An incinerator must be operated in
accordance with operating requirements
specified in the permit and meet the
applicable emissions standards at all
times that hazardous waste remains in
the combustion chamber. These will be
specified on a case-by-case basis as
those demonstrated (in a trial burn or in
alternative data as specified in
§ 264.344(b) and included with part B of
the facility's permit application) to be
sufficient to comply with the
performance standards of § 264.343.
*****
(g) ESV Openings. (1) Violation. If an
emergency safety vent opens when
hazardous waste is fed or remains in the
combustion chamber, such that
combustion gases are not treated as
during the most recent performance test,
it is a violation of the emission
standards of this subpart.
(2) ESV Operating Plan. The ESV
Operating Plan shall explain detailed
procedures for rapidly stopping waste
feed, shutting down the combustor,
maintaining temperature in the
combustion chamber until all waste
exits the combustor, and controlling
emissions in the event of equipment
malfunction or activation of any ESV or
other bypass system including
calculations demonstrating that
emissions will be controlled during
such an event (sufficient oxygen for
combustion and maintaining negative
pressure), and the procedures for
executing the plan whenever the ESV is
used, thus causing an emergency release
of emissions.
(3) Corrective measures. After any
ESV opening that results in a violation,
the owner or operator must investigate
the cause of the ESV opening, take
appropriate corrective measures to
minimize future ESV violations, and
record the findings and corrective
measures in the operating record.
(4) Reporting requirement. The owner
or operator must submit a written report
within 5 days of a ESV opening
violation documenting the result of the
investigation and corrective measures
taken.
4. Section 264.347 is amended by
adding paragraphs (e), (f), and (g).
§264.347 Monitoring and inspections.
*****
(e) Fugitive emissions. (1) Fugitive
emissions must be controlled by:
(i) Keeping the combustion zone
totally sealed against fugitive emissions;
or
(ii) Maintaining the maximum
combustion zone pressure lower than
ambient pressure using an
instantaneous monitor; or
(iii) Upon prior written approval of
the Administrator, an alternative means
of control to provide fugitive emissions
control equivalent to maintenance of
combustion zone pressure lower than
ambient pressure;
(2) The owner or operator must
specify in the operating record the
method used for fugitive emissions
control.
(f) Continuous emissions monitors
(GEMS). (1) Hazardous waste
incinerators shall be equipped with
CEMS for compliance monitoring.
(2) At all times that hazardous waste
is fed into the hazardous waste
incinerator or remains in the
combustion chamber, CEMS must be
operated in compliance with the
requirements of the appendix to subpart
EEE, part 63, of this chapter.
(g) Other continuous monitoring
systems. (1) CMS other than CEMS (e.g.,
thermocouples, pressure transducers,
flow meters) must be used to document
compliance with the applicable
operating limits.
(2) Non-CEM CMS must be installed
and operated in conformance with
§ 63.8(c)(3) of this chapter requiring the
owner and operator, at a minimum, to
comply with the manufacturer's written
specifications or recommendations for
installation, operation, and calibration
of the system.
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17531
(3) Non-CEM CMS must sample the
regulated parameter without
interruption, and evaluate the detector
response at least once each 15 seconds,
and compute and record the average
values at least every 60 seconds.
(4) The span of the detector must not
be exceeded. Span limits shall be
interlocked into the automatic waste
feed cutoff system.
PART 265—INTERIM STATUS
STANDARDS FOR OWNERS AND
OPERATORS OF HAZARDOUS WASTE
TREATMENT, STORAGE, AND
DISPOSAL FACILITIES
VI. In part 265:
1. The authority citation for part 265
continues to read as follows:
Authority: 42 U.S.C. 6905, 6912(a), 6924,
6925, 6935, and 6936, unless otherwise
noted.
2. Section 265.340 is amended by
redesignating paragraph (b) as paragraph
(c), and adding paragraph (b), to read as
follows:
§265.340 Applicability.
*****
(b) Incorporation of MACT standards.
(1) The requirements applicable to
hazardous waste incinerators under
subpart EEE, part 63, of this chapter are
incorporated by reference.
(2) When an owner and operator begin
to comply (i.e., submit a notification of
compliance) with the requirements of
subpart EEE, part 63, of this chapter,
those requirements apply in addition to
those of this subpart, and the provisions
of § 265.340(b) no longer apply.
*****
3. Section 265.347 is amended by
adding paragraphs (c), (d), and (e), to
read as follows:
§265.347 Monitoring and inspections.
*****
(c) Fugitive emissions. (1) Fugitive
emissions must be controlled by:
(i) Keeping the combustion zone
totally sealed against fugitive emissions;
or
(ii) Maintaining the maximum
combustion zone pressure lower than
ambient pressure using an
instantaneous monitor; or
(iii) Upon prior written approval of
the Administrator, an alternative means
of control to provide fugitive emissions
control equivalent to maintenance of
combustion zone pressure lower than
ambient pressure;
(2) The owner or operator must
specify in the operating record the
method used for fugitive emissions
control.
(d) Continuous emissions monitoring
systems (GEMS). (1) Hazardous waste
incinerators shall be equipped with
CEMS for compliance monitoring.
(2) At all times that hazardous waste
is fed into the hazardous waste
incinerator or remains in the
combustion chamber, CEMS must be
operated in compliance with the
requirements of the appendix to subpart
EEE, part 63, of this chapter.
(e) Other continuous monitoring
systems. (1) CMS other than CEMS (e.g.,
thermocouples, pressure transducers,
flow meters) must be used to document
compliance with the applicable
operating limits.
(2) Non-CEM CMS must be installed
and operated in conformance with
§ 63.8(c)(3) of this chapter requiring the
owner and operator, at a minimum, to
comply with the manufacturer's written
specifications or recommendations for
installation, operation, and calibration
of the system.
(3) Non-CEMS CMS must sample the
regulated parameter without
interruption, and evaluate the detector
response at least once each 15 seconds,
and compute and record the average
values at least every 60 seconds.
(4) The span of the detector must not
be exceeded. Span limits shall be
interlocked into the automatic waste
feed cutoff system.
PART 266—STANDARDS FOR THE
MANAGEMENT OF SPECIFIC
HAZARDOUS WASTES AND SPECIFIC
TYPES OF HAZARDOUS WASTE
MANAGEMENT FACILITIES
VH. In part 266:
1. The authority citation for part 266
continues to read as follows:
Authority: Sees. 1006, 2002(a), 3004, and
3014 of the Solid Waste Disposal Act, as
amended by the Resource Conservation and
Recovery Act of 1976, as amended (42 U.S.C.
6905, 6912(a), 6924, and 6934].
2. Section 266.100 is amended by
redesignating paragraphs (b), (c), (d), (e),
and (f) as paragraphs (c), (d), (e), (f), and
(g), adding paragraph (b), revising
introductory text to paragraph (d)(l),
revising paragraphs (d)(2) (i) and (ii),
revising the introductory text to
paragraph (d)(3), revising paragraphs
(d)(3)(i)(B) and (d)(3)(ii), and adding
paragraph (h), to read as follows:
§266.100 Applicability.
*****
(b) Incorporation of MACT standards.
(1) The requirements applicable to
cement kilns and lightweight aggregate
kilns under subpart EEE, part 63, of this
chapter are incorporated by reference.
(2) When an owner and operator begin
to comply (i.e., submit a notification of
compliance) with the requirements of
subpart EEE, part 63, of this chapter,
those requirements apply in addition to
those of this subpart.
*****
(d) * * *
(1) To be exempt from §§ 266.102
through 266.111, an owner or operator
of a metal recovery furnace or mercury
recovery furnace must comply with the
following requirements, except that an
owner or operator of a lead or a nickel-
chromium recovery furnace, or a metal
recovery furnace that burns baghouse
bags used to capture metallic dusts
emitted by steel manufacturing, must
comply with the requirements of
paragraph (d)(3) of this section, and
owners or operators of lead recovery
furnaces that are subject to regulation
under the Secondary Lead Smelting
NESHAP must comply with the
requirements of paragraph (h) of this
section.
*****
(2) * * *
(i) The hazardous waste has a total
concentration of nonmetal compounds
listed in part 261, appendix VIII, of this
chapter exceeding 500 ppm by weight,
as fired, and so is considered to be
burned for destruction. The
concentration of nonmetal compounds
in a waste as generated may be reduced
to the 500 ppm limit by bona fide
treatment that removes or destroys
nonmetal constituents. Blending for
dilution to meet the 500 ppm limit is
prohibited and documentation that the
waste has not been impermissibly
diluted must be retained in the records
required by paragraph (d)(l)(iii) of this
section; or
(ii) The hazardous waste has a heating
value of 5,000 Btu/lb or more, as fired,
and so is considered to be burned as
fuel. The heating value of a waste as
generated may be reduced to below the
5,000 Btu/lb limit by bona fide
treatment that removes or destroys
nonmetal constituents. Blending for
dilution to meet the 5,000 Btu/lb limit
is prohibited and documentation that
the waste has not been impermissibly
diluted must be retained in the records
required by paragraph (d)(l)(iii) of this
section.
(3) To be exempt from § 266.102
through 266.111, an owner or operator
of a lead or nickel-chromium or mercury
recovery furnace, except for owners or
operators of lead recovery furnaces
subject to regulation under the
Secondary Lead Smelting NESHAP,
* * *
(i) * * *
(B) The waste does not exhibit the
Toxicity Characteristic of § 261.24 of
-------
17532
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
this chapter for a nonmetal constituent;
and
*****
(ii) The Director may decide on a
case-by-case basis that the toxic
nonmetal constituents in a material
listed in appendix XI or XII of this part
that contains a total concentration of
more than 500 ppm toxic nonmetal
compounds listed in appendix VIII, part
261, of this chapter, may pose a hazard
to human health and the environment
when burned in a metal recovery
furnace exempt from the requirements
of this subpart. In that situation, after
adequate notice and opportunity for
comment, the metal recovery furnace
will become subject to the requirements
of this subpart when burning that
material. In making the hazard
determination, the Director will
consider the following factors;
(A) The concentration and toxicity on
nonmetal constituents in the material;
and
(B) The level of destruction of toxic
nonmetal constituents provided by the
furnace; and
(C) Whether the acceptable ambient
levels established in appendices IV or V
of this part may be exceeded for any
toxic nonmetal compound that may be
emitted based on dispersion modeling
to predict the maximum annual average
off-site ground level concentration.
*****
(h) Starting June 23,1997, owners or
operators of lead recovery furnaces that
process hazardous waste for recovery of
lead and that are subject to regulation
under the Secondary Lead Smelting
NESHAP, are conditionally exempt from
regulation under this subpart, except for
§ 266.101. To be exempt, an owner or
operator must provide a one-time notice
to the Director identifying each
hazardous waste burned and specifying
that the owner or operator claims an
exemption under this paragraph. The
notice also must state that the waste
burned has a total concentration of non-
metal compounds listed in part 261,
appendix VIII, of this chapter of less
than 500 ppm by weight, as fired and as
provided in paragraph (d)(2)(i) of this
section, or is listed in appendix XI, part
266.
3. Section 266.101 is amended by
revising paragraph (c)(l) to read as
follows:
§ 266.101 Management prior to burning.
*****
(c) Storage and treatment facilities. (1)
Owners and operators of facilities that
store or treat hazardous waste that is
burned in a boiler or industrial furnace
are subject to the applicable provisions
of parts 264, 265, and 270 of this
chapter, except as provided by
paragraph (c)(2) of this section. These
standards apply to storage and treatment
by the burner as well as to storage and
treatment facilities operated by
intermediaries (processors, blenders,
distributors, etc.)
*****
4. Section 266.102 is amended by
redesignating paragraph (a)(2) as (a)(3),
adding paragraph (a)(2), revising the
introductory text to paragraph (d)(4),
adding paragraph (d)(5), revising
paragraphs (e)(4)(i) (A) and (C), (e)(5)(i)
(A) and (C), (e)(6)(i) (A), (B), and (C),
and (e)(6)(iii), revising the introductory
text to (e)(7)(i), and revising paragraphs
(e)(7)(i)(C), (e)(8)(i) (A) and (C), and
(e)(10), to read as follows:
§ 266.102 Permit standards for burners.
(a) Applicability. (1) * * *
(2) Applicability of MACT standards
to cement and lightweight aggregate'
kilns. When an owner and operator of
a cement or lightweight aggregate kiln
that burns hazardous waste begin to
comply (i.e., submit a notification of
compliance) with the requirements of
subpart EEE, part 63, of this chapter:
(i) The emission standards provided
by §§ 266.104 through 266.107 are
superseded (i.e., replaced) by the
standards under subpart EEE, part 63,
except that the DRE requirement
provided by § 266.104(a) and the
enforcement provisions of those
sections (i.e., §§ 266.104(i), 266.105(c),
266.106(1), and 266.107(h)) continue to
apply;
(ii) The specific operating
requirements (and associated
monitoring requirements) provided by
paragraphs (e)(2)(ii), (e)(3), (e)(4), and
(e)(5) of this section are superseded by
the standards under subpart EEE, part
63, except that the provisions of
paragraphs (e)(2)(i)(G), (e)(3)(i)(E),
(e)(4)(ii)(J), (e)(4)(iii)(J), and (e)(5)(i)(G)
of this section continue to apply to
enable the permitting authority to
establish such other operating
requirements as are necessary to ensure
compliance with the standards of
subpart EEE, Part 63.;
(iii) An operating permit that is issued
or reissued under part 270 of this
chapter must ensure compliance with
the subpart EEE, part 63, standards as
well as those § 266.102 standards that
continue to apply.
*****
(d) * * *
(4) Except as provided by paragraph
(d)(5) of this section, * * *
(5) When a cement or lightweight
aggregate kiln becomes subject to the
standards of subpart EEE, Part 63, of this
chapter, the provisions of paragraph
(d)(4) of this section continue to apply,
except that the operating requirements
established under that paragraph will be
those sufficient to ensure compliance
with the emission standards of subpart
EEE and the DRE requirement of
§ 266.104(a).
(e) * * *
(4) * * *
(i) * * *
(A) Total feedrate of each metal in
every feedstream measured and
specified under provisions of paragraph
(e)(6) of this section;
*****
(C) A sampling and metals analysis
program for every feedstream;
*****
(5) * * *
(i) * * *
(A) Feedrate of total chloride and
chlorine in every feedstream measured
and specified as prescribed in paragraph
(e)(6) of this section;
*****
(C) A sampling and analysis program
for total chloride and chlorine for every
feedstream:
*****
(6) * * *
. (i) * * *
(A) One-minute average. The limit for
a parameter shall be established and
continuously monitored on a one-
minute average basis, and the permit
limit specified as the time-weighted
average during all valid runs of the trial
burn of the one-minute averages.
(B) Hourly rolling average. The limit
for a parameter shall be established and
continuously monitored on an hourly
rolling average basis. The permit limit
for the parameter shall be established
based on trial burn data as the average
over all valid test runs of the highest (or
lowest, as appropriate) hourly rolling
average value for each run.
(C) Instantaneous limit for
combustion chamber pressure.
Combustion chamber pressure shall be
continuously sampled, detected, and
recorded without use of an averaging
period.
(ii) * * *
(iii) Feedrate limits for metals, total
chloride and chlorine, and ash. Feedrate
limits for metals, total chlorine and
chloride, and ash are established and
monitored by knowing the
concentration of the substance (i.e.,
metals, chloride/chlorine, and ash) in
each feedstream and the flow rate of the
feedstreams. To monitor the feedrate of
these substances, the flowrate of each
feedstream must be monitored under the
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 /Proposed Rules
17533
monitoring requirements of paragraphs
(e)(6) (i) and (ii) of this section.
*****
(7)* * *
(i) Fugitive emissions. Fugitive
emissions must be controlled by the
following and it must specify in the
operating record the method used for
fugitive emissions control:
*****
(C) Upon prior written approval of the
Administrator, an alternative means of
control to provide fugitive emissions
control equivalent to maintenance of
combustion zone pressure lower than
ambient pressure.
*****
(8)* * *
(i) * * *
(A) If specified by the permit,
feedrates and composition of every
feedstream and feedrates of ash, metals,
and total chloride and chlorine;
*****
(C) Upon the request of the Director,
sampling and analysis of any
feedstream, residues, and exhaust
emissions must be conducted to verify
that the operating requirements
established in the permit achieve the
applicable standards of §§ 266.105,
266.106, 266.107, and 266.108.
*****
(10) Recordkeeping. The owner or
operator shall maintain files of all
information (including all reports and
notifications) required by this section
recorded in a form suitable and readily
available for expeditious inspection and
review. The files shall be retained for at
least 5 years following the date of each
occurrence, measurement, maintenance,
report, or record. At a minimum, the
most recent 2 years of data shall be
retained on site. The remaining 3 years
of data may be maintained on
microfilm, on a computer, on computer
floppy disks, on magnetic tape disks, or
on microfiche.
*****
6. Section 266.103 is amended by
redesignating paragraphs (a)(2) through
(a) (7) as paragraphs (a) (3) through (a) (8),
adding paragraph (a)(2), revising the
introductory text to paragraph (b)(2)(ii),
revising paragraphs (b)(2)(ii)(A),
(b)(2)(iii), and (b)(5)(i) and (iii), revising
the introductory text to paragraphs (c)
and (c)(4), revising paragraphs
(c)(4)(iv)(A) through (D), revising the
introductory text to paragraph (c)(7),
adding a sentence at the end of
paragraph (d), revising the introductory
text to paragraph (h), revising
paragraphs (h)(3) and (i), revising the
introductory text to paragraph (j)(l), and
revising paragraphs (j)(l){i) and (iii), and
(k), to read as follows:
§ 266.103 Interim status standards for
burners.
(a) * * *
(2) Compliance with subpart EEE, part
63. When an owner and operator begin
to comply (i.e., submit a notification of
compliance) with the requirements of
subpart EEE, part 63, of this chapter
(and that are incorporated by reference),
those requirements apply in lieu of the
requirements of paragraphs (b) through
(k) of this section.
*****
(b)* * *
(2) * * *
(ii) Except for facilities complying
with the Tier I or Adjusted Tier I
feedrate screening limits for metals or
total chlorine and chloride provided by
§§ 266.106(b) or (e) and 266.107(b)(l) or
(e), respectively, the estimated
uncontrolled (at the inlet to the air
pollution control system) emissions of
particulate matter, each metal controlled
by § 266.106, and hydrochloric acid and
chlorine, and the following information
supporting such determinations:
(A) The feedrate (Ib/hr) of ash,
chlorine, antimony, arsenic, barium,
beryllium, cadmium, chromium, lead,
mercury, silver, and thallium in each
feedstream;
*****
(iii) For facilities complying with the
Tier I or Adjusted Tier I feedrate
screening limits for metals or total
chlorine and chloride provided by
§§266.106(b) or (e) and 266.107(b)(l) or
(e), the feedrate (Ib/hr) of total chloride
and chlorine, antimony, arsenic,
barium, beryllium, cadmium,
chromium, lead, mercury, silver, and
thallium in each feedstream.
*****
(5) * * *
(i) General requirements. Limits on
each of the parameters specified in
paragraph (b)(3) of this section (except
for limits on metals concentrations in
collected particulate matter (PM) for
industrial furnaces that recycle
collected PM) shall be established and
monitored under either of the following
methods:
(A) One-minute average. The limit for
a parameter shall be established and
continuously monitored on a one-
minute average basis, and the permit
limit specified as the time-weighted
average during all valid runs of the trial
burn of the one-minute averages.
(B) Hourly rolling average. The limit
for a parameter shall be established and
continuously monitored on an hourly
rolling average basis. The permit limit
for the parameter shall be established
based on trial burn data as the average .
over all valid test runs of the highest (or.
lowest, as appropriate) hourly rolling
average value for each run.
(C) Instantaneous limit for
combustion chamber pressure.
Combustion chamber pressure shall be
continuously sampled, detected, and
recorded without use of an averaging
period.
*****
(iii) Feedrate limits for metals, total
chloride and chlorine, and ash. Feedrate
limits for metals, total chlorine and
chloride, and ash are established and
monitored by knowing the
concentration of the substance (i.e.,
metals, chloride/chlorine, and ash) in
each feedstream and the flow rate of the
feedstream. To monitor the feedrate of
these substances, the flowrate of each
feedstream must be monitored under the
monitoring requirements of paragraphs
(b)(5)(i) and (ii) of this section.
* * * * *:
(c) Certification of Compliance. The
owner or operator shall conduct
emissions testing to document
compliance with the emissions
standards of §§ 266.104(b) through (e),
266.105, 266.106, 266.107 and
paragraph (a)(5)(i)(D) of this section,
under the procedures prescribed by this
paragraph, except under extensions of
time provided by paragraph (c)(7).
Based on the compliance test, the owner
or operator shall submit to the Director
on or before August 21,1992, a
complete and accurate "certification of
compliance" (under paragraph (c)(4) of
this section) with those emission
standards establishing limits on the
operating parameters specified in
paragraph (c)(l).
*****
(4) Certification of compliance.
Within 90 days of completing
compliance testing, the owner or
operator must certify to the Director
compliance with the emission standards
of §§266.104(b), (c), and (e), 266.105,
266.106, 266.107 and paragraph
(a)(5)(i)(D) of this section. The
certification of compliance must include
the following information:
*****
(iv) * * *
(A) One-minute average. The limit for
a parameter shall be established and
continuously monitored on a one-
minute average basis, and the permit
limit specified as the time-weighted
average during all valid runs of the trial
burn of the one-minute averages.
(B) Hourly rolling average. The limit
for a parameter shall be established and
continuously monitored on an hourly
rolling average basis. The permit limit
for the parameter shall be established
based on trial burn data as the average
-------
17534
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
over all valid test runs of the highest (or
lowest, as appropriate) hourly rolling
average value for each run.
(C) Instantaneous limit for
combustion chamber pressure.
Combustion chamber pressure shall be
continuously sampled, detected, and
recorded without use of an averaging
period.
(D) Feedrate limits for metals, total
chloride and chlorine, and ash. Feedrate
limits for metals, total chlorine and
chloride, and ash are established and
monitored by knowing the
concentration of the substance (i.e.,
metals, chloride/chlorine, and ash) in
each feedstream and the flow rate of the
feedstream. To monitor the feedrate of
these substances, the flow rate of each
feedstream must be monitored under the
monitoring requirements of paragraphs
(c)(4)(iv)(A) through (C) of this section.
*****
(7) Extensions of time. If the owner or
operator does not submit a complete
certification of compliance for all of the
applicable emission standards of
§266.104, 266.105, 266.106, and
266.107 as specified in § 266.103(C)(1),
or as required pursuant to § 266.103(d),
he/she must either:
*****
(d) * * *. The extensions of time
provisions of paragraph (c)(7) of this
section apply to recertifications.
* * * * *
(h) Fugitive emissions. Fugitive
emissions must be controlled by one of
the following methods. The operator
must specify in the operating record the
method selected.
*****
(3) Upon prior written approval of the
Administrator, an alternative means of
control to provide fugitive emissions
control equivalent to maintenance of
combustion zone pressure lower than
ambient pressure.
(i) Changes. A boiler or industrial
furnace must cease burning hazardous
waste when changes in combustion
properties, or feedrates of any
feedstream, or changes in the boiler or
industrial furnace design or operating
conditions deviate from the limits
specified in the certification of
compliance.
(j) Monitoring and Inspections. (1)
The owner or operator must monitor
and record the following, at a minimum,
while burning hazardous waste. All
monitoring and recording shall be in
units corresponding to the units on the
operating limits established in the
certification of precompliance and
certification of compliance.
(i) Applicable operating parameters of
paragraphs (b) and (c) of this section
shall be monitored and recorded under
the requirements of paragraphs (b)(5) (i)
and (ii) of this section;
*****
(iii) Upon request of the Director,
sampling and analysis of any feedstream
and the stack gas emissions must be
conducted to verify that the operating
conditions established in the
certification of precompliance or
certification of compliance achieve the
applicable standards of §§ 266.104,
266.105, 266.106, and 266.107.
(k) Recordkeeping. The owner or
operator shall maintain files of all
information (including all reports and
notifications) required by this section
recorded in a form suitable and readily
available for expeditious inspection and
review. The files shall be retained for at
least 5 years following the date of each
occurrence, measurement, maintenance,
report, or record. At a minimum, the
most recent 2 years of data shall be
retained on site. The remaining 3 years
of data may be maintained on
microfilm, on a computer, on computer
floppy disks, on magnetic tape disks, or
on microfiche.
*****
7. Section 266.104 is amended by
removing paragraph (f), and
redesignating paragraphs (g) and (h) as
paragraphs (f) and (g), respectively.
8. Section 266.105 is amended by
revising paragraph (b), redesignating
paragraph (c) as paragraph (d) and
adding paragraph (c), to read as follows:
§266.105 Standards to control participate
matter.
*****
(b) An owner or operator meeting the
requirements of § 266.109(b) for the low
risk exemption is exempt from the
particulate matter standard. Owners and
operators of cement or lightweight
aggregate kilns are not eligible for this
exemption, however, upon compliance
with the emission standards of subpart
EEE, Part 63, of this chapter.
(c) Oxygen correction. (1) Measured
pollutant levels shall be corrected for
the amount of oxygen in the stack gas
according to the formula:
Pc=Pm x 14/(E-Y)
where PC is the corrected concentration
of the pollutant in the stack gas, Pm is
the measured concentration of the
pollutant in the stack gas, E is the
oxygen concentration on a dry basis in
the combustion air fed to the device,
and Y is the measured oxygen
concentration on a dry basis in the
stack.
(2) For devices that feed normal
combustion air, E will equal 21 percent.
For devices that feed oxygen-enriched
air for combustion (that is, air with an
oxygen concentration exceeding 21
percent), the value of E will be the
concentration of oxygen in the enriched
air.
(3) Compliance with all emission
standards provided by this subpart shall
be based on correcting to 7 percent
oxygen using this procedure.
*****
9. Section 266.108 is amended by
revising paragraph (a) (2), to read as
follows:
§ 266.108 . Small quantity on-site burner
exemption.
(a) * * *
(2) The quantity of hazardous waste
burned in a device for a calendar month
does not exceed 27 gallons.
*****
10. Section 266.109 is amended by
revising the introductory text to
paragraph (b) and adding paragraph
(b)(3), to read as follows:
§ 266.109 Low risk waste exemption.
*****
(b) Waiver of particulate matter
standard. Except as provided in
paragraph (b)(3) of this section, the
particulate matter standard of § 266.105
does not apply if:
*****
(3) When the owner and operator of
a cement or lightweight aggregate kiln
become subject to the standards of
subpart EEE, part 63, of this chapter
(i.e., upon submittal of the initial
notification of compliance), the source
is no longer eligible for waiver of the
PM standard provided by this
paragraph. At that time, the source is
subject to the PM standard provided by
subpart EEE, part 63.
11. Section 266.112 is amended by
adding a sentence at the end of the
introductory text to paragraph (b)(l),
adding a sentence at the beginning of
paragraph (b)(l)(ii), adding a sentence
before the last sentence of paragraph
(b)(2)(i), revising the first sentence of
paragraph (b)(2)(iii), redesignating
paragraph (c) as paragraph (d), and
adding paragraph (c), to read as follows:
§ 266.112 Regulation of residues.
*****
(b)* * *
(1) * * * For polychlorinated
dibenzo-p-dioxins and polychlorinated
dibenzo-furans, specific congeners and
homologues must be measured and
converted to 2,3,7,8-TCDD equivalent
values using the calculation procedure
specified in appendix IX, section 4.0 of
this part.
(ii) Waste-derived residue. Waste-
derived residue shall be sampled and
-------
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
17535
analyzed as required by this paragraph
and paragraph (c) of this section to
determine whether the residue
generated during each 24 hour period
has concentrations of toxic constituents
that are higher than the concentrations
established for the normal residue under
paragraph Cb)(l){i) of this section. * * *
(2)* * *
(i) * * * In complying with the
alternative levels for polychlorinated
dibenzo-p-dioxins and polychlorinated
dibenzo-furans, only the tetra-, penta-,
and hexa- homologues need to be
measured. * * *
*****
(iii) Sampling and analysis. Waste-
derived residue shall be sampled and
analyzed as required by this paragraph
and paragraph (c) of this section to
determine whether the residue
generated during each 24-hour period
has concentrations of toxic constituents
that are higher than the health-based
levels. * * *
(c) Sampling and analysis frequency.
(1) The owner or operator must sample
and analyze residues at least once each
24-hour period when burning hazardous
waste, unless written, advance approval
is obtained from the Regional
Administrator under paragraph (c)(2) of
this section for less frequent sampling
and analysis.
(2) Requests for approval for less
frequent sampling and analysis (that is,
less than once each 24-hour period)
must be based on and justified by a
statistical analysis.
(i) The Regional Administrator shall
not grant approval for a sampling and
analysis frequency of less than once
each month.
(ii) At a minimum, the following
information to support the request for
reduced sampling and analysis
frequency must be submitted to the
Regional Administrator and must be
contained in the facility's waste analysis
plan for residue sampling:
(A) The statistical methodology
selected, reason for selection, and the
statistical procedures for calculating the
sampling frequency;
(B) Analytical results used to generate
the statistical database; and
(C) A description of how the
statistical database is to be maintained
and updated.
*****
12. Appendix VIII to part 266 is
revised to read as follows:
Appendix VIII to Part 266—Organic Compounds for Which Residues Must Be Analyzed for Bevill Determinations
Volatiles
Semivolatiles
Benzene
Toluene ,
Carbon tetrachloride ,
Chloroform ,
Methylene chloride ,
Trichloroethylene ,
Tetra chloroethylene ,
1,1,1-Trichloroethane ,
Chlorobenzene ,
cis-1,4-Dich!oro-2-butene ,
Bromochloromethane ,
Bromodichloromethane ...,
Brornoform
Bromomethane
Methylene bromide
Methyl ethyl ketone ,
Bis(2-ethylhexyl)phthalate.
Naphthalene.
Phenol.
Diethyl phthalate.
Butyl benzyl phthalate.
2,4-Dimethylphenol.
o-Dichlorobenzene.
m-Dichlorobenzene.
p-Dichlorobenzene.
Hexachlorobenzene.
2,4,6-Trichlorophenol.
Fluoranthene.
o-Nitrophenol.
1,2,4-Trichlorobenzene.
o-Chlorophenol.
Pentachlorophenol.
Pyrene.
Dimethyl phthalate.
Mononitrobenzene.
2,6-Toluene diisocyanate.
Polychlorinated dibenzo-p-dioxins.
Polychlorinated dibenzo-furans.
13. In Appendix IX to Part 266,
Section 2.0 of the Table of Contents and
the Appendix is revised to read as
follows:
Appendix IX to Part 266—Methods
Manual for Compliance With the BIF
Regulations
Table of Contents
*****
2.0 Performance Specifications and
Quality Assurance Requirements for
Continuous Monitoring Systems
2.1 Continuous emissions monitors
(GEMS).
2.2 Other continuous monitoring systems.
Section 2.0 Performance Specifications and
Quality Assurance Requirements for
Continuous Monitoring Systems
2.1 Continuous emissions monitors
(GEMS).
2.1.1 BIFs shall be equipped with GEMS
for compliance monitoring.
2.1.2 At all times that hazardous waste is
fed into the BIF or remains in the combustion
chamber, GEMS must be operated in
compliance with the requirements of the
appendix to suhpart EEE, part 63, of this
chapter.
2.2 Other continuous monitoring systems.
2.2.1 CMS other than GEMS (e.g.,
thermocouples, pressure transducers, flow
meters) must be used to document
compliance with the applicable operating
limits provided by this section.
2.2.2 Non-CEM CMS must be installed
and operated in conformance with
§ 63.8(c)(3) of this chapter requiring the
owner and operator, at a minimum, to
comply with the manufacturer's written
specifications or recommendations for
installation, operation, and calibration of the
system.
2.2.3 Non-CEM CMS must sample the
regulated parameter without interruption,
and evaluate the detector response at least
once each 15 seconds, and compute and
record the average values at least every 60
seconds.
2.2.4 The span of the detector must not be
exceeded. Span limits shall be interlocked
into the automatic waste feed cutoff system.
PART 270—EPA ADMINISTERED
PERMIT PROGRAMS: THE
HAZARDOUS WASTE PERMIT
PROGRAM
VIII. In part 270:
.
-------
27536
Federal Register / Vol. 61, No. 77 / Friday, April 19, 1996 / Proposed Rules
1. The authority citation for part 270
continues to read as follows:
Authority: 42 U.S.C. 6905, 6912, 6924,
6925, 6927,6939, and 6974.
2. Section 270.19 is amended by
adding a sentence at the end of the
introductory text to the section.
§ 270.19 Specific part B information
requirements for incinerators
* * * When an owner and operator
begin to comply (i.e., submit a
notification of compliance) with the
requirements of subpart EEE, part 63, of
this chapter, specific requirements of
§§ 264.343, 264.345, and 264.347 are
superseded by the subpart EEE
standards as provided by § 264.340(b).
*****
3. Section 270.22 is amended by
adding introductory text to read as
follows:
§ 270.22 Specific part B information
requirements for boilers and industrial
furnaces burning hazardous waste.
When an owner and operator of a
cement or lightweight aggregate kiln
begin to comply (i.e., submit a
notification of compliance) with the
requirements of subpart EEE, part 63, of
this chapter, specific requirements of
§§ 266.104 through 266.107 are
superseded by the subpart EEE
standards as provided by
§266.102(a)(2).
*****
4. In Appendix I to § 270.42, an entry
is added to section L.
Appendix I to § 270.42—Classification
of Permit Modification
Modification
Class
Modification
Class
L. Incinerators, Boilers, and
Industrial Furnaces
9.2 Initial Technology Changes
Needed to Meet Standards under
40 CFR Part 63 (Subpart EEE—
National Emission Standards for
Hazardous Air Pollutants From
Hazardous Waste Combustors)"
1 Class 1 modifications requiring prior Agen-
cy approval.
2 Denotes that this section will be dropped
from Appendix I 4 years following promulga-
tion of this rule.
5. Section 270.62 is amended by
adding introductory text and revising
paragraph (b)(2)(vii), to read as follows:
§ 270.62 Hazardous waste incinerator
permits.
When an owner and operator begin to
comply (i.e., submit a notification of
compliance) with the requirements of
subpart EEE, part 63, of this chapter,
specific requirements of §§ 264.343,
264.345, and 264.347 are superseded by
the subpart EEE standards as provided
by § 264.340(b).
*****
(b)* * *
(2)* * *
(vii) Procedures for rapidly stopping
waste feed, shutting down the
combustor, maintaining temperature in
the combustion chamber until all waste
exits the combustor, and controlling
emissions in the event of equipment
malfunction or activation of any ESV or
other bypass system including
calculations demonstrating that
emissions will be controlled during
such an event (sufficient oxygen for
combustion and maintaining negative
.pressure), and the procedures for
executing the "Contingency Plan"
whenever the ESV is used, thus causing
an emergency release of emissions.
6. Section 270.66 is amended by
adding introductory text to read as
follows:
§ 270.66 Permits for boilers and industrial
furnaces burning hazardous waste.
When an owner and operator of a
cement or lightweight aggregate kiln
begin to comply (i.e., submit a
notification of compliance) with the
requirements of subpart EEE, part 63 of
this chapter, specific requirements of
§ 266.104 through 266.107 are
superseded by the subpart EEE
standards as provided by
§266.102(a)(2).
*****
7. Section 270.72 is amended by
adding paragraph (b)(8) to read as
follows:
§270.72 Changes during interim status.
(b) * * *
(8) Changes necessary to comply with
standards under subpart EEE, part 63, of
this chapter (National Emission
Standards for Hazardous Air Pollutants
From Hazardous Waste Combustors).
PART 271—REQUIREMENTS FOR
AUTHORIZATION OF STATE
HAZARDOUS WASTE PROGRAMS
IX. In part 271:
1. The authority citation for part 271
continues to read as follows:
Authority: 42 U.S.C. 9602; 33 U.S.C. 1321
and 1361.
Subpart A—Requirements for Final
Authorization
2. Section 271.1(j) is amended by
adding the following entries to Table 1
in chronological order by date of
publication in the Federal Register to
read as follows:
§ 271.1 Purpose and scope.
TABLE 1 .—REGULATIONS IMPLEMENTING THE HAZARDOUS AND SOLID WASTE AMENDMENTS OF 1984
Promulgation date
Title of regulation
Federal Register reference
Effective date
[Insert date of publication of final
rule in the Federal Register
(FR)]..
Revised Standards for Hazardous
Waste Combustion Facilities.
[Insert FR page numbers].
[Insert date of publication of final
rule].
[FR Doc. 96-7872 Filed 4-18-96; 8:45 am]
BILLING CODE 6560-50-U
-------
------- |