United States Office of Air Quality
Environmental Protection Planning and Standards
Agency Research Triangle Park NC 27711
Air
EPA-453/R-94-003C
March 1994
ERA Hazardous Air Pollutant
Emissions from Process
Units in the
Synthetic Organic Chemical
Manufacturing Industry--
Background Information
for Final Standards
Final
EIS
Volume 2C: Comments on
Emissions Averaging
-------
v
0
EPA-453/R-94-003C
Hazardous Air Pollutant Emissions
from Process Units in the
Synthetic Organic Chemical
Manufacturing Industry--
Background Information
for Final Standards
Volume 2C: Comments on
Emissions Averaging
Emission Standards Division
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
March 1994
-------
DISCLAIMER
This Report has been reviewed by the Emission Standards
Division of the office of Air Quality Planning and Standards,
EPA, and approved for publication. Mention of trade names or
commercial products is not intended to constitute endorsement
or recommendation for use. Copies of this report are
available through the Library Services Office (MD-35), U.S.
Environmental Protection Agency, Research Triangle Park, NC
27711, or from the National Technical Information Service,
5285 Port Royal Road, Springfield, VA 22161.
-------
ENVIRONMENTAL PROTECTION AGENCY
Background Information and Final Environmental
Impact Statement for Hazardous Air Pollutant
Emissions from Process Units in the Synthetic
Organic Chemical Manufacturing Industry
Volume 2C: Comments on Emissions Averaging
Prepared by:
Bruce Jordan (Date)
Directop-r Emission Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
1. The standards regulate emissions of organic hazardous
air pollutants (HAP's) emitted from chemical
manufacturing process units in the Synthetic Organic
Chemical Manufacturing Industry (SOCMI) and from other
processes subject to the negotiated regulation for
equipment leaks. Only those chemical manufacturing
process units that are part of major sources under
Section 112 (d) of the Clean Air Act (Act) will be
regulated. The standards will reduce emissions of 112
of the organic chemicals identified in the Act list of
189 HAP's.
2. Copies of this document have been sent to the following
Federal Departments: Labor, Health and Human Services,
Defense, Transportation, Agriculture, Commerce,
Interior, and Energy; the National Science Foundation;
and the Council on Environmental Quality. Copies have
also been sent to members of the State and Territorial
Air Pollution Program Administrators; the Association
of Local Air Pollution Control Officials; EPA Regional
Administrators; and other interested parties.
3. For additional information contact:
Dr. Janet Meyer
Standards Development Branch (MD-13)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Telephone: (919) 541-5254
-------
4. Paper copies of this document may be obtained from:
National Technical Information Service (NTIS)
5285 Port Royal Road
Springfield, VA 22161
Telephone: (703) 487-4650
5. Electronic copies of this document may be obtained from
the EPA Technology Transfer Network (TTN). • The TTN is
an electronic bulletin board system which is free,
except for the normal long distance charges. To access
the HON BID:
• Set software to data bits: 8, N; stop bits: 1
• Use access number (919) 541-5742 for 1200, 2400,
or 9600 bps modems [access problems should be
directed to the system operator at (919)
541-5384].
• Specify TTN Bulletin Board: Clean Air Act
Amendments
• Select menu item: Recently Signed Rules
-------
OVERVIEW
Emission standards under section 112(d) of the Clean Air
Act (Act) apply to new and existing sources in each listed
category of hazardous air pollutant (HAP) emission sources.
This background information document (BID) provides summaries
and responses for public comments received regarding the
Hazardous Organic National Emission Standard for Hazardous Air
Pollutants (NESHAP), commonly referred to as the HON. The HON
will primarily affect the Synthetic Organic Chemical
Manufacturing Industry (SOCMI). However, the provisions for
equipment leaks also apply to certain polymer and resin
production processes, certain pesticide production processes,
and certain miscellaneous processes that are subject to the
negotiated regulation for equipment leaks.
This BID comprises six volumes as follows:
• Volume 2A: Comments on Process Vents, Storage
Vessels, Transfer Operations, and Equipment Leaks
(EPA-453/R-94-003a);
• Volume 2B: Comments on Wastewater
(EPA-453/R-94-003b);
• Volume 2C: Comments on Emissions Averaging
(EPA-453/R-94-003C);
• Volume 2D: ' Comments on Applicability, National
Impacts, and Overlap with Other Rules
(EPA-453/R-94-003d);
• Volume 2E: Comments on Recordkeeping, Reporting,
Compliance, and Test Methods (EPA-453/R-94-003e);
and
• Volume 2F: Commenter Identification List
(EPA-453/R-94-003f) .
Volume 2A is organized by emission point and contains
discussions of specific technical issues related to process
-------
vents, storage vessels, transfer operations, and equipment
leaks. Volume 2A discusses specific technical issues such as
control technology, cost analysis, emission estimates,
Group I/Group 2 determination, compliance options and
demonstrations, and monitoring.
Volume 2B addresses issues related to controlling
emissions from wastewater. Specific technical issues include
control technology, cost analysis, emission estimates,
Group I/Group 2 determination, compliance options and
demonstrations, and monitoring.
Volume 2C contains the EPA's decisions regarding
emissions averaging. Specific issues include the scope of
emissions averaging in the HON, specific provisions related to
credits and banking, and enforcement of an emissions averaging
system for the HON.
Volume 2D discusses applicability of the HON in terms of
selection of source category, selection of source, and
selection of pollutants. Volume 2D also addresses the process
for determination of the MACT floor and selection of the
specific applicability thresholds for process vents, storage
vessels, transfer racks, wastewater operations, and equipment
leaks.
Volume 2E discusses the provisions for compliance,
recordkeeping and reporting. Volume 2E also discusses issues
related to the use of EPA test methods.
Volume 2F of each volume contains a list of commenters,
their affiliations, and the EPA docket and item number
assigned to each comment.
-------
TABLE OF CONTENTS
Section Page
ACRONYM AND ABBREVIATION LIST ix
LIST OF FREQUENTLY USED TERMS xiii
1.0 INTRODUCTION 1-1
2.0 EMISSIONS AVERAGING 2-1
2.1 COST 2-3
2.2 LEGALITY OF EMISSIONS AVERAGING 2-6
2.3 SCOPE 2-11
2.3.1 Source Definition 2-11
2.3.2 Averaging at New Sources 2-20
2.3.3 Averaging Between New and Existing
Sources 2-22
2.3.4 Emission Points Allowed in Trades . . . 2-24
2.4 COMPLEMENTARY LEGAL INTERPRETATION FOR BROADER
EMISSIONS AVERAGING 2-35
2.4.1 Legality of Broader Averaging 2-35
2.4.2 Policy and Practical Considerations . . 2-42
2.5 CREDITS 2-44
2.5.1 General Issues 2-44
2.5.2 Use of RCT Above Rated Efficiencies . . 2-46
2.5.3 Credits for Previous Actions 2-51
2.5.4 Credit for Pollution Prevention and
Recycling 2-60
2.5.5 Plant Shutdowns and Slowdowns 2-64
2.5.6 Approval Process for New Control
Technologies 2-66
2.6 CREDIT DISCOUNT FACTORS 2-70
2.7 COMPLIANCE PERIOD 2-81
2.7.1 Averaging Period 2-81
2.7.2 Preclusion of Administrative
Enforcement 2-92
2.7.3 Quarterly Emissions Check 2-93
2.7.4 Alternative Proposal for Quarterly
Limit 2-99
VII
-------
TABLE OF CONTENTS, CONTINUED
Section Page
2.8 IMPLEMENTATION AND ENFORCEMENT 2-102
2.8.1 General Issues 2-102
2.8.2 Monitoring, Recordkeeping, and
Reporting 2-104
2.8.3 Administrative Burden 2-117
2.8.4 State Discretion on Emissions
Averaging . 2-127
2.8.5 Number of Points Allowed in Averages 2-134
2.8.6 Title V/Implementation Plan 2-141
2.9 RISK AND INTERPOLLUTANT TRADING 2-146
2.9.1 Risk in Emissions Averaging 2-146
2.9.2 Interpollutant Trading 2-150
2.9.3 Legal Issues 2-152
2.9.4 Approaches for Toxicity Weighting . . 2-156
2.9.5 Problems with Toxicity Weighting . . 2-161
2.9.6 Inclusion of Risk in Averaging
Determinations 2-164
2.9.7 Broader Scope for Averaging 2-170
2.10 BANKING 2-171
2.10.1 General Issues 2-171
2.10.2 Period of Availability 2-174
2.10.3 Use for Quarterly Compliance .... 2-176
2.10.4 Miscellaneous Issues 2-177
2.11 GENERAL POLICY AND MISCELLANEOUS ISSUES . . . 2-177
2.11.1 Precedent for Future Rule Makings . . 2-177
2.11.2 Simplifying the Language of the Rule 2-179
2.11.3 The Intent of Section 63.112 .... 2-181
2.11.4 Emissions Estimation 2-185
2.11.5 Miscellaneous Issues 2-189
viii
-------
ACRONYM AND ABBREVIATION LIST
ACRONYM
Act
ALAPCO
ASPEN
BACT
BAT
BD
BID
BIF
CEM
CFR
CMA
CMPU
CO
CTG
CWA
DMS
DOT
ORE
EB/S
EDC
EFR
EO
E.O.
EPA
Fe
Fm
FR
Fr
FTIR
HAP
TERM
Clean Air Act
Association of Local Air Pollution
Control Officers
advanced system for process
engineering
best available control technology
best available technology
butadiene
background information document
boilers and industrial furnaces
continuous emissions monitoring
Code of Federal Regulations
Chemical Manufacturers Association
chemical manufacturing process unit
carbon monoxide
control techniques guideline
Clean Water Act
dual mechanical seal
Department of Transportation
destruction and removal efficiency
ethylbenzene/styrene
ethylene dichloride
external floating roof
ethylene oxide
Executive Order
Environmental Protection Agency
fraction emitted
fraction measured
FEDERAL REGISTER
fraction removed
Fourier transform infrared
hazardous air pollutant
IX
-------
ACRONYM
ACRONYM
HON
IFR
LDAR
LAER
MACT
MIBK
MR
NCS
NESHAP
NPDES
NRDC
NSPS
NSR
OCCM
OCPSF
OMB
OSHA
P.L.
PAY
POM
POTW
PRA
PRV
PSD
QIP
AND ABBREVIATION LIST, CONTINUED
TERM
hazardous organic national emission
standards for hazardous air
pollutants
internal floating roof
leak detection and repair
lowest achievable emission rate
maximum achievable control technology
methyl isobutyl ketone
mass removal (actual)
Notification of Compliance Status
national emission standards for
hazardous air pollutants
nitrogen oxides
National Pollutant Discharge
Elimination System
Natural Resources Defense Council
new source performance standards
new source review
Office of Air Quality Planning and
Standards Control Cost Manual
organic chemicals, plastics, and
synthetic fibers
Office of Management and Budget
Occupational Safety and Health
Administration
Public Law
product accumulator vessel
polycyclic organic matter
publicly owned treatment works
Paperwork Reduction Act
pressure relief valve
prevention of significant
deterioration
quality improvement program
-------
ACRONYM
ACRONYM
R & D
RCRA
RCT
RIA
RMR
SARA
SIP
SMS
SOCMI
STAPPA
TAG
TACB
TCI
THC
TIC
TOC
TRE
TRI
TSDF
VHAP
VO
VOC
VOHAP
AND ABBREVIATION LIST, CONTINUED
TERM
research and development
Resource Conservation and Recovery
Act
reference control technology
Regulatory Impact Analysis
required mass removal
Superfund Amendment and
Reauthorization Act
State Implementation Plan
single mechanical seal
synthetic organic chemical
manufacturing industry
State and Territorial Air Pollution
Program Administrators
total annual cost
Texas Air Control Board
total capital investment
total hydrocarbon
total industry control
total organic compound
total resource effectiveness
toxics release inventory
treatment, storage, and disposal
facility
volatile hazardous air pollutant
volatile organics measurable by
Method 25D
volatile organic compound
volatile organic hazardous air
pollutant
ABBREVIATION
bbl
BOE
UNIT OF MEASURE
barrel
barrels of oil equivalent
xi
-------
ACRONYM AND ABBREVIATION LIST, CONTINUED
ABBREVIATION UNIT OF MEASURE
BtU
Btu/kW-hr
oF
gal
gpm
hr
kg/hr
kPa
kW-hr/yr
£/hour»m2
£pm
gal
m3
Mg
mg
mg/dscm
MW
ppb
ppm
ppmv
ppmw
psia
scm/min
TJ
yr
British thermal unit
British thermal unit per
kilowatt-hour
degrees Celsius
degrees Fahrenheit
gallon
gallons per minute
hour
kilograms per hour
kilopascals
kilowatt-hour per year
liters per hour per square meter
liters per minute
gallons
cubic meters
megagrams
milligrams
milligram per dry standard cubic
meter
megawatts
parts per billion
parts per million
parts per million by volume
parts per million by weight
pounds per square inch absolute
standard cubic meter per minute
terajoules
year
xii
-------
LIST OF FREQUENTLY USED TERMS
Act means the Clean Air Act as amended in 1990.
Administrator means the Administrator of the U. S.
Environmental Protection Agency or his or her authorized
representative (e.g., a State that has been delegated the
authority to implement the provisions of part 63).
Enhanced monitoring rule means the rule to be located in
sections 64.1 through 64.9 of part 64 of title 40 of the Code
of Federal Regulations. This rule implements section 702(b)
of title VII of the 1990 Clean Air Act Amendments. This rule
establishes the criteria and procedures that owners or
operators must satisfy in evaluating, selecting and
demonstrating enhanced monitoring, and includes appendices
containing enhanced monitoring performance and quality
assurance requirements. The enhanced monitoring rule does not
apply to sources subject to 40 CFR part 63, and therefore does
not apply to sources subject to the HON. The proposed rule
was published in the Federal Register on October 22, 1993
(58 FR 54648) .
General Provisions means the general provisions located in
subpart A of part 63 of title 40 of the Code of Federal
Regulations. These General Provisions codify national
emission standards for hazardous air pollutants (NESHAP) for
source categories covered under section 112 of the Act as
amended November 15, 1990.
Implementing agency means the Administrator of the U. S.
Environmental Protection Agency or a State, federal, or other
agency that has been delegated the authority to implement the
provisions of part 63. Under section 112(1) of the Act,
States and localities may develop and submit to the
Administrator for approval a program for the implementation
and enforcement of emission standards. A program submitted by
xiii
-------
LIST OF FREQUENTLY USED TERMS, CONTINUED
the State under section 112(1) of the Act may provide for
partial or complete delegation of the Administrator's
authorities and responsibilities to implement and enforce
emission standards.
Operating permit program rule means the rule located in
sections 70.1 through 70.11 of part 70 of chapter I of
title 40 of the Code of Federal Regulations. This rule
implements section 502(b) of title V of the 1990 Clean Air Act
Amendments. Under this rule, States are required to develop,
and to submit to the EPA, programs for issuing operating
permits to major stationary sources (including major sources
of hazardous air pollutants listed in section 112 of the Act),
sources covered by New Source Performance Standards (NSPS),
sources covered by emissions standards for hazardous air
pollutants pursuant to section 112 of the Act, and affected
sources under the acid rain program. The final rule was
published in the Federal Register on July 21, 1992
(57 FR 32250).
Permitting authority rneans: (1) the State air pc .ation
control agency, local agency, other State agency, or other
agency authorized by the Administrator to carry out a permit
program under part 70; or (2) the Administrator, in the case
of EPA-implemented permit programs under part 71.
Section 112fg) rule means the rule to be located in subpart B
of part 63 of title 40 of the Co i of Federal Regulations.
This rule implements section 112(g) of the 1990 Clean Air Act
Amendments. This rule will impose control technology
requirements on "constructed, reconstructed or modified" major
sources of hazardous air pollutants not already regulated by a
section 112(d) or 112(j) MACT standard.
xiv
-------
LIST OF FREQUENTLY USED TERMS, CONTINUED
Section 112(1) rule means the rule located in subpart E of
part 63 of title 40 of the Code of Federal Regulations. Under
this rule, a State or locality may submit a program to the
Administrator to request partial or complete delegation of the
Administrator's authorities and responsibilities to implement
and enforce section 112 emission standards. The final rule
was published in the Federal Register on November 26, 1993
(58 FR 62262). '
Title III means title III of the 1990 Clean Air Act
Amendments. Section 112 of the Act authorizes the EPA to
establish MACT standards.
Title V means title V of the 1990 Clean Air Act Amendments,
which authorizes the EPA to establish the operating permit
program.
Title VII means title VII of the 1990 Clean Air Act
Amendments. Section 702(b) of the Act authorizes the EPA to
establish compliance certification procedures. The part 64
enhanced monitoring rule implements section 702(b).
xv
-------
2.0 EMISSIONS AVERAGING
Comment: Several commenters (A-90-19: IV-D-32; IV-D-33;
IV-D-48; IV-D-50; IV-D-55; IV-D-56; IV-D-57; IV-D-58; IV-D-59;
IV-D-62; IV-D-63; IV-D-67; IV-D-69; IV-D-71; IV-D-72; IV-D-73;
IV-D-74; IV-D-75; IV-D-77; IV-D-79; IV-D-80; IV-D-81; IV-D-82;
IV-D-83 and IV-F-1.3 and IV-F-5; IV-D-86; IV-D-92; IV-D-97;
IV-D-98; IV-D-104; IV-D-106; IV-D-108; IV-D-112; IV-D-113;
IV-F-1.1 and IV-F-3; IV-F-1.6 and IV-F-6; IV-F-7.41; IV-G-1;
IV-G-16; IV-G-17) supported the EPA's proposal to allow the
use of emissions averaging to comply with subpart G. Some of
the reasons listed by commenters include: emissions averaging
will reduce compliance costs and improve cost effectiveness;
it will encourage pollution prevention and the development of
innovative control technologies; and it is consistent with the
express requirement in section 112 of the Act to consider cost
in developing MACT standards.
Several commenters (A-90-19: IV-D-9; IV-D-10; IV-D-11;
IV-D-41; IV-D-45 and IV-F-7.7; IV-D-49; IV-D-70; IV-D-85 and
IV-F-7.39 and IV-F-12 and IV-G-6 and IV-G-8; IV-D-90; IV-D-93;
IV-D-96; IV-D-99; IV-D-100; IV-D-103 and IV-F-7.5; IV-D-103
and IV-F-7.40; IV-D-115; IV-D-117 and IV-F-7.43; IV-D-118;
IV-D-120; IV-D-122; IV-D-123; IV-D-124; IV-D-125; IV-F-1.5;
IV-F-7.1; IV-F-7.2; IV-F-7.6; IV-F-7.21; IV-F-7.23; IV-F-7.26;
IV-F-7.33; IV-F-7.34; IV-F-7.35; IV-F-7.36; IV-F-7.42;
IV-F-7.44; IV-F-7.45) opposed the EPA's proposal to allow the
use of emissions averaging to comply with subpart G or were
opposed to specific features of emissions averaging. Some of
the reasons listed by commenters include: emissions averaging
could increase risks to health and the environment; it will
result in emission reductions less than the maximum
2-1
-------
achievable, hence, it is inconsistent with section 112 of the
Act; and it raises enforcement concerns.
One commenter (A-90-19: IV-D-70) was concerned that
emissions averaging would: (1) create a needless third level
of regulatory issues; (2) be a source of problematic questions
which would slow down and undermine air pollution control
efforts; and (3) require needless continuing policy and
procedure development. The commenter (A-90-19: IV-D-70)
stated that if the driving force for the emissions averaging
program is a concern that the MACT standards as proposed may
not be economically reasonable or appropriate for certain
source types, then these issues should be resolved in the
standard itself by specific exemptions or cutoff levels.
Two commenters (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12; IV-F-1.5) suggested that even without emissions
averaging, the rule would provide reasonable flexibility for
sources that want to use alternative emissions reduction
techniques, but at the same time would ensure that real
reductions do occur.
Response: Emissions averaging has been maintained in the
final rule as an option fo-r sources to use to comply with
subpart G of the rule. This decision is in keeping with the
EPA's general policy of encouraging the use of flexible
compliance approaches where they can be properly monitored and
enforced. Under particular circumstances, emissions averaging
can provide sources the flexibility to comply in the least
costly manner while still maintaining a regulation that is
workable and enforceable. The EPA's goal in crafting the
emissions averaging provisions in the final rule has been to
make emissions averaging available to sources faced with some
emission points that are particularly difficult or costly to
control. At the same time, the EPA has simplified and
streamlined the emissions averaging provisions in order to
ease the enforcement burden on implementing agencies.
The rationale for the specific provisions of the
emissions averaging policy is detailed throughout this BID
volume. In general, the basic structure of the HON emissions
2-2
-------
averaging policy remains much the same as at proposal.
Fundamental elements such as the credit/debit system, kinds of
emission points allowed in averages, reference control
efficiency provisions, provisions for approval of new devices,
and an annual compliance period remain unchanged.
However, some provisions have been altered or added in
order to sharpen the focus of emissions averaging, ease
implementation and administration, and ensure at least the
same air quality benefit as point-by-point compliance. For
example, the number of emission points that can be included in
an average has been limited; banking of credits has been
disallowed; actions taken prior to November 15, 1990 will not
be credited; averaging will not be allowed at new sources; and
a discount factor of 10 percent will be applied to credits
generated by control other than pollution prevention measures.
In addition, sources must demonstrate, to the implementing
agency's satisfaction, that a proposed averaging plan will not
cause an increase in risk or hazard relative to point-by-point
controls. All of these changes are discussed in greater
detail throughout this BID volume.
2.I COST
Comment: Several commenters (A-90-19: IV-D-33; IV-D-48;
IV-D-55; IV-D-58; IV-D-59; IV-D-62; IV-D-67; IV-D-72; IV-D-73;
IV-D-74; IV-D-77; IV-D-83; IV-D-86; IV-D-98; IV-D-106;
IV-D-108; IV-D-112; IV-D-113; IV-G-1; IV-G-16; IV-G-17)
asserted that emissions averaging will allow sources to
achieve the mandated reductions more cost-effectively.
One commenter (A-90-19: IV-D-83 and IV-F-1.3 and IV-F-5)
predicted that emissions averaging will encourage the greatest
reductions as early as possible at significantly reduced
costs. Two commenters (A-90-19: IV-D-83 and IV-F-1.3 and
IV-F-5; IV-G-l) added that emissions averaging will provide an
additional incentive for sources to develop innovative control
technologies. Two commenters (A-90-19: IV-G-16; IV-G-17)
provided examples of where highly controlled emission points
fall short of meeting MACT, but through emissions averaging,
2-3
-------
can still achieve the required reductions in the most cost-
effective manner.
Nine commenters (A-90-19: IV-D-32; IV-D-33; IV-D-48;
IV-D-73; IV-D-83; IV-D-104; IV-D-112; IV-F-1.6 and IV-F-6;
IV-F-7.41) promoted emissions averaging for the instances
where MACT requirements will be "exceptionally high" for some
emission points or sources, and therefore not cost effective,
or where emissions averaging is the only "reasonable means" of
achieving compliance. One commenter (A-90-19: IV-D-83) noted
that the EPA draft RIA finds that HON compliance costs vary
widely from source to source, and in some "... cases, cost
increases can be in excess of 100 percent of market price."
Three commenters (A-90-19: -IV-D-33; IV-D-58; IV-F-1.6
and IV-F-6) maintained that emissions averaging may assist
facilities having unusually high MACT costs to improve cost
effectiveness and maintain a competitive edge relative to
other facilities. One commenter (A-90-19: IV-D-33) promoted
the advantage of maintaining competitiveness particularly for'
facilities whose products are sold in the worldwide
marketplace.
Response: The primary reason for allowing emissions
averaging as an alternative to point-by-point compliance with
RCT is that emission reductions equal to or greater than under
point-by-point compliance can still be achieved. At the same
time, emissions averaging can provide sources the flexibility
to comply in the least costly manner. As long as equivalent
reductions can be achieved, the EPA considers it appropriate
to increase regulatory flexibility.
Although the EPA appreciates the sentiments expressed
regarding cost savings, the EPA disagrees with the
implications that emissions averaging may be the only
"reasonable means" of achieving compliance. It is not
anticipated that emissions averaging would be the only
reasonable means of achieving compliance in any case. Even
though some owners or operators will realize significant cost
savings through emissions averaging, there is sufficient
flexibility provided in the point-by-point RCT compliance
2-4
-------
requirements that it will always be a reasonable strategy for
achieving reductions.
The draft RIA does contain the finding that compliance
costs can vary widely among manufacturers. However, as
emphasized throughout the draft RIA and other supporting
documents, compliance cost estimates were provided for the TIC
option. Under TIC, it is assumed that all emission points are
controlled without exclusions, and moreover, all emission
points are controlled individually, not ducted to common
control devices. Therefore, the wide variations in compliance
cost, especially those resulting in price increases in excess
of 100 percent of market price, represent a worst-case level
of variation. ' In reality, many emission points will be
classified as Group 2 points not requiring control, and the
total cost of control at most, if not all, facilities will be
less than predicted in the draft RIA.
Comment; Two commenters (A-90-19: IV-D-32; IV-D-86)
anticipated that the use of emissions averaging will be
limited to a few circumstances such as where use of RCT is
impracticable. One of the commenters (A-90-19: IV-D-32)
predicted that emissions averaging will be used primarily
where, due to special circumstances associated with a
particular Group 1 point, the cost of RCT for that point is
much in excess of the average relied upon by the EPA in
selecting that RCT. The commenter (A-90-19: IV-D-32)
predicted that emissions averaging will not enable the
industry to save money in comparison to the EPA's projected
costs for RCT. The other commenter (A-90-19: IV-D-86)
suggested that emissions averaging may have an insignificant
impact on the overall economic impact of the rule.
Response; As indicated previously, the EPA does not
anticipate that emissions averaging would be the only
reasonable means of achieving compliance for any case.
Neither commenter provided information on how likely it might
be that installing RCT would be impracticable. However,
emissions averaging was included in the rule to allow owners
2-5
-------
or operators the flexibility to make such a determination on a
site-specific basis.
The EPA agrees with the comraenters that emissions
averaging may be desirable for only a limited number of
emission points in any source and in fact, stated as much in
the proposal preamble. However, even though emissions
averaging may be used for only a small number of points, it
should still enable the industry to save money in comparison
to projected costs for RCT. Emissions averaging may not
enable sources to reduce their costs to or below the industry
average; nevertheless, sources will incur lower costs than
they would if point-by-point compliance were the only option
available. Otherwise, the source would be unwise to choose
emissions averaging.
The result of lowering control costs for some emission
points will be that the national average cost will be reduced."
The range of compliance costs experienced throughout the
industry will be reduced as well. The EPA cannot specifically
address the claim that cost reductions will be insignificant
because there is not sufficient data to make a specific
estimate of the extent to which emissions averaging will be
used. However, judging from the extensive comment supporting
the use of emissions averaging, the EPA anticipates that
industry will find ample opportunity for realizing more than
insignificant cost savings.
2.2 LEGALITY OF EMISSIONS AVERAGING
Comment: Several commenters (A-90-19: IV-D-32; IV-D-48;
IV-D-57; IV-D-62; IV-D-72; IV-D-74; IV-D-75; IV-D-77; IV-D-79;
IV-D-83 and IV-F-1.3 and IV-F-5; IV-D-86; IV-D-98; IV-D-104;
IV-D-106; IV-D-108) considered emissions averaging to be
consistent with section 112(d) of the Act. One commenter
(A-90-19: IV-D-104) also considered emissions averaging to be
consistent with section 112(i) of the Act. Another commenter
(A-90-19: IV-D-75) suggested that emissions averaging is
further supported by the statute in sections 112(h) and (j).
One commenter (A-90-19: IV-D-62) repeated the statement
in the proposal preamble that the EPA is not prohibited from
«
2-6
-------
allowing a source to meet MACT through use of emissions
averaging as long as every source in the category must comply
and the standard is at least as stringent as the MACT floor.
Seven commenters (A-90-1'9: IV-D-48; IV-D-62 ; IV-D-74;
IV-D-77; IV-D-83 and IV-F-1.3 and IV-F-5; IV-D-98; IV-D-108)
reiterated that section 112(d) expressly requires cost to be
considered in setting the MACT standard. Hence, two
commenters (A-90-19: IV-D-98; IV-D-83 and IV-F-1.3 and
IV-F-5) reasoned that because emissions averaging is a cost-
effective way of achieving the reductions required by the
standard, it is consistent with section 112(d).
Four commenters (A-90-19: IV-D-32; IV-D-57; IV-D-62;
IV-D-77) listed other factors that the Act requires be
considered in defining MACT such as non-air quality
environmental impacts and energy impacts. The commenters
(A-90-19: IV-D-32; IV-D-57; IV-D-62; IV-D-77) suggested that
because emissions averaging allows sources to take these
factors into account on an emission point-specific basis,
emissions averaging allows MACT to be fine-tuned and
implemented more completely. Two commenters (A-90-19:
IV-D-32; IV-D-57) referenced section 112(d)(2) and Senate
Report (S.Rep.) No. 228, 101st Cong., 1st Sess. 167 (1989).
Two commenters (A-90-19: IV-D-48; IV-D-83 and IV-F-1.3
and IV-F-5) further justified the use of emissions averaging
based on the direction to the EPA from Congress to implement,
whenever possible, market-based regulatory schemes for
achieving emissions reductions.
Four commenters (A-90-19: IV-D-32; IV-D-57; IV-D-74;
IV-D-106) interpreted the statute; specifically
section 112(h), as requiring the EPA to promulgate a numerical
emissions limit as MACT where feasible rather than design,
equipment, work practice, or operational standards, leaving it
to individual sources to meet that limit. Another commenter
(A-90-19: IV-D-98) found the same conclusion in Adamo
Wrecking v. United (1978) where section 112(d)(2) was
interpreted to mean that the EPA is authorized to establish
numerical limitations on air emissions to be achieved through
2-7
-------
the application of any control technology. One commenter
(A-90-19: IV-D-32) reasoned that although the proposed HON
specifies control requirements, because it allows sources to
achieve equivalent reductions through emissions averaging, it
is fully harmonious with section 112(h).
One commenter (A-90-19: IV-D-83) countered arguments
that emissions averaging would result in greater emissions.
The commenter (A-90-19: IV-D-83) pointed out that the
averaging provisions require sources to submit for approval an
Implementation Plan that demonstrates no net increase in HAP
emissions and that detailed monitoring is required. The
commenter (A-90-19: IV-D-83) therefore concluded that the HAP
reductions achieved under emissions averaging will be at least
as great as the total emissions reductions required on a
point-by-point basis. The commenter (A-90-19: IV-D-83)
argued that these emission reductions will translate into
substantial improvements in local air quality with or without
emissions averaging.
Two commenters (A-90-19: IV-D-92; IV-D-113) considered
emissions averaging to be neutral so that the total emissions
are no greater than what would be achieved with strict
application of the RCT.
In contrast, four commenters (A-90-19: IV-D-41; IV-D-45
and IV-F-7.7; IV-D-85 and IV-F-7.39 and IV-F-12; IV-F-7.43)
claimed that the proposed averaging scheme violates the law.
One commenter (A-90-19: IV-D-41) stated that the Act
does not promote emissions averaging. Another commenter
(A-90-19: IV-D-45) stated that although Congress instructed
the EPA to consider cost when evaluating MACT, they did not
intend to let polluters avoid control of point sources in
favor of a "bubble."
Three commenters (A-90-19: IV-D-45 and IV-F-7.7;
IV-D-70; IV-F-7.43) asserted that emissions averaging is not a
permissible application of MACT. One commenter (A-90-19:
IV-D-117) claimed that emissions averaging weakens the HON,
which is a violation of the MACT standard. Three commenters
(A-90-19: IV-D-85; IV-D-87; IV-D-96) contended that because
2-8
-------
of emissions averaging, the rule fails to achieve "maximum
achievable emissions reductions" as required under
section 112(d)(2) of the Act. Two commenters (A-90-19:
IV-D-90; IV-D-100) stated that it does not result in
continuous emission reductions achievable under MACT
standards.
Two commenters (A-90-19: IV-D-90; IV-D-100) disagreed
that the EPA has statutory authority to allow emissions
averaging to comply with MACT. The commenters (A-90-19:
IV-D-90; IV-D-100) reasoned that because Congress specified
the use of "offsets" in lieu of control technology
requirements for significant modifications in section 112(g)
and did not make any similar references in section 112(d), it
is unlikely that they intended to provide emissions averaging
as a compliance option for MACT standards. The commenters
(A-90-19: IV-D-90; IV-D-100) also disagreed with the
reasoning stated in the proposal preamble that emissions
averaging is allowed as long as every source "is required to
comply, averaging does not cross source boundaries, and the
standards are at least as stringent as the floor."
Ten commenters (A-90-19: IV-D-49; IV-D-51; IV-D-70;
IV-D-85 and IV-F-7.39 and IV-F-12; IV-D-93; IV-D-96; IV-D-99;
IV-D-115; IV-D-117 and IV-F-7.43; IV-F-1.5) doubted that an
emissions averaging system could actually achieve the same
level of emission reductions as a regulation based on RCT's
without emissions averaging. One commenter (A-90-19:
IV-D-103) contended that the rule fails to explain how
averaging will provide greater reductions than other programs,
such as traditional permit programs. The commenter (A-90-19:
IV-D-103) claimed that emissions averaging increases emissions
and an increase in emissions of a toxic chemical is a clear
violation of the intent of the law, and an increase without
adequate demonstration of any floor is contrary to
requirements under the law.
Response: The EPA has thoroughly reviewed all of the
comments received concerning the legality of averaging and has
concluded that emissions averaging is legally permissible
2-9
-------
under section 112 of the Act. Thus, the EPA agrees with the
conclusions of those commenters who contended that averaging
is permissible under the Act and disagrees with those who
contended that averaging was not permissible under
section 112.
Section 112(d) requires standards to be established for
each category or subcategory of sources listed under
section 112(c). Such standards shall then be applicable to
sources within those categories or subcategories. The statute
does not define source category, nor does it impose precise
limits on the Administrator's discretion to define source. In
this case, the Administrator has exercised that discretion to
define source so as to include all emission points related to
SOCMI production at a facility.
In setting the standard for a category or subcategory,
the Administrator is required to determine a floor for the
entire category or subcategory, and then set a standard
applicable to each source within that category that is at
least as stringent as the floor and requires the maximum
achievable emission reductions considering certain other
factors. In determining whether the standard should be more
-tringent than the floor and by how much, the Administrator is
-o consider, among other factors, the cost of achieving the
additional emission reductions. The Act does not limit how
the standard is to be set beyond requiring that it be
applicable to all sources in a category, be written as a
numerical limit wherever feasible, and be at least as
stringent as the floor. Therefore, the relevant statutory
language is broad enough to permit the Administrator to
exercise discretion to allow sources to meet MACT through the
use of emissions averaging provided the standard applies to
every source in the category, averaging does not cross source
boundaries, and .the standard is no less stringent than the
floor.
The averaging system established by this rule stays
within those legal parameters. The source has been defined to
include all SOCMI processes within a major source, and a
2-10
-------
standard has been written to apply to all sources in the
category as provided by sections 112(d)(1) and (2) of the Act.
This standard is no less stringent than the floor for the
category, calculated in accordance with section 112(d)(3), and
takes cost and other relevant factors into consideration. The
standard applies only to sources in the category, applies to
all such sources, and is written as a numerical limit where
feasible. Moreover, averaging can only be conducted within
the confines of each individual source, thus ensuring that the
standard, as applied to each source, is no less stringent than
the floor. In addition, a credit discount factor is applied
when averaging is used, which further ensures that averaging
will be at least as stringent as the rule without averaging.
Specific discussion of the discount factor is included in
section 2.6 of this BID volume.
The averaging system adopted in this rule will not result
in greater emissions of HAP's than the rule without averaging,
although the precise composition of the HAP's emitted from a
source may differ from that which would occur without
averaging. However, the provisions in the final rule
regarding a demonstration to the implementing agency that risk
will not be higher with averaging than without averaging will
ensure that the use of averaging does not increase risk.
Thus, the averaging system established by this rule will
result in neither greater emissions of HAP's nor an increase
in risk when compared with compliance without averaging.
Moreover, because averaging is not permitted between sources
or facilities (as discussed in section 2.4 of this BID
volume), emissions cannot be increased at one source or
facility as a consequence of reductions at another source or
facility. The EPA maintains that an averaging program such as
the one established by this rule is fully consistent with the
Act.
2 . 3 SCOPE
2.3.1 Source Definition
Comment; Three commenters (A-90-19: IV-D-45 and
IV-F-7.7; IV-D-70; IV-D-85 and IV-F-7.39 and IV-F-12 and
2-11
-------
IV-G-9) argued that the source definition picked to justify
emissions averaging violates the Act and is inappropriate as a
matter of policy, and urged the EPA to define the source as
the emitting unit for purposes of the HON rule.
One commenter (A-90-19: IV-D-85 and IV-G-9) stated that
the source definition in the rule conflicts with statutory
language, past practice under section 112, and the legislative
history of the 1990 Amendments, and presented an extensive
discussion of their interpretation of the definition of
source. The commenter stated that section 112(d)(2) requires
emissions standards for each "new or existing source," and
sections 112(a)(4) and (10) define the terms "new source" and
"existing source" by reference to the term "stationary
source." The commenter noted that section 112(a)(3) states:
"The term 'stationary source1 shall have the same meaning as
such term has under section 111(a)." [Emphasis added.] The
commenter reasoned that the term "stationary source" as
applied to the SQCMI under section 111 means emission unit,
rather than an entire plant or a collection of points
associated with certain kinds of processes.
The commenter stated that the meaning of the term
"stationary source" under section 111 is based on the judicial
construction in Asarco. Inc. v. EPA. 578 F.2d 319 (D.C. Cir.
1978), and maintained that Asarco rejected "bubbles." The
commenter stated that the EPA explained a few months prior to
passage of the 1990 amendments to the Act that the main
purpose of section 111 is to apply Best Demonstrated
Technology (BDT) to all new, modified, or reconstructed
sources, and that because of this, a much narrower stationary
source definition has applied to equipment within the SOCMI
under the NSPS program. The commenter added that the
legislative history confirms the literal meaning of the
statute, that source definitions under section 112 were to be
the same as source definitions under section 111.
The commenter stated that the term "major source" is
defined in section I12(a)(l) of the Act as "any stationary
source or group of stationary sources." The commenter further
2-12
-------
maintained that a major source refers to the plant as a whole
or any collection of stationary sources within a plant
emitting 10 tons or more of a toxic air pollutant. The
commenter contended that Chevron USA. Inc. v. NRDC (hereafter
referred to as Chevron), 467 U.S. 842-43 (1984) does not
authorize the EPA to ignore plain statutory language linking
section 112's definition of stationary source to prior
regulatory decisions under section 111(a).
The commenter stated that the legislative history
confirms that Congress intended a narrower stationary source
definition for those source categories involving different
kinds of emission points. The commenter quoted the Senate
Report on S1630 at 168 [emphasis added]:
. . . a particular VOC may be released from both a stack
and from non-point sources in the facility. In [this]
case, MACT will be determined for each type of emissions
point and not for the facility as a whole.
The commenter contended that the HON regulates plants in
precisely the kind of situation referred to in the Senate
Report, but the HON allows sources to determine MACT for its
HON process units as a whole or any part of it, rather than
determining "MACT for each type of emission point" and
requiring compliance as Congress intended. The commenter,
quoting the Senate Report at 101-228, stated that the Senate
Report specifically warned against this departure from past
practice under section 111 [emphasis added]:
Amendments to section 112(a) made by the bill also adopt
a definition of "stationary source" different than used
in current law. A stationary source is defined to
include any particular unit of a facility or installation
. . . in addition to the facility or installation itself.
This definition is intended to prevent "bubbling" within
facilities.
The commenter stated that the Senate Report's statement
refers to the source definition in the Senate bill, and the
House bill contained the language ultimately adopted. The
commenter contended that the adopted definition makes the
intention to exclude "bubbles," at least in this kind of
industry even clearer ,by referring to the meaning of the term
"stationary source" "under" section 111(a). The commenter
2-13
-------
(A-90-19: IV-D-85 and IV-G-9) stated that stationary source
generally means unit "under" section 111, at least when plant
sites are made up of different types of emission points.
The commenter contended that the stationary source
definition in the Senate bill which was intended to preclude
"bubbles" closely resembles the definition in
section lll(a)(3) upon which the EPA relies. The commenter
stated that the section 111(a)(3) definition refers to any
"building, structure, facility, or installation," and the
definition in the Senate bill refers to "any facility or
installation or unit of such facility or installation." The
commenter contended that even if Congress intended section 112
stationary source definitions to follow the language rather
than the practice of section 111, it can hardly be seen as a
repudiation of the Senate's intention to preclude "bubbles"
such as the one proposed in this rule.
The commenter discussed statements made by Senator
Durenberger, and contended that the Senator stated that a
broad definition would be inappropriate if the group of plant
lacked "similar configurations." The commenter stated that
the Senator compared two alternatives in a draft EPA paper
entitled "Definition of Source: Range of Alternatives" and
rejected alternative 3(b) identified in the paper, which would
focus MACT standards on entire plant sites. The commenter
stated that instead, the managers, speaking through Senator
Durenberger, endorsed alternative 3(a), which focuses MACT
standards "on a specific portion of a contiguous
facility ..." (Cong. Rec. S16927, October 27, 1990). The
commenter contended that the EPA acknowledges in the HON that
the SOCMI source category consists of plants using various
configurations of pollution-emitting units. The commenter
argued that Senator Durenberger states unequivocally that the
EPA should set standards "for logical parts" of plants, which
like the plants in the SOCMI category "consist of various
pollution-emitting units in a variety of configurations (Cong.
Rec. S16928).
2-14
-------
The coitunenter contended that the legislative history
refers to MACT standards over and over again as "technology-
based" standards, which evinces an intent to "ban the bubble"
and focus on logical parts of plants to which a technology is
applied.
The commenter maintained that Chevron only allows agency
discretion when the legislative history fails to speak to the
precise point at issue. The commenter argued that when the
legislative history speaks to the precise point at issue,
meshes perfectly with the literal statutory language, and the
position urged by the agency enjoys no explicit support
whatsoever, the EPA must heed Congress1 intent as revealed in
the legislative history and language.
The commenter stated that the EPA has argued in its Early
Reductions rule (57 FR 61970; December 29, 1992) that the
Senate managers' statement only meant to preclude plant-wide
definitions when plant-wide definitions would cause a small
source category. The commenter" argued that this is a
misreading focusing on one sentence taken entirely out of
context. The commenter stated that the comments as a whole
reveal that differently configured sites must have MACT
standards for their components.
The commenter stated that the EPA assumed that Congress
meant to say that the definition of stationary source under
section 112 need not comport with the meaning it has had under
section lll(a)(3) provided it comports with the statutory
language in section lll(a)(3). The commenter maintained that
the EPA relied on Chevron's holding in the title I context
that this language is ambiguous and that absent a specific
indication of intent by Congress, the EPA's policy judgement
demands deference.
The commenter contended that even if the language of
section 111(a)(3) is broad enough to refer to an entire plant
or an emitting unit and is unaccompanied by legislative
history speaking to the point at issue, it is not broad enough
to encompass the source definition in the HON proposal. The
commenter stated that the proposal defines the source as "the
2-15
-------
set of emission points in the organic HAP-emitting processes
used to produce synthetic organic chemicals that are in a
contiguous area under common control" (57 FR 62613). The
commenter maintained that this definition specified in the
proposal and reflected in the emissions averaging provisions
and applicability criteria does not require that the emission
points be contiguous or part of the same process train;
rather, they must be "in a contiguous area," i.e., in a plant,
but the points themselves may be far apart from each other and
not part of the same process. The commenter stated that this
definition does not describe a "building," a "structure," a
"facility," or an "installation;" rather, it describes several
unrelated parts of a plant, at least with respect to a plant
with more than one SOCMI process. The commenter concluded
that hence, the definition is inconsistent with the language
of section lll(a)(3).
The commenter maintained that the EPA has stated in the
Early Reductions rule that "an 'installation' suggests some
type of unit that undertakes a particular function, such as
wastewater treatment system." The commenter argued that this
conception, if it were correct and consistent with
Congressional intent, could not justify a system in which the
plant owner designs the source by choosing groups of emission
points from different process units or from different kinds of
emission points.
In contrast, five commenters (A-90-19: IV-D-51; IV-D-62;
IV-D-63; IV-D-69; IV-G-1) supported the definition of source
in the rule, which accommodates the concept of emissions
averaging. One commenter (A-90-19: IV-D-51) considered the
definition a valid approach based on the justification
presented by the EPA. Another commenter (A-90-19: IV-D-63)
agreed that defining source as a collection of emission points
incorporates the flexibility necessary to implement an
emissions averaging.program.
Ten commenters (A-90-19: IV-D-32; IV-D-48; IV-D-57;
IV-D-62; IV-D-74; IV-D-83; IV-D-92; IV-D-98; IV-D-104;
IV-D-113) considered emissions averaging consistent with
2-16
-------
section 112 of the Act because sections 112(d) and (i) require
sources, not individual emission points within sources, to
comply with MACT. Hence, four commenters (A-90-19: IV-D-32;
IV-D-57; IV-D-62; IV-D-113) regarded as without merit the
argument that allowing emissions averaging does not satisfy
the MACT floor. Two commenters (A-90-19: IV-D-32; IV-D-57)
contended that this argument confuses "sources" with "emission
points," and that "sources," not "emission points," must
comply with MACT.
One commenter (A-90-19: IV-G-1) presented specific legal
arguments in support of the EPA's definition of "source" to
accommodate emissions averaging. The commenter (A-90-19:
IV-G-1) stated that:
"MACT source" averaging is entirely consistent with the
Agency's historic discretion to define "source" based on
the overall purposes of the particular program, as well
as the Amendments' endorsement through silence of that
discretion. See, e.g.. Chevron, USA, Inc. v. NRDC, 467
U.S. 837 (1984). That is particularly true where a
"compliance bubble" which assures MACT-equivalent
reductions—not an "applicability bubble" which allows
otherwise-covered emission points to escape such
reduction requirements—is involved. Cf.. e.g., Asarco.
Inc. v. EPA. 578 F.2d 319 (D.C. Cir. 1978). Indeed, such
"MACT source" averaging is a fortiori supported by EPA's
repeated recognition that similar technology-based
requirements mean RACT- or NSPS-equivalent reductions,
not uniform controls on every regulated facility, point
or source. e.g.. NRDC v. EPA (American Cyanamid). 33 ERC
1657 (4th Cir. 1991) ; NSPS Compliance Bubble Policy.
(52 FR 28946, 28954; Aug. 4, 1987). See also Emissions
Trading Policy Statement. (51 FR 43829; December 4, 1986)
(generally authorizing VOC RACT trades raising HAP
implications within the same plant so long as any
proposed or final NESHAP is the baseline, or where the
HAP emissions stream is "traded down"). As Senator
Durenberger, the principal author of what became new
section 112, expressly noted, MACT was to function like
technology-based effluent guidelines under the CWA. See.
e.g.. 136 Cong. Rec. S516 (Jan. 30, 1990) . Those
guidelines have long allowed categorical averaging
between different outfalls at the same plant. See 49 FR
21024 (May 17, 1984); Krueger, "Implementing the Bubble
Policy Under the Clean Water Act," 4 Virginia J. Nat. R.
Law 155 (1984)."
2-17
-------
Response: The EPA has reviewed the comments relating to
the definition of "source" used in this rule, and has
concluded that no change to the definition is warranted.
The EPA began by creating a list of source categories as
required by section 112(c) of the Act. Section 112(c)
requires that "to the extent practicable, the categories and
subcategories listed under this subsection shall be consistent
with the list of source categories established pursuant to
section 111 and part C." As is clear from a review of those
existing lists, the categories listed are generally broadly
drawn. Listing SOCMI as a category on the section 112(c) list
(57 FR 31576, July 16, 1992) is consistent with the general
broad categorization of the section 111 and part C lists.
Section 112(d) directs the Administrator to set standards
for all "major sources" within every listed category. Major
sources are "stationary sources," or groups of stationary
sources, of a given size, as defined in section 112(a)(1).
The definition of "stationary source" included in section 112
is identical to the definition used in section 111(a) which is
"any building, structure, facility, or installation which
emits or may emit any air pollutant." 42 U.S.C. 7411(a).
However, section 112 as amended, does not require that the
standards set under section 112(d) be set for the same
components of the categories as was done under section 111.
Thus, there is no requirement that section 112(d) standards
for sources in the SOCMI be set for precisely the same
portions of the industry as the NSPS.
As the Supreme Court has recognized in Chevron. the EPA
has broad discretion to define "source." The Court recognized
in Chevron that if any Congressional intent can be discerned
from the statutory language of section lll(a)(3) (the
definition of "source" adopted in section 112), "the listing
of overlapping, illustrative terms was intended to enlarge,
rather than confine, the scope of the EPA's power to regulate
particular sources in order to best effectuate the policies of
the Act." Chevron. Thus, the court found that a "source" can
encompass "any discrete, but integrated operation, which
2-18
-------
pollutes." As such, it could also encompass an entire plant,
and the EPA has flexibility, within the broad definition of
"stationary source," to define the source for each
section 112(d) standard as broadly or narrowly as is
appropriate for the particular industry being regulated.
Several commenters supported the EPA definition of source
and disagreed with one commenter who argued that a source
should be limited to an emitting unit. The EPA disagrees with
the commenter who argued that the proposed definition of
"source" for this rule violates the Act and should have been
limited to an "emitting unit." The statute clearly states
that the EPA is to set standards for categories of "source."
It does not restrict the EPA's authority to emitting units.
As discussed above, the Chevron decision makes clear that a
source is a flexible term that the EPA has broad discretion to
define in the context of each rulemaking. The EPA also
disagrees with the commenter's argument that the EPA has
ignored the plain statutory language linking the definition of
"source" in section 112 of the Act to the definition in
section 111(a). The EPA believes that the definition of
"source" used in this rule is consistent with "any building,
structure, facility, or installation which emits or may emit
any air pollutant," [42 U.S.C. 7411(a)] and therefore does not
violate the Congressional mandate to apply the 111(a)
definition to sources under section 112.
For the HON, the EPA is defining "source" for the SOCMI
source category as the process vents, storage vessels,
transfer racks, wastewater collection and treatment
operations, and equipment leaks in the organic HAP emitting
chemical manufacturing processes that are located in a single
facility covering a contiguous areas under common control.
With this definition of source, all SOCMI portions of plant
sites that are major sources under section 112, approximately
350, are subject to the standard.
A commenter also argued that the EPA's proposed
definition of source was unlawful because it was inconsistent
with language in the Senate Report accompanying S1630, which
2-19
-------
discussed a definition of stationary source that was intended
to prevent "bubbling." However, the language in the Senate
Report referred to a statutory change in the definition of
"stationary source" that was later abandoned by Congress.
Therefore, the Senate Report language referred to by the
commenter is irrelevant.
2.3.2 Averaging at New Sources
Comment: Seven commenters (A-90-19: IV-D-85 and IV-G-6;
IV-D-87; IV-D-90; IV-D-99; IV-D-100; IV-D-115; IV-F-7.6)
recommended that if averaging is allowed, it should be
restricted to existing sources only.
Two commenters (A-90-19: IV-D-94; IV-D-115) stated that
new sources can and should be held to higher standards than
existing sources. Three commenters (A-90-19: IV-D-51;
IV-D-99; IV-F-7.6) maintained that historically, new and
modified sources have been held to a higher standard than
existing sources because, for example, it is most cost-
effective to integrate state-of-the-art controls into
equipment design and to install the technology during
construction. One commenter (A-90-19: IV-D-70) stated that
emissions averaging does not contribute to effective air
pollution control because it could have the effect of allowing
sources to be built or "substantially modified" without
technically practicable and economically reasonable emission
control technology.
Five commenters (A-90-19: IV-D-51; IV-D-85; IV-D-99;
IV-D-115; IV-F-7.6) argued that because new source MACT as
defined in the Act cannot be less stringent than the control
achieved by the best controlled similar source, the Act does
not allow new units to be undercontrolled, and hence,
averaging for new sources is inconsistent with the Act and
inadvisable under any circumstances.
Two commenters (A-90-19: IV-D-90; IV-D-100) opposed new
sources being involved in emissions averaging because it would
relax their State's current requirements and subvert the MACT
requirements that are intended to result in the continuous
reduction of HAP emissions.
2-20
-------
In contrast, one commenter (A-90-19: IV-D-73) suggested
that the economic benefit may be even more pronounced for new
sources because new source MACT may have very low thresholds
of applicability and hence, even wider ranges of cost
effectiveness than the several orders of magnitude range for
existing sources.
Response; The EPA agrees with the conunenters that it is
appropriate that emissions averaging be restricted to existing
sources only. Averaging is a mechanism designed to provide
each source the flexibility to comply with the MACT standard
in a way that is most practical and cost-effective for the
individual source. By employing averaging, a source is able,
for example, to avoid adding controls to an outlying emission
point that would be very expensive to control, or to avoid
replacing expensive control technology that does not achieve
enough emission reduction to meet the standard. These
concerns are applicable to existing sources. A new source can
be designed to avoid expensive outlying emission points, and
retrofitting is obviously not an issue. In addition, when a
new source is constructed, it can be designed to accommodate
the required MACT controls in the most practical and cost-
effective manner, thus reducing the need for the flexibility
of averaging.
The EPA does not agree with the commenters who argue that
prohibiting averaging at new sources would result in a more
stringent standard. The HON has been drafted to provide that
averaging is no less stringent than the standard without
averaging. Thus, allowing new sources to comply only via use
of the reference control technologies and not via averaging
does not require those sources to meet a more stringent
standard. Instead, it requires them to meet a more specific,
and thus more easily implemented standard. However, even if
prohibiting averaging at new sources would result in new
sources being held to a more stringent standard, such a result
would not be unlawful as the statute clearly provides that new
source standards may be more stringent than those for existing
sources.
2-21
-------
2.3.3 Averaging Between New and Existing Sources
Comment: Four commenters (A-90-19: IV-D-51; IV-D-85;
IV-D-94; IV-D-115) objected to allowing averaging between new
and existing sources for the same reasons they opposed
averaging within new sources (see previous comment).
Several commenters (A-90-19: IV-D-32; IV-D-56; IV-D-57;
IV-D-64; IV-D-69; IV-D-72; IV-D-73; IV-D-74; IV-D-75; IV-D-78;
IV-D-79; iv-D-80; IV-D-86; IV-D-92; IV-D-106; IV-G-1)
supported allowing averaging between new and existing sources
within the same plant. Five commenters (A-90-19: IV-D-32;
IV-D-57; IV-D-78; IV-D-79; IV-D-92; IV-G-1) argued that so
long as the plant as a whole achieves the reduction required
by MACT, including any increased level of reduction imposed on
new sources, it will comply fully with section 112(i). One
commenter (A-90-19: IV-D-64) also argued that including
reconstructed sources in averages with new and existing
sources should be allowed.
One commenter (A-90-19: IV-G-1) argued that once new
source MACT is set in accordance with the floor and
section 112(d), nothing in Title III appears to prevent those
reductions from being achieved through an average with points
subject to existing source MACT, and added that because
equivalent reductions would be achieved, the averaging
approach will equally well force technology. The commenter
(A-90-19: IV-G-1) cited as a precedent 40 CFR part 60.47
(1990) (combined average between two existing units and one
new unit to meet site-specific subpart D SO2 NSPS through
innovative coal cleaning technology that avoids scrubbing).
Two commenters (A-90-19: IV-D-32; IV-D-57) strongly
opposed the suggestion in the proposal preamble that new and
existing sources be made separate subcategories of SOCMI,
arguing that Congress clearly intended source categories to be
defined on the basis of types of operations and emissions and
to include both new and existing sources. Two commenters
(A-90-19: IV-D-32; IV-D-57) stated that this is apparent from
the organization of section 112(d), which establishes
requirements to be developed for "categories and
2-22
-------
subcategories" and then specifies special rules for new and
existing sources within these categories. Two commenters
(A-90-19: IV-D-32; IV-D-57)- also referenced House of
Representatives Report (H.-.Rep.) No. 490, P. 1, 101st Cong.,
2d Sess. 328 (1990), as evidence of Congress1 intent.
One commenter (A-90-19: IV-D-32) submitted that concern
over averaging between newatid existing sources could be a
"non-issue" depending on hov the EPA defines "new source."
The commenter (A-90-19: iy-rP-32) suggested that if the term
"new sources" is made equitrsAent to "source" for purposes of
section 112(i), there will^nwver be occasion to average
between a new source and another section 112(i) source because
the two sources are not with*in the same contiguous area and
under common control. The,cammenter (A-90-19: IV-D-32)
further asserted that if new*sources are defined as a source
in a section 112(c) source category, or even as something less
than the section 112(c) source, e.g., as a new process unit in
an existing source, there -is still no reason to disallow
averaging between new and existing sources as long as the
MACT-required reduction is achieved.
Response: The EPA does not consider it appropriate to
allow averaging between new and existing sources. Thus, the
EPA agrees with the conclusion of the commenters who objected
to allowing averaging between new and existing sources and
disagrees with those commenters who supported allowing
averaging between new and existing sources. To allow
averaging between new and existing sources would allow
averaging at separate sources, which the EPA has determined to
be beyond the bounds of permissible averaging under
section 112 of the Act. Whjile new and existing sources are
not separate subcategories*;qr categories of sources, they are
separate sources. There a^se. separate MACT standards with
separate floors for new and* existing sources under the HON.
Allowing averaging between new and existing sources would lead
to the likely consequence rttfiat one source would fail to meet
its applicable standard, a.-consequence that cannot be
reconciled with the statutory requirement that each source
2-23
-------
comply with the applicable standard. Consequently, averaging
between new and existing sources, even if located at the same
plant facility, is not permissible. This is fully consistent
with the EPA's view, explained in section 2.4 of this BID
volume, that averaging between sources is generally not
permissible under section 112 of the Act.
2.3.4 Emission Points Allowed in Trades
Comment; Several commenters (A-90-19: IV-D-9; IV-D-10;
IV-D-11; IV-D-49; IV-D-51; IV-D-70; IV-D-85 and IV-F-7.39 and
IV-F-12 and IV-G-6; IV-D-94; IV-D-99; IV-D-117 and IV-F-7.43;
IV-D-118; IV-D-122; IV-D-123; IV-D-124; IV-D-125; IV-F-1.5;
IV-F-7.6) objected to allowing sources to average across
different kinds of emission points.
One commenter (A-90-19: IV-D-49) was concerned that
allowing averaging across different kinds of emission points
would make it difficult for State and local agencies to
effectively analyze baseline calculations and monitor
emissions.
Six commenters (A-90-19: IV-D-51; IV-D-70; IV-D-90;
IV-D-99; IV-D-100; IV-F-7.6) stated that the impacts of
emissions from different kinds of points can vary
significantly if they have different emission characteristics
that influence dispersion such as elevation, distance from the
property line, volumetric _ow and stack gas temperature, and
the continuous or intermittent nature of emissions. Two
commenters (A-90-19: IV-D-70; IV-D-99) warned that ignoring
the differences in dispersion of pollutants emitted from
various kinds of sources in different parts of a facility
could result in increased adverse impacts on air quality. One
commenter (A-90-19: IV-D-70) gave examples of how dispersion
characteristics can impact air quality.
Response: The EPA agrees with the commenters that the
characteristics of emission points may affect the dispersion
and impacts of emissions. However, for several reasons, the
EPA does not consider these potential differences in
characteristics sufficient reason to limit averaging across
different kinds of points. First, the potential for
2-24
-------
variations in emissions exists if the points are controlled by
RCT with no averaging. The mix of controlled and uncontrolled
emission points under point-by-point compliance can also be
skewed towards one location on a plant site or one kind of
emission point.
Second, even among one kind of emission point (e.g.,
process vents), there is variation in height, exit velocity,
distance to fenceline, mix of HAP's, and other characteristics
that influence the environmental impact of the emissions.
These variations in characteristics would still exist
regardless of whether the rule allows emissions averaging
across different kinds of points, and differing impacts due to
the variation would still be possible.
Third, it is equally likely that emissions averaging
could result in decreased impacts if points closer to the
fenceline are controlled to a greater extent than required
under RCT. Finally, it is reemphasized that emissions
averaging will probably be used with only a few points in each
facility. As a result, averaging will probably only influence
a small proportion of the total emissions from a source.
In regards to the comment that it will be difficult for
State and local agencies to analyze baseline emissions and
monitor emissions, it is not necessary for implementing
agencies to perform these tasks. For compliance purposes,
debits and credits are based on the actual operation of the
emission points during each quarterly compliance period, and
must be calculated using actual operating data and consistent
estimation techniques. The only baseline decision that must
be made concerns when controls were applied. Controls applied
prior to November 15, 1990 are considered part of the source's
baseline control and cannot be used to generate credits.
Comment: Six commenters (A-90-19: IV-D-9; IV-D-70;
IV-D-85 and iv-F-7.39 and IV-F-12; IV-D-118; IV-D-124;
IV-D-125) claimed that allowing averaging across different
emission points and different types of process units
facilitates "game-playing." Five commenters (A-90-19:
IV-D-9; IV-D-85 and IV-F-7.39 and IV-F-12; IV-D-118; IV-D-124;
2-25
-------
IV-D-125) were concerned that sources can use inconsistent
emissions estimation techniques for credits and debits when
averaging across different types of sources and emission
points. One commenter (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12) specifically claimed that trades between different
process units will allow plant operators to claim credits that
reflect differences in production rate rather than added
controls.
Four commenters (A-90-19: IV-D-49; IV-D-51; IV-D-85 and
IV-F-7.39 and IV-F-12 and IV-G-6; IV-F-7.6) recommended that
trades, if allowed, should be limited to the same kinds of
emission points within a process unit. One commenter
(A-90-19: IV-D-85 and IV-F-7.39 and IV-F-12) added that this
would provide a check on the use of inconsistent emissions
estimation methodologies, especially if the EPA also required
application of identical replicable emissions calculation
methods, identical assumptions for credits and debits, and
stringent emissions monitoring.
Response; The EPA acknowledges that there is potential
for significant complexity in the emissions averaging policy
because it allows averaging across different kinds of points
and different process units. However, the EPA has decided to
maintain this scope for the final rule with the rationale that
the averaging program has sufficient structure to prevent
inconsistencies or inappropriate compliance scenarios from
arising. Specifically, the emissions averaging program makes
use of: (1) consistent emission estimation techniques; and
(2) actual operating data to calculate both debits and
credits.
Consistent emission estimation techniques are
incorporated into the rule. Many of these techniques are used
in other standards and have been found to be verifiable and
enforceable or have been updated for this rule. Only one
method for estimating debits and credits is allowed for each
kind of point, and these procedures are thoroughly prescribed
in the emissions averaging provisions. The EPA is confident
that use of these consistent estimation techniques for
2-26
-------
different kinds of points will check the potential for "game-
playing . "
In select cases, sources have some latitude in
determining certain parameters. For instance, the vent stream
flow rate needed to calculate process vent emissions can be
measured using .one of a number of similar methods. Also,
historical records or process knowledge may be substituted for
the determination of values for representative operating
parameters to establish average wastewater stream flow rates.
Use of one method versus another could be construed as using
different assumptions; however, the different methods have all
been determined to be interchangeable. As a result, the EPA
maintains that there is no opportunity for the use of
inconsistent estimation methodologies in this rule. Hence,
the EPA considers emissions monitoring as a check on the use
of inconsistent estimation techniques to be unnecessary. The
EPA has included provisions for appropriate monitoring in the
rule. These provisions are addressed in greater detail in
section 2.8.2 of this BID volume.
The use of actual operating data to calculate both debits
and credits will also limit "game-playing." Credits can be
derived only from a demonstrable reduction of emissions
achieved by either an approved control technology or pollution
prevention measure that performs better than what is required
under point-by-point compliance. Sources cannot claim credits
that reflect differences in production rates between
processes.
It is true that if a credit-generating point and a debit-
generating point both generate an equal amount of credits and
debits per unit of operation, the source can generate more
credits than debits by operating the credit generator more
than the debit generator. However, this cannot be construed
as the source generating credits because of a difference in
production rates. Rather, credits result from controlling the
credit generator to a level more stringent than what the rule
requires. Thus, for a given level of operation, the point is
emitting less than what it is allowed. As production
2-27
-------
increases, the difference between the allowed and actual
emissions for that emission point increases, and the credits
attributable to that point also increase.
Comment: Two commenters (A-90-19: IV-D-90; IV-D-100)
stated that any legal authority the EPA has in allowing
emissions averaging across all emission types must be based on
the demonstration of equivalency between emission types, the
ability to ensure compliance with permit conditions, and the
potential toxicity of HAP's emitted from these emission
points. Another commenter (A-90-19: IV-D-70) was concerned
with averaging across different process units because they
generally have different characteristic HAP's which have
varying levels of toxicity.
Response; The EPA maintains that it has ample legal
authority to allow averaging among different kinds of emission
points because the source (which is defined as the collection
of emission points) is required to reduce emissions to the
maximum level achievable.
The EPA holds that its legal authority is in no way
defined or constrained by the conditions the commenters
suggest. First, it is not clear what the commenters meant by
a demonstration of equivalency between emission types, but the
emission reductions from two points are generally considered
equivalent if the total mass quantities of reductions are
equal. Second, compliance with permit conditions is ensured
in the rule; comments regarding enforcement are further
addressed in section 2.8 of this BID volume. Finally, the
issue of toxicity in emissions averaging is addressed in
detail in section 2.9 of this BID volume.
Comment: Seven commenters (A-90-19: IV-D-9; IV-D-118;
IV-D-124; IV-D-125; IV-D-45 and IV-F-7.7; IV-D-70; IV-D-85 and
IV-F-7.39 and IV-F-12 and IV-G-6) specifically objected to
sources being allowed to include wastewater emissions in an
averaging scheme because they considered accurate or reliable
estimation of wastewater emissions to be unlikely or
impractical.
2-28
-------
One commenter (IV-D-45 and IV-F-7.7) -was concerned that
underestimates of emissions and inclusions in averages could
lead to undercontrol of emissions from wastewater plants.
Another commenter (A-90-19: IV-D-85) stated that the wide
variability over time of wastewater characteristics that
affect emissions, such as mass concentrations, wind speed,
oxygen content, surface configurations, temperature, flow
rate, etc., make the estimation of wastewater emissions
extremely uncertain and the inclusion of wastewater in
emissions averaging especially irresponsible. One commenter
(A-90-19: IV-D-70) was specifically concerned about the
uncertainties involved in calculating "fractions removed" by
steam stripping the various VOHAP's.
One commenter (A-90-19: IV-D-85) claimed that the EPA
has recognized that a numerical standard would not be feasible
for wastewater emissions. The commenter (A-90-19: IV-D-85)
argued that on the other hand, to include wastewater emissions
in averaging would require assigning a numerical emission
target to "overcontrolled" wastewater streams. The commenter
(A-90-19: IV-D-85) concluded that the same reasoning that
supported not setting numerical limitations for wastewater
requires its deletion from emissions averaging.
One commenter (A-90-19: IV-D-70) stated that if it is
not possible to exclude wastewater from emissions averaging,
then more rigorous monitoring and testing of wastewater
concentration and flow rate should be required to quantify the
emissions, along with a very conservative discount factor.
Response; The EPA considers the estimation of wastewater
emissions on an annual basis to be as reliable as for the
other kinds of points and hence, suitable for inclusion in
emissions averaging.
The EPA has recognized that the wastewater
characteristics cited by one commenter could make emissions
from areas such as surface impoundments changeable and
difficult to measure. Therefore, debits and credits for
wastewater streams, as well as HON applicability to wastewater
streams and Group status of streams, are determined at the
2-29
-------
stream point of generation. Also, if a wastewater stream is
being controlled as a credit generator, the stream must comply
with the standards for transport and handling equipment, which
require suppression to eliminate the influence of factors such
as wind speed, oxygen content, and surface configurations.
This ensures that the only emissions that need to be
considered are those from the control device.
As in the case of other emission points, characteristics
such as HAP concentration, temperature, and flow rate remain
relatively constant in wastewater streams so that
representative values can be used. The rule provides that if
operating conditions change such that previously measured
values are no longer representative, the values must be
redetermined.
The final rule now specifies that wastewater streams
treated in biological treatment units are not eligible for
emissions averaging. All other types of control are
acceptable as long as their reduction efficiency can be
determined. The EPA is confident that by making biological
treatment of wastewater ineligible for averaging, the
potential for underestimation of wastewater emissions will be
minimized.
It was not clear whether one commenter was questioning
the accuracy of the Fr's included in table 9 of subpart G of
the final rule, or whether the commenter was concerned about
the uncertainty in calculating Fr's for a steam stripper that
is not operated to the efficiency specified for the RCT. The
EPA is confident of the Fr's included in table 9, which are to
be used when the RCT is employed and which were updated for
the final rule to reflect new information improving their
accuracy. The discussion of how the factors were determined
and updated for the final rule can be found in section 5.1 of
BID volume 2B.
The EPA is also satisfied that sources can determine Fr's
accurately for a steam stripper that is not being operated to
the efficiency required for the RCT. A steam stripper that is
not the design steam stripper can be used to comply with the
2-30
-------
rule without averaging, and the rule specifies the procedures
and test methods to be used to demonstrate that the steam
stripper can achieve the required HAP removal efficiency.
These same sampling and analytical methods that are used to
demonstrate compliance are also appropriate for determining
the treatment efficiency of a steam stripper on a debit-
generating wastewater stream.
One commenter was mistaken regarding whether the EPA
considered a numerical standard feasible for wastewater
emissions. The proposal preamble stated that a numerical
standard would not be feasible for the provisions for
wastewater transport and handling equipment. On the other
hand, the provisions for reduction of VOHAP concentration in
the wastewater streams are in a numerical emission limit
format, specifically a percent emission reduction. However,
the wastewater RCT cannot be assigned a single reduction
efficiency because the different constituents in wastewater
streams will have different volatilities and strippabilities.
Nonetheless, the wastewater provisions for achieving
reductions are in a numerical format, which means that a
source has a "target" for overcontrolling wastewater streams
to generate emission credits.
Thus, the EPA considers the provisions for characterizing
and monitoring wastewater emissions suitable for emissions
averaging as well as for point-by-point compliance.
Similarly, the EPA considers a very conservative discount
factor for credits generated from wastewater unnecessary
because the estimation of wastewater emissions is as reliable
as for the other kinds of emission points. In summary, many
of the concerns for including wastewater emissions in
emissions averaging stem from misunderstandings about the
nature of its control in the rule, which is discussed in
greater detail throughout BID volume 2B.
Comment; Several commenters (A-90-19: IV-D-32; IV-D-50;
IV-D-56; IV-D-58; IV-D-62; IV-D-73; IV-D-75; IV-D-79; IV-D-86;
IV-D-89; IV-D-92; IV-D-113) urged the EPA to allow sources to
include equipment leaks and fugitive emissions in emissions
2-31
-------
averaging. One commenter (A-90-19: IV-D-89) complained that
not allowing credit for controlling fugitive emissions
discourages voluntary pollution prevention measures. One
commenter (A-90-19: IV-D-92) stated that fugitive emission
rates are required to be calculated for emissions in
nonattainment areas, permitting, and health effect reviews, so
facilities should be allowed to use fugitive emission
reductions for credits in emissions averaging. Another
conunenter (A-90-19: IV-D-75) added that quantification of
fugitive emissions is required to be submitted annually with
SARA title III section 313 reports. The commenter (A-90-19:
IV-D-75) stated that extensive work by industry and the EPA
has been conducted on quantifying emissions from fugitive
leaks and contended that there is sufficient data to develop a
protocol and calculation methodology to adequately estimate
emissions.
Four commenters (A-90-19: IV-D-32; IV-D-62; IV-D-75;
IV-D-113) suggested that it is already possible to quantify
emissions from equipment leaks sufficiently. Two commenters
(A-90-19: IV-D-32; IV-D-73) contended that the equipment
leaks provisions- address many emission points, such as
sampling systems, compression seal vents, closed-vent systems,
and product accumulator vessels, for which emissions can be
quantified through methods similar to those adopted for
emission points addressed in subpart G, and hence, these
points should be eligible for emissions averaging. Another
commenter (A-90-19: IV-D-58) added that the proposed
provisions for controlling equipment leaks enable the
inclusion of equipment leaks in emissions averaging. The
commenter (A-90-19: IV-D-58) elaborated that a facility would
be required to specify how it will achieve compliance to
further reduce emissions. The commenter (A-90-19: IV-D-58)
suggested that some emission points could be designated with a
lower leak rate definition than in the negotiated rule, or
with a lower percent leak rate to earn credit. The conunenter
(A-90-19: IV-D-58) advanced another possibility that other
2-32
-------
process streams currently excluded from the rule could be
added.
Three commenters (A-90-19: IV-D-32; IV-D-62; IV-D-113)
recommended that fugitive emissions be quantified according to
the methods published in the EPA's document, "Protocols for
Generating Unit Specific Emissions Estimates of Equipment
Leaks of VOC and VHAP." (1988) The commenters (A-90-19:
IV-D-32; IV-D-62; IV-D-113) suggested that this document
contains procedures that can be used to establish baseline
emissions resulting from the HON standards and "screening
value correlations" to calculate actual emissions and
potential credits.
One commenter (A-90-19: IV-D-73) suggested that the EPA
review the adequacy of an estimating procedure and statistical
data base being compiled by the CMA (the POSSEE data base) to
support quantification of equipment leak mass emission rate
estimates.
In contrast, two commenters (A-90-19: IV-D-51; IV-D-99)
agreed with EPA's decision not to include equipment leaks in
emissions averaging. Two commenters (A-90-19: IV-D-99;
IV-D-115) opposed equipment leak emissions being included in
future averaging rules because quantification o'f equipment
leak emissions will be problematic.
Response: The EPA acknowledges that methods are
available for quantifying emissions from equipment leaks;
however, this is not at issue in emissions averaging. As
stated in the proposal preamble, equipment leaks cannot be
included in emissions averages for two reasons. First, a
reference control efficiency cannot be established for the
negotiated standard for equipment leaks because the percent
reduction achieved by complying with subpart H of the rule
will vary depending.on the characteristics of the process and
the equipment being controlled. Second, no method currently
exists for determining allowable emissions for leaks, i.e.,
residual emissions from equipment controlled according to
subpart H. Without a reference control efficiency or the
2-33
-------
ability to assign allowable emissions, debits and credits
cannot be established for any kind of point.
Some commenters suggested methods for generating credits
from eguipment leaks. One proposal was to designate a lower
leak rate definition or a lower percent leak rate than in the
negotiated rule for some emission points. Such a policy could
allow a source to overcontrol equipment leaks, but it still
does not enable a source to estimate allowable emissions so
that debits and credits can be calculated.
The other suggestion was to use residual emissions after
complying with subpart H of the rule as allowable emissior
and "screening value correlations" to establish the actua-
emissions. However, credit and debit calculations must be
based on allowable and actual emissions from the same time
period. It is not acceptable to base averages on allowable
emissions from one time period and actual emissions from a
different period because the allowable emissions must be
calculated using the same operating rate data as the actual
emissions. Until suitable methods are developed to assign
reference control efficiencies and allowable emission for
particular leak points, eguipment leaks cannot be allowed in
emissions averages.
Comment: One commenter (A-90-19: IV-D-50) suggested
that a 100 percent credit should be allowed if facility-
specific emission factors are developed by bagging, and a
partial credit should be allowed if EPA-developed factors are
used. The commenter (A-90-19: IV-D-50) explained that
generally these factors have been found to be high, therefore,
applying a 50 percent credit for such emissions would provide
an ample margin of safety to assure an overall reduction.
Response; The commenter did not define their use of the
term "bagging." It is assumed that the reference is to the
technique of measuring emissions by enclosing an emission
point or area completely, allowing only one outlet for
sampling. This technique is used primarily for estimating
fugitive emissions, so it is assumed that the commenter is
2-34
-------
referring to a method for including equipment leaks in
emissions averages.
As stated previously, the issue is that neither a
reference efficiency nor allowable emissions can be
established for equipment leaks, not whether equipment leaks
can be quantified. Because a suitable method has not been
identified, equipment leaks cannot be included in emissions
averaging at this time.
2.4 COMPLEMENTARY LEGAL INTERPRETATION FOR BROADER EMISSIONS
AVERAGING
2.4.1 Legality of Broader Averaging
Comment: Seven commenters (A-90-19: IV-D-32; IV-D-58;
IV-D^-62; IV-D-74; IV-D-83 and IV-F-1.3 and IV-F-5; IV-D-86;
IV-D-108) maintained that ample legal authority exists to
support adopting a broad emissions averaging scheme. One
commenter (A-90-19: IV-D-83) stated that in the proposal the
EPA used a broad definition of "source," which includes both
SOCMI and other processes at a plant site, to determine
whether it is a major source and therefore subject to MACT
standards. The commenter (A-90-19: IV-D-83) contended that
the use of this broad source definition for determining
applicability of the HON provides the legal basis for allowing
emissions averaging within the entire plant site.
Seven commenters (A-90-19: IV-D-32; IV-D-62; IV-D-74;
IV-D-82; IV-D-98; IV-D-108; IV-G-1) agreed that the EPA has
broad discretion to define "source." Four commenters
(A-90-19: IV-D-32; IV-D-62; IV-D-82; IV-G-1) asserted that
the EPA's discretion to define "source" according to the
context has been firmly established in Chevron. One commenter
(A-90-19: IV-D-32) found further authority in Alabama Power
Co. v. Costle. 635 F.2d 323 (D.C. Cir., 1979). One commenter
(A-90-19: IV-G-1) cited the can coaters1 "bubble" authorizing
plant-wide RACT compliance (45 FR 80824, December-8, 1980) as
a precedent for the plant-wide average compliance
interpretation.
One commenter (A-90-19: IV-D-62) concluded that the
EPA's authority to define "source" stems from Congress not
2-35
-------
having clearly stated such a definition in the Act.
Furthermore the commenter (A-90-19: IV-D-62) declared that
the EPA's definition is based on a permissible construction of
the statute.
Seven commenters (A-90-19: IV-D-32; IV-D-73; IV-D-74;
IV-D-86; IV-D-98; IV-D-108; IV-G-1) considered the broader
averaging scheme to be consistent with sections 112(d) and (i)
of the Act. Three commenters (A-90-19: IV-D-32; IV-D-74;
IV-D-108) claimed that for the purposes of section 112(d), the
EPA may define "source" as only SOCMI operations; the "source"
under section 112(i) may then be defined as any entire
facility within a contiguous area and under common control to
which MACT standards are applicable. Another commenter
(A-90-19: IV-D-62) noted that the EPA has already adopted a
plant-wide definition of "source" in its Early Reductions rule
and its 1986 Emissions Trading Policy Statement. Three
commenters (A-90-19: IV-D-32; IV-D-74; IV-D-108) concluded
that with this dual definition of source, nothing should bar
emissions averaging across emission points that are within the
same section 112(i) source but in different section 112(d)
source categories so long as the section 112(i) source
achieves the reduction required by all applicable MACT
standards.
One commenter (A-90-19: IV-D-82) claimed that nothing in
the Act forbids a plant-wide approach to MACT compliance, and
several elements of the statute support it. The commenter
(A-90-19: IV-D-82) submitted that in fact, Congress deleted
language in the Senate bill that would have expressly
forbidden a "bubble" approach to MACT compliance. The
commenter (A-90-19: IV-D-82) also noted that under
section 112(g), a plant can "net out" of premature MACT by
making source-wide reductions, and because Congress expressly
allowed a plant-wide approach to postponing MACT, this should
also be an allowable approach to compliance with MACT after
MACT becomes applicable. The commenter (A-90-19: IV-D-82)
also cited a case, NRDC v. Thomas (1986), as supportive of
broader averaging in which the EPA allowed averaging across
2-36
-------
"engine families" to comply with Title II of the 1990
amendments to the Act.
One commenter (A-90-19: IV-D-51) did not agree with the
interpretation of the significance of the usage of "source"
between 112(d) and 112(i), and stated that section 112(a)
defines the word "source" as it is meant to apply in
section 112, and no language in subsequent paragraphs of
section 112 modifies the definition of "source." The
commenter (A-90-19: IV-D-51) added that the Chevron decision
does not give the EPA the authority to tamper with a basic
definition unless certain conditions are met.
One commenter (A-90-19: IV-D-85) stated that the
complementary legal interpretation as articulated in the
proposal preamble conflicts with explicit statutory language
fAccord Memorandum from Alan Eckert, EPA General Counsel, to
David Rivkin, President's Council on Competitiveness
(October 9, 1992) A-90-19, II-F-16]. The commenter maintained
in an extensive discussion that if the EPA wishes to redefine
"stationary source" to include the entire plant, it must
require MACT controls on the entire plant; otherwise, it will
violate sections 112(d) and (i) of the Act, which require
maximum achievable emissions reductions from the stationary
source.
The commenter stated that the complementary
interpretation posits that the EPA may define stationary
source differently for purposes of section 112(d) and
section 112(i) and that doing so allows achievement of the
result sought. The commenter (A-90-19: IV-D-85) maintained
that both suggestions are incorrect.
The commenter contended that the language in
section 112(i) does not allow a plant-wide stationary source
definition, and the language of section 112(d) by itself
precludes acceptance of the complementary legal approach. The
commenter stated that section 112(d)(2) speaks of standards
"applicable to new or existing sources," which means that
Congress intended the standards to apply to specific
stationary sources, not to subcategories of stationary sources
2-37
-------
and not to "major sources" including several subcategories of
sources.
The commenter contended that section 112(d)(2) further
states that MACT standards must "require" maximum achievable
emissions reductions for the "new or existing sources in the
category or subcategory to which the emission standard'
applies." The commenter maintained that the alternative
interpretation could require no emissions reductions from the
stationary source to which the standard applies, if all the
emission reductions came flrom elsewhere in the plant. The
commenter stated that any deviation from maximum achievable
emissions reductions from the stationary source would violate
section 112(d). The commenter stated that, on the other hand,
if the stationary source is defined to encompass the entire
plant, then the entire plant must achieve maximum achievable
emissions reductions, not just part of it.
The commenter contended that section 112(i)(3)(A)
requires the "source" to comply with the MACT standard. The
commenter stated that the complementary interpretation
suggests that Congress intended to use the word "source" in
this context to mean "major source." The commenter contended
that it is illogical to think that the schedule for compliance
created under section 112(i) applies to a different entity
than the emission standard created under section 112(d), and
section 112(d) speaks of standards "applicable to new or
existing sources" just as section 112(i)(3) speaks of
standards "applicable to a source." The commenter maintained
that Congress intended that: (1) MACT standards must require
maximum achievable reductions from each regulated stationary
source; and (2) each stationary source must comply with the
standard. The commenter maintained that a MACT level not
requiring maximum reductions from the entire stationary source
violates the law.
The commenter maintained that from a policy standpoint,
the alternative interpretation poses enormous problems. The
commenter stated that under that interpretation, States would
2-38
-------
be burdened with verifying estimation of emission credits from
all kinds of different processes and emission points.
The commenter stated that the preamble of the proposal
claims that the alternative definition would generate improved
controls at parts of the plant not addressed by the HON, thus
raising the level of control determining the floor levels for
future MACT standards. The commenter contended that the
claimed advantage of raising the floor will not materialize;
plant operators will simply claim credit for reductions they
are already making to meet other State or Federal standards or
to mollify angry citizens.
The commenter stated that the EPA has not used actual
emissions data in generating floors for the SOCMI. The
commenter maintained that improvements in the actual floor
level of control, even if they occurred, would not affect EPA
decision-making unless the EPA requires reports of: (1) the
emissions achieved at all non-HON points used in the average;
(2) an identification of the process unit of which the
controlled point is a part; and (3) a statement as to which
source category the point belongs to. The commenter stated
that the EPA would have to then enter this information in its
data bases for future rulemakings. The commenter concluded
that the alternative legal interpretation would simply lead to
less control from the facility as a whole.
On the other hand, one commenter (A-90-19: IV-D-32)
disagreed with the argument that averaging between source
categories could violate MACT floor requirements. The
commenter argued that MACT standards must achieve reductions
no less stringent than the floor, but section 112(i) requires
sources to comply with MACT, not with the MACT floor, and the
Act says nothing about how sources must go about complying
with a MACT standard once it is established. The commenter
stated that where a facility is subject to two or more MACT
standards, the overall degree of reduction that it must
achieve will be the sum of the reductions required under those
standards. The commenter contended that as long as the
facility achieves an overall level of reduction or the
2-39
-------
aggregation of emission points that are subject to the various
MACT standards, it has satisfied MACT. The commenter
suggested that at a minimum, the EPA should allow emissions
averaging to include all emission points that are within
source categories that are subject to MACT.
Response: After studying the arguments presented by the
commenters both for and against a broader averaging approach,
the EPA has decided to retain the narrower approach contained
in the proposed rule.
The EPA agrees with the commenters who argued that the
statute provides broad discretion to define "source," and does
not prohibit averaging in setting standards under
section 112(d) of the Act. However, the EPA has determined
that section 112 does provide some limits on the scope of
averaging, and that the broader averaging approach discussed
in the proposal preamble exceeds those limits.
As several commenters pointed out, the statute requires
the EPA to consider emissions from the entire facility in
order to determine whether it is a major source subject to a
given MACT standard. However, the EPA is also required to
develop a list of source categories, which are to be composed
of "sources" that are then subject to regulation under MACT
standards. Both the language of section 112(d) and the
legislative history indicate that sources in the category can
be coextensive with a major source, but are just as likely to
be merely a portion of a facility. Thus, a large facility
emitting more than 25 tons of multiple HAP's will, in most
cases, be composed of multiple sources in different source
categories subject to standards on different dates. It does
not follow that, because applicability under section 112
(i.e., whether a facility emits sufficient HAP's to be
considered a major source) is determined on a facility-wide
basis, compliance with specific standards written for sources
that comprise only a part of a facility should be permitted on
a facility-wide basis. The most that can be inferred is that
the entire facility is the largest entity that can be defined
as a source within any category, but that the source in a
2-40
-------
category can, and often will be, smaller than the entire
facility.
In accordance with section 112 (i) of the Act, all sources
in the category for which a standard is in effect must be in
compliance by a specified date. Commenters' arguments that
section 112(i) allows compliance with a standard that is set
for a source category to be achieved by a "source" that is
more extensive than the source in the category (i.e., the
entire major source that the source in the category is a
fraction of), is inconsistent with the specific language of
section 112(i). Section 112(i) provides different compliance
requirements for new and existing sources. New sources must
comply with an applicable standard earlier than existing
sources, which can be given up to three years to comply.
Moreover, section 112(i)(3) provides for compliance dates to
be established for "each category or subcategory of existing
sources." This provision clearly applies to compliance by
sources in a category rather than compliance with a standard
by any points within an entire major source. Therefore,
section 112(i) clearly provides for compliance by individual
sources within the relevant category rather than overall
compliance by a major source with a standard applicable to
only part of the major source.
Thus, the EPA is adopting the more limited approach to
averaging that was contained in the proposed rule. All
sources within a given source category must comply
individually with the standard either by application of the
reference control technology or by compliance with an approved
emissions average. Transferring emission reduction
obligations to points outside of the source within the
category would be inconsistent with the requirement of
section 112(d) of the Act that standards be set for sources in
a listed category, and the requirements of section 112(i) that
compliance with such standard be achieved by sources in the
category.
2-41
-------
2.4.2 Policy and Practical Considerations
Comment: Several commenters (A-90-19: IV-D-32; IV-D-33;
IV-D-43; 'IV-D-50; IV-D-56; IV-D-58; IV-D-62; IV-D-64; IV-D-69;
IV-D-72; IV-D-73; IV-D-74; IV-D-75; IV-D-79; IV-D-82; IV-D-83
and IV-F-1.3 and IV-F-5; IV-D-86; IV-D-89; IV-D-92; IV-D-98;
IV-D-106; IV-D-108; IV-D-113; IV-F-1.6 and IV-F-6; IV-G-1)
supported allowing a broader emissions averaging that includes
emission points located anywhere within a facility and not
subject to the HON. Some of the reasons listed by commenters
include: (1) it would enable sources to achieve the required
emission reductions earlier, with greater flexibility, or more
cost-effectively; (2) it would encourage the development of
alternative innovative control methods for the HON or for
emission points not covered under the HON, which could be used
to establish and potentially tighten the floors for future
MACT standards; (3) it could reveal emission points that may
have been otherwise overlooked by regulators, which could lead
to more accurate emission characterization in the future; and
(4) not allowing emissions averaging across source categories
would completely eliminate the incentives for emissions
averaging.
Four commenters (A-90-19: IV-D-58; IV-D-62; IV-D-113;
IV-F-1.6 and IV-F-6) cautioned against restricting emissions
averaging to only facility operations within the same 2-digit
SIC code.
One commenter (A-90-19: IV-D-92) recommended allowing
averaging among sources not under common ownership or control
as is currently allowed under the nonattainment area policy of
the Act.
Two commenters (A-90-19: IV-D-92; IV-D-98) supported
broader averaging that would allow HAP's not regulated by the
HON to be averaged as they become covered by subsequent MACT
standards. Another commenter (A-90-19: IV-G-1) stated that
plant-wide averaging should be limited to organic HAP's, but
not only organic HAP's covered by a MACT standard. The
commenter (A-90-19: IV-G-1) argued that equivalent reductions
of any organic HAP's within the fenceline should be allowed
2-42
-------
and cited the proposed Economic Incentive Program Rules as a
precedent.
One commenter (A-90-19: IV-D-58) was concerned that as a
result of the EPA definition of a "major source" under
section 112(a) to include all source categories at the same
location, all emission points at that location, even small
ones, would be required to install relevant MACT requirements
regardless of their emission rate because these points could
not be included in the emissions averaging program for the
predominant source category, unless broader averaging is
allowed.
One commenter (A-90-19: IV-D-32) recommended that if
broader averaging is allowed, the EPA should not specify
detailed requirements for the baseline level of control to be
used to determine credits from non-SOCMI emission points and
that the enactment date of the 1990 amendments to the Act is
an appropriate baseline date. Another commenter (A-90-19:
IV-D-58) suggested that provisions similar to those of the
Early Reductions Program promulgated under section 112(i)(5)
could be used to establish the baseline and enforce averages.
Another commenter (A-90-19: IV-D-89) suggested that the
data needed to establish a baseline can be identified in the
Implementation Plan, and non-SOCMI emission points could be
made mutually subject to the HON.
Two commenters (A-90-19: IV-D-74; IV-D-108) argued that
once an operating permit program is in place, the operating
permit must identify the applicable requirements for each
emission point, and credit could be allowed for the difference
between the otherwise applicable rules (e.g., RACT) and more
efficient controls. The commenters (A-90-19: IV-D-74;
IV-D-108) concluded that enforceability of broader averaging
could be ensured by the annual and quarterly calculations of
the average as well as by enforcement of the operating permit.
One commenter (A-90-19: IV-D-79) recommended that the
EPA develop specific procedures for establishing compliance
for non-SOCMI emission points included in averages that will
not conflict with a source's HON compliance, and that possibly
2-43
-------
group applications for approval of compliance procedures
through trade organizations or ad hoc groups could be allowed.
In contrast, eight commenters (A-90-19: IV-D-49;
IV-D-51; IV-D-85; IV-D-87; IV-D-90; IV-D-100; IV-D-115;
IV-F-7.6) opposed allowing broader averaging for reasons
including: (1) averaging between source categories, organic
and inorganic HAP's, or new and existing sources is
unacceptable; (2) broader averaging would significantly reduce
the impact of the HON on the originally intended emission
points and result in fewer reductions within the source
category; (3) it has no scientific or regulatory basis and
presents administratively burdensome requirements; and
(4) because non-SOCMI sources may not be covered by a MACT
standard, it could not be guaranteed that the standards for
non-SOCMI sources will be consistent with SOCMI standards and
that broader averaging would permit averaging in the same way.
Response: As discussed in the previous section,
emissions averaging can be permitted only among emission
points that are within the SOCMI source category. Hence,
although the comments submitted on the policy and practical
aspects of this issue are not without merit, due to the
finding that broader averaging cannot be allowed, these
comments are no longer applicable.
2.5 CREDITS
2.5.1 General Issues
Comment; Two commenters (A-90-19: IV-D-72; IV-D-106)
argued that credit should be allowed for all measures that
result in quantified emissions reductions beyond that required
by the standard. Another commenter (A-90-19: IV-D-59) argued
that credit should be allowed for the actual obtained
efficiency, not just for controls over the RCT's stated
efficiency because the latter penalizes industry for
installing the more efficient emission control device.
Response; Credit is allowed for all measures that result
in control levels more stringent than what the rule requires
for the relevant emission point or points. If a Group 2 point
is controlled, credit is allowed for the actual obtained
2-44
-------
reduction, not just for reductions over the RCT's nominal
efficiency. However, to generate credits from Group 1 points,
the emission reductions must be greater than what can be
achieved using RCT. Otherwise, emissions averaging would not
achieve the same reductions or represent an equivalent
alternative to point-by-point compliance. Allowing credit for
the difference between the actual obtained efficiency and the
RCT's nominal efficiency does not penalize a source for
installing the more efficient emission control device.
Instead, it gives a source the incentive to develop controls
that are more efficient than RCT, an incentive that would not
be present without emissions averaging.
Comment; One commenter (A-90-19: IV-G-17) foresaw only
a limited number of situations where cost-effective credits
might be available, which would force only the most difficult-
to-control points to be averaged. Hence, the commenter
(A-90-19: IV-G-17) anticipated that any averaging would
involve only a limited number of emission points. Moreover,
the commenter (A-90-19: IV-G-17) reported that since only
very special circumstances make an emission point worth
considering for averaging, there does not appear to be a
pattern of the particular kind of emission point that would be
a candidate for averaging, nor is a pattern expected.
Two commenters (A-90-19: IV-D-63; IV-D-71) argued that
the emissions averaging proposal is too constrained to be of
much use. One commenter (A-90-19: IV-D-63) concluded that
the stringency of the proposed program will limit its
applicability, and most sources will not find many
opportunities to generate credits because, for example,
controlling Group 2 points will generate only a small number
of credits.
Response; The EPA acknowledges that there are numerous
specific elements of emissions averaging that can be
considered constraints. This specificity is intended to
ensure that emissions averaging results in emission reductions
equivalent to point-by-point compliance and that there are
adequate records and reports to ensure enforceability. The
2-45
-------
EPA expects that emissions averaging will still allow sources
to avoid situations where point-by-point compliance would be
unusually expensive. The EPA anticipates that for the
majority of sources, the most cost-effective way to obtain the
required emissions reduction is to control the largest
emitters, i.e., the Group 1 points.
In addition, it should be noted that a source is not
limited to generating credits only from Group 2 points. Some
facilities will find instances where it is more cost effective
to overcontrol certain Group 1 points, perhaps with pollution
prevention, to generate credits.
2.5.2 Use of RCT Above Rated Efficiencies
Comment; Four commenters (A-90-19: IV-D-85 and
IV-F-7.39 and IV-F-12 and IV-G-6; IV-D-99; IV-D-115; IV-F-1.5)
opposed allowing credit for overcontrol, i.e., control to a
higher efficiency than the RCT" s rated efficiency.
Two commenters (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12; IV-F-1.5) argued that allowing extra credits for
reductions that go beyond a benchline standard is inconsistent
with the MACT concept. One commenter (A-90-19: IV-D-99)
stated that not allowing credits for overcontrol beyond the
RCT's rated efficiency would simplify the averaging process by
applying uniform credit for specific control measures.
One commenter (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12) noted that as stated in the draft rules sent to OMB
in December 1991, allowing credit for overcontrol is
inappropriate because the rated efficiency is intentionally
conservative, and understates the reductions the RCT will
actually achieve if properly operated. The commenter
(A-90-19: IV-D-85) argued moreover that to the extent the
reference control efficiency understates the actual emissions,
emissions averages will attain less than the maximum
achievable emission reduction, because of the gap between what
would have been achieved in reality without averaging and the
efficiency used to calculate a debit. Two commenters
(A-90-19: IV-D-85; IV-D-115) observed that the proposal does
not provide for debits when use of an RCT results in
2-46
-------
undercontrol, nor does it require CEM's in order to detect
*
undercontrol, which should be counted into an average. Thus,
one commenter (A-90-19: IV-D-85 and IV-F-7.39 and IV-F-12)
predicted that allowing credit for overcontrol will create
"paper credits," which will be used to balance real emission
increases.
Nine commenters (A-90-19: IV-D-33; IV-D-56; IV-D-72;
IV-D-74; IV-D-75; IV-D-92; IV-D-98; IV-D-106; IV-D-108)
supported allowing credit for reductions from the use of RCT
at a more efficient level than the efficiency rating
established for that RCT.
Two commenters (A-90-19: IV-D-74; IV-D-108) argued that
credit for many potentially significant emissions reductions
would be lost if credit is not given for reductions from the
use of RCT at a more efficient level than the rated
efficiency. Three commenters (A-90-19: IV-D-74; IV-D-75;
IV-D-108) warned that sources would have no incentive to
achieve greater efficiencies. Two commenters (A-90-19:
IV-D-72; IV-D-106) stated that credits for higher efficiency
operation will allow cost-effective compliance and will
encourage further development of existing control
technologies.
Two commenters (A-90-19: IV-D-74; IV-D-108) recommended
that credit be allowed for higher efficiencies based on the
amount of emission reduction which is measurable and
demonstrable (e.g., based on the accuracy of parametric
monitoring or other data) and not on an arbitrary efficiency
rating. Two commenters (A-90-19: IV-D-74; IV-D-108)
suggested that the high cost of tracking averages will
discourage sources from claiming insignificant increases in
efficiencies.
Response; Reference control efficiency ratings for RCT
were established because there is a minimum level of emissions
reduction that can be achieved by each RCT. It is
acknowledged that due to the different characteristics of
emissions to be controlled, RCT can sometimes achieve greater
emission reductions than predicted by the RCT's reference
2-47
-------
efficiency rating. However, the EPA still maintains that
providing credits for these instances of better RCT
performance is inappropriate for the same reasons stated in
the proposal preamble.
First, the magnitude of debits, not just credits, is
based on the RCT's reference efficiency ratings. Emission
debits are calculated as the difference between the actual
uncontrolled or undercontrolled emissions and the emissions if
RCT had been installed. Of course, because debit generators
are uncontrolled or undercontrolled, the actual control
efficiency that would have been achieved by the RCT cannot be
determined, so a reference control efficiency must be assumed.
It is impractical to require continuous testing of the debit
generator to determine the actual level of control that would
be achieved if RCT were applied.
If it could be determined that the RCT on a debit
generator could achieve greater reductions than its rated
efficiency, the magnitude of debits from the point would be
greater. Thus, to give credit for reductions above an RCT's
rated efficiency and not to increase the magnitude of debits
as well would represent a windfall from averaging. It would
also result in a net increase in emissions over the level that
would be expected if there were no emissions averaging. The
policy of reference control efficiency ratings for RCT is fair
as long as it is applied equally to debit and credit
generators.
Second, to grant credits for the small amount of emission
difference that might occur above a reference efficiency would
lead to significant enforcement problems. It would be very
difficult for a source to ensure that, on a continuous basis,
an RCT achieves an emissions reduction above its reference
efficiency rating. It would be even more difficult, if not
impossible, for sources to prove to inspectors that they are
in fact achieving these higher levels of efficiency. Use of a
reference control efficiency for each RCT allows inspectors to
simply check that the equipment is in place and operating as
planned. Then, the implementing agency can check records to
2-48
-------
examine the calculation of debits and credits in order to make
a compliance determination.
Hence, the use of reference efficiency ratings helps
ensure that the emissions averaging system will result in the
same or greater emission reductions as point-by-point
compliance. In addition, the use of reference efficiency
ratings simplifies the emissions averaging system, thus making
it more easily enforced.
Allowing credits for reductions that go beyond a
benchline standard (i.e., the reference control efficiency) is
consistent with the concept of MACT. Although reference
efficiencies have been established for the RCT's, the EPA does
not consider it inconsistent to allow credit for higher
efficiencies achieved by means other than the RCT's. If a
source can achieve a higher control efficiency than a RCT
through use of an alternative technology or pollution
prevention measure, it is achieving more emission reduction
than required by MACT. The source's alternative technology or
pollution prevention measure may not have been established as
MACT because MACT must be set for a source category, and as
such, must be universally available for that source category.
The fact that one source can employ control technologies that
exceed MACT does not mean all sources can use the same
technologies.
Comment: One commenter (A-90-19: IV-D-58) considered it
acceptable to disallow credit for the use of control equipment
above its designated reference efficiency rating except where
a storage vessel is controlled with a closed-vent system with
a control device and for process vents in certain
circumstances.
On the other hand, one commenter (A-90-19: IV-G-1)
considered it inconsistent to allow credit for 99.9 percent
control of vents and 98 percent control of storage vessels
using RCT's, but not to allow similar credits for other
emission points that can document and maintain a level of
control higher than the nominal efficiency. The commenter
(A-90-19: IV-G-1) argued that credits should be allowed for
2-49
-------
operating RCT's above the nominal efficiency, and such credits
are not "windfalls" because the source must commit to
enforceable measures to assure the reductions are continuously
achieved.
Response: The proposed rule allowed credit for the use
of RCT's at higher efficiencies than their nominal
efficiencies under certain conditions for process vents and
storage vessels. Comment was requested on whether to allow
credit for use of RCT's at higher efficiencies than their
rated control efficiencies for all of the emission points
allowed in emissions averaging. However, the EPA has decided
not to include the proposed allowance in the final rule for
the reasons stated in the previous response.
Comment; Two commenters (A-90-19: IV-D-74; IV-D-108)
argued that in the cases where credit is allowed for the use
of RCT at higher than rated efficiencies, parametric
monitoring, as well as continuous emission monitoring, should
be allowed.
One commenter (A-90-19: IV-D-34) objected to the
provisions in proposed §§63.150(h)(6)(iii) and (iv) that
require the control for process vents to achieve greater than
99.9 percent reduction to be allowable for a reduction
efficiency greater than the RCT. The commenter (A-90-19:
IV-D-34) complained that the requirement to achieve greater
than 99.9 percent reduction is overly conservative and will
make the emissions averaging program very difficult to use.
The commenter (A-90-19: IV-D-34) suggested that the sections
be changed to allow credit for achieving any efficiencies
greater than the rated efficiency of the RCT.
Response; For the reasons previously stated, the
proposed allowance to give credit for the use of RCT above its
designated reference efficiency rating was not included in the
final rule. Hence, the provisions for process vents to which
one commenter referred have been removed. However, as stated
in a previous response, if a control other than the RCT is
used, and that control has an approved nominal efficiency
2-50
-------
greater than that of the RCT, the additional reduction is
creditable in averaging.
2.5.3 Credits for Previous Actions
Comment: Several commenters (A-90-19: IV-D-9; IV-D-10;
IV-D-11; IV-D-45; IV-D-49; IV-D-85 and IV-F-7.39 and IV-F-12
and IV-G-6; IV-D-87; IV-D-90; IV-D-99; IV-D-100; IV-D-118;
IV-D-122; IV-D-123; IV-D-124; IV-D-125; IV-F-1.5; IV-F-7.23)
opposed allowing credits for previous actions.
Six commenters (A-90-19: IV-D-9; IV-D-85; IV-D-99;
IV-D-118; IV-D-124; IV-D-125) warned that allowing credits for
previous actions will lead to double-counting and the creation
of "paper credits." Six commenters (A-90-19: IV-D-9; IV-D-85
and IV-F-7.39 and IV-F-12 and IV-G-6; IV-D-87; IV-D-118;
IV-D-124; IV-D-125) claimed that allowing such credits
violates the maximum achievable reductions requirement.
In contrast, several commenters (A-90-19: IV-D-32;
IV-D-33; IV-D-50; IV-D-56; IV-D-62; IV-D-71; IV-D-73; IV-D-74;
IV-D-78; IV-D-79; IV-D-80; IV-D-92; IV-D-108; IV-G-1)
supported allowing credit for previous actions.
Three commenters (A-90-19: IV-D-32; IV-D-69; IV-D-73)
considered it unfair to penalize sources that installed
controls "too early" by not allowing credit, and argued that
such early actions should be rewarded instead. One of the
commenters (A-90-19: IV-D-32) declared that the test for
allowing credit for prior reductions should be whether a
reduction is otherwise required by another standard, not
whether it would otherwise exist. The commenter (A-90-19:
IV-D-32) complained that not allowing credit for preexisting
controls could lead to discontinued control of Group 2 points
because a source might decide to move the control equipment to
a Group 1 point instead of purchasing new equipment.
Response; Credit is not allowed in the final rule for
previous actions, i.e., actions taken prior to November 15,
1990, the date of passage of the 1990 Amendments to the Act.
As stated in the proposal preamble, emission reductions from
previous actions occurred for reasons unrelated to the
Amendments (such as other State requirements) or this rule and
2-51
-------
are included in the source's control on the baseline date. if
the EPA allowed reductions from previous actions to qualify
for credits, then the source would be able to generate more
debits and, thus, more total emissions than would be allowed
under point-by-point compliance.
For this reason, not allowing credit for previous actions
should not be considered unfair or a sort of penalty. Rather,
the provision is necessary to maintain emissions averaging as
an alternative means of compliance, achieving equal or greater
reductions than the rule without averaging. Likewise, it
cannot be considered a "reward" to allow credit for previous
actions, which then enables a source to emit more pollution
than would otherwise be allowed. Also, if a previous
reduction was required by another State or Federal rule, the
control can be used to meet the HON requirements for Group 1
points as lonq as the control is to the level that the HON
specifies. However, the control cannot be used to generate
emissions averaging credit.
It is possible that because no credit is allowed for
previous actions, some owners and operators may choose to
relocate existing controls from Group 2 points to other points
instead of installing new devices as long as the contro-ls on
the Group _ points were not required by o~_ ?r State or Federal
rules. However, as long as the higher-emitting Group 1 points
are controlled to the required level or reductions equivalent
to controlling Group 1 points are achieved, the objective of
the rule is ree ized.
Comment: vo commenters (A-90-19: IV-D-74; IV-D-108)
suggested that disallowing credit for prior controls indicates
that a source has a "baseline" level of control, and claimed
that the concept of a baseline is incompatible with a
technology-based standard. The commenters (A-90-19: IV-D-74;
IV-D-108) considered it a contradiction that a control can
meet MACT limits no matter when it was installed, but then is
ineligible to generate emissions averaging credit. The
commenters (A-90-19: IV-D-74; IV-D-108) did not agree that
credit should not be allowed for prior reductions because they
2-52
-------
occurred for reasons unrelated to the rule. The commenters
(A-90-19: IV-D-74; IV-D-108) also challenged the argument
that allowing credit for prior actions enables a source to
generate more emission debits and thus, more total emissions.
The commenters (A-90-19: IV-D-74; IV-D-108) labeled this a
risk-based analysis, which they claimed is also incompatible
with a technology-based standard because total emissions are
not properly the subject of a technology standard.
Response; It is true that the HON is a technology-based
standard; however, emissions averaging has been established as
an alternative means of compliance. In order to compare
reductions under averaging to reductions that would have been
achieved under point-by-point compliance, a baseline level of
control must be established. The EPA has established that
baseline level to be the controls existing at the time of the
passage of the 1990 Amendments.
One commenter is correct that some existing controls,
which may be used to comply with the rule without averaging,
cannot be used in emissions averaging. This is not a
contradiction, however, because averaging must achieve
equivalent or greater reductions than point-by-point
compliance. Also, in this case the requirement to achieve
equivalent or greater reductions is not associated with risk-
based analysis. Regardless of whether a standard establishes
a limit on a total mass quantity or a percent reduction basis,
total emissions must be compared to establish that averaging
represents a truly equivalent option to point-by-point
compliance.
Comment: Six commenters (A-90-19: IV-D-32; IV-D-33;
IV-D-69; IV-D-73; IV-D-79; IV-D-86) argued that there should
be no baseline date for credits. One commenter (A-90-19:
IV-D-73) stated that RCT's placed on Group 2 storage vessels
or transfer racks before November 15, 1990 are easily
verifiable and sources should be able to obtain approval to
credit them. Another commenter (A-90-19: IV-G-1) suggested
that sources that have applied controls before 1990 for
reasons other than the 33/50 Program (which is described in
2-53
-------
EPA Publication Number EPA-741-K-92-001) or Early Reductions
Program should be allowed to take credit for such controls as
long as they were voluntary. The commenter (A-90-19: IV-G-1)
added that no other cutoff date in Title III of the 1990
Amendments turns on enactment. One commenter (A-90-19:
IV-D-92) suggested that credit should be allowed for any non-
federally enforceable reduction after January 1, 1987, which
is the end of the first reporting year for SARA title III.
On the other hand, one commenter (A-90-19: IV-D-73)
supported a baseline date of November 15, 1990 or earlier and
stated that this date is far superior to the HON promulgation
date for a baseline. The commenter (A-90-19: IV-D-73) was
convinced that if the baseline date were the date of
promulgation, many voluntary emission reduction projects would
be put on hold until the relevant MACT standard is
promulgated. The commenter (A-90-19: IV-D-73) suggested that
the baseline date of November 15, 1990 would be
environmentally neutral in that voluntary projects will not be
inhibited, and credit would be allowed for many voluntary
emission reduction projects because the bulk of such projects
were implemented since 1990. The commenter (A-90-19:
•- •
IV-D-73) recommended that the same baseline date should be
applied to emissions averaging programs for other source
categories as well.
Response: As stated in the proposal preamble, credit is
allowed for controls put in place before the rule is
promulgated but after the Amendments were enacted. Because
the 1990 Amendments require the promulgation of emission
standards, many sources began installing controls in
anticipation of upcoming regulations. If these controls were
not creditable in averaging and the rule as a whole, these
sources would be at a disadvantage relative to other sources
that chose to postpone emission reductions until required by
rule. Thus, allowing credit for controls put in place since,
and presumably because of, passage of the Amendments creates a
more equitable emissions averaging system.
2-54
-------
The EPA appreciates one conunenter' s support of the
decision to establish November 15, 1990 as the baseline date.
However, it should be noted that decisions for future NESHAP
will be made on a specific source category basis. It should
not be assumed that the inclusion of emissions averaging in
this rule indicates that averaging will be allowed for other
source categories. If averaging is included other rules,
baseline dates will again be proposed, and public comment will
again be solicited.
Comment; One commenter (A-90-19: IV-D-108) suggested
that it was inconsistent to disallow credit for reductions
from programs the EPA had encouraged such as the 33/50
Program, individual company reduction programs, and possibly
the EPA's Early Reductions Program, all of which usually have
baseline dates in 1987 or 1988. The commenter (A-90-19:
IV-D-108) warned that disallowing credit will discourage
companies from making reductions that are not immediately
required by a rule. Another commenter (A-90-19: IV-D-69)
supported allowing credits for the 33/50 Program to continue
encouraging voluntary participation in future programs. A
third commenter (A-90-19: IV-D-73) recommended allowing
credit for emission points controlled as a part of the program
initiated by the EPA Administrator in August 1989, which was
the predecessor to the 33/50 Program. The commenter (A-90-19:
IV-D-73) argued that at the time of making commitments to the
predecessor program to the 33/50 Program, companies were
assured by the EPA that the reductions would be creditable to
the extent allowed by the Act. The commenter (A-90-19:
IV-D-73) warned that such cooperative efforts could be
undermined if credits are not allowed.
In contrast, two commenters (A-90-19: IV-D-90; IV-D-100)
stated that emission reductions from the 33/50 Program, Early
Reductions Program, or compliance with existing State
regulations do not reflect actual emission reductions required
by the HON and will result in double-counting of emission
reductions. Two more commenters (A-90-19: IV-D-49; IV-D-85)
considered the regulatory benefit of delayed compliance under
2-55
-------
the Early Reductions Program, combined with the public
relations benefits and economic benefits of pollution
prevention strategies sufficient incentive to encourage early
control.
One of the commenters (A-90-19: IV-D-85) maintained that
the EPA did not promise industry that voluntary reductions
would be creditable toward future requirements, and the EPA
does not have legal authority to do so. The commenter
(A-90-19: IV-D-85) stated that crediting pollution prevention
and 33/50 reductions conflicts with Congressional intent by
crediting non-enforceable prior reductions for enforceable
reductions made within certain dates. The commenter (A-90-19:
IV-D-85) further maintained that Congress did not intend to
authorize evasion of its limited Early Reductions policy by
allowing credits for reductions that were not formally part of
the program.
Response; As proposed, the rule disallowed credit for
previous actions with three exceptions: (1) pollution
prevention measures taken after 1987 and qualifying under the
EPA's Pollution Prevention Strategy; (2) 33/50 commitments; or
(3) Early Reductions •commitments other than equipment
shutdowns. However, in the final rule, these exceptions were
deleted for the sake of consistency and to reduce some of the
complexity of implementing the averaging program. One aspect
of the proposal was retained; controls applied as part of an
Early Reductions commitment can begin to generate credits only
if the points were not controlled to comply with other State
or Federal rules and only after the relevant point becomes
subject to the rule, i.e., after the expiration of the 6-year
extension for the Early Reductions source.
The proposal to allow three exceptions drew a great deal
of negative public comment. Moreover, the EPA concluded that
allowing credit for previous actions would actually provide
little benefit to industry. In order to get credit, the
measures taken under these programs would have to have been
overcontrol of a Group 1 point or control of a Group 2 point,
which are both unlikely. Instead, it is more common that a
2-56
-------
source controlled their largest-emitting Group 1 points to the
reference control efficiency under an Early Reductions or
33/50 commitment. Thus, the EPA has concluded that there are
probably very few previous actions taken under either program
that could generate emissions averaging credit.
The EPA disagrees that not allowing these potential
credits will discourage companies from making reductions that
are not immediately required by future rules. Setting the
baseline date as the Amendments enactment date instead of the
rule promulgation date should provide some motivation for
companies to make voluntary early reductions to comply with
future NESHAP. Also, the commitments made under the three
programs are creditable in the rule, but not in the strict
sense of emissions averaging credit. Previous actions under
these three programs or to comply with other State and Federal
rules are creditable if they achieve the required level of
emission reduction on a Group 1 point, that is if they satisfy
the requirements of point-by-point compliance. But, as
discussed previously, if these reductions were to be counted
as emissions averaging credit, the source would emit more
HAP's than would otherwise be allowed.
The EPA is committed to the success of the 33/50 and
Early Reductions Programs and encourages the use of pollution
prevention wherever feasible; this rule does not diminish that
commitment. However, to allow emissions averaging credit for
any prior reductions, regardless of the program with which
they are associated, would result in less stringent compliance
than the rule without averaging.
Comment; One commenter (A-90-19: IV-D-85) outlined an
example where a facility controlled a non-exempt vent in 1988
and reduced the emissions from 1000 pounds of HAP's to
500 pounds. The commenter (A-90-19: IV-D-85) further
hypothesized that under an emissions averaging plan the
facility would elect not to control the non-exempt vent in
exchange for additional control of an exempt emission point.
The commenter (A-90-19: IV-D-85) maintained that the value of
the debit would be 98 percent of 500 pounds (490 pounds), and
2-57
-------
if the plant had not made this prior reduction, its debit
would have been 98 percent of 1000 pounds (980 pounds). The
commenter (A-90-19: IV-D-85) concluded that the source's
500 pound pollution reduction reduces its obligation to the
public by 990 pounds.
Response; It is assumed that by "non-exempt," the
commenter refers to Group 1 emission points; "exempt" is taken
to mean Group 2 points. The final rule does not allow credits
for control measures taken prior to November 15, 1990, so this
example no longer applies. Even so, it should be noted that
the commenter calculated debits for this example incorrectly.
Debits are calculated as the difference between actual
and allowable emissions from a point. In this example, debits
would be generated by leaving a Group 1 process vent
uncontrolled or undercontrolled. Allowable emissions for
Group 1 points are the emissions that would result if RCT were
applied. Even though the example process vent was controlled
to 50 percent prior to November 15, 1990, because it is a
Group 1 vent, the rule requires that current emissions be
reduced by 98 percent. Hence, the allowable emissions from
this Group 1 point are 20 pounds, i.e., 2 percent of its
uncontrolled emissions, 1000 pounds. (See the provisions in
§63.150(g) of the rule for calculating emissions averaging
debits.)
If the existing control achieving a 50 percent reduction
is maintained, the actual emissions from the vent would be
500 pounds, generating a debit of 480 pounds (500 pounds of
actual emissions minus 20 pounds of allowed emissions). To
balance the debit, the source would have to overcontrol
another point or points by at least 480 pounds. If the
existing control device is removed entirely, the debits would
a 980 (1000 minus 20) pounds. In either case, it is
incorrect to conclude that if the point is designated as a
debit generator, the 50 percent control achieved in 1988 would
reduce the source's obligation to the public by 990 pounds.
Comment; One commenter (A-90-19: IV-D-51) concurred
with the conditions set for obtaining emission credits from
2-58
-------
previous actions, but could not support a plan that would
accumulate credits over a period greater than the averaging
compliance period.
Response: Presumably, the commenter was recommending
that the rule not allow a source to bank credit from previous
actions for use to balance future debits. The commenter's
concern has been addressed by not allowing credit for previous
actions and deleting credit banking from the final rule. The
discussion of the deletion of credit banking from emissions
averaging is found in section 2.11 of this BID volume.
Comment: One commenter (A-90-19: IV-D-45) questioned
why credits were being allowed for prior emissions reductions
on process units that are not covered under the HON rule.
Response; As discussed in section 2.4 of this BID
volume, emissions from points in process units and sources
that are not subject to the HON are not eligible for this
emissions averaging program.
Comment: One commenter (A-90-19: IV-D-69) considered
not allowing credit for prior reductions to be inconsistent
with the Act because the Act defines "new sources" based on
the date of NESHAP proposal.
Response: The commenter's claim was not clear regarding
the relationship between the date new sources are defined and
credit for previous actions. However, it is not inconsistent
with the Act to disallow credit for previous actions, and as
discussed in section 2.3.2 of this BID volume, in the final
rule averaging is not allowed at new sources.
Comment: One commenter (A-90-19: IV-D-89) suggested
that under the programs established by some States where a
facility can register and agree to not produce above a certain
level of emissions, the facility should be considered a credit
generator.
Response: An entire source cannot be a credit generator;
only emission points within sources can be used to generate
credits and debits. If a source can generate more credits
than debits overall, it still cannot be a net credit generator
because averaging is not allowed between sources as discussed
2-59
-------
in section 2.4 of this document. A source that participates
in a State program by agreeing to limit their total emissions
can use emissions averaging to comply with the HON and with
the program, but it is not within the scope of the HON
emissions averaging program for a source to be a net credit
generator.
2.5.4 Credit for Pollution Prevention and Recycling
Comment: Several commenters (A-90-19: IV-D-32; IV-D-50;
IV-D-57; IV-D-69; IV-D-71; IV-D-72; IV-D-79; IV-D-80; IV-D-83;
IV-D-86; IV-D-104; IV-D-106; IV-G-1) supported allowing credit
for pollution prevention measures as an additional incentive
for conducting such measures.
Two commenters (A-90-19: IV-D-32; IV-D-57) stated that
pollution prevention is almost always a superior means of
environmental protection. Three commenters (A-90-19: IV-D-83
and IV-F-1.3 and IV-F-5; IV-D-104; IV-G-1) suggested that
emissions averaging would encourage pollution prevention,
which is expressly authorized as a control measure for
reducing HAP emissions under section 112(d) of the Act. One
commenter (A-90-19: IV-D-83) stated that without averaging,
the use of pollution prevention could greatly decrease because
pollution prevention projects may not be able to achieve the
RCT efficiencies at each ^nd every emission point, whereas if
averaging is allowed, sy. :am-wide pollution pre- ntion
programs could be used to achieve compliance. The commenter
(A-90-19: IV-D-83) added that EPA Administrator Carol Browner
has, on several recent occasions, expressed her support for
pollution prevention.
One commenter (A-90-19: IV-D-74) urged the EPA to
develop a mechanism to allow the use of pollution prevention
projects to achieve MACT because such a use in MACT is
explicitly authorized in the section 112(d)(2) of the Act.
In contrast, one commenter (A-90-19: IV-D-103) opposed
allowing credit for pollution prevention, and stated that the
concept of pollution prevention is to prevent releases of a
pollutant, rather than shifting the effects or impacts in time
or space. One commenter (A-90-19: IV-D-44 and IV-F-7.28)
2-60
-------
objected to allowing the savings of pollution prevention
measures as "loopholes" and favored requiring pollution
prevention planning. Another commenter (A-90-19: IV-D-85 and
IV-G-6) recommended that pollution prevention should be
required in addition to reductions achievable through control
of emission points. The commenter (A-90-19: IV-D-85)
asserted that a system of comprehensive control with RCT that
accepts pollution prevention as an alternative control
mechanism will encourage environmentally beneficial pollution
prevention far more effectively than emissions averaging.
Response; Credit is allowed for reductions achieved by a
pollution prevention measure applied after November 15, 1990
to a Group 2 point or to a Group I point if the pollution
prevention measure achieves reductions greater than what could
be achieved using the RCT.
The EPA acknowledges that some of the emission reductions
from a pollution prevention measure will be offset by emission
increases elsewhere in the source if the pollution prevention
measure is used to generate credit for an average. However,
the EPA does not agree that emissions averaging interferes
with the intent of pollution prevention by allowing emissions
to be "shifted" instead of preventing their release
altogether. The intent of pollution prevention is to reduce
emissions in an economical and environmentally sound manner.
Under emissions averaging, it does not matter how emissions
are controlled so long as the level of reduction required by
the rule is achieved.
Pollution prevention is a method to reduce emissions that
is highly desirable because it often results in emission
reductions in several media. The EPA encourages its use to
the fullest extent; this emphasis in encouraging pollution
prevention is one o.f the reasons for allowing the use of
emissions averaging.
Comment: Two commenters (A-90-19: IV-D-32; IV-D-57)
encouraged allowing pollution prevention credit for cases in
which a source reduces its emissions by switching from
production of one chemical to another.
2-61
-------
One commenter (A-90-19: IV-D-85) recommended that
pollution prevention be carefully defined and that the current
exclusion of product switches from the definition is
essential. The commenter (A-90-19: IV-D-85) maintained that
companies will claim credits for product switches which would
have occurred in any case without taking debits for product
switches which increase pollution.
Response; The EPA solicited comment on whether credit
should be granted if a source reduces emissions by switching
from production of one chemical to another. For the final
rule, the EPA has maintained the policy that a process
conversion that qualif 3 as a pollution prevention measure as
defined in the EPA's Pollution Prevention Strategy
(56 FR 7849; February 26, 1991) and occurs after November 15,
1990 is eligible for credit. To qualify under the pollution
prevention strategy the process must be used to make the same
product before and ~er the pollution prevention conversion.
It should be noted -.iat only two commenters expressed interest
in generating credit by switching products and neither
provided sufficient information to convince the EPA that the
rule should differ from its Pollution Prevention Strategy.
Comment: Four commenters (A-90-19: IV-D-74; IV-D-98;
IV-D-104; IV-D-108) argued that for Group 1 emissidn points,
requiring pollution prevention projects to result in emission
reductions greater than that available through RCT was too
limiting, and recommended that credit be allowed for
reductions achieved through pollution prevention if they are
comparable to RCT or if they are substantial. Three
commenters (A-90-19: IV-D-74; IV-D-83; IV-D-108) explained
that a reduction of 98 percent or better from pollution
prevention is infrequent. Three commenters (A-90-19:
IV-D-74; IV-D-104; IV-D-108) suggested that the more likely
example is when an emission point with existing controls
undergoes pollution prevention to achieve a 98 percent
reduction. Four commenters (A-90-19: IV-D-74; IV-D-83;
IV-D-104; IV-D-108) argued that not allowing credit in such
2-62
-------
situations will discourage pollution prevention for already
controlled emissions sources.
Response: Pollution prevention measures at Group 1
points are treated the same as control measures. Where they
achieve less reductions than the RCT, they can be included in
an emissions averaging on debit-generating points. Where they
achieve reductions equivalent to the RCT, they can be used for
compliance but not as a credit-generating point. Where they
achieve greater reductions than the RCT, they can serve as a
credit generator. The system thereby encourages all types of
pollution prevention measures.
A pollution prevention measure in conjunction with add-on
controls achieving greater reductions than RCT is also
eligible for emissions averaging credit. An example of this
was presented in the proposal preamble in which a pollution
prevention process change reduces the annual amount of
wastewater a source generates by 50 percent. Then, an add-on
control (which happens to be the wastewater RCT in the
example, but need not be) is applied, and the emission
reduction from the two combined exceeds the reduction
achievable by the RCT only. The surplus reductions over that
achievable by the RCT can be used for credits in emissions
averaging.
There is a significant difference between the proposed
and final rule, however. In the example in the proposal
preamble, the pollution prevention measure was applied in
1988, prior to the baseline date. In the final rule, this
previous action is no longer creditable; it is counted in the
baseline level of control for the source. Thus, assuming the
wastewater stream remains a Group 1 point even with the 50
percent flow reduction achieved by the pollution prevention
measure, the RCT must still be applied.
Comment: Several commenters (A-90-19: IV-D-32; IV-D-57;
IV-D-69; IV-D-72; IV-D-79; IV-D-80; IV-D-86; IV-D-104;
IV-D-106; IV-G-1) supported allowing credit for recycling as
an incentive for conducting such measures. Five commenters
(A-90-19: IV-D-32; IV-D-57; IV-D-69; IV-D-98; IV-D-104)
2-63
-------
considered the definition of pollution prevention in the EPA's
Pollution Prevention Strategy too narrow, and suggested that
out-of-process as well as in-process recycling should be
considered pollution prevention and eligible as a credit
generator.
One commenter (A-90-19: IV-D-32) suggested that between
the enforceability aspects of including recycling in the
Title V operating permit program and the quantification
aspects of the HON and the Early Reductions Rule,
quantification of reductions from recycling is fully workable.
One commenter (A-90-19: IV-D-51) was hesitant to endorse
credits generated from recycling activities because recycling
activities may require collection activities off-site, and it
may be difficult to account for all emissions associated with
the recycling activities. The commenter (A-90-19: IV-D-51)
stated that recycling credit may work if it can be shown that
the recycling activity was self-contained.
Response; The EPA requested comment on the issue of
allowing credits for recycling activities that can result in
quantifiable emission reductions. In the final rule, since
in-process recycling is a pollution prevention measure, it can
be used to generate credits. Credits would be calculated as
provided in the rule for any pollution prevention measure.
On the other hand, it has been determined that emission
reductions from out-of-process recycling, which is not a
pollution prevention measure, cannot be included in emissions
averaging because out-of-process recycling is out of the
jurisdiction of this rule. Out-of-process recycling involves
waste management outside of the HON source, and is thus not
subject to this standard.
2.5.5 Plant Shutdowns and Slowdowns
Comment: Six commenters (A-90-19: IV-D-33; IV-D-56;
IV-D-58; IV-D-62; IV-D-92; IV-D-98) urged the EPA to allow
credit for production cutbacks or plant shutdowns. One
commenter (A-90-19: IV-D-56) claimed that allowing credit was
appropriate for cutbacks or shutdowns that are part of an
Early Reductions commitment.
2-64
-------
Two conunenters (A-90-19: IV-D-58; IV-D-62) disagreed
with the argument that credit should not be allowed because
shutdowns would have happened anyway. One commenter (A-90-19:
IV-D-58) argued that most shutdowns occur for a combination of
reasons. Another commenter (A-90-19: IV-D-62) added that it
is doubtful that a plant shutdown is ever a foregone
conclusion.
Two commenters (A-90-19: IV-D-58; IV-D-62) argued that
even if a shutdown were to occur regardless of any MACT
standard, there is still a net benefit from the reduction in
emissions. One commenter (A-90-19: IV-D-62) reasoned that
these reductions should qualify for credit just as they do
according to the 1986 Emissions Trading Policy Statement and
in the EPA's 33/50 Program. Moreover, the commenter (A-90-19:
IV-D-62) stated that neither the EPA's Pollution Prevention
Strategy nor the Act excluded or discouraged the use of
permanent shutdowns as a method of reducing emissions.
One commenter (A-90-19: IV-D-58) recommended that
permanent shutdowns after MACT promulgations should be
creditable with a 5 year lifespan, discounted at a straight-
line rate of 20 percent per year after the first year. The
commenter (A-90-19: IV-D-58) suggested that one mechanism to
achieve this is to credit shutdowns as part of "banking." The
commenter (A-90-19: IV-D-58) further recommend that shutdowns
associated with an approved Early Reductions, pollution
prevention, or 33/50 Program should be creditable for the
emissions averaging program based on recent actual emission
estimates.
One commenter (A-90-19: IV-D-51) commended the EPA for
not allowing shutdowns to generate credits. One commenter
(A-90-19: IV-D-85) stressed that neither permanent closures
nor maintenance shutdowns should generate credits because they
will occur from time to time regardless of environmental
decisions.
Response; It is not appropriate to allow credit in
emissions averaging for permanent shutdowns or slowdowns even
if they are part of an Early Reductions commitment under
2-65
-------
section 112(i)(5) of the Act. No matter what the motivation
for a shutdown or slowdown, the emission reductions from the
production curtailment are not made permanent if emissions
averaging credit is allowed. If credit were granted for the
emission reduction, the source could then emit an equal amount
of emissions from its debit generators. This is in contrast
to point-by-point compliance, where if a point is shut down,
the emissions reduction is permanent. To allow credit in
emissions averaging for permanent shutdowns and slowdowns
results in less stringent compliance and more total emissions
than point-by-point compliance, in which case emissions
averaging does not represent an equivalent compliance
alternative.
2.5.6 Approval Process for New Control Technologies
Comment: One commenter (A-90-19: IV-D-74) suggested
that where innovative control technology is submitted for
approval as a reference technology, the EPA should provide
procedures to account for the uncertainties encountered such
as during development of the technology from pilot to full
scale. The commenter (A-90-19: IV-D-74) added that if the
approved innovative technology does not achieve the assigned
efficiency rating, the source should be allowed adequate time
to come into compliance.
Response; Development of an innovative control
technology from pilot to full scale should be completed before
the technology is submitted for approval as an alternative to
RCT. The owner or operator seeking permission to take credit
for a new technology must be sufficiently confident of the
technology so that upon installation, it will immediately
comply fully with the rule. Moreover, the source must be able
to demonstrate ongoing compliance according to the provisions
of the rule. Development of new control technologies is, of
course, desirable; however, development cannot interfere with
attainment of the standard, especially when proven
technologies (i.e., the RCT's) are available.
Comment; Two commenters (A-90-19: IV-D-74; IV-D-108)
suggested that approval of new technologies for a specific
2-66
-------
source should be independent of the approval for nationwide
use because the control technology may be highly chemical-
specific, and its use in more than three applications at a
particular source does not necessarily mean that the
technology is widely applicable. The two commenters (A-90-19:
IV-D-74; IV-D-108) were concerned that a source should not be
delayed in its application of the technology pending review of
the technology for wider applicability.
Response; The effectiveness of innovative control
technologies that are different either in use or design from
RCT must be demonstrated prior to their use. If a source
wishes to use a new technology in more than three
applications, the EPA must approve the new technology and
assign it a nominal control efficiency. Also, EPA approval
constitutes approval of broad applicability, that is, use of
the new technology by any source subject to the rule.
If a new technology would be used in no more than three
applications in a given facility, the permitting authority,
instead of the EPA, can assign it a nominal control
efficiency. If the permitting authority feels the new
technology may have broad applicability beyond the three or
fewer applications for which it was submitted, the permitting
authority must forward the information about the technology to
the EPA. However, any subsequent EPA review of the technology
will be performed in parallel with the approval of the nominal
control efficiency by the permitting authority and will not
affect or delay the approval process. Once the permitting
authority has approved the nominal control efficiency of the
technology for compliance, the source can proceed with
installing and operating the technology.
Thus, if a new technology can be used in more than three
applications, the EPA will be involved in the nominal control
efficiency approval. Review for broad applicability is a
separate issue and again, it will never delay the nominal
control efficiency approval process. But for other than site-
specific innovations, the EPA intends to maintain a close
2-67
-------
oversight of new technologies that have promise as future
reference control technologies.
Comment: Two commenters (A-90-19: IV-D-56; IV-D-74)
argued that the procedure for approving new technologies
provided an excessive amount of time, 120 days, to determine
data sufficiency. One commenter (A-90-19: IV-D-56) noted
that the completeness and adequacy review periods for PSD are
not even this, long. The other commenter (A-90-19: IV-D-74)
recommended that the period should be reduced from 120 days to
within 30 days. The commenter (A-90-19: IV-D-74) recommended
that the rule should allow a source to proceed with an
alternative control upon approval rather than requiring the
source to wait for the notice in the Federal Register.
Response: The EPA considers it necessary and responsible
to be deliberate in considering new technologies especially
when proven RCT's are available. As such, it is not excessive
to reserve 120 days to consider the sufficiency, of the data
and 120 days more to approve the nominal control efficiency
for never-before-seen technologies. Moreover, the EPA
considers it reasonable to use a published notice in the
Federal Register to alert the regulated community that a new
RCT is available when the new technology has broad
applicability because a widespread understanding of new
technologies could lead to further control advancements.
However, use of the approved new technology will not be
delayed.
Comment: One commenter (A-90-19: IV-D-74) recommended
that the use of more than one control in series should be
allowed for MACT compliance if the total reduction equals or
exceeds the required percentage reduction.
Response: A combination of control devices or techniques
is allowed for compliance. If a source can achieve or exceed
an RCT's rated efficiency by using a combination of control
devices, such a strategy can be used to generate credits in
emissions averaging from Group 1 points. If further
reductions can be obtained by adding more controls to a
Group 1 point that already has RCT installed, the surplus
2-68
-------
reductions are eligible for emissions averaging credit as
well. Of course, any combination of technologies or the use
of RCT below its rated efficiency may be used to control a
Group 2 point for emissions averaging credit.
Comment; One coimnenter (A-90-19: IV-D-51) requested
further guidance in determining when the permitting authority
(rather than the EPA) has the authority to assign reference
efficiency ratings for new control technologies. The
commenter (A-90-19: IV-D-51) was concerned that the mechanism
for determining the reference efficiency must be a timely
process in order not to discourage innovative methods for HAP
control. The commenter (A-90-19: IV-D-51) suggested that a
mechanism and schedule for new control technology approval be
contained in the final rule.
Response; As discussed previously, the rule provides
that where a new control technology is to be used in no more
than three applications at a single plant-site, the permitting
authority can approve its use and assign its nominal
efficiency. The permitting authority shall refer a technology
to the EPA if it believes that the new technology has broad
applicability. The determination of "broad applicability" is
a case-by-case decision, and the elements of that
determination are left to the discretion of the permitting
authority.
When a new technology is forwarded to the EPA for further
review of broader applicability, the rule provides that the
EPA's review shall not affect the permitting authority's
approval of the nominal efficiency. Whether the permitting
authority or the EPA is responsible for reviewing a particular
application of a new technology, the same amount of time is
allocated for review of the submittal, i.e., 120 days to
determine if sufficient information has been provided and
120 days more to review and approve the new technology for
use. This process is intended to encourage innovation in
control technologies by establishing a relatively low approval
hurdle. Thus, the commenter's concern for a timely approval
process and schedule should be satisfied.
2-69
-------
Comment; One commenter (A-90-19: IV-D-56) disagreed
wirh the proposed approach for approval of new technologies
claiming it duplicates the anticipated State construction
permit program. The commenter (A-90-19: IV-D-56) recommended
that the EPA defer all construction approval review and
approval responsibility to the appropriate State in an effort
to avoid duplicative regulatory programs.
Response: The commenter's reference to an "anticipated
State construction permit program" was not clear. The HON has
been designed to be compatible with the operating permit
program rule. If a State wishes to use an existing State
construction permit program to approve new technologies for
the HON, it is free to do so, assuming all of the procedures
specified in the rule are followed. Thus, the approval
process need not be duplicative. However, the EPA will not
delegate the authority for approval of new technologies to
States that do not have an approved operating permit program
in place.
2.6 CREDIT DISCOUNT FACTORS
Comment: Eight commenters (A-90-19: IV-D-9; IV-D-45;
.IV-D-85 and IV-F-7.39 and IV-F-12 and IV-G-6; IV-D-99;
IV-D-115; IV-D-118; IV-D-124; IV-D-125) supported the use of
discount factors in emissions averaging.
S /eral commenters (A-90-19: IV-D-32; IV-D-50; IV-D-57;
IV-D-58; IV-D-62; IV-D-69; IV-D-71; IV-D-72; IV-D-73; IV-D-74;
IV-D-75; IV-D-77; IV-D-78; IV-D-79; IV-D-82; IV-D-83 and
IV-F-1.3 and IV-F-1.5; IV-D-86; IV-D-89; IV-D-92; IV-D-97;
IV-D-104; IV-D-106; IV-D-108; IV-F-1.6 and IV-F-6; IV-G-1)
opposed the use of discount factors (i.e., supported a zero
percent discount factor) in emissions averaging.
Response: A discount factor of 10 percent is required in
calculating credits in the final rule. An exception is
provided for reductions accomplished by the use of pollution
prevention measures. For pollution prevention measures, full
credit with no discounting is allowed. At proposal, the EPA
sought comment on whether it is appropriate to require the use
of a credit discount factor and what value between 0 to 20
2-70
-------
percent should be selected for the discount factor. Specific
aspects of the decision to include a credit discount factor
are addressed in the remainder of this section.
Comment: A number of commenters (A-90-19: IV-D-32;
IV-D-57; IV-D-58; IV-D-62; IV-D-71; IV-D-72; IV-D-74; IV-D-75;
IV-D-77; IV-D-83 and IV-F-1.3 and IV-F-5; IV-D-95; IV-D-97;
IV-D-98; IV-D-106; IV-D-108; IV-F-1.6 and IV-F-6) argued that
discount factors would reduce, and could completely eliminate,
the incentive to achieve compliance through emissions
averaging. One commenter (A-90-19: IV-D-58) stated that
facilities with very high cost MACT installation requirements
compared to other facilities in the same source category are
the facilities for which emissions averaging was primarily
intended. The commenter (A-90-19: IV-D-58) contended that a
discount factor could defeat the purpose of the emissions
averaging program by preserving the competitive disadvantage
of these facilities. Two commenters (A-90-19: IV-D-32;
IV-D-57) maintained that the cost savings associated with the
use of averaging for those points is potentially critical to
individual sources, even if the actual number of points and
quantity of emissions involved are relatively small.
Two commenters (A-90-19: IV-D-32; IV-D-57-) argued that
there will not be many opportunities to generate credits
because Group 2 points are by definition the ones with the
lowest emissions, and the burden of the recordkeeping and
reporting requirements provides an additional disincentive.
One commenter (A-90-19: IV-D-62) claimed that when the EPA
mentioned three variable discounting options in the proposal
preamble, it stated the most compelling argument against
discounting, which is that discounting "would greatly increase
the administrative complexity of emissions averaging, reducing
its workability." Three commenters (A-90-19: IV-D-32;
IV-D-33; IV-D-57) predicted that sources will reserve the use
of averaging for when RCT is not practicable because these
instances will be the most costly for the marginal emission
reduction. Two commenters (A-90-19: IV-D-32; IV-D-57)
concluded that the additional social and economic cost imposed
2-71
-------
by a restriction or loss of the averaging alternative will far
outweigh any marginal emission reduction derived from the use
of a discount factor.
Response: The EPA acknowledges that a credit discount
factor will make averaging of points with marginal differences
in cost effectiveness unlikely. However, the EPA disagrees
with commenters that a discount factor could completely
eliminate the incentive to achieve compliance through
emissions averaging.
The goal of emissions averaging is not to enable sources
to reduce their overall compliance costs to the industry
average, or to gain a competitive advantage. Rather, the
purpose of averaging is to allow sources to comply with the
rule in the least costly manner for their site-specific
situation. Sources will definitely realize cost savings using
emissions averaging instead of installing RCT; otherwise, they
will not use emissions averaging. The purpose of a discount
factor, then, is to ensure that the emission points selected
for averages are the ones where truly significant cost savings
can be realized and to share this savings with the
em ^nment.
omment: Two menters (A-90-19: IV-D-85; IV-D-99)
supported a discount factor because industry will enjoy a cost
savings from complying through averaging and the environment
shoul- also benefit from the flexibility and cost savings.
One commenter (A-90-19: IV-D-85) stated that if a given level
of reductions is deemed achievable without emissions
averaging, then a higher level must be possible with emissions
averaging, because emissions averaging supposedly lowers
costs.
On the other hand, several commenters (A-90-19: IV-D-32;
IV-D-57; IV-D-62; IV-D-69; IV-D-74; IV-D-75; IV-D-92; IV-D-97;
IV-D-104; IV-D-108; IV-G-1) considered a discount factor to be
inconsistent with the statutory intent that MACT be
implemented in a flexible and cost-effective fashion. Several
commenters (A-90-19: IV-D-32; IV-D-57; IV-D-62; IV-D-72;
IV-D-75; IV-D-77; IV-D-89; IV-D-104; IV-D-106; IV-G-1)
2-72
-------
submitted that as long as the EPA correctly identifies the
appropriate level for MACT in the first place, no additional
"price" or penalty should be imposed for allowing sources to
achieve that level in the most efficient manner possible.
Three commenters (A-90-19: IV-D-32; IV-D-57; IV-D-73) did not
regard a discount factor as an appropriate "price" for savings
gained from emissions averaging. The commenters (A-90-19:
IV-D-32; IV-D-57; IV-D-73) stated that averaging will be used
for emission points whose control is impractical or
substantially more costly than the average, so emissions
averaging with a discount factor will not provide
opportunities to avoid the normal costs of applying RCT.
Response: As stated previously, the EPA accepts the
rationale for using a credit discount factor that the
environment should also benefit from cost savings achieved
through emissions averaging.
The use of a discount factor is not inconsistent with the
Act nor does it represent a "price" or penalty for using
averaging. Emissions averaging is an alternative method for
complying with the MACT standard that offers flexibility and
the opportunity to apply a more cost-effective control option
for compliance. Sources are able to lower their control costs
for the points included in the average below the cost required
to comply on a point-by-point basis. The decision to include
a discount factor recognizes that a portion of the cost
savings could be used to benefit the environment, i.e., to
achieve more emission reductions than is required under point-
by-point compliance.
The EPA does not consider sharing a 10 percent portion of
savings with the environment to be so great a disincentive to
dissuade many sources from choosing to use averaging. Sources
will always realize lower control costs under averaging versus
point-by-point compliance. If this were not so or if the
source does not consider the cost savings substantial enough,
the option of emissions averaging would not be selected.
Comment: Five commenters (A-90-19: IV-D-58; IV-D-72;
IV-D-83; IV-D-106; IV-G-1) predicted that a discount factor
2-73
-------
will discourage the implementation of innovative control
technologies. One commenter (A-90-19: IV-D-58) was concerned
that discouraging innovation would slow the progression of
MACT standards over time. Two commenters (A-90-19: IV-D-83;
IV-D-89) added that pollution prevention would also be
discouraged. One of the commenters (A-90-19: IV-D-89)
claimed that voluntary pollution prevention would be
discouraged because a discount factor would penalize sources
that spend money on pollution prevention efforts and cause
competitive disadvantage. One commenter (A-90-19: IV-G-1)
recommended that no discount factor should be applied to
pollution prevention measures.
In contrast, one commenter (A-90-19: IV-D-85) stated
that a lower discount factor for pollution prevention is not
appropriate, because pollution prevention can be more
difficult to quantify and less expensive than other types of
pollution control.
Response: Credits generated by pollution prevention
measures are not discounted in the final rule. The EPA is not
concerned that a discount factor would discourage the use of
pollution prevention or any other type of control that could
achieve significant cost savings. Rather, no discount factor
is being applied to pollution prevention to identify it as the
preferred method of achieving emission reductions and thus
encourage its use.
Only measures that qualify as pollution prevention
activities according to the EPA's Pollution Prevention
Strategy are considered pollution prevention measures under
the rule and therefore are not discounted. The emissions
reductions from these measures are fully quantifiable. The
EPA cannot confirm one commenter's suggestion that pollution
prevention measures are less expensive to implement than other
types of controls; the commenter provided no accompanying
data. The EPA does not share the concern that the discount
factor selected for the final rule will discourage the
development of innovative control technologies because the
value of the discount factor is small. The EPA expects that
2-74
-------
new technologies that can reduce emissions more than existing
technologies, and do so more cost-effectively, will be
developed and implemented regardless of the application of a
small discount factor.
Comment: Four commenters (A-90-19: IV-D-32; IV-D-57;
IV-D-74; IV-D-77) suggested that because of the differences
between the current situation and past situations in which
discount factors have been used in emissions averaging, such
as offsets to avoid new source control requirements in
nonattainment areas, the discount factor is not relevant here.
One commenter (A-90-19: IV-D-97) added that because the use
of discount factors is covered in the section of the Act
dealing with nonattainment of NAAQS, applying one in this rule
amounts to double discounting and an unfair penalization. The
commenter (A-90-19: IV-D-97) suggested that a discount factor
should only apply where good performance has not been
maintained. One commenter (A-90-19: IV-G-1) argued that a
discount factor is inconsistent with the proposed Economic
Incentive Program rules (58 FR 11110, February 23, 1993). One
commenter (A-90-19: IV-D-70) stated that basing the discount
factor on the Emission Trading Policy Statement for
nonattainment areas is not appropriate because impacts of HAP
emissions are localized and result in short-term, acute
effects, whereas nonattainment issues are a more general
problem in areas with long-term goals of attaining the NAAQS.
Response; Although discount factors are used in other
programs, the rationale for their use and their implementation
can vary from program to program. Thus, differences between
this and past situations are acknowledged and expected. Even
though the reasons for using a discount factor in this rule
may be different from other regulatory programs, its use is
still relevant. As such, the use of a discount factor need
not comport with any provisions of the New Source Review
program, the Emission Trading Policy Statement, or the
proposed Economic Incentive Program rules.
2-75
-------
Comment; One commenter (A-90-19: IV-D-99) supported a
discount factor because emissions estimates are highly
imprecise.
However, four commenters (A-90-19: IV-D-32; IV-D-57;
IV-D-74; IV-D-108) disagreed with the argument that discount
factors are needed to address uncertainties in the averaging
calculation that could lead to control less stringent than
MACT. Four commenters (A-90-19: IV-D-32; IV-D-57; IV-D-74;
IV-D-108) argued that the extreme detail required to calculate
an average and the increased monitoring and reporting required
will result in more certainty, not less. Two commenters
(A-90-19: IV-D-72; IV-D-106) suggested that it would be
technically more valid for a source to include considerations
of estimation uncertainty in the credit and debit calculation.
One commenter (A-90-19: IV-G-l) stated that discount factors
should apply only in cases of demonstrated greater-than-
average uncertainty.
Response: The EPA is confident that the estimation .
methodologies are suitable for calculating debits and credits
and are equivalent and fully interchangeable. The EPA does
not consider the estimates highly imprecise as one commenter
suggested. The EPA does not rest the justification for using
a discount factor nor the value of the discount factor chosen
on the need to account for estimation uncertainty. A more
detailed discussion of the emission estimation procedures for
averaging can be found in section 2.11.4 of this BID volume.
Comment: Six commenters (A-90-19: IV-D-32; IV-D-50;
IV-D-57; IV-D-73; IV-D-74; IV-D-108) argued that a great deal
of inherent conservatism is built into the emissions averaging
rules, and that this conservatism acts as a built-in discount
factor to compensate for any possible uncertainty in debit and
credit calculations. Five commenters (A-90-19: IV-D-32;
IV-D-57; IV-D-73; IV-D-74; IV-D-108) listed not allowing
credit in most instances for control efficiencies above RCT as
one example of inherent conservatism. Eight commenters
(A-90-19: IV-D-32; IV-D-50; IV-D-57; IV-D-69; IV-D-73;
IV-D-74; IV-D-78; IV-D-108) suggested another example of
2-76
-------
inherent conservatism is that most sources will strive to
maintain excess credits for use in balancing their annual
averages. Two commenters (A-90-19: IV-D-74; IV-D-108)
indicated that for an operation with significant complexity
and variability, such as a batch operation, the excess credit
cushion may need to be substantial. Another commenter
(A-90-19: IV-D-73) suggested that banking of excess credits
would-also be a built-in safety factor.
Response; The EPA recognizes that conservative practices
are sometimes employed by sources when complying with
standards. Disallowing credit for RCT operating above its
rated efficiency is not an example of inherent conservatism
because the higher efficiency is not attributed to the debit
generator as well. Also, one commenter's suggestion that
banked credits can act as a built-in safety factor is not
valid since banking is disallowed in the final rule. The
exclusion of banking is discussed in section 2.10 of this BID
volume.
The use of conservative practices is prudent, and sources
are encouraged to avoid any possibility of compliance
violations. However, since conservative practices are not
required by the rule, the entire industry cannot be counted on
to use them. Inherent conservatism also does not address the
issue of whether cost savings realized through the use of
emissions averaging should be shared with the environment.
Even if it could be assumed that all sources would build in a
safety factor, a discount factor ensures that a specific
amount of emissions reductions will go to the environment in
exchange for cost savings.
Comment: Two commenters supported the use of a small
discount factor in the range of 0-5 percent (A-90-19:
IV-D-67), and 5 percent (A-90-19: IV-D-56), to address the
general concerns cited in the proposal preamble while
maintaining both the incentive for emissions averaging and a
simple program to implement. One commenter (A-90-19:
IV-D-56) added, however, that if sources are actually
monitored with CEM's, no discount factor should apply.
2-77
-------
Eight commenters (A-90-19: IV-D-9; IV-D-45; IV-D-70;
IV-D-85 and IV-F-7.39 and IV-F-12; IV-D-99; IV-D-118;
IV-D-124; IV-D-L25) stated that if an averaging scheme is
retained, it must use a substantial discount factor. One
commenter (A-90-19: IV-D-85 and IV-F-7.39 and IV-F-12)
maintained that the maximum achievable emissions reduction
standard requires that discount factors be set as high as
possible without discouraging trading completely.
Response; A discount factor of 10 percent was selected
for the final rule as one that provides a benefit to the
environment yet maintains the incentive for emissions
averaging. No discount factor is to be applied for pollution
prevention measures.
The commenter who suggested that a discount factor is
unnecessary if CEM's are used was probably referring to the
proposal to allow credit for the use of RCT above its rated
efficiency on process vents. The proposal, which was deleted
from the final rule, required the use of CEM's in order to get
credit. The suggestion of CEM's pertains to the issue of
uncertainty of emissions estimation in averaging, and as
stated previously in this section, the discount factor is not
included to address uncertainty.
Comment; Several commenters (A-90-19: IV-D-9; IV-D-45;
IV-D-51; IV-D-70; IV-D-85 and IV-F-7.39 and IV-F-12; IV-D-99;
IV-D-118; IV-D-124; IV-D-125) suggested that the discount
factor should take into account considerations such as: the
range of uncertainty of the emissions estimation method
(especially for wastewater); the toxicity of pollutants;
wastewater emissions; emission points with large marginal cost
of control differences; cost savings; engineering estimates or
non-specialized monitoring; differences in the frequency and
type of monitoring among emission points; differences in
dispersion characteristics between emission points; or
granting a benefit to the environment.
One commenter (A-90-19: IV-D-51) suggested that the rule
establish a procedure for determining discount factors on a
site-specific basis, which can then be applied to each
2-78
-------
individual emission point. The commenter (A-90-19: IV-D-51)
contended that considering other factors when developing a
discount factor for individual emission points would not
increase the regulatory burden.
Response: Designing the discount factor to account for
all of the considerations suggested by commenters would have
necessitated a very complicated mechanism. Instead, a single
value for the discount factor was selected with simplicity of
implementation in mind. It should not add any complexity to
include the discount factor in the calculations of credits.
Also, it should not be an added burden for the source or
authorizing agency to single out pollution prevention measures
as credits that are not discounted. Pollution prevention
measures must be specifically approved for use anyway, and
calculation of credits from their use will probably be
highlighted in the emissions averaging plan.
Comment: One commenter (A-90-19: IV-D-85) suggested
that the discount factor be based on three components: the
"marginal cost differential component", the "gaming
component", and the "volume discount component." The
commenter (A-90-19: IV-D-85) described their "marginal cost
differential component" as a system where operators retain the
minimum savings over what they would achieve by point-by-point
compliance necessary to allow the trade to occur and that this
savings rate should be a constant. The commenter (A-90-19:
IV-D-85) stated that the "gaming component" factor should be
higher when trades between different processes occur, taking
into account differences in hours of operation. The commenter
(A-90-19: IV-D-85) maintained that the "volume discount
component" was needed when the number of emission points
involved in a trading scheme increases. The commenter
(A-90-19: IV-D-85) contended that a volume discount factor
will focus industry efforts on the points offering real
savings instead of simply using the rule's emissions averaging
provisions to avoid enforcement of emission standards. The
commenter (A-90-19: IV-D-85) summarized the suggested
discount factor in the following equation:
2-79
-------
D = D(l) + P + D(w) + D(v) + D(tr) ••• D(st) + KV
where:
D = The overall discount factor.
D(l) = The discount factor component based on marginal cost
differentials.
P A percentage factor to adjust the rates between the
highest and lowest producing process trains in the
average.
D(w) = A factor compensating for uncertainties of
wastewater emissions estimation.
D(v) = An uncertainty factor for vents.
D(tr) = An uncertainty factor for transfer racks.
D(st) = An uncertainty factor for storage vessels.
K = Constant.
V = The number of emission points in the average greater
than two.
Response; The concerns embodied in the commenter's
suggestions have been addressed elsewhere in the rule so that
the commenter's recommended methodology for developing a
discount factor is unnecessary. The EPA acknowledges the
basis for the three areas of concern: sharing cost savings
with the environment, minimizing gaming, and limiting the
number of points involved in averages. The main rationale for
the selection of 10 percent for the discount factor is that it
represents a reasonable portion of the cost savings to share
with the environment without discouraging the use of emissions
averaging. The potential for gaming is minimized through the
use of consistent emissions estimation techniques, which is
discussed further in section 2.11.4 of this BID volume.
Finally, as explained in section 2.8.5 of this BID volume, a
provision has been added to the final rule limiting the number
of points allowed in averages to 20, or 25 if pollution
prevention is used.
Hence, the commenter's concerns have been addressed
without adding tremendous complexity to the administration of
the emissions averaging program by requiring a discount factor
calculation. The EPA considers its selection and use of a
single value for the discount factor a simple but effective
means of sharing some cost savings with the environment.
2-80
-------
Comment; Two commenters (A-90-19: IV-D-90; IV-D-100)
maintained that the discount factor range of 0 to 20 percent
does not take into account the potential interactive effects
from exposure to chemical mixtures and potential for
underestimating public health impacts.
Response; The EPA has concluded that a discount factor
is not the appropriate mechanism for accounting for health
risks because of the potential complexity that such a
mechanism could introduce. Moreover, the issue of health risk
that might be posed by emissions averaging has been addressed
by including a new provision that sources must demonstrate to
the satisfaction of the implementing agency that their
emissions average will not increase risk or hazard.
Discussion of whether and how risk should be taken into
account in emissions averaging can be found in section 2.9 of
this BID volume. Specific discussion of the new provision in
the final rule for making risk equivalency demonstrations is
included in section 2.9.6 of this BID volume.
2.7 COMPLIANCE PERIOD
2.7.1 Averaging Period
Comment; Several commenters (A-90-19: 'IV-D-32;
IV-D-33; IV-D-57; IV-D-58; IV-D-62; IV-D-72; IV-D-73 and
IV-G-11; IV-D-74; IV-D-79; IV-D-82; IV-D-83 and IV-F-1.3 and
IV-F-5; IV-D-86; IV-D-92; IV-D-106; IV-D-108; IV-F-1.6 and
IV-F-6; IV-G-1; IV-K-2; IV-K-6; IV-K-7; IV-K-14; IV-K-19;
IV-K-20; IV-K-21; IV-K-25; IV-K-27; IV-K-33; IV-K-34; IV-K-35;
IV-K-39; IV-K-42; IV-K-45; IV-K-47; IV-K-49; IV-K-50; IV-K-53;
IV-K-56; IV-K-61; IV-K-62; IV-K-66) on the proposed rule and
supplemental notice supported an annual compliance period for
balancing averages.
Four commenters (A-90-19: IV-K-19; IV-K-21; IV-K-39;
IV-K-66) argued that an annual period was needed for source
flexibility. Several commenters (A-90-19: IV-D-58; IV-D-73
and IV-G-11; IV-D-74; IV-D-83 and IV-F-1.3 and IV-F-5;
IV-D-108; IV-G-1; IV-K-6; IV-K-19; IV-K-21; IV-K-42; IV-K-56;
IV-K-62; IV-K-66) promoted it as necessary to accommodate
fluctuations in operational processing and production levels
2-81
-------
at facilities. One commenter (A-90-19: IV-G-11) provided a
record of the monthly production of one SOCMI chemical at one
of their plants to illustrate fluctuation in production rate
and support this claim. Two commenters (A-90-19: IV-K-33;
IV-K-66) advised that annual periods are enforceable, and that
batch processes, which result in variable emissions, must have
a longer compliance period than those proposed.
Two commenters (A-90-19: IV-D-32; IV-D-57) claimed
that it would be impossible in many situations to compute
debits and credits over periods shorter than 30 days. Four
commenters (A-90-19: IV-D-32; IV-D-57; IV-D-58; IV-D-92)
claimed that because of,variability in operating conditions
and rates, a compliance period of significantly longer than
30 days is required to make averaging a practical option. One
commenter (A-90-19: IV-G-11) was concerned that because
credits could decline during periods of lower production (and
lower emissions) at the credit-generating process unit,
unplanned decreases or stoppage of production could cause
violation of the average with a shorter averaging period. The
commenter (A-90-19: IV-G-11) added that on the other hand, a
longer averaging period provides an opportunity for a source
to take steps to increase credits or decrease debits to regain
the required balance.
Two commenters (A-90-19: IV-D-32; IV-D-57) considered
it well within the EPA's discretion to establish quarterly and
annual compliance periods because section 112(d) of the Act
says nothing about the period over which the required
emissions reductions must be achieved and demonstrated.
In contrast, several commenters (A-90-19: IV-D-9;
IV-D-41; IV-D-49; IV-D-51; IV-D-70; IV-D-85 and IV-F-7.39 and
IV-F-12 and IV-G-6; IV-D-87; IV-D-99; IV-D-115; IV-D-117 and
IV-F-7.43; IV-D-118; IV-D-124; IV-D-125; IV-F-1.5; IV-F-7.2;
IV-F-7.29; IV-F-7.36) criticized the compliance periods for
averaging in the proposed rule and the supplemental notice as
too long.
One commenter (A-90-19: IV-D-85) stated that the
annual compliance period reflects an intention to accommodate
2-82
-------
rather than limit emissions and suggested that it sends the
wrong message to toxic polluters. Two commenters (A-90-19:
IV-D-87; IV-D-99) were concerned about the enforcement and
administrative problems that a long period could cause. Two
commenters (A-90-19: IV-D-87; IV-D-115) stated that the
quarterly and annual averaging periods do not meet the
criteria for "federally enforceable" because these are not
"the shortest practicable time periods" and exceed 30 days.
One commenter (A-90-19: IV-D-51) stated that an
averaging period for compliance not longer than quarterly was
reasonable. Four commenters (A-90-19: IV-D-49; IV-K-10;
IV-K-30; IV-K-44) endorsed a monthly averaging period. One of
the commenters (A-90-19: IV-D-49) suggested that a compliance
period no longer than 30 days was consistent with the
statement in the proposal preamble that a 30-day compliance
period could reasonably be applied to all the kinds of points
that can be included in averages.
Response; The compliance period for averaging that was
proposed, an annual period with quarterly checks, has been
maintained for the final rule. Allowing averaging over a
year's time instead of just one quarter provides flexibility
for sources whose production rates vary over time. The
additional requirement that debits cannot exceed credits by
more than 30 percent in any one quarter should assure that
wide-ranging fluctuations in HAP emissions will not occur.
The EPA concurs with commenters that a shorter
averaging period than annual would preclude the use of some
emission points in averages. An annual period allows
inclusion of points that: (1) do not have the same emission
rates during some periods of the year; and (2) must undergo
temporary maintenance shutdowns at different times during the
year. Hence, an annual period provides sources the necessary
latitude to construct the most cost-effective averages.
Moreover, the EPA considers it within their authority under
the Act to establish the averaging period as any length that
can be demonstrated to be enforceable.
2-83
-------
The EPA is satisfied that the annual period will not
pose any significant enforcement and administrative problems.
As explained in section 2.7.2 of this BID volume, it is true
that the annual averaging period could reduce the EPA's
ability to use administrative enforcement actions. However,
the requirement of a quarterly emissions check enables use of
the administrative enforcement mechanism and allows more
frequent enforcement than just once a year. Judicial
proceedings can also be undertaken against sources violating
the annual average or the quarterly check.
Some commenters were mistaken in citing criteria for
Federal enforceability. The requirement for employing the
shortest practicable time period relates to monitoring and
does not apply to the compliance period for averaging. A more
complete explanation of the difference between compliance
periods for monitoring and emissions averaging is provided in
the response to the fourth comment in this section.
The EPA acknowledges that a 30-day averaging period can
be applied to all the kinds of emission points subject to the
rule. However, as just discussed, a 30-day period is simply
not workable because it would preclude averaging of points in
processes with even slightly different production and
maintenance schedules, thus discouraging averaging and
decreasing the emission reduction benefits and cost savings
that can be gained from averaging.
Comment: Two commenters (A-90-19: IV-K-9; IV-K-37) on
the supplemental notice supported a quarterly block averaging
period, asserting that it would be manageable and enforceable.
Four commenters (A-90-19: IV-K-1; IV-K-18; IV-K-30; IV-K-44)
concluded that a quarterly block averaging period was the best
of the options proposed in the supplemental notice, although
they preferred the elimination of emission averaging from the
HON regulation.
Two commenters (A-90-19: IV-K-34; IV-K-46) supported
the option of a quarterly block averaging period with banking
for up to one or two additional quarters.
2-84
-------
Four commenters (A-90-19: IV-K-28; IV-K-29; IV-K-52; •
IV-K-54) supported a semiannual block averaging period with
banking for an additional six months. Two commenters
(A-90-19: IV-K-21; IV-K-25) preferred the semiannual block
averaging period with banking of the four options given,
although they preferred an annual averaging period.
However, several commenters (A-90-19: IV-K-2; IV-K-14;
IV-K-20; IV-K-27; IV-K-34; IV-K-35; IV-K-39; IV-K-45; IV-K-47;
IV-K-49; IV-K-62) preferred the approach originally stated in
the proposed rule, of an annual compliance period with
quarterly checks. Six of the commenters (A-90-19: IV-K-27;
IV-K-34; IV-K-35; IV-K-39; IV-K-49; IV-K-50) argued that such
an approach was both flexible and enforceable. Four
commenters (A-90-19: IV-K-27; IV-K-39; IV-K-49; IV-K-50)
argued that it was consistent with the compliance periods in
regulations under titles I and IV of the Act. Two commenters
(A-90-19: IV-K-2; IV-K-47) claimed that the annual period
would allow emissions to fluctuate, and the quarterly check
would ensure that debits and credits balance. Two more
commenters (A-90-19: IV-K-2; IV-K-35) reasoned that changing
the proposal provisions would limit emissions averaging.
One commenter (A-90-19: IV-K-22) stated that an annual
averaging period was not necessary, and that the EPA should
instead set an emissions cap based on the maximum allowable
emissions for the aggregate of the emission points being
averaged. Another commenter (A-90-19: IV-K-30) asserted that
the EPA should evaluate other options than those proposed in
the supplemental notice.
Response: For the reasons explained in the previous
response, an annual averaging period with quarterly checks was
selected as the most appropriate compliance period for the
final rule. Proponents of the other options proposed in the
supplemental notice did not provide a sufficient explanation
of how a period other than annual would better address the
concerns for emissions averaging compliance. One commenter's
recommendation to evaluate options other than those in the
supplemental notice could not be followed as the commenter did
2-85
-------
not suggest any specific types of compliance periods to
consider.
One commenter's suggestion of an emissions "cap" based
on maximum allowable emissions is not appropriate for
standards under section 112(d) of the Act, which should set
emission limitations that are based on use of the maximum
achievable control technology. These standards can take
various forms such as percent reductions, concentration
levels, or emissions per unit of production, for example.
However, an emissions cap in units of mass (e.g., total
megagrams over some time period) is not consistent with
section 112(d) because a cap would limit a source's production
rate and prohibit expansions. The intent of NESHAP is to
require the best controls, not to limit production. Moreover,
even if a cap were allowed, a compliance period must still be
established for enforcement purposes.
Comment: Three commenters (A-90-19: IV-D-49; IV-D-70;
IV-D-85 and IV-F-7.39 and IV-F-12) argued that lengthy
averaging times would allow increased peak emissions of some
pollutants with serious health effects and increase annual
emissions. Nine commenters (A-90-19: IV-K-1; -IV-K-10;
IV-K-17; IV-K-30; IV-K-41; IV-K-44; IV-K-55; IV-K-63; IV-K-64)
endorsed even shor-3r averaging periods than what was proposed
in the supplemental notice, maintaining that they were
necessary to protect the public health.
Seven commenters (A-90-19: IV-F-7.2; IV-F-7.29;
IV-F-7.35; IV-F-7.44; IV-D-85 and IV-F-7.39 and IV-F-12;
IV-D-99; IV-D-120) were concerned that emissions averaging
would allow peak exposures because plants can maximize
exposures in a short period of time as long as they average
over a long period with low exposures. One commenter
(A-90-19: IV-F-7.2) suggested that the whole idea of allowing
credits is under attack nationally, in programs such as the
acid rain program because of potential peak exposures. One
commenter (A-90-19: IV-D-49) suggested that the EPA assess
whether the lengthy averaging times will increase health risk.
2-86
-------
Two commenters (A-90-19: IV-D-90; IV-D-100) stated
that a three-month averaging time would not provide sufficient
short-term data to evaluate the potential cancer and non-
cancer health effects associated with exposure to emissions
from the facility, which is an assessment required by State
programs and the EPA. One commenter (A-90-19: IV-K-30)
cautioned that even a monthly averaging period was too long to
ensure public safety from exposure to hazardous pollutants.
One commenter (A-90-19: IV-K-55) stated that three compliance
periods, hourly, daily, and annual, were needed to assess the
short- and long-term health impacts from exposure to HAP's.
Response; The EPA reemphasizes that neither the
averaging period nor any other emissions averaging provision
will allow increases in annual emissions compared to
compliance without averaging. The requirement of a 30-percent
quarterly emissions check is intended to prevent exposures to
peaks of HAP emissions from occurring during the annual
averaging period. Furthermore, there are other mechanisms to
protect against peak releases. Malfunction plans and
reporting of malfunctions are required by the General
Provisions. Additionally, in the event that an accidental
release occurs, the source will be subject to the proposed
accidental release prevention rule,(58 FR 5702; January 19,
1993) .
The EPA took health risk into account by requiring a
quarterly check along with the requirement that debits and
credits balance annually. The EPA does not consider it
necessary to perform a formal assessment of the averaging
period's effect on health risk, or to account for the
averaging period in the risk equivalency demonstrations now
required by the final rule (discussion of this new requirement
is located in section 2.9.6 of this BID volume). However, if
a State takes the time period into account in their own risk
assessment methodologies, they are free to continue
considering it in the hazard or risk equivalency
demonstration.
2-87
-------
The claim that quarterly reporting of emission debits
and credits provides insufficient short-term exposure data to
evaluate health effects is not relevant to emissions
averaging. Neither the proposed approaches for toxicity
weighting (which were not adopted) nor the hazard or risk
equivalency demonstration now required depend on short-term
emissions or health effects data gathered after an average is
approved and in effect. The commenters also stated
incorrectly that hazard or risk assessments are required by
the EPA; the EPA does not require them.
Comment; One commenter (A-90-19: IV-D-87) stated that
the compliance period should be set at a minimum period such
as hourly and daily, instead of quarterly and annual. Another
commenter (A-90-19: IV-D-117 and IV-F-7.43) submitted that
the shorter the averaging time, the more stringent and
efficient the controls will be, and suggested that continuous
monitoring technology can measure an hourly or a rolling
15-minute average. One commenter (A-90-19: IV-D-49)
contended that any averaging program must require monitoring
periods achievable by the most technically advanced monitoring
equipment currently available.
Five commenters (A-90-19: IV-K-1; IV-K-10; IV-K-17;
IV-K-41; IV-K-63) supported a 1-hour averaging period,
asserting that even short-term exposures to HAP's can pose
substantial risk. One of the commenters (A-90-19: IV-K-1)
counselled that anything less than a 1-hour averaging period
for process vents would violate the intent of the Act, as more
than 12 percent of the industry currently achieve reductions
on an hourly or continuous basis, and the EPA must make sure
that it is proposing the shortest achievable compliance
period. The commenter (A-90-19: IV-K-1) also stated that
compliance periods for other kinds of emission points should
be the shortest that are achievable.
One commenter (A-90-19: IV-D-85 and IV-G-6) noted that
the proposal preamble does not explain why a compliance period
stricter than 30 days could not be applied, and therefore,
argued that the EPA should require hourly balancing of credits
2-88
-------
and debits calculated by computer and reported on a monthly
basis for most emission points. The commenter (A-90-19:
IV-D-85) stated that the annual compliance period will not
produce reductions equal to those achievable through point-by-
point compliance, because the total yearly emissions of a
source complying with a standard on an hourly or daily basis
will generally be less than a source complying with an
obligation to comply on only a yearly or quarterly basis.
Response: Some commenters seem to have mistaken the
compliance period for balancing averages with the period over
which the operating parameters used to calculate emissions are
measured. It was determined that regardless of the use of
emissions averaging, calculation of hourly emissions and
continuous emission monitoring are not technically feasible
for this rule, as discussed in sections 2.3 and 3.2.4,
respectively, of BID Volume 2E. Instead, in almost all cases, •
the rule requires that operating parameters that reflect the
effectiveness of a control device (rather than emissions) be
monitored every 15 minutes. In addition, the daily average of
the operating parameter data is reported if it is outside a
specified range. Hence, daily compliance is the shortest
period required in the rule, even for points that are not in
an emissions average.
This same parameter monitoring and reporting of daily
average values outside their ranges is also required for
points in averages, and if such excursions are not excused,
they are considered violations of permitted operating
conditions (for a more detailed discussion of excursions, see
section 3.2.5 of BID Volume 2E). However, to also require
daily balancing of debits and credits would be so restrictive
as to render averaging useless.
The EPA discussed at proposal why an averaging period
shorter than monthly could not be applied. The control and
monitoring equipment available has only limited ability to
distinguish short-term fluctuations in emissions from some
kinds of points, such as transfer racks and storage vessels.
Furthermore, emissions from these two kinds of emission points
2-89
-------
vary daily depending on factors such as temperature and
loading schedule. Compliance periods shorter than monthly
would preclude their inclusion in averaging. Hence, it was
concluded that 30 days was the shortest averaging period that
could reasonably be applied to all the kinds of points that
can be included in averages. The reasons why an annual
compliance period with quarterly checks was chosen instead of
a monthly or quarterly compliance period are explained in the
first response in this section.
An annual compliance period for averaging will not
affect the emission reductions that will be achieved compared
to point-by-point compliance. Emission points included in
emissions averages must comply with the same monitoring
requirements as the other points in the source, including
taking data samples every 15 minutes and reporting excursions.
If a debit or credit generator emits more than what was
planned, the source will be in violation unless the average is
balanced with more credits or fewer debits from other points.
Comment; Five commenters (A-90-19: IV-D-9; IV-D-85
and IV-F-7.39 and IV-F-12; IV-D-118; IV-D-124; IV-D-125)
considered it unacceptable to greatly lengthen the averaging
times beyond current State practice for compliance with
emissions limitations. As an example, one commenter (A-90-19:
IV-D-85 and IV-F-7.39 and IV-F-12) contended that under
ordinary State smog control rules, process vents are subject
to averaging times of one hour. Four commenters (A-90-19:
IV-D-9; IV-D-118; IV-D-124; IV-D-125) suggested that averaging
times for MACT standards for each emitting unit should be at
least as strict as the most stringent state VOC rules.
One commenter (A-90-19: IV-D-85) claimed that because
all process vents can comply with an hourly emission limit,
they could comply with a reduction requirement on an hourly
basis, and thus, the EPA must require hourly balancing. The
commenter (A-90-19: IV-D-85) added that emission points that
cannot comply with an hourly emission limitation should not be
included in averages with vents. The commenter (A-90-19:
IV-D-85) argued that allowing such trades would relax the
2-90
-------
stringency of the standard through a relaxation of the
cumulative averaging time.
Response: As discussed in the previous response, the
!
compliance period for averaging should not be confused with
the period over which operating parameters are monitored to
ensure a control device's effectiveness.
Previous NSPS and CTG's for VOC emissions from process
vents required controls to achieve 98 percent reduction on a
3-hour basis. Compliance was determined by an initial
performance test (conducted over a 3-hour period), and
continuous parameter monitoring was required. State VOC rules
may use similar compliance procedures. However, such State
rules are developed under different programs with different
goals from the federal NSPS and NESHAP programs.
The HON compliance approach is generally consistent
with the previous NSPS for process vents, and results in the
use of MACT. Process vents in emissions averages (as well as
Group 1 vents that are not in averages) must perform an
initial test over a 3-hour period to demonstrate the control
efficiency achieved, and then must apply continuous parameter
monitoring to ensure the control device's effectiveness. The
parameter monitoring results are summarized on a daily basis
for compliance determinations for reasons explained in
section 2.3.1 of BID Volume 2E. These monitoring requirements
are the same for both averaged and non-averaged points. Thus,
the EPA is not relaxing the standards or its monitoring
requirements to accommodate averaging.
Comment; One commenter (A-90-19: IV-D-85) suggested
that for any compliance period, the EPA should either state
when the period begins and ends or require the source to
choose a period in advance and stick to it. The commenter
(A-90-19: IV-D-85) was concerned that otherwise, sources
could play "games," claiming that they were in compliance with
respect to a period figured from a starting date selected
after the fact to evade enforcement.
Response: For the source that intends to use emissions
averaging immediately, the compliance period of the average
2-91
-------
will begin on the same date that the source as a whole must
comply with subpart G of the rule. That date is defined in
§63.100(k)(2) of subpart F of the final rule to be no later
than three years after the date the rule is published in the
Federal Register. The periodic (quarterly) reporting
provisions in §63.152 of subpart G clarify that the first
quarter begins on the compliance date, and the second quarter
would begin when the first quarter ends, etc.
If the owner or operator of a source decides to use
emissions averaging after the source has begun complying with
the rule, the change must be made as a permit amendment
subject to all review and comment provisions. If the source
is operating under an Implementation Plan instead of an
operating permit, the plan must be updated and approved
according to the procedures in §63.151(i) of the rule. In
either case, the date that the compliance period for the
emissions average begins will be established when the
operating permit amendment or Implementation Plan update is
approved. Of course, until the emissions average goes into
effect, the points in the average must have been in compliance
with the rule on a point-by-point basis.
2.7.2 Preclusion of Administrative Enforcement
Comment: Six commenters (A-90-19: IV-D-9; IV-D-85 and
IV-F-7.39 and IV-F-12; IV-D-117 and IV-F-7.43; IV-D-118;
IV-D-124; IV-D-125) warned that an annual averaging period may
preclude administrative enforcement, because violations more
than a year old cannot be enforced through the Act's
administrative enforcement mechanism. One commenter (A-90-19:
IV-D-85) reasoned that because the EPA considers
administrative enforcement a relatively inexpensive
enforcement tool (as stated in the proposal preamble),
constructing a scheme that makes use of administrative
enforcement difficult conflicts with the Congressional intent
to make it available.
Three commenters (A-90-19: IV-D-32; IV-D-57; IV-D-86).
disagreed and suggested that an annual compliance period would
not unduly impede the EPA's ability to enforce the standard.
2-92
-------
One commenter (A-90-19: IV-D-32) predicted that the only
potential area in which the EPA's enforcement authority could
be limited would be the imposition of administrative penalties
under section 113(d) of the Act, which imposes a one-year
statute of limitations, because such a penalty action
instituted following a report of a compliance failure could
not seek penalties for the entire year. The commenter
(A-90-19: IV-D-32) suggested, however, that if the EPA
initiated the penalty action reasonably promptly, the period
of "lost" penalties would be relatively insignificant. The
commenter (A-90-19: IV-D-32) also stated that the compliance
periods will remain fully enforceable under sections 113(b)
and (c) of the Act, which provides for civil and criminal
penalties with a longer statute of limitations.
Response: As stated at proposal, the EPA recognized
that an annual averaging period could limit its authority to
take administrative enforcement actions because under
section 113(d) of the Act, assessment of administrative
penalties is limited to violations that occur no more than
12 months prior to the initiation of the administrative
proceeding. This concern was one of the reasons that a
quarterly emissions check was proposed in addition to the
annual period and included in the final rule. The quarterly
check enables the EPA to use its administrative enforcement
authority by providing a shorter period in which to verify
compliance. Further details of the quarterly check are
discussed in the next section. The commenter is also correct
in stating that the one-year statute of limitations does not
apply to judicial proceedings for civil and criminal
penalties.
2.7.3 Quarterly Emissions Check
Comment: Several commenters (A-90-19: IV-D-32;
IV-D-33; IV-D-56; IV-D-58; IV-D-62; IV-D-72; IV-D-74; IV-D-75;
IV-D-79; IV-D-83 and IV-F-1.3 and IV-F-5; IV-D-92; IV-D-106;
IV-F-1.1 and IV-F-3; IV-F-1.6 and IV-F-6; IV-F-1.7; IV-G-1)
concurred with the EPA that a quarterly emissions check is
reasonable.
2-93
-------
One commenter (A-90-19: IV-F-l.l and IV-F-3) warned
that with a shorter time frame than quarterly, if credit-
generating points are shut down for a period of time, a source
might be out of compliance even though total emissions from
the facility may be lower than if the credit generator had
been operating. Two commenters (A-90-19: IV-D-32; IV-D-74)
suggested that the quarterly emission limitation should allay
concerns that an annual compliance period would allow high
emissions for some shorter period.
Three commenters (A-90-19: IV-D-72; IV-D-106; IV-G-1)
supported a 35 percent quarterly debit-to-credit ratio limit.
One commenter (A-90-19: IV-K-6) advocated that debits be
allowed to exceed credits by at least 35 percent. Another
commenter (A-90-19: IV-K-62) stated that if the compliance
period were shortened from what was proposed originally,
debits should be allowed to exceed credits by 100 percent
instead of 35 percent.
Response; The EPA appreciates commenters1 support for
establishing an additional quarterly emissions check to enable
the use of administrative enforcement and to preclude the
possibility of peak HAP emissions. The requirement that
debits not exceed credits by more than 30 percent in any
quarter has been included in the final rule. A range of 25 to
35 percent was proposed for the amount of debit exceedance to
be allowed in any quarter. The midpoint of the proposed
range, 30 percent, was selected as a way of balancing industry
concerns about operational flexibility with other concerns
about protection from peak emissions.
Comment: One commenter (A-90-19: IV-D-78) was opposed
to a quarterly limitation as long as the source meets the
annual limitation and recommended that only quarterly
reporting should be required. The commenter (A-90-19:
IV-D-78) suggested that if quarterly debit-to-credit ratios
must be limited, the EPA should at a minimum- allow use of
banked credits for quarterly compliance.
One commenter (A-90-19: IV-D-89) submitted that a
limitation on the usage of banked credits not to exceed 25 to
2-94
-------
35 percent quarterly is counterproductive and disadvantageous
to industry. The commenter (A-90-19: IV-D-89) claimed that
it establishes a dual set of limitations, which discourages
ongoing efforts to control other emissions. The commenter
(A-90-19: IV-D-89) complained that the quarterly limitation
assumes that industry will make only a one-time or infrequent
effort to control emissions beyond regulatory requirements,
which may not be the case nor should it be encouraged by the
EPA. The commenter (A-90-19: IV-D-89) argued that because
the rule requires monthly emissions averaging records,
establishing a quarterly limitation as a shorter period to
enable the EPA to verify compliance is not important, and the
documentation associated with this and other compliance
limitations places a great burden on industry and the
reviewing agency.
Response; The EPA is satisfied that establishing a
dual compliance period of annual averaging with quarterly
emissions checks is justified. One commenter's recommendation
of requiring only quarterly reporting has already been met;
Periodic Reports for emissions averages must be submitted
every quarter. However, the commenter's further suggestion of
allowing the use of banked credits to meet quarterly
compliance is not appropriate. For reasons discussed in
section 2.10 of this BID volume, credit banking is not allowed
in the final rule. Moreover, allowing the use of banked
credits to meet the quarterly requirement runs counter to one
main reason for establishing the requirement: to preclude the
possibility of peaks of HAP emissions.
The commenter opposed to a quarterly limit on the use
of banked credits appears to have misinterpreted the proposed
rule. The commenter is correct that the provisions establish
a dual compliance period, but did not explain why such a dual
limit should discourage control of other emissions. The EPA
does not assume that any source's controls will be one-time or
infrequent and predicts that emission reductions will occur
continuously throughout the averaging period.
2-95
-------
However, many commenters have cited the likelihood that
operational variability or different maintenance schedules on
points in an average can lead to short-term periods when
debits and credits do not balance, despite the fact that
credits outweigh debits on an annual basis. Other commenters
have expressed concern that peak releases of HAP's could occur
and could have health impacts. Allowing annual averaging with
quarterly checks accommodates operational variability, but
prevents wide-ranging fluctuations in HAP emissions over time.
A quarterly check of 30 percent debit exceedances strikes a
reasonable balance between operating flexibility and
protection from peak emissions.
The frequency of recordkeeping does not substitute for
establishing a period for verifying compliance or for guarding
against peak emissions. The justification for requiring
monthly records and discussion of the burden associated with
this requirement can be found in sections 2.3 and 2.5,
respectively, of BID volume 2E. The only bearing
recordkeeping frequency has on the compliance period issue is
whether sufficient data will be available to verify
compliance, and monthly recordkeeping is entirely suitable for
the averaging periods that have been established.
Comment; Five commenters (A-90-19: IV-D-51; IV-D-90;
IV-D-99; IV-D-100; IV-D-115) opposed allowing emission debits
to exceed credits by 25 to 35 percent in a quarter.
Three commenters (A-90-19: IV-D-90; IV-D-100;
IV-D-115) contended that debits should never be allowed to
exceed credits. One commenter (A-90-19: IV-D-85) argued that
the EPA should not allow 25 to 35 percent exceedances from its
standards because doing so violates the maximum emissions
reduction achievable standard. One commenter (A-90-19:
IV-D-85 and IV-F-7.39 and IV-F-12) complained that a quarterly
limit permitting administrative enforcement when sources
produce more than 25 percent in extra emissions does not
provide an adequate check because sources that stay within the
quarterly limit, but violate the annual limit might be immune
from administrative enforcement. Two commenters (A-90-19:
2-96
-------
IV-D-90; IV-D-100) stated that the 25 to 35 percent exceedance
was in direct conflict with permit conditions which require
the owner or operator to comply with emission standards.
One commenter (A-90-19: IV-D-115) contended that the
25 to 35 percent debit exceedance provision makes it
impossible for an inspector to determine whether or not a
source is in compliance. The commenter (A-90-19: IV-D-115)
stated that any exceedances should be reported and reviewed by
the administering agency.
Three commenters (A-90-19: IV-D-51; IV-D-99; IV-D-115)
opposed debits exceeding credits by 25 to 35 percent because
of concern that owners or operators would accumulate so much
"debt" they would not be able to comply with the annual
average. Two of the commenters (A-90-19: IV-D-51; IV-D-99)
stated that hence, debits should not be allowed to
significantly exceed emission credits. One commenter
(A-90-19: IV-D-51) suggested that the permitting authority be
informed through a facility permit or Implementation Plan of
how a source intends to comply with the MACT standard in order
to prevent owners or operators from accumulating too much
emissions "debt."
Three commenters (A-90-19: IV-D-51; IV-D-99; IV-D-115)
opposed debits exceeding credits by 25 to 35 percent because
of concern for public health impacts caused by short-term
exposures. One commenter (A-90-19: IV-D-87) stated that
emission debits should not exceed emission credits at any time
without an evaluation of the impact to potential adverse
effects to human health and the environment.
Response: The commenters1 suggestion not to allow
debit exceedances in any quarter would be tantamount to
establishing a quarterly averaging period, which would
restrict flexibility too severely as discussed in
section 2.7.1 of this BID volume. The provision for a
30-percent quarterly debit exceedance does not violate the
maximum emissions reduction standard because the source must
achieve the same or greater emission reductions on an annual
basis as it would under point-by-point compliance and make the
2-97
-------
same annual compliance certification. Furthermore, controls
applied to points in emissions averages are subject to the
same requirements for continuous monitoring to assure proper
operation of control technology as other emission points.
It is possible that a source could always meet the
quarterly limit, but not comply annually and still avoid
administrative penalties. However, noricompliance with the
annual limit is the more serious violation, invoking much more
substantial penalties than the administrative ones. Judicial
proceedings could be undertaken in such a situation. The
concern over conflicts with permit conditions is also
unfounded as emissions averaging (and the rule as a whole) is
consistent with the operating permit program rule. Both the
quarterly and annual limits can be incorporated as enforceable
requirements in operating permits.
The quarterly check will not impair compliance
inspections as suggested. Typically, an inspection is used
primarily to ensure that control devices are operating as
specified in the operating permit. In addition, monthly
records of debit and credit calculations would also be
available during an inspection. Finally, the quarterly check
will be demonstrated in the Periodic Report. As long as a
source's debit and credit calculations are acceptable
(i.e., the control devices have been operating correctly, the
values in the emissions estimation equations are accurate, and
operating rates were as planned) and debits do not exceed
credits by more than 30 percent, the source would be
considered to be in compliance for the quarter.
Regarding the concern for sources accumulating too much
debit exceedance, allowing the flexibility of quarterly
exceedances does not absolve the source of its responsibility
to comply with the annual average. As stated previously,
annual noncompliance is the most serious violation carrying
the most severe penalty, which should deter sources from
accumulating too much "debt."
A quarterly check was incorporated into the compliance
scheme out of concern for public health and short-term
2-98
-------
exposures. As stated previously, the quarterly limit will
protect against emission peaks so that potential health and
welfare effects are avoided. The 30-percent differential
between debits and credits should not result in a significant
increase in emissions from a plant during any given quarter
because only the net emissions from the few points in the
average would increase.
2.7.4 Alternative Proposal for Quarterly Limit
Comment: Several commenters (A-90-19: IV-D-32;
IV-D-33; IV-D-56; IV-D-58; IV-D-62; IV-D-73 and IV-G-11;
IV-D-74; IV-D-75; IV-D-79; IV-D-83 and IV-F-1.3 and IV-F-5;
IV-D-86; IV-D-92; IV-F-1.6 and IV-F-6; IV-F-1.7) expressed
support for the industry proposal identified in the preamble,
which would establish a quarterly emissions check based on the
source's allowable emission levels. One commenter (A-90-19:
IV-D-83 and IV-F-1.3 and IV-F-5) recommended a quarterly cap
of 35 percent of total annual allowable emissions as
sufficient to provide adequate protection against potential
short-term adverse air quality impacts. Two commenters
(A-90-19: IV-D-32; IV-D-73) included separate attachments
illustrating examples of the industry fixed cap approach.
Two commenters (A-90-19: IV-D-58; IV-D-62) stressed
that the cap should be applied to allowable emissions as set
in a source's operating permit. Two commenters (A-90-19:
IV-D-32; IV-D-73) supported the alternate "fixed cap" approach
for a quarterly limitation claiming that it satisfies the
intended purposes of the quarterly compliance requirement by
precluding short-term spikes in emissions. Four commenters
(A-90-19: IV-D-32; IV-D-62; IV-D-74; IV-D-92) suggested that
the alternate approach provides for further reductions because
it avoids situations under which an emission point is operated
simply to generate needed credits. Three commenters (A-90-19:
IV-D-32; IV-D-73; IV-D-75) also supported it because it does
not cause a source to be in violation if a credit-generating
operation is unavoidably curtailed for some part of the
quarter.
2-99
-------
Two commenters (A-90-19: IV-D-58; IV-D-75) preferred
the industry proposal because it gives sources more certainty
as to what the allowable emissions are for a fixed period and
allows for an easier compliance determination by both the
source and the State. One of the commenters (A-90-19:
IV-D-58) warned that under the proposed approach for a
quarterly check, a quarterly limit on the debit-to-credit
ratio could be less than or greater than allowable emission
limits included in permits depending on the circumstance of
the averaging program, and could be a conflicting compliance
requirement.
One commenter (A-90-19: IV-D-85) opposed the industry-
proposed alternative for a quarterly emissions limitation.
The commenter (A-90-19: IV-D-85) suggested that an emissions
cap that excludes consideration of the emissions from
uncontrolled Group 1 points is even less defensible than the
debit-to-credit ratio proposal. The commenter (A-90-19:
IV-D-85) asserted that the industry-proposed quarterly
limitation bears no discernible relationship to the emissions
level sought to be achieved.
Response: The EPA did not adopt the industry-proposed
alternative for the quarterly emissions check because of
concerns about an absolute emissions limit based on
projections. Operating levels for calculating allowable
emissions are based on representative predictions of realistic
operating scenarios. The use of such a system creates an
incentive to "game," i.e., to project higher operating rates
for credit-generating points than is representative or
realistic. In contrast, the quarterly check included in the
final rule depends on the actually demonstrated operating rate
during the quarter, not projections.
Under the industry-proposed alternative, it would make
no difference whether the emissions from a debit generator
increase or the emissions from a credit generator decrease; as
long as the total emissions are below the cap, the facility
remains in compliance. However, in order for a source to be
in compliance on an annual basis, credits from overcontrol
2-100
-------
must equal or exceed debits from undercontrolled points in
order to result in the same or greater emission reductions as
would have occurred under point-by-point compliance. A
quarterly limit on the debit-to-credit ratio is more
consistent with this approach. If the emissions from a debit
point increase and/or the emissions from the credit point
decrease significantly, it could impact whether or not the
facility is in compliance. A large increase of emissions from
a debit generator or decrease in emissions from a credit
generator (i.e., a deviation greater than 30 percent from the
emissions that would have occurred under a point-by-point
compliance) is significant. Therefore, the debit-to-credit
ratio limit represents a better check on potential annual
noncompliance.
The possibility of an emission point being operated
simply to generate needed credits is not of great concern
because it can be demonstrated that there is not much
difference in a source's total emissions whether a credit
generator is operated or shut down. Moreover, the industry-
proposed cap would not shrink along with unexpected decreases
in production, which could allow a much greater exceedance of
debits over credits, resulting in more opportunities for
emission spikes. Or, if production increased dramatically in
one quarter, a source could be significantly out of compliance
on a quarterly basis but could stay in compliance for the
year. However, it would be difficult for implementing
agencies to recognize either of these situations without a
detailed knowledge of both actual and projected production
levels. The selected debit-to-credit ratio limit based on
actual operation allows sources and implementing agencies to
recognize a quarterly violation easily and immediately.
In this rule, the source does not need to know what its
total allowable emissions are in any period because the total
emissions are not limited. The source must either maintain
RCT's properly or ensure that debits are balanced by an equal
number of credits with a leeway of 30 percent each quarter.
The EPA maintains that instead of allowing for an easier
2-101
-------
compliance determination, a system of assigning credits based
on allowable emissions requires a great deal more scrutiny of
the source's prediction of operating levels. As stated
previously, the entire rule is designed to be consistent with
the operating permit program rule. There should be no
conflict between the HON and the operating permit because the
quarterly check, as well as the annual credit/debit balance
and the monitoring requirements will be stipulated as permit
conditions.
2.8 IMPLEMENTATION AND ENFORCEMENT
2.8.1 General Issues
Comment; Nine commenters (A-90-19: IV-D-9; IV-D-41;
IV-D-49; IV-D-70; IV-D-85 and IV-F-7.39 and IV-F-12; IV-D-117;
IV-D-118; IV-D-124; IV-D-125) claimed that the emissions
averaging provisions in the proposal make the HON weak and
unenforceable. Nine commenters (A-90-19: IV-D-41; IV-D-90;
IV-D-99; IV-D-100; IV-D-103 and IV-F-7.5; IV-D-115; IV-D-117
and IV-F-7.43; IV-F-1.5; IV-F-7.21) contended that there could
'~>e seriou:. ^ractical enforcement problems in an averaging
cheme. c commenter (A-90-19: IV-D-99) stated that
recordkeeping and enforcement problems are compounded by the
long averaging period. The commenter (A-90-19: IV-D-99)
added that allowing banking for an extended period (i.e., two
to five years) contributes to recordkeeping and enforcement
problems as well.
Response: The EPA has structured the emissions
averaging provisions to be enforceable. Some aspects of the
proposed rule have been changed to simplify emissions
averaging and its enforcement. For example, banking has been
removed, and averaging is not allowed at new sources or across
source categories. Credits are not allowed for any control
applied prior to 1990. The rule has been clarified to
stipulate that wastev^ter treated in a biological treatment
unit cannot be inclv i in averaging. The total number of
emission points that can be included in an average has been
limited to no more than 20 points or 25 points if pollution
prevention is used. Finally, the change offering the greatest
2-102
-------
administrative ease is that State and local agencies have been
granted the discretion to not include emissions averaging in
their implementation of the HON without having to go through
the delegation process established in the section 112(1) rule.
The rule provides clear mechanisms for enforcement of
averaging. Detailed procedures are prescribed for credit and
debit estimation, and credits must outweigh debits. This
assures that emissions are estimated on a consistent basis and
that emission reductions under averaging will be at least as
great as if all Group 1 points had been controlled. If
credits and debits do not balance, this is a clear and
enforceable violation of the emission standard. Furthermore,
monitoring is required for emission points included in an
emissions averaging. If continuous parameter monitoring
results are outside the established range for more than a
limited number of excused excursions, this is a violation of
the requirements for proper operation, and enforcement actions
can be taken. Provisions have also been added for calculation
of credits and debits during monitoring parameter excursions
to further clarify enforcement as discussed in section 2.8.2
of this BID volume.
Comment: Two commenters (A-90-19: IV-D-90; IV-D-100)
contended that emissions averaging is not cost effective
because of the costs associated with enforcing the provisions.
One commenter (A-90-19: IV-D-85 and IV-F-7.39 and IV-F-12)
predicted that a review of the cost of enforcing "bubbles"
should lead the EPA to abandon emissions averaging in the
proposed rule. The commenter (A-90-19: IV-D-85) provided a
rule-effectiveness study of the aerospace coating industry
conducted by EPA Region 9, which the commenter claimed
concluded that "almost all large sources" operating under
source "bubbles" failed to achieve required emission levels of
control and that the "bubble" was extremely difficult to
enforce. The commenter (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12) further claimed that this study reveals how costly
"bubbles" can be in regulating agency staff time and in lost
emissions reductions, and because the proposed HON is far more
2-103
-------
complex than the Aerospace Coating Standards evaluated in this
study, the EPA should undertake a thorough review of its prior
rule effectiveness studies before including emissions
averaging.
Response: The number of emission points that can be
included in an emissions average is now limited to only
20 points, 25 if pollution prevention is used. Therefore,
review of Implementation Plans and quarterly reports will not
be overly time-consuming for implementing agencies. Because
the HON provides specific equations and procedures for credit
and debit calculations, the review to determine whether
calculations are correct will be relatively straightforward.
The parameter monitoring for emission points in averages is
the same as for other Group 1 emission points. Therefore,
averaging will not increase the burden of reviewing monitoring
results. As explained in the previous response, averaging has
been simplified since proposal, which will reduce the
complexity and, therefore, the cost of enforcement. Averaging
is also designed to ensure equivalent emission reduction to
control of all Group I points, and is structured differently
from previous "bubble" rules. With respect to the study
conducted by Region 9 of the EPA provided by one commenter, a
perceived defect of the aerospace coating rule that thwarted
enforcement efforts was inconsistent emission estimation
methodologies and procedures used by different sources.
Accordingly, the HON carefully prescribes the procedures and
equations that must be used to estimate debits and credits,
and sources may not deviate from their use, which eliminates
inconsistencies.
2.8.2 Monitoring. Recordkeeoina. and Reporting
Comment: Four commenters (A-90-19: IV-D-50; IV-D-59;
IV-D-63; IV-D-71) argued that the monitoring, recordkeeping
and reporting requirements for demonstrating compliance are
overly burdensome and could negate any potential savings from
emissions averaging. One commenter (A-90-19: IV-D-33) urged
the EPA to minimize the monitoring, recordkeeping and
reporting costs for ongoing compliance.
2-104
-------
Response; The EPA recognizes that some additional
monitoring, recordkeeping, and reporting is necessary for
emissions averaging. For example, credits and debits must be
calculated monthly and reported quarterly to ensure that the
required emission reductions are achieved, and Group 2 points
being used to generate credits must apply the same control
device monitoring as Group 1 points. Owners or operators
should take the recordkeeping and reporting requirements into
account when deciding whether to utilize emissions averaging.
The EPA considers the monitoring, recordkeeping, and
reporting requirements to be the minimum necessary to
demonstrate compliance. Prior to and since proposal, the EPA
has considered ways to reduce the general recordkeeping and
reporting burden without sacrificing enforceability. For
example, the proposed and promulgated rules require reporting
of monitored parameter values only when they are outside the
established range. Since proposal, provisions have been added
to §63.151 of subpart G allowing case-by-case requests to use
data compression and other alternative monitoring and
recordkeeping systems that may allow continued use of current
or more cost-effective systems at plants. Another change
allows retention of hourly rather than 15-minute average
values of monitored parameters for days when there is not an
excursion. Other recordkeeping and reporting changes are
described in chapter 2.0 of BID volume 2E. The effect of
these changes will be to reduce the burden for all plants,
including those that utilize emissions averaging.
Comment; Seven commenters (A-90-19: IV-D-9; IV-D-85
and IV-F-7.39 and IV-F-12; IV-D-117 and IV-F-7.43; IV-D-118;
IV-D-124; IV-D-125; IV-F-1.5) considered the emissions
averaging scheme unenforceable because it does not provide for
adequate monitoring. One commenter (A-90-19: IV-D-85)
declared that allowing emissions averaging without adequate
monitoring violates the enhanced monitoring requirements, the
maximum achievable emission reduction standard, and the
Congressional intent to increase, not decrease enforceability
of emission standards.
2-105
-------
Six conunenters (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12; IV-D-90; IV-D-99; IV-D-100; IV-D-103; IV-F-7.6)
claimed that emissions monitoring plays an even more crucial
role in an averaging scheme than under a technology-based
approach. One commenter (A-90-19: IV-D-103) stated that
cross-chemical trading over time would require implementation
of comprehensive monitoring of all chemicals, and that there
is no assurance that the EPA or the private sector can
implement and enforce a complex trading system. One commenter
(A-90-19: IV-D-85) maintained that emissions averaging
increases the monitoring needs compared to the non-averaging
approach because to verify compliance, monitoring must not
only show that the required reduction has been attained at
controlled points, but also that the reductions from these
points exceed the emissions from uncontrolled Group 1 points.
The commenter (A-90-19: IV-D-85) added that monitoring the
uncontrolled emissions from debit-generating points is
critical in emissions averaging because achievement of the
standard is dept ident on proper measurement of the debit as
well as the credit.
Hence, six commenters (A-90-19: IV-D-9; IV-D-85 and
IV-F-7.39 and IV-F-12; IV-D-117; IV-D-118; IV-D-124; IV-D-125)
contended that under the proposal, no monitoring checks the
accuracy of estimates of emissions from uncontrolled, debit-
generating points nor the baseline emissions from the credit-
generating point. One commenter (A-90-19: IV-D-85 and
IV-G-6) asserted that the EPA should bar emissions averaging
wherever the amount of both debits and credits cannot be
adequately monitored.
Response; There are several mechanisms for enforcement
of emissions averaging. Monthly credits and debits must be
calculated based on measured and recorded values for different
parameters depending on the kind of emission point, such as
HAP concentration, flow rate, and monthly operating hours for
process vents and rack throughputs for transfer operations.
Values for some of these parameters (e.g., concentration and
flow) are determined initially rather than measured
2-106
-------
continuously, but the rule requires a re-determination when
process or operating changes are made to a debit or credit
generator that could cause the previously measured values to
be no longer representative. Other values that vary from
month to month, such as operating hours for process vents and
throughput for transfer racks, are recorded for each month,
and the monthly values are used to calculate debits and
credits. These procedures and equations in the rule allow
sufficiently accurate estimation of monthly credits and debits
to determine compliance. If credits do not equal or exceed
debits in a year's time, or if debits exceed credits by more
than 30 percent in any quarter, this is a violation of the
emission standard, and enforcement action can be taken.
Furthermore, the controls applied to most Group 1 and
Group 2 points in an emissions average must be monitored
continuously. If these monitored average parameter values are
outside the established range for more than the allowed number
of excused excursions, this is a violation of the requirements
for proper operation, and enforcement actions can be taken.
Finally, provisions have been added to the final rule to
require conservative estimation of credits and debits during
excursions. These procedures will assure debits are not
underestimated and credits are not overestimated during
monitoring excursions.
The EPA considered emission monitoring, but determined
that it was not technically feasible or necessary to use CEM's
to determine credits and debits. This issue is discussed in
the next response.
Comment: Two commenters (A-90-19: IV-D-85 and IV-G-6;
IV-D-99) recommended that if emissions are averaged, CEM's be
required wherever technically feasible to better ensure that
control operate at the expected levels. One of the commenters
(A-90-19: IV-D-85) considered CEM's even more essential to
emissions averaging than the rule without averaging for the
accurate and reliable measure of emissions and reductions, and
suggested as a comparison that CEM's are essential in the acid
rain program. The commenter (A-90-19: IV-D-85) suggested
2-107
-------
that even if parameter monitoring assures that the control
technology on credit-generating points is working perfectly,
without continuous emissions monitoring, increases in
emissions above estimated values will create undetected
violations of the standard.
Response: The EPA considered various means of
determining credits and debits, and concluded that it is not
technically feasible or necessary to use CEM's. To measure
emissions continuously,, both CEM's to measure HAP
concentrations and continuous flow monitors would be needed at
every emission point. There are no CEM's available for
measurement of some organic HAP's. Where CEM's are available,
they are generally more costly and more complex to calibrate
and operate than operating parameter monitors, and may have
greater downtime and greater uncertainty in their
measurements. Further information regarding CEM's is included
in section 3.2.4 of BID volume 2E. It was determined that the
combination of credit and debit calculations based on
representative operating conditions and records of process
operation such as monthly operating hours and throughputs,
along with continuous monitoring of control device operating
parameters would be a more reliable and efficient means of
enforcing emissions averaging than requiring CEM's. This
selected system is described in the previous response.
Comment; One commenter (A-90-19: IV-D-85) complained
that enforcement of the emissions averaging program is based
on emissions estimation under representative operating
conditions and warned that these estimates cannot substitute
for monitoring and allows "gaming." The commenter (A-90-19:
IV-D-85) explained that the phrase "representative operating
conditions" is vague enough to encompass fairly wide
variations in operating conditions, which gives plant
operators the incentive to choose the operating conditions
most likely to minimize the debits and exaggerate credits.
Therefore, the commenter (A-90-19: IV-D-85 and IV-G-6)
recommended that the EPA require conservative assumptions and
2-108
-------
eliminate gaining possibilities in the estimation of future
credits and debits.
The commenter (A-90-19: IV-D-85) further complained
that if the operating conditions change such that they are no
longer "representative," the operator must conduct a new
performance test, but need not report a violation of the
standard. The commenter (A-90-19: IV-D-85) did not consider
this an adequate check on emissions increases when operating
conditions change especially with respect to uncontrolled
Group 1 points, since no monitoring applies to those points.
The commenter (A-90-19: IV-D-85) recognized that for
the purposes of NSR, the EPA has used representative operating
conditions to assess whether a change of emissions has
occurred, and then traditionally required the operator
predicting no future emissions increases to specify operating
conditions and to accept enforceable permit limitations
including those operating conditions. However, the commenter
(A-90-19: IV-D-85) did not consider this methodology an
adequate means of measuring actual compliance with a standard.
Moreover, the commenter (A-90-19: IV-D-85) warned that the
EPA has not proposed to specify operating conditions as
enforceable limitations; therefore, a source could increase
emissions from every uncontrolled point in an average while
holding emissions from credit generators constant without a
violation of the standard occurring.
Response; Representative operating conditions are
determined on a case-by-case basis, and usually the source and
the implementing agency discuss and agree on performance test
conditions. Thus, the implementing agency can have direct
input in establishing those conditions. It is expected that
operating conditions for points generating debits, as well as
points generating credits will be specified in a source's
operating permit or as part of the approval process for
emissions averaging Implementation Plans. If operating
conditions required in the permit are violated, the
implementing agency could take enforcement action. Changes in
operation would be governed by the operating permit
2-109
-------
modification process, or the requirement for Implementation
Plan updates, if an operating permit is not yet in effect.
Furthermore, as the commenter mentions, if operating
conditions for process vents, transfer operations, and
wastewater streams change such that previously measured
parameters are no longer representative, new representative
values must be determined, and the new measurements must be
used to calculate debits and credits from the time of the
change forward. If the quarterly or annual credit/debit
balances are not met, this is clearly a violation of the
emission standard, and enforcement action can be taken.
Comment; One commenter (A-90-19: IV-D-49) contended
that in no case should monitoring requirements in an averaging
program be less stringent than existing State requirements.
Response: The HON establishes monitoring requirements
that are necessary to determine compliance for emission points
in emissions averages. The HON does not preclude a State from
establishing additional requirements that the State determines
are necessary to establish compliance with other State or
Federal programs that affect the source. The operating permit
program can be used to establish detailed requirements for
each source.
Comment; One commenter (A-90-19: IV-D-117) suggested
that companies which select emissions averaging be required to
pay $1.00 per pound of HAP emissions annually, and that the
money be used to purchase ambient air monitoring systems,
which should be placed in the community closest to the HAP
emission source.
Response; Fee schedules for HAP emissions are set by
States as part of their operating permit programs. The EPA
does not have the authority to mandate ambient monitoring in
communities near sources through this rule. The purpose of
the monitoring required by the HON is to establish compliance
with the rule. Ambient air monitoring is not necessary to
determine compliance with the rule because the HON does not
establish an ambient air target concentration. Further
2-110
-------
information on this topic is contained in a response in
section 3.2 of BID volume 2E.
Comment: Seven commenters (A-90-19: IV-K-1; IV-K-9;
IV-K-10; IV-K-17; IV-K-30; IV-K-44; IV-K-64) supported the
proposal in the supplemental notice to assign no credits and
maximum debits when monitoring data are missing, the monitor
is not functioning, or the monitor indicates that the
operating parameter values are outside the permissible range.
One of the commenters (A-90-19: IV-K-9) asserted that
these provisions would provide incentives for sources to
properly maintain, operate, and monitor equipment.
Another commenter (A-90-19: IV-K-1) reemphasized their
previous claim that full monitoring was required and that
emission estimating and parameter monitoring are not adequate
for averaging. Two commenters (A-90-19: IV-K-17; IV-K-63)
recommended that the EPA mandate the use of real-time ambient
monitoring in emissions averaging through the use of an FTIR
instead of allowing parameter monitoring. Two commenters
(A-90-19: IV-K-1; IV-K-17) considered the proposal defective
because it does not allow enforcement against sources using
emissions averaging that operate their monitors, and control
equipment properly, but overestimate credits and underestimate
debits.
In contrast, ten commenters (A-90-19: IV-K-2; IV-K-7;
IV-K-19; IV-K-22; IV-K-27; IV-K-33; IV-K-34; IV-K-35; IV-K-49;
IV-K-66) opposed the proposal to assign no credits and maximum
debits, citing the following reasons: (1) parameter values
outside of the specified range do not necessarily indicate
complete failure of the control device; (2) process
instrumentation is designed and installed to allow for
continued successful operation when monitors or other
instruments are out of service; and (3) backup devices to
monitors are available and utilized. One commenter (A-90-19:
IV-K-66) presented examples in which monitors were out of
service, but the source was not out of compliance.
Two commenters (A-90-19: IV-K-21; IV-K-22) stated that
the provisions for emission points that are averaged should
2-111
-------
not be more stringent than those for other emission points,
citing §63.152(c)(2)(ii)(A) of the proposed rule, which
references excusable periods during the operation of the
control device.
Response; After rule proposal, the concern was raised
that if a point included in an emissions average experienced
an excursion, the emissions could be different from what was
expected because of the change in the control device's
operation. As discussed in section 3.2.5 of BID volume 2E, an
excursion occurs when either: (1) there are insufficient
monitoring data; or (2) the parameter values are outside the
permitted range. Because of the effect an excursion could
have on a control device's effectiveness, a source that
experiences excursions might be in violation of the standard.
Hence, in the supplemental notice, comment was solicited on an
approach whereby, when points in an average experience
excursions, no credits would be assigned to a credit generator
and maximum debits would be assigned to a debit generator for
the period of the excursion.
After considering the comments submitted, this approach
was added to the final rule. The presumption Is that the
excursion is caused by a significant problem in control device
operation and the device is not achieving emission reductions.
However, if the source has data indicating that some partial
credits or debits may be warranted, the rule provides that the
source can submit that information to the implementing agency
with their next Periodic Report. Partial credits and debits
can be assigned with the approval of the implementing agency.
These provisions are necessary to ensure that averaging
achieves equivalent reductions to point-by-point compliance at
all times. It is also true that the rule now provides sources
with additional incentive to maintain monitoring equipment in
proper working condition. However, this change in the
monitoring provisions for emissions averaging does not
indicate that the emissions estimation methodologies and
parameter monitoring are suspect, or that real-time ambient
monitoring is required. Also, these new provisions are not
2-112
-------
intended to be used to enforce against incorrect debit and
credit estimations. As long as a source uses the equations
specified in the rule correctly and determines the inputs to
the equations according to the stipulated methods, there is no
reason to doubt the accuracy of the debit and credit
estimations.
The EPA agrees with commenters that it may be possible
that an emission point might still be in compliance or the
control might be achieving partial reductions even though an
excursion was reported, and has therefore included provisions
for sources to demonstrate that full or partial debits or
credits are warranted during an excursion. However, if
compliance during excursions cannot be satisfactorily
demonstrated, any other assumption than a full failure of the
control device during the excursion would result in estimated
emission reductions that could not be verified or adequately
enforced. Emissions averaging depends on the demonstration
that debits and credits balance based on the actual operating
conditions after the fact. Compliance on a point-by-point
basis requires only that the source demonstrate that the RCT
was operated at the proper design specifications. Hence, the
' averaging provisions are not more stringent. Rather, they are
more detailed to ensure the consistency of the debit/credit
estimation.
Comment: Four commenters (A-90-19: IV-K-2; IV-K-10;
IV-K-37; IV-K-49; IV-K-50) identified specific situations in
which missing data and exceedances should lead to assigning
maximum debits and no credits. Two commenters (A-90-19:
IV-K-2; IV-K-37) advocated assigning no credits and maximum
debits only if the exceedances were numerous, repetitive, the
result of negligence, or if other operating data indicated
large deviations. Another commenter (A-90-19: IV-K-50) said
no credits and maximum debits should be assigned if the data
were missing or exceedances occurred over an extended period
of time.
Four commenters (A-90-19: IV-K-2; IV-K-30; IV-K-37;
IV-K-49) noted that it would seem reasonable to excuse missing
2-113
-------
data if the occurrences were infrequent and without pattern,
and if 'her data do not -J icate a deviation from normal
operati. ..
Response: The EPA agrees- that at a minimum, these
situations described by conunenters are ones that must be
corrected, but does not agree that they should be the only
situations that warrant taking the approach included in the
final rule. Because of the nature of emissions averaging,
i.e., substitution of control of some points for others, every
step must be taken to ensure that this alternate compliance
approach remains equivalent to the compliance scheme that
would otherwise be required. Hence, it is the EPA's position
that even a single excursion must be corrected and accounted
for in calculating debits or credits, not just extended or
repetitive violations; the source can reestablish the average
balance in the succeeding quarters.
It should be noted that some data can be missing and
still not qualify as an excursion, thus the concerns of
several commenters are accommodated by this policy.
Comment: Several commenters (A-90-19: IV-K-14;
IV-K-17; IV-K-18; IV-K-21; IV-K-25; IV-K-27; IV-K-30; IV-K-33;
IV-K-34; IV-K-39; IV-K-46; IV-K-54; IV-K-55; IV-K-62; IV-K-63;
IV-K-64) offered alternatives for accommodating missing data
and parameter exceedances.
Six commenters (A-90-19: IV-K-21; IV-K-35; IV-K-47;
IV-K-49; IV-K-56; IV-K-66) favored allowing sources to use
other available data to calculate full credits and debits when
monitoring data are missing. One of the commenters (A-90-19:
IV-K-66) further indicated that the source should be required
to notify the permitting authority when the monitor (not the
control unit) is malfunctioning, identify an alternative
monitoring parameter that could be used, and follow other
requirements of the General Provisions.
Three commenters (A-90-19: IV-K-54; IV-K-62; IV-K-64)
advocated including provisions for partial or full credits in
certain circumstances. Two commenters (A-90-19: IV-K-18;
IV-K-21) recommended that no credits and maximum debits be
2-114
-------
assigned only in situations that trigger the option of
quarterly reporting. Another commenter (A-90-19: IV-K-14)
supported requiring data for less than 100 percent of the
operating time, and allowing the use of portable analyzers.
One commenter (A-90-19: IV-K-30) advocated the maximum
allowable emission rates be used to calculate maximum debits.
Another commenter (A-90-19: IV-K-62) recommended
allowing the source to assume the last emission rate measured
if the monitoring equipment was down for less than 24 hours,
but assigning maximum debits and no credits if the monitor
indicates that the operating parameter values exceed allowable
ranges. A third commenter (A-90-19: IV-K-25) agreed that no
credits or debits should be assigned when monitoring data is
missing.
One commenter (A-90-19: IV-K-33) supported case-by-
case assessment of the significance of missing data and
parameter exceedances. Another commenter (A-90-19: IV-K-55)
recommended that State agencies have broad latitude to assess
compliance if the source exceeds the permit limit, but agreed
that no credits should be allowed if monitoring data are
suspect.
One commenter (A-90-19: IV-K-39) advocated using the
approach of the program under Title IV of the Act, which
provides incentives for continuous data and does not
completely eliminate credits for missing data.
Response: The EPA agrees that there may be some cases
or conditions under which the implementing agency can be
satisfied that granting partial or full credits and debits is
still warranted. For example, the emission point may be
routed to a backup control device, or there may be evidence
that a control device is operating even if a particular
monitor is out of service. Therefore, the final rule provides
that the evaluation and issuance of credits and debits during
questionable periods shall be at the discretion of the
implementing agency.
It is stipulated in §63.150(1) of the final rule, that
along with notifying the implementing agency that excursions
2-115
-------
have occurred, a source may demonstrate that other types of
monitoring data or engineering calculations are appropriate to
establish that the control device for the emission point was
operating in such a fashion that partial, if not complete,
reduction was being achieved. Documentation of these other
types of monitoring data or engineering calculations must be
provided to the implementing agency at the time the excursion
is reported. The demonstration must be made to the
implementing agency's satisfaction according to the procedures
that the agency has established. Some of the alternatives
suggested by commenters for accommodating missing data and
parameter exceedances workable, and an implementing agency may
allow for the suggestions in their procedures. The
implementing agency may then assign full or partial credits
and debits upon review of the information provided.
As noted in the previous response, no credits and
maximum debits are assumed (unless proven otherwise) for
excursion periods, which are determined on a daily basis. If
a monitor is out of service or a parameter is out of range for
such a short period that it does not cause a daily excursion,
then the calculation of credits and debits are not affected.
The definition of an excursion is the same for points in
emissions avera ss as it is for other points at the source.
Use of a consistent definition of an excursion ^s necessary so
that the averaging provisions achieve the same reductions as
the point-by-point provision.
Comment: Three commenters (A-90-19: IV-K-27; IV-K-34;
IV-K-46) recommended that emission credits and debits should
be independent of each other, so that facilities are not
doubly penalized when one monitor was not working. The
commenters (A-90-19: IV-K-27; IV-K-34; IV-K-46) also
advocated limiting emission credits and debits to the emission
points addressed by the monitor in question.
Response: The EPA thanks the commenters for
recognizing this important point, which was not made clear in
the supplemental notice. The only emission point that is
affected by this new provision is the point exhibiting the
2-116
-------
excursion. If that point is a credit generator, it will be
assumed that the point generated no credits for the duration
of the excursion. No other points are affected, and the
source will not be doubly penalized.
2.8.3 Administrative Burden
Comment; Three commenters (A-90-19: IV-D-70; IV-D-90;
IV-D-100) were concerned about the burden and difficulty
emissions averaging .would pose to the implementing agency.
One commenter (A-90-19: IV-D-70) contended that for each of
the hundreds or thousands of emission points at a facility,
three numbers would have to be evaluated and checked
(baseline, required, and proposed). The commenter (A-90-19:
IV-D-70) also stated that emissions averaging would require
many OEM's to be certified and reports reviewed and concluded
that emissions averaging would require three to four times
more resources to administer than HON without emissions
averaging.
Two commenters (A-90-19: IV-D-90; IV-D-100) were
concerned about the implementing agencies' abilities to
determine compliance of a HON facility that averages emissions
because of the complexity involved in estimating and tracking
emissions from various kinds of points. The commenters
(A-90-19: IV-D-90; IV-D-100) claimed that extensive
monitoring and recordkeeping will be required to ensure that
sources using averaging achieve the same emission reduction as
would be achieved without averaging. The commenters (A-90-19:
IV-D-90; IV-D-100) stated that extensive resources would also
be required to track emission fluctuations and associated
adjustments to the emissions average due to modifications or
simple routine maintenance of points within the average.
Hence, the commenters (A-90-19: IV-D-90; IV-D-100) contended
that the resource requirements for emissions averaging far
exceed what is currently available or anticipated by State
agencies in implementing the operating permit program.
Response; The administrative burden of implementing
the emissions averaging program of the rule is one issue
voiced by all concerned parties to which the EPA paid
2-117
-------
particular attention. Many of the changes to the final rule
were made in response to comments and with the express purpose
of easing perceived administrative burdens. These changes are
discussed in greater detail throughout this BID volume. Not
all changes that were suggested could be made, and some
provisions were added to improve enforceability or to ensure
public health protection, which may contribute to the
administrative burden.
Some commenters were mistaken about the number of
emissions estimates that would be required in emissions
averaging. Calculation of emission debits and credits are
required for only the points included in an emissions average,
not for all emission points at a source. One change in the
final rule is that an average can contain no more than
20 points, 25 if pollution prevention is used (this new
prc sion is discussed in section 2.8.5 of this BID v._ume).
So, :he concern over tne number of points for which emission
estimates are required has been addressed.
Finally, the source need only calculate two emission
values for a debit generator and two values for the credit
generator. For a debit generator, actual emissions based on
the controls in place (if any) and emissions if the RCT had
been applied need to be calculated and compared. These values
can be easily calculated using estimates of uncontrolled
emissions and the reduction efficiencies of controls that were
demonstrated in initial performance tests, and specific
procedures for making estimates and carrying out performance
tests are provided in the rule. For a credit generator, the
emissions that are allowed under the rule and the actual
emissions are calculated using procedures specified in the
rule as well.
The CEM's to which the commenters referred are not
required for eraging or other parts of the rule; rather
continuous parameter monitoring is allowed. It is true that
emissions averaging requires quarterly reporting whereas under
the rule without averaging, Periodic Reports need be submitted
only twice a year. Still, the EPA does not foresee the
2-118
-------
administration of an emissions average requiring three to four
times more resources than if the points were complying on a
point-by-point basis, and no compelling evidence was provided
to substantiate the claim.
The EPA does not consider the estimation methodologies
for averaging to be too complex; the appropriateness of the
methodologies is discussed in greater detail in section 2.11.4
of this BID volume. Tracking emissions is not unduly complex
either as the points included in averages must be identified
separately in the Implementation Plan or the operating permit.
Debits and credits are calculated monthly based on limited
inputs such as monthly operating hours and previously measured
values. Adjustment of emission calculations for fluctuations
is reguired only if an excursion occurs, and specific
procedures have been included in the final rule to address
such situations. It was acknowledged earlier that some
additional monitoring, recordkeeping, and reporting is
necessary to implement emissions averaging and to ensure
proper operation. But, again, the EPA maintains that with the
new limits on averaging, any additional burden has been
limited as well and does not far exceed that associated with
compliance on a point-by-point. If, however, ah implementing
agency does realize greater costs in administering averages,
the cost could be addressed by applying a higher permit fee
for the points included in emissions averaging.
Comment; Five commenters (A-90-19: IV-D-85 and
IV-F-7.39 and IV-F-12; IV-D-87; IV-D-99;" IV-D-115; IV-F-7.6)
testified that State and local agencies found the emissions
averaging programs implemented in the past complicated to
enforce because it is very difficult to determine whether all
points involved in an average are maintaining emissions below
the required levels. One commenter (A-90-19: IV-D-115)
maintained that this can only be determined in hindsight
through review of records, which means that equipment may
continue to operate out of compliance for prolonged periods of
time. Four commenters (A-90-19: IV-D-87; IV-D-99; IV-D-115;
IV-F-7.6) predicted that tracking, recordkeeping, and
2-119
-------
enforcement will be a long-term resource and financial strain
for industry as well as for State and local agencies.
Response: Under the emissions averaging program in
this rule, it is not difficult to determine whether points are
maintaining their required emissions levels. An inspector
must simply check whether controls have been installed and are
operating properly for credit-generating points, just as all
the Group 1 points would be inspected under point-by-point
compliance.
Then, sources are required to calculate debits and
credits using the equations specified in the rule and to
report results. Because the calculations must be made
according to specific procedures and because the inputs on
which the calculations are based and data on any monitoring
parameter excursions must be provided, the calculations can be
checked relatively easily. Thus, it should not be true that
noncompliance situations would be allowed to continue for
prolonged periods.
Comment: One commenter (A-90-19: IV-D-85 and
IV-F-7.39 and IV-F-12) suggested that State agencies in areas
with large numbers of chemical plants with potentially huge
numbers of emission estimates lack the resources to oversee an
emissions averaging system that depends on estimates and one-
time performance tests rather than reported monitoring of
emissions. The commenter (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12) further suggested that States will have to check
carefully whether prior reductions credited are real, evaluate
requests for alternative monitoring, and compare alternative
technologies to RCT. The commenter (A-90-19: IV-D-85) added
that after creating monitoring programs, States will have to
redo these programs every time a source amends their
Implementation Plan. The commenter (A-90-19: IV-D-85 and
IV-F-7.39 and IV-F-12) complained that few of these items are
required under the proposed rule and, if they were, an
enormous increase in permit fees would be required to fund the
work.
2-120
-------
Response; As indicated previously, the administration
of an emissions average is not much different than point-by-
point compliance. The same types of monitoring are required
for points in averages as for other Group 1 points. Beyond
inspections for proper operation of control devices, averaging
compliance determinations are based on monitoring data and
debit and credit calculations submitted in Periodic Reports.
The final rule does not allow credit for controls applied
prior to November 15, 1990, resulting in a decrease in the
administrative burden, as discussed in section 2.5.3 of this
BID volume. Many of the other burden items mentioned by the
commenter are not unique to emissions averaging. An
implementing agency will have to evaluate requests to use
alternative monitoring and control technologies under point-
by-point compliance as well as under averaging, and it is not
anticipated that there will be many such requests. Moreover,
requests for alternative means of emission limitation
(requested in place of an equipment or work practice standard
under point-by-point compliance) and requests for nominal
efficiencies for new control technologies that are more
efficient than the RCT will usually have to be approved by the
EPA, not the implementing agency. Any change to the
Implementation Plan must be reviewed regardless of the use of
averaging. Contrary to the commenter's claim, these items
were all included in the proposed rule and have been retained
for the final rule.
Comment: One commenter (A-90-19: IV-D-70) stated that
emissions averaging was in total conflict with their State's
NSR requirements. The commenter (A-90-19: IV-D-70) added
that they had implemented a comprehensive NSR program for over
twenty years, and their State concluded early in this history
that emissions averaging or "bubbling" did not contribute to
effective air pollution control and was a detriment to these
efforts.
Response; The EPA recognized that the averaging
provisions might conflict with some states' existing programs
for regulating HAP's or other air quality programs. Hence,
2-121
-------
the provision has been added to the final rule for States to
exclude averaging from their implementation of the rule
outside of the rule delegation process provided under
section 112(1) of the Act. The discussion of this new
provision is contained in section 2.8.4 of this BID volume.
However, another change to the rule prohibiting new sources
from using emissions averaging, which is discussed in
section 2.3.2 of this BID volume, may make averaging in this
rule compatible with the State's existing air programs.
Comment; One commenter (A-90-19: IV-D-115) maintained
that the emissions averaging provisions should not apply to
sources that are subject to other applicable requirements, or
that such sources should only be allowed to undercontrol down
to the levels permitted under the existing Federal, State, or
local requirements.
Response: If another State or Federal regulation
applies to an emission point subject to the HON, the more
stringent of the requirements takes precedence. As such, if
another rule requires control more stringent than the RCT
established by the HON, the point cannot be left uncontrolled
or undercontrolled as a debit generator in an emissions
average. However, if controls are installed after 1990 and
achieve more stringent control than is required by the other
State or Federal rule, the emission point is eligible as a
credit generator in an emissions average, but only for the
control above what is required by the other rule.
Even if the HON RCT is the more stringent of two
requirements, the source must maintain the control established
by the other requirement. If the point were controlled with
the HON RCT, both requirements would be met. However, if the
source plans to use the point as a debit generator, the point
must still meet the non-HON requirement. The emission point
can be used as an undercontrolled (according to the HON) debit
generator for which the difference in control between the HON
and the other requirement is the basis for the debits.
The EPA is considering allowing a limited exception for
Federal RACT requirements that apply to points subject to the
2-122
-------
HON. If the policies published in draft guidance
(58 FR 54136) are implemented, points to which RACT
requirements apply can be left completely uncontrolled as
debit generators as long as both the HAP and non-HAP portion
of the VOC emissions are balanced by credit generators.
Comment; One commenter (A-90-19: IV-D-99) stated that
the sources should be required to submit emissions averaging
proposals to State and local agencies that have delegated air
toxics programs, who could then approve or disapprove the
averages.
One commenter (A-90-19: IV-D-70) maintained that
emissions averaging should not prohibit a State's authority to
require review of plant modifications and emissions increases
under its State permitting program. Another commenter
(A-90-19: IV-D-115) was concerned that the emissions
averaging proposal in the RON would allow sources to be
constructed without the controls required for every other
source in their district.
Response: All of the commenters' recommendations are
already required or allowed in the rule. Implementation Plans
for emissions averaging are to be submitted to the
implementing agency for approval. The HON will not in any way
prevent States from enforcing other regulations. One
commenter's concern about newly constructed sources has been
addressed by excluding the use of emissions averaging by new
sources.
Comment: One commenter (A-90-19: IV-F-7.5) stated
that a detailed strategy for enforcement must be demonstrated
by the EPA so that the delegated authority will be able to
construct the necessary legal and monitoring strategies.
Another commenter (A-90-19: IV-D-70) requested that the EPA
provide detailed guidance to address policy and procedural
questions that will arise in implementing an emissions
average. The commenter (A-90-19: IV-D-70) also suggested
that the agencies that must implement emissions averaging
should have significant input concerning the approach of the
program.
2-123
-------
Response; The provisions for calculating debits and
credits, monitoring, recordkeeping and reporting, and
enforcement for emissions averaging are quite detailed in the
rule in order to eliminate questions and confusion. By
"baseline," it is assumed that one conunenter is referring to
the uncontrolled or undercontrolled emissions for determining
debits or to the emission controls in place on Group 2 points
on November 15, 1990, which are used in determining credits.
Again, the specific procedures and equations for calculating
these emissions are included in the rule, and these are the
only equations that can be used. Finally, implementing
agencies that will be delegated authority for administering
the rule and the emissions averaging program have had ongoing
opportunities to provide input on how the averaging program
was designed, including work group representation, roundtable
discussions, and conference calls with the EPA prior to
proposal, as well as the opportunity to speak at two public
hearings and to submit written comment on the proposed rule
and supplemental notice.
Comment; One commenter (A-90-19: IV-D-115) requested
that the rule require a demonstration that sources electing to
emissions average are in compliance. The commenter (A-90-19:
IV-D-115) stated that a procedure for demonstrating compliance
must be worked out such that it is easily verified and not
overly burdensome, which may not be possible.
One commenter (A-90-19: IV-D-85 and IV-G-6) suggested
that a certification should be required that the summation of
debits and credits is accurate and that they equal the
emissions that would have been emitted had all Group 1 points
in the average been controlled.
Response: The source must demonstrate compliance each
quarter and submit the demonstration in each Periodic Report.
The rule specifies n §63.152(c)(5)(iv) that every fourth
quarterly report - ~1 include a demonstration that annual
credits are greatt than or equal to annual debits and a
certification of compliance with all the emissions averaging
2-124
-------
provisions in the rule. This would be the basis of the annual
compliance certification required under Title V of the Act.
Comment: One commenter (A-90-19: IV-D-98) requested
clarification regarding penalties imposed by 1990 amendments
to the Act of $25,000 per "incident" for violations of
emissions limit. The commenter (A-90-19: IV-D-98) claimed it
was not clear whether these penalties would be assessed on the
basis of the overall plant limit or with respect to each
emissions point, nor what the implications of using emissions
avexaging as a means of compliance are with respect to a
violation of the emissions limit.
Response; The rule does not establish an overall plant
limit; rather each emission point in a source subject to the
HON must comply with control or operating requirements
established for each kind of point. If any individual
emission point experiences an unexcused excursion, this
constitutes a violation that could be subject to the maximum
penalty of $25,000 per day of violation. This penalty may be
assessed for each violation at each control device per day.
(If more than one rule applies to a point or control device,
more than one violation may be cited for each point or control
device found to be out of compliance.)
These same provisions apply to emission points involved
in an emissions average. If any controlled point in an
average experiences an unexcused excursion, the point is
liable for up to the $25,000 maximum penalty per violation per
day. Moreover, if the violation also results in noncompliance
with the quarterly averaging check or the annual averaging
balance, it counts as yet another violation, which is subject
to the penalty. Therefore, the source may be penalized up to
a maximum of $25,000 for every day a point experiences an
unexcused excursion and another $25,000 for every day of the
quarter or year that the average is out of balance. It should
be pointed out, however, that the EPA will exercise its
enforcement discretion in assessing penalties.
Comment; Two commenters (A-90-19: IV-D-74; IV-D-108)
supported the provision that excludes periods of start-ups,
2-125
-------
shutdowns, and malfunctions from the calculation of monthly
credits and debits because of the difficulty and burden of
quantifying emissions under such conditions and because they
are not representative of operations.
Response: The commenters' support of the EPA's
position is appreciated.
Comment: On commenter (A-90-19: IV-D-78) recommended
allowing compliance extensions in cases where a credit-
generating point shutdown or slowdown occurs or a debit-
generating point increases emissions. The commenter (A-90-19:
IV-D-78) argued that the extension would provide time to find
other credits.
Response: Compliance extensions will not be allowed
under the Act (other than an initial case-by-case extension of
up to 1 year provided for in the Act if there is prior
justification and approval). The provision that sources will
be found in violation if quarterly checks or annual averages
are violated will motivate the operators to select emission
points on units where shutdowns and slowdowns are unlikely.
Moreover, the operators should be prudent enough to
have additional credits built into their averages to avoid
being in violation. Many commenters have claimed that such
conservatism will arise in an emissions averaging scheme.
Comment: One commenter (A-90-19: IV-D-74) complained
that the emissions averaging calculations place excessive
reliance on actual operating conditions. Two commenters
(A-90-19: IV-D-74; IV-D-108) noted that proposed
§63.150(f)(2)(ii)(B) requires a re-determination of
representative values for flow, concentration, stream
molecular weight, and temperature every time there is a change
in capacity utilization or in the vent stream flow rate,
concentration, molecular weight or discharge temperature. The
commenters (A-90-19: IV-D-74; IV-D-108) submitted that this
ape -oach will not be useful for batch operations, such as in
ph-: .-maceutical manufacturing, because the equation assumes an
absolutely constant operation during the month. Hence, the
commenters (A-90-19: IV-D-74; IV-D-108) suggested that
2-126
-------
predictive calculations of emissions should be allowed to
demonstrate emissions.
Response; Process vents from batch operations are not
subject to the HON and therefore are not eligible for
emissions averaging. The equations in emissions averaging
allow for variation in operating hours, as monthly operating
hours are an input to the equations; however, it is assumed
that during periods of operation, the other operating
conditions that influence emissions such as flow and
concentration are relatively constant. This is generally true
for the kinds of emission points allowed in averaging. In
situations where operating conditions vary, a source would be
prudent to test a number of different likely operating
conditions initially and include alternative operating
scenarios in their Implementation Plan or operating permit
application as specified in §63.151(h) of subpart G of this
rule. However, if representative conditions are difficult to
establish, these units may not be good candidates for
emissions averaging.
2.8.4 State Discretion on Emissions Averaging
Comment: Several commenters (A-90-19: IV-D-49;
IV-D-85 and IV-F-7.39 and IV-F-12; IV-D-87; IV-D-90; IV-D-99;
IV-D-100; IV-D-115; IV-F-7.6) on the proposed rule recommended
that State and local agencies, be allowed to implement MACT
standards without the emissions averaging provisions. Two
commenters (A-90-19: IV-D-87; IV-D-115) requested that States
be provided with maximum flexibility in implementing and
enforcing regulations that are at least as stringent as the
EPA standard, and if a State elects not to allow emissions
averaging, its equivalent program should be considered more
stringent under the program required under Title III of the
1990 Amendments.
Response; The EPA announced in the supplemental notice
that it was considering adding language to the HON that would
grant State or local agencies the discretion to exclude
emissions averaging from their implementation of the HON
without having to go through the rule adjustment process
2-127
-------
specified under section 112(1) of the Act. The final rule has
been revised to grant this discretion, thus providing States
more flexibility in implementing the HON.
Comment: Nine commenters (A-90-19: IV-K-9; IV-K-10;
IV-K-18; IV-K-30; IV-K-37; IV-K-44; IV-K-55; IV-K-63; IV-K-64)
supported the provisions proposed in the October 15, 1993
supplemental notice that would grant State and local agencies
the discretion to not include emissions averaging in their
implementation of the rule without having to go through the
rule adjustment process under section 112(1) of the Act.
Four commenters (A-90-19: IV-K-10; IV-K-37; IV-K-55;
IV-K-64) maintained that State and local agencies should be
allowed to include or exclude emissions averaging provisions,
without any EPA review, when implementing the rule. Eight
commenters (A-90-19: IV-K-9; IV-K-10; IV-K-18; IV-K-30;
IV-K-37; IV-K-41; IV-K-44; IV-K-55) asserted that emissions
averaging creates an administrative burden for States, so they
should have the discretion to exclude averaging. Five
commenters (A-90-19: IV-K-17; IV-K-30; IV-K-37; IV-K-41;
IV-K-55; IV-K-63) argued that States should have discretion
because averaging is not enforceable, and will not protect
public and environmental health. Six commenters (A-90-19:
IV-K-l; IV-K-10; IV-K-17; IV-K-37; IV-K-55; IV-K-64) stated
that allowing the States increased flexibility to implement
the rule is desirable. Two commenters (A-90-19: IV-K-9;
IV-K-10) indicated that flexibility to implement regulations
consistent with State or local agency policy was necessary
because the use of limited resources must be maximized to
implement the HON.
Three commenters (A-90-19: IV-K-9; IV-K-44; IV-K-64)
maintained that States should have discretion not to include
emissions averaging because averaging would make the rule less
stringent. Four commenters (A-90-19: IV-K-l; IV-K-30;
IV-K-37; IV-K-55) considered it critical to include the State
discretion provision because otherwise, States would be forced
to adopt the less stringent Federal regulation, which includes
emissions averaging.
2-128
-------
Several commenters (A-90-19: IV-K-2; IV-K-6; IV-K-7;
IV-K-17; IV-K-19; IV-K-20; IV-K-21; IV-K-25; IV-K-27; IV-K-28;
IV-K-33; IV-K-34; IV-K-35; IV-K-39; IV-K-40; IV-K-41; IV-K-42;
IV-K-45; IV-K-46; IV-K-47; IV-K-50; IV-K-54; IV-K-56; IV-K-61;
IV-K-62; IV-K-66) opposed allowing State discretion to not
include emissions averaging without going through the
section 112(1) rule adjustment process.
Several commenters (A-90-19: IV-K-7; IV-K-20;
IV-K-26; IV-K-27; IV-K-34; IV-K-39; IV-K-42; IV-K-46; IV-K-47;
IV-K-50; IV-K-54; IV-K-61; IV-K-66) argued that such a
provision would allow States to not adopt emissions averaging,
which would limit a source's ability to select cost-effective
control options. Eight commenters (A-90-19: IV-K-2; IV-K-34;
IV-K-42; IV-K-46; IV-K-61; IV-K-56) maintained that States
should not be allowed to deny the flexibility that emissions
averaging affords. Three commenters (A-90-19: IV-K-34;
IV-K-45; IV-K-46) stated that the rulemaking process under
section 112(1) gives States ample flexibility to address State
equivalency determinations.
Three commenters (A-90-19: IV-K-21; IV-K-42; IV-K-54)
argued that allowing State discretion would create an uneven
playing field, and that facilities in States without these
provisions would be penalized. Another commenter (A-90-19:
IV-K-56) suggested that sources would be subject to different
HON rules if State discretion not to include emissions
averaging were allowed.
One commenter (A-90-19: IV-K-25) opposed allowing
State discretion claiming that emissions averaging eases the
administrative burden on the State implementing agency.
Another commenter (A-90-19: IV-K-62) agreed that emissions
averaging places an administrative burden on the State, but
this was not sufficient reason to disallow averaging. One
commenter (A-90-19: IV-K-39) disagreed with allowing State
discretion claiming if emissions averaging costs were not
considered, the cost calculation for the regulation was
incorrect.
2-129
-------
Two commenters (A-90-19: IV-K-39; IV-K-62) argued that
States should not be allowed to exclude the emissions
averaging provisions if they are not allowed to exclude other
provisions. Three commenters (A-90-19: IV-K-1; IV-K-34;
IV-K-46) stressed that having the State discretion provision
may create regulatory promulgation difficulties for some
States. One of the commenters (A-90-19: IV-K-1) suggested
that emissions averaging be eliminated in order to avoid the
regulatory uncertainty created by the State discretion
provision.
Response: The EPA maintains that States should have
discretion on whether to allow emissions averaging for a
nu ir of reasons. First, the EPA acknowledges that averaging
ca oe more complex to Iminister than the rule allowing only
point-by-point compliance, so allowing averaging could
increase the administrative burden, which is an especially
important concern for implementing agencies with limited
personnel and resources. However, the determination of what
constitutes too much administrative burden will differ from
State to State. Some States may consider emissions averaging
an acceptable strategy for compliance and will retain the
program.
Second, the EPA recognized that averaging in the HON
could be inconsistent with some States' ongoing air pollution
control programs. The EPA supports the use of emissions
averaging where it may be appropriate, and maintains again
that the program has been designed to be enforceable and
protective of health and welfare. However, the EPA also
acknowledges that its use must be balanced by the individual
needs of State and local agencies that bear the responsibility
for administering and enforcing the rule. Furthermore, with
the inclusion of these provisions, the EPA does not consider
the stringency of the rule with or without averaging is to be
an issue. Stringency is discussed in greater detail in the
next response in this section.
Allowing this discretion will not create an uneven
"playing field" because without this provision, most States
2-130
-------
already have the ability to exclude emissions averaging
through the section 112(1) rule adjustment process encoded in
40 CFR 63.92, 63.93, and 63.94. Rather, the EPA has decided
to make excluding averaging more simple by exempting the
decision from the section 112(1) rule adjustment process.
Including this provision in the HON will reduce paperwork
burdens on States, expedite delegation of the rule to States,
and remove a potential source of uncertainty for sources
subject to the HON.
The EPA does not agree that providing for State
discretion in the HON itself is either unnecessary or
burdensome for States. While the section 112(1) rule
adjustment process would also permit States to choose to
implement the HON without averaging, providing for that choice
in the HON itself streamlines the process by eliminating EPA
review of the choice. In addition, since the section 112(1)
rule permits States to make the choice, providing for the
exercise of such discretion in the HON itself cannot be viewed
as placing any new burdens on States. The provision of an
option will not impose a burden or impose new requirements; it
increases choice and flexibility. Furthermorer if emissions
averaging is removed by a State, the calculation of cost and
economic impacts of the rule are not affected because the
impacts do not reflect possible use of averaging. The cost
impacts presented in the proposal preamble were based on
applying an RCT to each Group 1 emission point, while the
economic analysis at proposal was based on applying control to
every emission point. Specific comments on the cost analyses
are addressed in BID volumes 2A and 2B.
Because emissions averaging is an alternative
compliance method to the primary control strategy, States
should have the discretion to exclude it as opposed to other
provisions that are essential to the rule and for which no
alternative compliance mechanism has been provided.
Finally, the EPA predicts that instead of creating
promulgation difficulties and uncertainties, providing the
clarifications in this provision at this time will benefit
2-131
-------
sources as well as States. Without this provision, sources
might be uncertain during the section 112(1) rule adjustment
process about whether averaging ultimately would be allowed or
not in their State, yet would be given no added time for
compliance. The EPA predicts that because of their complex
nature, many HON sources will need the full time period
allowed for compliance.
Comment; Seven commenters (A-90-19: IV-K-19; IV-K-21;
IV-K-22; IV-K-28; IV-K-40; IV-K-47; IV-K-66) warned that to
allow States discretion to exclude emissions averaging
provisions would conflict with the clear mandate of
section 112(1) of the Act. Two commenters (A-90-19: IV-K-35;
IV-K-50) reasoned that the section 112(1) process was
necessary to ensure that more stringent State programs are
consistent with the Act.
Two commenters (A-90-19: IV-K-25; IV-K-26) argued that
the EPA should not circumvent the mandate of State
legislatures barring their States from enacting more stringent
requirements such as the rule without emissions averaging.
On the other hand, three commenters (A-90-19: IV-K-7;
IV-K-39; IV-K-62) considered the rule with emissions averaging
to be more stringent.
Response; The section 112(1) process requires States
to make a demonstration that the State rule is of equivalent
or greater stringency to the Federal rule. For a State rule
without averaging, one component of this demonstration would
be to show that the lack of averaging did not result in the
State rule being less stringent than the Federal rule. The
EPA has determined that requiring States to make this
demonstration would be a needless exercise for the following
reason. The final rule defines both point-by-point compliance
and averaging as acceptable ways of achieving a MACT level of
control. If all sources in a State use the point-by-point
compliance method — as would be the case in a State that
implemented HON requirements without averaging — all sources
would be achieving the MACT level of control required by the
rule. Under the final rule, no source is required to achieve
2-132
-------
emissions reductions greater than would be achieved by point-
by-point compliance, and no source is required to use
averaging. Therefore, a State rule that implements
requirements of the HON rule without averaging is equivalent
in stringency to the Federal HON rule.
Based on this equivalency finding and the final rule,
the EPA is allowing States to implement the HON unchanged
without averaging through the same processes available to
States that wish to implement the HON unchanged with
averaging. Before an operating permit program is in effect in
a State, the States may implement the HON without change
through a streamlined procedure in §63.91 of the
section 112(1) rule. After the State's operating permit
program is in effect, the State may implement the HON either
with or without averaging without going through any of the
section 112(1) rule adjustment process. Also based on this
equivalency finding and the final rule, a State seeking EPA
approval for a State rule that differs from the HON and lacks
averaging will not have to make a demonstration related to
averaging as part of their equivalency demonstration.
By providing State discretion in the rule, the EPA is
not circumventing any State laws or overriding the decisions
of State legislatures that limit the ability of implementing
agencies to adopt requirements more stringent than Federal
requirements. The EPA maintains that implementing the rule
without averaging would not be a decision to implement a more
stringent program. Moreover, if a State law or constitution
contained provisions, that, in the States' view, prevented the
State from adopting the rule without averaging, nothing in the
rule would override that provision, i.e., in that situation,
the implementing agency would not have the authority to
implement the rule without averaging, and the provision
allowing the State to choose would not change that.
Comment: Eight commenters (A-90-19: IV-K-6; IV-K-22;
IV-K-27; IV-K-34; IV-K-39; IV-K-46; IV-K-47; IV-K-56) argued
that allowing State discretion to exclude emissions averaging
is contrary to the intent of President Clinton as expressed in
2-133
-------
Executive Order 12866 and EPA Administrator Carol Browner.
Eight commenters (A-90-19: IV-K-6; IV-K-22; IV-K-27; IV-K-34;
IV-K-39; IV-K-46; IV-K-47; IV-K-56) maintained that allowing
State discretion: (1) defeats the intent of the Executive
Order by specifying compliance behavior? (2) ignores the
directive that regulations be cost effective; or (3) stifles
the ability of the source to use innovative methods.
Response; Allowing State discretion to exclude
emissions averaging is not contrary to the executive order or
remarks made by the Administrator as suggested by the
commenters. The commenters have neglected to point out that
in both the order and the Administrator's remarks, the goals
of designing cost-effective and flexible regulations, which
stimulate innovative control responses, should be met within
the larger context of achieving and enforcing the emission
reductions required by the Act. Simply by including emissions
averaging as an option, both the order and the Administrator's
policies have been addressed.
As discussed previously, most States already had the
discretion through the rule adjustment process developed
pursuant to section 112(1) of the Act to exclude emissions
averaging; these new provisions only make the process of doing
so simpler. The Act cannot be contrary to the executive order
nor can the reverse be true. Moreover, the HON contains many
more provisions to increase flexibility and innovation. More
than one control technology or method of compliance is
available for each kind of emission point. The source is free
to develop a unique, innovative method so long as it meets
agency approval. Also, the establishment of Group status for
emission points focuses the rule on the points that are the
most cost-effective to control. All-in-all, the HON, which
has been in development since well before the executive order
or the Administrator's stated policies were issued, is still
in keeping with both sets of guidelines.
2.8.5 Number of Points Allowed in Averages
Comment; Nine commenters (A-90-19: IV-K-1; IV-K-9;
IV-K-10; IV-K-18; IV-K-29; IV-K-37; IV-K-44; IV-K-52; IV-K-54;
2-134
-------
IV-K-55) supported the proposal in the October 15, 1993
supplemental notice to limit the number of emission points
allowed in averages. Three commenters (A-90-19: IV-K-9;
IV-K-10; IV-K-37) reiterated that allowing too many emission
points in an average would be burdensome for State authorities
and would not be enforceable.
One commenter (A-90-19: IV-K-18) suggested that the
number of points that can be included in averages should be
limited to no more than 10 percent of the emission points in
the source. Another commenter (A-90-19: IV-K-10)
recommended that the maximum number of points in the average
be limited to 20 if the source has more than 400 points. Two
commenters (A-90-19: IV-K-1; IV-K-44) indicated that if
emissions averaging were included, it should be limited to
5 points or 5 percent of the points, whichever is less.
Another commenter (A-90-19: IV-K-54) advocated limiting the
number of uncontrolled or undercontrolled points to 5 or 10,
and allowing 5 to 10 overcontrolled points for each
uncontrolled or undercontrolled point. One commenter
(A-90-19: IV-K-52) supported limiting the number to 15 points
or 5 percent of the total number of points in the source,
whichever is greater.
On the other hand, several commenters (A-90-19:
IV-K-2; IV-K-7; IV-K-14; IV-K-19; IV-K-21; IV-K-22; IV-K-25;
IV-K-26; IV-K-27; IV-K-28; IV-K-33; IV-K-34; IV-K-35; IV-K-39;
IV-K-42; IV-K-46; IV-K-47; IV-K-48; IV-K-49; IV-K-50; IV-K-56;
IV-K-61; IV-K-64; IV-K-66) opposed limiting the number of
points that can be included in an emissions average.
Ten commenters (A-90-19: IV-K-6; IV-K-14; IV-K-21;
IV-K-22; IV-K-26; IV-K-35; IV-K-39; IV-K-48; IV-K-56; IV-K-66)
declared that limiting the number of points in the average
would limit sources' flexibility. Three commenters (A-90-19:
IV-K-2; IV-K-30; IV-K-37) warned that it would limit
flexibility especially for sources with large numbers of
emission points. Six commenters (A-90-19: IV-K-14; IV-K-17;
IV-K-26; IV-K-35; IV-K-42; IV-K-66) stated that limiting the
number of points would hinder the ability of a source to
2-135
-------
select cost-effective controls, and warned that the proposed
limit would discourage averaging in situations where it was
especially useful, specifically those in which one large
emission point could be overcontrolled for credit and many
smaller points could be left undercontrolled as debit-
generators .
Two commenters (A-90-19: IV-K-47; IV-K-66) claimed
that only a limited number of facilities would be using
emissions averaging, and they would need to include a large
number of points for averaging to be effective. Four
commenters (A-90-19: IV-K-27; IV-K-34; IV-K-46; IV-K-47)
argued that the number of points that can be averaged had
already been severely limited by the design of the program,
and should not be further reduced. One commenter (A-90-19:
IV-K-14) considered a limit to be unfair to sources that have
already reduced their emissions.
Six commenters (A-90-19: IV-K-7; IV-K-33; IV-K-39;
IV-K-47; IV-K-56; IV-K-66) maintained that implementing an
emissions average was more of a burden for the source than the
implementing agency, and therefore a limit was not necessary
to decrease the implementing agency's burden.
Five commenters (A-90-19: IV-K-26; IV-K-27; IV-K-34;
IV-K-46; IV-K-47) opposed limiting the number of points
because equipment leaks could not be accommodated in the
future. Two commenters (A-90-19: IV-K-14; IV-K-28) cautioned
that the limit would decrease the incentive to use pollution-
preventing technologies.
Response; The EPA announced in the supplemental notice
that it was proposing to restrict the number of points allowed
in an emissions average to address concerns for the
administrative burden posed by the use of averaging. The
proposal requested comment on the feasibility of including
such a limit and on what form the limit should take: (1) a
restriction on the percentage of total emission points in the
source in the range of 5 to 15 percent; or (2) a restriction
on the total number of points that can be included in averages
in the range of 5 to 15 emission points.
2-136
-------
After considering the public comment on the proposal,
the final rule has been revised to limit a source to including
no more than 20 Group 1 and Group 2 emission points in an
emissions average. Where pollution prevention measures are
used to control emission points to be included in an average,
no more than 25 points can be included. For example, if two
points to be included in an average are controlled by the use
of a pollution prevention measure, the source can include up
to 22 points in their emissions average. However, if six or
more points in the average are controlled by pollution
prevention, the source can include no more than 25 points in
their average.
The EPA concurs that most sources will not find a large
number of opportunities to generate cost-effective credits.
Hence, it can be anticipated that most averages will involve a
limited number of emission points, and imposing a limit should
not affect most sources. The EPA rejected the choice of a
fixed percentage of points at a source because for larger
sources, this could result in hundreds of emission points in
averages, which is unacceptable from an enforcement
perspective.
The limit of 20 points, 25 if pollution prevention is
used, was chosen because the EPA anticipates that most sources
will rarely want to include more than 20 points in an average.
A higher number of points is allowed where pollution
prevention is used in order to encourage pollution prevention
strategies, and because the same pollution prevention measure
may reduce emissions from multiple points. Otherwise,
allowing much more than 20 to 25 points would make enforcement
increasingly untenable. Thus, the competing interests of
flexibility for sources and enforceability were balanced in
this decision.
There may be situations where overcontrolling a point
could generate enough credits to offset emissions from a
number of smaller debit-generating points, but the limit on
the number of points should not discourage averaging in these
cases. If one credit generator could balance more than
2-137
-------
19 debit generators, *'-.a limit would ensure that the source
had credits to spare. However, it should be pointed out that
this is not the situation for which emissions averaging was
designed. The more likely situation is where a source finds
it more cost-effective to control some Group 2 points or
overcontrol other Group 1 points than it is to apply the RCT
to a Group 1 point that would otherwise be required. In other
words, averages will probably be constructed by identifying
debit generators first and then locating enough credit
generators to offset the debit generators' emissions.
The EPA does not agree that the implementing agency
would not bear much of the burden of averaging. The source's
effort to comply with monitoring, recordkeeping and reporting
requirements will be matched equally by the agency's oversight
and approval. Nor is future inclusion of equipment leaks in
averages a sufficient reason to not restrict averages. The
limit addresses present concerns. If equipment leaks can be
addressed in averaging at a later date, the limit may be
reexamined at that time.
Comment: Two commenters (A-90-19: IV-K-55; IV-K-64)
proposed that the number of emission points, as well as which
points can be included in the emissions average, should be
determined by the State. One of the commenters (A-90-19:
IV-K-64) elaborated that the State implementing agency had the
best information on what it could monitor or enforce.
Two commenters (A-90-19: IV-K-30; IV-K-37) identified
administrative issues that the EPA should address if the
number of points is limited. One commenter (A-90-19:
IV-K-42) proposed that points to be included in averages be
selected on the basis of number of applicable MACT standards,
the intermittent nature of operations, the ease or difficulty
of controlling the unit, the size and type of control
selected, and the proximity of the points.
Several commenters (A-90-19: IV-K-2; IV-K-7; IV-K-19;
IV-K-21; IV-K-22; IV-K-25; IV-K-27; IV-K*28; IV-K-33; IV-K-34;
IV-K-42; IV-K-46; IV-K-47; IV-K-49; IV-K-62; IV-K-64) argued
that there is no rational basis upon which to select points
2-138
-------
for averaging, and that the EPA's proposal of 5 to 15 percent
of the total points appeared to be arbitrary.
Response ; A substantial restriction on the rule's
implementation such as placing a limit on the number of points
to be allowed should not be left to the State without
providing them proper authority in the rule itself. The new
requirement of a numerical limit provides that authority.
In response to the first of three issues concerning two
commenters, as explained in section 2.3.4 of this BID volume,
all emission points except for equipment leaks are appropriate
for emissions averaging at any source subject to the rule. If
the source has more than 20 to 25 points that they wish to
include in an average, the source should choose the ones that
offer the greatest cost savings, operating flexibility, or
that will ensure ongoing compliance.
The other two issues regarding how to change averages
are specifically addressed and detailed in §63. 151 (i) of the
final rule. Further discussion of the procedure for making
changes to averages may be found in section 2.8.6 of this BID
volume. Furthermore, to limit the inclusion of certain points
based on their characteristics under case-by-case decisions as
one commenter suggested would add unreasonable complexity for
both the source and the implementing agency.
Finally, the EPA disagrees that the selection of
20 points, 25 if pollution prevention is used, has no rational
basis. The EPA submits that the reasoning presented in the
previous response, that any more than 20 to 25 points is
untenable from an enforcement perspective, is wholly rational
and defensible.
Two commenters (A-90-19: IV-K-30; IV-K-37)
urged the EPA to limit the use of emissions averaging by
establishing a mass emission limit, not by limiting the number
of emission points.
One commenter (A-90-19: IV-K-10) asserted that
wastewater emissions should be excluded from emissions
averaging instead of limiting the number of points, as
wastewater emissions are difficult to quantify, and the
2-139
-------
reference ccatrol technology, steam stripping, will not
guarantee consistent, quantifiable HAP removal.
Response; The EPA considers that placing a mass limit
on a source would be difficult to enforce and also add
complexity to the rule. This was the experience in past
situations where the total mass of emissions was limited, and
where the limit was established by predictions of allowable
emissions from anticipated operating rates. And as stated
previously, all emission points except for equipment leaks are
considered appropriate for emissions averaging, including
wastewater emission points. The suitability of wastewater
emission points for averaging is discussed in greater detail
in section 2.3.4 of this BID volume.
Comment; Two commenters (A-90-19: IV-K-1; IV-K-17)
argued that no emission points should be averaged, stating
that instead emissions averaging should be prohibited. Each
of the commenters (A-90-19: IV-K-1; IV-K-17) maintained that
emissions averaging was not enforceable, and that the public
health would be endangered by allowing averaging, even across
a small number of points. One of the commenters (A-90-19:
IV-K-17) further counselled that if emissions averaging were
allowed at all, the States would be pressured to include too
many points in the average, resulting in public exposure to
toxic pollutants.
Response: The appropriateness of including emissions
averaging in the final rule is discussed throughout this BID
volume. Averaging is enforceable and protective of public
health as it is an equivalent alternative to compliance on a
point-by-point basis. By limiting the number of points
allowed and by requiring risk or hazard equivalency, there
will be no significant difference in the emissions and the
risk or hazard under averaging versus compliance without
averaging, even across 20 to 25 points. Finally, with this
new provision, a State cannot be pressured into allowing more
than 20 to 25 points; such an average cannot be approved at
all.
2-140
-------
2.8.6 Title V/Implementation Plan
Comment; Two commenters (A-90-19: IV-D-72; IV-D-106)
stated that the requirements of Title V of the Act will assure
the enforceability of emissions averaging.
Response: Proper implementation of the monitoring,
recordkeeping and reporting, and compliance provisions of the
rule will ensure enforceability. Part 70 operating permit
programs will likely be the vehicle by which the rule is
implemented, but an operating permit program, in and of
itself, will not ensure enforceability. Hence, the provisions
of the HON are as detailed as they are to establish the
applicable requirements the rule places on subject sources.
Comment; Seven commenters (A-90-19: IV-D-9; IV-D-85
and IV-F-7.39 and IV-F-12; IV-D-117 and IV-F-7.43; IV-F-1.5;
IV-D-118; IV-D-124; IV-D-125) objected to the feature in the
emissions averaging proposal that would allow sources to
change their emissions averaging scheme at any time. Two of
the commenters (A-90-19: IV-D-85 and IV-F-7.39 and IV-F-12;
IV-F-1.5) complained that allowing sources to change their
Implementation Plans without prior approval of the State
regulatory agency or opportunities for public comment could
allow sources to change their Implementation Plans after
violations had occurred in order to avoid detection after the
fact. Four commenters (A-90-19: IV-D-9; IV-D-118; IV-D-124;
IV-D-125) stated that this feature makes it difficult to know
what plants have committed to do, and asserted that no State
will be able to effectively monitor all the game-playing under
the rule. Another commenter (A-90-19: IV-D-85 and IV-G-6)
recommended a correction for this feature.
Response; The final rule has been revised to require
approval for changes to emissions averages after the
Implementation Plan has already been approved. As specified
in §63.151(i) of the final rule, a planned change cannot be
made until a written update requesting the change has been
submitted and approved. Two kinds of changes may be approved
after they occur: (1) a change in the Group status of any
emission point in an average caused by a process change; and
2-141
-------
(2) a change in parameter value such that the value is
outside the ran .-. specified in the Implementation Plan and the
change causes a decrease in the projected credits or an
increase in the projected debits. These two kinds of changes
must be reported within 90 days after the change becomes known
to the source or in the next Periodic Report.
The update to the Implementation Plan would need to
include a . projection of debits and credits based on the
changes, and would need to demonstrate that credits will still
balance debits. The reason the two kinds of changes can be
reported af they occur is that a source may not know that
they have oc arred until after the fact. For example, if a
process change is made, the process vent TRE must be
recalculated to determine Group status, but measurements
needed to perform the calculation cannot be made until after
the process change is completed. If either kind of change is
not approved, the source may be found in violation.
If the commenter who recommended requiring enforceable
limitations on each point was referring to a limitation on
total emissions, such a provision was not included in the rule
because a mass emission limit could restrict production. The
intent of the NESHAP program is to require the maximum
achievable level of control on emission points, but not to set
a limit on production or prohibit production increases.
Therefore, the form of the standard—specified percent
reductions from each kind of emission point—is more
appropriate. This emission standard established in the rule
must certainly be considered an enforceable limitation. The
averaging Implementation Plan must specify the reductions to
be achieved on each point in the averages, and the provisions
of §63.151(i) fully address the process for making and
approving changes to an emissions average. Public review and
comment are not warranted for approval of Implementation Plan
updates because the Implementation Plan is a temporary
document that is only relevant until a source's operating
permit, which is subject to public review is approved. If an
operating permit is already in place, and a source wants to
2-142
-------
make a change in their emissions average, they would need to
follow the procedures specified in the operating permit
program.
Comment; One commenter (A-90-19: IV-D-85) was
concerned that the current proposal may allow the evaluation
and approval of emissions averages in Implementation Plans
before the information necessary to check credits and debits
estimates is provided. The commenter (A-90-19: IV-D-85)
complained that plant operators are not clearly required in
§63.151 of the proposed rule to complete the performance test
and measurements before the Implementation Plan is submitted.
The commenter (A-90-19: IV-D-85 and IV-G-6) stated that the
Implementation Plan must be accompanied by the results of all
applicable performance tests in order to make meaningful
evaluation of emissions estimates technically possible.
The commenter (A-90-19: IV-D-85) did not consider the
subsequent Notification of Compliance Status, which does
include the results of performance test, as a sufficient
substitute for proper information at the time a State
evaluates an Implementation Plan, which should be prior to the
actual compliance date. The commenter (A-90-19: IV-D-85)
complained that the Notification of Compliance Status creates
no opportunity for a public hearing and no obligation for the
EPA or the State to approve or disapprove a plan, hence, the
entire procedure is contrary to the intent of Congress as
expressed in Title V of the 1990 amendments to the Act.
Response; Contrary to the commenter's claims,
§63.151(d)(6) through (d)(8) of the final rule stipulates all
of the information that is needed to check the estimates of
projected debits and credits and that must be submitted in the
Implementation Plan.
It is true that performance tests are not to be
performed by the time the Implementation Plan (or an operating
permit application, for that matter) must be submitted. It is
not appropriate to require results of performance tests before
obtaining Plan approval and receiving permission to construct
controls because a source would not have applied controls at
2-143
-------
the time the Plan is due. Hence, the performance test results
are not required until the Notification of Compliance Status
is due. If the test results at this time indicate that the
source is not operating according to its Implementation Plan,
and that debits and credits do not balance, the source will be
in violation if it begins or continues such operation. Such a
scenario should result in an enforcement action.
The emissions averaging Implementation Plan must be
approved by the implementing agency before the source can
proceed. However, the opportunity for public notification and
review of the average is at the time an operating permit
application is being reviewed. The operating permit
application will also have to be approved by the implementing
agency, and therefore, the entire procedure has been designed
according to the intent of Title V of the Act.
Comment; One commenter (A-90-19: IV-D-74) complained
that the provision in §63.151(b)(2)(ii) of the proposed rule
requiring that the Initial Notification be submitted 180 days
in advance of construction or reconstruction could produce
substantial delays for some projects, depending on the
definition of construction and reconstruction. The commenter
(A-90-19: IV-D-74) noted that the Initial Notification
requires some substantial technical information and is
required under proposed §63.151(c)(2)(i) and (ii) to be
submitted with the Implementation Plan if the source elects to
use emissions averaging. The commenter (A-90-19: IV-D-74)
was concerned however, that the specific control technology
for each point and the definition of each point, for that
matter, may not be specified 180 days in advance, and
construction permits may be granted based on a generic control
efficiency requirement without specification of device, hence,
the decisions on control device may be made after construction
has commenced.
Response: In §63.151(b)(2)(i), the final rule states
that existing sources shall submit the Initial Notification
within 120 days after the date of promulgation. Furthermore,
the substantial technical information referred to by the
2-144
-------
commenter, specifically control technologies and
identification of individual points, is not required in the
Initial Notification and instead must be included in the
Implementation Plan. The commenter's concern about the
provisions for new sources is no longer applicable because new
sources cannot use averaging as discussed in section 2.3.2 of
this BID volume.
The Implementation Plan for existing sources that plan
to use emissions averaging must be submitted at least
18 months prior to the compliance dates specified in
§63.100(k) of subpart F of the final rule. For this
submittal, it is true that the provisions require submittal of
necessary technical information prior to construction of some
controls. The intent of the provisions is to require advanced
planning for emissions averaging, which is subject to
approval. It is not possible for a source or an implementing
agency to determine whether an emissions average will balance
without knowing specifics of the emission points and planned
control devices. Furthermore, a source would need to know
detailed control specifications well in advance of the
compliance date in order to have the controls constructed,
installed, and operating by the compliance date.
Comment: One commenter (A-90-19: IV-D-74) complained
that the provision in §63.151(d)(2) of the proposed rule
requiring a projection of debits and credits does not specify
the level of detail needed to substantiate the projection.
The commenter (A-90-19: IV-D-74) recommended that instead of
submitting masses of supporting data, the source can maintain
such data and should only be required to project debits and
credits one or two years in the future. The commenter
(A-90-19: IV-D-74) further recommended that sources should be
allowed and encouraged to base projected averages on the years
immediately prior to the averaging request.
In contrast, one commenter (A-90-19: IV-D-85)
suggested that a system that depends on predictions of future
emissions is inherently unenforceable, and that even if plant
operators could estimate emissions perfectly, they could
2-145
-------
plausibly second-guess their estimations in court and thwart
any attempt at showing a violation of the average.
Response; The rule specifies in great detail the
information required to be submitted in the Implementation
Plan. The source must supply estimates of all values needed
to check the estimates of projected debits and credits. The
emissions estimates for averaging submitted in the
Implementation Plan can be updated if needed according to the
procedures of §63.151(i) of the final rule, so the projections
can be based on one or many years. It would not be surprising
for the projections to reflect the operation from preceding
years; in some cases, the values for making emission estimates
can be drawn from historical operating rates.
However, compliance for an emissions average is not
based on projections of future emissions. As stated
previously, the Implementation Plan can only contain estimates
of future instead of actual emissions because the source has
not begun operating according its averaging plan. Once the
source must be in compliance, the average must be calculated
from records of actual production. If the average does not
balance each quarter or for the year, the actual data from the
compliance period for the average will be the basis for
enforcement actions, not the estimates in the Implementation
Plan.
2.9 RISK AND INTERPOLLUTANT TRADING
2.9.1 Risk in Emissions Averaging
Comment; Three commenters (A-90-19: IV-D-90;
IV-D-100; iv-D-115) were concerned that emissions averaging
will not provide sufficient public health protection. Two
commenters (A-90-19: IV-D-90; IV-D-100) were concerned that
the EPA did not assess the health risks of emissions averaging
or consider the public health and environmental impacts of not
controlling or undercontrolling HAP emissions. The commenters
(A-90-19: IV-D-90; IV-D-100) considered the absence of a
requirement to determine the potential health impacts from a
facility that averages to be a significant diversion from
State program requirements as well as sound public health
2-146
-------
policy, and contended that any emissions increase associated
with averaging must be accompanied by a public health and
environmental impact analysis.
Response; As stated at proposal, the emissions
averaging program is designed to result in equal or lesser
total emissions from any one source compared to point-by-point
compliance with the rule. At first, the EPA reasoned that
because of the residual risk evaluation required under
section 112 (f) of the Act, a source would have an incentive to
avoid increases in emissions of highly toxic HAP's.
However, the comments received were sufficient to
convince the EPA that a demonstration of risk or hazard
equivalency is warranted when using averaging. The EPA agrees
that since emissions averaging is allowed as an alternative
compliance option, it must represent an equivalent strategy in
more aspects than just mass reductions. The EPA also
recognizes that even though the overall health impacts may not
change, risk or hazard that is decreased through averaging at
one source cannot be viewed as balancing the possible
increased risk or hazard from averaging at another source.
Finally, the EPA acknowledges that many States already
have programs for considering risk, or hazard in HAP control,
which are suitable for evaluating emissions averages. Thus,
in the final rule sources are required to demonstrate to the
satisfaction of the implementing agency that their use of
emissions averaging will not result in any greater risk or
hazard than compliance without averaging.
One commenter (A-90-19: IV-D-103) stated
that allowing industry to select points in a facility to
control or not to control in emissions averages may have
significant unintended effects on worker or community
exposures due to the locations of the emission points. One
commenter (A-90-19: IV-D-115) warned that overcontrolling a
point on one side of a facility and undercontrolling one on
another side may actually increase the impact at the off site
receptor. Another commenter (A-90-19: IV-D-85) stated that
HAP's emitted near a residence or worksite could be balanced
2-147
-------
with a point a mile away, which could pose a health threat.
Another commenter (A-90-19: IV-D-117 and IV-F-7.43) was
concerned that some emission points at or near ground level,
such as wastewater and solid waste impoundments, could be
undercontrolled, which could increase exposures close to a
source.
One commenter (A-90-19: IV-D-45 and IV-F-7.7)
suggested that emissions averaging would create high risk
areas around industrial areas, and Congress had deleted
language that would have created so-called "dead zones" around
facilities, so the EPA should not attempt to resurrect a
provision which was deleted by Congress.
Response; The EPA maintains that there is an equal
likelihood that the opposite of the situations described by
the commenters could occur in emissions averages as well. If
so, these situations would result in lower impacts and risk to
receptors near the source than if the source complied point by
point. However, the EPA agrees that trades should not result
in increased hazard or risk from any source. In the final
rule, the implementing agency can prevent any of the
situations described by commenters from occurring by
restricting or rejecting emissions averaging plans that do not
demonstrate hazard or risk equivalency to the EPA's
satisfaction.
It should be noted that one commenter's concern for HAP
emissions from solid waste impoundments is addressed because
solid waste is outside of the purview of this rule, and thus
could not be included in an average.
gommept; One commenter (A-90-19: IV-D-103 and
IV-F-7.5) maintained that the concept of trading toxic air
pollutants to gain some health or environmental advantage is
fundamentally flawed and has not been demonstrated to be an
efficient mechanism to achieve an environmental goal. The
commenter (A-90-19: IV-D-103) considered emissions averaging
flawed because it does not take into account a variety of risk
factors and their effects.
2-148
-------
Response; The health and environmental benefit
associated with the reduction in HAP emissions that will be
achieved by the implementation of this rule nationwide is
substantial. Emissions averaging will enable this same
benefit to be realized at a lower cost and with greater
flexibility for the eligible sources. The EPA maintains that
the use of innovative strategies such as emissions averaging
that take cost or market principles into account is
appropriate for achieving environmental goals.
The EPA acknowledges that the averaging program that
was proposed without provisions for considering risk or hazard
was deficient. Risk equivalency must now be demonstrated
according to State procedures or Federal guidelines that will
published.
Comment; One commenter (A-90-19: IV-D-117) summarized
a modeling exercise they performed regarding emissions
averaging and interpollutant trading utilizing four example
emission points, where two points were assigned varying
emissions reductions and two points had no emissions
reduction. The commenter concluded from their study that
there was a residual risk of greater than 1 in 1,000,000.
Hence, the commenter (A-90-19: IV-D-117) stated that
emissions averaging and interpollutant trading will result in
greater risk to citizens than compared to controlling all
points within a facility.
Response: The commenter did not provide enough
supporting information for the analysis they performed to
respond adequately to specifics of the claim. In any case,
risk analysis is highly dependent on site-specific
assumptions. Thus, a case study could be formulated just as
easily to show lower risks after averaging and interpollutant
trading. No single hypothetical situation can be used to
generalize about the outcome from use of emissions averaging
in the limited way allowed under the final rule. However, the
final rule does require that no emissions average can result
in greater risk or hazard than control without averaging.
2-3,49
-------
2.9.2 Interoollutant Trading
Comment:; Several commenters (A-90-19: IV-D-9;
IV-D-10; IV-D-11; IV-D-35; IV-D-41; IV-D-49; IV-D-51; IV-D-70;
IV-D-72; IV-D-85 and IV-F-7.39 and IV-F-12 and IV-G-6;
IV-D-87; IV-D-90; IV-D-93; IV-D-94; IV-D-96; IV-D-99;
IV-D-100; IV-D-103 and IV-F-7.5; IV-D-103 and IV-F-7.40;
IV-D-106; IV-D-115; IV-D-117 and IV-F-7.43; IV-D-118;
IV-D-120; IV-D-124; IV-D-125; IV-F-1.5; IV-F-7.1; IV-F-7.2;
IV-F-7.6; IV-F-7.7; IV-F-7.10 and IV-F-9; IV-F-7.23;
IV-F-7.26; IV-F-7.27-and IV-F-10; IV-F-7.29; IV-F-7.35)
objected to the interpollutant trading feature in the
emissions averaging proposal for one or more of the following
reasons: (1) it does not take toxicity into account; (2) it
would endanger public health; (3) it is not protective of
worker health and safety; (4) it is inappropriate given
various limitations in the scientific understanding of
pollutants; and (5) because of the variability of effects, the
pollutants covered by the HON are not fungible and cannot be
treated as such.
In contrast, several commenters (A-90-19: IV-D-32;
IV-D-33; IV-D-57; IVrD-58; IV-D-62; IV-D-68; IV-D-69; IV-D-73;
IV-D-74; IV-D-77; IV-D-79; IV-D-82; IV-D-83 and IV-F-1.3 and
IV-F-5; IV-D-92; IV-D-98; IV-D-104; IV-F-1.6 and IV-F-6;
IV-G-1) supported the use of nonweighted emissions averaging,
whereby HAP's from a source may be averaged on a one-to-one
basis without regard to toxicity. Seven commenters (A-90-19:
IV-D-32; IV-D-57; IV-D-69; IV-D-77; IV-D-79; IV-D-86;
IV-D-104) submitted that the EPA should not restrict averaging
of different pollutants, so long as the pollutants are listed
in section 112(b) of the Act. Six commenters (A-90-19:
IV-D-62; IV-D-68; IV-D-73; IV-D-83 and IV-F-1.3 and IV-F-5;
IV-D-98; IV-F-1.6 and IV-F-6) claimed that an emission-
weighted scheme would add complexity to the program making it
very difficult to implement and determine compliance with the
HON. One commenter (A-90-19: IV-G-1) warned that toxicity
screens would consume intolerable resources and make averaging
impractical. Two commenters (A-90-19: IV-D-32; IV-D-57)
2-150
-------
cited various limitations in the scientific understanding of
pollutants, which they claimed makes adjusting trades
according to toxicity impossible or inappropriate.
Two commenters (A-90-19: IV-D-58; IV-D-62) contended
that there is no.evidence that facilities may choose to
overcontrol less hazardous pollutants and undercontrol more
hazardous pollutants. One commenter (A-90-19: IV-D-62)
stated that operational considerations govern a facility's
process management decisions and there is no incentive in the
rule to engage in aberrant behavior. Another commenter
(A-90-19: IV-G-17) suggested that a decision to average will
be based on technical and economic criteria, and toxicity is
not a factor in whether an emission point is difficult and/or
costly to control. Six commenters (A-90-19: IV-D-32;
IV-D-57; IV-D-62; IV-D-73; IV-G-1; IV-G-17) claimed that
sources have other strong incentives to control the most
hazardous emissions such as: protecting community and worker
safety; increasing product safety; controlling pollution in
other media; addressing State air toxics laws; reducing
residual risks; and upcoming rulemakings as in section 112(g).
Response: As stated at proposal, the EPA considers it
appropriate to allow interpollutant trading, i.e., to allow
emissions of different HAP's to be included in emissions
averages. To restrict averaging to only points emitting the
same HAP would be excessively restrictive in this industry
where emission streams are generally a mixture of different
HAP's. The requirement in the final rule of a risk or hazard
equivalency demonstration should help to allay concerns for
public health and welfare. Worker health and safety continues
to be guarded by other Federal statutes, and allowing
averaging of different HAP's will not compromise that
protection.
The EPA is also sensitive to the charges that a HAP-
speciated averaging system would consume additional resources
and increase the administrative burden for both sources and
implementing agencies. However, many States already require
risk or hazard examinations, and so would not consider the
2-151
-------
demonstration of risk or hazard equivalency an additional
burden. Moreover, the limit on the number of points that can
be included in averages should minimize any additional burden
and cost.
The EPA agrees with the claims that sources have no
incentive to propose emissions averages that could increase
risk or hazard, and stated as much at proposal. However, the
EPA was equally persuaded that a source's decision to average
will be based largely on technical and economic criteria, and
so recognized the necessity of elevating risk or hazard as a
consideration in averaging as well. If sources will control
the most hazardous emissions first for the reasons commenters
stated, then they need not fear that a risk or hazard
examination would severely limit their averages.
The EPA acknowledges that some limitations in the
scientific understanding of HAP toxicity exist. However, the
EPA does not believe the limitations are substantial enough to
make interpollutant trading impossible or to bar implementing
agencies from making adequate risk and hazard evaluations.
2.9.3 Legal Issues
Comment: Four commenters (A-90-19: IV-D-49; IV-D-85
and IV-F-7.39 and IV-F-12; IV-D-96; IV-F-1.5) maintained that
there is no legal justification for interpollutant trading.
Two commenters (A-90-19: IV-D-49; IV-D-85 and
IV-F-7.39 and IV-F-12) argued that Congress carefully
considered the issue of interpollutant trading when it passed
the 1990 Amendments and did not authorize it under section
112(d) of the Act as amended. One of the commenters (A-90-19:
IV-D-85 and IV-F-7.39 and IV-F-12) recommended that the EPA
should not put itself on scientifically untenable terrain by
allowing interpollutant trading without a clear legal mandate
to do so. The other commenter (A-90-19: IV-D-49) added that
section 112(g) of the Act permits trading between different
pollutants; however, Congress required that pollutant
reductions used to meet the offset provision come from
pollutants "deemed more hazardous" than the pollutant(s) being
offset. The commenter (A-90-19: IV-D-49) further stated that
2-152
-------
section 112(g)(l)(B) specifically disallows interpollutant
trading in certain circumstances. The commenter (A-90-19:
IV-D-49) suggested that if the EPA were to apply the
section 112(g) provision that allows interpollutant trading to
section 112(d), such a program must conform to the 112(g)
restrictions that: (1) an increase in a HON pollutant must be
compensated for by a reduction of equal or greater amount of a
HON pollutant of greater toxicity; and (2) pollutants for
which the EPA cannot demonstrate a safety threshold can only
be traded by requiring an equal or greater reduction in the
amount of the same pollutant(s).
One commenter (A-90-19: IV-D-103) stated that the MACT
program, while technology-based, is driven by the principles
of protecting the environmental and human health. Another
commenter (A-90-19: IV-D-115) stated that although they
recognize that factors contributing to exposure and health
risks are generally associated with risk-based programs rather
than technology-based programs such as MACT, the standard
technology-based program requires control across the board,
and emissions averaging introduces the risk components.
Three commenters (A-90-19: IV-D-49; IV-D-51; IV-D-99)
were not convinced by the argument given in the proposal
preamble that the potential for additional regulation under
section 112(f)(2)(A) would provide sufficient incentive to
ensure that increases in the emissions of more toxic
substances do not occur. Three commenters (A-90-19: IV-D-51;
IV-D-94; IV-D-99) stated it is more likely that owners or
operators will choose the most economical operation rather
than minimizing residual risks. One commenter (A-90-19:
IV-D-115) acknowledged that residual risk will ultimately be
examined but still considered 8 years of exposure
unacceptable.
In contrast, several commenters (A-90-19: IV-D-32;
IV-D-57; IV-D-58; IV-D-68; IV-D-73; IV-D-74; IV-D-82; IV-D-83
and IV-F-1.3 and IV-F-5; IV-D-98; IV-D-104; IV-F-1.6 and
IV-F-6; IV-G-1) considered any adjustment to emissions
averaging on the basis of risk or hazard to be inconsistent
2-153
-------
with the statutory requirement in section 112(d) of the Act to
base MACT standards on achievability of control technologies
instead of risk to public health or the environment.
Four commenters (A-90-19: IV-D-32; IV-D-57; IV-D-74;
IV-D-104) rejected restricting interpollutant trading on the
basis of risk. Three commenters (A-90-19: IV-D-32; IV-D-57;
IV-D-58) claimed that the statute conspicuously excludes risk
from the list of factors to be considered in the establishment
or implementation of MACT under section 112(d). Three
commenters (A-90-19: IV-D-32; IV-D-73; IV-G-1) asserted that
other sections of the Act, sections 112(g) and 112(i)(5),
specifically call for some consideration of risk while
section 112(d) does not, and these considerations should not
be imported where not authorized. Three commenters (A-90-19:
IV-D-82; IV-D-98; IV-G-1) claimed that the whole point of
technology-based standards was to replace the old risk-based
approach to HAP control. Four commenters (A-90-19: IV-D-32;
IV-D-57; IV-D-68; IV-D-104), referring to the proposal
preamble, pointed out that the EPA recognizes that it does not
have an appropriate scientific foundation on which to impose
risk-based limits on interpollutant trading.
Five commenters (A-90-19: IV-D-32; IV-D-74; IV-D-77;
IV-D-92; IV-D-104) noted that the Act requires MACT to control
categories of sources, not particular pollutants. Three
commenters (A-90-19: IV-D-32; IV-D-74; IV-D-104) claimed that
Congress intentionally changed the NESHAP program to control
sources, not pollutants; one of the commenters (A-90-19:
IV-D-32) referenced Senate Bill S.1630 in making.this claim.
Hence, two commenters (A-90-19: IV-D-32; IV-D-57) argued that
the proposed RCT's should apply to all SOCMI sources, even
though no two such sources emit precisely the same quantities
and mixes of pollutants.
One commenter (A-90-19: IV-D-82) interpreted the
legislative history of the 1990 amendments to the Act to
suggest that any analysis of emissions averaging should take
place when the MACT standard is set, and not on a case-by-case
basis. The commenter argued that once the EPA concludes that
2-154
-------
emissions averaging will produce better environmental results,
its use should be freely allowed without any subsequent case-
by-case assessment. Moreover, the commenter suggested that
once the assessment is made that emissions averaging will
reduce risk, the burden of showing otherwise should be placed
on those who oppose such a finding. The commenter added that
if such a showing is made, averaging should not be banned,
rather, safeguards should be imposed to assure that the trades
would be environmentally beneficial. In addition, the
commenter claimed that the legislative history suggests that
any assessment of non-technology factors should be more of a
"screening" review than the kind of hyper-detailed analysis
proposed by the EPA.
Ten commenters (A-90-19: IV-D-32; IV-D-57; IV-D-58;
IV-D-62; IV-D-68; IV-D-73; IV-D-74; IV-D-82; IV-D-83 and
IV-F-1.3 and IV-F-5; IV-D-104) pointed out that toxicity of
pollutants and risk will be taken into consideration when the
EPA addresses residual risk under section 112(f) of the Act.
Five commenters (A-90-19: IV-D-32; IV-D-57; IV-D-62; IV-D-73;
IV-D-82) suggested that because of section 112(f), sources
have little incentive for emissions averaging to result in
less risk reduction than application of RCT. "
Response; The EPA considers interpollutant trading to
be permissible under an emissions averaging program. The
floor determination and MACT standards under section 112(d) of
the Act are technology-based and are not based on an
evaluation of the relative toxicity of the pollutants being
emitted. Thus, even without averaging, the applicable
standards do not differentiate on the basis of the toxicity of
the pollutants being emitted and do not take into account the
differing composition of streams of pollutants being emitted
from the emission points subject to control. For example,
process vents are subject to the same MACT standards
regardless of the toxicity of the HAP or combination of HAP's
being emitted. As the basic standard is a technology-based
standard that does not vary with the specific HAP's being
emitted, there is no statutory requirement to impose
2-155
-------
restrictions on interpollutant trading when averaging is
permitted as an alternative compliance option.
While the EPA does not find that the Act requires that
interpollutant trading be restricted when averaging is
permitted, the EPA believes it has the authority under the Act
to establish provisions as part of the alternative averaging
system that wall assure that there is no increase in risk or
hazard as a result of a source's election of the averaging
mechanism. The fact that section 112(f) of the Act
contemplates that residual risk will be evaluated at a later
time and that other provisions specifically call for the
consideration of risk does not mean that the EPA is precluded
from considering risk or hazard in other contexts.
Consequently, the EPA maintains that it has the authority to
address risk and hazard in the averaging program through i
procedure such as the one adopted in the final rule—the
requirement that sources that elect to use averaging must
demonstrate, to the satisfaction of the implementing agency,
that compliance through averaging would not result in greater
risk or hazard than compliance without averaging.
With respect to the comments regarding section 112(g)
of the Act, the EPA notes that section 112(g) is designed to
fill a gap before a MACT standard is in effect and to minimize
increases in risk during that period. Consequently, once a
MACT standard has been promulgated, as is being done with the
HON, the provisions of section 112(g) do not limit
interpollutant trading. Moreover, the EPA is confident that
the requirement adopted in the final rule regarding the
demonstration by the source to the implementing agency that
compliance through averaging will not increase risk or hazard,
will ensure that the use of averaging by an individual source
will not result in an increase in risk or hazard attributable
to the emissions from that source.
2.9.4 Approaches for Toxicity Weighting
Comment; Four commenters (A-90-19: IV-D-70; IV-D-87;
IV-D-93; IV-D-99) stated that interpollutant trading should
not be allowed until a system is developed for ensuring that
2-156
-------
trading will not increase risk to public health. Two
conmenters (A-90-19: IV-D-51; IV-D-94) stated that toxicity
of the emitted HAP's must be taken into account for emissions
averaging to be as stringent as the non-averaging HON MACT
standard.
Three commenters (A-90-19: IV-D-70; IV-D-99; IV-F-7.6)
declared that if interpollutant trading must be a part of the
HON, the EPA must base it on a credible system for hazard
ranking that accounts for the varying quality of data
regarding health effects. Two commenters (A-90-19: IV-D-90;
IV-D-100) recommended that the EPA not propose emissions
averaging until an acceptable scheme that accounts for the
range of adverse health effects and exposures associated with
stationary source emissions has been developed, peer reviewed
by external scientists, and subject to public review process
through the Federal Register notification. The commenters
(A-90-19: IV-D-90; IV-D-100) also stated that a significant
amount of research is still required on the non-cancer health
effects of carcinogens before it can be assumed that emissions
averaging is, in fact, equivalent to controlling toxic
emissions. Two commenters. (A-90-19: IV-D-85; IV-D-99) urged
the EPA to take an extremely conservative approach to toxicity
weighting.
Response: The EPA agrees that emissions averaging
should not pose any increased health risk or hazard, a concern
which should be considered in determining whether an emissions
average is an equivalent compliance alternative. Hence, the
final rule was revised to require sources to demonstrate that
their use of averaging will not result in any greater risk or
increased hazard relative to compliance without averaging.
The use of the term "hazard" encompasses consideration of the
toxicities of the different HAP's.
The EPA maintains that adequate methodologies for
assessing and comparing risk or hazard are available.
Equivalency can be demonstrated according to either procedures
used by the implementing agencies or a Federal technical
support document that will be published.
2-157
-------
Comment: Several commenters (A-90-19: IV-D-45;
IV-D-51; IV-D-56; IV-D-58; IV-D-62; IV-D-72; IV-D-82; IV-D-85;
IV-D-87; IV-D-89; IV-D-99; IV-D-106; IV-D-115; IV-D-117 and
IV-F-7.43; IV-F-7.6) discussed approaches for adjusting for
toxicity in interpollutant trading.
Six commenters (A-90-19: IV-D-51; IV-D-85; IV-D-87;
IV-D-99; IV-D-115; IV-D-117 and IV-F-7.43) urged that a hazard
ranking or risk equivalency system should account for
carcinogenicity, non-carcinogenic toxicity, different exposure
pathways, target endpoin.:s; half-lives in the environment; and
bioaccumulation. One commenter (A-90-19: IV-D-87) stated
that credits should be adjusted based on a pollutant hazard
ranking rather than factors such as kinds of emissions, number
of emissions, weight of emissions released, or a lower factor
for other credit-generating activities. Three commenters
(A-90-19: IV-D-51; IV-D-85; IV-D-115) stated that debits of a
hazardous chemical should not be balanced by credits of a less
hazardous chemical. One commenter (A-90-19: IV-D-115) stated
that trades should not be allowed between carcinogenic and
non-carcinogenic compounds, nor between organic and inorganic
HAP's. Another commenter (A-90-19: IV-D-117 and IV-F-7.43)
submitted that points emitting carcinogenic, mutagenic, or
teratogenic pollutants should be controlled before non-
carcinogenic ones, but no trading of a HAP should be allowed
unless sufficient data is collected to fully characterize its
impact. One commenter (A-90-19: IV-D-99) presented a STAPPA
and ALAPCO resolution on Interpollutant Trading under
Title III of the 1990 Amendments, which recommended
characteristics that a credible hazard ranking system should
include.
Two commenters (A-90-19: IV-D-72; IV-D-106)
recommended that the EPA use the same process for developing
relative hazard potential for HAP's or one similar to the
process used to determine offsets under section 112(g).
Five commenters (A-90-19: IV-D-56; IV-D-58; IV-D-62;
IV-D-82; IV-D-89) suggested that one possible approach for
adjusting for toxicity could be based on the high risk
2-158
-------
pollutant list and toxicity-weighting factors used in the
EPA's Early Reductions Program promulgated pursuant to
section 112(i)(5). One commenter (A-90-19: IV-D-62)
considered this a superior approach to toxicity ranking
because it provides a consistent means of addressing the
"relative toxicity" issues across the many section 112
programs the EPA will be implementing.
Response: The EPA thanks commenters for their input on
this issue, and intends to take these and other
recommendations into account before issuing the final
technical support document for making hazard or risk
equivalency determinations. Some of the factors or procedures
may already be taken into account in existing State risk
evaluation procedures.
Comment: Five commenters (A-90-19: IV-D-70; IV-D-85;
IV-D-93; IV-D-99; IV-F-7.6) recommended the first approach for
toxicity weighting suggested in the HON proposal preamble,
which is based on the "more hazardous pollutant." One
commenter (A-90-19: IV-D-70) recommended the "more hazardous
pollutant" approach if a greater than 1:1 reduction is
required. Another commenter (A-90-19: IV-D-93) further
declared that the first approach must be incorporated if
interpollutant trading is allowed because emissions averaging
can only work if it provides not only for equal or greater
emission reductions, but also for equal or greater public
health protection.
In contrast, one commenter (A-90-19: IV-D-82) opposed
the first approach because of its complexity, arguing that
"hazard" is not a linear concept, rather, it depends on the
type and quality of the data and the type of health effect at
issue.
One commenter (A-90-19: IV-D-89) was inclined to
support the second alternative approach, which is based on the
"more hazardous quantity," and suggested that it could be
based on weighting factors similar to the Early Reductions
Program where each HAP is assigned a factor based on toxicity
and baseline.
2-159
-------
However, four commenters (A-90-19: IV-D-70; IV-D-82;
IV-D-99; IV-F-7.6) opposed the second approach discussed in
the proposal preamble for reasons including: (l) it would be
difficult to establish a more hazardous quantity without case-
by-case me ^-ling and review of impacts; (2) it relies far too
heavily or. che small data base that exists to characterize the
toxicity of pollutants; (3) it is virtually impossible to
determine factors for different toxic endpoints; and (4) it
would require a value judgement regarding which health effects
are most critical which is especially problematic because each
HAP has multiple health endpoints. One of the commenters
(A-90-19: IV-D-99) added that the fourth problem listed would
exist in the "more hazardous pollutant" approach as well,
although the uncertainty is smaller because the extra step of
deciding how much more of an emission is necessary to obtain a
more hazardous quantity is not required. Two of the
commenters (A-90-19: IV-D-99; IV-F-7.6) claimed that the EPA
has already determined the "more hazardous quantity" approach
to be flawed and is no longer considering it for inclusion
under the regulations for section 112(g).
One commenter (A-90-19: IV-D-82), who opposed both
approaches discussed in the proposal preamble, argued further
that emissions streams often occur as mixtures of different
HAP's, which makes determining the "more hazardous" of two
streams under the first approach as difficult as determining
the overall "hazard balance" under the second. The commenter
(A-90-19: IV-D-82) was concerned that in both cases, the
complexity and effort of performing the calculation and the
chances of disputes with regulatory agencies will deter
facilities from attempting an emissions trade.
Response; Two approaches for addressing toxicity and
hazard in interpollutant trading were presented at proposal
for comment. The first approach was to restrict averages
based simply on "the more hazardous pollutant." The second
approach proposed coupling "the more hazardous pollutant" with
the actual mass of emissions to establish "the more hazardous
quantity" as the basis on which to restrict interpollutant
2-160
-------
trades. As stated in the supplemental notice, after receiving
input from public comment, it was concluded that neither of
the two approaches could be developed in sufficient detail to
provide the basis for final rulemaking. This conclusion
stimulated the decision to seek additional comment on whether
to require risk or hazard determinations, which would be made
according to State procedures or a Federal technical support
document. As discussed previously, the final rule now
contains this provision for demonstrating risk or hazard
equivalency, and guidance for making these determinations will
be published.
Comment: Two commenters (A-90-19: IV-D-45; IV-D-51)
suggested that classes of chemicals be defined with a discount
factor assigned based on the relative toxicity between the
credits and debits, and suggested specific discounting
scenarios. One commenter (A-90-19: IV-D-115) suggested that
a discount factor could be used to mitigate the impact of
interpollutant trades, but added that the factors alone will
not guarantee that the impact does not increase.
Response; A credit discount factor has been included
in the rule for reasons other than to adjust for toxicity as
discussed in section 2.6 of this BID volume. Although the
form of a technical support document for making equivalency
determinations has not yet been established, the EPA does not
consider discount factors to be the appropriate method to
address interpollutant trading. This is not to say that a
risk or hazard weighting factor for different HAP's in the
form of a percentage will not be allowed, but to avoid
confusion, the term "discount factor" will not be associated
with the risk or hazard equivalency determination.
2.9.5 Problems with Toxicity Weighting
Comment; Several commenters (A-90-19: IV-D-32;
IV-D-51; IV-D-57; IV-D-85 and IV-F-12; IV-D-90; IV-D-99;
IV-D-100; IV-D-103 and IV-F-7.5; IV-D-103 and IV-F-7.40;
IV-D-115; IV-D-120; IV-F-7.6; IV-F-1.5; IV-F-7.27 and IV-F-10)
cited various limitations in the scientific understanding of
pollutants including: varying health and environmental
2-161
-------
effects; varying public health threat; the amount and quality
of data available to categorize risk; lack of toxicological
understanding of toxicity mechanisms; the lack of a
methodology to compare such dissimilar pollutants; or the lack
of a means of equating various toxic end-points including
immunotoxicity, fetotoxicity, reproductive and developmental
toxicity, and others that can vary according to age, sex, and
other factors.
Response; It is acknowledged that the scientific
understanding of the many aspects of HAP toxicity listed by
commenters is incomplete in certain respects. However, the
EPA does not agree that the limitations are so great as to
prevent agencies from making acceptable risk or hazard
equivalency comparisons for the purpose of averaging. A
number of States have designed and implemented their own
programs requiring risk assessments of sources before
approving permits to operate. In some cases, these programs
have been in place for a number of years.
The States have drawn on EPA and other Federal guidance
and their own resources to make risk or hazard determinations
and comparisons in permitting sources of HAP emissions. The
EPA agreed with comments included elsewhere in 'this section
that States that already have their own programs for
evaluating risk should be able to use them for emissions
averaging in the HON. Moreover, these States' experience in
addition to Federal experience and resources are transferable
to all other States in the form of guidance, which can
continually be updated as the scientific understanding
continues to improve.
One comment er (A-90-19: IV-D-85) contended
that even if a scientifically acceptable toxicity weighting is
possible, it would not provide a stable framework for
regulation because the weighting would have to be adjusted as
more is learned about the regulated pollutants.
One commenter (A-90-19: IV-F-7.5) noted that the
current methods for comparing carcinogens are so weak and
bounded by uncertainty that an efficient "marketplace" based
2-162
-------
just on cancer risk is inconceivable. One coramenter (A-90-19:
IV-D-103) claimed that the notion that toxicity can be equated
using some measure based on the risk-specific cancer potency
(i.e., the slope of the dose response curve for cancer) is
absurd. The commenter (A-90-19: IV-D-103) added that using
only cancer potency over the lifetime of an individual ignores
all the other health effects, interactive effects, and non-
human endpoints. The commenter (A-90-19: IV-D-103 and
IV-F-7.5) stated that because changing the ratio of chemicals
in complex mixtures can change the health effects, allowing
trades based on one valuation, even if it were accurate for
cancer, ignores the possibility of other effects being induced
due to the change. The commenter (A-90-19: IV-D-103 and
IV-F-7.5) predicted that in a very few years, through advances
in molecular biomonitors, it will be possible to ascribe cause
and effect to particular diseases and chemical exposures, and
so trading pollutants will in time be traceable in terms of
the mixture of chemicals which contributes to the diseases.
Another commenter (A-90-19: IV-F-7.40) added that to
account for toxicity, all exposure routes must be known,
including the atmospheric direct exposure route and the
atmospheric deposition routes onto soil, water, and food that
reach human and nonhuman endpoints. However, the commenter
(A-90-19: IV-F-7.40) maintained that this data is also not
available, so a toxicity-based approach is not possible.
Response; The first commenter's concern emphasizes the
advantage of issuing guidance for making risk or hazard
determinations at this time instead of including provisions
for adjusting interpollutant trades in the formal rulemaking.
As new information is made known, implementing agencies can
take advantage of it immediately without having to amend
rules.
The concerns about an efficient "marketplace" and
adjustments to interpollutant trading based on cancer risk but
not other health endpoints, and exposure routes that are not
taken into consideration are understandable for large
quantities and complex mixtures of HAP emissions. However,
2-163
-------
because the scope of emissions averaging has been limited to
decrease administrative complexity, the emissions associated
with averaged points will also be limited compared to the
source as a whole. Hence, there should be little cause for
the specific concerns stated here especially with the added
provision to consider risk or hazard.
2.9.6 Inclusion of Risk in Averaging Determinations
Comment; One commenter (A-90-19: IV-D-99) on the
proposed rule urged that the regulation in no. way prohibit
State and local agencies from requiring risk assessments or
other procedures as part of the process for reviewing
averages.
Seven commenters (A-90-19: IV-K-1; IV-K-10; IV-K-30;
IV-K-37; IV-K-44; IV-K-55; IV-K-64) supported the proposal in
the supplemental notice to require sources that elect to use
averaging to demonstrate, to the satisfaction of the agency
implementing the rule, that compliance through averaging would
not result in greater risk than compliance without averaging.
The commenters (A-90-19: IV-K-1; IV-K-10; IV-K-30; IV-K-37;
IV-K-44; IV-K-55; IV-K-64) maintained that consideration of
risk would better ensure that public health is protected.
In contrast, several commenters (A-90-19: IV-K-2;
IV-K-7; IV-K-14; IV-K-19; IV-K-20; IV-K-21; IV-K-22; IV-K-25;
IV-K-26; IV-K-27; IV-K-33; IV-K-34; IV-K-39; IV-K-42; IV-K-46;
IV-K-47; IV-K-48; IV-K-49; IV-K-54; IV-K-56; IV-K-61; IV-K-62;
IV-K-66) opposed the proposal to require risk equivalency
demonstrations in emissions averaging.
Several commenters (A-90-19: IV-K-2; IV-K-14; IV-K-19;
IV-K-21; IV-K-22; IV-K-25; IV-K-26; IV-K-27; IV-K-33; IV-K-34;
IV-K-39; IV-K-42; IV-K-45; IV-K-46; IV-K-47; IV-K-49; IV-K-54;
IV-K-56; IV-K-62; IV-K-66) argued that it is inappropriate to
address risk under a section 112(d) standard, as the Act
specifies that the NESHAP are to be technology-based, and risk
determinations should be made under section 112(f) following
the implementation of the technology-based standards.
Four commenters (A-90-19: IV-K-34; IV-K-39; IV-K-46;
IV-K-48) objected to the risk equivalency demonstration
2-164
-------
requirement on the basis that there is no compelling evidence
that the use of emissions averaging will create additional
risk.
Six commenters (A-90-19: IV-K-6; IV-K-26; IV-K-27;
IV-K-34; IV-K-46; IV-K-50) opposed requiring a risk
equivalency demonstration because it would be burdensome to
the source. Four of the commenters (A-90-19: IV-K-6;
IV-K-26; IV-K-34; IV-K-46) further stressed that the burden
posed by risk equivalency demonstrations would be a
disincentive for using emissions averaging.
Three commenters (A-90-19: IV-K-19; IV-K-27; IV-K-54)
stated that the EPA and the National Academy of Science are
reviewing current risk assessment approaches as required by
section 112(0), and that it would be useless to base a
regulation on current methodologies, as they are likely to
change as a result of these studies.
Three commenters (A-90-19: IV-K-21; IV-K-27; IV-K-66)
•advocated the development of a threshold level below which
risk equivalence would not need to be demonstrated. Two
commenters (A-90-19: IV-K-34; IV-K-46) complained that there
is no threshold of risk provided in the risk equivalence
proposal, even small changes in risk would need to be
addressed, regardless of the insignificance of the baseline
risk level or of the change in risk.
Response; The EPA introduced the issue of including
risk determinations in averaging to a large part in response
to public comment such as the first commenter. The support
for the proposal in the supplemental notice was sufficient to
warrant revising the final rule to include provisions
requiring risk or hazard equivalency demonstrations for
emissions averages.
It is appropriate to introduce the consideration of
risk in emissions averaging. The floor and the RCT's for the
rule were determined without any consideration of risk. On
the other hand, averaging represents an alternative to the
technology-based system of point-by-point compliance, and as
an alternative, must be demonstrated to result in equivalent
2-165
-------
control. This demonstration can consider risk without
violating the intent of section 112(d) of the Act.
It is possible that in some cases having to make a risk
equivalency demonstration may so increase the cost of
averaging that it is no longer more cost-effective to average,
but the EPA does not think this is likely in most cases
because of the limited size of most averages. Even though it
is difficult to predict whether averaging would be more likely
to increase or decrease risk, any possibility of increased
risk would represent HAP control that is not completely
equivalent to point-by-point compliance.
The Act contains no requirement that risk or hazard
considerations be delayed until after the study of risk
assessment by the National Academy of Sciences is completed.
By issuing guidance, the suggested methodologies and
procedures can be revised when the study becomes available.
The statutory requirements and deadlines remain in effect, the
study notwithstanding.
The issue of threshold levels can be addressed in the
guidance as well or be determined by the implementing agency.
If the agency is satisfied that a de minimis risk level should
be established, it can be included in individual State
programs.
Comment; Seven commenters (A-90-19: IV-K-1; IV-K-17;
IV-K-44; IV-K-49; IV-K-55; IV-K-63; IV-K-66) recommended that
instead of requiring a risk equivalency demonstration, the EPA
should eliminate emissions averaging to protect public health.
The commenters (A-90-19: IV-K-1; IV-K-17; IV-K-44; IV-K-49;
IV-K-55; IV-K-63; IV-K-66) cautioned that there is no adequate
methodology for assessing risk, and that the lack of
information regarding exposure to toxic chemicals prevents the
determination of toxic equivalency for the purpose of
averaging HAP's. Four of the commenters (A-90-19: IV-K-1;
IV-K-30; IV-K-44; IV-K-55) supported the risk equivalency
demonstration, but preferred that emissions averaging be
eliminated.
2-166
-------
Response; As just discussed throughout the previous
sections, the EPA considers risk assessment methodologies and
toxicological information to be developed sufficiently to make
adequate risk and hazard equivalency determinations. The
rationale for allowing the use of emissions averaging is
repeated throughout this BID volume. The, flexibility that is
afforded sources by its use is important, and the safeguards
such as this new requirement and others maintain the program
as a reasonable and responsible one.
Comment; Eight commenters (A-90-19: IV-K-2; IV-K-10;
IV-K-18; IV-K-19; IV-K-28; IV-K-37; IV-K-40; IV-K-64) stated
that the EPA should provide standard guidance for risk
equivalency determinations. One commenter (A-90-19: IV-K-10)
listed items that the EPA guidance should address. Another
commenter (A-90-19: IV-K-28) recommended that the guidance
should resemble the EPA's air quality modeling guidance, which
defines different approaches applicable in different
circumstances. Two commenters (A-90-19: IV-K-37; IV-K-55)
stressed that States should be involved in developing the
guidance. One commenter (A-90-19: IV-K-44) declared that EPA
guidance should establish a presumptive minimum process that
States must meet in assessing the risk equivalency
demonstration.
Response: A Federal technical support document will be
available after rule promulgation. As discussed previously,
all of the recommendations received will be taken into careful
consideration in drafting the guidance, and if need be further
input from different resources such as State agencies may be
solicited at a later time. The EPA will not establish a
presumptive minimum process for making determinations,
however. The provisions of the final rule are that risk or
hazard equivalency demonstrations are to be made to the
satisfaction of the implementing agency. As such, the process
is left entirely at the discretion of the implementing
agencies. They are free to use whatever methodologies and
procedures they choose including the guidance to be issued.
2-167
-------
Comment; Nine commenters (A-90-19: IV-K-7; IV-K-9;
IV-K-19; IV-K-30; IV-K-37; IV-K-39; IV-K-44; IV-K-55; IV-K-64)
agreed that State and local agencies should be allowed to use
or develop their own policies and assessment tools for
analyzing risk equivalence. One commenter (A-90-19: IV-K-19)
favored allowing the source to select whether to comply with
State or Federal risk assessment guidelines.
One commenter (A-eo-19: IV-K-39) noted that states
should be authorized to set a ceiling on risk by using their
existing fenceline monitoring programs for BACT for toxics
(T-BACT) or related assessments. The commenter (A-90-19:
IV-K-39) asserted that risk assessment could not be based on a
single emission point, but must be based on all points.
One commenter (A-90-19: IV-K-35) preferred to address
the issue of averaging with different pollutants as discussed
in the original proposal, according to the relative hazard of
the pollutant.
Response: The EPA appreciates the support for the
position that States should be allowed to continue using
already established processes or to select and develop their
own programs. The source is not allowed to follow examples in
the Federal technical support document over established State
procedures unless the implementing agency provides for such a
choice.
However, a State cannot use the HON as its authority to
place a ceiling on risk unless the risk associated with point-
by-point compliance is the ceiling to which the commenter is
referring. That is, as long as a source can demonstrate to
the satisfaction of the agency that an emissions average poses
no greater risk or hazard than if the same points had been
controlled point-by-point, the average could be approved. As
su risk assessments would not be based on a single point,
bv rather all the points included in an average.
Comment; Four commenters (A-90-19: IV-K-18; IV-K-30;
I*; :<-37; IV-K-64) supported requiring the identification of
al- HAP's in emission streams when assessing risk. Three of
the ommenters (A-90-19: IV-K-30; IV-K-37; IV-K-64) stated
2-168
-------
that the expertise to speciate HAP's in emission streams is
available, and that it is currently required for permits,
annual emission fees, and major source applicability
determinations. The commenters (A-90-19: IV-K-30; IV-K-37;
IV-K-64) further argued that speciating HAP's would not entail
new data collection and would not be overly burdensome. One
of the commenters (A-90-19: IV-K-37) considered speciation of
HAP's necessary to reduce public exposure to risk.
In contrast, two commenters (A-90-19: IV-K-21;
IV-K-26) opposed a requirement to require the identification
of all HAP's in assessing risk. Four commenters (A-90-19:
IV-K-21; IV-K-26; IV-K-54; IV-K-55; IV-K-62) stated that it
would be very burdensome to identify and quantify all the
HAP's in an emission stream, because of: (1) the variability
of feedstocks contributing to a varying pollutant content;
(2) the many different analytical methods required; or (3) the
absence of approved methods thus requiring the development of
individual methods of identification.
Two commenters (A-90-19: IV-K-28; IV-K-50) questioned
whether complete speciation was technically feasible. One of
the commenters (A-90-19: IV-K-28) noted that speciation from
reactor vents and for HAP's below detection limits may be
impossible. Two commenters (A-90-19: IV-K-21; IV-K-66)
maintained that although high concentrations may be
measurable, it would be difficult to measure small
concentrations. The commenters (A-90-19: IV-K-21; IV-K-66)
noted that reactions in sample containers, interference in the
gas chromatographic method, and problems with analytical
sampling methods could prevent accurate measurement of HAP
concentrations.
One commenter (A-90-19: IV-K-54) advocated minimizing
the number of HAP's that a source must test for. The
commenter (A-90-19: IV-K-54) suggested, for example, if a
tank only has nonhalogenated VOC, a source should not have to
test for halogenated VOC, semi-volatiles, or particulate
matter. The commenter (A-90-19: IV-K-54) also proposed that
a source should not have to speciate HAP's if the control
2-169
-------
technology utilized was known to control other HAP's of the
same type.
Response; In the supplemental notice, the EPA
recognized that to satisfy an implementing agency that an
averaging plan would not increase risk, a source might have to
identify and quantify all the HAP included in the average.
Hence, comment was requested on whether identifying all the
HAP's in the emissions streams would pose difficulties for
sources, and, if so, what those difficulties would be.
It is acknowledged that individual HAP's emitted in
mixtures must already be identified in some States' permitting
programs and that individual HAP's must be identified to a
certain extent to make the risk equivalency demonstration.
However, the EPA also recognizes that it may not be
technically feasible to identify HAP's at levels below some
minimal concentration and that establishing some level for
identification may be in order or allowing engineering
judgement, in some cases. The concerns over HAP
identification have been noted and will be considered in
developing the guidance.
2.9.7 Broader Scope for Averaging
Comment; One commenter (A-90-19: IV-D-83 and IV-F-1.3
and IV-F-5) stressed that if a broader emissions averaging
program that allows averaging across different source
categories is adopted, it should remain a nonweighted scheme
or have safeguards that are easy to implement.
Two commenters (A-90-19: IV-D-58; IV-D-73) suggested
that if a broad averaging program is adopted, it may need to
take toxicity of different HAP's into account.
Two commenters (A-90-19: IV-D-58; IV-D-83 and IV-F-1.3
and IV-F-5) suggested that risk-neutral averaging among
sources and HAP's outside the scope of the HON could be based
on the existing list of 47 "high risk" pollutants and
associated weighting factors included in the Early Reductions
rule under section 112(i)(5) of the Act.
One commenter (A-90-19: IV-D-83) stated that under a
broader averaging scheme, the EPA's proposal to limit trading
2-170
-------
to organic HAP's covered under the HON could be workable; but
added that the EPA should specify that for future MACT
standards covering organic and inorganic HAP's, one-for-one
averaging among inorganic HAP's would also be allowed. The
commenter (A-90-19: IV-D-83) added that trading between
organic and inorganic HAP's could also be addressed by the
high risk pollutant list and weighting factors discussed
previously.
Response; As discussed in section 2.4 of this BID
volume, broader emissions averaging, i.e., averaging between
sources and source categories, is not allowed in the final
rule. As such, these comments are no longer applicable. It
is reemphasized that although the HON can be regarded as
precedent-setting, it should not be assumed that any elements
of this rulemaking will be duplicated in other standards. Nor
should it be assumed that any aspect of the HON imposes any
restrictions on the design and development of future rules.
2.10 BANKING
2.10.1 General Issues
Comment; Six commenters (A-90-19: IV-D-51; IV-D-85
and IV-F-7.39 and IV-F-12 and IV-G-6; IV-D-99; IV-F-1.5;
IV-F-7.1; IV-F-7.33) opposed allowing banking of emission
credits.
One commenter (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12) stated that emissions credit banking violates the
maximum achievable emission reductions standard by reducing
future performance based on past differences between actual
and allowable emissions, and that sources will plan less
carefully when they have emissions credits in the bank. The
commenter (A-90-19: IV-D-85) stated that banked credits will
also reflect actions taken to comply with State rules, many of
which are stricter than the proposed HON standard. The
commenter (A-90-19: IV-D-85) stated that plants with credits
in the bank will rely on Implementation Plans, which appear
adequate on the surface but which the plant operator knows
will likely prove inadequate. The commenter (A-90-19:
IV-D-85) added that if the State suspects the standards are
2-171
-------
not being achieved, the company will come forward with
windfall banked credits.
One commenter (A-90-19: IV-D-51) contended that
banking of credits is unacceptable until the EPA develops
guidance on acceptable ambient exposure levels of these
chemicals. Another commenter (A-90-19: IV-D-99) opposed
banking because of administrative difficulties, negative
effects on future emission reductions, and a possible increase
in public exposure to toxic emissions.
Several commenters (A-90-19: IV-D-32; IV-D-33;
IV-D-48; IV-D-50; IV-D-58; IV-D-62; IV-D-69; IV-D-72; IV-D-73;
IV-D-74; IV-D-78; IV-D-79; IV-D-82; IV-D-83 and IV-F-1.3 and
IV-F-5; iv-D-86; IV-D-92; IV-D-106; IV-D-108; IV-F-1.6 and
IV-F-6; IV-G-1) supported emissions banking.
Two commenters (A-90-19: IV-D-83 and IV-F-1.3 and
IV-F-5; IV-D-92) considered banking an essential element of a
successful averaging program. One commenter (A-90-19:
IV-F-1.6 and IV-F-6) stated that banking offers facilities
needed flexibility in conducting operations that may vary with
annual changes in business and productivity and further
claimed that banking would encourage early implementation of
emission controls. Another commenter (A-90-19! IV-D-58)
stated that banking rewards facilities that reduce emissions
early in averaging program.
Three commenters (A-90-19: IV-D-32; IV-D-78; IV-D-92)
asserted that banking will increase the likelihood of success
of the averaging program by providing a safety valve for
unexpected events that may throw a balance off. Another
commenter (A-90-19: IV-D-48) suggested that emissions will
tend to be less than under RCT as sources adopt more stringent
controls to accumulate banked credits. One commenter
(A-90-19: IV-D-62) supported banking because of the
flexibility it provides in compliance and for the
environmental benefit.
One commenter (A-90-19: IV-D-58) suggested that
allowable emission limits in operating permits will limit
emissions of any particular source, therefore the use of
2-172
-------
banked emission credits will not jeopardize environmental
protection in any emissions averaging program.
Response; Banking of extra credits generated in one
compliance period for use in a future compliance period is not
allowed in the final rule. Several commenters mentioned the
likelihood of significant administrative burden resulting from
tracking the generation and use of banked credits, which was
the primary reason for not including the proposed banking
provisions. With the goal in mind of keeping the
administration of the rule as simple as possible, credit
banking represents a complication that would affect the source
and implementing agency alike. Another reason for deleting
banking from the final rule was the possibility that
communities near sources could experience peak HAP exposures
if banked credits were allowed to offset unexpected increases
in emission debits. Any additional flexibility offered by
banking is offset by the increased administrative burden and
potential for peak exposures such that little overall
advantage can be gained from allowing credit banking.
The EPA disputes contentions that banking of credits is
essential to emissions averaging. Simply allowing emissions
averaging as an alternative to comply with the rule provides a
great deal of flexibility in and of itself. Several
commenters stated that the annual compliance period is
sufficient to accommodate the normal fluctuations in operating
rates and unexpected events (see section 2.7.1 of this BID
volume for summaries of these comments). So, further
insurance from banked credits is not necessary.
Several commenters argued that the prudent source will
include a margin of safety in their credit/debit balance in
order to ensure compliance (see section 2.6 of this BID volume
for summaries of these comments). The EPA agrees that to do
so is prudent and anticipates that sources will present
conservative emissions averaging plans for approval. Again,
with the prospect of such safety measures promised by industry
sources, banking of credits was deemed an unnecessary
provision providing little or no additional advantage overall.
2-173
-------
It is possible that allowable emission limits
established in operating permits could be used in some cases
to prevent peak HAP emissions from occurring due to the use of
banked emission credits. However, not all sources will have
allowable emission limits established in their operating
permits, nor does this rule require that specific numerical
emission limits be applied. Moreover, some facilities located
in areas in attainment of NAAQS may not be required by Federal
or State rules to limit emissions either. Therefore, unless a
permitting authority requires allowable emission limits in
operating permits, the commenter's claim that limits in
operating permits can afford environmental protection does not
apply. In any case, if a source does accept emission limits,
their ability to bank would be reduced so much as to make
banking virtually impossible to use.
2.10.2 Periodof Availability
Comment: One commenter (A-90-19: IV-D-50) suggested
that there be no time limit on the use of banked credits.
Several commenters (A-90-19: IV-D-32; IV-D-58;
IV-D-62; IV-D-69; IV-D-72; IV-D-73; IV-D-78; IV-D-83 and
IV-F-1.3 and IV-F-5; IV-D-89; IV-D-106; IV-G-1X proposed that
banked credits be available for five years or more. Five
commenters (A-90-19: IV-D-58; IV-D-62; IV-D-83 and IV-F-1.3
and IV-F-5; IV-D-89; IV-G-1) claimed that this will encourage
early reductions of HAP's. Three commenters (A-90-19:
IV-D-58; IV-D-62; IV-G-1) explained that the longer period
encourages sources to make extra reductions earlier because
they know the banked credits will be available for a longer
period; on the other hand, if banked credits expire too
quickly, there will be more incentive to postpone reductions
for as long as possible. One commenter (A-90-19: IV-D-62)
contended that this undermines the intent of the MACT controls
which should encourage emissions reductions as early as
possible. Two commenters (A-90-19: IV-D-32; IV-D-73)
suggested that because excess credits would be continually
generated, and in most years banked credits would not be
needed and would lapse, banking would result in additional
2-174
-------
emissions reductions compared to an emissions averaging
program without it. Two commenters (A-90-19: IV-D-72;
IV-D-106) stated that the banking provisions encourage sources
that use emissions averaging to generate more credits than
necessary to balance debits annually. One commenter (A-90-19:
IV-D-89) worried that a time range shorter than 5 years could
result in significant emission variations as industry rushed
to use banked credits. Another commenter (A-90-19: IV-D-78)
argued that the lower the allowable period for banking, the
more difficult it will be for sources to justify emission
reduction projects beyond that required for regulatory
compliance.
One commenter (A-90-19: IV-D-58) disagreed with
concerns that a five year banking period could interfere with
enforcement and recordkeeping. The commenter (A-90-19:
IV-D-58) argued that because credits can only be generated
from the time of MACT promulgation onward, historical data
should not be a problem. The commenter (A-90-19: IV-D-58)
further suggested that quarterly reporting will provide plenty
of notice of the need and use of banked credits in advance of
the annual compliance period, enabling verification in a
timely manner and also adequate time for enforcement actions
if needed.
Three commenters (A-90-19: IV-D-58; IV-D-74; IV-D-108)
suggested that the term of availability of banked credits
should be similar to banking provisions in other regulations,
such as PSD netting and future section 112(g) offset
provisions.
One commenter (A-90-19: IV-D-82) recommended allowing
reductions to be banked if they were achieved after enactment
of the 1990 amendments. One commenter (A-90-19: IV-D-69)
added that credit for previous reduction projects should be
available for at least five years after promulgation of the
rule.
Response; The proposal to allow banking of surplus
credits was not included in the final rule for the reasons
discussed in the previous response. Hence, comments regarding
2-175
-------
the period over which banked credits should be available are
no longer relevant.
2.10.3 Use for Quarterly Compliance
Comment: Four commenters (A-90-19: IV-D-32; IV-D-79;
IV-D-86; IV-D-92) urged the EPA to permit the use of banked
credits on a quarterly basis to reduce the likelihood of last-
minute compliance problems. One commenter (A-90-19: IV-D-32)
suggested that concerns about potential significant quarterly
spikes in excess debits could be addressed by allowing no more
than 10 percent of credits used in a quarterly average to be
banked credits.
One commenter (A-90-19: IV-D-32) stated that although
the "fixed-cap" quarterly average approach is preferred, it is
less flexible, and if the alternate "fixed cap" approach is
used, allowing the use of banked credits on a quarterly basis
would be important to restore flexibility.
In contrast, one commenter (A-90-19: IV-D-58) opposed
allowing the use of banked credits on a quarterly basis. The
commenter (A-90-19: IV-D-58) was concerned that if the
alternative quarterly emissions limit based on allowable
emissions is implemented, a compliance inconsistency could
result if banked emission credits are allowed for quarterly
compliance.
Response; As stated in the first response in this
section, surplus credits cannot be banked for future use.
Because the source has a full year over which to average
emissions, the prudent source will create more credits than
are needed early in the compliance period, which can be used
to balance unexpected debits later in the same compliance
period.
In general, the best way to reduce the likelihood of
last-minute compliance problems is to select emission points
whose operating histories can ensure reliable averages at any
given time. The EPA also agrees that it is prudent to
incorporate a safety margin of extra credits into averages.
Conservative planning should ensure that quarterly exceedances
beyond the 30 percent allowable quarterly debit exceedance
2-176
-------
will be avoided and that a source's compliance with the rule
will not likely be in danger.
2.10.4 Miscellaneous Issues
Comment ; One commenter (A-90-19: IV-G-1) stated that,
beyond the methods stated in the proposal preamble,
enforcement concerns regarding banking could be addressed by:
(a) requiring appropriately limited waiver of the statute of
limitations as a condition of averaging approval; or (b)
construing the "violation" to occur when a defective credit is
used.
One commenter (A-90-19: IV-D-87) stated that banked
credits should be discounted 20 percent per year.
Two commenters (A-90-19: IV-D-90; IV-D-100) were
concerned that EPA did not evaluate the risks associated with
increasing emissions of more toxic HAP's by banking credits.
Two commenters (A-90-19: IV-D-74 ; IV-D-108) suggested
that banked credits should be defined to be created on a
specific date, such as the last date of the quarter in which
they are banked.
Response: Although the suggestions advanced by one
commenter may satisfy the concern over the statute of
limitations, they do not address the other prob'lems that
banking poses for adequate enforcement of the rule. The fact
remains that allowing credit banking would increase the
complexity of the emissions averaging program and allow for
peak emissions. Hence, the EPA concluded that deleting the
provisions for banking produces the most workable structure
for the final rule, and because banking of credits for future
use is not allowed in the final rule, the remaining comments
are no longer applicable.
2.11 GENERAL POLICY AND MISCELLANEOUS ISSUES
2.11.1 Precedent for Future Rule Makings
one commenter (A-90-19: IV-D-58) supported
the emissions averaging program and stated it sets an
important precedent for subsequent MACT proposals on other
source categories.
2-177
-------
Five commenters (A-90-19: IV-D-85 and IV-F-7.39 and
IV-F-12; IV-D-90; IV-D-99; IV-D-100; IV-F-7.6) were concerned
that including the emissions averaging provisions could set a
dangerous precedent for future air toxics rules. One
commenter (A-90-19: IV-D-85 and IV-F-7.39 and IV-F-12) warned
that including emissions averaging in the HON will tarnish the
reputation of economic incentive programs, which have the
potential to contribute to environmental progress if designed
properly and applied appropriately to other kinds of
pollution.
Response: As stated in the proposal preamble, this
rule describes the first application of emissions averaging
for compliance with standards developed under section 112(d)
of the Act. Many interested groups and parties subject to
NESHAP have interpreted, and will continue to interpret this
rule as an indication of the types of requirements that may be
incorporated into future standards. However, although the EPA
does consider this rule to be precedent-setting, decisions for
future NESHAP must be made on a specific source-category
basis. If the use of emissions averaging is allowed for other
source categories, public comment will be solicited again as
part of that rulemaking effort. It should not be assumed that
emissions averaging will be allowed in other standards. Also,
where emissions averaging is included in future rules, it may
not be comprised of the same specific requirements as in this
rule.
Comment; One commenter (A-90-19: IV-D-103) discussed
basic conditions for a successful market-based trading system
for regulating environmental releases. The commenter
(A-90-19: IV-D-103) concluded that none of the conditions for
a successful market-based trading system are met in the
proposed averaging provisions and outlined why.
Response: The conditions outlined by the commenter for
a successful market-based trading system do not apply to the
emissions averaging program in this rule because emissions
averaging is not a market-based trading system in the sense
described by this commenter. In emissions averaging, trades
2-178
-------
occur strictly within the source; there is no opportunity to
"market" surplus emission reductions outside of the source.
The emissions averaging program in the rule does not meet
these conditions because the conditions do not pertain to this
type of program at all.
2.11.2 Simplifying the Language of the Rule
C9mpig]rt: Several commenters (A-90-19: IV-D-59;
IV-D-67; IV-D-68; IV-D-71; IV-D-83; IV-D-90; IV-D-100;
IV-F-1.1 and IV-F-3; IV-F-7.21; IV-F-7.36; IV-F-7.41) were
concerned about the complexity of the final emissions
averaging provisions. Two commenters (A-90-19: IV-D-72;
IV-D-106) urged the EPA to structure the program to maintain
maximum flexibility for sources.
Response; Allowing the use of emissions averaging
increases the flexibility of sources to comply with the rule
overall. However, any additional flexibility in emissions
averaging must be balanced by the need to maintain the
enforceability of the program without unduly burdening the
authorizing agency.
As discussed in section 2.8.2 of this BID volume, the
nature of emissions averaging requires some provisions for
monitoring, recordkeeping, and reporting that are not needed
under point-by-point compliance. Also, the calculation of
debits and credits has been specified in great detail to
ensure that emissions are estimated consistently. The
provisions are required to maintain enforceability.
The emissions averaging provisions have also been
modified to decrease complexity. Banking has been removed,
and credit for prior controls and averaging at new sources are
not allowed. With these simplifying changes, concerns about
the complexity of the proposed rule and the emissions
averaging program should be minimized.
Comment: Four commenters (A-90-19: IV-D-32; IV-D-62;
IV-D-64; IV-D-113) suggested that equations for emissions
averaging should not be written into the regulation. Two
commenters (A-90-19: IV-D-32; IV-D-62) were concerned that
emissions equations can become outdated quickly and that other
2-179
-------
docume-ts or publications containing the equations could be
revised more easily than the regulation.
Two commenters (A-90-19: IV-D-32; IV-D-73) recommended
restructuring the presentation of the averaging provisions to
make key subsections more prominent and identifiable by
relocating detailed calculation procedures and tables
referencing AP-42 to an appendix to the rule. The commenter
(A-90-19: IV-D-73) further recommended incorporating recent
changes to AP-42 in the suggested appendix to keep calculation
procedures up-to-date.
One commenter (A-90-19: IV-D-33) recommended that
tables 20 through 31 in proposed §63.150 should not be
promulgated as part of the HON; rather, they should be
referenced as coming from AP-42 because the information in the
tanles is changed periodically as the EPA updates AP-42
factors. The commenter (A-90-19: IV-D-33) further
recommended that a clarifying statement should be added to
§§63.150(f)(3) and (g)(3) of the proposed rule that the AP-42
tables that are to be used are those in existence at the time
the Implementation Plan or operating permit application is
submitted. The commenter (A-90-19: IV-D-33) suggested that
doing so would provide certainty that once the factors were
used, those factors would be the ones that would apply
throughout the term of an emissions average. However, the
commenter (A-90-19: IV-D-33) provided that as new
Implementation Plans or operating permit applications for
additional emissions averaging groups are submitted in the
future, tables and relevant factors present at that time
should be used.
Response; All tables in subpart G of the final rule
have been moved to the end of the subpart. The tables and
equations cannot be removed as suggested by commenters because
cross-referencing of other EP documents or publications is
not allowed. "--.stead, data - i equations obtained from other
Federal documc .cs must be du:.j.icated as part of the
regulation.
2-180
-------
It is not to the benefit of the source or the
implementing agency to revise a rule every time estimation
equations and factors change. If a rule were changed, a
source that was in compliance based on older estimation
methods could be judged out of compliance later. The Act
provides that NESHAP shall be reviewed and revised as
necessary no less often than every eight years. Significant
changes that may have accumulated can be incorporated at that
time.
In any case, the EPA is confident of the estimation
methodologies for the emission points subject to this rule.
The equations and data are not expected to change
substantially anytime in the future.
Comment; One commenter (A-90-19: IV-D-113) found that
when attempting to verify the equations in the emissions
averaging provisions, it was difficult and sometimes
impossible to cancel units. The commenter (A-90-19:
IV-D-113) cited an equation for process vents as an example of
this problem where a constant used undocumented units.
Response; The units of all constants have been
specified, and the units of measurement for the parameters in
the equations have been verified as appropriate in the final
rule.
2.11.3 The Intent of Section 63.112
Comment: Four commenters (A-90-19: IV-D-32; IV-D-64;
IV-D-73; IV-D-81) recommended clarifications of §63.112 of the
proposed rule.
Three commenters (A-90-19: IV-D-32; IV-D-73; IV-D-81)
recommended including in §63.112(c)(2) the statement that
emission points not included in an emissions average may
comply in accordance with §63.112(c)(1).
One commenter (A-90-19: IV-D-64) requested that the
EPA clarify that the intent of §63.112(c)(2)(ii) is to provide
sources the choice of emissions averaging or complying with
§§63.113 through 63.147 of the proposed rule. The commenter
(A-90-19: IV-D-64) further suggested that the calculation of
the allowable emission rate specified in proposed §63.112(a)
2-181
-------
and required by §63.112(c)(2) is unnecessary, because if the
source elects to use emission averaging, all it should be
required to do is specified in §63.150.
Response; Several commenters misunderstood the
provisions in §63.112 of the rule, so the provisions have been
revised to clarify their intent. As stated in the proposal
preamble, the rule establishes a control requirement for each
kind of regulated emission point in a source. However, to
facilitate the use of emissions averaging, it is necessary to
recognize that compliance by the source as a whole is
accomplished by achieving an allowable emissions level. This
allowable emissions level is the sum of emissions from all
points (excluding equipment leaks) in the source if the
required controls are applied, and the level is represented by
the equation of §63.112(a) of the final rule.
The equation of §63.112(a) is simply a mathematical
representation of the allowed emissions when a source complies
with the rule. The provisions now state that owners or
operators are not required to calculate the allowable
emissions level for compliance purposes. It has also been
made clear that the allowable emissions level is established
for a given collection of emission points and is never fixed.
The level represented by the equation of §63.112(a) will be
different from source to source, and the level for a
particular source can change if the number or kinds of
emission points constituting the source changes or as
production changes.
Section 63.112(c) of the final rule introduces the two
compliance options available only to existing sources: point-
by-point compliance or emissions averaging. The provisions
now state specifically that emissions averaging alone cannot
be used to comply with the rule. Emissions averaging is to be
used for groups of points, not all of the points in a source,
and compliance for the points not involved in emissions
averages will still be determined on a point-by-point basis.
Thus, using the emissions averaging compliance option
for some points does not require that the emissions for all
2-182
-------
points in the source must be quantified. Emissions
quantification is required only for the points included in
averages in order to calculate emission debits and credits,
and the quantification of debits and credits is based on the
equations in §63.150. Under point-by-point compliance, the
source needs only to install RCT where it is required. But
regardless of the compliance option that the source chooses
for each of its emission points, the source must still achieve
the overall emission level.
This last point must be emphasized. While the equation
of §63.112(a) of the final rule was included to enable the use
of emissions averaging, it represents the total emissions
allowed regardless of how a source complies with the rule. As
a representation of overall emissions, the equation need not
indicate how aspects of emissions averaging are to be
incorporated just as it does not indicate how point-by-point
compliance is to be incorporated. Instead, appropriate
references are made to the specific compliance provisions for
each kind of emission point and to the provisions for
emissions averaging, which are detailed in §63.150.
Comment: One commenter (A-90-19: IV-D-74) considered
the description of emissions averaging in the proposal
preamble and the actual language in the rule to be
inconsistent. The commenter (A-90-19: IV-D-74) stated that
in the proposal preamble, emissions averaging is described as
an "allowable emission level set for a given mix of emission
points," in other words, a fixed-mass cap or a fixed percent
reduction of overall emissions (57 FR 62613-14). The
commenter (A-90-19: IV-D-74) complained that the process-
specific emphasis in the language of the proposed rule would
be very difficult to implement in pharmaceutical
manufacturing, which is typically accomplished by batch
processes. The commenter (A-90-19: IV-D-74) was concerned
that the level of recordkeeping and analysis required in
emissions averaging for a batch processor would be far more
burdensome than for a continuous process operation if
2-183
-------
emissions averaging is expressed as debits and credits against
a process-specific allowable emission rate.
Response; The form of the standard is described as the
allowable emissions level set for a given mix of emission
points. The allowable emissions level represented by the
equation of §63.112(a) is not a "fixed cap," i.e., a limit on
total emissions. The rule does not in any way bar the source
from changing the number or kinds of emission points or
restrict their operation. The rule only requires that a
certain percent reduction be achieved at Group 1 points.
Thus, the rule does not limit how much can be emitted from a
point at any time? it only requires for Group 1 points that
the emissions be the residual from achieving the reference
control efficiency designated for that kind of point.
For example, a Group 1 process vent capable of emitting
10 tons of uncontrolled HAP emissions must apply RCT with a
reference efficiency of 98 percent and therefore, cannot emit
mere than 0.: tons of emissions. The rule does not bar the
source from increasing production at any time at the process
unit containing the Group 1 vent. If production is increased
such that the uncontrolled HAP emissions from the vent are
dc oled to 20 tons, the allowable emissions are also doubled
tr ).4 tons (2 percent of 20 tons). The allowable emissions
f -. the source are not fixed.
If this process vent were to be used as a debit
generator in an emissions average, production increases are
still allowed, and the emissions in this example are still
allowed to double from 0.2 tons to 0.4 tons. However, when
production is increased, the debit also doubles from 9.8 (10
minus 0.2) to 19.6 (20 minus 0.4) tons. The source is
required to maintain the average's balance after the
production increase. Hence, when debits increase, the source
must find an equal number of new credits. Thus, it can be
seen that the rule does require a fixed percent reduction of
emissions from applicable points, but the rule does not impose
a fixed-mass cap.
2-184
-------
The pharmaceutical manufacturing to which the ccamenter
refers is not subject to this rule, and the EPA cannot predict
at this time what future NESHAP may require for the
pharmaceutical industry. Process vents in SOCMI batch
operations are also not subject to this rule. Including other
batch emission points in averages and complying with the
monitoring, recordkeeping and reporting requirements is no
different than for continuous processes. However, if the
source considers emissions averaging burdensome for their
specific situation, point-by-point compliance may be the
preferred compliance option.
Comment : One commenter (A-90-19: IV-D-33) suggested
that the equation in proposed §63. 112 (a) improperly defines
terms SEPV1 , ZES^ and SETI^ , and that they should not be
preceded by numbers (0.02, 0.05, and 0.02 respectively).
Response: The allowable emissions from Group 1 process
vents, storage vessels and transfer racks are correctly
written in the equation of §63. 112 (a) of the final rule. By
themselves, SEPV^ ZES,, and SETR, represent the sum of
uncontrolled emissions from the respective Group 1 points.
The numbers preceding these terms in the equation denote the
percent reductions required for each kind of Group 1 point,
and it is correct mathematically to place the numbers outside
the summation symbol. Thus, the terms as written in the
equation are properly defined as the sum of residual emissions
from all such points in a source.
2.11.4 Emissions Estimation
One commenter (A-90-19: IV-D-85) predicted
that even if "representative operating .conditions" were
precise and never varied, enforcement of the emissions
averaging program would still be inadequate because the
procedures for estimating credits and debits invite gaming.
The commenter (A-90-19: IV-D-85) warned that a wide menu of
estimation techniques are allowed in many situations, and
operators are not required to use the most accurate
techniques. The commenter (A-90-19: IV-D-85) argued that the
same technique and assumptions should be used on all emission
2-185
-------
points to the extent that is technically practicable, and
conservative assumptions (i.e., low for credits, high for
debits) should be required in making estimates. Otherwise,
the commenter (A-90-19: IV-D-E5) predicted that operators
could substitute lower estimates of emissions on high debit-
generating points through substitution of some other technique
(e.g., historical flow rates from units that have produced
less in the past) and boost paper credit generation through
use of another technique (e.g., flow rates based on design
capacity for wastewater streams) on another point.
Response; Similar charges were made in the second
comment in section 2.3.4 of this BID volume regarding allowing
averages between different kinds of emission points. The
response to the comment in section 2.3.4 pertains to all of
the claims made here as well.
Comment; Two commenters (A-90-19: IV-D-34; IV-D-78)
noted that the overall source emission limit equation in
§63.112 and the debit and credit equations in §63.150 assume
that Group 1 emission points are all controlled to the rated
RCT efficiencies (e.g., 95 percent or 98 percent). The
commenters (A-90-19: IV-D-34; IV-D-78) suggested that for the
facility that elects to use the outlet concentration option
(e.g., 20 ppmv) or other EPA-approved control technologies,
the EPA should state that these and other equations based on
the assumption of 95 percent or 98 percent control should be
modified to include the 20 ppm component where appropriate
because they may not need to achieve the full 95 percent or
98 percent removals in these cases.
Response; The commenters noted correctly that the
equations of §§63.112 and 63.150 of the rule do not reflect
the option to control to a exiting HAP concentration of 20
parts per million by volume.
It is not necessary to revise the general equation of
§63.112(a) of the final rule, which is not specifically used
for calculations. The option of controlling to 20 parts per
million by volume is not an issue for emissions averaging
either. When a Group 1 point is left uncontrolled as a debit
2-186
-------
generator, it simplifies matters to use the RCT's nominal
efficiency to calculate debits in all cases. When a Group 2
point is controlled to generate emission credits, the percent
reduction must be determined to calculate the credits,
regardless of the exit concentration that results from
control.
Comment; One commenter (A-90-19: IV-D-32) suggested
that the equation in §63.150(f)(2)(ii) of the proposed rule
for calculating uncontrolled emission rates from process vents
is invalid as written because it includes a temperature
adjustment that is not needed since the vent stream flow rate
and HAP concentrations are already expressed at standard
conditions. The commenter (A-90-19: IV-D-32) recommended
rewriting the equation to remove the temperature correction.
Response; The commenter's observation is correct, and
the temperature parameter has been removed from the equation
of §63.150(9)(2)(ii) °f the final rule.
Comment; One commenter (A-90-19: IV-D-74) suggested
that ambient annual temperature is improperly used in
§63.150(f)(3)(i) of the proposed rule for calculating storage
tank emissions because no provision is made for indoor storage
tanks, where color is not particularly relevant and the
ambient temperature depends on the air conditioning. The
commenter (A-90-19: IV-D-74) further recommended that the
calculation allow the source the option to take advantage of
cold outdoor temperatures in winter by using the average
monthly ambient temperature, corrected for tank color. The
commenter (A-90-19: IV-D-74) reasoned that with this option,
the advantage of overcontrolling a tank farm in the summer is
greater than it is in the winter, which could be a relevant
matter for some plants' production schedules.
Response; Ho specific provision was made for indoor
storage vessels because it is unlikely that SOCHI vessels with
capacities greater than 20,000 gallons (the vessels likely to
be classified as Group 1) are located indoors. The commenter
is correct that if a storage vessel is located indoors, the
source probably will not need to correct the average storage
2-187
-------
temperature (Tg) for the tank paint color. The final rule now
provides that in cases where a vessel is located indoors, the
paint factor (Fp) can be taken to be 1.
The issue for indoor storage vessels is not so much
whether the tank paint color is relevant in calculating
emissions. The main issue is whether a storage vessel will
still experience a diurnal temperature change even though it
is located indoors. It is possible that the air temperature
may not be regulated where the vessel is located, and the
vessel may still experience a diurnal temperature change. In
this case, breathing losses will still occur and must be
estimated for emissions averaging. However, if by being
indoors, Tg can be held constant, breathing losses would not
have to be calculated.
It would not be to the advantage of most sources to use
average monthly ambient temperatures for calculating credits
from storage vessels. The commenter is correct that storage
vessels will have greater emissions in the summer and hence
can generate more credits in the summer. But, the opposite is
true during the winter when the vessels would generate fewer
credits. This would have no effect on the annual compliance,
of course. However, the difference could cause the source to
be out of compliance with the quarterly emissions limit. In
this case, the use of the annual average temperature can
protect the source from potential quarterly compliance
violations.
In effect, the use of the annual average temperature
averages the emissions from storage vessels over the
compliance period. There should be no difference between
total annual emissions calculated using the annual average and
the monthly average temperature. The use of the annual
average ambient temperature in calculating debits and credits
for storage vessels is more appropriate because it simplifies-
emissions averaging calculations. It frees the source and the
authorizing agency from contending with variations in debits
and credits that would result from seasonal temperature
fluctuations.
2-188
-------
Comment; One commenter (A-90-19: IV-D-62) claimed
that the emissions averaging equations for fixed-roof storage
vessels are outdated since the EPA has just issued a new
section on "Storage of Organic Liquids" in their AP-42
document with new equations different from those in the
proposed rule.
Response; A provision was added to the final rule
allowing the use of updated AP-42 equations for estimating
evaporation (breathing) losses from fixed-roof storage
vessels. Breathing losses must be estimated in emissions
averaging to calculate the total uncontrolled emissions from a
fixed-roof storage vessel. The updated procedures and
equations have been incorporated by reference from American
Petroleum Institute Publication 2518, which contains them in
the identical form to that of AP-42. The provision also
stipulates that if the updated equations are to be used, they
must be used for all of the storage vessels to be included in
an emissions average as debit or credit generators. The new
equations cannot be used for some vessels, and the equations
in the rule used for other vessels at the same time.
2.11.5 Miscellaneous Issues
Comment: One commenter (A-90-19: IV-D-85 and IV-G-6)
considered exemptions from control based on cost effectiveness
incompatible with emissions averaging, which does not require
control of any particular emission point. Hence, the
commenter (A-90-19: IV-D-85) concluded that including
emissions averaging in the rule eliminates the justification
for cost-effectiveness cutpoints. The commenter (A-90-19:
IV-D-85) suggested that if the supposed flexibility of
emissions averaging produces significant benefits through
technological innovation and pollution prevention, then these
improvements should provide sufficient extra reductions to
offset the small quantities of emissions allowed through
reasonable exemptions. The commenter (A-90-19: IV-D-85)
declared that to the extent emissions averaging becomes part
of the final rule, any cost-effectiveness exemptions are
arbitrary and capricious.
2-189
-------
Moreover, the commenter (A-90-19: IV-D-85) considered
the imposition of outpoints one of their worst fears about
emissions trading. The commenter (A-90-19: IV-D-85)
suggested that to the extent credits are allowed for exceeding
standards, regulated companies and ideologically committed
government agencies have an incentive to weaken standards to
fuel trading. The commenter (A-90-19: IV-D-85 and IV-G-6)
declared that relaxing stringency through exemptions in order
to encourage trading is unacceptable.
Response; The EPA considers it consistent with the Act
to use applicability criteria to distinguish Group 1 and 2
points that are subject to different levels of control.
Discussion of the applicability criteria for group status is
included in section 5.2 of BID volume 2D. Cost effectiveness
was used along with other factors to determine the control
options above the MACT floor for the different kinds of
Group 1 points. The EPA considers basing applicability on
cost effectiveness, among other criteria, to be consistent
with the Act and compatible with emissions averaging as well.
The use of cost effectiveness to determine control options
above the floor is similarly discussed in section 5.2 of BID
volume 2D.
On the average for the industry, Group 1 points can be
controlled more cost-effectively than Group 2 points.
However, some sources may have Group 1 points that are much
more expensive to control than the national average.
Emissions averaging is provided for these select cases where
it is more cost-effective to control some Group 2 points to
achieve the required emission reductions.
Emissions averaging provides sources the flexibility to
comply with the rule in a less costly manner on a site-
specific basis. But, the possibility that some Group 2 points
may be less costly to control than on average does not mean
that more cost-effective control of all Group 2 points is
possible. The use of emissions averaging will not eliminate
the difference on average in the cost effectiveness of
controlling Group 1 and 2 points. The reasons for designating
2-190
-------
Group 2 points still exist for the majority of points that are
not involved in emissions averaging.
Comment; One commenter (A-90-19: IV-D-86) stated that
additional flexibility is required for an emissions averaging
program to be viable for batch operators because a wide range
of different products is made in the same batch equipment,
which can result in changing, intermittent emissions. The
commenter (A-90-19: IV-D-86) suggested that a significant
amount of emissions from batch operations results from the
equipment cleaning that is required at changeovers from
manufacturing one product to the next. The commenter
(A-90-19: IV-D-86) stated that in many cases, however, batch
operators have the flexibility to campaign runs of a specific
product; for example, instead of manufacturing a product
during the first week of each month of the year, the operator
may be able to manufacture over a 12-week period and inventory
the product. The commenter (A-90-19: IV-D-86) explained that
campaigning product runs minimizes changeovers and hence
clean-outs and their attendant emissions; emissions during
production do not change whether the product is made
intermittently or all at once, while overall emissions are
reduced. The commenter (A-90-19: IV-D-86) complained that
with quarterly limits or a compliance period shorter than a
year, campaigning product runs could result in a violation
because the emissions are concentrated in one quarter even
though campaigning lowers overall emissions. The commenter
(A-90-19: IV-D-86) submitted that the flexibility needed by
batch operators can be accomplished through the provisions
recommended by the CMA.
Response; The EPA appreciates the commenter's concerns
for flexibility in emissions averaging, but a number of
factors argue against redesigning the emissions averaging
program to further accommodate batch operations. In the first
place, there are few batch processes that are subject to the
rule. Most batch operations are associated with source
categories other than SOCMI. It also does not appear that a
batch operation can contribute much to an emissions average.
2-191
-------
A large portion of emissions from batch operations, emissions
from batch process vents, are not subject to the rule and
hence, are not eligible for emissions averaging.
The point is noted that campaigning batch product runs
may reduce overall emissions by reducing cleanings at
changeouts, and that emissions averaging may discourage
production campaigning. It is not desirable that emissions
averaging should ever encourage greater emissions than would
otherwise occur. The EPA encourages the commenter to submit
data illustrating the significance of emissions from cleanings
for use ir, future standards affecting batch operators.
Cleaning batch process equipment is also considered a
maintenance turnaround, so any wastewater that may be
generated by the cleaning is not subject to subpart G.
Rather, such wastewaters are subject to the provisions of
subpart F, and their emissions are not suitable for averaging.
The batch operator must determine which compliance
alternative is best for their site-specific situation, point-
by-point compliance or emissions averaging as provided in the
final rule. If the batch operator can associate the operation
of their credit-generating points with their debit generators,
then regardless of how emissions may be concentrated in a
particular period, the average will always stay balanced.
Otherwise, product campaigns in batch operations may not be
compatible with emissions averaging.
2-192
-------
TECHNICAL REPORT DATA
frlease read Instructions on llic went be/ore completing)
1 REPORT NO
EPA-453/R-94-003C
4 TITLE AND SUBTITLE
Hazardous Air Pollutant IBiaaiona from Proc
Chaaical Manufacturing Industry — Backgroune
volum* 2C: cooMnti on Baiaaiona Avaraginc
7 AUTHOR(S)
9 PERFORMING ORGANIZATION NAME AC,
Office of Air Quality Planr
U.S. Environmental Protect:
Research Triangle Park, Noi
12. SPONSORING AGENCY NAME AND ADO
Director, Office of Air Qu«
Office of Air and Radiatioi
U.S. Environmental Protect.
Research Triangle Park, Noi
2- 3 RECIPIENTS ACCESSION NO
5. REPORT DATE
•aa Unita in th« Synth.trc Organic MAlTCh 1994
1 Information fo«- »tn«l st»ndarda ...
( 6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO
D ADDRESS 10. PROGRAM ELEMENT NO
ling and Standards
ion Agency n. CONTRACT/GRANT NO
rth Carolina 27711 68D10117
RESS 13. TYPE OF REPORT AND PERIOD COVERED
ility Planning and Standards
1 14. SPONSORING AGENCY CODE
ion Agency EPA/200/04
rth Carolina 27711 «»/ *««/««
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A final rule for the regulation of emissions of organic hazardous air
pollutants (HAP's) from chemical processes of the synthetic organic chemical
manufacturing industry (SOCMI) is being promulgated under the authority of
sections 112, 114, 116, and 301 of the Clean Air Act, as amended in 1990. The
emission standards were proposed in the Federal Register on December 31, 1992
(57 FR 62608). Public hearings were held. A supplemental notice was
published in the Federal Register on October 15, 1993 (58 FR 53478). This
volume of the background information document summarizes all comments and
presents the agency's responses on emissions averaging.
17.
a. DESCRIPTORS
Air pollution
Pollution control
SOCMI
Hazardous air pollutant
National impact*
18. DISTRIBUTION STATEMENT
KEY WORDS ANO DOCUMENT ANALYSIS
b. IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air pollution control
19. SECURITY CLASS (Tins Report) 21. NO. OF PAGES
208
20. SECURITY CLASS , This page i 22. PRICE
UNCLASSIFIED
EPA f<»m 2220-1 (R«v. 4-77) previous EDITION is OBSOLETE
------- |