EPA-453/R-95-014
United States Office of Air Quality
Environmental Protection Planning and Standards
Agency Research Triangle Park, NC 27711
EPA^53/R-95-014
July 1995
Air
FEDERAL STANDARDS FOR MARINE TANK
VESSEL LOADING OPERATIONS AND
NATIONAL EMISSION STANDARDS FOR
HAZARDOUS AIR POLLUTANTS FOR MARINE
TANK VESSEL LOADING OPERATIONS
Technical Support Document for Final
Standards:
Summary of Public Comments and
Responses
ENVIRONMENTAL
PROTECTION
AGENCY
DALLAS, TEXAS
-------
FEDERAL STANDARDS FOR MARINE TANK
VESSEL LOADING OPERATIONS
AND
NATIONAL EMISSION STANDARDS
FOR HAZARDOUS AIR POLLUTANTS FOR
MARINE TANK VESSEL LOADING OPERATIONS
TECHNICAL SUPPORT DOCUMENT FOR
FINAL STANDARDS
g
O
-------
FEDERAL STANDARDS FOR MARINE TANK VESSEL LOADING OPERATIONS
AND NATIONAL EMISSIONS STANDARDS FOR HAZARDOUS AIR POLLUTANTS FOR
MARINE TANK VESSEL LOADING OPERATIONS--
TECHNICAL SUPPORT DOCUMENT FOR FINAL STANDARDS
Summary of Public Comments and Responses
Emission Standards Division
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
July 1995
-------
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, North Carolina 27711;
from the Office of Air Quality Planning and Standards Technology
Transfer Network, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina 27711; or, for a fee, from the
National Technical Information Services, 5285 Port Royal Road,
Springfield, Virginia 22161.
Publication No. EPA-453/R-95-014
iii
-------
ENVIRONMENTAL PROTECTION AGENCY
Background Information for Final Standards
Marine Tank Vessel Loading Operations
Prepared by:
Jordan /" (Date)
Director, Emission Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
1. The final federal standards and national emissions standards
limit emissions of VOC and HAP from existing and new marine
tank vessel- loading operations. The final standards
implement Section 183 (f) and Section 112 of the Clean Air
Act as amended in 1990.
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; the Council on
Environmental Quality; members of the State and Territorial
Air Pollution Program Administrators; the Association of
Local Air Pollution Control Officials; EPA Regional
Administrators; Office of Management and Budget; and other
interested parties.
3. For additional information contact:
Mr. David Markwordt (MD-13)
Emission Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Telephone: (919) 541 0837
4. Copies of this document may be obtained from:
U.S. EPA Library (MD-35)
Research Triangle Park, North Carolina 27711
Technology Transfer Network
Telephone No. for data transfer: (919) 541-5742
Telephone No. for information: (919) 541-5384
National Technical Information Service
5285 Port Royal Road
Springfield, Virginia 22161
IV
-------
TABLE OF CONTENTS
Page
1.0 SUMMARY 1-1
1.1 SUMMARY OF CHANGES SINCE PROPOSAL 1-2
l.l.l Removal of unloading operations
(ballasting) from the source
category 1-2
1.1.2 Extension of the compliance
schedule for section 183(f)
("Title I") and section 112
("Title III") standards 1-3
1.1.3 Addition of new subcategories
under the section 112 regulations
for offshore terminals and for
the Alyeska Pipeline Service
Corporation's Valdez Marine
Terminal 1-5
1.1.4 Expansion of the petroleum
refineries source category
to include marine tank
vessel loading operations
collocated at petroleum
refinery operations 1-8
1.1.5 Revision to the subcategories
established based on annual HAP
emissions from 1 ton per year to
10 tons per year of any single HAP
or 25 tons per year of total HAP ... 1-10
1.1.6 Incorporation of minimum vapor
pressure limit 1-11
1.1.7 Recalculation of the MACT floors ... 1-12
1.1.8 Incorporation of additional
flexibility to the monitoring
requirements and compliance
provisions 1-16
1.1.9 Minor Changes 1-17
1.2 SUMMARY OF IMPACTS OF PROMULGATED ACTION ... 1-18
1.3 REFERENCES 1-20
2.0 SUMMARY OF PUBLIC COMMENTS 2-1
2.1 APPLICABILITY 2-12
2.1.1 Loading/Unloading Operations 2-12
2.1.2 Potential to Emit 2-14
2.1.3 Titles I and III 2-17
2.1.4 Vapor Pressure Limit 2-30
2.1.5 Commodity Type 2-34
2.1.6 Valdez Marine Terminal (VMT) 2-36
2.1.7 Ballasting and Bunkering 2-41
2.1.8 Regeneration of Carbon Beds 2-46
-------
TABLE OF CONTENTS (continued)
Page
2.2 INCLUSION WITH THE PETROLEUM REFINERY SOURCE
CATEGORY AND CONTIGUITY 2-47
2.2.1 General 2-47
2.2.2 Contiguity and Collocation 2-50
2.2.3 Emissions Averaging 2-52
2.2.4 Exposure Spikes 2-56
2.2.5 Other Impacts 2-58
2.3 SUBCATEGORIZATION 2-61
2.3.1 Subcategories Based on Size 2-61
2.3.2 Offshore Terminals/Offshore
Production 2-65
2.3.3 Parcel Tankers 2-70
2.3.4 VMT 2-71
2.4 RACT/MACT 2-73
2.4.1 MACT Floor 2-73
2.4.2 Stringency of MACT/RACT 2-80
2.4.3 Partial Control of Commodities .... 2-89
2.4.4 New/Existing Sources and Requirements 2-90
2.4.5 Area Sources 2-93
2.4.6 Recovery 2-94
2.4.7 Incineration 2-95
2.4.8 Format of the Standards 2-96
2.4.9 Regulation of Two Berths within VMT . 2-98
2.4.10 RACT and MACT in Same/Separate Rules . 2-102
2.4.11 Parallel With The HON 2-103
2.4.12 Alternative Standard 2-104
2.5 COMPLIANCE SCHEDULE FOR TITLES I AND III . . . 2-106
2.6 COMPLIANCE, PERFORMANCE TESTING, AND
MONITORING 2-119
2.6.1 Compliance 2-119
2.6,2 Performance Testing Requirements
and Procedures 2-125
2.6.3 Establishing Operating Parameters . . 2-128
2.6,4 Monitoring Requirements ....... 2-135
2.6,5 Emission Estimating Procedures .... 2-139
2.6.6 Emissions Averaging 2-151
2.6,7 Exemptions from Performance Testing . 2-154
2.1 VAPOR TIGHTNESS REQUIREMENTS 2-156
2.8 LEAK DETECTION AND REPAIR 2-168
2.9 REPORTING AND RECORDKEEPING REQUIREMENTS ... 2-173
2.10 GENERAL PROVISIONS . ............ 2-179
2.11 WORDING OF REGULATION . . . . , 2-181
2.11.1 General ....... 2-181
2.11.2 Definitions . 2-188
2.11.3 Overlap with Other Standards 2-196
2.12 ADMINISTRATIVE RECORD/SOURCES OF INFORMATION . 2-201
2.13 COST EFFECTIVENESS/IMPACTS 2-211
2,13.1 Cost Effectiveness 2-211
-------
TABLE OF CONTENTS (continued)
Page
2.13.2 Other Cost-Related Impacts 2-218
2.14 MISCELLANEOUS 2-224
2.14.1 Private Citizens (Groups) 2-224
2.14.2 Publication of Regulation in the
Federal Register 2-225
2.14.3 National Security 2-229
2.14.4 Other Miscellaneous Comments 2-231
2.15 COMMENTS ON PROPOSED APPENDICES TO
40 CFR PART 64 2-235
VI1
-------
LIST OF TABLES
Page
TABLE 1-1 MACT FLOORS, EMISSION REDUCTIONS, AND
NATIONAL COSTS 1-13
TABLE 2-1. COMMENTS RECEIVED ON THE PROPOSED MARINE
TANK VESSEL LOADING OPERATIONS FEDERAL RULES
AND NESHAP 2-4
Vlll
-------
1.0 SUMMARY
On May 13, 1994, the Environmental Protection Agency (EPA)
proposed federal rules and national emissions standards for
hazardous air pollutants (NESHAP) for marine tank vessel loading
operations (59 FR 10591) under authority of Section 183 (f) and
Section 112 of the amended Clean Air Act (Act). Public comments
were requested on the proposal in the Federal Register. The
Agency also reopened the public comment period for specific
comments on two occasions -- August 31, 1994 (59 FR 44955) and
March 8, 1995 (60 FR 12723). One hundred forty commenters,
composed of States, environmental groups, control equipment
vendors, trade groups, marine tank vessel owners/operators, and
marine tank vessel terminal owners/operators responded to the
request.
The comments that were submitted and the responses to these
comments are summarized in this document. The summary of
comments and responses serves as the basis for the revisions made
to the standards between proposal and promulgation.
1-1
-------
1.1 SUMMARY OF CHANGES SINCE PROPOSAL
Several changes have been made since the proposal of these
standards. A discussion of these changes in provided in the
preamble to the final rule. The majority of the changes have
been made to clarify portions of the rule that were unclear to
the commenters. A summary of the major changes is presented
below.
1.1.1 Removal of unloading operations (ballasting) from the
source category
In the proposed rule, the Agency included regulations
proscribing emissions from ballasting of vessels following
unloading of vessels. Comments asserted that ballasting
operations are performed by vessel operators, not by the
regulated terminal sources (see Section 2.1.1). The Agency
agreed with this interpretation of the affected source. The
Agency also agreed that regulating ballasting operations would be
difficult to enforce. Moreover, EPA agreed that the relatively
low amount of actual emissions associated with ballasting did not
justify dual regulation of ballasting.1 As discussed in the
Section 2.1.1, the total VOC emissions from crude oil tankship
ballasting were estimated to be approximately 950 Mg/yr. Based
on the portion of HAP in crude oil vapor, total HAP emissions
from ballasting would be less than 120 Mg/yr. The Agency noted
that ballasting emissions are likely to diminish in the future
because tankships built- since 1980 are required by domestic law
and international agreement to use segregated ballast tanks that
do not emit vapors during ballasting. Therefore, in order to
prevent confusion in the regulated community, the Agency decided
to not address ballastiaq or bunkering emissions in the final
regulation, The Agency elected to defer to the U.S. Coast
Guard's existing 'standards (3.3 CFR parts 155 and 157; and 46 CFR
parts 30 et al.).
1-2
-------
1.1.2 Extension of the compliance schedule for section 183(f)
("Title I") and section 112 ("Title III") standards
The Agency proposed to establish compliance deadlines of
2 years for the Section 183(f) standards, and 3 years for the
Section 112 standards. The Agency received numerous comments in
response to the proposed rule and the Agency subsequently
published a request for comment dealing specifically with whether
the compliance periods for the RACT and MACT regulations should
be extended [published in the Federal Register on March 8, 1995,
(60 FR 12723)]. Most of these commenters stated that the length
of the compliance periods was insufficient to comply with the
standards (see Section 2.5). The Agency agreed with the
commenters that permitting and safety approvals from permitting
authorities and the Coast Guard, the dearth of skilled
engineering and construction firms, and the history of facilities
being unable to comply with existing regulations compelled the
Agency to extend the date for full compliance with the RACT and
MACT rules.
In the final standards, EPA allowed sources regulated under
section 183 (f) 3 years to be in full compliance with the emission
control requirements promulgated under section 183(f). In
addition, RACT sources may request a waiver of up to 1 year to
achieve full compliance with the requirements if they can show
that the additional period is necessary for the installation of
controls. The Agency believes that this result was consistent
with section 183(f) requirements for the application of
"reasonably available" control technology, considering costs, any
non-air quality benefits, environmental impacts, energy
requirements, and safety factors. The Agency also believes that
the imprecision of the term "effective date" provided EPA with
the ability to allow compliance after 2 years. [The distinction
between "effective dates" of regulations and "compliance dates"
is important and has been a clear part of administrative
procedure for many years. See, e.g.. section 112 (i) (3); Natural
Resources Defense Council v. Environmental Protection Agency.
1-3
-------
22 F. 3d 1125, 1138 (D.C. Cir. 1994).] In the final rule, the
Agency required regulated RACT sources to provide proof that they
have commenced construction of vapor collection systems and air
pollution control devices within 2 years after promulgation of
the final standards. The Agency believes that these actions
could reasonably be achieved within 2 years of promulgation.
The EPA allowed existing sources regulated solely under
section 112 four years to be in full compliance with the emission
control requirements promulgated under section 112.
Section 112(i) of the Act specifically allows EPA to provide
sources with a waiver of up to 1 year to achieve full compliance
with the requirements if they can show that the additional period
is necessary for installing the controls. The Agency determined
that the sources controlled under section 112 not controlled
under section 183(f) should automatically receive a waiver of
1 year allowing a total of four years to comply with the MACT
emission reduction requirements. The Agency believes that this
total of 4 years is sufficient time for the estimated 20 sources
presently uncontrolled to design and install control technologies
sufficient to meet the MACT standards.
The Agency belie^res that the staggered compliance schedule
(i.e., 3 years for- RACT terminals and 4 years for MACT terminals)
coupled with the reduced number of terminals required to control
emissions under the final rule would alleviate commenters'
concerns about che scarcity of qualified installation consultants
and vendors. This extended schedule was also expected to address
concerns regarding permitting delays.
The Agency provided the VMT with 30 months to be in full
compliance with these regulations. The Agency believes, per its
discussions with APSC, that this extension provided sufficient
time to comply with the promulgated rule.
1-4
-------
1.1.3 Addition of new subcategories under the section 112
regulations for offshore terminals and for the Alyeska Pipeline
Service Corporation's Valdez Marine Terminal
In the proposed rule, the Agency established two
subcategories based on size for the section 112 MACT
regulations.2 The Agency also solicited comments on whether
additional subcategories should be established under the MACT
regulations for specific types of terminals based on particular
characteristics of those types of terminals of which the Agency
had no information at that time. Based on the public comments,
the Agency determined that two additional subcategories exist
within the marine tank vessel loading operation source category.
1.1.3.1 Subcategory for offshore terminals. In the
proposed rule, the Agency stated that it did not believe a
facility at least one-half mile offshore is part of a land-based
contiguous site, and believed that such offshore terminals
present unique regulatory challenges such as costs, environmental
impacts, and/or size constraints. The Agency requested
information regarding the feasibility and costs of controlling
emissions from offshore terminals, and inquired whether offshore
terminals should be grouped into a separate subcategory and what
the control status of terminals in such a subcategory should be.
Several comments were received in response to this request
(Section 2.3.2). These comments indicated that these types of
vessel loading operations face significant and unique challenges
in controlling emissions that were different from land-based,
contiguous loading operations. Based on these comments, the
Agency determined that a subcategory for these types of terminals
was justified and based its definition for offshore terminals on
a minimum distance of one-half mile from the terminal's furthest
loading point to the shore, regardless of the existence of subsea
lines.
1.1.3.2 Subcateqory for Alyeska Pipeline Service Company's
Valdez Marine Terminal. In the proposed rule, the Agency
1-5
-------
solicited comment on the possibility of placing the VMT in a
separate subcategory. Comments were received from APSC and
several other commenters representing State and local
governments, industry, private citizens, and environmental
groups. The Agency determined that the VMT should be placed in a
separate subcategory for the following reasons: (l) the VMT is
the largest (by a significant amount) crude oil loading operation
in the U.S.; (2) special circumstances, including climatic and
economic conditions, require keeping the oil moving through the
pipeline; (3) severe meteorological conditions result in
increased loading irregularity; and (4) the VMT throughput
projections show declining throughput over the next several
years. Several comments from environmental groups, State and
local agencies, and private citizens in the Valdez area did not
object to placing VMT in a separate subcategory provided that
emissions were controlled. After review of all of these factors,
the Agency placed APSC's VMT in a separate subcategory.
Emissions from all throughput loaded from at least two berths at
VMT will be reduced by 98 percent efficiency under both MACT and
RACT standards. Maximum limits for total throughput and
throughput at uncontrolled berths (above which all VMT loading
berths would be required to reduce emissions by 98 percent) were
included as part of these standards. These throughput limits
address the projected decreasing throughput that would
necessitate the use of only two berths for routine loading after
2001. Provisions to allow for scheduled maintenance of the
controlled berths were also established in the VMT standards.
The selected approach for regulating VMT under MACT was more
stringent than the MACT floor.3 The Agency selected this
approach after review of the cost effectiveness of controlling
VOC at this terminal (i.e., approximately $1,050 per Mg for
controlling VOC or approximately $8,000 per Mg for controlling
HAP). (Note: the data for this cost effectiveness analysis were
supplied by the APSC in their public comments.) The Agency
1-6
-------
initially received comments stating that the cost of controlling
VOC at this terminal was high relative to the benefits of
controlling VOC at a remote site in an Arctic ozone attainment
area. Additionally, some commenters initially stated that the
benefits of controlling HAP did not appear to justify the costs.
However, later comments from the APSC, the State of Alaska, the
Prince William Sound Regional Citizens Advisory Committee (a
local citizens group) and private citizens agreed that a Federal
rule mandating control of the primary emissions at the APSC was
acceptable.
1-7
-------
1.1.4 Expansion of the petroleum refineries source category to
include marine tank vessel loading operations collocated at
petroleum refinery operations
[Note: Please refer to the Summary of Public Comments and
Responses: National Emission Standards for Hazardous Air
Pollutants -- Petroleum Refineries -- Background Information for
Final Standards, EPA-453/R-95-015b for additional information on
the expansion of the petroleum refineries source category.]
The preamble to the proposed petroleum refinery NESHAP
published in the Federal Register on July 15, 1994 (59 FR 36130)
requested comments on whether marine tank vessel loading
operations at refineries should be included in emissions
averaging. On August 31, 1994, the EPA also reopened the comment
period for the proposed NESHAP for Marine Tank Vessel Loading
Operations (59 FR 44955) to request comment on whether marine
terminals collocated at refineries should be moved to the
petroleum refinery source category. Several commenters
responding to the marine tank vessel loading operations proposed
standards supported averaging of refinery process unit emissions
with emissions from marine terminals and gasoline distribution
operations that are located at refineries. The commenters cited
more cost effective emission reduction as the advantage of
including these emission, points in emissions averaging and
specifically commented that the costs per Mg emission reduction
of the marine tank vessel loading emission controls are high.
These commenters also claimed that emission calculation
procedures for loading are well established and that adding
marine loading to the averaging provisions will not appreciably
increase the complexity of enforcement. Other commenters opposed
including marine tank vessel loading and gasoline distribution in
emissions averaging. Some commenters claimed that these are
separate source categories, and the Act does not permit averaging
across source categories. Others were concerned that including
marine loading in averages could result in uncontrolled peak
emissions.
1-8
-------
In the final rules, emissions from marine tank vessel
loading operations as well as those from bulk gasoline terminal
or pipeline breakout station storage vessels, and bulk gasoline
terminal loading racks at petroleum refineries are allowed to be
included in emissions averages. The petroleum refinery source
category and source definitions were changed to include marine
tank vessel loading operations among others. However, this
redefinition only applied to MACT marine tank vessel operations;
all terminals subject to section 183(f) regardless of location
cannot average emissions with petroleum refinery sources.
In redefining the petroleum refinery source category, the
Agency determined that a marine tank vessel loading operation
that is collocated at a petroleum refinery could be considered
part of the same source as the refinery subject to this rule.
Because these operations were redefined to be part of the source
subject to the rule, the prohibition against intersource
averaging was not violated. This decision was made to provide
more opportunities to average and allow owners or operators of
petroleum refineries to find the optimal control strategy for
their particular situation.
The Agency noted that including emissions from marine tank
vessel loading operations in emissions averages will result in
equivalent or greater overall HAP emission reduction at each
refinery. The averaging provisions (see Petroleum Refinery
NESHAP final rule) are structured such that "debits" generated by
not controlling an emission point that otherwise would require
control must be balanced by achieving extra control at other
refinery emission points covered by the NESHAP.
1-9
-------
1.1.5 Revision to the subcategories established based on annual
HAP emissions from 1 ton per year to 10 tons per year of any
single HAP or 25 tons per year of total HAP
The proposed standards grouped major source terminals into
two subcategories based on HAP emissions: terminals with HAP
emissions of 1 ton per year or more and terminals having HAP
emissions of less than 1 ton per year. In the preamble of the
proposed rule, the Agency requested comment on establishing these
subcategories based on size (i.e., HAP emissions). In the final
rule, supported by general support from the comments, the Agency
has retained the use of HAP emissions to group major source
terminals into two subcategories. However, the subcategories
were changed to terminals having emissions of 10 tons per year or
more of any single HAP or 25 tons per year or more of total HAP
and terminals having HAP emissions of less than 10 tons per year
of all single HAP or less than 25 tons per year of total HAP (see
Section 2.3.1, and Reference 3).
1-10
-------
1.1.6 Incorporation of minimum vapor pressure limit
The proposed rule enabled individual facilities to determine
which products to control to achieve a 93 percent mass limit.2
However, the format of the final MACT standards has been altered
to explicitly exempt low vapor pressure liquids (i.e., liquids
having a vapor pressure below 1.5 psia).3 Commenters noted that
EPA had based its calculation of the MACT floors on existing
State rules2'3 and stated that existing State regulations exempt
these liquids from emissions control requirements. Therefore EPA
decided that only liquids having a vapor pressure 1.5 psia or
greater should be controlled under the final rule. The format of
the final standards were changed to an efficiency format to
reflect the new approach.
1-11
-------
1.1.7 Recalculation of the MACT floors
The MACT floors determined for this final rulemaking are
different than those in the proposed rule. These final rule MACT
floors reflected changes in the Agency's regulation of marine
tank vessel loading including: (1) the establishment of
subcategories for offshore terminals and the VMT terminal; (2)
the incorporation of a 1.5 psia minimum vapor pressure limit
instead of the weighted average as was proposed; and (3) the
increase of the levels of the subcategories based on size (i.e.,
HAP emissions) from 1 ton per year to 10/25 tons per year. The
MACT floors for the final rule also reflected comments on the
proposed rule, including comments on the number of controlled and
affected terminals. However, the Agency did not change the way
in which the MACT floors for the final rule were calculated.2-3
The MACT floors for existing and new sources in the marine tank
vessel loading source category are shown in Table 1-1.3 The MACT
floors for existing and new major source terminals included with
the petroleum refineries source category are also shown in
Table i-l.
The MACT floors for existing sources were calculated using
the criteria established in section 112(d)(3)A) of the Act the
Additional information on the determination of these MACT floors
is in the docket,3 There are approximately 44 major source
terminals (not including the VMT) that emit 10 tons per year or
more of any one hazardous air pollutant (HAP) or 25 tons per year
or more of any combination of HAP. Twenty-three of these
terminals are controlled. The resulting MACT floor level of
control (i.e., average of the best performing 5 sources) is a 97
percent reduction of HAP emissions. There are approximately
1,435 terminals that emit less than 10 tons per year of all
individual HAi? and less than 25 tons per year of combined HAP.
Seventy-nine of these terminals reduce emissions from marine tank
vessel loading. The resulting MACT floor level of control is no
reduction in HAP emissions (Note: The average of the best
1-12
-------
o
§
1
1
1
U
w
5
8
1
1
1
Z
9
1
§
o
D
o
w
a!
§
u
0
z
u
£
I
eduction in
8
fc>
r-
OS
on in
red
g
o
red
red
n
a
o
u.
H
|
le
el
'i
4)
a.
S
S
Bl
g
of
ls.
ssi
un
te
e
ts
*
X t.
as o
Ex
ha
loa
Ga
•
I
1-13
-------
performing 172 sources resulted in 74 percent control. Since
this did not match a known control technology, the median of the
best performing 12 percent of sources, the 86th facility, was
used to calculate the MACT floor. This terminal has a calculated
emission reduction of 0 percent.)3
The Agency estimated that there are less than 30 offshore
terminals (i.e., loading terminals located 0.5 miles or more from
shore). The Agency was aware of only 2 controlled offshore
terminals. The resulting MACT floor is no reduction in HAP
emissions. (Note: Since the mean of the best performing 5
sources produces a result that did not match a control technology
-- 38 percent -- the median of the best performing 5 sources was
used to calculate the MACT floor. This calculation resulted in a
MACT floor of no control.) The VMT was uncontrolled. Since this
was the only terminal in the VMT source category, the MACT floor
level of control is no control.3
The MACT floors for new source marine terminal subcategories
subject to regulation under Title III of the Act were also
calculated following the criteria in section 112(d)(3)(A) of the
Act. The resulting new source MACT floors are listed in
Table l-l. There was no benefit in establishing two MACT floors
for new sources above and below the 10/25 subcategory for onshore
major sources since the MACT floors would be identical.
Therefore, terminals in these subcategories were combined for new
source MACT in the regulation. For new major source onshore
terminals (not including the VMT) regardless of the marine tank
vessel loading HAP emissions, the best performing source achieved
a 98 percent reduction of HAP emissions. Therefore, the
resulting MACT floor for these sources is 98 percent reduction of
HAP emissions. For new major source offshore terminals whose
marine tank vessel loading HAP emissions exceeded the limits for
a major source (i.e., 10 tons of any one HAP, or 25 tons of total
HAP), the best controlled similar source achieved a 95 percent
reduction of HAP emissions. The resulting MACT floor for new
1-14
-------
offshore major sources is therefore a 95 percent reduction in
emissions. Since the VMT subcategory only contains a single
source, and it is not possible for an additional source to be
added to this subcategory, no new source MACT floor was
calculated for the VMT subcategory.3
1-15
-------
1.1.8 Incorporation of additional flexibility to the monitoring
requirements and compliance provisions
The proposed rule required parametric monitoring or
continuous emissions monitoring (CEM) as a means of showing
compliance with the standards. Any exceedance of the parameters
or concentration limits established during a performance test
would have resulted in a violation of the standard. Comments
indicated that this approach was too severe and warranted
additional flexibility. In consideration of the lack of test
data establishing the relationship between marine tank vessel
loading emissions control efficiency and parametric monitoring
and because of the batch nature of marine tank vessel loading
operations, the Agency thoroughly revised the monitoring
requirements and compliance provisions of the final rule. A
requirement for an operation and maintenance (0 & M) plan was
added to the final regulation to ensure proper operation of the
air pollution control and monitoring equipment. The O & M plan
contains an inspection schedule for each component of the control
and monitoring equipment. The "compliance" language that
appeared in § 6J.563 of the proposed rule was removed. In its
place, the final rule contained provisions that require an
unscheduled inspection and corrective actions when operating
parameters exceed the applicable baseline parameters.
Flexibility was also added to the methods for determining
baseline parameters. Owners or operators of a source required to
reduce emissions may establish baseline parameters during a
performance test or may choose to set the applicable baseline
based on a manufacturer's recommended baseline operating
parameter.
Additional operating parameters were added to the final
regulation for several control devices. A cross reference to the
general provisions found in 40 CFR part 63 (containing
requirements for establishing alternative monitoring procedures)
was provided to assist sources seeking approval of alternative
monitoring procedures.
1-16
-------
1.1.9 Minor Changes
(1) Revisions were made to definitions and phrasing to
clarify the regulation.
»
(2) The reporting requirements were changed to require the
following one-time reports as specified in the general provisions
found in 40 CFR part 63, subpart A: report of startup,
construction or modification; notification and report of
emissions tests and results and/or initial notification of
compliance status; notification and report of
physical/operational changes; notification and report of waiver
applications; and an engineering report describing the vent
system used to vent each vent stream to a control device. Owners
or operators must submit yearly summary reports and yearly
reports of excess emissions and monitoring system performance
reports instead of the quarterly reports required under the
proposed rule. However, good-performing sources need only submit
the summary report found in 40 CFR part 63 subpart A instead of
both the summary report and the full excess emissions and
continuous monitoring system performance report found in
§ 63.567(d)(2).
(3) A table identifying the relationship of the final
General Provisions requirements was added to the final
regulation. Language similar to that in the General Provisions
was added to subpart Y in cases where a direct reference to the
General Provisions was not appropriate.
(4) References to the proposed Performance Specifications
101 and 102 were updated to incorporate the Agency's promulgation
of Performance Specifications 8 and 9 (see 59 FR 64580, Docket
Number A-90-44, Item Number IV-I-4).
1-17
-------
1.2 SUMMARY OF IMPACTS OF PROMULGATED ACTION
These standards will reduce nationwide emissions of
hazardous air pollutants (HAP) from marine tank vessel loading
operations by approximately 4,150 Mg (4,565 tons) after 1999
compared to the emissions that would result in the absence of the
standards.5 These standards will reduce emissions of volatile
organic compounds (VOC) from marine tank vessel loading
operations by approximately 39,000 Mg (42,900 tons) after 1999
compared to the emissions that would result in the absence of the
standards.5 No significant adverse secondary air, water, solid
waste, or energy impacts are anticipated from the promulgation of
these standards (59 FR 25004).
The implementation of this regulation is expected to result
in nationwide annualized costs for existing marine tank vessel
loading operations of $60 million to $100 million beyond baseline
based on an analysis of applying controls to all existing
facilities not currently controlled to the level of the
standards.5 Nationwide capital costs expected to result from
these regulations are approximately $266 million to $440
million.5
The economic impact: analysis performed for this rulemaking
showed that the estimated maximum price increases for the
affected products varied, but were not large (less than 1%).
These price-increase est.i mates reflect the control cost increases
for transporting crude and products. Because these increases are
small and because the elasticities of demand for petroleum
products are small, estimated percent output reductions were
minimal Correspondingly, estimated employment reductions were
also relatively small.4
Potentially significant economic impacts on some of the
smaller affected terminal operations were identified, although
the decision not to require emission controls for existing
smaller operations great!7 reduces the potential for adverse
economic impacts on small terminal operations. These potential
1-18
-------
impacts would result from the high per barrel control cost
differential between the smaller and larger terminal operations
that would need to control emissions. Some of these smaller
terminal operations, to the extent that they are competing with
nearby larger or unaffected terminal operations, could have had
difficulty raising prices to cover cost increases and could have
been significantly adversely impacted by this rule.4
The potential economic impact on marine tank vessel owners
was substantially reduced because of the decision not to require
emission controls on small existing terminals in this rulemaking.
Because only a small percentage of U.S. marine transported volume
of products will be impacted by the standard, only a relatively
small percentage of U.S. marine tank vessels will need to
retrofit. Thus, only the vessels that will need the least cost
to retrofit (most likely the larger, newer, double-skin vessels)
will do so, leading to some degree of dedicated service. It is
expected that vessel owners that do retrofit will be able to pass
most retrofit costs forward in terms of higher prices.4
1-19
-------
1.3 REFERENCES
1. U.S. Environmental Protection Agency. Technical Support
Document For the Development of a VOC Rule for Marine Vessel
Loading Operations. Research Triangle Park, North Carolina.
May 1992 (Revised December 1993) . Docket Number A-90-44,
Docket Item II-A-3.
2. Memorandum from T. Campbell, Midwest Research Institute, to
D. Markwordt, U.S. Environmental Protection Agency.
Calculation of a MACT Floor for Title III Rulemaking for
Marine Vessel Loading Operations. September 30, 1993.
Docket Number A-90-44, Item Number II-A-34.
3. Memorandum from D. Hearne, Midwest Research Institute, to
D. Markwordt, U.S. Environmental Protection Agency.
Recalculation of MACT Floors for Title III Rulemaking for
Marine Vessel Loading Operations. July 14, 1995. Docket
Number A-90-44, Item Number IV-B-2.
4. U.S. Environmental Protection Agency. Final Federal
Standards for Marine Tank Vessel Loading Operations and
Final National Emission Standards for Hazardous Air
Pollutants for Marine Tank Vessel Loading Operations:
Regulatory Impact Analysis. July 1995. Docket
Number A-90-44, Item Number IV-A-2.
5. U.S. Environmental Protection Agency. Marine Tank Vessels
Data Base: Final Federal Standards for Marine Tank Vessel
Loading Operations and Final National Emission Standards for
Hazardous Air Pollutants for Marine Tank Vessel Loading
Operations. July 1995. Docket Number A-90-44, Item
Number IV-A-1.
1-20
-------
2.0 SUMMARY OF PUBLIC COMMENTS
A total of 140 letters (including six duplicate entries)
commenting on the proposed standards and the technical support
document (TSD) for the proposed standard were received. A list
of commenters, their affiliations, and the EPA tracking number
assigned to their correspondence is given in Table 2-1.
For the purpose of presentation, the comments have been
categorized under the following topics:
2.1 APPLICABILITY
1. Loading/unloading operations
2. Potential to emit
3. Titles I and III
4. Vapor pressure limit
5. Commodity type
6. Valdez Marine Terminal (VMT)
7. Ballasting and bunkering
8. Regeneration of carbon beds
2.2 INCLUSION WITH THE PETROLEUM REFINERY SOURCE CATEGORY AND
CONTIGUITY
l. General
2. Contiguity and collocation
3. Emissions averaging
4. Exposure spikes
5. Other impacts
2-1
-------
2.3 SUBCATEGORIZATION
1. Subcategories based on size
2. Offshore terminals/offshore production
3. Parcel Tankers
4. VMT
2 .4 RACT/MACT
1. MACT floor
2. Stringency of MACT/RACT
3. Partial control of commodities
4. New/existing sources and requirements
5. Area sources
6. Recovery
7. Incineration
8. Format of the standards
9. Regulation of two berths within VMT
10. RACT and MACT in same/separate rules
11. Parallel with the HON
12. Alternative standard
2.5 COMPLIANCE SCHEDULE FOR TITLES I AND III
2.6 COMPLIANCE, PERFORMANCE TESTING, AND MONITORING
1. Compliance
2. Performance testing requirements and procedures
3. Establishing operating parameters
4. Monitoring requirements
5. Emission estimating procedures
6. Emissions averaging
7. Exemptions from performance testing
2.7 VAPOR TIGHTNESS REQUIREMENTS
2.8 LEAK DETECTION AND REPAIR
2.9 REPORTING AND RECORDKEEPING REQUIREMENTS
2.10 GENERAL PROVISIONS TO 40 CFR Part 63
2.11 WORDING OF REGULATION
1. General
2. Definitions
3. Overlap with other standards
2-2
-------
2.12 ADMINISTRATIVE RECORD/SOURCES OF INFORMATION
2.13 COST EFFECTIVENESS/IMPACTS
1. Cost effectiveness
2. Other cost-related impacts
2.14 MISCELLANEOUS
1. Private citizens (Groups)
2. Publication of regulation in the Federal Register
3. National security
4. Other miscellaneous comments
2.15 COMMENTS ON PROPOSED APPENDICES TO 40 CFR Part 64
2-3
-------
TABLE 2-1. LIST OF COMMENTERS ON PROPOSED STANDARDS OF
PERFORMANCE FOR MARINE TANK VESSEL LOADING OPERATIONS
Docket
item"
Commenter and affiliation
Docket
item"
Commenter and affiliation
IV-D-01 Paul McCollum IV-D-09
P.O. Box 2016
Homer, Alaska 99603
IV-D-02 Michael J. Wax, Ph.D.
Institute of Clean Air Companies
1707 L. Street, NW IV-D-10
Suite 570
Washington, D.C. 20036-4201
IV-D-03 Elsie L. Munsell
Department of the Navy
Office of the Assistant Secretary IV-D-11
(Installations and Environment)
Washington, D.C. 20360-5000
IV-D-04 John W. Casey IV-D-12
Shell Oil Company
One Shell Plaza
P.O. Box 4320
Houston, Texas 77210-4320 IV-D-13
IV-D-05 David W. Haugen
Lynden Incorporated
1029 W. Third Ave., Suite 150 IV-D-14
Anchorage, Alaska 99501-1981
IV-D-06 Stan Stephens
Prince William Sound Regional IV-D-15
Citizens'
Advisory Council
Royal Center, 310 Egan Street
Room 210, Box 3470 IV-D-16
Valdez, Alaska 99686
IV-D-07 Becky L. Gay
Resource Development Council for IV-D-17
Alaska, Inc.
121 West Fireweed Lane, Suite 250
Anchorage, Alaska 99503-2035
IV-D-18
IV-D-08 Martin King
The Alliance
4220 'B' Street, Suite 200
Anchorage, Alaska 99503-5911
Terry Hennach
Prince William Sound Conservation
Alliance
P.O. Box 1697
Valdez, Alaska 99686
Robert Colby
Donald Theiler
STAPPA/ALAPCO
444 North Capitol Street, Northwest
Washington, D.C. 20001
Kari Gustafson
P.O. Box 750962
Fairbanks, Alaska 99775-0962
Dan Moeller
Box 571
Valdez, Alaska 99686
James G. Lachance, Jr.
P.O. Box 1224
Valdez, Alaska 99686
Lilian D. Dillon
Box 1297
Valdez, Alaska 99686
Priscilla and Terry Gregg
P.O. Box 1458
Valdez, Alaska 99686
William and Cleo Wyatt
P.O. Box 93
Valdez, Alaska 99686
Kristine Gresham
881 No. Cottonwood
Valdez, Alaska 99686
Jenna Stephens
P.O. Box 1297
Valdez, Alaska 99686
2-4
-------
TABLE 2-1. (continued)
Docket
item"
Commenter and affiliation "
Docket
item*
Commenter and affiliation
IV-D-19 Gary and Roberta Thurow FV-D-27
P.O. Box 2569
Valdez, Alaska 99686
IV-D-20 David S. Kircher
Puget Sound Air Pollution Control
Agency
110 Union Street, Suite 500 IV-D-28
Seattle, Washington 98101-2038
IV-D-21 John W. Walton, P.E.
Tennessee Air Pollution Control Board
Department of Environment and IV-D-29
Conservation
401 Church Street
L & C Annex, 9th Floor
Nashville, Tennessee 37243-1531
IV-D-30
IV-D-22 W. N. Tuttle, P.E.
John Zink Company
International Headquarters
P.O. Box 21220
Tulsa, Oklahoma 74121-1220
IV-D-23 Arthur Lee
Texaco, Inc.
P.O. Box 509
Beacon, NY 12508
IV-D-24 Allen R. Elicit
BP Oil Company
200 Public Square
Cleveland, Ohio 44114-2375
IV-D-31
IV-D-32
IV-D-33
IV-D-25 David C. Copeland
Occidental Chemical Corporation
Occidental Chemical Center
360 Rainbow Boulevard South IV-D-34
P.O. Box 728
Niagara Falls, New York 14302-0728
IV-D-26 Ann Farner
Tosco Refining Company IV-D-35
2300 Clayton Road
Suite 1100
Concord, California 94520-2100
William R. Ludt
Hvide Shipping, Incorporated
2200 Eller Drive, Building 27
P.O. Box 13038 Port Everglades
Station
Fort Lauderdale, Florida 33316
William S. Dickinson
ARCO Products Company
P.O. Box 61004
Anaheim, California 92803-6104
Charles D. Malloch
Monsanto Company
800 N. Lindbergh Boulevard
St. Louis, Missouri 63167
Duane B. Bordvick
Tosco Refining Company
2300 Clayton Road
Suite 1100
Concord, California 94520-2100
Toby A. Threet
The Dow Chemical Company
2030 Dow Center
Midland, Michigan 48674
Alyeska Pipeline Service Company
1835 South Bragaw Street
Anchorage, Alaska 99512
Captain E. N. George
West Coast Shipping Company
1201 West 5th Street
Los Angeles, California 90017
Paul Bailey
American Petroleum Institute
1220 L. Street, Northwest
Washington, D.C. 20005
Richard Halluska
OMI Corp.
90 Park Avenue
New York, New York 10016
2-5
-------
TABLE 2-1. (continued)
Docket
item"
Commenter and affiliation
Docket
item*
Commenter and affiliation
IV-D-36 Andrea Grant
Counsel to the Independent Fuel
Terminal Operators Association
Verner, liipfert, Bemhard,
McPherson and Hand
901-15th Street, NW
Washington, D.C. 20005-2301
IV-D-37 W. T. Flis
Exxon Company, U.S.A.
Post Office Box 2180
Houston, Texas 77252-2180
IV-D-38 Beverly Hartsock
Texas Natural Resource Conservation
Commission
P.O. Box 13087
Austin, Texas 78711 -3087
IV-D-39 Barbara J. Price
Phillips Petroleum Company
Bartlesville, Oklahoma 74004
IV-D-40 Duplicate Entry
IV-D-41 Melanie S. Kelley
Total Petroleum, Inc.
P.O. Box 500
Denver, Colorado 80201-0500
IV-D-42 William J. Doyle, Ph.D.
Marathon Oil Company
539 South Main Street
Findlay, Ohio 45840-3295
IV-D-43 Gregory A. Green
Department of Environmental Quality
State of Oregon
811 SW Sixth Avenue
Portland, Oregon 97204-1390
IV-D-44 D. A. Woodring
BP Chemicals, Inc.
200 Public Square
Cleveland, Ohio 44114-2375
IV-D-46
IV-D-47
IV-D-48
IV-D-49
IV-D-50
IV-D-51
IV-D-52
Norbert Dee, Ph.D.
National Petroleum Refiners
Association
Suite 1000, 1899 L Street, N.W.
Washington, D.C. 20036
Deborah W. Gates
Ashland Petroleum Company
Division of Ashland Oil, Inc.
P.O. Box 391
Ashland, Kentucky 41114
James Udelhoven
Udelhoven Oilfield System Services,
Incorporated
11401 Olive Lane
Anchorage, Alaska 99515
Richard N. Schok
Flowline Alaska
1881 Livengood
Fairbanks, Alaska 99701
Harry McDonald
Carlile Enterprises, Inc.
1524 Ship Avenue
Anchorage, Alaska 99501
George A. Walker
Unocal Corporation
1201 West 5th Street
P.O. Box 7600
Los Angeles, California 90051
David Beck, et. al.
Valdez Chamber of Commerce
P.O. Box 512
Valdez, Alaska 99686
John A. Sandor
State of Alaska
Department of Environmental
Conservation
410 Willoughby Avenue, Suite 105
Juneau, Alaska 99801-1795
2-6
-------
TABLE 2-1. (continued)
Docket
item1
Commenter and affiliation
Docket
item*
Commenter and affiliation
IV-D-53 Judith M. Brady
Alaska Oil and Gas Association
121 West Firewood Lane, Suite 207
Anchorage, Alaska 99503-2035
IV-D-54 John Prokop
Independent Liquid Terminals
Association
1133 16th Street, N.W., Suite 650
Washington, D.C. 20005
IV-D-55 Wesley P. Nason
H.C. Price Company
301 W. Northern Lights Blvd.,
Suite 300
Anchorage, Alaska 99503
IV-D-56 Brian M. Hamey
Mobil Oil Corporation
3225 Gallows Road
Fairfax, Virginia 22037-0001
IV-D-57 Val J. Molyneux
Norcon, Inc.
Post Office Box 190947
Anchorage, Alaska 99519
IV-D-58 Nathan S. Bergerbest
Doyon, Limited
201 First Avenue, Suite 300
Fairbanks, Alaska 99701
IV-D-59 James P. Varley
Stolt Parcel Tankers, Inc.
15635 Jacintoport Boulevard
Houston, Texas 77015-6534
IV-D-60 Margaret Kaigh Doyle
The Chemical Carriers' Association,
Inc.
1700 North Moore Street, Suite 1805
Arlington, Virginia 22209
FV-D-61 R. James Lethcoe
P.O. Box 1313
Valdez, Alaska 99686
IV-D-62 Nancy R. Lethco
Alaska Wilderness Recreation and
Tourism Association •
P.O. Box 1353
Valdez, Alaska 99686
IV-D-63 Connie Rodgers
Prince William Sound College Dorms
Valdez, Alaska 99686
IV-D-64 Dawn Box
Valdez, Alaska 99686
IV-D-65 Mary Helen Stephens
P.O. Box 1272
Valdez, Alaska 99686
IV-D-66 Riki Ott, Ph.D.
Box 1430
Cordova, Alaska 99574
IV-D-67 Richard A. Fineberg
P.O. Box 416
Ester, Alaska 99725
IV-D-68 Walter Quanstrom
Amoco Corporation
200 East Randolph Drive
Post Office Box 87703
Chicago, Illinois 60680-0703
IV-D-69 Jennifer A. Kelly
The American Waterways Operators
1600 Wilson Boulevard
Suite 1000
Arlington, Virginia 22209
FV-D-70 Arthur J. Volkle, Jr.
Mari trans
One Logan Square
Philadelphia, Pennsylvania 19103
IV-D-71 Norman L. Morrow
Exxon Chemical Americas
P.O. Box 3272
Houston, Texas 77253-3272
2-7
-------
TABLE 2-1. (continued)
Docket
item'
Commenter and affiliation
Docket
item*
Commenter and affiliation
IV-D-72 P.M. Bitten IV-D-80
The Chevron Corporation
1401 Eye Street NW, Suite 1200
Washington, D.C. 20005
IV-D-73 M. L. Mullins
Chemical Manufacturers Association
2501 M Street, N.W.
Washington, D.C. 20037
IV-D-74 Nancy A. Wildeboer
Sun Company, Inc.
Ten Penn Center
1801 Market Street
Philadelphia, PA 19103-1699
IV-D-75 Sarosh J. H. Manekshaw
Pennzoil Company
Pennzoil Place
P.O. Box 2967
Houston, Texas 77252-2967
IV-D-76 William C. O'Brien
BF Goodrich
P.O. Box 527
Calvert City, Kentucky 42029-0527
IV-D-81
IV-D-82
IV-D-83
IV-D-84
IV-D-85
IV-D-86
IV-D-87
IV-D-88
IV-D-77 Ellen Siegler
American Petroleum Institute
1220 L Street, Northwest
Washington, D.C. 20005 IV-D-89
IV-D-78 J. C. Hovious
Union Carbide Corporation
Health, Safety and Environment
39 Old Ridgebury Road IV-D-90
Danbury, Connecticut 06817-0001
IV-D-79 William O'Sullivan, P.E.
State of New Jersey
Department of Environmental IV-D-91
Protection and Energy
401 East State Street, CN027
Trenton, New Jersey 08625
Milton Feldstein
Bay Area Air Quality Management
District
939 Ellis Street
San Francisco, California 94109
Kenneth Adams
Cordova, Alaska 44214
Duplicate Entry
Duplicate Entry
Duplicate Entry
Joseph J. Cox
American Institute of Merchant
Shipping
1000 16th Street, N.W., Suite 511
Washington, D.C. 20036-5705
Duplicate Entry
Duplicate Entry
Normon L. Morrow
Exxon Chemical Americas
P.O. Box 3272
Houston, Texas 77253-3272
Robert P. Stricter
American Petroleum Institute
1220 L Street, Northwest
Washington, D.C. 20005
Susan F. Tierney
U.S. Department of Energy
Office of Policy
Washington, D.C. 20585
William J. Doyle
Marathon Oil Company
539 South Main Street
Findlay, Ohio 45840-3295
2-8
-------
TABLE 2-1. (continued)
Docket
item1
Commenter and affiliation
Docket
item*
Commenter and affiliation
IV-D-92 Paul Bailey
American Petroleum Institute
1220 L Street, Northwest
Washington, D.C. 20005
IV-D-93 Brian M. Haraey
Mobil Oil Corporation
3225 Gallows Road
Fairfax, Virginia 22037-0001
IV-D-94 Stan Stephens
Prince William Sound Regional
Citizen's Advisory Council
601 West Fifth Avenue
Suite 500
Anchorage, Alaska 99501-2254
IV-D-95 Ann Farmer
Tosco Refining Company
2300 Clayton Road
Suite 1100
Concord, California 94520-2100
IV-D-96 Philip T. Cavanaugh
Chevron Corporation
1401 Eye Street, NW
Suite 1200
Washington, D.C. 20005
IV-D-97 Norbert Dee
National Petroleum Refiners
Association, Suite 1000
1899 L Street, NW
Washington, D.C. 20036
IV-D-98 David Driesen
Natural Resources Defense Council
1350 New York Avenue, NW
Washington, D.C. 20005
IV-D-99 Deborah W. Gates
Ashland Petroleum Company
Post Office Box 391
Ashland, Kentucky 41114
IV-D-100 Beverly Hartsock
Texas Natural Resource Conservation
Commission
Post Office Box 13087
Austin, Texas 78711-3087
IV-D-101 Norman Ingram
Alyeska Pipeline Service Company
1835 South Bragaw Street
Anchorage, Alaska 99512
IV-D-102 Stan Stephens
Prince William Sound
Regional Citizens' Advisory Council
601 West Fifth Avenue
Suite 500
Anchorage, Alaska 99501-2254
IV-D-103 Ellen Siegler
American Petroleum Institute
1200 L Street Northwest
Washington, D.C. 20005
IV-D-104 Philip T. Cavanaugh
Chevron Corporation
1401 Eye Street, NW
Suite 1200
Washington, D.C. 20005
IV-D-105 Charles Keffer
Monsanto Company
800 N. Lindbergh Boulevard
St. Louis, Missouri 63167
IV-D-106 Philip Cavanaugh
The Chevron Companies
1401 Eye Street, NW, Suite 1200
Washington, D.C. 20005
IV-D-107 Bliss Higgins
Louisiana Department of Environment
Quality
Office of Air Quality
Post Office Box 82135
Baton Rouge, Louisiana 70884
2-9
-------
TABLE 2-1. (continued)
Docket
item'
Commenter and affiliation
Docket
item*
Commenter and affiliation
IV-D-108 Duane B. Bordvick IV-D-117
Tosco Refining Company
2300 Clayton Road
Suite 1100
Concord, California 94520-2100
IV-D-109 Marc Phillips IV-D-118
Enron Operations Corporation
Post Office Box 1188
Houston, Texas 77251
IV-D-110 NorbertDee
National Petroleum Refiners
Association IV-D-119
Suite 1000
1899 L Street, NW
Washington, D.C. 20036
IV-D-111 Paul Bailey IV-D-120
American Petroleum Institute
1220 L. Street, Northwest
Washington, D.C. 20005
IV-D-112 William J. Doyle IV-D-121
Marathon Oil Company
539 South Main Street
Findlay, Ohio 45840-3295
IV-D-113 C. Barry Gipson
Ashland Petroleum Company
Post Office Box 391 IV-D-122
Ashland, Kentucky 41114
IV-D-114 J.W. Collins
4223 W. Telephone Road
Martinsburg, West Virginia IV-D-123
IV-D-115 Norman Ingram
Alyeska Pipeline Service Company
1835 South Bragaw Street
Anchorage, Alaska 99512 IV-D-124
IV-D-116 Robert Colby
Donald Theiler
STAPPA/ALAPCO
444 North Capitol Street, Northwest
Washington, D.C. 20001
John Prokop
Independent Liquid Terminals
Association
1133 15th Street, NW, Suite 650
Washington, D.C. 20005
William Walker
Prince William Sound Regional
Citizen's Advisory Council
750 West 2nd Avenue
Suite 100
Anchorage, Alaska 99501
Brian Harney
Mobil Oil Corporation
3225 Gallows Road
Fairfax, VA 22037
Ellen Siegler
American Petroleum Institute
1200 L Street Northwest
Washington, D.C. 20005
Bliss Higgins
Louisiana Department of Environment
Quality
Office of Air Quality
Post Office Box 82135
Baton Rouge, Louisiana 70884
Paul Bailey
American Petroleum Institute
1220 L. Street, Northwest
Washington, D.C. 20005
Robert Freeman
SeaRiver Maritime, Inc.
Post Office Box 1512
Houston, Texas 77251
John Harris, Mayor
City of Valdez, Alaska
Post Office Box 307
Valdez, Arkansas 99686
2-10
-------
TABLE 2-1. (continued)
Docket
item"
Commenter and affiliation
Docket
item*
Commenter and affiliation
IV-D-125 Norman Ingram
Alyeska Pipeline Service Company
1835 South Bragaw Street
Anchorage, Alaska 99512
IV-D-126 Norman Ingram
Alyeska Pipeline Service Company
1835 South Bragaw Street
Anchorage, Alaska 99512
IV-D-127 Thomas Allen
New York State Department of
Environment Conservation
50 Wolf Road
Albany, New York 12233
IV-D-128 Robert Reges
State of Alaska Department of Law
Capitol Building
Post Office Box 110300
Juneau, AK 99811
IV-D-129 Norman Ingram
Alyeska Pipeline Service Company
1835 South Bragaw Street
Anchorage, Alaska 99512
IV-D-130 Ronald Kiracofe
ARCO Products, Company
1055 West Seventh Street
Post Office Box 2570
Los Angeles. California 9u051
IV-D-131 Frank Murkowski
Ted Stevens
United States Senate
Washington, D.C. 20510
IV-D-132 V. M. Gonzales
ARCO Products Company
1055 West Seventh Street
Post Office Box 2570
Los Angeles, California 90051
IV-D-133 David Smith
ARCO Products Company
1055 West Seventh Street
Post Office Box 2570
Los Angeles, California 90051
IV-D-134 Ellen Siegler
American Petroleum Institute
1200 L Street Northwest
Washington, D.C. 20005
IV-D-135 James Sample
2608 E. Meredith Drive
Vienna, Virginia 22181
IV-D-136 Mike Steinbrecher
Chevron Corporation
Federal Relations
575 Market Street, Room 968
San Francisco, California 94105
IV-D-137 Del Fogelquist
Western States Petroleum Association
2201 Sixth Avenue
Seattle, WA 98121
IV-D-138 Norman Ingram
Alyeska Pipeline Service Company
1835 South Bragaw Street
Anchorage, Alaska 99512
IV-D-139 H. Nelson Meeks
ARCO Pipe Line Company
5900 Cherry Avenue
Post Office Box 787
Long Beach, California 90801
IV-D-140 Robert Ehrlich
Amerada Hess Corporation
1 Hess Plaza
Woodbridge, New Jersey 07095
2-11
-------
2.1 APPLICABILITY
2.1.1 Loading/Unloading Operations
Comment; While the rule indicates applicability to "loading
and unloading operations," several commenters (04, 23, 29, 36,
43, 44, 59, 60, 73, 78) contended that unloading operation
emissions occur at storage tanks, which are regulated under other
regulations; this proposed rule does not establish any standards
nor does it discuss emissions from unloading operations. The
commenters stated the EPA should clarify the applicability of the
regulation. Eight of the commenters (04, 23, 29, 43, 44, 59, 73,
78) suggested that the word "unloading" should be deleted from
the package. One commenter (80) noted that the term "loading and
unloading operations" is not defined in the rule. The commenter
requested clarification on the definition of "major source" and
believed that nonmajor individual loading and unloading
operations within a major- or area-source facility should be
subject to the standard.
Two commenters (34, 39) indicated the proposed rule does not
specify that ballast operations are the only nonloading operation
subject to control, leaving open the potential for further
product unloading requirements; the EPA should clarify this
position (39) . Two commenters (31, 34) stated that the rule
needs to specify that ballasting operations are the only
nonloading operation required to be controlled and that product
unloading should be deleted from further regulation. One
commenter (42) requested that EPA provide clarification in the
regulation requiring unloading emissions to be considered only in
cases where ballast material is added to product storage tanks.
Also, the emission factor for unloading operations that includes
ballasting into the product storage tank is not appropriate for
unloading operations that do not include the addition of ballast
into the product storage tank.
Another commenter (23) stated that EPA should have requested
comments on the feasibility of closed loading by certain barges.
The commenter suggested reopening the comment period to address
2-12
-------
the question of whether barges that are open-loaded can be
closed-loaded as stated in the preamble.
Response: The Agency concurs with the comments stating the
proposed standards were only applicable to marine tank vessel
loading operations and ballasting operations. After further
consideration, emissions from ballasting operations will not be
covered by the standards (for further explanation, see 2.1.7).
This exclusion will help clarify any confusion about
applicability to any nonloading operations and limit
applicability to loading activities only. Therefore, the word
"unloading" was deleted from the final rule. The Agency has also
included a definition for marine tank vessel loading operations
in the final rule. The Agency has also clarified its
interpretation of the term "major source" (see Sections 2.1.2 and
2.1.3 of this document for additional discussion of major source
determination).
With regard to the issue of requiring certain open-loaded
barges to be closed-loaded, the Agency believes that the
commenter has not shown that the open loaded barges are different
from other types of vessels loaded at affected terminals. The
Agency disagrees with the comment that the loading of these types
of barges is technically or economically infeasible.
2-13
-------
2.1.2 Potential to Emit
Comment; One commenter (20) stated that the EPA needs to be
explicit about how to determine (and to limit the said
determination of) the potential to emit from marine tank vessel
loading operations. The commenter recommended federally
enforceable emissions limitations, including specific throughput
limits for each product, emission standards for vapor recovery
systems, the American Petroleum Institute's (API's) latest
emission factors for storage tanks and fitting losses, and
default HAP/VOC ratios for common products such as gasoline,
distillate oil, residual oil, and crude oil.
Three commenters (28, 42, 56) indicated a marine facility's
actual emissions should be regarded as its potential to emit when
determining whether or not it is a major source. Requiring
facilities to assume a 10,000 barrels/hr, 24 hr/d, 365 d/yr
loading rate is unrealistic. The commenter stated a more
realistic approach needs to be developed to limit the
applicability of the "potential to emit" rule contained in the
General Provisions to 40 CFR Part 63. Three commenters (34, 42,
56) suggested a tiered approach as developed for the Gasoline
Marketing NESHAP. This approach bases major source status on
actual emissions and by the facility committing to federally
enforceable limitations established through the MACT standard.
Six commenters (34, 50, 56, 68, 71, 75) pointed out that the
proposed regulation's MACT floor determination and regulatory
impact analysis have been developed based on actual emission
estimates and actual throughput of products at facilities, not
potential to emit. The commenters claimed that it is
inconsistent to set standards based on one criterion and
suggested that those standards be based on another, much broader,
criterion. These commenters also noted the calculated impacts
are understated because, under a "potential to emit" definition
the regulation would apply to far more major sources than
estimated.
2-14
-------
One commenter (34) recommended arranging marine tank vessel
facilities into one of three categories based on actual emission
rates (of product loading and ballasting operations). The three
categories included facilities exceeding the major source
definition based on actual emissions, facilities with actual
emissions of between 50 and 100 percent of the major source
definition, and facilities with actual emissions of less than
50 percent of the major source definition. Two other commenters
(50, 56) supported the three-category approach proposed by
commenter 34 to overcome their problems with the potential to
emit definition.
One commenter (24) stated the final rule needs to be
modified to explicitly exempt marine terminals that actually emit
less than l.l tons per year (tons/yr) of HAP from all controls,
even if its potential to emit is greater than the definition of a
major source. In addition, the commenter stated that any source
that has installed controls to reduce emissions of VOC due to a
State VOC rule should be allowed to calculate their actual
emissions after such controls. This commenter noted that the
Agency's current definition for "potential to emit" would result
in even the smallest terminal being classified as a major source.
It is not clear if sources having actual HAP emissions of greater
than 1 ton/yr but less than 10 tons/yr would be subject to the
emission reduction requirements of this regulation. The
commenter asserted that the EPA should clarify its position on
the potential to emit as it regards applicability of these
smaller sources. Two commenters (45, 46) noted if EPA proceeds
with promulgating this regulation under Section 112(d), the
complicated issue of "potential to emit" becomes a factor in
determining applicability If EPA were to re-propose this
regulation under Section !33(f); this issue would be avoided
completely.
Response; The Agency has classified all major sources
having actual HAP emissions from marine tank vessel loading
operations of 10 tons/yr or mora of any one HAP or 25 tons/yr or
2-15
-------
more of total HAP as sources subject to the emissions standards
under Section 112. Major sources having actual emissions from
marine tank vessel loading operations of less than 10 tons/yr of
all single HAP or less than 25 tons/yr of combined HAP are only
required to comply with the recordkeeping and reporting
requirements documenting that the source's emissions have not
increased above the 10/25 tons/yr limits. Nonmajor sources
(i.e., sources whose source-wide potential emissions are below
10 tons/yr of all individual HAP or 25 tons/yr of total HAP) are
not affected by these standards. Since most facility-wide
activities are non-batch operations, the major source/nonmajor
source decision is based on a source's potential to emit as
required by Section 112(a) and as defined in the General
Provisions to 40 CFR Part 63 (59 FR 12408, March 16, 1994) rather
than on actual emissions. If a facility is classified as a major
source, actual emissions from marine loading operations are used
to determine the subcategory to which the source belongs and,
consequently, the applicability of the MACT standard to that
source.
Several commenters supported the API-proposed, three-
category approach based on actual emissions. EPA believes that
Congress intended to distinguish between actual emissions and
potential to emit in its decision to use both phrases in
Section 112(a). EPA believes that the definition of Potential to
Emit in the General Provisions is a better fulfillment of
Congress's intent than API's proposal. (Any use of actual
emissions to distinguish between major and nonmajor sources would
have to be federally enforceable.) For a more detailed
discussion of EPA's "potential to emit" definition, see the
General Provisions rule.
2-16
-------
2.1.3 Titles I and III
Comment: Ten commenters (04, 23, 30, 34, 45, 46, 56, 75,
120, 134) believed that EPA should not regulate marine loading
operations under Section 112. According to commenters, while
Section 183(f) is directed specifically at marine loading
facilities and includes specific procedural and substantive
requirements, Section 112 is a general provision that applies to
many different sources of HAP. Furthermore, there is no
indication from the legislative history that Congress intended,
nor wanted, EPA to regulate marine loading operations under
Section 112. The commenters noted that the EPA itself reached
that conclusion when it developed the initial source category
listing. One commenter (32) stated that if EPA had stuck with
its decision to regulate vessel loading emissions exclusively
under Section 183 (f), the Agency would have reason to argue that
reference to "any other air pollutant" in Section 183(f), thus
giving EPA the authority to regulate indirectly those HAP
emissions that are a subset of the VOC emissions. However, EPA
abandoned that argument. The commenter added that it will be
infeasible to comply with the specified compliance periods. As
such, the Section 112 portion of the rule should be withdrawn.
One commenter (04) elaborated on the conflicts between the two
sections, maintaining that Congress meant for EPA to consider
factors such as safety and cost effectiveness in 183(f) but that
Section 112(d), in conflict, constrains EPA's consideration of
cost and safety.
In addition, one coxiimenter (134) claimed that the Agency is
unable to give full consideration to legal issues regarding the
appropriateness of Agency regulation under Section 112 because an
existing consent decree requires EPA to take final action with
regard to Section 112(d) by April 30, 1995. [This date has been
revised to July 28, 1995,]
Three commenters (32, 39, and 75) stated that, in the
proposed rule, there is no demonstration that VOC emissions in
attainment areas for marine loading operations are detrimental to
2-17
-------
human health. For this reason, applicability of Title I
provisions to attainment areas should be dropped from the rule
until EPA demonstrates the hazards to human health. Two
commenters (42, 75) indicated that limiting the rule's
requirements to terminals in nonattainment areas where RACT
controls are likely to provide the greatest benefit and fewer
subject facilities will improve the feasibility of these
terminals to comply with the Title I regulations within
24 months. Another commenter stated that the final rule should
consider a terminal's attainment/nonattainment status when
thresholds for a terminal's applicability under RACT are
established. The commenter also suggested using a terminal's HAP
cost effectiveness rather than its VOC cost effectiveness when
determining RACT for a given terminal.
One commenter (04) stated that Congress did not intend for
EPA to apply marine vapor controls to all affected facilities
without regard to their location relative to ozone attainment
status. One commenter (23) took this assertion a step further by
stating that it is clear that Congress directed EPA to issue
marine tank vessel loading and unloading standards to reduce
ozone only in nonattainment areas. The commenter stated that EPA
exceeded its authority under the Act by requiring RACT standards
for any existing or new marine tank vessel loading and unloading
operations. One commenter (28) stated that if the ozone standard
is being attained, the justification for imposing additional
regulations on existing sources of ozone precursors is not
apparent. This commenter's company operates a facility in an
ozone attainment area that is 3 miles from the nearest community
and that serves a refinery that is 1.5 miles away. The commenter
believed that neither VOC controls or HAP controls should be
imposed because the ozone standard is being attained and because
facilities that are remote from populated areas should be in a
separate subcategory and not combined with facilities that have
an impact on communities. In any case, the regulation should
allow for case-by-case review, and facility owners should have a
2-18
-------
mechanism for securing relief from these rules through
presentation of sound, technical justification.
One commenter (34) suggested EPA should do one of the
following for the Section 183(f) rule: (1) limit the scope of the
rule to facilities located in ozone nonattainment areas;
(2) revise the technical requirements to control options that can
be implemented within 2 years; or (3) defer issuing the final
rule until the proposed control options can be achieved within
2 years from promulgation.
One commenter (56) asked if EPA should decide to keep marine
tank vessel loading operations on the Section 112 source category
list, that regulations not be promulgated before the year 2000.
Another commenter (90) suggested EPA reconsider issuing
regulations under Section 183 (f) exclusively. Marine terminal
VOC controls that apply to smaller terminals only in the more
serious, nonattainment areas would be consistent with cost-
benefit considerations. The commenter agreed with EPA that the
result could be a complex regulation. According to the
commenter, the best course might be to limit Federal regulation
to RACT alternative I and leave to the States the option of
regulating smaller terminals, as may be needed for ozone
attainment demonstration, reasonable further progress, or
maintenance under Title X.
One commenter (90) provided a possible basis for omitting or
delaying MACT by uising its recommended interpretation of the
terms "best controlled" and "best performing," which leads to a
conclusion that the "best:" control is no control. Two other
alternatives were also offered: either regulate emissions from
marine tank vessel Loading facilities solely under Section 183(f)
rather than under both Sections 183(f) and 112 of the Clean Air
Act, or defer promulgation of the Section 112 rule. The
reasoning and justification for these alternatives were explained
at length and involved the CAA sections differing significantly
in purpose, scope, and regulatory approach.
2-19
-------
Response; As stated in the preamble to the proposed rule,
the Agency first intended to issue standards for marine tank
vessel loading operations solely under Section 183 (f) of the
Clean Air Act. However, the Agency decided to issue standards
under both Section 183 (f) and Section 112 of the Act because EPA
now believes that the best interpretation of the Clean Air Act
requires that standards be issued under both. The language of
Section 112 of the Act is clear: "[T]he Administrator shall
publish. . . a list of all categories and subcategories of major
sources and area sources of [HAP]." Clean Air Act
Section 112(c)(l), 42 U.S.C. § 7412(c)(l) [emphasis added].
Further, the Administrator "shall promulgate regulations
establishing emission standards for each category or subcategory
of major sources and area sources of hazardous air pollutants
listed for regulation pursuant to subsection (c)." Clean Air Act
Section 112(d)(l), 42 U.S.C. § 7412(d)(1). The tank vessel
loading operations source category is clearly a category of major
sources, as defined in the Act. The Act is thus clear on its
face that this source category be regulated under Section 112.
It is true that another section of the Act, Section 183 (f),
also regulates this category of sources. However, the fact that
two separate sections of the Act regulate the same source
category does not necessitate that one of the sections should be
ignored. Unless the regulations promulgated under one section
create an inescapable conflict with regulations promulgated under
the other section, both sections must be followed. The
regulations promulgated under the two sections are not in
conflict. The safety provisions that EPA and the Coast Guard
have required are equally applicable to sources regulated under
either Section 183(f) or Section 112.
In addition, though Section 183(f) anticipates RACT
standards, taking into account costs and other factors, whereas
Section 112 requires MACT standards, which provide the Agency
with less discretion to take costs into account, the regulations
promulgated under Section 112 are not in inescapable conflict
2-20
-------
with those promulgated under Section 183(f). The Agency believes
that any source regulated under both sections would have no
problem meeting the requirements of both standards at the same
time. Therefore, there is no conflict between the two sets of
requirements.
The Agency does not believe that it can ignore the specific
requirements of Section 112 merely because regulations
promulgated under that section are likely to be more stringent
than those promulgated under Section 183(f). The fact that one
set of regulations is more stringent than another does not mean
that the two regulations are in inescapable conflict. Only when
compliance with one set of necessary regulations prevents
compliance with another should the Agency consider ignoring the
clear language of the statute.
Moreover, the language of the statute itself and of the
legislative history indicates that Congress did not intend, by
its silence, to prevent regulations of marine tank vessel loading
terminals under Section 112. First, Congress explicitly said
where it wanted one section of the Clean Air Act to be exclusive
of further regulation under Section 112(d): see
Sections 129(h)(2) and 112(d)(9), Thus, Congress could have
added specific language to Section 183(f) and prevented the
Agency from regulating this source category under Section 112;
however, it failed to do so
Neither the statute nor the legislative history indicates
that Congress intended EPA regulations under Section 183 (f) to be
the exclusive regulation of these sources. In fact,
Section 183(f) explicitly provides that states may regulate
marine tank vessel loading processes and, in fact, requires that
any such regulations be as stringent or more stringent than the
Agency's regulations under Section 183(f).
Moreover, the legislative history indicates that Congress
was well aware that sources could be subject to dual regulation
under the Act and presumed that Section 112 standards would be
2-21
-------
more stringent than other standards. Page 167 of the Senate
Committee Report (Report 101-228) states the following:
An emissions limitation based on Section 112(d) will,
in most cases, be more stringent than a new source
performance standard for the same category of sources
or pollutants. It will also generally be more
stringent than an emissions limitation based on
reasonably available control technology imposed under
part D as the emission limitation here must be
equivalent to the most stringent degree of control
achieved in practice. That is appropriate as this
program is for control of extremely harmful air
pollutants. . . (emphasis added).
Therefore, the legislative history seems clearly to indicate
that regulations under Section 112 be in addition to any RACT
regulations required under Part D of Title I (which includes
Section 183(f)). (In any case, the commenter notes that EPA has
previously stated that it has the discretion to regulate these
sources under Section 112 of the Act)
In reference to the comment about EPA's ability to provide
full consideration of comments regarding the appropriateness of
regulation under Section 112, the Federal Register notice
announcing EPA's decision to regulate this source category under
Section 112 of the Act (58 FR 60021) stated that the public would
be afforded an opportunity to comment on the listing of this
source category under Section 112. The Agency notes that the
comments summarized above include comments on the necessity and
appropriateness of regulating these sources under the MACT
standards of Section 112. The Agency believes that it has
provided a complete and fair hearing on these issues and that it
approached these issues fully prepared to provide full
consideration to the comments. If EPA had been convinced by the
comments regarding the appropriateness of regulation under
Section 112, it could have requested revision of the consent
decree to take such comments into account. In addition, it is
arguable whether the consent decree requires promulgation of the
standards under Section 112(d) or whether a final action
determining that such standards are not appropriate would have
2-22
-------
satisfied the mandate. In any case, after a full hearing on the
issue of concern to the commenter, EPA continues to believe that
regulation of marine tank vessel loading sources under
Section 112 is appropriate. It is therefore unnecessary to
broach the issue of whether the consent decree would have needed
to be revised had EPA come to a different determination. It
should be noted that, although the Agency did receive comment
opposing the filing of the consent decree at the time the consent
decree was issued, no commenter suggested that the consent decree
be revised to require only that EPA reach a final decision
regarding the appropriateness of regulation under Section 112.
The EPA disagrees with the comments stating that
Section 183(f) should not apply in ozone attainment areas. The
EPA notes that nothing in the text of Section 183(f) indicates
that the regulations promulgated under Section 183(f) should be
restricted to nonattainment areas. In fact, the references in
subsection (5) to Section 111 indicate that Congress intended
standards under Section 183(f) to be similar in scope to those
under Section ill, which apply to any new sources in the United
States, not 'just those in nonattainment areas.
Moreover, Section L8MO requires EPA to "promulgate
standards applicable to the emission of VOC and any other air
pollutant from loading and unloading of tank vessels. . . which
the Administrator find causes or contributes to air pollution
that may be reasonably anticipated to endanger public health or
welfare," [emphasis added] This language clearly presumes that
EPA regulation would apply to ozone pollution and to other air
pollution problems associated with emissions from marine tank
vessel loading operations. The air pollution problem most
associated with such emissions, aside from ozone pollution, is
emissions of HAP.
Considerable evidence indicates that the HAP emissions
associated with marine tank vessel loading operations cause or
contribute to air pollution that can be reasonably anticipated to
endanger public health or welfare. Current emissions of VOC from
2-23
-------
marine tank vessel loading operations are estimated at
approximately 75,000 megagrams per year (Mg/yr), of which
approximately 8,000 Mg are HAP. The Agency has determined that
public exposure to these HAP is likely to occur in many areas of
the United States. As discussed below, exposure to such
substances has the potential to cause adverse health effects,
ranging from the minor symptoms of dizziness, fatigue, and nausea
to more serious and significant manifestations, such as
nonlymphocytic leukemia.
Approximately 53 different HAP are emitted from marine tank
vessel loading operations. For example, benzene, hexane,
toluene, xylenes, ethylbenzene, iso-octane, methyl tertiary butyl
ether (MTBE), naphthalene, and cumene are HAP emitted when
gasoline and crude oil are loaded into marine tank vessels.
These nine chemicals represent approximately 96 percent of the
total HAP emitted from marine tank vessels. Approximately 44 HAP
comprise the remaining 4 percent of toxic emissions. Although
liquids that are at least 70 percent by weight benzene are
already regulated by EPA, roughly 700 Mg/yr of benzene are
emitted from marine tank vessel loading operations that are not
subject to current regulations; one facility alone emits
approximately 410 Mg/yr of benzene. Approximately 6,900 Mg/yr of
hexane, toluene, xylene compounds, ethyl benzene, iso-octane,
MTBE, naphthalene, and cumene are emitted from marine tank vessel
loading operations.
The following presents a discussion of different HAP and
their adverse health affects.
Benzene is an important component of gasoline and crude oil.
It is also a HAP for which adverse health effects are well
documented. The EPA has classified benzene as a verified
Group A, or known human carcinogen. This classification means
that there is sufficient evidence to conclude that benzene causes
an increased risk of cancer in humans. Chronic inhalation of the
chemical has been demonstrated to increase the incidence of
nonlymphocytic leukemia in occupationally exposed individuals.
2-24
-------
Leukemias and lymphomas, as well as other tumor types, have been
observed in experimental animals that have been exposed to
benzene by inhalation or oral administration. A number of
adverse, noncaricer health effects have also been associated with
exposure to benzene. Benzene exposure has been associated with
significant increases in chromosomal aberrations of bone marrow
cells and peripheral lymphocytes. Benzene is also known to cause
disorders of the blood. People with long-term exposure to
benzene at low to moderate concentrations may experience harmful
effects on the blood-forming tissues, especially the bone marrow.
These effects can disrupt normal blood production and cause a
decrease in important blood components, such as red blood cells
and blood platelets, leading to anemia and a reduced ability to
clot. Chronic exposure to benzene at comparable or even lower
levels can be harmful to the immune system, increasing the chance
for infection and perhaps lowering the body's defense against
tumors by altering the number and function of the body's white
blood cells. Studies performed with pregnant animals show that
benzene may also cause adverse effects to the developing fetus,
including low birth weight, delayed bone formation, and bone
marrow damage.
Although benzene is only one of several HAP emitted from
marine tank vessel loading operations, it is the only one for
which the threat of adverse health effects has been quantified.
However, the other HAP identified above can exhibit adverse
health effects that may potentially endanger human health and
welfare. Toluene, another important component of marine tank
vessel loading operations emissions, is one such example.
Although the carcinogenic effects of this HAP cannot be evaluated
at this time due to insufficient evidence (toluene is classified
as a Group D carcinogen). toluene exposures have been associated
with several adverse, noncarcinogenic effects. The primary
impact from human inhalation of toluene is depression of the
central nervous system, including central nervous system
dysfunction, narcosis, impaired neuromuscular and cognitive
2-25
-------
function, and permanent damage over longer periods of exposure.
Case studies of exposed workers suggest that toluene can also
cause upper respiratory tract and eye irritation. Toluene may
also induce adverse developmental effects in laboratory animals,
such as skeletal abnormalities, low fetal weights, and retarded
skeletal growth.
Naphthalene, xylenes, and ethylbenzene--other HAP in
gasoline vapors--are considered Class D carcinogens because the
available data are inadequate to determine their carcinogenic
potential. Available evidence suggests that some adverse,
noncancer health effects are associated with exposure to
naphthalene, xylenes, and ethylbenzene. Cumene, still another
HAP in gasoline vapors, has yet to be investigated for its
carcinogenicity and has no classification at this time.
The health effects associated with naphthalene inhalation
have been examined on a limited basis. Exposure to acute levels
has been documented to cause acute hemolytic anemia in exposed
individuals. Recent studies have begun to reevaluate the
carcinogenic effects resulting from naphthalene inhalation.
Laboratory tests suggest that female mice exposed to elevated
levels of naphthalene develop pulmonary alveolar/bronchiolar
adenomas, and that for female mice there is some evidence of
carcinogenic activity. In addition to these inhalation effects,
many studies have examined health effects associated with oral
administration of naphthalene. Naphthalene taken orally has been
shown to produce cataracts and bronchiolar epithelial cell
necrosis in laboratory animals. When inhaled, some naphthalene
may enter the systemic circulation and eventually follow the same
metabolic path as naphthalene taken orally. Given this
information, it is possible that inhalation of naphthalene may
also lead to cataracts and pulmonary cell damage.
Accounts of exposure to xylene vapor indicate there is
little difference in toxicity between individual xylene isomers
and xylenes (mixed). Occupational exposure to xylene at high
levels indicates that xylene can have a narcotic effect as well
2-26
-------
as cause liver and kidney damage. Additional toxicity tests
support the findings that short-term xylene inhalation may lead
to central nervous system depression and reversible kidney and
liver damage. Xylene vapor has also been shown to be irritating
to the eyes, nose, skin, mucous membranes, and throat. At high
concentrations, these effects may be expressed as dizziness,
staggering, drowsiness, and unconsciousness.
Subchronic and chronic exposures to ethylbenzene appears to
target the liver and kidneys and is indicated through increased
liver and kidney weight in animal inhalation studies.
Additionally, studies have indicated that exposure to
ethylbenzene at acute levels appears to target the central
nervous system, causing loss of coordination, narcosis, and
convulsions. Exposure to levels over 100 ppm can lead to
fatigue, sleepiness, headaches, and eye and respiratory
irritation. The limited reproductive studies done on
ethylbenzene suggest that subchronic exposure may lead to
testicular degeneration in some test animals.
Both cumene and MTBE are used as additives in high-octane
gasoline and vary in concentration depending on the octane of the
gasoline. Limited information is available for cumene; however,
estimates of MTBE concentrations in gas can range from 2 to 8
volume percent, although concentrations reaching 11
volume percent have been approved.
The available information on cumene suggests that it may
affect the central nervous system, causing dizziness., narcosis,
and faintness, Exposxrre may also be linked to narcosis and
faintness and to damage to the kidneys, liver and lungs. Over
the short term, cumene has beer reported to irritate the eyes,
skin, nose, and throat.
The EPA continues to study the health effects associated
with MTBE; however, results of acute and subchronic exposure
studies suggest a potential peripheral neurotoxicity. This
evidence is supported by animal experiments that report a
disruption in motor activities in all acutely exposed rat groups.
2-27
-------
At higher levels, MTBE has been shown to adversely affect fetal
development in mice. Observed effects include reduced body
weight, increases in incidence of cleft palate, and skeletal
variations.
The EPA has determined that HAP emissions from marine tank
vessel loading operations cause or contribute to air pollution
that may be reasonably anticipated to endanger public and health
welfare and, therefore, has decided to regulate HAP emissions
from tank vessel loading operations as well as VOC emissions. As
described above, such emissions have broadly been found to be a
health hazard. The Agency need not show that emissions from each
marine tank vessel loading facility regulated under
Section 183(f) cause or contribute to such air pollution.
The Agency received a comment that the language in
Section 183(f) referring to a finding that an emission "causes or
contributes to air pollution that may be reasonably anticipated
to endanger public health" is a phrase triggering regulation only
of criteria pollutants, not hazardous pollutants. However, this
argument is belied by the language of Clean Air Act
Section 112(b), which is applicable to HAP and which is quite
similar, to the language of Section 183(f). Moreover, prior to
amendment in 1990, Clean Air Act Section 122 required EPA to
review information on emissions of certain HAP to determine
whether such emissions "cause or contribute to air pollution that
may be reasonably anticipated to endanger public health. Similar
language also appears in other sections of the Clean Air Act that
require nationwide standards (See Sections 115, 202, 211, and
231.)
The Agency also notes that the location of Section 183(f) in
a broad subpart called "Additional Provisions for Ozone
Nonattainment Areas" does not demonstrate that Section 183(f)
applies only in nonattainment areas. The title of an Act cannot
contradict the more specific language contained in the body of
the Act. See Mohegan Tribe v. Connecticut. 638 F.2d 612 (2d
Cir.), cert. denied 452 U.S. 968 (1980). In fact, other sections
2-28
-------
of Subpart 2 (e.g., Section 184) apply in areas beyond ozone
nonattainment areas. While it is likely that Section 183(f)'s
location is a result of Congress's concern about VOC emissions
from tank vessel loading operations, the reference to "any other
air pollutant" shows that such emissions were not Congress's only
concern.1
Regarding the comment that EPA should delay regulations
under Section 112, EPA believes that it is most appropriate to
promulgate regulations under Section 112 for marine tank vessel
loading facilities at the same time it is promulgating
Section 183 (f) regulations. This provides regulated sources with
a single, all-encompassing regulatory action and also allows EPA
to coordinate the two regulations as much as possible.
*The Agency notes that Representative Jones mentioned the
possible association between emissions from tank vessel loading
operations and hazardous air pollutants in the floor debates on
the amendments. 149 Cong. Rec. H12929 (Oct. 26, 1990).
2-29
-------
2.1.4 Vapor Pressure Limit
Comment; One commenter (04) stated the EPA should limit
applicability to only volatile HAP; those with a negligible vapor
pressure should be exempted from control on the basis of cost-
effectiveness alone. The commenter suggested that there is
currently no control in the industry of these compounds and
challenged the Agency to prove otherwise.
Three commenters (34, 50, 75) recommended that the rule
include a list of relevant volatile organic HAP (VOHAP) that
should be controlled and for which testing and monitoring
requirements would apply. The commenters suggested that the list
of relevant VOHAP contained in the Technical Support Document be
considered. The commenters also cited precedent set in the
hazardous organic NESHAP (HON), in which EPA specified HAP of
interest. One commenter (04) suggested that EPA examine existing
State requirements upon which the Agency chose to base its floor
determination to ascertain where the correct volatility limit
should be. The commenter suggested that EPA set a minimum vapor
pressure of no less than 0.5 psia.
Five commenters (34, 56, 71, 75, 78) stated that the
Agency's MACT floor analysis failed to consider that the floor
level of control does not apply to products with low vapor
pressures. The commenters added that no statewide requirements
are in existence for the loading of low vapor pressure materials.
If EPA wishes to use three State rules as models to establish a
MACT floor, it must also use all exemptions in the State rules.
The commenters stated that EPA made a false assumption that
products of all vapor pressures are controlled equally in
facilities that are equipped with vapor control systems. The
commenters also noted the inconsistency in regulatory approaches
between this rule and in both the HON and Petroleum Refinery MACT
standards, where EPA considered product vapor pressures in
establishing floors. The EPA should eliminate control
requirements for low vapor pressure products because the MACT
floor for loading of these materials is zero and because it is
2-30
-------
not cost effective to control these products. Also, new source
MACT was incorrectly determined because it does not account for
vapor pressure or HAP concentration.
Two commenters (34, 78) believed that for a vapor pressure
limit to work, the vapor pressure exemption would need to be
based on the partial pressure of HAP present. For the HON rule,
EPA set for transfer operations a 10.3 kilopascal (1.5 psia)
vapor pressure limit for transfer racks that contain organic HAP.
In addition, EPA issued final rules for the floor definition,
which set a vapor pressure limit of 13.1 kilopascals (1.9 psia)
for Group 1 storage tanks of 75 to less than 151 cubic meters
capacity and 5.2 kilopascals (0.75 psia) for Group 1 storage
tanks of 151 cubic meters capacity or greater.
One commenter (78) noted that a low vapor pressure limit
would be far simpler to use than the 93 percent overall HAP
reduction efficiency approach selected in the proposal. No
complicated calculations and projections of future material to be
loaded in the compliance interval would be necessary.
Five commenters (39, 42, 56, 72, 75) stated EPA should
eliminate requirements for low vapor pressure products because
they account for very little of total HAP emissions from vessel
loading and because it is not cost effective to control emission
streams from the loading of these materials for very little
environmental benefit. Additionally, some commenters cite a
study conducted by Radian Corporation on the cost effectiveness
of controlling low vapor pressure compounds and statt that the
results of the analysis indicate that the costs would lead EPA
not to control such products„ The results of this study have
been offered to EPA in support of a low vapor pressure limit.
Four commenters (23, 68 ,71, 78) believed that the cut-off
value of 1.5 psia (at the loading temperature of the liquid) in
Louisiana's marine tank vessel loading regulations is
appropriate. Four commenters (28, 34, 50, 73) recommended that a
low vapor pressure limit of 2 psia be adopted.
2-31
-------
Six commenters (28, 34, 50, 56, 73, 75) stated recovery
systems are primarily applicable to products with high vapor
pressures, generally greater than 2 psia. Consequently,
facilities loading products with lower vapor pressures would most
likely install combustion devices. However, the extremely small
emission reduction obtained in controlling vapors from low vapor
pressure product loading operations would be offset by the
increase in criteria pollutants associated with the combustion
device.
One commenter (42) stated a calculation procedure needs to
be established to estimate loading losses from low vapor pressure
materials, such as No. 6 oil and asphalt.
One commenter (71) added that without adding a MACT
applicability vapor pressure minimum, all of the 100 existing,
controlled terminals that EPA indicates have uncontrolled loading
emissions and all the terminals in Louisiana that just meet the
State requirement of 90 percent control will have to add
additional controls of little or no incremental benefit.
One commenter (23) stated that Section 63.565(d) does not
list any exempted commodities nor does it contain explicit
procedures that would apply in determining the exemption of any
commodity, even though it is cross-referenced by 63.562(b) as
having that purpose. The commenter urged EPA to provide explicit
procedures whereby an owner or operator may seek exemption of a
commodity in addition to those that should be exempted by a vapor
pressure limit.
Two commenters (36, 46) suggested EPA narrow the scope of
the rule to apply only to those facilities that load gasoline or
crude oil. This limitation would be consistent with existing
State rules and would be cost-effective. One commenter (90)
stated that this issue is easily resolved because products other
than gasoline and crude oil are not transferred in quantities
over 1 million or 10 million barrels per year respectively at an
individual terminal.
2-32
-------
Response; The Agency acknowledges and agrees with those
comments concerning no State regulations existing for loading low
vapor pressure material. The proposed rule enabled individual
facilities to determine which products to control to achieve the
93 percent mass limit. Therefore, facilities would not have had
to control low vapor pressure liquids under the proposed rule if
higher vapor pressure liquids were available for control. The
promulgated rule explicitly exempts low vapor pressure liquids
consistent with State requirements and recalculated the control
requirement for liquids above the vapor pressure limit.
Therefore the MACT floor is no control for liquids having a vapor
pressure below 1.5 psia and 97 percent control for liquids having
a vapor pressure 1.5 psia or greater. The format of the standard
was changed to an efficiency format to reflect the new approach.
The issue of cost effectiveness to control emission streams
from the loading of these materials is also a realistic concern.
As the MACT floor for regulation of such activities is no
control, (because no State has regulated these activities), EPA
has discretion, based on Section 112(d)'s criteria for going
beyond the floor to institute a vapor pressure limit. Because of
the high costs cited by cotnmenters, the Agency elected not to
require controls more stringent than the MACT floor for these low
vapor pressure HAP, The Agency, therefore, selected a vapor
pressure limit of 1.5 psia to be used in determining the HAP
emissions reduction for the final standards. Control of HAP
having vapor pressures below this limit is not required to meet
the standards.
2-33
-------
2.1.5 Commodity Type
Comment: One commenter (22) stated the proposed new
standard apparently does not exclude methane and ethane, which
can be present in blanketing systems, from the total VOC
definition. Because vapor recovery units are not designed to
remove these compounds, compliance with the emission standard
could be difficult if these compounds are not excluded from the
total VOC definition.
Response; The Agency's definition for VOC has not changed
and does exclude methane and ethane (based on the fact that these
compounds are not ozone precursors).
Comment: One commenter (78) stated the proposal does not
consider loading of mixed materials or materials that may include
small quantities of HAP as impurities.
One commenter (80) stated that although the HAP emission
threshold may act as a surrogate in some cases, this threshold
only applies to the MACT standard, and EPA has proposed a MACT
level of control that is lower than the RACT level of control.
An existing marine loading terminal at a SOCMI facility would
therefore be subject to less stringent standards than a marine
loading facility that only processes crude oil, for example, even
if the products loaded at the SOCMI terminal were substantially
more hazardous. The commenter urges EPA to consider total
loading throughput for the purpose of determining applicability
and suggests that products can be grouped according to relative
toxicity.
Response: The Agency evaluated means to rank materials
based on relative toxicity. However, there is currently no
established procedure for ranking or comparing VOC and HAP.
Therefore, for purposes of this standard, all HAP are treated
similarly. Since many of the compounds on the HAP list are also
VOC, there is no way to differentiate or group facilities for
MACT and/or RACT applicability.
The Agency agrees with the commenter regarding the control
of liquids with very low concentrations of HAP. A definition of
2-34
-------
"impurities" similar to that used in the HON has been included in
the final regulation, and an exemption from the MACT standards
has been established for loading berths that only transfer
liquids containing HAP as impurities.
Comment; One commenter (30) stated that, in setting
standards, EPA should not penalize terminals that handle
gasolines containing oxygenates (such as MTBE) for the express
purpose of meeting the oxygenated and reformulated gasoline
requirements of the CAA.
Response: The Agency is aware of the increasing use of
gasolines containing oxygenates (such as MTBE). The Agency notes
that Section 112 of the Act does not provide a mechanism for
discounting HAP emissions because their presence is mandated by
other regulations. Rather, the Act requires calculation of total
HAP emitted by the source.
2-35
-------
2.1.6 Valdez Marine Terminal (VMT)
Comment; One commenter (06) believed that the Alyeska
Pipeline Service Company's Valdez Marine Terminal (VMT) allowable
benzene emissions "may reasonably be anticipated to endanger
public health or welfare," and, therefore, it is within the
authority of the EPA under 183(f) to regulate HAP emissions under
the provisions of the Section 183 RACT requirements. The
commenter stated that previous efforts by the Alyeska Pipeline
Service Company (Alyeska) to demonstrate that the ambient impacts
from benzene emissions at the VMT are insignificant reinforce
this point.
Another commenter (07) felt the VMT should be regulated
under Title III of the regulations as opposed to both Titles I
and III. Regulation under Title III would solve many of the
concerns expressed by Alyeska, including the ambitious 2-year
schedule.
One commenter (32) stated there is no basis in the
Administrative Record to support a determination that emissions
from the VMT present a risk to human health and welfare. Until
EPA makes a determination, it cannot impose standards on the VMT
under Section 183 (f). Regulation under Section 112 does not
eliminate the requirement to evaluate the health and welfare
risks under Section 183(f). The EPA cannot simply ignore the
threshold requirements of Section 183(f) by stating that it
"would require intensive effort" to make threshold findings or
that sources are covered by Section 112, especially since (1) the
Section 183 (f) RACT standard is more stringent than the
Section 112 MACT standard, and (2) Section 183(f) has a shorter
timetable for compliance. Finally, EPA must consider that the
VMT will have to meet the Section 112 MACT standards for HAP
within 1 year after the compliance deadline for the RACT standard
under Section 183 (f). Under any set of assumptions, the
incremental risk reduction achieved by controlling emissions for
the 1-year period prior to the MACT standards will be negligible.
2-36
-------
This commenter (32) further stated that ozone nonattainment
is not a basis for regulating the VMT under Section 183 (f) since
the VMT is located in an ozone attainment area. The Valdez Air
Study Review (VASR) concluded that benzene emissions would pose
an average lifetime cancer risk of no greater than 2.5 chances in
l million assuming that none of the VMT tanker loading berths had
a vapor control system. Cancer risks drop to 0.72 in 1 million
if berths 4 and 5 were controlled and berth 3 had operational
limitations, and to 0.66 in 1 million when all three berths
employed vapor control technology. The commenter concluded that:
(1) the risks from the VMT, without controls, are negligible and
within EPA's tolerable range of risk; (2) Alyeska's proposed
vapor control strategy will reduce health risks even further; and
(3) the incremental risk benefit of controlling Berth 3 is
sufficiently small as to be of no medical consequence. The
commenter disagreed with the VASR, which concluded that terminal
emissions from the VMT contribute to 95 percent of the total
benzene emissions in the Port of Valdez region, and insisted that
if the VMT contribution of benzene emissions were as high as that
estimated by the VASR, then the control strategy put forth by
Alyeska would reduce benzene emissions to levels lower than the
original, uncontrolled emissions estimated by Alyeska. Finally,
the commenter recommended that EPA adopt a RACT standard for the
VMT that requires Alyeoka'B proposed strategy.
One commenter (43.) recommended that EPA not give Alyeska
special treatment. Any concessions granted to Alyeska should be
granted to the industry as a whole. Another commenter (06)
stated that benzene emissions from crude oil marine terminals
should be regulated under the Section 183 (f) RACT provisions.
Response; The Agency reviewed all of the comments related
to the Alyeska Pipeline Service Company's Valdez Marine Terminal
(VMT). The Agency first notes that the Title III standards are
technology driven and are not based on the risk posed by a
particular source. In some ways, the comments regarding the
applicability of the Title I regulation are similar to those
2-37
-------
summarized in Section 2.1.3 of this document. As discussed above
in Section 2.1.3 of this document, EPA need not make a
determination of risk for each regulated source in order to
regulate that source. As discussed above, marine tank vessel
loading facilities emit several types of HAP, including benzene,
that have been found to be harmful to human health or that can be
reasonably anticipated to endanger public health or welfare. The
VMT is by far the largest emitter of such HAP in this source
category. In any case, a health study performed by the Alyeska
Pipeline Service Company indicated a risk above 10"*, which is
higher than the acceptable risk used by the Agency in considering
the delisting of source categories under Section 112. Therefore,
EPA believes that regulation under both Section 112 and
Section 183 (f) is appropriate.
However, the Agency believes that this particular facility
warrants careful consideration for the following reasons:
(1) the VMT is the only oil-loading facility of its size that
would be subject to these standards; (2) special circumstances
require keeping the oil moving (i.e., the limitations on storage
capacity and the need to keep the oil moving through the
pipeline); (3) periods of extreme bad weather complicate the
loading schedule; and (4) information submitted by several
commenters shows that the volume of oil loaded at the VMT has
decreased and will continue to decrease in the future; therefore,
infrequent and declining use of VMT's auxiliary loading berths
(i.e., berths other than the two berths for which installation of
control technologies is under development) is anticipated. These
considerations will be discussed in more detail below under
subcategorization.
Because of these reasons, and particularly because of the
declining throughput anticipated at VMT, the Agency has
determined that a phased-in approach is warranted for the Title I
regulation of the VMT. This approach requires control of loading
emissions at the primary loading berths (2 berths) and allows for
2-38
-------
continued, uncontrolled use of remaining berths for a limited
time, with throughput limits on the use of such berths to reflect
declining throughput. By selecting this approach, the Agency
reasserts its position detailed in the preamble to the proposed
rule and in this document that the distinction of attainment area
and nonattainment areas in these regulations is not warranted and
is not the basis for this decision. The selected approach
requires a reduction in emissions by 98 percent efficiency of all
throughput loaded from at least two of the terminal's loading
berths. Maximum throughput limits (above which other VMT loading
berths would be required to reduce emissions by 98 percent) are
included as part of the standards. These throughput limits
address the projected decreasing throughput that would only
necessitate the use of two berths for routine loading after 2001.
Provisions to allow for scheduled maintenance of the controlled
berths are also established in the VMT standards. The Agency
notes that the allowance for scheduled maintenance is provided
for by existing mechanisms that allow States to grant waivers for
maintenance purposes. Indeed, such maintenance waivers are
embedded in any MACT floor for such large marine tank vessel
loading operations since these MACT floors were developed (in
part) based on existing State regulations. The Agency
established the 40 day maintenance allowance for VMT in order to
provide needed flexibility to this source and to reduce the
burden of requesting waivers for anticipated maintenance.
Some commenters initially noted that the cost of controlling
VOC may be high relative to the benefits of controlling VOC at a
remote site in an Arctic ozone attainment area. Additionally,
some commenters initially stated that the benefits of controlling
HAP would not appear to justify the costs. However, the Agency
has also considered later comments from the APSC, the State of
Alaska, the Prince William Sound Regional Citizens Advisory
Committee (a local citizens group) and private citizens in
determining MACT/RACT for the VMT. These commenters agreed that
a Federal rule mandating control of primary emissions at the APSC
2-39
-------
was acceptable. After careful consideration of the costs, the
environmental impacts and the comments, the Agency decided that
MACT for this subcategory was control beyond the level of the
MACT floor (see Docket A-44-90, Item Number IV-B-2).
2-40
-------
2.1.7 Ballasting and Bunkering
Comment: One commenter (80) agreed with the prohibition on
ballasting and cited a similar prohibition in California's Bay
Area. Another commenter (38) agreed that ballasting operations
should be prohibited and stated that appropriate recordkeeping
and, if possible, adequate monitoring should be required to
certify that ballasting emissions do not occur; the final rule
should also prohibit nonsegregated ballasting of HAP and VOC
containing segments of transport vessels. This commenter
indicated the regulations need to provide guidance as to how
ballasting operations will be prevented by the facility. One
commenter (79) suggested that the standard regulation state that
only marine tank vessels with segregated ballast tanks can be
used for affected facilities.
Several commenters (04, 23, 24, 27, 34, 35, 36, 59, 85) did
not agree with the proposed ballasting requirements and provided
the following arguments:
1. The marine terminal owner or operator should not be
responsible for enforcing ballasting requirements because
ballasting is done by the vessel operators (Commenter 24);
2, Coast Guard requirements already appropriately address
this issue (Commenter 04); EPA should grandfather marine tank
vessels built prior to 1980 to allow ballasting without the use
of segregated compartments as has been done by the Coast Guard
(Commenter 23); the rule should allow for the uncontrolled
loading of ballast in accordance with the rules established by
the Coast Guard until the phase-out of nonsegregated ballast
vessels is complete {Commenters 34, 85, 59); the compliance
period for ballasting control is not addressed, resulting in a
3-year compliance period which makes the rule incompatible with
the second ballasting alternative for meeting Coast Guard
Standards that do not go into effect until 2003 (Commenter 34);
3. The vapor control requirements will force many small and
medium sized facilities out of business because they cannot
afford vapor control systems but serve vessels that ballast
2-41
-------
(Commenter 36); ballasting operations on all tank vessels would
alter the competitive relationship between integrated tank barges
(ITBs) and barges and cause an undue economic burden for certain
operators because very few barges ballast, and ITBs and smaller
tankships usually must ballast by design (Commenters 27, 35);
4. Emissions from ballasting are not significant .because
vessel tanks that do not contain segregated ballast are filled
with inert gas, and no more than 50 percent of the tank is loaded
with water during a ballasting operation, which means the last
20 percent of the headspace vapor thought to be most saturated
with volatile compounds is not emitted (Commenter 36); the
occurrence of ballast water entering a cargo tank is so rare not
to warrant EPA's attention and only clutters the regulation
(Commenter 04);
5. Ballasting is necessary for some vessels to acquire the
necessary draft to maneuver through areas (Commenter 36);
ballasting is an essential requirement for ship safety and cannot
be avoided, and the option that the vessel not perform ballasting
at any time is not realistic (Commenter 34); the only occasions
when water enters cargo tanks is when safe operations of vessels
require it, and EPA should not interfere in these matters
(Commenter 04); at some time, generally shortly after departure,
the ship has to be fully ballasted to limit stresses in heavy
weather (Commenter 123); and
6. The rule fails to specify that ballasting operations are
the only nonloading operation controlled, leaving open the issue
of product unloading requirements (Commenter 34);
One of these commenters (23) recommended that EPA provide
data to support their assertion that significant HAP emissions
are generated during ballasting operations. One commenter (04)
recommended that EPA delete all references to the control of
vapor emissions due to the ballasting of cargo tanks. Two of the
commenters (27, 35) recommended that vessels that are not
required to have segregated ballast tanks, such as ITBs and small
tankships (less than 30,000 dwt), be excluded from ballasting
2-42
-------
vapor control requirements. Another commenter (36) recommended
that the rule require that during a ballasting operation, no more
than 50 percent of the vessel volume be filled with water. This
commenter stated that the ballasting prohibition requirement
should be eliminated altogether. One commenter (123) suggested
an addition for the end of § 63.562(h): "while at the affected
source."
Two commenters (31, 34) stated EPA should clearly indicate
that unloading operations other than ballasting are exempt from
the rule and should clearly indicate that calculation of major
source status should include estimation of only ballasting
emissions during unloading. These commenters also recommended
that facilities not be required to assume that all vessels have
nonsegregated ballast tanks at their facilities for purposes of
determining applicability in the regulations. One commenter (06)
explained that the preamble states that the decision to regulate
emissions from ballasting was made as part of the RACT
provisions. Therefore, the commenter (06) stated there should be
corresponding sections limiting ballasting VOC emissions from
ballasting operations 1 year sooner for sources subject to both
RACT and MACT provisions.
Two commenters (23, 34} stated that recordkeeping for
ballasting in vessels with segregated ballast tanks is
unnecessary and a violation of the Paperwork Reduction Act, since
there can be no environmental benefit from the requirement of
obtaining documentation describing ballasting procedures.
Another commenter (32'! stated that ballasting records, if kept at
all, would be more appropriately recorded and maintained by the
vessel owner or operator, rather than the affected source;
therefore, the requirement, should be deleted from the proposed
regulation. One commenter (31) suggested the ballasting
requirement should specify what documentation will suffice for
compliance with the ballasting requirement. The commenter feared
that § 63.567(h)(10) could be misinterpreted to require a more
detailed statement than i-.he three general descriptions listed in
2-43
-------
the preamble. The commenter (31) feared that enforcement
personnel might interpret the rule as requiring something akin to
the standard operating procedures (SOPs) used to run chemical
plants. This commenter suggested rewording of § 63.567(h)(10) as
follows: "Description of ballasting procedures. This paragraph
is satisfied if the owner or operator has documentation either
that the vessel does not perform ballasting at any time, that the
vessel meets the Coast Guard standards, or that ballasting
emissions are ducted to a control device."
One commenter (31) stated this rule should exempt cargo
consolidation and bunkering activities on board ships or barges
that are completed without interaction with the terminal.
Response: The Agency agrees that ballasting operations are
performed by vessel operators, not by the regulated terminal
sources. The Agency also agrees that regulating ballasting
operations would be difficult to enforce. The Agency's intent in
prohibiting ballasting emissions in the proposed regulation was
to provide a cross-reference with existing Coast Guard
regulations addressing ballasting in nonsegregated vessels. The
Coast Guard rules require vessels to have segregated ballast
tanks for crude oil loadings. The Agency sees no benefit to
restating Coast Guard requirements for ballasting. Moreover, EPA
agrees that the relatively low amount of actual emissions
associated with ballasting does not justify dual regulation of
ballasting. As discussed in the proposal TSD, the total VOC
emissions from crude oil tankship ballasting were estimated to be
approximately 950 Mg/yr. Based on the portion of HAP in crude
oil vapor, total HAP emissions from ballasting are less than 120
Mg/yr. Ballasting emissions will diminish in the future because
tankships built since 1980 are required by domestic law and
international 'agreement to use segregated ballast tanks (SET)
that do not emit vapors during ballasting. Therefore, in order
to prevent confusion in the regulated community, the Agency does
not address ballasting or bunkering emissions in the final
2-44
-------
regulation. The Agency defers to the U.S. Coast Guard's existing
standards.
2-45
-------
2.1.8 Regeneration of Carbon Beds
Comment: One commenter (41) recommended that EPA clarify
the language in the final rule to prohibit direct emissions from
steam stripping. The proposed rule prohibits emissions from
steam stripping used to regenerate carbon beds. When these
regeneration vapors are taken to either recovery or destruction,
however, there will be an efficiency associated with the recovery
or destruction device. One commenter (34) suggested that
Section 63.562(i) state that the carbon bed regeneration vapors
should be routed through the on-line bed and not vented to the
atmosphere. One commenter (73) stated that other methods, such
as steam stripping back to the process or to combustion devices,
should be allowed as a means of determining compliance for carbon
bed regeneration. Another commenter (38) noted there are steam
stripping designs that route the noncondensables from the
regenerating units back to absorbing units.
One commenter (06) explained that the preamble states that
the decision to regulate emissions from carbon regeneration was
made as part of the RACT provisions. Therefore, the commenter
stated there should be corresponding sections limiting carbon
regeneration VOC emissions carbon regeneration operations 1 year
sooner for sources subject to both RACT and MACT provisions.
Response; The Agency agrees with the comments recommending
changes to the steam-regeneration prohibition. The final rule
states that the carbon bed regeneration vapors must be routed
through the on-line bed and not vented to the atmosphere. The
Agency also agrees with the commenter regarding the applicability
date for sources subject to RACT.
2-46
-------
2.2 INCLUSION WITH THE PETROLEUM REFINERY SOURCE CATEGORY AND
CONTIGUITY
2.2.1 General
The EPA requested comments on the concept of expanding the
petroleum refinery source category to include marine tank vessel
loading and unloading operations. Under this approach, marine
tank vessel loading operations subject to MACT would be split
into two subparts-- those operations that are collocated at
refineries and those that are not. The operations collocated at
refineries would be combined with and become part of the refinery
source category. Sources subject to RACT are not allowed to
average emissions with petroleum refinery operations.
Comment: Five commenters (91, 92, 93, 96, 99) support the
revision of the refinery source category to include collocated
marine tank vessel loading operations. Three commenters (94, 98,
100) opposed such a revision and two others (95, 97) recommended
that EPA allow an individual refiner the option of complying with
the marine tank vessel loading and refinery MACT rules separately
or complying with a combined terminal/refinery MACT rule to
comply with the emission standards.
One commenter (94) was concerned that the proposed amendment
could result in a relaxation of the proposed HAP emission control
standards applicable to the VMT. The specific concerns involved:
- possible lower MACf floor requirements applicable to
marine terminals not colJocated at refineries;
- inconsistency with the requirements of Section 112 (c) (1)
of the CAA regarding identification of major source categories;
- relaxation of: proposed HAP control standards applicable to
nonrefinery marine termsaals by allowing certain commodities or
operating berths to be partially controlled; and
- allowing the WT to average their marine tank vessel
loading HAP emission reductions with on-site storage tanks and
process vents in such a way that total emissions from the
facility are higher than if the Marine Tank Vessel Loading Rule
were implemented as currently proposed.
2-47
-------
Response: The Agency agrees with the commenters who
supported the position that the petroleum refinery source
category should be expanded to include marine tank vessel loading
operations at petroleum refineries. This expansion will enable
emission reduction by either direct control of loading emissions
or emission averaging with other emissions points at the
refineries. Several commenters to the marine tank vessel loading
operations proposed NESHAP supported averaging of refinery
process unit emissions with emissions from marine terminals and
gasoline distribution operations that are located at refineries.
The commenters cited more cost effective emission reduction as
the advantage of including these emission points in emissions
averaging, and specifically commented that the costs per Mg
emission reduction of the marine tank vessel loading emission
controls are high. These commenters also claimed that emission
calculation procedures for loading are well established and that
adding marine loading to the averaging provisions will not
appreciably increase the complexity of enforcement. Other
commenters opposed including marine tank vessel loading and
gasoline distribution in emissions averaging. Some commenters
claimed that these are separate source categories and the Act
does not permit averaging across source categories. Others were
concerned that including marine loading in averages could result
in uncontrolled peak emissions.
In the final rules, emissions from marine tank vessel
loading operations, bulk gasoline terminal or pipeline breakout
station storage vessels, and bulk gasoline terminal loading racks
at petroleum refineries are allowed to be included in emissions
averages. The definition of the petroleum refinery source
category and source has been changed to include marine tank
vessel loading operations, bulk gasoline terminal and pipeline
breakout station storage vessels, equipment leaks, and bulk
gasoline terminal loading racks classified under SIC codes 5171
(Petroleum Bulk Stations and Terminals) and 4613 (Refined
Petroleum Pipelines) that are located at refinery plant sites.
2-48
-------
Note that these operations are closely connected with refinery
process unit operations since they transfer products of the
refinery process units.
A marine tank vessel loading operation or gasoline terminal
or pipeline breakout station that is collocated at a petroleum
refinery can be considered part of the same source as the
refinery subject to this rule. Because these operations are
redefined to be part of the source subject to the rule, the
prohibition against intersource averaging is not violated.
In keeping with the EPA's stated goal of increasing
flexibility in rulemakings, this decision has been made to
provide more opportunities to average. This optimizes the
opportunities for refiners to find cost-effective emission
reductions from overall facility operations on-site. Costs and
cost-effectiveness of controlling a particular kind of emission
point, such as marine tank vessel loading, will vary depending on
many site-specific factors. Emissions averaging allows the owner
and operator to find the optimal control strategy for their
particular situation.
Regarding the comments received on the impact of the
redefinition of the refineries source category on the regulation
of the Alyeska Pipeline Service Company's Valdez Marine Terminal
(VMT), the Agency notes that the VMT is not collocated at a
refinery and would not be affected by this change.
2-49
-------
2.2.2 Contiguity and Collocation
Comment: One commenter (92) stated that the use of
emissions from the entire contiguous source to determine a
terminal's major source status is not an optimal resolution,
since it is likely that relatively few marine tank vessel
terminals are collocated with refineries. The commenter (92)
requested the term "collocated" be better defined, and two
commenters (92, 96) suggested that additional clarification is
needed for terminals that are collocated but noncontiguous.
Another commenter (130) stated that its tank vessel loading
operations are not contiguous to its other facilities and that
EPA has not provided a definition of major source or contiguous
and requested that a definition of "contiguous" be provided. Two
commenters (28, 133) stated that the contiguous definition should
be clarified for terminals that are not geographically close to
the refineries or plants (sometimes greater than a mile apart)
but are currently assumed to be "contiguous." Another commenter
(72) suggested language for the definition of "contiguous area."
Response: Several comments indicated various levels of
confusion about "collocated" and "contiguous". Contiguous means
located on the same or adjacent properties. It includes
properties that are divided by highways and rights-of-way and
generally does not require that facilities physically touch one
another. Please refer to the discussions in the General
Provisions rule (59 FR 12412, March 16, 1994) and the HON
(59 FR 19458, 19459, April 22, 1994). The EPA intended the term
"collocated" to be synonymous with "part of the same major
source," which presumes contiguity. Therefore, to avoid
confusion, EPA is deleting the term collocated from the final
regulations.
Regarding the comments on contiguity of offshore facilities
to shoreside major sources, EPA has determined that any offshore
tank vessel loading facility that is over one-half mile offshore
is not contiguous to the onshore facility. The Agency believes
that the considerable amount of distance between shore and land
2-50
-------
and the numerous issues that arise as a result of this
considerable distance make it clear that these facilities are, as
a class, not contiguous with shoreside facilities.
However, EPA is not stating that, as a class, offshore
terminals less than one-half mile from shore are contiguous to
shoreside major sources. This determination must be done on a
case-by-case basis, based on the factors pertinent to the
definition of contiguous. In particular, EPA believes that key
factors for this decision involve the extent to which the area
between the shoreside facility and the offshore facility is used
exclusively or generally by the shoreside facility, the extent to
which others use the area, and the actual distance between the
facilities.
Regarding the comments (130) from the company whose tank
vessel loading facility is 0.45 miles from the shoreline and
1.5 miles from the refinery, which is inland, EPA agrees that the
sources are not contiguous. The sources are over one mile apart,
they do not share a boundary, and the intervening land is neither
owned nor operated by the company.
2-51
-------
2.2.3 Emissions Averaging
Comment; Six cornmenters (23, 91, 92, 97, 99, 101) supported
and four commenters (94, 98, 100, 102) opposed the averaging of
emissions between marine tank vessel loading facilities and
petroleum refineries. Those opposed to such averaging stated
that EPA should not allow averaging between separate source
categories. One commenter (91) stated that in the long run,
emission averaging will increase the operator's flexibility and
increase cost effectiveness in complying with the emission
standards promulgated under Section 112. Emissions averaging
will encourage innovative control strategies which will likely
result in more efficient (cost effective) control strategies and
greater reductions in HAP emissions.
Three commenters (23, 92, 93) recommended expanding the
averaging program to all HAP emission points within the site,
regardless of source category. Two commenters (92, 101) stated
that marine operations located at facilities other than
refineries should be allowed to use emissions averaging as well.
While two commenters (92, 93) generally supported the
concept of emissions averaging, they had many objections to this
specific program. The strongest objections to the specifics of
the proposed emissions averaging program are summarized as
follows:
1. The equivalent risk demonstration has no place in a
technology-based regulation;
2. The 10 percent discount on emission credits unfairly
penalizes those who would utilize emissions averaging and will
act as a deterrent to use;
3. Allowing State and local discretion to preclude
averaging might effectively eliminate the utility of emissions
averaging; and
4. The limitation on the number of emission points to be
included in an averaging program has no sound basis, and its
elimination will not produce significantly burdens on
implementing agencies.
2-52
-------
One commenter (93) stated EPA should recognize that there
are marine loading operations at some facilities that will be
subject to the Gasoline Distribution MACT rule. For the same
reasons stated above, EPA should allow emissions averaging
between these sources as well.
One commenter (98) opposed the proposal to average
intermittent marine loading emissions with more constant
emissions throughout the facility, stating, "the Act does not
allow emissions averaging." In addition, EPA has not
demonstrated that a 10 percent discount factor adequately
compensates for the likely imprecision in estimation of
emissions. The EPA has also not demonstrated that the emission
reductions contemplated in the marine loading rule with a
10 percent discount factor is the maximum achievable emission
reduction when averaging is employed.
The commenter (98) continued the argument by stating that
the EPA has not provided for real-time monitoring of controlled
or uncontrolled marine loading emissions. Hence, the rule offers
no basis for accurately and reliably checking source estimates of
likely future credits and debits.
One commenter (100) thmight it was inappropriate to lessen
the requirements for HAP or VOC reduction since averaging two or
more emission points could lead to an emissions increase.
Two commenters (94, 102) expressed concern that emission
averaging will allow the VMT to average marine tank vessel
loading HAP emission reductions with reductions at on-site
storage tanks and process vents in such a way that total
emissions from the facility will be higher than if the marine
tank vessel loading rule were implemented as currently proposed.
The commenters stated any averaging provisions, including
on-shore equipment, would fo<= difficult to accurately monitor and
enforce.
One commenter (102) specifically opposed an emissions
averaging plan that would allow the Alyeska Pipeline Service
Company to emissions average and not be required to control
2-53
-------
emissions or be able to increase emissions that would be
otherwise controlled under the proposed standards. This
commenter was concerned that an emissions averaging approach
could lower the MACT floor, confuse the requirements for specific
source categories, result in partial control of commodities and
uncontrolled berths, and delay compliance with the standards.
Response; The Agency agrees with commenters who stated that
additional flexibility was justified and expanded the petroleum
refineries source category to include HAP emissions from marine
vessel loading operations contiguous with petroleum refineries.
However, this expansion includes those sources subject only to
the Title III standards (i.e., MACT standards). Because source
categories are irrelevant to the applicability of Section 183 (f),
sources subject to the RACT standards are not eligible for
emissions averaging. In order to minimize confusion between the
two standards, all related issues for MACT sources
(i.e., emissions averaging) were directed to the petroleum
refinery NESHAP. Please refer to docket A-93-48 and/or the
promulgation TSD (Volume II) for the petroleum refinery NESHAP
for specific comments and responses. Regarding the request for
emissions averaging at other collocated source categories, the
Agency does not believe that averaging with other source
categories is appropriate at this time [note that the averaging
of HAP emissions from marine tank vessel loading with emissions
from petroleum refineries is accomplished through a redefinition
of the respective source categories], nor have any commenters
produced information justifying any such change in source
categories or emissions averaging.
Regarding the commenters who stated that the emissions
averaging approach would allow the Alyeska Pipeline Service
Corporation's Valdez Marine Terminal to average emissions with
other emissions sources, the Agency notes that this terminal is
not collocated at a petroleum refinery and therefore would not be
addressed by this emissions averaging approach. The final
regulations for VMT require control of the primary loading berths
2-54
-------
and ultimately limits the uncontrolled throughput at other
berths. If throughput subsequently increases beyond the
projected levels, controls are required.
The Agency disagrees with the commenter regarding the
allowance in the Act for emissions averaging. The Agency
believes that the emissions averaging approach promulgated in the
petroleum refinery standards represents the maximum achievable
reduction in emissions in consideration of the costs of
controlling these emissions and is therefore consistent with
Section 112(d) of the Act. The Agency asserts that allowing
emissions from marine tank vessel loading operations, bulk
gasoline terminal or pipeline breakout station storage vessels,
and bulk gasoline terminal loading racks to be included in
emissions averages will result in equivalent or greater overall
HAP emission reduction at each refinery. The averaging
provisions are structured such that "debits" generated by not
controlling an emission point that otherwise would require
control must be balanced by achieving extra control at other
refinery emission points covered by the NESHAP. (Please refer to
the Petroleum Refineries Background Information Document
(EPA-453/R-95-015b), Section 10 for additional discussion of the
development of the emissions averaging approach.)
2-55
-------
2.2.4 Exposure Spikes
Comment; One commenter (100) stated that intermittent
process control may be allowed to offset a continuous process,
but continuous processes should never be allowed to offset an
intermittent process. The resulting exposure spike would likely
exceed threshold criteria levels for health and environment
equivalency. An intermittent process occurring once a month
would average its emissions to a daily rate for that month. The
commenter stated that if the intermittent daily average is offset
by the daily emissions of a continuous process, the actual excess
emissions for the 1-day intermittent process are 30 times the
offset. This emissions spike would dramatically increase acute
exposure and would occur repeatedly throughout the year.
Unfavorable weather conditions and the location of the loading
operation could severely increase the risk of public exposure.
One other commenter (92) stated that concerns about potential
exposure spikes were inappropriate in the context of the current
rulemaking. The commenter further stated there are many factors
that would need analysis before any conclusions could be made
regarding exposure spikes and adverse health effects.
Response; The Agency notes that the standards developed for
this source category are technology driven and have been
developed to reflect control technologies currently in use.
These control technologies account for exposure spikes inherent
with batch operations, such as are found in marine tank vessel
loading operations. The comments regarding the relationship of
potential exposure spikes with the emissions averaging approach
developed for marine loading operations collocated at petroleum
refineries are addressed in the justification for the emissions
averaging found in the petroleum refineries standard [citation
provided above]. The emissions averaging approach allows the
owners or operators to develop and implement their own emissions
averaging plan subject to approval by the Administrator.
Concerns with exposure spikes would be addressed under this plan.
Furthermore, the quarterly cap on the ratio of debits to credits
2-56
-------
established under the emissions averaging approach in the
petroleum refineries standards is intended to limit the
possibility of exposure peaks. Because loading operations occur
fairly frequently and emissions from an individual vessel filling
or loading event are relatively small, such emissions are not
expected to cause significant exposure peaks. Moreover, no
evidence has been presented that emissions averaging would permit
a very different mix of emissions to occur than would point-by-
point compliance. That is, peaks of exposures from batch
streams, storage, and loading operations should be equally likely
under point-by-point compliance as under emissions averaging, so
emissions averaging does not represent a less effective control
strategy. Furthermore, in order to receive approval for an
emissions average, the owner or operator is required to
demonstrate that the emissions average does not increase the risk
or hazard relative to compliance without averaging.
2-57
-------
2.2.5 Other Impacts
Comment: One commenter (91) stated that the addition of
certain marine vessel loading operations to the petroleum
refineries source category would delay promulgation of the
emission standards applicable to collocated marine tank vessel
loading operations, therefore providing additional time to
install emission controls. One commenter (94) believed it was
not appropriate to delay the compliance date for nonrefinery
marine tank vessel loading operations in order to be consistent
with compliance dates associated with the refinery rule unless
there is a substantial, technical reason for doing so. The
compliance date should follow rule promulgation by some
reasonable time-frame (in this case, 3 years) which allows the
source operators to design, purchase, and install control
equipment.
One commenter (100) stated, in the event the EPA allows
cross averaging between marine tank vessels and refineries, all
deadlines based on promulgation should remain distinct for the
two MACT's except the averaging deadlines. Averaging plans and
compliance deadlines should not start until promulgation of the
second MACT.
Response; The Agency asked for and received three
additional months to promulgate the final standards for marine
tank vessel loading operations. This decision was based on the
significant issues that were contained in comments submitted by
commenters on the proposed rule. Even with the allowance for
emissions averaging between the two sources, the standards for
marine tank vessel loading operations and petroleum refinery
operations each contain their own separate compliance schedules.
Comment: One commenter (100) stated that there would be
additional regulatory and enforcement complexities from the
integrated source approach. If the U.S. Coast Guard finds that
land-based emissions were to affect marine safety, how far into
the refinery could they exercise influence? Another issue was
raised involving setting a precedent for reopening MACT standards
2-58
-------
to re-establish the affected source categories. This precedent
could lead to overly burdensome administrative procedures that
would hamper the efficiency of the State's operating permits
program.
Response: The Agency notes that the current U.S. Coast
Guard regulations affecting these sources address safety issues
for marine loading. These safety regulations are independent of
emissions reduction requirements established by the Agency. Any
future decisions on the part of the Coast Guard to promulgate
safety regulations for land-based refinery operations would take
place regardless of EPA's redefinition of the petroleum refinery
source category to include marine tank vessel loading.
Regarding the comments on potential reduction in a state's
operating permit program, the Agency is not redefining any source
categories for which standards have already been promulgated.
Rather, as part of the regulatory development process for these
two standards, modifying the definition of the two source
categories would provide additional flexibility to sources
subject to the emissions standards while providing no increase in
emissions. The Agency, therefore, requested comment on the
redefinition of the source categories and has decided to proceed
with this definition in the promulgated standards for both
rulemakings
Comment: Two cornmenters (95, 101) stated that EPA's marine
vessels database does not identify which loading and unloading
operations occur at refineries as opposed to other ;:ypes of
plants. Therefore, the commenter stated that there is not
sufficient detai] in EPA's database to determine if the movement
of terminals associated with refineries out of the marine tank
vessel loading MACT source category would impact the
determination of the MACT floor for the marine tank vessel
loading MACT rule,
Response; The Agency agrees with the commenter that the
Agency is unable to determine with certainty which facilities in
the data base are refineries Therefore, the Agency could assume
2-59
-------
that a certain percentage of controlled and uncontrolled
facilities are refineries and then calculate a separate MACT
"floor" for the refinery terminals. However, there is no
assurance that this estimate would be any more correct than
assuming that the floors for marine terminals in both source
categories would be the same. Because of this uncertainty, and
because there is no inherent reason to believe that there is any
substantial difference in the MACT floors for the two types of
marine terminals, the Agency is assuming that the MACT floor for
marine terminals in the petroleum refinery source category is
equal to that of the marine tank vessel loading source category.
Therefore, marine tank vessel loading at refinery terminals will
be referred to Subpart Y for control requirements or emission
averaging limits. (See Docket No. A-90-44, item No. IV-B-2 for
additional information on the determination of MACT floors for
the final standards.)
2-60
-------
2.3 SUBCATEGORIZATION
2.3.1 Subcategories Based on Size
Comment; One commenter (29) stated the l Mg/yr HAP
emissions threshold is appropriate. Another commenter (44)
stated that emissions reductions achieved by control devices
should be included in the 1 Mg/yr applicability cut-off for
sources currently equipped with emission control devices. The
commenter asserted that controlled and uncontrolled sources
having HAP emissions of less than 1 Mg/yr should be exempt from
all other requirements of Subpart Y.
One commenter (04) noted that the preamble to the proposed
rule did not elaborate on why the upper limit of 1.0 Mg/yr was
selected as the final applicability determination. The commenter
noted that EPA's cost-effectiveness data indicated that the
1.0 Mg/yr limit would result in costs to industry of
approximately $99,000 per ton of HAP removed. This commenter
questioned why the Agency did not develop cost data for higher
limits. The commenter encouraged EPA to raise the applicability
limit.
One commenter (79) stated that the subcategorization of
facilities emitting less than 1 Mg/yr is not necessary because
these facilities are probably area sources. One commenter (80)
suggested subcategories should be drawn around sources with
similar operating cycles, not an arbitrary threshold. Some
sources under 1 ton/yr may operate sporadically, and some may be
area sources, not ma joe sources.
One commenter (45) stated that the applicability
determination (and therefore, the subcategory basis] the for this
regulation should be 10 tons per year of a single HAP or 25 tons
per year of a combination ot HAP irrespective of where the marine
facilities are located
Response: In the preamble of the proposed rule, the Agency
requested comment on the establishment of these subcategories
based on size (i.e., HAP emissions). In the public comments, the
Agency found general, though not universal, agreement on the
2-61
-------
establishment of subcategories based on size for this source
category. However, some of the comments encouraged the Agency to
raise the HAP emissions level of the controlled subcategory. The
final standards continue to group major source terminals into
subcategories based on HAP emissions; however these subcategories
were changed to terminals with HAP emissions of 10 tons/yr or
more of any single HAP or 25 tons/yr or more of total HAP, and
terminals having HAP emissions of less than 10 tons/yr of all
individual HAP and less than 25 tons/yr of combined HAP. The
Agency based this decision on information found in the comments
received and the high incremental cost effectiveness associated
with requiring controls on these terminals having the lowest HAP
emissions.
This decision raises the limit to the definition of major
source found in the Act (i.e., 10 tons/yr or more of any single
HAP or 25 tons/yr or more of combined HAP), and was suggested by
several commenters. Other commenters favored leaving the cutoff
as proposed or lowering it. Several commenters provided
rationales based on economic concerns and prior regulation.
However, none of the commenters provided any data to support
their recommendations. Commenters noted that prior State
regulations generally (though not invariably) had distinguished
between large tank vessel loading facilities that are responsible
for the vast majority of emissions and small tank, vessel loading
facilities that are substantially less cost-effective to
regulate. Commenters also noted that the smaller marine
terminals that would be regulated under the proposed delineation
at 1 Mg would face considerable risk and may have to close. The
commenters stated that such disruption is not justified given the
small emissions benefit of regulating smaller terminals. [As
discussed elsewhere in the docket, the incremental cost
effectiveness shown when moving from the 25 ton per year limit to
a 1 ton per year HAP emissions limit is approximately $180,000
per Mg (i.e., approximately $100 million in annual costs
resulting in control of approximately 550 Mg of HAP). This ratio
2-62
-------
is contrasted with the $27,000 to $53,000 per Mg cost
effectiveness associated with the control of sources emitting
10 tons/yr or more of any single HAP or 25 tons/yr or more of
total HAP (i.e, approximately $20 million to $40 million in
annual costs resulting in control of approximately 750 Mg of
HAP).]
Though Section 112 does not provide any language indicating
the criteria for subcategorization, Section 112(d)(l) of the Act
states that EPA may distinguish among classes, types, and sizes
of sources in establishing standards. EPA believes that dividing
this source category into two subcategories based on size is
appropriate, especially given issues of frequency of use, cost-
effectiveness, and amount of emissions connected with the
subcategory of smaller facilities. (See Section 2.3.1 of the
promulgation TSD for additional discussion of the subcategories
based on size.)
The Agency agrees with the commenters who support changing
the subcategories so that marine tank loading operations that
emit more than 1 Mg/yr of HAP but that are not themselves major
emitters of HAP are appropriately grouped with smaller
facilities, rather than with the large facilities with which they
had previously been grouped. However, EPA is unwilling to raise
the subcategory limit above the 10 ton limit for determining
major source status for single HAP emissions or above the 25 ton
limit for determining major source status for multiple HAP
emissions. The Agency believes that Congress's determination of
10 tons of emissions of any single HAP and 25 tons of emissions
of total HAP as the outer limits of the major source
determination is an indication that it; is appropriate to
distinguish sources that are above this level from those that are
below this level and to ensure regulation of sources that emit
beyond that level (regulation of sources having emissions below
that level is based on a separate MACT determination).
Therefore, the Agency selected 10 tons/yr as the HAP emissions
2-63
-------
limit for single HAP emissions and 25 tons/yr as the HAP emission
limit for total HAP emissions.
Comment; One commenter (45) indicated that no additional
source categories should be added to the applicability
determination. As proposed, marine tank vessel loading
operations, which are located at major source refineries and
chemical plants, are only subject to the proposed standards if
they emit more than 1 ton/yr of HAP; however, independent marine
tank vessel facilities are only subject to the rule if they emit
10 tons/yr of one HAP or 25 tons/yr of a combination of HAP. The
commenter stated that the applicability determination section of
this regulation should be 10 tons/yr of a single HAP or
25 tons/yr of a combination of HAP, irrespective of where marine
facilities are located.
Response: Section 112 of the Act requires MACT for major
sources of HAP, which are generally a collection of several
emissions points, some of which individually may be below the
10/25 tons/yr limits contained in the Act but which collectively
exceed these limits. The language of the Act and the legislative
history clearly shows that any group of sources located within a
contiguous area and under common control that collectively emit
more than the 10/25 tons/yr limit shall be treated as a major
source, and each emissions point within the major source shall be
subject to major source MACT requirements, even if the emissions
point does not itself exceed the 10/25 tons/yr limit. Area
sources that only perform loading, while possibly having the same
loading emissions as a terminal collocated at a major source,
would not necessarily have the same health impacts as the
collective emissions from all points at the major source. Under
the urban air toxics program (Section 112(k) of the Act), the
emissions from these sources will be analyzed to determine if
their collective emissions pose a risk to the public that would
necessitate future emissions standards.
2-64
-------
2.3.2 Offshore Terminals/Offshore Production
Comment; Six commenters (28, 30, 34, 73, 127, 133)
supported the creation of a separate subcategory for offshore
facilities. One commenter (28) added that a significant reason
not to require control at offshore facilities would be the lack
of any population that would be impacted by HAP emissions from
these operations. This commenter applied the same standard to
facilities located in remote areas. Another commenter (30)
described a host of factors relative to differing economic
considerations associated with offshore terminals and provided a
definition of "offshore terminal".
One commenter (36) recommended that terminal facilities that
operate docks or platforms extending one-half mile or more into
the water be classified as "on-shore terminals," consistent with
Federal "law, but that such terminals be exempted from the vapor
control requirements proposed in this rulemaking.
Another commenter (130) stated that the grouping of offshore
terminals into a subcategory is appropriate but stated that the
Agency's use of a one-half mile limit is arbitrary. This
commenter's terminal is less than one-half mile from shore, but
is over two miles from a refinery. The commenter requested that
the final regulations address such terminals so that they would
not be required to contro] emissions. Another commenter (137)
also stated that the one-half mile limit is arbitrary and
maintained that the main consideration should be if a terminal is
contiguous to a refinery.
One commenter (41) supported the proposed requirement that
offshore terminals meet the same requirements as onshore
terminals. This commenter continued by stating "unless it can be
shown that emissions from such terminals have no risk to the most
exposed person," Another commenter (87) did not support grouping
offshore facilities into source categories unless the MACT
requirements are at least as stringent as those for onshore
facilities.
2-6b
-------
One commenter (73) stated HAP emissions from offshore
terminals have a much lower potential for human exposure due to
the distance from the emissions source to any receptors. Only if
the terminal by itself is determined to be a major source should
the terminal be subject to any applicable MACT standards.
Two commenters (28, 73) stated that noncontiguous marine
loading/unloading facilities should not be regulated under the
proposed MACT rules unless the marine terminal by itself is
determined to be a major source.
One commenter (80) disagreed with the separate facility
designation and cited two facilities with loading operations that
occur more than 0.5 miles offshore; the commenter considers these
sources "contiguous."
Several commenters (34, 72, 73) supported EPA's separate
facility designation for offshore terminals that are one-half
mile or greater from shore. One commenter (30) suggested EPA
create subcategories within the offshore terminal source category
to reflect the different sizes of terminals, the quantities of
products handled, and the fundamental differences between crude
oil production platforms and product terminal platforms. This
commenter and one other (72) suggested language for the
definition of the term "offshore terminal."
One commenter (127) stated that offshore terminals face
significant control costs that are several times the national
ceiling and many times higher than the national average. The
commenter also stated that control requirements would result in
significant technological difficulties that are not yet fully
understood. The commenter stated that the State regulations
governing the Riverhead, Long Island terminal do not require
control of this offshore terminal and that Federal regulation
should take a similar approach.
One commenter (136) stated that none of the offshore
terminals of which the commenter is aware presently control
emissions. Therefore, the commenter stated that the MACT floor
for a subcategory of offshore terminals would be uncontrolled.
2-66
-------
One commenter (140) favored the establishment of a
subcategory for offshore terminals provided that the definition
of "offshore terminal" be structured to include docks or piers
having loading arms that are at least one-half mile from shore
and may not necessarily incorporate subsea lines. The commenter
stated that these types of terminals would face similar costs and
technical difficulties in installing control equipment as the
offshore terminals with subsea lines.
Response: The Agency is in agreement with many of the
comments received. The Agency maintains its position as stated
in the proposed rule that a marine tank vessel loading operation
that is at least one-half mile offshore is not part of a land-
based contiguous site. The Agency agrees with commenters that
these offshore terminals should be considered separate (stand
alone) sources because many are supplied solely by subsea lines
and others, by definition, have at least one berth that is one-
half mile or more beyond the shore line. Offshore loading
operations with subsea lines in particular require the permitting
of either additional, subsea lines to carry vapors or permitting
of docks or platforms. If permits are unavailable for these
offshore terminals, compliance with the standards would be
impossible. These factors result in significantly higher costs
compared to onshore terminals, Additionally, these terminals
pose less of a. health risk to any surrounding population. The
Agency also agrees with several of the commenters that these
sources warrant the establishment of a separate subcategory
because of the differences in the feasibility and cost of
controlling emissions at these offshore terminals.
Once the Agency determined that a subcategory existed for
these offshore terminals a MACT floor determination was made and
MACT selected. The data made available to the Agency indicates
that there are fewer than 20 offshore terminals having subsea
lines. None of these terminals presently control emissions from
marine tank vessel loading. The Agency was also made aware of
additional offshore terminals that do not have subsea lines. Two
2-67
-------
of these terminals for which the Agency has information presently
control emissions. Based on the information available to the
Agency, the MACT floor for existing sources in this subcategory
is no control of HAP emissions. The MACT floor for new offshore
sources is 95 percent reduction of emissions. (See Docket A-90-
44, Item Number IV-B-2 for additional discussion of the MACT
floor determination for this subcategory). Data submitted by
commenters show that the costs associated with the control of
offshore terminals are between two and five times more expensive
than comparable onshore control techniques (see Docket A-90-44,
Item numbers IV-D-108 and IV-D-136). Because of the poor cost
effectiveness resulting from these significantly higher costs, as
well as the environmental and technical challenges (discussed
above) associated with requiring control more efficient than the
MACT floor, the Agency has selected the MACT floor level of no
control for offshore marine tank vessel loading operations.
The Agency also determined that offshore terminals loading
10 million barrels per year or more of gasoline or 200 million
barrels or more of crude oil should not be required to control
emissions of VOC or HAP. The comments that noted the
significantly higher costs and poor cost effectiveness of these
sources (see previous paragraph) would make control requirements
unreasonable for these offshore terminals (See Docket
No. A-90-44, item No. IV-B-2 for additional information on the
determination of MACT floors for the final standards.).
Regarding the comment from Commenter 36, EPA is using the term
"offshore terminals" solely for the sake of simplicity and
practicality in this rulemaking. The use of the term in this
rulemaking should not have any effect on other programs. This
approach is more straightforward and practical than the approach
suggested by Commenter 36.
Comment; One commenter (30) suggested EPA create
subcategories within the offshore terminal source category to
reflect the different sizes of terminals, the quantities of
products handled, and the fundamental differences between crude
2-68
-------
oil production platforms and product terminal platforms. The
commenter was concerned that EPA, in setting standards, should
not penalize terminals that handle gasolines containing
oxygenates (such as MTBE) to the oxygenated and reformulated
gasoline requirements of the CAA.
Response: The Agency agrees that offshore terminals could
be further subcategorized based on the types of commodities
loaded, the size of the terminal, or by the type of operation
with which the terminal is associated. However, the comments
received did not contain sufficient information to justify
further subcategorization of offshore terminals.
2-69
-------
2.3.3 Parcel Tankers
Comment; One commenter (31) stated the loading and
unloading of parcel tankers presents unique challenges and costs
that have not been addressed by EPA. There are economic and
technical difficulties associated with the potential mixing of
contained vapor streams from the simultaneous loading/unloading
of multiple products (up to 15 types). Additionally, most parcel
tankers are not owned by the same companies that own the
terminals, therefore the burden of compliance with a rule is
shifted from the terminal to the ship owners. Lastly, the EPA's
own technical support document indicated that emissions from
chemical product parcel tankers are small compared to those from
petroleum tankers and that the costs to retrofit product tankers
are the highest of all types of marine tank vessels. Until
further studies have been done, the rule should not cover the
loading and unloading of parcel tankers.
Response: The Agency is aware that parcel tankers as well
as river barges may carry more than one product. The commenter
did not provide sufficient data demonstrating the infeasibility
of a control strategy system to handle multiple commodity
streams. The Agency believes that a combustion device is capable
of destroying multiple chemicals to a level sufficient to meet
the final standards. Additionally, the commenter did not provide
information to suggest inherent safety problems with the control
of emissions from terminals loading these vessels. Therefore,
the Agency has no technical basis to distinguish parcel tankers.
However, if after promulgation, the U.S. Coast Guard determines
that safety problems do exist with specific loading scenarios,
then the Agency would reevaluate the standards applicable to
terminals loading those types of vessels.
2-70
-------
2.3.4 VMT
Comment: Three commenters (06, 10, 80) were opposed to
creating a separate subcategory for the VMT because Alyeska has
not presented any information that warrants a separate
subcategory. However, in a later submittal (118), Commenter 06
agreed with the subcategorization of VMT. The mere fact that it
may be more difficult and costly to control large volumes of
crude oil vapors in a sub-Arctic climate does not warrant the
creation of a new subcategory. The commenters supported
reasonable, cost-effective controls for the VMT that are
consistent with the cost effectiveness levels used for other MACT
and RACT controlled sources. The commenters disagreed with all
of the arguments put forth by EPA for considering the VMT as a
separate subcategory. One commenter (06) further elaborated
that, under Section 112, any control beyond the MACT floor must
consider the cost of achieving reductions, as well as any nonair
quality health and environmental impacts. If the VMT is a
separate source category, then the MACT floor for the source
category will be zero, and any control beyond this floor would
require that EPA consider costs and nonair impacts. None of the
issues VMT has raised (high vapor flow, complexity of vapors,
declining throughput, low loading temperatures) differentiates
VMT from other terminals with respect to health, environmental,
or energy impacts of the proposed control measures. Therefore,
there appears to be no reason to justify MACT controls at the VMT
that are less stringent- than other terminals.
One commenter (10'; staled if EPA insists on a separate
subcategorization for che VMT, regulatory alternatives should be
negotiated with the State or local permitting authority,
particularly if that agency has more stringent requirements.
Several, commenters (07, 28, 32, 34, 49, 50, 53, 55, 58, 68,
80, 118, 129, 131) stated the VMT should be regulated under a
separate source category due to its numerous unique
characteristics, including size, technical and economic
feasibility of and availability of controls, type of product
2-71
-------
loaded, safety considerations associated with the control
systems, and projected declining throughput.
Response; The Agency reviewed all of the comments received
related to the potential subcategorization of Alyeska Pipeline
Service Company's Valdez Marine Terminal (VMT). The Agency
agrees with some of the commenters that this particular facility
warrants the establishment of a subcategory for the following
reasons: (1) special circumstances require keeping the oil
moving (i.e., the limitations on storage capacity and the need to
keep the oil moving through the pipeline); (2) periods of extreme
bad weather complicate the loading schedule; (3) the amount of
throughput loaded at VMT is much higher than other facilities in
EPA's data base, dwarfing by several times the throughput loaded
at other sources and creating special circumstances regarding the
control devices that may be used at VMT; and (4) information
submitted by several commenters shows that the volume of oil
loaded at the VMT has decreased and will continue to decrease in
the future; therefore, infrequent and declining use of VMT's
third loading berth is anticipated. Because of these
differences, the Agency has established a subcategory for
Alyeska's Valdez Marine Terminal for the purposes of both the
RACT and MACT standards.
2-72
-------
2 .4 RACT/MACT
2.4.1 MACT Floor
Comment: One commenter (32) stated that the preamble and
supporting record show that MACT was calculated as an overall
emission limitation rather than a limit on individual emission
points. However, the commenter noted that the proposed standard
applies to berths as individual emission points and not as an
overall limitation.
Response: The Agency disagrees with the commenter's
statement that the standards apply to berths as individual
emission points. The standards apply to sources loading marine
tank vessels. The emission limitation for a given commodity may
be determined over several emission points.
Comment; One commenter (32) stated that the proposed
regulations should be changed to permit terminals to demonstrate
compliance with the requisite emission reduction standard by a
combination of control technology and operating limitations.
Response; The proposed and final RACT standards contain
throughput limits that are used in determining the applicability
of these standards and would, therefore, allow sources to accept
an operational limit on loading to avoid the additional control
requirements associated with the RACT requirements. Operational
limits could also be employed by a source to avoid having to
control emissions under the MACT standards. The proposed and
final MACT standards divide the marine tank vessel loading
operations source category into two subcategories iuased on HAP
emissions. The final rule used 10 tons/yr of any single HAP or
25 tons/yr of all HAP as the dividing line based on actual HAP
emissions. A source demonstrating emissions below these limits
and incorporating such limits in their Title V permit would avoid
having to install additional control equipment. Additionally, a
source could incorporate federally enforceable operational
limitations on its HAP emissions and become a nonmajor source
under its Title V permit.
2-73
-------
Comment: One corranenter (10) stated that the MACT floor in
the proposed regulation was based on a weighted average of
controlled and uncontrolled emission points at 43 facilities that
comprise the best performing 12 percent of the terminals emitting
greater than 1 ton/yr. The commenter stated that this average
should not include the uncontrolled emission points at these
facilities, since these points are not among the best performing
12 percent of sources.
Response: The Agency based the calculation of the MACT
floors for the final standards on the available data for marine
tank vessel loading sources subject to the standards. The Agency
determined the overall emission reduction for the sources based
on total throughput and the emissions reductions required by
State regulations in order to avoid over-estimating the required
control of these emissions. The Agency then ranked these sources
in order of control. The MACT floors, therefore, represent the
average emission limit achieved by the best performing 12 percent
of similar sources. In cases where this average control did not
approximate a control technology, the source at the
94th percentile was used to determine the MACT floor. (See
Docket No. A-90-44, item No. IV-B-2 for additional discussion of
the calculation of MACT floors for the final standards).
Comment; Three commenters (23, 30, 32) stated that the
interpretation yielding an 88th percentile performance level is
the interpretation intended by Congress for determining the MACT
floor. One commenter (23) believes the use of an average
emission reduction is incorrect because calculating the average
of a set of numbers that are already representative of the
average performance of each control technology being considered
results in a technology that is no longer meaningful. The
commenter stated that control technology considered at the
88th percentile would represent the emission control that is
achieved in practice for the entire group.
Two commenters (10, 78) stated that if the mean level of
control does not correspond to an actual technology, the MACT
2-74
-------
limit should be selected from control technologies capable of
performing better than the mean. The commenters generally
objected to the finding of an uncontrolled MACT floor for the
subcategory of sources having loading emissions of less than
1 ton per year.
One commenter (80) stated that the best reading of the Act
requires the determination of the MACT floor based on the
arithmetic mean.
Response; The Agency considered the approach suggested by
the commenter regarding the use of the 88th percentile in
calculating the MACT floor level of control. In a final rule
published in the Federal Register on June 6, 1994 (59 FR 29196),
the Agency discussed its conclusions regarding the best reading
of the applicable statutory language. The Agency determined at
that time that the use of either the arithmetic mean or the
median of the best performing 12 percent of sources was the
method intended by Congress for determining the MACT floor.
Although the Agency also stated in this notice that it retained
discretion within the statutory framework to set MACT floors at
appropriate levels for other source categories, the Agency does
not believe that comments submitted in response to the proposed
rule for marine tank vessels loading and unloading operations
have contained any new information or data sufficient to cause
the Agency to alter its position as presented in the June 6, 1994
Federal Register notice. The use of the arithmetic mean and
median of the best performing 12 percent of sources has,
therefore, been used in determining the MACT floors in the final
rule.
Regarding the use of the median of the best performing
12 percent of sources in calculating the MACT floor for sources
having annual HAP emissions of less than 1 ton found in the
proposed rule, the Agency stated in the preamble to the proposed
rule that this approach best suits this subcategory of sources.
During the development of the proposed rule, the Agency
considered regulatory alternatives involving control beyond the
2 75
-------
MACT floor level of control. These alternatives are discussed in
the preamble to the proposed rule and in the docket (see docket
reference II-A-34). The Agency did not receive data in the
public comments that would cause it to alter this position on the
calculation of the MACT floor for this subcategory, even though
the level for this subcategory increased from 1 ton per year in
the proposed rule to 10 tons/yr of any single HAP or 25 tons/yr
of total HAP in the final rule. The calculation used to develop
the MACT floor for this subcategory and the decision not to
require control beyond the floor for this subcategory are found
in Docket No. A-90-44, Item No. IV-B-2.
Comment; Four commenters (28, 45, 46, 50) stated that the
data upon which the MACT floor is based were not included in the
proposal or its technical support document. In some states cited
as already requiring marine vapor recovery, VOC control systems
at many facilities remain in the design stages and, thus, have
never actually operated. It is not appropriate to include such
facilities that do not have a proven performance history in the
population of sources for which the MACT floor is determined.
Three commenters (04, 34 71) stated that the floor determination.
was based on unfounded assumptions regarding the number of
facilities that would have to install controls and the number
that currently have controls in place. Two commenters (04, 71)
noted that EPA did not develop the numerical basis to indicate
that 12 percent of the impacted facilities are actually located
in the referenced states. These commenters also noted that most
State requirements have exemptions including, but not limited to,
the loading of low vapor pressure materials, which would
dramatically decrease the reduction efficiencies noted by EPA.
Response; The information on control technologies that was
available to the Agency during the development of these standards
is included in the docket supporting this rulemaking. A
memorandum detailing the Agency's calculation of the MACT floors
for this source category was included in the docket prior to
proposal (docket reference II-A-34 in Air Docket A-90-44). The
2-76
-------
data used in calculating the MACT floor was among the information
included in the docket. Additionally, the Agency's entire data
base (consisting of the data found in the docket for the proposed
rule as well as a great deal of data that was not related to the
calculation of the MACT floors or impacts of the standards) is in
the docket (see Docket Number A-90-44, Item Number IV-A-1) and
was made available to interested parties upon request. As stated
in the preamble to the proposed rule, the Agency based its
calculation of the MACT floor, in part, using State regulations
that were effective prior to the development of the proposed
rule. Other data on the extent of controls for this source
category were not available.
Information submitted by commenters that specifies (by
State) the numbers of controlled facilities was incorporated in
the data base and was used to calculate the MACT floors for the
final rule. These data (see Docket No. A-90-44, item No. IV-D-34
and IV-D-135) indicate that emissions from at least
114 facilities in several States are controlled. These data were
combined with information on controls required by State
regulations and were used in determining the MACT floors for the
final standards (see Dockei .Mo. A-90-44, item No. IV-B-2) .
Comment; Two coramenters (34, 71) stated that the
subcategorization scheme the Agency has developed has led to the
erroneous conclusion that e floor exists for the source category
as a whole. Another commencer (80) stated that the
subcategorization of existing sources subject to the MACT
requirements in this proposal is arbitrary and not defensible.
Response; As was stated in the proposed rule, the Agency
based its decision to establish a subcategory based on size
because sources emitting less than l ton of HAP emissions per
year were likely to be area sources or facilities that would be
represented by relatively minimal, episodic emissions.
Therefore, there was no need to investigate the existence of a
MACT floor for the entire source category. However, if area
sources could have been excluded from the calculations, it is not
2-77
-------
improbable that the entire source category may have been subject
to a controlled floor. In the final rule, additional
subcategories have been added for offshore terminals and for the
Valdez Marine Terminal. Additionally, the petroleum refineries
source category has been expanded to include marine tank vessel
loading operations at petroleum refineries. In addition, based
on data submitted by commenters, the subcategories based on size
now contain sources whose HAP emissions are 10 tons/yr or more of
any one HAP or 25 tons/yr or more of total HAP and less than
10 tons/yr for all single HAP and less than 25 tons/yr for total
HAP. The MACT floors for all of the subcategories have,
therefore, been recalculated based on these changes.
The Agency disagrees that the use of a measure of annual HAP
emissions to delineate subcategories is arbitrary and not
defensible. In establishing the subcategories, the Agency
attempted to provide flexibility to the owners and operators of
marine loading facilities located at major sources that load
marine tank vessels sporadically or in small quantities. The use
of annual HAP emissions as a deciding factor was based on the
fact that emissions are a reasonable surrogate for size and, in
fact, are likely the most appropriate measure to distinguish
large HAP emitters from small HAP emitters, which is relevant to
the purpose of Section 112. In addition, the Agency believes
that this size distinction will provide relief from requirements
to install expensive controls that would provide relatively
insignificant environmental benefit.
Comment; Two commenters (34, 71) stated that the floor
analysis fails to consider that the floor level of control does
not apply to products with low vapor pressures.
Response; In response to this and other comments on low
vapor pressure commodities, the Agency incorporated a low vapor
pressure limit of 1.5 psia in the final rule. This vapor
pressure limit was used when calculating the MACT floors for the
final standards.
2-78
-------
Comment; One commenter (34) believed EPA should address any
requirements for existing sources under Section 112(d) based on
cost effectiveness; EPA should reassess the proposed rule and
find that there is no MACT floor. As such, any requirements for
existing sources should be based on an assessment of cost-
effectiveness because such controls would exceed the MACT floor
for existing sources.
Response; The proposed standards were based on MACT floor
level of control for all sources in this source category except
for the VMT source (see discussion of the VMT standards in
section 2.1.6 of this document). The Agency has reviewed data
submitted by commenters and has included these data in the MACT
floor determination for the final rule. Cost effectiveness was
evaluated only as a criterion for selecting a regulatory
alternative with a level of control more stringent than the MACT
floor. For all subcategories except VMT, poor cost-effectiveness
values were used as the bases for not requiring more stringent
control beyond the MACT floor levels of control. The cost
effectiveness of the final standards also was included as a means
of determining the impacts of the final rule.
2-79
-------
2.4.2 Stringency of MACT/RACT
Comment; Several commenters (38, 79, 80) submitted comments
regarding the apparent higher stringency of RACT versus MACT.
One commenter (38) noted that the RACT and MACT requirements may,
in practice, be identical given that the associated equivalence
of 95 percent efficient recovery devices and 98 percent efficient
combustion devices is accepted. The commenter opposed lowering
the RACT requirements to 93 percent control because it could lead
to a lessening of control requirements. Several commenters (10,
38, 79, 80) opposed lowering the RACT control limit because MACT
requirements would be less stringent than RACT requirements. One
commenter (38) stated that MACT for one HAP should equal MACT for
all. Therefore, HAP loaded at an affected facility should be
subject to the same levels of control whether or not they are
considered generic volatile organic compounds (VOC). Two
commenters (79, 80) believed that existing sources subject to
both RACT and MACT should be required to meet the 95 percent
reduction in the RACT standard. One commenter (10) believes that
the MACT standard should: (1) reflect the maximum achievable
control efficiency of existing technologies, (2) not allow the
exemption of certain commodities from regulation, and (3) be no
less stringent than RACT. Three commenters (32, 34, 68)
suggested that EPA reduce the RACT standard to 93 percent control
to match the stringency of the MACT standard. One commenter (32)
stated that the intent of the Clean Air Act was for RACT
standards to be set at a more lenient level than MACT standards.
This interpretation would also harmonize the two sections because
smaller facilities that would not be covered by the MACT standard
would be covered by the RACT standard. Under the current
version, it is possible that some sources might not be covered by
the less stringent MACT standard, yet could be covered by the
more stringent RACT standard. The commenter stated that Congress
did not intend such a result. One commenter (06) stated that a
lower VOC standard should not be allowed as an excuse to operate
recovery equipment improperly. The commenter stated that Alyeska
2-80
-------
must demonstrate that a recovery system operating at a lower VOC
control level would not result in a HAP control efficiency of
less than 95 percent across the control system.
Response: The standards developed for RACT apply to the
largest marine tank vessel loading sources (sources having an
annual throughput of greater than 10 million barrels of gasoline
or 200 million barrels of crude oil--approximately 18 sources)
and have been developed to comply with Section 183(f) of the
Clean Air Act. The MACT standards apply to both the facilities
affected by Section 183(f) and to smaller sources. Therefore,
the MACT standards are more stringent in a broader sense because
a greater number of facilities would be affected. In addition,
due to changes in the MACT floor determination caused by
additional data received from commenters, the MACT standard
appears to have been made more stringent on an individual
facility basis by requiring an overall emissions reduction of at
least 97 percent (Docket No. A-90-44, Item No. IV-B-2). However,
the final standard for this subcategory also reflects a minimum
vapor pressure limit of 1.5 psia that does not require control of
low vapor pressure commodities. This change in format makes it
appear as if the final standards are more stringent. The Agency
anticipates that the same control technologies will be used to
comply with the final standard as were expected to be used to
comply with the proposed standards. The Agency also notes that
the promulgated MACT standard is nationally less stringent than
the proposed standard because it affects fewer facilities and
does not regulate loading of products with vapor pressures below
1.5 psia.
Comment; One oommeuter r?9) suggested lowering the
threshold for RACT to make if consistent with New Jersey's
adopted RACT rule, which is that facilities exceeding a threshold
of 6 million gallons or greater annual throughput of 60,000 gal/d
during the ozone season are required to attain a minimum VOC
emission reduction of 95 percent by weight. Another commenter
(24) stated that State requirements, which were driven by the
2-81
-------
need to reduce VOC emissions, may require less than the RACT
requirements specified in this rule. For example, Pennsylvania
requires a 90 percent reduction in VOC emissions from marine tank
vessels. A source that has installed controls to meet this
requirement may be unable to meet the 93 percent reduction for
MACT, much less the 95 to 98 percent reduction for RACT. The
commenter stated that it is unfair for EPA to enter the
regulatory arena after sources have spent money to comply with a
State rule to reduce VOC emissions and require those sources to
control further. One commenter (46) stated that EPA should
establish a control efficiency standard no greater than that
contained in many existing State rules--90 percent. One
commenter (38) recommended that EPA consider a limitation of 2 Ib
per 1,000 barrels loaded and stated that fugitive emissions
should be considered in determining uncontrolled and controlled
emissions. One commenter (32) stated that EPA should consider
not setting the same RACT standards on every marine tank vessel
loading operation in the United States and, instead, allow for a
case-by-case determination of whether a particular control
technology is "technologically and economically feasible".
One commenter (80) indicated that the throughput thresholds
for the RACT standard are too high. The 5 million barrels
correspond to 100 tons/yr if it is assumed that the cargo tanks
were all gas-free upon loading. The commenter cited an emission
factor of 3.4 Ibs VOC per 1,000 gallons loaded for loading of
gasoline into a cargo tank that has not been purged of the
previous cargo. The commenter supported the applicability
thresholds as outlined in alternatives III or IV of Table 1 in
the preamble to the rule. One commenter (43) recommended a RACT
limit of 154 million liters (l million barrels) of gasoline per
year and 3 billion liters (20 million barrels) of crude oil. The
higher limit proposed by EPA (5 million barrels/yr gasoline)
would exclude all sources in Oregon and may require the issuance
of individual RACT determinations. The commenter explained that
several other states also have RACT criteria that require
2-82
-------
controls at lower throughput levels. This commenter pointed out
that excluding terminals below 1 tons/yr HAP emissions may
conflict with existing State regulations that apply to small HAP
sources. One commenter (36) suggested that the MACT and RACT
requirement apply only to facilities that load more than
5 million barrels per year of gasoline or more than 100 million
barrels per year of crude oil.
One commenter stated that the threshold for Title I
applicability for gasoline terminals should be raised to 15
million barrels or more per year. The commenter supplied cost
data for several terminals in California that showed higher costs
than reported in the proposed rule.
Response: There are several State regulations that require
control of loading emissions from marine tank vessel loading
operations. The EPA agrees that there may be instances where the
Federal RACT emissions standards could vary from an existing
State or local standard. These Federal standards are intended to
be representative of reasonably available control on a national
basis, and the Agency recognizes that in some areas the standards
may not address air pollution control needs of individual states
or localities, While States are prohibited from adopting
standards that are less stringent than a Federal requirement,
they may go beyond the Federal requirement and adopt standards
that are more stringent, Certain States use Federal rules as a
baseline for their ov/u regulations, and it is at the State's
discretion to go beyond the Federal requirement.
The fact that a State currently regulates a particular
source that is also the subject of Federal regulation does not
predispose those facilities to exemption from the Federal
regulation and its requirements. If a source is currently
controlling emissions from, marine tank vessel loading operations
to comply with a State or local rule, they may have sufficient
control in place to meet Federal standards as well. In addition,
the reporting required for a State rule may also be submitted to
fulfill the reporting requirements for these Federal rules given
2-83
-------
that all of the appropriate information is contained in the State
report. States will likely be delegated the authority for
implementing rules for Part 63.
Regarding comments asserting that the RACT level of control
found in the proposed rule is more stringent than the level of
control found in some State regulations, the Agency notes that
the economic analysis performed during its development of the
proposed standards has indicated that the proposed RACT emissions
reductions are reasonable for the facilities affected by the RACT
standards. The Agency also notes that in some of the States
where the emissions reduction is lower, the loading limits are
either lower than those in the proposed rule or are nonexistent
(i.e., more facilities in these States would be affected by the
State regulations than under the Federal rule).
Regarding the comments that the emissions reduction required
by RACT should be lowered, the Agency believes that the emission
reduction requirements for RACT are justified. No commenter has
provided sufficient data to indicate that the technology levels
used to determine RACT are incorrect or unreasonable.
Comment; One commenter (79) cited current State permit
requirements to control emissions by 95 to 99 percent (for
gasoline), which is more stringent than the proposed MACT for
existing sources, and stated that a 95 percent control level for
loading operations is reasonably achievable for existing sources;
another commenter (80) concurred and stated that it is equally
cost-effective and as technically feasible to achieve 95 percent
control rather than the proposed 93 percent. Another commenter
(02) stated that control equipment suitable for controlling HAP
emissions from marine tank vessel loading operations is widely
available and that this equipment is capable of meeting or
exceeding the proposed standard.
Response; The proposed MACT standard referred to by the
commenter was developed from existing State regulations and is
based on a mass emissions reduction calculation that produced a
weighted average emission reduction of 93 percent. The
2-84
-------
comparable MACT standards contained in the final rule are based
on a 97 percent reduction in emissions for HAP having a vapor
pressure of 1.5 psia or more. The Agency believes, based on the
information provided in the docket, that the final standards
reflect the maximum emission reduction currently achieved by
similar sources (i.e., MACT floor). The Agency disagrees with
the commenters regarding the higher emission reduction efficiency
values that were cited and notes that these cited efficiency
estimates do not indicate what compounds are being controlled
(i.e., the estimates do not indicate the vapor pressure limit
incorporated in the requirements). Because of this lack of data,
the Agency was not compelled to raise the control level for these
sources beyond the 97 percent level contained in the final rule.
Comment: One commenter (78) supported the inclusion of the
93 percent overall reduction as one method of considering the
relative volatility of materials loaded and suggested that the
concept be retained in the final rule. However, the commenter
noted that this provision would be problematic for facilities
loading a wide variety of materials because it requires a complex
calculation of both past and future projected loading to estimate
whether market demand for materials will exist such that high
volatility materials loaded will achieve sufficient reductions to
account for demand of low volatility materials that need to be
loaded. This also occuirj because 63.563 (b) (4) indicates that
partial control is not allowed as a means of achieving 93 percent
control. The commenter noted that no averaging time limit is
given for the 93 percent compliance requirement and suggested the
use of a yearly averaging time. One commenter (10) stated that
there is no distinction between Regulatory Alternative A
(93 percent) or B (95 percent) in terms of the number of
terminals, capital costs, annual costs and cost effectiveness.
The commenter recommended that EPA require 95 percent control and
not allow sources to exclude certain vessels or process lines
from controls to reduce the overall level of control to
93 percent. One commenter (32) stated that there is no valid
2-85
-------
reason for allowing some terminals to meet a 93 percent MACT
standard while requiring others in the same subcategory to meet
95 percent or 98 percent. One commenter (31) supported the
flexibility of the 93 percent cap and suggested that EPA refine
the option by having the 93 percent control apply only to
substances with a vapor pressure greater than 0.5 psia. Another
commenter (06) suggested that 63.562(c) of the final regulation
should require 93 percent of overall [emphasis added] HAP
reductions.
One commenter (32) stated that the loss in efficiency
between the capabilities of the technology operating at
95 percent or 98 percent efficiency and the MACT standard
(93 percent efficiency) occurs before the emissions are captured,
not after. As a result, it is incorrect to express the MACT
standard in terms of captured emissions. The commenter suggested
revising the MACT standard to require a 93 percent reduction of
all HAP emissions rather than of captured emissions.
Response; The Agency agrees with commenters that a
definitive vapor pressure limit is consistent with the proposed
MACT floor determination for this source category. The proposed
rule with a 93 percent mass limit and no vapor pressure limit is
equivalent to the promulgated standard with the 97 percent
emissions reduction requirement and a 1.5 psia vapor pressure
limit. The Agency incorporated a 1.5 psia vapor pressure limit
in the final rule and has recalculated the MACT floors in the
final rule to include this approach. This approach clearly
defines product categories not requiring control rather than
setting a lower, overall MACT floor control level that
incorporates no-control estimates for low-volatility products.
Comment; One commenter (61) stated that EPA should word its
Title V permit in such a way as to guarantee that the berths not
required to have emission capture equipment not be employed if
their use would allow a terminal to violate the minimum capture
requirement of 93 percent.
2-86
-------
Response: The final rule generally requires the control of
overall emissions rather than berth-by-berth emission reduction
requirements. A terminal is in compliance with the emission
reduction standards provided that the overall emission reductions
contained in the standards are met, regardless of the control
status of individual loading berths.
Comment: One comtnenter (29) stated that the wording of
§ 63.562(b) does not reflect the intent of the standard discussed
in the preamble. The preamble language involving compliance with
the proposed MACT standard is clear and unambiguous; however, the
language in the regulation at proposed 63.562(b) does not reflect
this preamble discussion and should be reworded. Nowhere in the
63.562 section of the regulation has EPA clearly required either
the 95 percent efficient recovery device or the 98 percent
efficient destruction device as a mode of the standard. Only the
93 weight-percent captured HAP emissions standard is mentioned at
proposed 63.562(c).
On the other hand, at proposed 63.562(g), the Agency does
address a 98 weight-percent/95 weight-percent value in relation
to captured VOC emissions under the RACT standards,
These percentages are misleading relative to the preamble
discussion for MACT, hence the MACT standard needs to be further
clarified.
For the purpose of clax~ity and to accurately reflect the
preamble discussion, the commenter suggested rewording
§ 63.562(b) as follows:
"(b) The owner or operator of an affected source under
§ 63.560(a), MACT standards, shall comply with one of
the requirements described in paragraph(b)(i), (b)(11),
or (b)(iii) of this section,
(i) Limit the loading of marine vessels to those
vessels that are vapor-tight and are connected to the
vapor collection system which Is routed to either a
95 percent efficient recovery device or a 98 percent
efficient destruction device.
2-87
-------
(ii) Limit the loading of marine vessels to those
vessels that are vapor-tight and are connected to the
vapor collection system which is routed to either a
95 percent efficient recovery device or a 98 percent
efficient destruction device unless the owner or
operator is loading a vessel with a commodity for which
the owner or operator has made a determination,
pursuant to Section 63.565(d), that control of
emissions from that commodity is not required to meet
the emission limit specified in paragraph (c) of this
section.
(iii) The loading of nonvapor-tight vessels may occur
under the conditions specified in
63.563(a)(2)(iii)(B)."
Response; The Agency agrees with the commenter and has
revised the final standards so that the perceived ambiguity has
been corrected, although the commenter's approach of linking the
standards to vapor collection was not selected. Section 63.563
of the final regulation has been clarified to address the
commenter's concerns where such changes are practicable given
other changes in the compliance provisions found in the final
rulemaking.
2-88
-------
2.4.3 Partial Control of Commodities
Comment; One commenter (72) stated that EPA should revise
requirements proscribing partial control of a given commodity to
incorporate additional flexibility. The commenter suggested
either disallowing partial control of a given commodity only
during a given loading operation or disallowing partial control
of a given commodity at any dock or berth where vapor recovery
equipment has been installed.
Another commenter (32) stated that since the penalty for
using a particular control once would be a requirement to use
that control forever after, the requirements to control all
emissions from an emissions stream creates a substantial
disincentive to control. The commenter recommended that the
relevant language in 63.563(b)(4) be modified as follows: As a
condition of compliance with 63.562(c), each control device must
operate at an efficiency of at least 95 percent for recovery
technologies and 98 percent for combustion technologies. Another
commenter (78) suggested that the final rule be modified to allow
existing sources with installed control systems in place that do
not meet the 95 or 98 percent reduction efficiencies to use those
installed controls in the 93 percent overall emission reduction
calculation.
Response; The Agency agrees with commenters that partial
control of commodities should be allowed provided that the
overall emissions reductions are achieved by the source,
However, the Agency does not believe that such an approach will
be common given the overall average emission reduction used in
the final rule and the incorporation of a minimum vapor pressure
for controlled HAP, The Agency has removed the requirements for
complete control of any given commodity from the final rule and
has deferred the determination of what commodities will be
controlled at a given facility to the designated authority to be
developed as part of the Title V permitting program.
2-89
-------
2.4.4 New/Existing Sources and Requirements
Comment; Several commenters (04, 06, 10, 23, 25, 31, 34,
38, 50, 68, 71, 73, 78, 80, 90) stated that EPA should allow new
sources to use recovery devices to reduce HAP emissions. The
commenters objected to the requirement that new sources use only
combustion and must achieve 98 percent control. While this
efficiency requirement is appropriate for those sources opting to
use combustion, the commenters contend that sources should have
the option to recover vapors and achieve a lower efficiency. The
commenters noted that the use of recovery devices fulfills EPA
support for pollution prevention practices. One commenter (90)
stated that MACT less than 98 percent reduction could be allowed
under Section 112(d)(3) if there are extenuating circumstances
because of the statutory language that states MACT is to be based
on the "best controlled similar source." One commenter (68)
suggested that a two-tiered (destruction/recovery) standard be
adopted for new source MACT, similar to that set for RACT. The
recovery, new source standard should be set at 95 percent. Two
commenters (23, 31) stated that EPA should allow new sources to
use recovery devices, even if they may only achieve a 95 percent
performance level, because the recovery of pollution is
hierarchically superior to the treatment (or, in this case,
destruction) of pollution pursuant to the Pollution Prevention
Act. Further, product recovery is inherently an attractive
cost-savings measure to the petrochemical industry, which would
soften the impact of the retrofit cost. One commenter (71)
believes that "best" control is clearly control by vapor
balancing-condensing or absorbing which achieves a high level of
reduction without generating new, secondary emissions and which
recovers the emissions rather than destroying them. Particularly
for low vapor pressure materials, the environmental benefit of
recovering 3 percent less HAP emissions is easily outweighed by
the environmental damage associated with burning 98 percent of
the emissions rather than recovering 95 percent. The criteria
pollutants generated and the emissions involved in obtaining or
2-90
-------
manufacturing supplemental fuel and replacement product clearly
make a recovery control system the "best" similarly controlled
source. One commenter (04) stated that allowing sources to
recover emissions is consistent with EPA policy decisions made
relative to the HON, whereby EPA allowed sources far better terms
in the emissions averaging provisions if the source achieved
reductions through pollution prevention measures.
One commenter (38) opposed the adoption of condenser control
technology as new source MACT because the commenter does not
believe that condenser technology is a reliable form of control.
Response: The Agency has decided that the proposed new
source MACT standard of 98 percent reduction should be finalized.
The Agency believes that the best controlled similar sources in
all subcategories except the VMT subcategory are those that
reduce HAP emissions by 98 percent. The EPA understands that, at
least for gasoline loading, it appears to be possible to
construct a recovery system that meets the 98 percent reduction
requirement. The Agency does not intend to prevent the use of
recovery devices for new sources, but intends that such devices
meet the 98 percent reduction threshold. The Agency has included
alternative provisions for vapor balancing. These requirements
do not require control of emissions that are reduced by the use
of a vapor balancing system. However, the Agency has required
owners or operators where vapor balancing systems are used to
comply with the provisions pertaining to the vapor collection
system, ship-to-shore compatibility, and vapor tightness of
marine tank vessels. The Agency believes that such provisions
for vapor balancing systems are appropriate because such systems
typically result in near-zero emissions during loading, recover
the emissions rather than destroy them, and reduce secondary
emissions.
Comment; One commenter (31) stated that the proposed rule
did not mention what controls would apply following modifications
of existing marine tank vessel loading sources. The commenter
stated that the modification of existing marine tank vessel
2-91
-------
loading and unloading operations is covered by the 112(g) rule
and will be subject to "existing source" MACT as defined by the
Marine tank vessel loading standards.
Response: The General Provisions of 40 CFR Part 63 describe
the applicability of new source requirements to modified sources.
The final rule has incorporated a cross reference to these
General Provisions to assist the regulated community in
determining the applicable requirements.
2-92
-------
2.4.5 Area Sources
Comment: One commenter (10) stated that a lack of
information to determine adverse effects to human health or the
environment does not justify not addressing area sources. Both
the nature and scope of HAP emitted from marine tank vessel
loading operations and the relatively large number of facilities
(approximately 1,200) that may be considered area sources require
that the impact of these area sources be further investigated.
Response; It is not the intent of this rulemaking to imply
that area sources will not be considered in future regulatory
activities. However, the Agency is unwilling to go beyond the
MACT floor of no control for many of the subcategories for major
sources. It is likely a similar conclusion would be reached
concerning requirements for control of area sources. Moreover,
Section 112 (c) (3) requires a finding that a particular category
of area sources presents a threat of adverse effects to human
health or the environment (by such sources individually or in the
aggregate) warranting regulation under Section 112. The
Administrator has not made such a finding with regard to this
category of area sources.
Comment; Three commenters (50, 75, 34) supported EPA's
decision not to regulate area sources at this time.
Response; The Agency is in agreement with these commenters.
The final rule (as in the proposed rule) does not require
controls for area sources.
2-93
-------
2.4.6 Recovery
Comment; One commenter (34) stated that the short
compliance deadlines would potentially impact the technology
chosen for complying with the vapor control requirements. As
outlined in the proposed rule, vapor control can be achieved by
recovery techniques that capture the displaced vapors or by
destructive techniques that combust the vapors. Destructive
techniques are simpler to install, as well as initially less
expensive than recovery methods. With a short compliance
deadline, facilities will probably opt for the destructive
control techniques to aid in shortening the time needed to
install the equipment. As a result, the short compliance periods
will prejudice facilities to use destructive control techniques,
a result that is contrary to EPA's pollution prevention goals.
Response; The Agency has extended the compliance date for
affected sources, regardless of control technology employed, to
4 years for terminals effected by MACT requirements and 3 years
for terminals effected by RACT requirements, with a 1-year waiver
provision available for RACT sources.
Comment: One commenter (123) stated that the 95 percent
requirement will probably discourage recovery, particularly for
crude oil loading because the economics of recovery are not
supportive of the technology until an emission reduction
approaching 90 percent due to the investment necessary to recover
the light ends. Even at 90 percent the economics are marginal.
Response: The Agency agrees with the commenter that in some
cases site-specific factors at a given terminal may result in the
selection of a nonrecovery control technology. However, the
Agency notes that because these standards are technology driven
and because the selected format of the standards is based on
emissions reduction and not an equipment standard, the owners or
operators of a source required to reduce emissions may select any
control technology provided that the source achieves the
standards.
2-94
-------
2.4.7 Incineration
Comment; One commenter (39) stated that confusion exists as
to what defines each of the combustion devices identified by the
Agency. The commenter suggested that EPA define each combustion
device and any and all applicable requirements in the final rule.
Response: The Agency agrees with the commenter regarding
possible confusion associated with the proposed monitoring
requirements and test methods and has streamlined these sections
in the final rule. Separate requirements are included for
combustion devices and flares, and the term combustion device is
defined in section 63.560 of the final rule. In addition, there
are provisions for owners or operators of affected sources
utilizing other control devices to obtain approval for
alternative test methods and monitoring protocols.
2-95
-------
2.4.8 Format of the Standards
Comment: One commenter (36) objected to the use of a mass
emissions reduction standard because compliance is difficult to
measure, and the composition of gasoline constituents will change
due to reformulations over the next few years, therefore creating
lower volatility fuels and reduced total emissions resulting from
loading operations. The commenter recommended that EPA adopt for
both MACT and RACT a definitive numerical standard instead of
the percentage of emission reductions. The standard for MACT
should be a concentration of HAP in the exhaust stream of
6,300 ppm for new facilities (98 percent reduction) and
22,000 ppmv for existing facilities (95 percent reduction). The
standard for RACT should be a concentration of VOC in the exhaust
stream of 6,300 ppmv for new and existing terminals. Two
commenters (20, 22) stated that emission standards for vapor
control systems could be established for specific commodities in
units of milligrams of HAP (or VOC) per liter of product
transferred. Another commenter (36) suggested that EPA eliminate
the use of a percentage of mass emissions reduction as a
compliance standard and should instead mandate compliance with a
definitive numerical standard, such as parts per million by
volume of emissions concentration in the exhaust stream of a
vapor control unit.
Two commenters (26, 34) noted that the limit of 1,000 ppmv
on emissions on a dry basis corrected to 3 percent oxygen does
not make sense for a recovery device and is only applicable to a
combustion device. The commenter recommended that this sentence
either be revised to a maximum vent concentration of 1,000 ppmv
or correlated to mg of VOC's per volume of product loaded, as is
the case in California standards. These comments also apply to
Sections 63.563(b) (3) (ii), (iii), and (iii) (B) and 63.565(c) (ii)
and 63.566(b)(1)(ii).
One commenter (78) supported the decision not to propose a
mass per unit loaded alternative. The commenter stated that a
standard based on removals using specified control or recovery
2-96
-------
technologies is much preferred for simplicity to the regulated
community.
Response; The Agency agrees with the commenter that there
are various formats by which to structure the compliance test.
The Agency has limited data from which to develop definitive mass
rate requirements. Therefore, the Agency maintained the approach
developed for the benzene transfer NESHAP (40 CFR Part 61,
Subpart BB). This approach requires an efficiency determination
during the last 20 percent of compartment loading (i.e., test
during the highest concentration of pollutants in the emissions
stream).
Regarding the comment about the correction to 3 percent
oxygen, the Agency notes that the final rule requires Method 25A
for determining and/or calculating efficiency. Method 25A does
not require a correction to 3 percent oxygen.
2-97
-------
2.4.9 Regulation of Two Berths within VMT
Comment: Several commenters (07, 08, 32, 53, 55, 58, 129,
131) supported Alyeska's proposal to install controls on two
berths and to implement Federally-enforceable operational
limitations on a third berth, which would have limited and
temporary use due to declining throughput. One commenter (32)
stated that because the VMT should be placed in a separate
subcategory and because the MACT floor for this subcategory would
be zero, regulatory alternatives above the floor must be
considered. The commenter stated that the incremental costs and
benefits of controlling VMT Berth 3 should, therefore, be
considered when evaluating requirements for a vapor control
system on Berth 3. The commenter recommended that the MACT
standard require operating limitations on that berth.
The commenter (32) stated that Federally enforceable
limitations on use of berth 3 would provide an adequate mechanism
to ensure that the VMT will achieve compliance with applicable
emission standards for three reasons: First, a limit on loaded
quantity would correspond to the justification for relying on an
operational limit. Second, the use of Berth 3 would not easily
correlate to an emission reduction ratio because both total
loading and loading over Berth 3 are declining. Finally, loaded
quantity is easily quantifiable, and data on loading are already
collected and are readily available, resulting in a readily
enforceable loading limit.
Three commenters (06, 08, 62) were opposed to the exemption
from control of operating berth No. 3 at the VMT. One commenter
(62) requested that Alyeska be required to meet the 93 percent
requirement and that emergency and upset conditions be calculated
as part of this 93 percent. One commenter (06) stated that if
Alyeska could demonstrate that the incremental cost effectiveness
of controlling a third berth at VMT is disproportionate relative
to the incremental cost-effectiveness values used in the EPA's
MACT analysis then it may be appropriate to allow an uncontrolled
berth. However, this berth would have to be subject to
2-98
-------
throughput limits consistent with the values used in the cost
effectiveness analysis and to stringent throughput monitoring
requirements, including some form of non-resettable flow meter on
the crude oil pipeline leading to the berth. Any violation of
the regulation throughput limit should require that controls be
installed within a specified period of time, regardless of berth
throughput.
One commenter (67) noted that most North Slope oil field
production forecasts understate future production prospects and
that the Agency should not relax emission reduction standards
based on these declining throughout data. The commenter also
stated that the North Slope operations are so profitable--even at
current oil prices--that there are no financial barriers to
requiring full vapor recovery at Alyeska's VMT.
One commenter (52) noted that, in making a decision to
regulate the third berth, EPA should ideally consider many
factors, including those outside of the air program. The
commenter does not believe that control of the third berth is the
best use of capital funds given the scope of the environmental
challenges facing the North Slope oil field, pipeline, and marine
tank vessel loading terminal. The commenter suggested that if
the third berth is riot controlled in the final standards, EPA
should provide for re-evaluation of the standards in the event of
increased throughput from the North Slope oil field.
One commenter (3^) suggested that, as part of approving
Alyeska's proposed vapor control strategy, EPA should adopt a
rolling 12-month average for measuring compliance with the RACT
requirements. By using this approach, EPA would be able to
determine on a monthly bas.ls whether the annual loading averages
delivered over Berth 3 are consistent with the throughput caps on
the berth.
Response: The Agency has determined that the Alyeska
Pipeline Service Company's (APSC's) Valdez Marine Terminal (VMT)
is a separate subcategory (see discussion in Section 2.3.4 of
this document). For this subcategory, the MACT floor is no
2-99
-------
control, as one commenter stated in Section 2.3.4 of this
document. However, as pointed out in the preamble to the
proposed rule and by several of the commenters, the cost
effectiveness for controls beyond the MACT floor is not
prohibitive and is lower for the VMT than for other terminals in
the source category. Therefore, the Agency has elected to
require controls more stringent than the MACT floor for this
subcategory (see Docket No. A-90-44, item No. IV-B-2).
In adopting this more stringent option for this subcategory,
the Agency also recognizes that the same factors that contributed
to the establishment of a subcategory also affect the selection
of MACT (i.e., declining throughput, shortened construction
season). Some commenters initially noted that the cost of
controlling VOC may be high relative to the benefits of
controlling VOC at a remote site in an Arctic ozone attainment
area. Additionally, some commenters initially stated that the
benefits of controlling HAP would not appear to justify the
costs. However, the Agency has also considered later comments
from the APSC, the State of Alaska, the Prince William Sound
Regional Citizens Advisory Committee (a local citizens group) and
private citizens in determining MACT/RACT for the VMT. These
commenters agreed that a Federal rule mandating control of
primary emissions at the APSC was acceptable. After careful
consideration of the costs, the environmental impacts and the
comments, the Agency decided that MACT for this subcategory was
control beyond the level of the MACT floor (see Docket A-44-90,
Item Number IV-B-2).
Because the VMT would be above any Title I applicability
limits, the Agency reviewed both RACT and the selection of
technology more stringent than the MACT floor for MACT (see
Docket No. A-90-44, item No. IV-B-2).
The Agency concluded that both RACT and MACT require control
of the primary berths (two berths) but allow uncontrolled
emissions at the remaining berths; throughput at the uncontrolled
berths would be gradually reduced to a minimal level (to allow
2-100
-------
for maintenance) above which controls would be required. The
average cost effectiveness of both VOC and HAP control at the
primary berths was approximately $1,000 and $7,000 per Mg of
pollutant reduced respectively. The Agency rejected control of
the emissions at the remaining berths due to the high incremental
cost effectiveness of approximately $3,000 to $6,000 per Mg of
VOC and $28,000 to $55,000 per Mg of HAP and the limited time
(estimated at 4 years) that a significant amount of pollution
will be emitted from the uncontrolled berths. If at any time
actual throughput from the uncontrolled berths are greater than
that permitted under the regulations, emission control equipment
must be put in place for the affected uncontrolled berth. In
addition, controls at the VMT must be in place by December, 1997,
six months earlier than at other sources.
Comment; One commenter (32) noted that there still will be
the occasional need to use Berth 3 after it is retired from
routine operations (either during maintenance on one of the other
berths or due to malfunctions.) Therefore, Alyeska should be
permitted to load using an operationally limited berth in cases
of emergency or for short periods as necessary fox. maintenance.
The commenter requested that EPA authorize the limited use of
Berth 3 at the VMT dviring maintenance and malfunctions for a
period not to exceed 40 days of loading time in any year. Such a
limit should also be exclusive of any emergency needs.
Response: The Agency agrees that some limited use of the
uncontrolled loading berths would be necessary following the
implementation of the strictest emissions limits under the final
rule. Therefore, the fina!! rule allows the continued use of the
uncontrolled berths for maintenance beginning in 2000 provided
that the uncontrolled throughput limits (which take affect in
1998) established in the standards for this subcategory are
adhered to by VMT.
2-101
-------
2.4.10 RACT and MACT in Same/Separate Rules
Comment; One commenter (29) stated that attempting to
understand the regulatory language in this proposal has been very
difficult due to the intermix of the requirements for a RACT
standard with the requirements for a MACT standard. The
commenter recommended that the Agency separate these two rules in
the final promulgation and place the MACT standards in Part 63,
Subpart Y and the RACT standards in a different part/subpart.
Response; The Agency has streamlined the final regulation
to make it more user friendly. The Agency does not believe that
there is a need to separate the RACT and MACT rules given that
they basically only differ in applicability.
2-102
-------
2.4.11 Parallel With The HON
Comment: One commenter (29) noted that since the control
devices, monitoring, performance testing, and other requirements
contained in this proposal are related most directly to identical
or similar requirements in the HON, the Agency should establish
similar requirements to the HON in Subpart Y so that owners or
operators are confronted with a consistent set of requirements
for similar operations.
Response; The Agency agrees with the commenter's basic goal
of making rules consistent where possible. The Agency has used
approaches finalized in the HON rule where possible.
2-103
-------
2.4.12 Alternative Standard
Comment; Two commenters (22, 36) stated that the
requirement of reducing the vent to 1,000 ppmv is extremely
severe and represents 99 + percent removal efficiency. One
commenter (22) suggested a limit of 10,000 ppmv total hydrocarbon
measured as propane and averaged over the test period. Another
commenter (36) recommended that EPA adopt a definitive numerical
standard for both RACT and MACT standards instead of the
proposed percentage of emission reductions. The standard for
MACT should be a concentration of HAP in the exhaust stream of
6,300 ppm for new facilities (98 percent reduction) and
22,000 ppmv for existing facilities (95 percent reduction). The
standard for RACT should be a concentration of VOC in the exhaust
stream of 6,300 ppmv for new and existing terminals. Also, EPA
should specifically include "vapor balancing" as an appropriate
means of controlling vapor emissions in the text of the final
regulations.
Response; The Agency has provided as much flexibility in
the final rule as possible given the underlying data available to
the Agency. The commenter correctly states that the 1,000 ppmv
concentration limit may only be achievable at facilities with
gasoline vapors. The 1,000 ppmv limit was derived from test data
from a facility controlling gasoline vapors. The Agency wanted
to provide an alternative to the percent emission reduction
requirements but had test data from only one facility from which
to derive an alternative. Although the commenter has provided
alternative concentration limits to the proposed 1,000 ppmv
limit, no additional test data was provided to demonstrate the
effectiveness of the alternative concentration limits. The
limits imply a certain average inlet concentration, but no data
were provided to substantiate the recommendations. Therefore,
the Agency has retained the 1,000 ppm limit in the final rule for
gasoline loading. The Agency has also included alternative
compliance provisions in the final rule whereby an owner or
operator of an affected source could, subject to the
2-104
-------
Administrator's approval, demonstrate compliance with the
emissions reductions standards at an alternative concentration
limit during an initial performance test and employ continuous
emissions monitors to show compliance thereafter. Sources
required to reduce emissions also have the option for all control
devices to establish a baseline VOC concentration limit during
the performance test.
The Agency agrees with the commenter's request for
requirements to address vapor balancing as a means of controlling
emissions from marine tank vessel loading. The Agency has
included requirements that do not require control of emissions
that are reduced by the use of a vapor balancing system.
However, the Agency has required owners or operators where vapor
balancing systems are used to comply with the provisions
pertaining to the vapor collection system, ship-to-shore
compatibility, and vapor tightness of marine tank vessels. The
Agency believes that such provisions for vapor balancing systems
are appropriate because such systems typically result in
near-zero emissions during loading, recover the emissions rather
than destroy them, and reduce secondary emissions.
2-105
-------
2.5 COMPLIANCE SCHEDULE FOR TITLES I AND III
Comment; Many commenters (04, 23, 24, 28, 29, 30, 31, 34,
36, 39, 41, 42, 47, 50, 51, 56, 68, 71, 75, 78, 102, 103, 104,
106, 107, 108, 109, 111, 112, 113, 115, 117, 119, 121) stated
that the length of the compliance periods for RACT and MACT are
not adequate. A number of these commenters stated that these
rules do not allow adequate time for all affected marine loading
and unloading operations to come into compliance. Three
commenters (34, 75, 107) stated that compliance concerns are not
limited to the installation of equipment; the lead time for
facilities to meet permitting and safety approvals from
permitting authorities and the U.S. Coast Guard within the
limited compliance periods must also be considered. Potential
delays could be associated with State permitting requirements,
where other MACT rules, operating permits, and public review
could further delay permit approvals. These administrative and
permit review requirements will add to the time needed to comply
with the provisions as proposed. Up to 6 months have been
required to obtain safety certification and approvals from the
Coast Guard for the vapor control system. Based on the proposed
rule, the commenters believe that a significant number of
facilities may need to curtail operations while waiting for
construction completion, permit approvals, and safety
certifications. Moreover, commenters stated that there are a
limited number of contractors experienced in installing control
equipment for marine tank vessel loading facilities. Given the
limited availability of such contractors to work on numerous
projects at once, terminals will need to wait for an experienced
contractor to become available. Efforts to rush numerous
facilities into compliance using inexperienced contractors would
raise safety concerns. The commenters suggested that EPA show
that the RACT requirements under Section 183(f) are exceeding
(or, at 93 percent, matching) the proposed MACT requirements and
grant a longer compliance period, at least matching the MACT
compliance period of 3 years. The commenters also suggested that
2-106
-------
if EPA promulgates the Section 112 rule in 1995, the Agency
should grant all facilities subject to the control requirements
4 years from the promulgation date to comply.
Commenters also noted that over 2,000 vessels would likely
require the installation of retrofit equipment to be compatible
with marine terminals that have installed vapor collection
equipment and emission controls. Commenters also noted that
their experience with State regulation of marine tank vessel
loading facilities indicated that, even with the smaller number
of facilities regulated by individual states, deadlines of
2 years were routinely missed, and States were forced to grant
extensions. In general, commenters believed that the 2-year
deadline for RACT standards and the 3-year deadline for MACT
standards are unrealistic and essentially impossible to meet,
especially given the large number of sources that will need the
services of a limited number of control equipment design and
installation contractors.
Other commenters (36, 71, 110, 111) proposed a 4-year
compliance date for MACT and RACT. One commenter (110) supported
a 4-year compliance date for RACT and believes that the MACT
standards are not defensible, One commenter (109) favored a
1-year extension available upon a source's determination that
they were unable to meet the compliance date.
Three commenters (04, 23, 34) recognized the statutory
constraint of a 2-year compliance period under RACT but believed
that EPA has the discretion to extend this period. One
commenter (23) stated that EPA's analysis of the retrofit
activities fails to adequately consider the number of skilled
engineering and construction firms experienced in doing such
retrofits and believed that EPA has the discretion to relax the
2-year compliance period under che criterion of "reasonably
available" by considering tha supply of experienced engineering
and construction personnel. The commenter requested a 3-year
compliance period for RACT standards, with the allowance for
sources to apply for an additional 1-year extension. Two
2-107
-------
commenters (04, 34) suggested that Congress intended that the EPA
address only those sources located in ozone nonattainment areas
by locating the requirements of Section 183(f) in the part of the
Act that concerns ozone nonattainment. The EPA should at least
allow those sources located in attainment areas a longer
compliance period. For MACT, one commenter (23) recognized the
statutory constraint of a 3-year compliance period under MACT,
but believed EPA has the discretion to set policy that would not
unduly disrupt the operation of the marine tank vessel loading
industry. Therefore, the commenter urged EPA to allow a 4-year
compliance period. The commenter added that EPA should recognize
that the 1-year extension application under the General
Provisions to 40 CFR Part 63 is a cumbersome procedure with too
many onerous, enforceable requirements in a 1-year period to
serve as a realistic alternative. If the automatic extension is
not granted, the EPA should streamline and relax the requirements
for obtaining a 1-year extension under the MACT standards of the
proposed rule. One commenter (105) stated that the extension
language found in the General Provisions of 40 CFR 63.6(i) was
sufficient for facilities requiring additional time for
compliance. Another commenter (116) stated that since the
standards are based on "achievable" control technologies, then
the compliance dates as proposed should stand unless individual
terminals require additional time to achieve compliance. If
needed, this additional time should be provided through State and
local regulations under the Title V permit program. Commenters
(105, 108) also suggested a phased-in approach or a determination
that "compliance" at the 3-year deadline could be interpreted to
mean having contracts in place. However, one commenter (116)
opposed the use of a phased-in regulatory approach that is
related to a source's attainment/nonattainment status.
One commenter (32) pointed out that the Preamble says the
deadline for compliance with MACT is 2 years after promulgation,
yet the proposed regulation states that the deadline is 3 years
after promulgation.
2-108
-------
One commenter (78) recommended that EPA provide ways to
allow at least a 3-year compliance timeframe for any marine
facility required to install vapor control equipment. This
commenter requested that EPA clarify in the preamble to the final
rule that an extension is available when necessary.
One commenter (103) provided options whereby the Agency
could extend the compliance date. The commenter's preferred
options limited the number of facilities having to comply with
the standards by eliminating the Section 112 rule and restricting
the scope of the Section 183 (f) rule to facilities located in
nonattainment areas. The commenter also suggested that the
Agency could interpret the "effective date" to mean that progress
must be made towards achieving compliance at a future "compliance
date." The commenter suggested that the Agency examine the
approach used in a Section 211 rulemaking (59 FR 33047).
Commenters also suggested phasing-in compliance so that
Section 183 (f) facilities in ozone nonattainment areas would be
required to comply first, followed by Section 183 (f) facilities
in ozone attainment areas, then by smaller facilities subject to
the Section 112 standards only. Commenters also suggested the
deferred promulgation of thes^ standards until 2000 so that
contractors and vendors could gain sufficient expertise and
capacity to satisfy the demand for control equipment. Another
commenter (108) also favored delay of the standards, the
establishment of additional subcategories, and a phased-in
compliance schedule for the different subcategories as a means of
improving the ability of sources to comply with the standards in
the given time frame.
On March 8, 1995, the Agency reopened the comment period to
solicit additional comments on the issue of whether the
compliance periods for the RACT and MACT regulations should be
extended. Numerous commenters indicated support for the
extension of the compliance periods, generally reiterating the
views expressed in earlier comments. They also indicated that
EPA has the requisite legal authority to extend the compliance
2-109
-------
periods, noting particularly Section 183(f)'s language concerning
"reasonably available" technology, the imprecision of the term
"effective date," and EPA's general authority to grant a l-year
waiver from MACT compliance under Section 112. Some comtnenters
also noted that greater environmental benefits can be obtained by
granting longer compliance periods. The commenters stated that
longer compliance periods can sometimes allow for better
designed, more robust, safer, and more advanced technologies.
The commenters also noted that, in this instance, an extended
compliance period could result in greater use of recovery
technologies and less use of incineration. One commenter (139)
provided a list of several marine loading terminals in California
that had installed emission control equipment and indicated that
almost all of the projects required a minimum of 3 years to
complete.
Response; The Agency agrees with the commenters that
permitting and safety approvals from permitting authorities,
potential delays associated with State permitting requirements,
the potential lack of skilled engineering and construction firms,
and the history of facilities unable to comply with existing
State regulations compel the Agency to extend the compliance
times for both the RACT and MACT rules.
Therefore, the Agency recognizes the need to provide a
3-year compliance date for Title I facilities to achieve full
compliance with the standards; facilities are required to meet
commencement requirements within 2 years after promulgation.
Further, the agency granted a blanket 1-year waiver for
facilities subject only to the MACT requirements. This waiver
recognizes the limited personnel to design and begin construction
of all the facilities in the first 6 to 18 months after
promulgation and enables facilities having the highest emissions
to comply first and the smaller facilities second, as resources
become available after the anticipated initial period of
activity.
2-110
-------
The EPA shall allow sources regulated under Section 183(f)
3 years to be in full compliance with the emission control
requirements promulgated under Section 183 (f). In addition,
sources may request a waiver of up to 1 year to achieve full
compliance with the requirements if they can show that special
circumstances exist that require additional time for compliance
with the RACT standards. The EPA believes that this result is
consistent with Section 183(f). Section 183 (f) requires the
application of "reasonably available" control technology that
considers costs, any nonair quality benefits, environmental
impacts, energy requirements, and safety factors. Section 183(f)
also requires that the effective date of any regulation shall be
no more than 2 years after promulgation of such regulations.
The overwhelming evidence received by the Agency indicates
that most, if not all, of the sources that must install emission
control devices cannot do so within 2 years. States that have
attempted to establish such a requirement have been forced to
provide waivers to the regulated sources. In addition, given the
relative scarcity of qualified contractors and the permitting and
other requirements necessary for such construction, it is clear
that the emission control technologies required by this rule will
not be "reasonably available" within 2 years of the promulgation
of this rule, Moreover, the information provided to the Agency
indicates that a 2-year deadline may force regulated sources to
install equipment that is less reliable and may cause safety
concerns. Given the emphasis that Congress put on safety in this
regulation and the fact that the Coast Guard will need to review
such installations prior to operation, a 2-year deadline seems,
based on the evidence, to be contrary to Congress's broad intent
and may result in conflicts with Coast Guard requirements. Also,
the 2-year deadline may force sources to install incineration
devices, rather than recovery devices, because incineration
devices are more straightforward to install, though they can
cause more secondary pollution. Given the Agency's and Congress
determination to encourage pollution prevention and other
-------
programs to lessen pollution, EPA believes that forcing the use
of incineration is not an environmentally beneficial approach.
The Agency is requiring that regulated sources undertake a
continuous program of construction or that these sources enter
into a contractual obligation to undertake and complete a
continuous program of construction within 2 years after
promulgation. Based on comments received on the proposed rule,
the EPA believes that these actions can reasonably be achieved
within 2 years of promulgation.
The EPA has in the past provided sources with reasonable
time to complete actions required by the Clean Air Act. For
example, in a rulemaking published on June 27, 1994
(59 FR 33042), EPA allowed regulated fuel/fuel additive
manufacturers 6 years to complete certain testing required under
Section 211. Section 211 requires that "requisite information"
be provided within 3 years of promulgation. However, due to
Agency concerns that the extensive testing requirements and
limited laboratory facilities would mean that the regulated
parties would not be able to complete testing in 3 years, the
Agency allowed 6 years for completion of "Tier 2" testing but
required "Tier 1" data and other materials to be provided within
3 years.
Moreover, EPA believes that the imprecision of the term
"effective date" could also provide EPA with the ability to allow
full compliance later than 2 years after promulgation. The
effective date of this regulation will be the date of publication
in the Federal Register, which obviously meets the 2-year
requirement of the statute. Though it can be argued that the
language of Section 183 (f) may imply that the term "effective
date" actually is referring to the date of source compliance, the
distinction between "effective dates" of regulations and
"compliance dates" is an important one and has been a clear part
of administrative procedure for many years. See, e.g..
Section 112 (i) (3); Natural Resources Defense Council v.
2-112
-------
Environmental Protection Agency. 22 F. 3d 1125, 1138 (B.C. Cir.
1994) .
The EPA believes that the 3-year deadline in particular is
appropriate because the terminals that will be subject to the
RACT standards that are not yet controlled (based on commenter
data) represent some of the largest of the terminals controlled
under this final rule. It is expected that these terminals
should be able to install control measures within the three year
time period allotted under the final RACT standards. However,
should additional time be required, there is an available 1-year
waiver that would provide a total of 4 years to be in complete
compliance with the final standards. The Agency believes that
most RACT terminals will be able to meet the emissions reduction
requirements contained in the final standards within the three
years following the promulgation date. The Agency estimates that
only 8 terminals subject to the RACT requirements are not
presently controlling emissions to the level specified in the
standards. These terminals are among the largest terminals in
the U.S., and can reasonably be expected to have in-house staff
capable of assisting in the design and installation, of control
technology. Fiirthermore, the Agency is aware that some of these
terminals are already designing control equipment in anticipation
of these final RACT requirements.
The EPA shall allow existing sources regulated under
Section 112 4 years to be ID full compliance with the emission
control requirements promulgated under Section 112. Sources must
generally comply with M&CT standards under Section 112 within
3 years of promulgation. However, Section 112 (i) of the Act
specifically allows EPA to provide sources with a waiver of up to
1 year to achieve full compliance with the requirements if they
can show that the additional period is necessary for the
installation of controls. The Agency agrees with commenters who
stated that standards containing similar compliance dates for a
large number of sources would result in numerous facilities
competing for a limited number of experienced contractors in
2-113
-------
order to meet the standards at the same time. Several of these
commenters suggested a staggered compliance schedule for the
sources affected by the standards. The Agency notes the
information supplied by commenters indicating that many source
would require more than 3 years to install the required control
equipment given the limited number of contractors experienced in
installing control equipment in marine loading facilities and the
lead time needed to meet permitting and safety requirements from
permitting authorities and the U.S. Coast Guard.
The Agency agrees with the commenters that many MACT sources
would probably require 1-year waivers if there was a three-year
compliance date for MACT sources in the final rule. The Agency
notes that these sources are typically smaller than the sources
regulated under RACT, and would not be as likely to have in-house
staff capable of assisting in the design and installation of
control technology. Therefore, the Agency believes that the
sources controlled under Section 112 that are not controlled
under Section 183 (f) should receive a waiver for 1 year. The
Agency believes that this total of 4 years is sufficient time for
the estimated 20 sources presently uncontrolled to design and
install control technologies sufficient to meet the MACT
standards. The Agency believes that the staggered compliance
schedule (i.e., 3 years for RACT terminals and 4 years for MACT
terminals) coupled with the reduced number of terminals required
to control emissions under the final rule should alleviate
commenters' concerns about the scarcity of qualified installation
consultants and vendors. This extended schedule is also expected
to address concerns regarding permitting delays.
The Agency is providing the Valdez Marine Terminal with
30 months to be in full compliance with these regulations. The
Agency believes, per its discussions with Alyeska, that this
extension provides sufficient time to comply with the promulgated
rule.
Comment: One commenter (29) stated that the compliance date
for existing sources is appropriate, but clarifications are
2-114
-------
needed to § 63.560(d)(l) [as well as to § 63.567(b)] regarding
when to file an initial notification. The MACT applicability
thresholds would allow 2 years to comply with the existing source
MACT standards. The wording of the rule at 63.560(d)(1)(i) and
(ii), however, is confusing. Proposed 63.560(d)(1)(i) and (ii)
appropriately provide that compliance be achieved in "3 years,"
not "2 years." The preamble discussion, however, does not
reflect the "3 year" compliance timing of the CAA at CAA
Section 112(i)(3)(A). It appears that the Agency intended to
trigger the 3-year compliance date for an existing source after
such source exceeded the 1 Mg/yr of HAP threshold required at
proposed 63,560(b). The commenter concurred that this is an
appropriate means of establishing compliance, particularly
because most sources will not exceed the 1 Mg/yr of HAP emissions
on the exact date that this rule is promulgated.
Two commenters (29, 30} proposed that the Agency eliminate
63.560(d)(l)(i), renumber the subsequent paragraphs, and modify
the last paragraph for new sources so that the wording is
consistent with the wording in the remaining paragraph for
existing sources. The regulated community then will have a clear
understanding of the requirements. Therefore, 63.560(d)(l)
should be modified as shown below.
(d) Compliance dates.
(1) MACT standards compliance dates.
(i) Any source that exists as of the effective date of the
standards that subsequently exceeds the applicability thresholds
specified in § 63.560(b) shall be subject to existing source
requirements and must be in compliance with the provisions of
this subpart within 3 years from the month in which the source
exceeded the threshold levels.
(ii) Any source affected under § 63.560(a) constructed or
reconstructed after the date of promulgation shall be subject to
new source requirements and must be in compliance with the
provisions of this subpart upon startup.
2-115
-------
Response; The Agency agrees with the commenter that the
requirements for new and existing sources could be clarified and
has done so in the final rule.
Comment; Several commenters submitted comments specifically
applicable to the compliance schedule for VMT.
Seven commenters (06, 07, 32, 34, 115, 118, 131) stated that
the compliance schedules are not sufficient for VMT to comply.
Two commenters (06, 07) stated that the compliance schedule is
too aggressive because VMT must install the largest vapor control
system ever undertaken, handling large volumes of explosive
vapors. One commenter (06) urged EPA to consider the means by
which three construction seasons (ending September, 1997) could
be allowed for the VMT to come into compliance with the
provisions of Section 183(f). Another commenter (52) urged EPA
to consider the shortened construction season when developing
standards for Alaskan marine tank vessel loading operations.
Two commenters (32, 34) stated that EPA should exercise its
discretion to regulate the VMT only under Section 112 in order to
provide a reasonable and safe period for Alyeska to design and
construct a vapor control system for Berths 4 and 5. One
commenter (32) stated that a number of vapor control projects
(for projects much smaller than VMT) completed in the last
2 years required 24 to 48 months to permit, design, construct,
and begin operations. Projects much smaller than the proposed
VMT vapor control system have required considerably more than
24 months from final rule promulgation to the compliance
deadline. The EPA Administrative Record shows that a number of
terminals in the United States that have installed vapor control
equipment did not install their equipment within a 2-year time
frame.
Five commenters (08, 09, 55, 58, 78) stated that the
compliance schedules for the VMT are adequate and should not be
changed. One commenter (09) strongly disagreed with any time
extension on completing the VMT vapor recovery system, noting
that the entire pipeline and terminal were built in 4&-years.
2-116
-------
Furthermore, those residents who are concerned about long-term
health are not willing to compromise their health another summer;
1996 is still too long from now. Four commenters (08, 55, 58,
78) stated that the Regulations under Title III will result in a
3-year compliance timeline, rather than the 2-year time frame
proposed under Title I. Because of the short construction and
maintenance season in Valdez, they believe that the 3-year
timeline is more appropriate.
Response: As stated previously, the Agency has extended the
compliance time for VMT to 30 months. The Agency believes, per
its discussions with the Alyeska Pipeline Service Company, that
this extension is sufficient time for the VMT to comply with the
promulgated rule.
Comment: Two commenters (34, 75) stated that another
problem with the near-term deadline for a. final rule, combined
with the short term compliance deadline, is that many vessels
from foreign locations will not be equipped with appropriate
vapor control equipment, thereby creating safety and distribution
problems. As a result, marine tank vessels will be turned away,
or they may attempt cargo transports with unsafe equipment.
Obviously, more time is needed for foreign-flagged vessels to
achieve the compliance requirements for this rule.
Response: The compliance schedule has been extended, and
the Agency considers the schedule to allow adequate time for
retrofit of vessels.
Comment; Two couanenters (59, 60) recommended that
implementation of the regulations be delayed until the U.S. Coast
Guard can confirm that due consideration has been given to the
safety aspects of controlling vapors other than crude oil,
benzene, and gasoline, Chemical tankers differ from crude or
product tankers. The constraints set forth in this proposed rule
do not readily apply to the loading operations aboard these types
of vessels.
Response; Pursuant to the language of Section 183(f) of the
amended Clean Air Act, the Agency has worked closely with the
2-117
-------
U.S. Coast Guard during all phases of the development of these
standards. The final rule has been reviewed by the Coast Guard
and addresses their safety concerns.
2-118
-------
2.6 COMPLIANCE, PERFORMANCE TESTING, AND MONITORING
2.6.1 Compliance
Comment; One commenter (12) strongly suggested that the
final regulation address the loading cycle startup and shutdown
periods. Because the marine loading operation is a batch
process, a time interval needs to be incorporated to allow for
the normal control response for these process changes. The EPA
has similar language and logic for process control response in
the Refinery MACT preamble.
The recovered vapors will have varying BTU content and flow
rates depending upon several factors: loading rate, product Reid
vapor pressure (RVP) and temperature, and vapor growth during the
loading cycle. Because of these variations, at least a one,
5-minute interval is required to allow for adjustments to the
auxiliary fuel or quench air to stabilize the stack temperature
during any loading cycle of 3 hours or less, or a maximum of
1 hour if the loading cycle is more than 3 hours. In addition to
the normal startup and shutdown periods (except for excursions
during startup), shutdown that meets the criteria for a
malfunction specified in the General Provisions to 40 CFR Part 63
should also be included in the final rule. The temperatures
during these startup and shutdown periods should not be included
in the average temperature requirement for the loading cycle.
Response: Regarding concerns about excursions during
loading cycle startup and shutdown periods, the Agency believes
that short-term excursions should not affect the source's ability
to meet the emissions standards. Monitoring of parameters (VOC
concentration, temperature, pressure, flowrates, etc.) occurs at
15-minute intervals and is averaged over 3 cycles or 3 hours
(depending on the baseline parameter established in § 63.565); it
is not anticipated that a 5-minute excursion, as mentioned by the
commenter, would prevent the source from meeting the baseline
parameter. The Agency would like to note that the compliance
violations related to exceedances of baseline parameters that
these commenters are concerned about have been deleted from the
2-119
-------
final regulation and have been replaced with requirements to
perform an unscheduled inspection of the control equipment and
monitoring device(s) when exceedances of the baseline parameters
occur. The inspection is a component of the operation and
maintenance plan specifically related to the control device and
monitoring equipment developed to ensure that pollution control
equipment and monitoring equipment functions properly and that
variances of the control equipment and monitoring equipment are
minimal [see § 63.562(e)]. The Agency also notes that the
frequency of this component inspection should be determined based
on the degree to which the improper operation of the component
affects the efficiency of the control system, whether or not the
component can be inspected while the control device is in
operation, and the frequency with which the component must be
repaired or replaced. The Agency has provided for a vendor-
recommended inspection schedule (standard operating procedure) or
an applicable schedule published in a technical publication
regarding the operation of the control system that serve as a
guide for establishing this part of the operation and maintenance
plan.
As specified in § 63.6(e)(3) of the General Provisions, all
sources in this source category have been waived from the
requirement to develop a startup, shutdown, and malfunction (SSM)
plan. While the Agency has not required sources to develop a SSM
plan, a source may choose to voluntarily develop a startup,
shutdown, and malfunction plan if they have a concern regarding
the source's ability to meet the standards due to HAP being
emitted during startups, shutdowns, and malfunctions.
Comment: One commenter (34) recommended that the final rule
include provisions for a maintenance allowance when facilities
need to conduct repairs to the loading berth or vapor control
systems or are faced with unique compliance circumstances not
otherwise addressed in this regulation. Such a variance would
allow EPA to make case-by-case assessments, thus allowing
facilities to continue operations, either at the berth being
2-120
-------
repaired or, on a limited basis, at a stand-by berth at the
facility not equipped with vapor controls. Further, a variance
mechanism could provide a facility with a compliance extension
when the compliance period available cannot be achieved due to
limited equipment availability or other delays beyond the control
of the facility owner or operator. Current regulations in New
Jersey and California provide for such variances.
Response; The Agency agrees with the commenter that
maintenance allowances may be requested and obtained. Since the
State regulations do not apply during periods of repair (as
mentioned by the commenter), the Agency has included language
pertaining to a maintenance allowance in Sections 63.562(b)(5)
and (c)(6) of the final rule whereby an owner or operator of an
affected source may apply to the Administrator for a maintenance
allowance. Additional provisions for relief are found under the
General Provisions in 40 CFR Part 63. Under these General
Provisions, each facility has the option to voluntarily develop a
SSM plan that prescribes actions the facility will take if it is
operating under conditions of malfunction (40 CFR 63.6(e)(3)), as
discussed above.
The Agency agrees that additional time will be required for
MACT sources to comply with the standards. The compliance date
has been extended to four years after the date of promulgation.
However, in doing this, the Agency has in essence already granted
a one-year waiver per the General Provisions. No additional
extensions of the compliance date are available for MACT sources.
Comment; One commenter (32) stated that, of the
technologies available to recover liquid hydrocarbons, the Cool
Sorption process would have the highest HAP and VOC recovery
efficiencies and the lowest costs to build and operate. The Cool
Sorption technology provides about 96 percent HAP and 79 percent
VOC removal from the ship vapor collected in the two or three
berth configuration. Another commenter (32) stated that
SORBATHENE adsorption systems have a 99 percent HAP removal
efficiency, and a VOC removal efficiency ranging from 32 to
2-121
-------
79 percent. The capital cost of these configurations is
relatively high. Also, the impact of hydrogen sulfide and other
ship gas contaminants on activated carbon adsorbent performance
can be significant. As a result, configurations based on the
adsorption process are not practical for the VMT.
One commenter (34) stated that Section 63.564(c) of the
proposed regulation (requirement for a vacuum pump for carbon bed
regeneration) is unnecessary given the performance requirements
for the system. The commenter believes that EPA should follow a
performance standard approach for these regulations and avoid
specifying technology requirements.
One commenter (32) stated that processes based on single
stage refrigeration systems fail to achieve the target of
95 percent HAP removal efficiency, and they provide a very low
overall VOC removal efficiency. Two stage refrigeration systems
satisfy the 95 percent HAP removal efficiency but only provide
about 50 percent overall VOC removal efficiency. All
refrigeration based processes have the added risk of requiring a
means to inhibit ice formation or providing a freeze-thaw design.
The commenter's poor experience with glycol dehydration
operations to remove water vapor reflects serious technology
risks associated with this process. Refrigeration technology
poses high process uncertainty due to freezing of the chiller
because of poor dehydration and because of the remaining high
moisture content of the collected ship vapor. Another commenter
(38) noted the reliability of condenser/refrigeration units is
highly suspect due to possible needs to defrost both the
intercooler tubes within the condenser and the refrigeration
equipment providing the cooling (if a dedicated refrigeration
unit is applied). Vent condensers using cooling provided by unit
or plant-wide refrigeration systems are often subject to downtime
unrelated to the process being controlled. The commenter stated
that the use of vent condensers as final control devices is
unlikely to consistently provide what the reference control
techniques call for and poses significant reliability problems.
2-122
-------
Therefore, the commenter does not believe vent condensers should
be an approved approach to provide emissions control under the
Hazardous Organic NESHAP (HON) - National Emission Standards for
Hazardous Air Pollutants. The commenter also stated that
refrigeration units depend on the use of chlorofluorocarbons
(CFC's), hydrochlorofluorocarbons (HCFC's), or ammonia (NH3) .
The CFC's and HCFC's will not be in great supply in the coming
years, and the use of pressurized NH3 may lead to highly
dangerous releases if NH3 lines rupture during operation.
Response; The promulgated rule specifies performance
standards rather than equipment standards or work practice
standards. Therefore, facilities may select any technology that
meets the performance standards. The Agency agrees with the
commenters that, where possible, the Agency should refrain from
specifying technology. However, while not specifying control
technologies to be used, the Agency has included monitoring
requirements and test methods for several types of control
technologies currently used to control emissions at marine tank
vessel loading operations. Requirements for developing and
seeking approval of alternative monitoring protocols and test
methods for alternative control technologies are also included in
the final rule. For carbon bed regeneration systems, the Agency
has specified that desorbed hydrocarbons must be vented to the
on-line carbon bed.
Comment; One commenter (22) stated that elevated,
open-flame flaring is the most cost effective pollution control
technology for many applications, such as large emergency
releases of combustible gases and vapors from petroleum refining
and chemical manufacturing plants. It is a technology that
certainly has a place in the overall plan to protect the
environment. However, the commenter stated that open flame
flaring is not a recommended technology for Marine Vapor Control
applications because it appears to be inconsistent with the
2-123
-------
intent of the proposed marine tank vessel loading and unloading
emission standards.
Response: The Agency does not prohibit the use of flares in
the promulgated standard. However, flares must be operated in
accordance with standard operating and maintenance requirements
specified in Section 63.11 of Subpart A of the General Provisions
and the promulgated standards (see Sections 63.563(b)(5),
63.564(f) and 63.565(e) of the final rule).
Comment: One commenter (123) stated that any attempt by
State or local authorities to regulate the normal operation of
properly functioning pressure/vacuum (P/V) valves is inconsistent
with U.S. Coast Guard authority and safety regulations. The
commenter endorsed the statement that such inconsistent State or
local regulations are "not likely to satisfy the objectives of
the Act and would therefore be disapproved by the Administrator."
However, some California districts have already had regulations
approved that, under their interpretation of the regulations,
prevent the normal operation of the P/V valves, even when no
cargo or ballast operations are taking place. The commenter has
recently paid settlements for Notices of Violations with one
district for this type emission for liftings that occurred while
at anchor or underway to the dock, even though the effect of the
liftings did not cause a violation of the 95 percent recovery
requirement.
A limitation of deliberate venting or purging of tanks,
except under special conditions such as the need to enter tanks
for emergency repairs, within the affected air shed might be
appropriate as separate rule making if there is evidence of
substantial impact from this practice.
Response; The Agency agrees with the commenter, and the
final rule is consistent with the Coast Guard requirements for
P/V vents.
2-124
-------
2.6.2 Performance Testing Requirements and Procedures
Comment; Four commenters (23, 31, 36, 73) stated that the
75-day advance written notification and submission of a test plan
proposed in 63.566(a) is too long and should be reduced to
30 days as set forth in the HON, 63.102(a). Another
commenter (12) stated that the 75 days in § 63.566(a) should be
changed to 60 days to be consistent with § 63.7(b) of the final
General Provisions to 40 CFR Part 63 published in the Federal
Register on March 16, 1994. These commenters also requested a
period of 60 days or 90 days to report the performance test
results rather than the proposed 30 days, consistent with
§ 63.7(g) of the General Provisions to 40 CFR Part 63, to allow
adequate time for report preparation by contractors, review, and
final compilation of the report by the owner/operator. One
commenter (31) also stated that § 63.566(a) should not require
the use of "registered mail" to submit the results. The General
Provisions to 40 CFR Part 63 do not require registered mail (also
the result of comments), which costs more and takes longer to
arrive than first-class mail. Express services (such as Federal
Express) or personal delivery should also be permitted; personal
delivery is common for documents delivered to a State permitting
authorities. In the future, facsimile transmission or other
similarly speedy methods might be used. These other methods are
permitted by the General Provisions at 40 CFR 63.1(a)(11),
59 FR 12431 (March 16, 1994). Section 63.566(a) should be
amended by deleting the reference to registered mail. The
General Provisions to 40 CFR Part 63 will then govern, thus
allowing a variety of delivery methods.
Response; The notification time, report submittal time, and
report submittal method(s) in the final rule are consistent with
the General Provisions except in cases where additional time is
allotted for report submittal or certain reports are not required
under this final rule (e.g., the deadline for the initial
notification requirements has been extended). The use of
2-125
-------
delivery services and devices other than registered mail has also
been incorporated into the final rule.
Comment: One cornmenter (32) stated that § 63.565 (c) (2) , as
proposed, requires performance testing to include the last
20 percent of loading of a tank or compartment. When loading
large, multicompartment vessels, the stability of the vessel
often requires partial and sequenced loading of the compartments,
such that filling the last 20 percent of a compartment may not be
a single, continuous operation. The commenter recommended that
the wording be changed as follows: (proposed revisions in BOLD)
(2) All testing shall be performed to include the last
20 percent of loading of a tank or compartment vessel.
Response: The Agency is requiring testing during the last
20 percent of compartment loading to ensure that control
equipment can handle peak concentrations. The Agency, therefore,
disagrees with the commenter's suggestion.
Comment; Two commenters (59, 60) stated that the maximum
loading pressure indicated in § 63.563(a)(3) of the proposed rule
will have an adverse effect on loading rates. The proposed rule
requires "0.8 times the highest setting of any vessel relief
valves." Maximum pressure settings assume a common tank system.
Parcel tankers have individual tanks with separate vent and
pressure/vacuum (P/V) valves for each tank. Operational
pressures should be based on the P/V setting for individual tanks
and not on the highest or lowest valve setting on the vessels.
Further, three commenters (69, 70, 78) stated that the
proposed procedures for pressure testing in § 63.565(a)(1)(i),
which would require that each product tank be pressurized with
dry air or inert gas to not less than 1.0 psig and not more than
the pressure of the lowest relief valve setting, is problematic
because many of the tank barges that would be affected by the
proposed rule are equipped with a 1.0 psig pressure/vacuum (P/V)
relief valve. These valves begin to relieve at a pressure below
the rated setting and are typically at their full rated relief
capacity at the stated relief pressure. Thus, the technical
2-126
-------
requirement to lift at 1 pound, but not more than 1 pound, will
be very difficult, if not impossible, for these barges to meet.
They recommended that this provision be modified to require a
test pressure of 80 percent of the lowest P/V valve setting.
This recommendation is consistent with U.S. Coast Guard
regulations (46 CFR Part 39), which require that barge vapor
recovery systems be designed and operated so that the pressure at
any location on the barge is 80 percent or less of the P/V valve
setting. A tightness test conducted at the reasonably
anticipated worst-case operational condition is more practical
than a test performed at an unanticipated condition. One
commenter (35) agreed and stated that any vapor-tight testing
should incorporate these standards, not create a new set of
standards in addition to them.
Response; The Agency agrees with intent of this comment.
Therefore Section 63.565(c)(1) of the final rule has been revised
to address the commenter's concerns. The pressure test requires
that each product tank be pressurized with dry air or inert gas
to not more than the pressure of the lowest pressure relief valve
setting. The U.S. Coast Guard has reviewed these changes and any
safety concerns with these requirements have been addressed.
2-127
-------
Response: The final rule incorporates additional
flexibility regarding the selection of monitoring parameters and
monitoring approaches by providing other options in
Sections 63.563 and 63.564 of the final regulation. The Agency
has provided additional options for monitoring parameters where
commenters submitted the necessary information to allow the
Agency to judge the suitability of a give parameter for this
source category. The Agency notes that a monitoring parameter
that is appropriate for another source category (e.g., the
synthetic organic chemical manufacturing industry for which the
hazardous organic NESHAP was written) may not be appropriate for
the marine tank vessel loading source category.
In addition to the additional choices for monitoring
parameters added to the final rule, owners and operators of
sources subject to the emissions standards may use alternative
monitoring parameters and monitoring approaches upon application
and approval by the Administrator per § 63.564 (j) of the
regulation.
Comment; Seven commenters (26, 31, 32, 34, 36, 71, 74)
stated that values for operating parameters that define
compliance should not be established based on the values from the
initial performance test. The initial performance test will
establish parameters that correspond to control device operation
in an as-new condition, which will likely exceed the standard,
and the procedures in the proposed rule allow for no variations
in performance based on normal variations during a single loading
operation and normal deterioration. One commenter noted, for
example, that carbon adsorbers operate over a range of
efficiencies during various stages of the loading operation and
are affected by variations in cargo concentrations; carbon bed
performance also degrades over time. For these reasons,
facilities are designed with extra capacity to assure continuous
compliance. Another commenter (26) cited an example from test
data in the docket that, if used as an initial compliance test,
would establish a VOC reduction requirement of 99.4+ percent,
2-130
-------
markedly higher than the 95 percent required by the standard.
Another commenter (74) noted that the rule as proposed calls for
establishing an average comprised of high, low, and intermediate
values from the performance test; one of these values would be
above the average and would, therefore, be out of compliance.
Two of these commenters pointed out that the final HON rule
includes "Transfer operations - reporting and recordkeeping for
performance tests and notification of compliance status," in
63.126(c) provision for "each parameter monitored ... the owner
or operator shall establish a range for the parameter that
indicates proper operation of the control device." The EPA
should include the HON provision in the final rule for marine
tank vessel operations and delete the need to maintain the same
operating parameters as attained in the initial performance test.
The commenters recommended that the performance test be performed
over a range of product loading rates and that the test data be
used to establish a parameter value or range of values that
correspond to compliance levels.
One commenter (26) questioned what exactly constitutes the
"baseline VOC outlet concentration" (§ 63.653(b)(3)(i)(A)); for
example, for a given set of test data, would it be the low value,
the high value, some type of average value, or a time dependent
value? Clearly, the inherent variability of the outlet VOC
concentration means it cannot be used as a reliable indicator of
compliance with a standard that calls for 95 weight percent
reduction over a major portion of the loading operation.
Specific recommendations for revisions to the rule:
1. Delete the last sentence at 63.564(b)(2) and add words
allowing a facility to determine the "VOC outlet concentration"
from an examination of the performance test. A facility should
be allowed to conduct performance tests at varied loading rates
and determine from this data the VOC concentration at the loading
rate representing 95 percent efficiency. Any future VOC
concentration above this value would represent noncompliance,
while a value below it would represent compliance (Commenter 74).
2-131
-------
2.6.3 Establishing Operating Parameters
Comment; Many commenters felt that the proposed regulation
does not provide enough flexibility for choosing parameters to
use for parameter monitoring and test procedures to establish
limits for these parameters that determine compliance. Two
commenters (44, 68) stated that the rule should be flexible to
allow operators to propose alternative monitoring schemes. One
commenter (68) noted that §§ 63.563(b)(3)(v) and 63.564(f)
require continuous monitoring of temperature and specific gravity
of absorber systems. The commenter does not believe that these
are necessarily the appropriate parameters for all absorber
systems or that it is always necessary to continuously monitor
these parameters. One commenter (44) stated that
paragraph 63.564 (f) should specify scrubber liquid flow rate as
an acceptable parameter to monitor instead of liquid temperature
and specific gravity and that 63.564 (f) be rewritten to allow
alternative monitoring parameters. More monitoring options would
eliminate the need for additional approval steps otherwise
required by 63.564(g).
Six commenters (04, 25, 34, 56, 73, 78) do not believe that
there is any justification for mandating the use of CEM's
[§ 63.563(b)(3)(i)] to monitor carbon adsorber performance by
continuous measurement of VOC concentrations at the exhaust of
each unit. The requirement to compute VOC concentrations every
5 minutes is unnecessary to demonstrate compliance and is not
consistent with the monitoring intervals for other control
devices. As allowed in the HON, CEM's should be a monitoring
option, but parameter monitoring should also be acceptable for
demonstrating compliance. The source should be allowed to
establish appropriate parameters for monitoring carbon unit
performance during the performance test. These parameters would
then be used to ascertain compliance. One option is a monitor of
vacuum on the carbon bed in the recovery cycle (i.e., the vacuum
would certify that vapors are being extracted and recycled from
the carbon bed during the recovery cycle following adsorption).
2-128
-------
Conversely, one commenter (79) noted that the control device
parameters and monitoring requirements set forth in Subpart Y are
virtually the same as required by the NJDEP. These parameters
include (but are not limited to): (1) installing, calibrating,
maintaining and operating a continuous flow indicator to record
the vent flow, or (2) a continuous emission monitor system (CEM)
to monitor VOC concentrations at the exhaust of a carbon adsorber
unit or the exhaust stream temperature on a refrigerated
condenser unit. Based on the commenter's experience, these
parameters are appropriate and enforceable. Additionally, the
records that are to be maintained and the requirement that they
be maintained for a minimum of 5 years are the same requirements
as set forth by the NJDEP.
One comtnenter (22) agreed with the Agency that continuously
monitoring an important process parameter of vapor processing
devices may provide some assurance that the device is operating
within the proposed emission standard. However, monitoring any
single operating parameter (e.g. vent temperature, vent
hydrocarbon concentration, firebox temperature, etc.), while
useful as a guide to warn of potential problems and to gauge
efficient operation, is not sufficient to assure compliance with
the proposed emission standard. Failure to attain the specified
value of the monitored process parameter would be cause for
investigation, but a violation of the emission standard could not
be determined without further data.
One commenter (38) stated that boilers and heaters should be
required to establish an appropriate destruction and removal
efficiency (DRE) through initial performance testing, regardless
of the unit's heat duty. Testing should establish firebox exit
temperature at 1200°F or greater and carbon monoxide (CO)
concentration in the outlet stack gas at under 100 ppm(v), on a
dry basis. Operational parameter monitoring should include
firebox exit temperature and oxygen (02) and CO stack gas
concentrations for the entire time of operation of the process.
2-129
-------
2. To allow loading rates, etc., to be varied during the
performance test, the EPA should delete the words "that meets the
emission limits" from § 63.564(c)(2) and from any other section
that uses those or similar words. (Commenter 31)
3. Add definitions of "baseline outlet VOC" and "baseline
temperature" to 63.561 as follows:
"Baseline Outlet VOC Concentration" means the VOC
concentration at the outlet of a control device that
indicates the required VOC or HAP removal is being achieved.
"Baseline Temperature" means the temperature that indicates
that a control system is achieving the required VOC or HAP
removal. This temperature may be a maximum, a minimum, or a
range as appropriate (Commenter 71).
4. Delete the language in Section 63.564(c)(2) that
requires operating parameters to remain the same until a new
performance test is conducted (Commenter 32).
Four commenters (32, 38, 34, 41) saw no reason for an upper
limit on the temperature of a combustion device as required in
Section 63.563(b)(3)(ii) because higher combustion temperature
indicates increased control efficiency. The commenters stated
that EPA should change the requirement to operate combustion
devices within +/-5.6°C to a requirement for incinerators to
achieve a minimum operating temperature and residence time when
captured vapors are being combusted by the incinerator.
The commenters recommend the following wording for
Section 63.563(b)(3)(ii):
(ii) Thermal or catalytic incinerators (deleted
"Combustion device, except flare"). For sources
complying with Section 63.562(c), (d), or (g) through
use of a thermal or catalytic incinerator, the control
device shall achieve at least a 98 percent destruction
efficiency during the initial performance test. In
addition, the owner or operator shall establish as an
operating parameter the baseline temperature using the
procedures described in Section 63.565(c). Operation
of the affected source more than 5 percent (Fahrenheit
scale) below the baseline temperature, as monitored in
Section 63.564(c)(1), when captured vapors are being
incinerated (deleted "in deviation of the baseline
2-132
-------
temperature, as monitored in Section 63.564(c)(1), in
excess of +/-5.6°C (10°F)"), shall constitute
noncompliance with the standard.
Response; The Agency continues to believe that parametric
monitoring can be used to determine compliance if sufficient test
data are available to establish the relationship between control
performance and the associated parameters. The Agency carefully
reviewed all the comments and determined that the docket contains
no test data establishing the relationship between marine tank
vessel loading emissions control and parametric monitoring.
Because of the intermittent nature of marine tank vessel loading
operations (i.e., wide variation in parameter readings is
systemic for any process that cycles on and off routinely), the
Agency has decided not to use parametric monitoring to determine
compliance.
The Agency has eliminated the requirements directly linking
parametric monitoring for marine tank vessel loading operations
with compliance. Instead, the Agency has included a requirement
for an operation and maintenance (0 & M) plan to ensure proper
operation of the air pollution control equipment and monitoring
equipment. The 0 & M plan contains an inspection schedule for
each component of the control and monitoring equipment. The
"compliance" language that appeared in § 63.563 of the proposed
rule has been removed. In its place, the final rule contains
provisions that require an unscheduled inspection and corrective
actions when continuously measured operating parameters exceed
the applicable baseline parameters established during a
performance test.
Comment; Two commenters (25, 41) stated that there is no
justification for a lower limit on the temperature of a
condenser. Therefore, the EPA should remove this end of the
limit because lower condenser temperature indicates increased
control efficiency.
Response: The EPA agrees with this comment. The final
rule, therefore, does not require the establishment or monitoring
2-133
-------
of lower temperature limit when a condenser is used to control
emissions from marine tank vessel loading operations; an upper
limit only is established for the baseline temperature .
2-134
-------
2.6.4 Monitoring Requirements
Comment: One commenter (38) stated that average parameter
values should be determined hourly and based on data gathered
every 15 minutes. Compliance should be based on hourly averages.
Response; The Agency agrees with the commenter and has
incorporated 15-minute data collection requirements in the final
rule where applicable. Hourly averages or averages over a cycle
(based on the same cycle as the performance test) are calculated
and 3-hour or 3-cycle averages are then calculated for comparison
to the baseline operation parameter.
Comment; One commenter (20) inquired how minimum operating
temperatures can be established for boilers exempted from the
performance test requirements.
Response: In the final rule, boilers with design heat input
capacity of 44 MW or less where the vent stream is the primary
fuel, boilers with design heat input capacity of 44 MW or
greater, and boilers subject to 40 CFR Part 266, Subpart H are
exempted from the requirements to establish baseline operating
parameters for monitoring purposes.
Comment; Two comtnenters (06, 46) stated that EPA should
require an annual or biannual source test to confirm the
compliance status of control devices. One commenter (06)
requested such scheduled testing for vapor combustion equipment,
especially on larger incineration devices at marine tank vessel
loading operations subject to the RACT requirements. Currently,
§ 63.564(c)(2) requires a performance test only when the
operating temperature is established or changed. Variations in
burner performance can cause mixing problems that lead to lower
destruction efficiencies that may not show up on a temperature
monitor. One commenter (46) stated that, as proposed,
Section 63.565(b)(3) is unclear on the scheduling requirement for
marine tank vessel performance tests and suggested it be revised
as follows:
A performance test shall be conducted annually and
properly documented for this pressure measurement by,
2-135
-------
recording the pressure every 5 minutes while the test
marine tank vessel is being loaded, and recording the
highest instantaneous pressure that occurs during the
test loading cycle.
Response; The Agency disagrees with the commenter regarding
the need for annual performance testing. The commenter did not
provide cost/benefit information supporting the request for
additional testing requirements. The Agency believes that the
ability to show compliance with the standards once a year during
an annual compliance test do not justify the costs associated
with annual tests (typically $11,000 per facility in labor costs
alone). Under the final regulation, after an initial performance
test, either CEM's or parametric monitoring of the control system
is required to indicate on-going compliance with the standards.
Further, an operation and maintenance plan developed by the
owners or operators of the affected source to assure proper
operation of the control systems must be followed. Although
nothing in the final standards precludes an owner or operator
from conducting annual or periodic performance tests in addition
to the (required) initial performance test, the Agency does not
believe that the costs of requiring annual tests are warranted
given the monitoring and compliance provisions (outlined above)
contained in the final rule.
Comment: Two commenters (32, 71) noted that
Section 63.564(c)(1) of the proposed regulations requires monthly
accuracy checks of the temperature monitor with a monitor
traceable to the National Institute of Standards and Technology.
One commenter (32) stated that annual checks are sufficient for
the following reasons:
1. Temperature measurement instrumentation is a mature and
proven technology;
2. Experience shows that temperature measurement is
reliable;
3. Calibration drift of temperature monitors is not
significant;
2-136
-------
4. Normal failure mode is burnout of the sensing element,
which results in loss of temperature signal and would be promptly
repaired;
5. No benefit is gained by monthly calibration; and
6. Monthly reference checks would jeopardize safe work
practice because of need to access heights on the side of the
incinerator stack, potentially during inclement weather.
Also, the language in Section 63.564 (c) (1) should reference
thermal or catalytic incinerators, rather than combustion
devices.
Two commenters (71, 88) recommended replacing the
requirement for monthly thermocouple checks with a requirement
for dual thermocouples, which are to be repaired within 5 days if
the temperature difference between them at operating conditions
exceeds an amount established during the performance test. One
of these commenters stated that the quality assurance provisions
presented in the proposed rule were unnecessary and would, in
fact, harm the devices, reduce the accuracy of the device, and
could present a safety risk.
Response; The Agency agrees with the commenter. The final
rule requires that owners or operators monitor and record
continuously the presence of the flare pilot flame. The flare
must meet the criteria specified in Section 63.11 of the General
Provisions. Annual (instead of monthly) checks of the accuracy
of thermocouples are required in the final rule.
Comment; One commenter (71) noted that proposed 40 CFR
Part 64 will impose enhanced monitoring requirements on
applicable RACT rules. However, the proposed Part 64 would not
apply to MACT rules because each MACT rule will consider required
monitoring enhancements during its development. The proposed
Subpart Y has done that, but because it also implements RACT,
there may be confusion about the applicability of Part 64. The
Agency should clarify, preferably in 63.560 but possibly in the
preamble, that Part 64 requirements do not apply to MACT or RACT
facilities affected by Subpart Y.
2-137
-------
Response: The Agency has revised the applicability of
enhanced monitoring requirements as part of its thorough revision
of the monitoring requirements and compliance provisions of the
final rule. The Agency agrees that the requirements of 40 CFR
Part 64 should not apply to sources only affected by
Section 183(f) of the Act.
Comment; One commenter (20) asked why the proposed VOC
monitoring data for carbon adsorbers has to be corrected to
3 percent oxygen and 0 percent water? Another commenter (22)
stated the practice of correcting the vent stream to a given
oxygen concentration is done for combustion systems to compensate
for the amount of excess air drawn into the device by the
combustion process. This correction is done for combustion
equipment in-order to "place" each device on an equal basis for
comparison. The commenter does not believe it makes sense to
correct the vent from any noncombustion system (recovery systems)
to an arbitrary oxygen concentration.
Response; The final rule requires the use of Method 25A for
these measurements. Method 25A does not require correction to
3 percent oxygen and 0 percent water.
2-138
-------
2.6.5 Emission Estimating Procedures
Comment: Two coiranenters (06, 38) noted that EPA is
proposing to allow either actual source test data or API emission
factors to calculate compliance with the 93 percent overall MACT
control level. Actual emissions data should be the preferred
method for determining emission factors, especially with multi-
component liquids like crude oil. It should be clarified in the
regulation and the preamble that actual test data acquired from
EPA-approved test methods must be used when available and that
API factors are to be used only when actual test data is not
available.
Response: The Agency agrees with these commenters that
actual emissions data are preferable to the use of emissions
factors. Clarification has been included in the final rule
stating that actual emissions data collected using EPA test
methods will be used unless shown to misrepresent the emission
stream/control device combination. The rule does not require
emission testing to determine emission factors because no one has
demonstrated that the improvement in accuracy merits the high
costs associated with testing the vessels. Because of the
complex emission characteristics of this source category, the
Agency has provided sources with the flexibility to select the
emission estimation factors and emissions estimates to be used at
the source provided that these emission estimation factors and
emission estimates are based on test data or on generally
accepted industry standards of measurement or estimating
techniques for maximum operating conditions at the source.
Comment: Five commenters (32, 34, 37, 39, 79} objected to
the requirement to determine aggregate HAP emissions from loading
on a calendar quarter basis. The commenters believe that the
determination of compliance should be based on a calendar year
efficiency requirement, with quarterly reporting that will allow
the permit authority to monitor a facility's performance. The
commenters stated that EPA did not take into account seasonal
product changes, variations in crude and product slates, and
2-139
-------
turnarounds that can impact these facilities. The coiranenters
stated that an annual average period is necessary to provide
operational flexibility and noted that quarterly reporting during
the calendar year would provide the permit authority with the
necessary information to monitor a facility's performance.
One commenter (32) stated that compliance should be based on
annual averaging of HAP emissions and that a baseline period of
either 2 or 5 years prior to implementation of controls should be
established. For purposes of determining compliance with 40 CFR
Part 63, Subpart Y, the commenter recommended that
Section 63.565(d)(l), (d)(2), and (d)(3) be changed as follows:
(1) The average amount loaded in the baseline years shall
be multiplied by the appropriate emission factors
and percent HAP to determine [deleted "aggregate"]
initial HAP emissions from loading [deleted "per
calendar quarter"] during the baseline years. The
equation to be used is as follows:
HT = ZQjX FjX Pj (i = 1 to m)
where:
HT = Average total HAP emissions, [deleted
"Mg/quarter"] Mg/year
Qi = Average quantity of commodity loaded (per
vessel type) in the baseline years, liters
Fj = Average emission factor in the baseline years,
Mg/liter
P, = Average percent HAP in the baseline years
m = number of combinations of commodities and
vessel types loaded during the baseline years
(2) The equation to calculate the quantity of HAP emissions
controlled is as follows:
HC = Qj x Fj x Pj x CE (j = 1 to n)
2-140
-------
where:
HC = Quantity of HAP emissions controlled, [deleted
"Mg/quarter "] Mg/year (rolling 1 2-month
period)
Qj = Quantity of commodity loaded during the year
(rolling 12-month period) (per vessel type)
while connected to control device, liters
Fj = Emission Factor, Mg/liter
Pj = Percent HAP
CE *= Removal/destruction efficiency of control
device, as calculated in Section 63.565(c)(7)
n = Number of combinations of commodities and
vessel types loaded during the year (rolling
12-month period)
(3) The HAP control efficiency of the source during the
year (rolling 12 month period) shall be calculated by
dividing the controlled quantity of HAP by the [deleted
"estimated quantity of uncontrolled emissions"] HAP
emissions during the baseline year as follows:
Hp = Source HAP control efficiency, percent
Hc = Quantity of HAP emissions controlled, [deleted
•Mg/quarter"] Mg/year (rolling 1 2-month
period)
HT = Total HAP emissions, [deleted "Mg/quarter"]
Mg/baseline year
For periodic reporting on a quarterly basis, this
commenter suggested that EPA retain the language proposed in
Section 63.565(d)(1), (2), and (3) but clarify that compliance is
determined on an annual basis, not a quarterly basis.
Response: The Agency agrees with the commenters that the
compliance determination should be based on an annual basis.
However, the change in the format of the final standards
(i.e., no control for HAP having vapor pressures less than 1.5
psia, 97 percent control for HAP having vapor pressure of 1.5
2-141
-------
1.0 percent resolution generate quite acceptable results. Having
to purchase new instruments to meet a tighter specification is an
unnecessary and very costly burden. A 1.0 percent resolution
should more than meet the requirements of the system.
Response: The Agency agrees with the commenter. The final
rule refers to Performance Specification 8 for CEM resolution
requirements. This performance specification was finalized on
December 15, 1994 (59 FR 64593).
Comment; One commenter (32) questioned why § 63.564(c)(2)
references the test methods in § 60.614 to determine the removal
efficiency of devices, while the balance of § 63.565 references
the test methods in § 63.564 and in Appendix A of Part 60. In
addition, the commenter stated that it would be more clear to
refer to Section 63.565 for specific test methods and procedures
for this subpart, rather than referring to Appendix A, where
there are many test methods and procedures. This reference to
Section 63.565 for test methods and procedures should be
incorporated into Section 63.564(b)(2), (c)(2), (d)(2), and
(f)(2). The commenter requested that wording in
Section 63.564(c)(2) be changed as follows:
(2) Using the test methods and procedures in
Section 63.565, the owner or operator shall
determine the destruction efficiency of the
combustion device....
Response; The references in the test methods have been
revised in the final rule to incorporate the commenters
recommended approach.
Comment; One commenter (80) was concerned that Method 25A
will not be able to quantify VOC emissions; it quantifies
emissions of TOC. The commenter stated that the distinction is
very important when testing thermal oxidation systems because the
USCG requires the inlet air to vapor stream ratio to exceed
1.5 times the upper explosive limit (for safety reasons). This
ratio is normally accomplished using natural gas injection. The
commenter recommended using Method 25 to speciate methane or
2-144
-------
using a carbon scrubber with Method 25A to allow the
determination of VOC by subtraction.
Response; The Agency agrees with the commenter. The final
rule makes this distinction in applying the two methods.
Comment; One commenter (32) recommended the following
changes in Section 63.565(c)(5): (revisions in bold)
The VOC mass [deleted "emitted"] at the [deleted "inlet
and outlet"] sampling points of the combustion or
recovery device during each testing interval shall be
calculated as follows ....
In addition, the meaning of Cvoc should be changed to:
Cvoc = VOC concentration (as measured) at the [deleted
"exhaust vent"] sampling points of the combustion or
recovery device, parts per million by volume (ppmv),
dry basis.
In Section 63.565(c)(6), commenter (32) requested the
wording be changed to:
The VOC mass emission rates at sampling points [deleted
"the inlet and outlet"] of the combustion or recovery
[deleted "or destruction"] device shall be calculated
as follows ....
where:
Eif E0 = mass flow rate of VOC at the sampling points
[deleted "inlet (i) and outlet (o)"] of the
combustion or recovery [deleted "or
destruction"] device, kg/hr
My, MOJ = mass of VOC at the sampling points [deleted
"inlet (i) or outlet (o)"] during testing
interval j, kg.
In Section 63.565(c) (7), commenter (32) requested the wording be
changed to:
Where Method 25 or 25A is used to measure the percent
reduction in VOC, the percent reduction across the
combustion [deleted "destruction"] or recovery device shall
be calculated as follows ....
where:
= mass emitted or mass flow rate of VOC at the
incoming sampling point [deleted "the inlet"] to
the combustion [deleted "destruction"] or recovery
device as calculated under paragraph (c)(6) of this
section, kilogram/hour.
2-145
-------
psia or more) has made the commenters' suggested approach a moot
point.
Comment: Four commenters (4, 31, 32, 71) requested that
revisions to the referenced control methods be used for analyzing
VOC concentrations and mass flows. One commenter (32) noted an
inconsistency between § 63.565(c)(4) and the preamble
(page 59-stating that Method 25 is to be used for initial
performance tests of combustion devices) and requested that
§ 63.565(c)(4) be revised as follows: (revisions in bold)
The average VOC concentration upstream and downstream
of the control device in the vent shall be determined
using Method 25 or Method 25A of Appendix A of Part 60
of this chapter for combustion or recovery devices,
respectively.
Three commenters (04, 31, 71) stated that Methods 25 and 25B
should also be acceptable for determining the average VOC
concentration upstream and downstream of the control device. One
commenter (31) further noted that Methods 18, 25, 25A, and 25B
are available for analyzing VOC's and requested that all of these
test methods be acceptable, in particular Method 18 because it
can be used for inorganics such as Hcl.
Response: The perceived inconsistency between the preamble
and proposed rule has been corrected. The final rule contains
provisions to use either Method 25 or 25A to measure the percent
reduction in VOC. Additionally, the final rule contains
provisions that allow owners or operators who desire to use
alternative test methods to apply for approval to use such
methods.
Comment: One commenter (20) questioned why the proposed
test methods are so different from those required under
subparts XX and R of Part 60, and cites the following examples:
1. Emission standards for vapor control systems could be
established for specific commodities in units of milligrams of
HAP (or VOC) per liter of product transferred.
2. A minimum pressure relief setting determined safe by the
Coast Guard could be established, and annual pressure/vacuum
2-142
-------
tests could be required for marine tank vessel loading
operations.
3. A limit on the back pressure of the vapor control
systems could be established based on the minimum pressure relief
setting for marine tank vessels.
Response: The Agency lacks the requisite test data to
establish a mass rate limit as recommended by the commenter, and
the information submitted by commenters is insufficient to allow
the Agency to determine such a rate. In regards to the minimum
pressure relief setting, the Agency is satisfied that the
promulgated rule, developed to reflect current Coast Guard
standards, will ensure good capture of displaced vapors.
Comment: One commenter (31) thought EPA should allow
alternative test methods with prior approval by the permitting
authority, without the burdensome procedures of Method 301. The
commenter stated that Method 301 is so excessively complicated
and time-consuming that new methods could not be approved
expeditiously, if at all. Section 63.565(c)(8) should be revised
as follows: (revisions in bold)
(8) Use of methods [delete "other than Method 25 or
25A"] not specified in this Subpart shall be validated
pursuant to Method 301 of Appendix A or Part 63 of this
chapter, or shall be approved by the Administrator.
Response; The final regulation is consistent with the final
General Provisions to 40 CFR Part 63 regarding the use of
alternative test methods. Section 63.565(m) of the final rule
refers interested parties to the specific requirements regarding
alternative test procedures. These requirements allow the use of
alternative test methods after validation with Method 301 or upon
approval by the Administrator.
Comment; One commenter (31) stated that the 0.5 percent
resolution requirement on instrument spans is excessively
restrictive and prohibits the use of existing acceptable
instrumentation. The commenter recommended changing this
requirement to a more reasonable level of 1.0 percent. This
commenter stated that many instruments used today with a
2-143
-------
E0 = mass emitted or mass flow rate of VOC at the
outgoing sampling point [deleted "outlet"] of the
combustion [deleted "destruction"] or recovery
device, as calculated under paragraph (c)(6) of
this section, kilogram/hour.
Response: The Agency agrees that the use of "emitted" in
this case could be confusing and has, therefore, made changes in
the final rule similar to those suggested by the commenter. The
Agency does not agree that the use of "inlet" and "outlet" is
confusing and has not changed the use of these terms in the final
rule. The Agency has also referred to both combustion and
recovery devices where appropriate.
Comment; One commenter (54) stated that at least one of
their members currently uses EPA's Tank 2 program to estimate
emissions and would like to continue doing so. Although the
regulations state that "[a]n owner or operator may use a
procedure different from API 2514A upon approval from the
Administrator," obtaining special approvals can be very time-
consuming, labor-intensive, and frustrating. One commenter (54)
recommended that other generally-accepted alternative means of
estimating emissions be included in the regulation itself,
obviating the need for Agency approval.
Two commenters (31, 34) stated EPA should use AP-42 instead
of API publication 2514A as the reference protocol for estimating
emissions under § 63.565(d) for the loading of materials other
than crude oil and gasoline. The commenters stated the API
publication 2514A document is used solely for estimating
emissions from loading of crude oil and gasoline. It may not be
suitable for estimating emissions of other HAP. A more
appropriate calculation method is the AP-42 procedures for
Transportation and Marketing of Petroleum Liquids found in
Section 4.4. These procedures are used extensively by the
chemical industry and regulatory agencies for estimating
emissions for air permitting, air dispersion analysis, and
Superfund Amendments and Reauthorization Act (Section 313,
Reporting). In addition, the API publication 2514A, Atmospheric
2-146
-------
Emissions for Marine Vessel Transfer Operations, is used as a
reference in AP-42.
Additionally, it appears to commenter (31) that EPA's
reference to the API publication may be an error. In the
preamble to the proposed regulation on page 25018, Col. 2, EPA
states that:
Regarding the emission estimation procedures to be
followed in determining compliance with the proposed
standards, the Agency is proposing that facilities use
either actual test data or AP-42 emissions factors to
identify emissions from various commodities and streams
loaded.
Therefore, it appeared to the commenter (31) that EPA intended to
make AP-42 the reference method.
On the other hand, three commenters (31, 34, 72) support
EPA's proposal to allow other methods with prior approval. In
many instances, the AP-42 factors will greatly overestimate
emissions. One commenter (31) has sometimes identified other
methods that produce more accurate results; for example, State
regulators have been willing to accept these alternative methods
once their accuracy has been substantiated. One commenter (41)
stated AP-42 only lists VOC factors for loading crude and
gasoline into marine tank vessels and no HAP. If the EPA wants
to use such factors, a larger set of more accurate factors need
to be developed.
One commenter (72) stated that the provision would also
allow emission control techniques such as "short loading" of a
marine tank vessel to prevent expulsion of the highly
concentrated vapors that form just above the liquid surface in a
vessel's storage area. In that situation, proposed 63.565(d)
would allow owners to do source testing to demonstrate the actual
emission factor for loading a specific product into a specific
vessel type when the vessel's tanks are only filled to a
specified, measured, and recorded level. Details concerning the
required measurements and documentation would be specified by the
2-147
-------
EPA as part of the approval process for the specific emission
factor.
For all these reasons, the commenters recommend changing the
relevant portion of Section 63.565(d) as follows:
Emission estimates and emission factors shall be based
on test data or emission estimation procedures found in
Compilation of Air Pollution Emission Factors (Fourth
Edition)(AP-42). Section 4.4 Transportation and
Marketing of Petroleum Liquids.
Response: In the proposed rule, the Agency selected API
Publication 2514A as the reference protocol for estimating
emissions of commodities. Provisions were also proposed to allow
owners or operators of affected sources to use alternative
methods, provided that the use of these alternative methods is
approved by the Administrator in advance. The Agency agrees with
commenters that other methods of estimating emissions are
routinely used, and has changed the emissions estimation
requirements in the final rule. Emissions estimates and emission
factors used by owners or operators of sources required to
calculate an estimate of their HAP emissions must be based on
test data or on generally accepted industry practices of
measurement or estimating techniques for maximum operating
conditions at the source.
The Agency has not been persuaded by the commenter's
assertion that "short loading" is an equivalent method of
emissions reduction. Although the emissions appear to be reduced
the first time the vessel is short-loaded, this determination
does not account for the large volume of concentrated VOC
(i.e., additional emissions) in the cargo space that are
eventually emitted.
Comment: One commenter (34) stated that toluene is given a
tanker emission factor of 0.40 lb/1,000 gallons of product
loaded; in contrast, xylene is given a factor of 0.000075 lb/
1,000 gallons and 0.00024 is given for "xylenes" in the same
table. The toluene emission factor is, therefore, 5,300 times as
high as the factor for xylene, even though toluene and xylene
2-148
-------
have relatively similar vapor pressures (1.0 and 0.3, psia,
respectively). The commenter stated that the emission factors in
the Technical Support Document and in the proposal merit further
review.
Response; The Agency disagrees with the commenter's
statement on the emissions factors used in estimating emissions
even though the requirements for estimating the emissions have
changed in the final rule (see previous comment). Information
received from commenter's was insufficient to develop alternative
emissions estimates. The format of the final standards
(incorporating a limit for HAP having a vapor pressure of less
than 1.5 psia) makes the commenter's concern a moot point since
the vapor pressure of both toluene and xylene are both below 1.5
psia. The HAP mentioned by the commenter would not be included
in emissions reduction calculations.
Comment: One commenter (80) stated that it is inappropriate
to assume that the average control efficiency of the abatement
device is representative of the reduction that will be achieved
for all HAP's. Carbon adsorption will achieve a higher control
efficiency for polar molecules with a high molecular weight.
Incineration of certain HAP's will give rise to secondary HAP
emissions (polycyclics, for example). Refrigeration has
different efficiencies for specific compounds. The HAP reduction
should be established during the initial compliance test for the
individual control unit using Method 25 or an alternate
speciating procedure.
Response: The MACT floor control efficiency was calculated
based on efficiencies for either VOC or benzene. The compliance
test determines the efficiency based on VOC control and is,
therefore, consistent with the approach used to calculate the
MACT floor. The Agency could have developed performance data for
individual HAP, at great expense, and then determined the MACT
floors based on these determinations. The commenters provided no
data indicating that there is any difference in these approaches
2-149
-------
or in the HAP emissions reductions required by the promulgated
rules.
2-150
-------
2.6.6 Emissions Averaging
Comment; One commenter (72) stated that terminals having
more than one loading dock should not be required to install
multiple vapor recovery systems and should be able to average
emissions between multiple docks to meet the emission reduction
requirements.
Response: The Agency agrees with the commenter. The final
rule requires an overall emissions reduction that may be achieved
by controlling all or partial loading of commodities and all or
some loading berths.
Comment: One commenter (81) noted that the proposed MACT
level of control for existing sources sets a "floor" at
93 percent and allows sources to over-control some products while
"excluding" other products from control, provided the facility
meets the overall control level of 93 percent. This approach is,
in essence, emissions averaging. The Bay Area has commented
extensively on a number of proposals under Section 112 that
provide some form of averaging or off-setting of emissions. They
have the same concerns about averaging in this proposal. If
facilities propose to "exclude" the loading of certain products
from control requirements, the exclusion should be restricted
according to the toxicity of the components. Carcinogenic
compounds that have a high-unit risk value should not be excluded
from control in favor of controlling the emissions of compounds
with a lower unit risk value. This inclusion is especially
critical for marine loading of products from the SOCMI industry.
The rule should specifically allow State or local agencies to
exclude the "averaging" portion of the rule without requiring
review under Subpart E.
Response; The format change of the standards (i.e.,
incorporating a limit for HAP having a vapor pressure less than
1.5 psia) has made much of the commenter's concern a moot point.
However, the Agency has provided emissions averaging approaches
in several areas of this final rule (e.g., VMT standards, marine
tank vessel loading operations collocated at petroleum
2-151
-------
refineries, etc.) to provide flexibility to sources in achieving
the overall emissions reductions that are required by the
standards. In keeping with the EPA's stated goal of increasing
flexibility in rulemakings, this decision has been made to
provide more opportunities to average. This optimizes the
opportunities for refiners to find cost-effective emission
reductions from overall facility operations on-site. Costs and
cost-effectiveness of controlling a particular kind of emission
point, such as marine tank vessel loading, will vary depending on
many site-specific factors. Emissions averaging allows the owner
and operator to find the optimal control strategy for their
particular situation.
Allowing emissions from marine tank vessel loading
operations, bulk gasoline terminal or pipeline breakout station
storage vessels, and bulk gasoline terminal loading racks to be
included in emissions averages will result in equivalent or
greater overall HAP emission reduction at each refinery. The
averaging provisions are structured such that "debits" generated
by not controlling an emission point that otherwise would require
control must be balanced by achieving extra control at other
refinery emission points covered by the NESHAP. The Agency has
not incorporated the commenter's request to insert language in
the regulation that would have provided State or local agencies
the authority to exclude emission averaging from terminals
operating in the respective State or locality. States or
localities may promulgate more stringent regulations that do not
provide for emissions averaging, but they cannot selectively
modify Federal regulations without independent promulgation of
their own regulations.
Comment; One commenter (104) suggested that EPA provide
alternative standards to RACT that utilize the Economic Incentive
Program rules. The commenter stated that these alternatives
would offer sources flexibility to trade emissions from marine
tank vessel loading operations with other VOC emissions
reductions programs undertaken by the source.
2-152
-------
Response: The Agency agrees with the commenter that sources
should have flexibility in meeting the overall RACT emissions
reduction standards. However, the Agency believes that the
commenter's suggested program should be carried out via a global
regulation promulgated under the Agency's Economic Incentive
Program.
2-153
-------
2.6.7 Exemptions from Performance Testing
Comment: One commenter (34) requested clarification
regarding the performance test exemption proposed in
63.563(b)(2)(i) and (ii) for boilers and process heaters with
input capacities above and below the 44 MW breakpoint. The EPA
should clarify in the final rule to what exactly the exemption is
intended to apply.
Response; The Agency believes that the section has
sufficient clarification. The commenter may not have noted the
phrase "and the vent stream is used as the primary fuel" in
Section (i).
Comment: A number of commenters stated that the rule should
not require performance testing that would be redundant to
testing required by current regulations. Facilities that have
recently conducted tests under current regulations should not
have to bear the cost of duplicative testing. Three commenters
(04, 25, 31) stated that in those instances where a source uses a
RCRA incinerator to control marine tank vessel loading emissions,
the EPA should exempt the incinerator from the performance
testing requirements of § 63.563(b)(2). One commenter (73) noted
that, where the proposed rule would apply to operations already
covered by the Benzene Transfer Operations NESHAP, there should
be a performance test exemption for facilities where no
additional controls will be necessary in order to comply with the
proposed rule. Two commenters (36, 31) recommended that, if an
affected terminal has already conducted a performance test for
its marine vapor control equipment to the satisfaction of the
State in which it is operating, the EPA should accept that test
and not require any additional testing.
Three commenters (31, 71, 78) stated that other units, such
as industrial furnaces and vapor balancing systems, should be
exempt from performance testing. Therefore, Section 63.563
(b)(2) should be amended as follows:
2-154
-------
(2) Performance test exemptions. An initial test is not
required in the following cases: (the proposed additions are
noted in BOLD)
(i) When a boiler or process heater with a design heat
input capacity of 44 Megawatts or less is used to comply with
63.562(c), (d), or (g) and the vent stream is used as the primary
fuel;
(ii) When a boiler or process heater with a design heat
input capacity of 44 Megawatts or greater is used to comply with
63.562(c), (d), or (g);
(iii) When a boiler, incinerator, or industrial furnace is
used to comply with §63.562(c), (d), or (g); and
(A) The owner or operator has been issued a final permit
under 40 CFR Part 270 and complies with the requirements of
40 CFR Part 266, Subpart H; or
(B) The owner or operator has certified compliance with the
interim status requirements of 40 CFR Part 266, Subpart H (note:
reference to demonstrating 99.99 percent destruction or recovery
efficiency has been deleted); or
(C) The owner or operator has been issued a final permit
under 40 CFR Part 270 and complies with the requirements of
40 CFR Part 264, Subpart O; or
(D) The owner or operator has certified compliance with the
interim status requirements of 40 CFR Part 265, Subpart O; or
(iv) When a vapor balancing system is used; or
(v) When emissions are recycled to a chemical manufacturing
process-unit.
The vapor balancing definition from the EON could be incorporated
into this rule as well. See 59 FR 19459 (April 22, 1994).
Response: The Agency agrees that duplicative testing should
be avoided. Sources that have completed recent testing where
that testing meets or exceeds the testing requirements of the
final rule from the performance test requirements and may,
subject to approval by the Administrator, use these data in
establishing the baseline parameters to be monitored.
2-155
-------
2.7 VAPOR TIGHTNESS REQUIREMENTS
Comment: Comments were received from several commenters
related to vessel tightness testing. Two commenters (79, 80)
supported EPA's proposal to require an annual tightness test for
vessels. Eight commenters (23, 31, 32, 33, 59, 60, 69, 78)
requested revisions to the requirements for testing vessel
tightness, maintaining records, and repairing leaks.
Six commenters (23, 31, 32, 33, 69, 78) stated that, because
ships are not defined as affected sources by the EPA, all the
burden of compliance and assuring compliance is borne by the
marine tank vessel loading terminal owner or operator. The
commenters noted that terminal operators have no authority over
vessel operators, and it is totally inappropriate for the
proposed rule to require the terminal to be responsible for
vessel equipment or documentation.
One commenter (138) stated that the leak test method would
be the only available alternative for their terminal to use.
One commenter (32) stated that there is no practical way for
the terminal operator to be obligated or take responsibility for
vessel testing. The commenter requested the wording of
Section 63.565(a) be changed as follows: (revisions in bold)
(a) When testing a vessel for vapor tightness to
comply with the marine tank vessel vapor-tightness
requirements of Section 63.563(a), the marine vessel
owner or operator (delete "of an affected source")
shall use one of the following methods:
Similarly, changes should be made to Section 63.565(a)(2) and (3)
as follows:
(2) Leak test. Each marine vessel owner or operator
loading at an affected facility complying with....
(3) Negative pressure loading. Each marine vessel
owner or operator loading at an affected facility
complying with....
This commenter also stated that additional leak test requirements
should not be applied to other portions of the vapor control
system, including any existing equipment that may be integrated
into an overall vapor control system.
2-156
-------
Four commenters (23, 33, 69, 78) stated that leak testing
ought to be incorporated into the vessel's annual U. S. Coast
Guard Inspection and its Mid-Period Inspection, which both
require a marine tank vessel to be sent to a shipyard. The
12-month vessel testing interval should be extended to coincide
with routine Coast Guard exams, and the procedures in
63.565(a)(2) should not be duplicative of any similar U.S. Coast
Guard procedures. Testing at this interval would appropriately
put the requirement for maintaining leak-free vessels on the
vessel owner. However, if a vessel were pressed into a service
for which it had not been tested, the provisions of
63.563(a)(2)(iii) would still be available for emergency use.
Further, a simple certification from the vessel owner or operator
should be sufficient documentation for the owner or operator of
the affected source, who should not have the onus of conducting
"vapor-tightness" testing of marine tank vessels. In fact,
enforcement of these procedures should lie in the proper domain
of the U.S. Coast Guard. Therefore, the provisions of 63.563
should be streamlined to require only a simple certification that
the vessel has passed the inspection and certification of the
U.S. Coast Guard.
One commenter (33) questioned whether the vessel can be
assured of receiving a copy of the documentation of the vapor
tightness test that can then, in turn, be presented to other
terminals at which the vessel may call in the next 12 months.
One commenter (23) stated that it would be more efficient for
"vapor tightness" certification to be maintained with the
vessel's documentation (i.e., U.S. Coast Guard Certificate of
Inspection or U.S. Coast Guard Tank Vessel Examination Letter)
rather than each terminal maintaining separate documentation.
The inspection and certification of vessels should be conducted
in accordance with standards developed by the U.S. Coast Guard.
To the extent that the certification and inspection of vessels
are inherent parts of the safety of loading tank vessels, the EPA
must consult with the Coast Guard to make the certification and
2-157
-------
inspection requirements of vessels consistent pursuant to
Section 183 (f) (2) of the Act. The commenter urged EPA to
coordinate more closely with the Coast Guard on developing the
specific procedures of inspection and certification of vessels.
Requiring each marine tank vessel loading operation to have
emission testing equipment and to enforce emission testing
requirement is both inefficient and improper. The specified
testing is complicated to perform and requires specialized
equipment and specially trained personnel. These requirements
would be a significant expense and an enormous burden to the
marine tank vessel loading operation owner or operator. This
testing would be used rarely because most vessels would have the
documentation and not need the testing. Further, the role of
testing should belong to government trained personnel such as the
Coast Guard.
Two commenters (59, 60) stated that the regulations do not
clearly define who shall conduct and certify the test for vapor
tightness documentation and certification. The reporting
requirements indicate a "witnessing inspector" of some
"affiliation" shall be included in the documentation. The
commenters believe the tests can be conducted by the vessel's
-personnel without the need for outside inspectors. Chemical
parcel tankers have numerous small tanks. It is not unusual for
a vessel to have as many as 58 independent tanks and systems.
Outside inspectors would put a considerable financial burden of
chemical parcel tankers because the tests would then have to be
done in port prior to operations. Although the rule allows for
testing during loading, due to the risk of nonvapor tightness and
additional record keeping requirements, it is doubtful terminals
will allow vessels to berth unless they can certify they are
vapor tight (or unless the terminal loads under vacuum).
Logistically, it is not unreasonable for it to take over 24 hours
to conduct tests as prescribed by 63.565(a)(1) on a vessel with
58 tanks. If these tests must be witnessed by a surveyor, in
port prior to loading, there will be a substantial cost to
2-158
-------
conduct these tests on chemical parcel tankers. One
commenter (59) stated that company representatives routinely
conduct tightness tests for vessel lines and hoses. Addition-
ally, for particular cargos (propylene oxide for example), tests
are conducted prior to loading and the pressure is monitored
throughout the voyage. Class regulations mandate tank tightness
tests every 5 years (at time of special survey). The types of
tank tightness tests conducted, in addition to the pressure test
stated in the regulations (without comparison to the pressure
drop formula), are water tests and use of an ultra sound gun to
find leaks. Testing has been conducted by company shore staff
and vessel officers without the "benefit" of a certifying
company. They are conducted as the opportunity presents itself.
Their motives for having vapor tight vessels are their own self-
interests. Many products they carry are water sensitive, and
their indemnity coverage, in the event of product contamination,
would be jeopardized if they had knowledge of a leak that was not
repaired. Self-certification has been found acceptable by the
USCG, the vessels' classification society, and their customers.
The commenters, therefore, requested that the proven practice of
self certification be permitted.
One commenter (23) viewed leak testing of foreign vessels as
a potential problem. Certification of equipment or material
condition is usually issued by the vessel's flag state either via
the flag state's regulatory agency or a classification society
such as Lloyds, ABS, etc. Whatever the means decided upon by the
EPA, in consultation with the U.S. Coast Guard, to certify
foreign vessels, it is imperative that U.S. flag vessels are not
penalized either financially or operationally by standards more
stringent than those required for foreign vessels.
Response: The Agency has revised the leak-test requirements
to provide additional flexibility to vessel owners or operators.
The final rule contains requirements for owners or operators to
conduct a pressure test, or leak test. The Agency agrees that
U.S. Coast Guard procedures should not be superseded or
2-159
-------
duplicated. Self-certification from the vessel owner or operator
that the vessel has passed U.S. Coast Guard inspection and
certification requirements has been incorporated into the final
rule. Vessel owners or operators are responsible for performing
the Method 21 tightness test and the vessel pressure test, and
the terminal owners or operators are responsible for the shore-
side control equipment and assurance of negative pressure system
operation (if applicable). The terminal is responsible for
maintaining records of vessels that dock and/or load.
In regards to comments that the vessel tightness
requirements would require significant expense or time, the
Agency notes that the vessel owners or operators have a minimum
of 3 years (4 years for vessels loaded at MACT facilities) to
initially obtain the proper certification and demonstrate
compliance. After the compliance date, this information must be
obtained annually. Based on the information received in the
comments, the Agency does not believe that this requirement poses
an unreasonable burden to vessel owners or operators.
Comment: One commenter (43) noted that many references are
made to specific responsibilities for the "owner or operator" in
§§ 63.562, 63.563, and 63.564. Industry consultants have
requested that each requirement be specifically assigned to
either the owner of the facility or vessel, or the operator of
the facility or vessel. For example, in 63.562(a), it should be
the owner's responsibility to equip the terminal with a vapor
collection system, and in 63.563 (a)(4), it should be the
operator's responsibility to obtain a copy of the ballasting
documentation. Due to the myriad of leasing and joint operating
agreements that exist in the maritime industry, enforcement would
be enhanced by further defining the party responsible for
compliance.
Response; The final rule clarifies the specific
responsibilities of owners and operators of facilities and owners
and operators of vessels. The Agency's use of the terms "owner
2-160
-------
or operator" is consistent with the use of the terms in the Act
and the General Provisions to Part 63.
Comment; Two commenters (32, 80) submitted perceived
deficiencies with the vapor tightness test procedures. One
comtnenter (80) concluded that the proposed pressure-test
procedure may bias the results towards compliance. The procedure
permits the use of air to pressurize the vessel during testing.
The presence of oxygen can result in vapor growth, and the test
results would indicate an absence of leaks when, in fact, leaks
are present. The use of inert gas to pressurize the cargo tank
would minimize the bias, provided the vessel is equipped with an
inert gas system or could be cost-effectively retrofitted with
one. If an inert gas system is not feasible, the volumetric
flowrate of the leak could be quantified by measuring the
flowrate of air that is required to maintain a specified
pressure. EPA would need to establish a standard for an
allowable leak rate at a given pressure. There should be no
cargo in the tank during testing, and the headspace temperature
should be monitored and recorded to allow for correction due to
thermal expansion. This method also has the benefit of providing
a means of calculating fugitive emissions during loading.
Fugitive emissions would be equal to the concentration times the
flowrate, integrated over the loading event.
One commenter (32) noted that vapor-pressure test methods
presented in Sections 63.565(a)(1)(i) through (vi) are
impractical to execute for the types of marine tank vessels that
call at the VMT. The pressure calculation method is inadequate
to indicate a vapor tight vessel. The commenter requested this
section of the proposed regulations be reworked to provide
guidelines or requirements that are reasonable and achievable for
pressure testing. In Section 63.565(a)(1)(i), the proposed
regulation requires use of dry air or inert gas for the pressure
test. Dry air is not suitable because the marine tank vessels
that call at the VMT are required to be inerted (i.e. the crude
oil cargo compartments are supplied with inert gas from the ship
2-161
-------
boiler exhaust so there is not a combustible vapor atmosphere in
the cargo compartments). Therefore, using air for a pressure
test is not allowable because it would present serious safety
hazards and is contrary to the design and operation of the
vessels. Inert gas from the ship boilers would be available and
suitable from a safety standpoint, but the gas will not be dry
because flue gas from the ship boilers typically passes through
an S02 scrubber and liquid deck seal before entering the cargo
compartments. The inert gas may be close to saturation with
water vapor. Any calculation method should allow for an inert
gas with varying amounts of moisture. In
Sections 63.565(a) (1) (iii) through (vi), the proposed regulation
gives the calculation method and pass/fail criteria for the
pressure test. However, this calculation and criteria could lead
to incorrect conclusions, such as a vapor-tight vessel would be
indicated as not vapor tight and in need of repair. Significant
weakness in the calculation method exists by not accounting for
the potential changes in temperature during the test, variation
of temperature within the cargo compartments, and the potential
for condensation of moisture from the vapor. This potential for
condensation would exist, even if a dry gas was used for
pressurization, because the initial vapor in the vessel will
likely include water vapor. The following hypothetical, but
realistic, scenario demonstrates the weakness of the proposed
calculation and criteria. In this scenario, a vapor-tight vessel
for the VMT significantly fails the vapor-tightness test when, in
fact, it is vapor tight.
Pj = 434.9 inches WC, (15.7 psia) - Initial test pressure,
inches WC
Pf = 430.7 inches WC, (15.55 psia) - Final test pressure,
inches WC, due to 5°F temperature change of vapor from 60°F to
55°F
P = Pj- Pf = 4.2 inches WC, (0.15 psia) - Change in
pressure, inches WC
2-162
-------
Pia = 15.7 psia - Initial pressure, pounds per square inch
absolute
L = 100,000 barrels per hour - maximum loading rate; (Actual
VMT conditions)
V = 1.100.000 barrels - total volume of vessel; (Typical for
a 165 MDWT ship at VMT)
PM = 0.861 (Pj) (LN) =1.23 inches WC - Max. allowable
pressure change, inches WC
According to the proposed regulations, if P > PM, the vessel
is not vapor tight and in need of repair. Therefore, a vapor-
tight vessel would substantially fail the pressure test as
proposed if the temperature changed just a few degrees. It is
reasonable to expect changes in vapor temperature during the
test. Also, one uniform vapor temperature throughout the vapor
of a ship cargo compartment does not exist. During ship-vapor
testing at the VMT, temperature variations of greater than 10°F
have been measured among different locations within a
compartment, at a single point in time. The pressure changes,
for example, would be more exaggerated if condensation of vapor
were considered. This scenario is by no means a worst-case
example, but it is a way of demonstrating the concern and need
for attention to this section of the proposed regulation.
Response: The Agency disagrees that temperature changes
and/or oxygen presence significantly affect the results of leak
testing, are impracticable for particular marine vessels, or
involve safety concerns. The commenter did not supply test data
documenting that temperature variances or oxygen presence
significantly affects pressure-test results. The Agency notes
that the pressure-leak test is not required by the rule but
rather offered as one of several alternatives for showing that a
vessel is vapor tight. Furthermore, alternative methods are
provided and others may be submitted to the Administrator for use
following the Administrator's approval. Regarding the use of
inert gas loading, the Agency notes that the final rule contains
provisions for leak testing of inert blanketed vessels.
2-163
-------
Comment; One commenter (71) noted that the leak detection
requirements of these sections are often redundant with equally
or more stringent requirements under Federal and State VOC and
HAP rules, and these requirements should be waived in such cases.
Recommendation: Add 63.563(f) as follows:
(f) Closed vent systems and control devices subject to
federally enforceable VOC or HAP fugitive emission monitoring
programs need not comply with 63.563(c) or (d) or 63.567(i).
Response; The compliance provisions found in Section
63.563(c) of the final rule address leak detection and repair
requirements for marine tank vessel loading operations. The
commenter did not submit sufficient information for the Agency to
determine that these requirements were redundant or unnecessary.
If the procedures currently employed at a given terminal meet or
exceed the requirements found in Subpart Y, then the owner or
operator of the affected source may apply to the Administrator
for an exemption from the leak detection and repair compliance
provisions.
Comment; Two commenters (34, 85) supported the requirement
in Section 63.563(a)(2)(iv) for measuring pressure at the
shore-vessel interface. One commenter (34) stated that the
second sentence of § 63.563(a)(2)(iv) of the proposed rule should
be revised to be consistent with U.S. Coast Guard requirements
for the location of pressure sensors as detailed in 33 CFR
Part 154, Subpart E, 154.814(h) or (j) that address requirements
for vessel vapor over pressure and vacuum protection. Shore side
location of pressure sensors must be allowed for documented
safety reasons and practical concerns.
Response; The final rule is consistent with the U.S. Coast
Guard requirements and has been reviewed by the Coast Guard to
address safety concerns.
Comment; Three commenters (32, 34, 38) requested changes to
Section 63.563 (a) (2) (ii). Two commenters cited additional
circumstances under which a vessel that is not vapor-tight should
2-164
-------
be allowed to be loaded. One corranenter disagreed with any
provision for loading such a vessel.
Two commenters (32, 34) pointed out that repairs that do not
require dry-docking a vessel and that are made at a marine-tank
vessel-loading terminal create greater emissions than if the
repairs were postponed. Emissions of purged material from the
immediate clearing of a compartment or pipe can be greater than
emissions resulting from delay of the repair. In this case, it
is more practical to load immediately and delay repair until
dry-docking, even though immediate repair may be technically
feasible. One commenter (34) stated that this provision is
consistent with the NSPS and NESHAP fugitive equipment leak
regulations. Another commenter (32) requested the language in
Section 63.563(a)(2)(ii) be changed as follows: (Revisions in
bold)
(ii) ... If the owner or operator of the vessel can
document that repair is technically infeasible without
clearing the compartment or piping system [delete "dry
docking the vessel], the owner or operator of the
affected source may load the marine vessel.
Conversely, one commenter (38) stated that § 63.563(a)(2) is
unclear with regard to the facility owner's responsibility of
whether to load or not load a vessel in need of repairs that are
technically infeasible without dry-docking the vessel. While an
alternative to loading such a vessel is available (i.e., loading
at negative pressure), the vessel needs to be repaired at some
point in time. The rule should state that loading a vessel
without control of emissions constitutes a violation.
Response: The Agency has revised the regulation to address
the commenter's concern about loading of vessels following the
detection of leak(s). The Agency strongly agrees with the
commenter that requirements for repair of equipment should not
generate more emissions than would be saved by the repair
operations. Regarding the comment concerning the future loading
of vessels failing the leak test, the final rule contains
requirements for the owner or operator of the terminal not to
2-165
-------
load the vessel again unless the marine tank vessel owner or
operator documents that the equipment responsible for the leak
has been repaired.
Comment: One commenter (85) supported the definition of a
vapor-tight marine tank vessel in Section 63.561 and added that
leaks found on vessels meeting the definition of vapor tightness
should be corrected in accordance with the standard; however,
they should not be considered a violation of the standard.
Response; The Agency agrees with the commenter. Under the
final rule, a violation of the standard does not occur if leaks
are detected and the cause for those leaks is repaired or if the
cause of the detected leaks requires the vessel to be dry-docked.
Comment; Five commenters (35, 60, 69, 70, 78) took issue
with the testing requirements for vessel vapor tightness. One
commenter (60) stated that the pressure drop formula (PM = 0.861
Pu L/V) favors vessels with large tanks. Chemical parcel tankers
with numerous small tanks are unfairly penalized because there is
very often a difference in pressure on a closed tank over time.
This pressure difference may or may not be a reflection of the
tank's vapor tightness (i.e., leak free). The sun and
atmospheric pressure will affect the test results (sometimes over
short periods of time). If Pu is very small (but not zero) and V
is small compared to L, it may be difficult to pass this test
even if the tank is in fact "vapor tight." The relationship is
with respect to the pressure at actual load rate; no relationship
exists between vapor tightness and tank size. In addition, two
commenters (69, 70) noted that the proposed equation for
calculating vapor tightness supports a lower pressure drop
tolerance at low loading rates, in effect discouraging a vessel
operator from establishing a lower rate. Because low loading
rates are safer than high rates, both from an engineering and a
human factors perspective, the commenters find the equation
perplexing. They recommended that the equation be modified so as
2-166
-------
not to penalize operators who elect to establish a lower loading
rate.
One commenter (60) noted that the regulations favor negative
pressure loading. By definition, a vessel is vapor tight if
loaded under negative pressure. Compliance and performance
testing is extensive for transfers not done under negative
pressure. The safety issues associated with these types of
transfers under closed conditions need to be taken into account.
Safety is an inherent concern when considering any control
technology alternatives.
Response: The equation in the regulation of concern to the
commenters was incorporated from the Benzene Transfer Operations
NESHAP (40 CFR Part 61 Subpart BB). The commenters taking
exception to this equation proposed no alternative equation or
specific revisions for the agency to consider; therefore, the
equation will be left unchanged.
Regarding safety issue concerns, again no specific
deficiencies were submitted that the Agency could consider. The
final rule has been developed with the full cooperation and
review of the U.S. Coast Guard in consideration of safety
concerns. The Agency is confident that the safety concerns
expressed during the development of these standards have been
addressed in these final standards and the Coast Guard standards
found in 33 CFR part 154 et al. and in 46 CFR part 30 et al.
2-167
-------
2.8 LEAK DETECTION AND REPAIR
Comment; Four commenters (32, 34, 38, 123) commented on the
leak detection and repair procedures in § 63.563(c)(2). Three
commenters (32, 38, 123} requested revisions or clarifications to
the definition of a leak. Two commenters (32, 34) requested
revisions to the scope of monitoring required upon leak
detection.
One commenter (32) agreed with the 10,000 ppmv threshold for
the definition of a leak and requested a clarification that a
reading of less than 10,000 ppmv require no further action. One
commenter (38) stated that the requirements of § 63.563(c)(2)
should be consistent with 40 CFR Part 63, paragraph 63.148(d) HON
requirements recently promulgated by the EPA, which define a leak
as greater than 500 ppmv above background levels. Conversely,
one commenter (123) stated that the 10,000 ppm standard is not by
itself a reliable indication of a leak and recommended adding to
the end of the last sentence of § 63.565(a)(2) "unless subsequent
flow measurement using a flow testing device capable of measuring
minimum gas flows of 1 SCFH (as air) indicates that flow from the
indicated leak is below detectable limits."
One commenter (123) stated that, at best, Method 21 provides
only an indication of a possible leak and that most OVA
indications of 10,000 ppm on ships and barges do not result in
detectable flow rates, even when using test rigs capable of
detecting 0.5 scfh. Organic vapor analyzer (OVA) detections have
exceeded 10,000 ppm for crude oil stains on deck. Defining a
10,000 ppm detection as an actionable leak without allowing the
operator to prove otherwise will make using these instruments for
preventative maintenance surveying prohibitive. This commenter
submitted results of surveys of ships and barges indicating that
of 31 occurrences where OVA indicated a leak based on ppm levels
of 8,000 or higher, only 4 (3.5 percent) of these indicated leaks
were found to have a measurable flow rate of 1.0 SCFH or higher.
Results of fugitive emissions testing of 5,035 fittings in
refinery service were also submitted. Forty-two showed OVA
2-168
-------
readings above 10,000 ppm, and only 5 of these 42 fittings
(11.9 percent) were found to have a flow rate of 1.0 scfh or
higher. Further, this standard requires an instrument such as
the OVA, which is not suited to marine use due to frequent and
complicated calibration requirements and lack of marine grade
construction and weatherproofing. This commenter's experience
has been that any measurable leak (>1 SCFH) will be detected with
a standard marine explosive gas detection meter, such as the MSA
Gascope, which is found on virtually all vessels for use in tank,
pump room, and void entry. A reading of 20 percent LEL on these
instruments has been found to be a good criteria for further
investigation and measurement for flow.
In general, this commenter's experience, as documented in
the analysis of the most recent 8 months of fugitive emissions
testing, has shown this emissions source to be virtually
insignificant with respect to marine vapor recovery. The highest
combined flow rate occurred on February 24, 1994, with a mass
flow rate of about 28 Ib/d, compared to 8.67 tons of recovered
vapor for that lightering. The data also shows that, while
0.8 percent of the over 5,000 fittings surveyed gave an
indication of over 10,000 ppm, only one-eighth of these had a
measurable flow rate, and even these few averaged a VOC emission
rate of only about 0.28 Ib/hr when detected.
Almost all measurable leaks seem to be from P/V valves,
which, by their design, offer no opportunity for preventive
maintenance such as tightening flanges or glands. Thus, P/V
valves can never be guaranteed to be gas tight at all times. The
commenter pointed out that when this type leak is discovered it
is usually the result of foreign matter that has built up on the
seating surfaces of the valve. Simply rotating the valve disk
almost always removes the foreign matter and allows the valve to
reseat.
One commenter (06) stated that the leak-monitoring
requirements of 63.563(c) are inadequate to assure continuous
compliance with the required control efficiencies of the rule
2-169
-------
when vapor balance systems are used. Visual, audible, and
olfactory leak detection methods will not identify leaks on a
consistent basis. A specific monitoring schedule should be
required with leak detection made by portable VOC monitoring
equipment.
Response: The Agency does not agree that the requirements
to define a leak should be made consistent with the HON. The HON
contains a program for reduction of emissions from storage tanks,
whereas these standards use Method 21 as a surrogate for a
tightness test. The Agency agrees that measurable leaks can be
detected through other detection methods, however, the commenter
did not supply sufficient test data showing that comparable
tightness would be achieved at multiple sources using these
alternative methods. The Agency has incorporated provisions in
the final rule for the use of alternative methods to be used in
the determination of compliance. The Agency believes that these
provisions in section 63.565 (m) of the final rule provide a
sufficient mechanism to owners and operators at terminals that
desire to use alternative devices.
Comment; Two commenters (32, 34) took exception to the
requirements in §§ 63.563(c) and 63.567(j) for complete
reinspection and monitoring of a closed vent system within 5 days
if a single leak is observed. The commenters stated that only
the leaking components should be reinspected. Complete
monitoring of closed vent systems should only be required in the
initial performance test and annually thereafter. One commenter
(32) requested that Section 63.563(c) be revised as follows:
(Proposed revisions are shown in BOLD)
(c) Leak detection for closed vent systems and control
devices.
(1) If evidence of a potential leak is found by visual,
audible, olfactory, or any other detection method, the potential
leak shall be monitored while in use within 5 days by the method
specified in Section 60.485(b) of this chapter. Each detection
of a leak shall be recorded, and the leak shall be tagged until
repaired. (This revision substitutes "the potential leak" for
"all ductwork and piping connections to covers and control
2-170
-------
devices", thereby limiting monitoring to only the detected leak
rather than the complete ductwork and piping system.)
(2) If a leak detected by (c)(1) is due to a design flaw or
material failure, then all ductwork and piping and connections to
covers and control devices that could be expected to experience
similar failure shall be monitored within a time reasonably
practical, by the method specified in Section 60.485(b) of this
chapter. Each detection of a leak shall be tagged until
repaired. The reasonableness of the monitoring period shall
consider such circumstances as location, weather, safety, and
extent of work to be accomplished. (Could the required
(3) If a leak detected by (c)(1) is due to operator or
equipment failure, then all components the operator has serviced
or equipment which has failed shall be monitored within 5 days of
confirmation of the original leak by the method specified in
Section 60.485(b) of this chapter. Each detection of a leak
shall be tagged until repaired.
(4) A reading of 10,000 ppmv or greater as methane (as
determined using the test methods in Part 60, Appendix A,
method 21) shall constitute a leak. A reading of less than
10,000 ppmv shall require no further action.
(5) Piping systems and equipment which normally operate
under vacuum are exempt from the requirements of
Sections 63.563(c)(l) through (4).
Response: The Agency agrees with the commenter that annual
leak detection and repair of the closed vent system is
appropriate. In addition to detecting leaks through the annual
leak detection and repair program, any potential leaks that are
detected by visual, audible, or olfactory methods should be
inspected to positively identify the potential leak and the
potential leak will be monitored within 5 days. If a leak is
detected, a first attempt at repair must be made within 15 days
or prior to the next loading, whichever is later.
Comment; Six commenters (29, 44, 46, 69, 70, 71) requested
revisions to the leak repair requirements. The Agency should
recognize that marine tank vessels are not loaded every day at
many affected sources. Monitoring requirements in 63.563(c)(1)
and leak repair requirements in both 63.563(a)(5)(ii) and
2-171
-------
63.563(d) are reasonably required within days of leak detection
if the source is continuing to load marine tank vessels.
However, the Agency should allow additional time for the
monitoring and repair of leaks in cases where the affected source
does not continue to load marine tank vessels.
In each of these paragraphs, the phrases "5 days" or
"5 calendar days" should be replaced with "before the next vessel
is loaded or 5 days, whichever time is later."
One commenter (71) noted that virtually all fugitive
emission programs require leak repair within 15 days, with a
first attempt within 5 days. Imposing different requirements for
this case will lead to confusion and inefficiencies, with no
significant, offsetting environmental benefit. The commenter,
therefore, recommended that Section 63.563(d) be made consistent
with other fugitive rules.
Response; The Agency agrees with these commenters.
Section 63.563(c)(1) of the final rule was revised to require
monitoring of the leak before the next vessel is loaded or within
5 days, whichever time is later. Sections 63.563(a)(5)(ii) and
63.563(d) were revised to require a first attempt at leak repair
before the next vessel is loaded or within 15 days, whichever
time is later.
2-172
-------
2.9 REPORTING AND RECORDKEEPING REQUIREMENTS
Comment: Four commenters (25, 29, 32, 73) responded to the
initial notification requirements of § 63.567(b). Three
commenters (25, 32, 73) stated that § 63.567(b) and § 63.9(b)(2)
are inconsistent. As currently proposed, 63.567(b) requires that
all existing sources file an initial notification within
13 months of the promulgation date, compared to § 63.9, which
requires an initial notification within 120 days after the date
of promulgation. These commenters stated that the perceived
intent is for the 13 months stipulated in § 63.567(b) to
supersede the 120 days in § 63.9; clarification of the intent is
requested. These commenters further stated that the 13 month
time period is very appropriate because time is needed for
potentially affected facilities to gather the data to properly
evaluate their applicability to this regulation.
Conversely, one commenter (29) stated that tying the initial
notification filing deadline to promulgation erroneously presumes
that all existing sources would be subject to these requirements.
Therefore, the Agency should utilize the time period specified at
63.9(b)(2) of "120 calendar days" and begin this time period for
filing the initial notification on the date that an owner or
operator becomes subject to this subpart. This commenter further
stated that, because the initial notification will contain
information only for a portion of the year (that is, up until
such time as the source exceeds the applicability thresholds),
then only the information for that portion of the year should be
required to be submitted as part of the initial notification.
Other reporting requirements that will apply to the source once
it has achieved compliance should cover additional information
requirements. Accordingly, 63.567(b) should be rewritten as
follows: (Revisions in bold)
(b) The owner or operator of any affected source shall
include the following information in the initial
notification report required by 63.9 of Subpart A of
this part. This initial notification report shall be
submitted within 120 calendar days after the end of the
2-173
-------
month in which the source exceeds the applicability
threshold levels specified in S 63.560 (b) [deletes
"within 13 months (after date of promulgation). After
the first year, the owner or operator shall continue to
record the information requested and keep the
information readily available for inspection."]. The
information shall include, as a minimum:
(1) The affected source's name and address;
(2) For sources subject to § 63.562(c) or (d), MACT
standards, the annual HAP emissions from the affected
source as calculated according to the procedures in
§ 63.565(d). Emissions will be reported by commodity
and type of vessel (barge or tanker) loaded for the
portion of the year up through the end of the month in
which the source exceeds the applicability threshold
levels specified in § 63.560(b).
(3) As an alternative to reporting the information in
paragraph (2), the affected source may submit
documentation showing that all HAP-containing marine
tank vessels loading occurred using vapor tight vessels
that comply with the procedures of § 63.563(a) and that
the emissions were routed to control devices meeting
the requirements specified in § 63.563(b).
Response: The Agency has incorporated the initial
notification requirements from the General Provisions found in
40 CFR Part 63. These requirements require affected sources
having a startup date prior to the date of promulgation to submit
an initial notification within 365 days after promulgation.
Sources having a startup date after the promulgation date or that
become affected sources after the promulgation date must submit
an initial notification report within 365 days or 120 days of
startup, whichever occurs before notification of the initial
performance test (refer to § 63.9 of the General Provisions to
40 CFR Part 63 and § 63.567(b) for additional information on the
specific requirements for initial notification).
Comment; Five commenters (27, 32, 34, 41, 75) recommended
revisions or requested clarifications to Sections 63.563(a)(5)
and 63.566(c) to make them more workable and less burdensome.
Section 63.563(a)(5)(i) requires that valves leading from the
vent stream to the atmosphere be securely closed using a
2-174
-------
car-seal, a lock-and-key type configuration, or be equipped with
a flow indicator (except for pressure/vacuum relief valves). One
of these commenters (27) interpreted these requirements as not
applicable to a vessel's equipment as it refers to "collected
vapors." The commenter noted that the specific requirements
could not work with respect to many pieces of vessel equipment
and requested clarification. Three commenters (32, 34, 41)
believed this requirement should be applied only to valves
controlling or diverting the major vapor flow streams and then
only while the vapor control system is in use. The EPA should
exclude all small portions of the vapor stream, such as streams
for analyzers, instrumentation devices, purging, venting portions
of piping for maintenance, depressurizing isolated equipment or
devices, and sampling equipment used for laboratory analysis.
One commenter (32) requested the language in Section 63.563(a)(5)
be reworded as follows: (revisions noted in bold)
(5) Major vent stream by-pass requirements
(i) Each valve in the vapor collection system that would
lead the major vent stream (i.e., collected vapors) to the
atmosphere, either directly or indirectly, shall be secured
closed while in use, using a car-seal, a lock-and-key type
configuration, or shall be equipped with a flow indicator, except
for pressure/vacuum relief valves. Loading product with an open
by-pass line shall constitute noncompliance with the standard.
Additionally, Section 63.567(c) requires that continuous
records be kept for 5 years concerning periods when flow bypasses
the control device, when maintenance is performed on car-sealed
valves, when car-seals are broken, and when valve positions are
changed. Two commenters (34, 75) stated that these recordkeeping
and reporting requirements should be deleted because the loading
frequency and the significant number of valves at facilities with
multiple berths loading multiple products make reporting
information on locking mechanisms extremely burdensome. Another
commenter (32) requested the following changes in
Section 63.567(c) to make these recordkeeping requirements more
practical and effective:
2-175
-------
(1) All periods when major vapor stream flow bypassing the
control device is indicated if flow indicators are installed
under Section 63.564(a)(1), and
(2) At times when maintenance is performed on car-sealed
valves, when the car-seal is broken, and when the valve position
is changed (i.e., from open to closed for valves in the vent
piping to the control device and from closed to open for valves
that vent the major vapor stream directly or indirectly to the
atmosphere bypassing the control device) if valves are monitored
under Section 63.564(a)(2).
Response; The final rule was clarified to state that the
requirements of § 63.563(a)(5)(i) of the proposed rule apply only
to shore-side or terminal piping carrying vapors to an emissions
control device. The Agency agrees with the commenters that small
portions of the vapor stream should be excluded from the valve
closure requirements and has included in the promulgated
regulation additional exceptions mentioned by the commenters for
valves that would divert collected vapors away from the control
device or to the atmosphere; the exemptions in the regulation
include valves for pressure/vacuum relief, analyzers,
instrumentation devices, sampling, and venting for maintenance.
With this clarification, the Agency believes the reporting and
recordkeeping requirements in § 63.567 of the final rule are
reasonable.
Comment; One commenter (44) stated that the recordkeeping
requirements of 63.567(g) should be revised. Section 63.567(g)
says that vapor-tightness documentation required under 63.563(a),
which includes five subparagraphs, must be on file at the
affected source in a permanent form available for inspection.
Section 63.567(h), however, defines documentation requirements
for Section 63.563(a)(2). Section 63.567(g), therefore, should
only address the documentation required under 63.563(a)(1), (3),
(4), and (5).
In addition, the language in 63.567(g) is confusing. No
record retention time is specified, and the term "in permanent
form" does not seem to be defined. This section should be
consistent with Section 63.10(b)(1) of the General Provisions of
2-176
-------
40 CFR Part 63, which specifies that files shall be retained for
5 years; 2 years must be on site, and the remaining 3 years may
be retained offsite. Such files may be maintained on microfilm,
computer, computer floppy disks, magnetic tape disks, or
microfiche. The commenter also saw no need to keep a historical
record of the information required under 63.563(a). Only the
most recent information is of use. The commenter recommended
that 63.567(g) be revised to read as follows: (revisions in bold)
(g) The owner or operator of an affected source subject
to the provisions of this subpart shall keep the vapor-
tightness documentation required under 63.563(a) (1),
(3), (4), and (5) on file at the affected source to
reflect current information (deleted "in a permanent
form available for inspection"). Such documentation
may be on microfilm, on a computer, on computer floppy
disks, on magnetic tape disks, or on microfiche, but
must be available for inspection.
Response: The Agency agrees that revision of the
recordkeeping requirements is warranted. The only historical
records that need to be maintained by sources are records of the
need for repairs to emission control equipment and data regarding
vapor tightness. The requirements for storage medium are
consistent with the General Provisions of 40 CFR Part 63. The
final rule allows owners or operators to maintain files on
microfilm, computer, computer floppy disks, magnetic tape/disks,
microfiche, or paper.
Comment: One commenter (44) stated that continuous
monitoring, recordkeeping, and reporting of control device
parameters should not be required if the design of loading
facility prevents loading without a properly functioning control
device (e.g., loading pumps interlocked with control device
parameter sensors).
Response; The Agency has considered allowing interlock
systems in lieu of the monitoring requirements presented. While
the EPA wishes to encourage innovative technologies such as
interlock systems, the EPA has insufficient information on the
variety of designs and applications of interlock systems to
2-177
-------
specify alternative monitoring, recordkeeping, and compliance
procedures that would be appropriate for all such systems.
Sources wishing to use interlock devices may apply to the
Administrator as described in the General Provisions § 63.8(f)
and in § 63.564 (j) of the final rule.
Comment; One commenter (41) stated that the proposed rule
[§ 63.566(b)(1)(ii)] requires submission of quarterly reports of
the source's HAP control efficiency. The commenter questioned if
compliance with an outlet concentration on an adsorbing system
automatically indicates 95 percent efficiency since the
monitoring standards only require outlet concentration.
Response; The Agency notes that the final rule no longer
requires quarterly reporting of HAP control efficiency but
requires annual reporting of exceedances. The Agency believes
that these monitoring and reporting requirements address multiple
commenters' concerns about the suitability of the requirements to
the control systems used. The Agency believes that this
commenter's specific comment becomes a moot point given the
changes in the monitoring and reporting requirements contained in
the final rule.
Comment: One commenter (73) noted that Section 63.567 (j) (2)
requires a source to repair and document information related to
any leak detected in the closed vent system or control device.
However, this section also contains a conflicting requirement to
document "findings" even if a leak is not discovered. The
requirement for a negative declaration is unreasonable and
conflicts with the applicability criteria of the proposed rule.
Response: The Agency agrees with the commenter and has
removed the requirements for recording that no leak was found.
The Agency believes that this addresses the commenter's concern
regarding the submission of "negative" information by minimizing
the amount of information submitted while allowing the Agency to
obtain an indication of a source's compliance.
2-178
-------
2.10 GENERAL PROVISIONS TO 40 CFR Part 63
Comment: Seven commenters (04, 23, 29, 31, 72, 73, 78)
raised the issue of referencing parts of the General Provisions
to 40 CFR Part 63 and recommended applicable sections of the
General Provisions be referred to within the MACT standard.
Three commenters (04, 72, 73) believed this would reduce
confusion in determining which requirements of these General
Provisions are superseded and which are not. Two commenters (04,
73) believed this would reduce the likelihood of wrong decisions
that may result in involuntary noncompliance. Three commenters
(23, 73, 78) recommended that the Agency refer to only those
applicable sections of the General Provisions. Five commenters
(23, 31, 72, 73, 78) encouraged the Agency to incorporate a table
or list of applicable/not applicable sections to clarify the
applicability of the General Provisions (the same approach used
in the HON or Petroleum Refinery MACT regulations). One
commenter (23) pointed out that Section 63.560 does not clearly
state the applicability of the General Provisions to major
sources with emissions less than 1 Mg. One of these
commenters (29) also stated the proposed Subpart Y provisions are
predominantly self-contained with only several cross-references
to needed subparagraphs of Subpart A. The commenter recommended
the Agency not rely on Subpart A except where absolutely
necessary and make Subpart Y a stand-alone MACT standard.
One commenter (78) suggested EPA create a mechanism for
quick resolution (i.e., within 30 days) of questions that
facility owners/operators will have regarding the overlap between
individual MACT requirements and the requirements of the General
Provisions to 40 CFR Part 63.
Response: The General Provisions to 40 CFR Part 63 were
finalized on March 16, 1994, less than 1 month prior to proposal
of these standards. Sufficient time was unavailable at proposal
to incorporate the requirements of these final General Provisions
into this regulation and identify specific applicable sections;
however, the final regulation contains a table identifying the
2-179
-------
applicable and nonapplicable requirements of these General
Provisions. Addition of this table eliminates confusion about
what requirements are appropriate and eliminates compliance
concerns for subject sources. Because Subpart A was developed to
avoid reiteration of similar requirements in each of the Part 63
regulations, this regulation references the General Provisions
and is not a "stand-alone" regulation. Sources subject to
Subpart Y are subject to the General Provisions of 40 CFR Part 63
in its entirety, except for individual instances identified as
not applicable in this table.
2-180
-------
2.11 WORDING OF REGULATION
2.11.1 General
Comment: Five commenters (06, 10, 71, 72, 79) stated the
control standards should be clarified to reflect the fact that
not all HAP vapors displaced from marine tank vessel loading will
ultimately be captured by controls. Paragraphs 63.562(a), (c),
(d), and (g) should clarify the distinction between the capture
requirement and the destruction requirements of the control
system. Paragraph (a)(1) specifies that vapor collection systems
must be "designed to collect all HAP vapors displaced from marine
tank vessels during loading ..." However, it is believed that
EPA's intent is to allow a facility to be able to meet the
standard by collecting and treating only those HAP vapors
necessary to meet the standard. Paragraph (c) specifies that
captured HAP emissions must be reduced by 93 weight-percent,
whereas paragraph (d) states that HAP emissions be reduced by
98 weight-percent. It is unclear in Paragraph (d) if the total
capture and reduction combined is to be 98 percent or if the
reduction of captured emission alone is to be 98 percent.
Additionally, the standard should specify what the minimum
capture efficiency of the capture system should be or prohibit
leakage of VOC from the vessel and control system during
transfer, consistent with other LDAR rules.
Response; The approach to marine tank vessel vapor control
employed under these final standards is to ensure satisfactory
capture of vapors associated with bulk liquids subject to rule
requirements and require technology that satisfies the
requirements of both Sections 183 (f) and 112 of the Act.
Satisfactory capture is not to be equated with 100 percent
capture of displaced vapors. The Agency has provided in the rule
various approaches to demonstrate an acceptable level of vapor
"tightness." Once vapor tightness is demonstrated, and the
requirements of §§ 63.562(b)(1), (c)(2), or (d)(1) are satisfied
(i.e., owners or operators show that vapors are delivered to
shore), then the focus becomes the ability to demonstrate
2-181
-------
compliance with the emission reduction requirements of the RACT
or MACT standards as determined in § 63.563.
Comment; One commenter (10) stated that the regulation
should clearly indicate that State and local agencies may adopt
rules more stringent than Federal regulations.
Response; The Agency agrees that State or local agencies
may promulgate more stringent regulations than the final
standards promulgated for these sources. However, such
regulations must comply with the U.S. Coast Guard safety
regulations.
Comment: One commenter (20) believed that, as proposed,
this regulation seems unduly complex, especially in comparison to
the new source performance standard for bulk gasoline terminals
(40 CFR 60, Subpart XX) and the recently proposed MACT standard
for gasoline distribution facilities (40 CFR 63, Subpart R).
Response: The Agency has revised several sections of the
final rule to make it more streamlined and less confusing. The
Agency has also included several options for owners or operators
to choose means of demonstrating compliance with the standards or
monitoring the performance of control equipment.
Comment; One commenter (31) stated that EPA should clarify
that the option to load at negative pressure applies to some
vessels, not to all vessels as a group. The EPA offers an option
of loading at negative pressure as an alternative to the other
provisions intended to prevent losses from leaking vessels. This
option is appropriate. However, the language of the proposed
rule makes it sound like this approach is an all-or-nothing
approach, in which we must either load every vessel (even those
that are documented not to be leaking) at negative pressure, if
we decide to use the negative pressure option at all.
Section 63.563(a)(2)(iv) says that, as an alternative to the
normal procedures, "the owner or operator of an affected source
shall ensure that each marine tank vessel is loaded at negative
pressure." The EPA should change this statement to say, "the
2-182
-------
owner or operator may ensure that a marine tank vessel is loaded
at negative pressure."
Response; The Agency agrees with this commenter and has
clarified the final rule to show that negative pressure loading
is one alternative to show compliance with the requirements to
load vapor tight vessels.
Comment; One commenter (31) believed EPA should clarify
what is meant by "gas CEM" and the difference between gas CEM's,
VOC CEM's, and GC CEM's as used in the appendices.
Inconsistencies exist in the Appendices' specifications for
CEM's. For example, Appendix A (1.4.1) says all CEM's must
conform to the performance specifications provided in 40 CFR
Part 60 Appendix B. However, later in Appendix A of this
proposed rule, performance specifications for VOC and GC CEM's
differ significantly from those in 40 CFR Part 60 Appendix B.
The commenter provided an explanation; it appears that, in
Appendix A (1.4.1), EPA means that "gas CEM's" must conform to
the Part 60 Appendix. However, this raises a different problem;
the term "gas CEM" is never defined. Thus, it is unclear how
owners or operators should proceed in order to comply with the
rule. To resolve these problems, the relevant portion of
Appendix A (1.4.1) should be amended by adding the following as
the first sentence:
1.4.1 Gas CEM's and COMS's. Gas CEM's are defined as those
CEM's whose performance specifications are contained in
Appendix B of Part 60 of this chapter. All gas CEM's and
COM3 that are ...
... chapter. Unless a gas CEM's or COMS is subject ...
The relevant portion of Appendix A (3.2) should be amended by the
following:
... as stringent as the applicable RA required requirement
for a gas CEM's pursuant to ...
2-183
-------
The relevant portions of Appendix A Performance Specification 101
should be amended by the following:
(1.1.1) ... acceptability of the VOC OEM's at the time ...
... evaluate VOC CEIS performance at the time ...
(1.2) ... conformance of the VOC CEM's with these ...
(3.1) 3.1 VOC CEM's installation ...
... The VOC CEM's shall be ...
(3.2) ... locate the VOC CEM's probe at a ...
(4.) 4. VOC CEM's Performance and ...
(4.1) ... installing the VOC CEM's.
(4.3) For a VOC CEM's intended ...
... For a VOC CEM's installed to ...
(4.6) ... within the VOC CEM's measurement ...
(4.7) ... time for the VOC CEM's must not ...
(4.8) The VOC CEM's must allow ...
..., VOC CEM's calibration response ...
(5.1) Install the VOC CEM's, prepare the ...
... prepare the VOC CEM's for operation and ...
... of the VOC CEM's, the owner ...
(6.3.2) Calculate the VOC CEM's relative error ...
Cm = Average VOC CEIS response, ...
The relevant portions of Appendix A Performance Specification 102
should be amended by the following:
(l.l) ... acceptability of the GC CEM's at the time ...
... evaluate CEM's performance at the time ...
(1.2) ... conformance of the GC CEM's with these ...
(4.6) The GC CEM's must allow ...
..., the GC CEM's calibration response ...
(6.1.2) ... between the GC CEM's responses ...
(7.1) x = GC CEM's response
(8) ... calculations, GC CEM's data records, ...
The relevant portion of Appendix C (1.1) should be amended by the
following:
... provided in either Appendix B of Part 60 or Appendix A
of this part of this chapter as appropriate. ...
2-184
-------
The relevant portion of Appendix D (1.1) should be amended by the
following:
1.1 Gas CEM's and COM3 ...
Response: The Agency has clarified the use of OEM's in the
final rule and has provided alternative compliance provisions.
Furthermore, the references to performance specifications 101 and
102 have changed to performance specifications 8 and 9
respectively (Docket No. A-90-44, Item No. IV-I-4).
Comment; One commenter (31) suggested Appendix A, PS-101,
Section 4.5 be modified to facilitate the use of audit gases from
sources other than EPA and to allow the use of any accurate audit
gas regardless of source. Appendix A, PS-101, Section 4.5
currently requires an EPA audit gas as a first choice for a CEM
calibration. A "Protocol 1" gas is the only other choice listed
as acceptable. These restrictions are unnecessary and add to the
burden of the rule.
The commenter states that in many instances no EPA audit gas
exists for the compound(s) of interest. And, even when they are
available, the EPA audit gases are very expensive and may take
months to obtain. Thus, even though the commenter has a high
degree of confidence that the gas concentration would be
accurate, the EPA gases would not be a competitive product in an
open market. The commenter hopes that the Agency will not
institute a monopoly of its own creation, to further the sale of
these gases. Indeed, this section creates the appearance of a
conflict of interest. The EPA should simply note that these
gases are among the acceptable choices. Additionally, the
commenter indicated that although "Protocol 1" gas mixtures are
more readily available at more reasonable prices, there is no
reason why they should be the only acceptable alternative to
EPA's own audit gases. The whole idea of an audit gas is to
provide a reference concentration. An audit gas of verified
concentration should be acceptable, no matter who prepared it.
Even a local laboratory or the company itself may be able to
produce an acceptable audit gas and should be allowed to do so.
2-185
-------
Thus, EPA should allow any acceptable audit gas regardless of who
prepared it. Protocol 1 specifications more than meet the
accuracy requirements of the CEM. Allowing the use of a
Protocol 1 gas mixture would also minimize the costs and delay
times that have been associated with the EPA audit gases, while
still achieving acceptable results. This commenter recommended
that EPA revise Section 4.5 as follows: "4.5 Performance audit
gas. An audit gas approved by the permitting authority shall be
used. A Protocol 1 gas mixture within the calibration range, or
certified EPA audit gas, if used, shall be deemed approved. For
any other gas, the owner or operator shall demonstrate that the
concentration has been verified."
One commenter (31) thought EPA should clarify what is meant
by "span" and "range" in the appendices and make several
revisions to assure that these terms are used consistently.
In a number of places in the Appendices, the words "span"
and "range" are used. These terms can be used interchangeably or
differently, depending on the industry. One commenter (31)
stated that EPA may be using the terms differently at different
places in the Appendices (e.g. Appendix A (5) measurement Span
versus Appendix A, PS 101 2.1 Instrument Range, and Appendix A,
PS 101 2.2 Instrument Span or Span Value), thus causing confusion
and making compliance difficult.
The EPA should define these terms. For example, under a
different statute, EPA has defined "Instrument Measurement Range"
as the difference between the minimum and maximum concentration
that can be measured by a specific instrument, 40 CFR Part 266
App. LY 2.1.2.3. Similarly, in other contexts, EPA has defined
"Span or Span Value" as the upper end of the Instrument
Measurement Range. 40 CFR Part 266 App IX 2.1.2.4, 40 CFR
Part 60 Appendix B PS-2 (2.4), and 40 CFR Part 60 Appendix F
(2.3). These definitions would be acceptable for the Enhanced
Monitoring rule as well. Then, once EPA has defined these terms,
the Agency should be consistent in using them. Therefore, the
2-186
-------
relevant portions of the appendices should be amended by the
following:
Appendix A (l.l.d) -
(5)
(5.1)
(5.2)
(5.2.1) -
Appendix B (2.1.1) -
(2.2)
Appendix D (2.1.2) -
(d) Instrument range;
d) Instrument range;
5. Measurement range is...
Measurement range is ...
... consider the measurement range in
the ...
... a proposed BMP measurement range
which meets any existing measurement
range requirement ...
... two types of measurement range
specifications:
5.2.1 Measurement ranges...
5.2.2 Measurement ranges...
... over the measurement range...
... over the measurement range....
... at the low and high levels. ...
Response; The Agency notes that the references to
performance specification 101 have been changed in the final rule
to performance specification 8. Performance specification 8 does
not require the use of audit gases. The terms "range" and "span"
are also no longer used in the manner stated by the commenter in
performance specification 8.
Comment; One commenter (71) stated Section 63.563(a)(5) and
(a)(5)(i) virtually duplicate 63.564(a) and that these sections
should therefore, be combined for simplicity and clarity. The
following recommendations were made:
1) Delete 63.563(a)(5) and (5)(i).
2) Add the "except for pressure/vacuum relief valves" phase
from 63.563(a)(5)(i) to63.564(a).
3) Move and renumber 63.563(a)(S)(ii) to 63.564(a)(3).
Response: The Agency has streamlined the requirements of
the final rule. The specified sections are among those
clarified.
2-187
-------
2.11.2 Definitions
Comment: One commenter (10) believed that EPA should expand
the definitions section to include the definition of "State and
local permitting authority" and "delegated State authority,"
which is not mentioned until the last section of the regulation.
The regulation should be revised and made consistent throughout.
Response: The definition for State and local permitting
authority and delegated State authority is given in the General
Provisions to 40 CFR Part 63. These definitions have not been
modified in the final rule; therefore, the Agency has not
included these definitions in the text of the final rule.
Comment; One commenter (20) asked what is meant by "an hour
block average."
Response; The Agency agrees with the commenter and has
described how to determine an hour block average in Section
63.564 of the final rule as 4, 15-minute averages.
Comment: One commenter (24) requested clarification on the
definition of "major source." Because the rule states in
63.560(a) that the provisions apply to any existing or new marine
tank vessel loading or unloading operations that meet the
definition of a major source in Section 112(a)(1) of the Clean
Air Act as amended, it is not clear why a source would even need
the 1 metric ton exemption. However, it is clear in the preamble
that this rule applies to any marine tank vessel loading
operation that is associated with a facility that is a major
source as defined in 112(a)(1) of the Clean Air Act, as amended.
Section 63.560(a) should state that the rule applies to any
marine tank vessel loading and unloading operations that are
major sources or are associated with a major source as defined in
112(a)(1) of the Clean Air Act.
Response; The Agency agrees that only major sources are
affected by the MACT standards. However, the Agency disagrees
with the commenter regarding the need for establishing two
subcategories based on HAP emissions. A terminal having very
small marine tank vessel loading emissions may be classified as a
2-188
-------
major source because of source-wide HAP emissions. The Agency
believes that a subcategory for such sources is justified (see
Section 2.3.1 of this document for additional discussion). The
selected MACT {i.e., the MACT floor) for the subcategory of major
sources emitting less than 10 Mg/yr of all individual HAP and
less than 25 Mg/yr of total HAP is no control.
Comment: One commenter (29) requested that a definition of
"affected source" be added to § 63.561 because the term is not
defined for purposes of proposed Subpart Y. It appears that EPA
intended that marine tank vessel loading operations that are
major sources by themselves or located at major sources are to be
the affected source. The preamble to the proposed rule specifies
that "the source category to be regulated is major source marine
tank vessel loading operations," (59 FR 25007), but this
discussion is inadequate.
Only the term "affected source," as clearly defined by a
definition in 63.561, should be used in the final rule. The use
of the term "affected facility" as found in 63.563(a)(2)(iii)(B),
"the owner or operator of the affected facility shall require
that the vapor-tightness failure be documented," should be
eliminated.
The definition of "affected source" should be added to
63.561 as follows:
"Affected source for purposes of MACT is a marine tank
vessel loading operation which meets the applicability
requirements in § 63.560(a)."
Response: The Agency has added a definition for the term
"affected source" to the final rule.
Comment: One commenter (31) stated the definition of
"recovery device" should be changed to include all devices that
remove vapors, regardless of the ultimate disposition of the
recovered material. As currently written, the definition of
"recovery device" in Section 63.561 includes only those units
that are "used for the purpose of removing vapors and recovering
liquids or chemicals." The phrase "recovering liquids or
2-189
-------
chemicals" is unclear. Although EPA does not define the term
"recovering," this commenter thought enforcement personnel might
interpret it to mean "obtaining for recycling or reuse." In this
case, adsorbers, condensers, etc. would not be considered
"recovery devices" and also would not be "combustion devices."
Consequently, these pieces of equipment would not be "control
devices" as defined under this rule and would not qualify to use
a removal efficiency of 95 percent.
Some of these devices will likely demonstrate a control
efficiency of 98 percent or better. Other devices, such as
scrubbers, will only be able to achieve the 95 percent control
level. It is important that EPA clarify that these devices may
be used to comply with the standard. The EPA should also revise
the rule to indicate that any device (whether or not it is a
combustion device) that achieves a removal efficiency of
98 percent is acceptable, regardless of the disposition of the
removed material.
Another commenter (39) requested clarification of combustion
vapor control device requirements and recommended that EPA define
in the proposed rule each combustion device and all applicable
requirements.
Response: The Agency has clarified the definition of
"recovery device" in the final rule by using a definition similar
to one established in the HON. The Agency has provided test and
monitoring protocols for standard control devices available in
the marketplace. The Agency has also provided for the use of
alternative control technologies and has provided owner/operators
provisions to test and monitor these devices.
Comment; One commenter (31) suggested that EPA add two new
definitions and change the definition of "source," as follows:
(1) Cargo consolidation means transfer of product from a barge or
ship to another vessel by way of loading hoses without utilizing
shore equipment; (2) Bunkering operation means the fueling of a
vessel from a barge or other vessel without utilizing shore
equipment; and (3) Source means any location where there exists a
2-190
-------
dock or berth capable of bulk loadings on marine tank vessels
except offshore drilling platforms, offshore lightering
operations, cargo consolidation, and bunkering operations.
Another commenter (114) requested that these standards apply to
vessel-to-vessel transfers.
Another commenter (34) disagreed with the proposed rule
definition of "source," particularly with regard to Section 112
requirements. The EPA proposes "source" in 63.561 as "any
location where there exists a dock or berth capable of bulk
loadings on marine tank vessels except offshore drilling
platforms and offshore lightering operations." The commenter
considered the definition too restrictive and in conflict with
the proposed Section 112(d) provisions in the proposed rule. The
definition should be revised to state "any location where there
exists a dock or berths, or multiple docks or berths, capable of
bulk loadings on marine tank vessels except offshore drilling
platforms and offshore lightering operations."
Another commenter (85) made suggestions regarding the
definition of "source." The commenter supported the exclusion of
lightering operations and noted the word "offshore" should be
dropped as it is defined in the definition of "lightering." This
commenter also suggested the addition of "bunkering" to the
exclusion list, because bunkering operations are not meant to be
covered, i.e., the definition refers to tank vessels only. Also,
the phrase "a marine tank vessel" should be "of marine tank
vessels."
Response; These standards address bulk loading operations
at terminals. Loading operations between vessels including
lightering and bunkering operations are not addressed by these
standards. The Agency has clarified the determinations and
applicability sections accordingly. The Agency did not agree
that the definition revisions suggested by commenter 34 were
needed in the final rule.
Comment: One commenter (34) also requested that the
definition of "terminal," included in 63.561, be modified to
2-191
-------
include "any land or sea based permanent structure, or structures
located on contiguous property, capable of loading liquids in
bulk onto marine tank vessels."
Another commenter (69) recommended that Section 63.561 be
amended to clarify that the definition of "terminal11 does not
include marine fueling facilities. This definition should be
modified to read "... capable of loading liquid cargoes in
bulk onto marine tank vessels." The commenter believes this
modification will prevent unnecessary confusion regarding the
scope of the proposed regulations.
Response; The proposed regulation covers bulk loading
operations and not fueling operations. Additional clarification
of the definition is not necessary. The Agency disagrees that
the suggested revision to the definition of "terminal" is needed
given the Agency's use of "contiguous." The Agency also believes
that the permanence of a terminal should not dictate a limitation
on the requirements to install controls if that terminal exceeds
the applicability levels for the final rule.
Comment: One commenter (34) requested that the final rule
include the definition for steam generating units contained in
NSPS Subpart D-Dc in place of the proposed definition for
boilers. Some affected facilities will likely use combustion
devices that heat a process gas or fluid as well as generate
steam, and these combustion devices may also be used as a control
device, using marine loading vapors as a primary fuel.
Response: The definition of boilers has been revised to
reflect the commenters suggestion.
Comment: One commenter (36) requested clarification on the
term "percentage of HAP" used in the emission estimation formula.
The commenter believed that the percentage of HAP's is
the percentage of HAP in the vapors emitted by the facility, not
the HAP in the liquid commodity loaded. These percentages are
often substantially different and would produce different
determinations of HAP emissions from a given facility. For
example, HAP may be 2 percent by weight in the liquid product but
2-192
-------
only 0.67 percent by weight in the vapor. It appears from both
the preamble and the proposed regulations that the Agency is
attempting to reduce the emissions, which relates, of course, to
the content of the vapor.
Response; The emissions estimation procedures in the final
rule no longer specify the calculation of the percentage of HAP.
Comment; One commenter (46) believes the term "dry-docking"
is unclear as used in Section 63.563(a)(2)(ii) and (iii)(B). To
eliminate potential confusion, the commenter suggested "dry-
docking" should be replaced by the phrase "gas freeing for hot
work" or be defined in Section 63.561 as "having to gas free a
vessel."
Response: The references specified by the commenter now
refer to "cleaning and gas freeing or drydocking."
Comment; One commenter (71) stated that "continuous" is
used throughout the rule and needs to be defined. Industry
comments on the HON pointed out that modern technologies for
monitoring process parameters are not continuous, like archaic
strip-chart recorders, but are digital. Digital technologies
make measurements frequently but not continuously. The commenter
recommended adding to this subpart the following definitions from
the final 40 CFR Part 63 Subpart G rule: (1) "Continuous record
means documentation, either in hard copy or computer readable
form, of data values measured at least once every 15 minutes and
recorded at the frequency specified in § 63.152(f) of this
subpart"; and (2) "Continuous recorder means a data recording
device that either records an instantaneous data value at least
once every 15 minutes or records 15-minute or more frequent block
average values."
Response; Use of the word "continuous11 has been made
consistent with its use in the General Provisions to 40 CFR
Part 63. A definition similar to one used in the HON has been
incorporated in the final rule.
Comment; One commenter recommended EPA modify the
definition of primary fuel and the wording of the performance
2-193
-------
test exemption so that the vent stream need not be the only fuel
burned in a combustion device exempt from testing. The proposed
definition of primary fuel requires that it be "sustaining
operation of the device without the addition of other fuels".
However, this limitation is not necessary to assure destruction
of any vent material included with the primary fuel and is
inconsistent with the definition used in other regulations.
Normally a vent stream is too low in volume to support combustion
on its own and is mixed with primary fuel or supplemented with
other fuel. Destruction adequacy is based on how the vent stream
is introduced to the flame zone, not what portion of the fuel it
represents. The commenter urged the Agency to adopt the
definition in the HON of primary fuel instead of the definition
proposed here and to modify the wording of 63.563(b)(2)(i).
Recommendations:
(1) Replace the proposed definition of primary fuel with
that from 63.111, specifically:
"Primary fuel means the fuel that provides the principal
heat input to the device. To be considered primary, the
fuel must be able to sustain operation without the addition
of other fuels."
(2) Modify 63.563(b)(2)(i) as follows:
...comply with 63.562(c), (d), or (g) and the vent stream is
used as or introduced with the primary fuel;" [add
underlined]
Response; The Agency agrees with the commenter and has
incorporated similar language.
Comment: One commenter (38) requested that the baseline
outlet VOC concentration (under operating parameters for carbon
adsorbers) be defined.
Response; The Agency has defined "baseline operating
parameter" in the final rule. The Agency does not believe that
it is necessary to define each specific baseline parameter
(e.g., temperature, pressure, VOC, etc.) used in the final rule
2-194
-------
since the generic term is defined. The Agency has clarified the
use of the baseline outlet VOC concentration in the final rule.
Comment: One commenter (31) stated the definition of
"calibration precision" is incorrect and should be modified. In
Performance Specification 102 (2.2), the term "calibration
precision" is defined as an "agreement." However, calibration
precision is an indication of error resulting from slightly
differing readings for the same calibration gas concentration,
rather than the "agreement" of the readings. Therefore, the
relevant portion of PS 102 (2.2) should be amended by the
following: "The error between triplicate injections of each
calibration standard."
Response: The term "calibration precision" is no longer
used in the final rule. The Agency did not make the suggested
change to the promulgated PS- 8 and PS-9.
2-195
-------
2.11.3 Overlap With Other Standards
Comment: Several commenters (04, 10, 68, 71, 73) were
concerned about the overlap of the marine tank vessel standards
with other existing standards. One commenter (04) stated EPA
should specify whether this proposal overrides other overlapping
requirements. The commenter stated that this proposal does not
provide adequate information or guidance for a source to
determine which provisions of this proposal or other similar
standards are the most stringent. The commenter contended that
the overlap between this and other regulations will lead to
considerable uncertainty in both the regulated and regulatory
communities. The EPA should clearly specify which provisions of
other related regulations are overridden by the marine tank
vessel loading rule and which are not. The commenter suggested
that EPA employ the methodology used in the HON in which the
Agency clarified which requirements were to be complied with when
an emissions point was subject to other standards. Another
commenter (10) was concerned that this regulation may be unduly
complex, especially in comparison to the New Source Performance
Standards for bulk gasoline terminals and the recently proposed
MACT standard for gasoline distribution facilities. The EPA has
historically established standards for terminals in units of
milligrams of HAP (or VOC) per liter of product transferred. The
EPA's stated rationale is that control efficiencies are dependent
on the inlet concentration to the control device.
Five commenters (25, 29, 68, 71, 73) raised the issue of
overlap with the Benzene Transfer NESHAP. One commenter (68)
believed the regulation should clearly state in the Standards
Section (63.562) that any source currently subject to and
compliant with the benzene loading NESHAP (40 CFR 61 Subpart BB)
has discharged all of its compliance obligations under this
proposal. One commenter (73) recommended that loading operations
and equipment subject to the benzene loading NESHAP be considered
in compliance with the proposed marine tank vessel rule. Thus,
benzene loading operations, which must achieve 98 percent
2-196
-------
control, should not be required to modify, retest or otherwise
trigger requirements from a different standard. Two
commenters (71, 73) recommended modifying 63.560(e) as follows:
"(e) A loading operation that is subject to the provisions of
this subpart and also to the provisions of 40 CFR Part 61
Subpart BB, is only required to comply with the provisions of 40
CFR Part 61, Subparts A and BB."
One commenter (29) stated that 63.563(b) does not give
recognition to situations in which an existing marine tank vessel
loading facility is utilizing a combustion device or a recovery
device to achieve compliance with the benzene transfer operations
NESHAP at 40 CFR Part 61, Subpart BB. From references throughout
the preamble to this proposal, it is apparent that the Agency has
utilized many of the procedures specified in this [i.e., benzene
transfer] NESHAP. Accordingly, an owner or operator should not
be required to run performance tests when the control device
being used is in conformance with the benzene transfer operations
NESHAP.
To reflect this exemption, 63.563(b)(2) should have the
following additional exemption added to it:
(iv) When the control device being used for the marine
vessel loading is in compliance with the 40 CFR Part 61,
Subpart BB requirements.
Response; The Agency agrees that potential confusion exists
regarding the overlap of the marine tank vessel loading standards
with the benzene loading standard. In the final rule, the Agency
has clarified that benzene emissions controlled under the benzene
transfer NESHAP are not subject to the final standards for the
marine tank vessel loading operations standard. However, the
Agency notes that terminals subject to the benzene loading NESHAP
may be required to control the emissions from other commodity
loading operations (i.e., nonbenzene loading operations) under
the final marine tank vessel standards.
Comment: One commenter (34) noted that in 63.562(a)(1), EPA
duplicated existing Coast Guard regulations of 33 CFR Part 154,
2-197
-------
Subpart E 154.810 concerning interlocking valves on the line(s)
from the vessel. The commenter recommended this duplicative
requirement be eliminated from the final rule.
Response: The Agency has removed duplicative requirements
for interlocking valves where the Agency's requirements were not
needed for monitoring or compliance assurance purposes.
Comment: One commenter (36) stated that a concentration of
emissions in the exhaust stream of 6,300 ppmv is generally
equivalent to 10 milligrams per liter of gasoline -- the very
standard proposed in EPA's rulemaking on National Emission
Standards for Hazardous Air Pollutants for Source Category:
Gasoline Distribution (Stage 1) , Docket No. A-92-38, which is
also being promulgated under Section 112 of the Clean Air Act.
In addition, a concentration level of 6,300 ppmv is equivalent to
approximately a 98 percent by weight reduction of HAP emissions.
Also, a concentration of 22,000 ppmv is equivalent to about
35 milligrams per liter of gasoline or a reduction of
approximately 95 percent by weight of HAP emissions.
Response: The Agency notes that displaced vapors from
marine tank vessels vary from the beginning of loading where
concentration can approach zero (in a clean vessel) to saturation
at the end of loading. The information cited by the commenter
refers to displaced vapors from the loading of tank trucks. The
emissions from these tank trucks are typically uniform
(i.e., saturated) because of the much smaller volume of the tank
trucks. The Agency, therefore, does not agree with the
commenter's conclusions regarding the relationship between
concentration level and emissions reduction.
Comment; One commenter (59) stated the USCG has written
regulations (33 CFR 154, 155, 156 and 46 CFR Parts 30, 32, 35,
39) to address the safety aspects with regard to transfer of
crude oil, benzene and gasoline. These regulations do not apply
to other products; instead, the Commandant has conveyed a policy
for the collection of vapors for other products. This action was
done in recognition of the hazards associated with vapor
2-198
-------
collection of these other products. Safety aspects concerning
control of vapors other than crude oil, benzene and gasoline have
yet to be addressed by USCG regulation. Implementation of the
regulations should be delayed until USCG can confirm that due
consideration has been given to the safety aspects of controlling
vapors other then crude oil, benzene, and gasoline.
Response: The Agency agrees with the commenter's concern
that some control vapors are not currently addressed by U.S.
Coast Guard regulations. The final marine tank vessel loading
regulations have been developed in cooperation with the U.S.
Coast Guard so that the safety concerns expressed by the
commenter could be addressed.
Comment; One commenter (80) was concerned about Federal
requirements for sources of HAP because some existing permitting
and risk management programs address many of the same toxics
concerns. States and districts have numerous Rules and
Regulations in place which limit emissions from the significant
source categories in the commenter's region. In particular, the
cornmenter's district has Regulation 8, Rule 44 limits emissions
from Marine Vessel Loading Terminals and Regulation 8, Rule 46
limits emissions from Marine Tank Vessel to Marine Tank Vessel
Loading. The commenter believed it is critical that the
rulemaking efforts at EPA recognize the importance of established
local programs and provide flexibility to minimize negative
impacts on these programs wherever possible.
Response; The Agency agrees with the commenter and has
considered known local and district requirements when developing
these standards. The final rule contains provisions where owners
or operators of affected source may utilize test data and
reporting information used to comply with State or local
regulations where such data and information meet or exceed the
requirements contained in the final rule.
Comment: One commenter (80) raised the issue of "fugitive
sources of HAP." The preamble to the proposed rule specifically
states that fugitive sources are not considered part of the
2-199
-------
marine loading operation (organic HAP storage, piping, pumping,
vapor leaks from seals, hatch covers, and pressure relief
valves). Some of these sources may be subject to regulation
under the HON, the Refinery MACT, or the Stage I Gasoline
Distribution MACT; however, it is unclear how applicability is
divided. For example, at what point is the storage part of the
HON facility versus the marine loading terminal? The same
question applies to piping, pumping, etc. The commenter
recommended that fugitive emissions from marine loading
operations be addressed in the rule in order to provide greater
clarity of the total marine loading requirements to the
regulatory and regulated community.
Response; The source category addresses only emissions
displaced from the loading of marine vessels and does not address
shore-side emission points from other sources. Therefore, the
final rule regulates only emissions from vessel loading.
Comment; One commenter (38) stated that emissions from the
vessels that are under jurisdiction of the USCG must be
controlled. In order to load or unload, the vessel must be
compatible with the vapor recovery equipment on shore. It seems
that both agencies are interested in the common goal, control of
emission from vessels. A start can be achieved by requiring the
vessel to be vapor tight and prohibiting emissions from
ballasting operations; the commenter believed these changes
should be reflected in the final rule.
Response; The Agency believes that the basic approach
reflected in the promulgated rule satisfies the commenter's
recommendation. However, the Agency notes that ballasting
requirements have been removed from the final rule.
2-200
-------
2.12 ADMINISTRATIVE RECORD/SOURCES OF INFORMATION
Comment: Four commenters (30, 34, 42, 50) stated that the
administrative record is inadequate to support the proposed
regulations and that EPA should withdraw the Proposed Regulations
to correct and supplement the administrative record. The
commenters stated that the preamble to the proposed regulation,
the TSD, and EPA's administrative record are based on incomplete,
flawed, and outdated information. Additionally, the Preamble
fails to explain the rationale and provide the necessary
information to evaluate the Proposed Regulations. The
inadequacies in the administrative record demonstrate that EPA
has failed to meet the requirements of the Clean Air Act and the
Federal Administrative Procedures Act by proposing these
regulations. Two commenters (30, 32) stated that the Technical
Support Document refers to an "Appendix F" with data tables
showing throughput versus cost information, yet the TSD contains
no such Appendix. The commenter stated that this omission makes
it impossible for the public to assess the accuracy of EPA's
calculations. This commenter stated that the lack of access to
the underlying data and comprehensive documentation of the
information contained in the data base prevents the public from
knowing if other errors exist in the database. The commenter
stated that this omission renders the Proposed Regulations
insupportable. The fact that the public has uncovered numerous
glaring defects that EPA did not discover casts serious doubt on
the overall adequacy of the administrative record.
Three commenters (32, 34, 89) stated that EPA's extracted
data from the 1988 version of the Waterborne Commerce of the
United States (WCUS) data base is outdated and that more recent
data is available to the Agency through the 1990 or 1992
U.S. Army Corps of Engineers Marine Loading Database. The
commenter noted that the Corps was unwilling to release this
updated information to those affected by the Proposed Regulations
until mid-June of 1994. The commenter noted that other published
preliminary reports indicate total product throughput increased
2-201
-------
10 percent between 1988 and 1990. The commenter stated that this
statistic could have a significant effect on the number of
facilities subject to controls under the Proposed Regulations.
One commenter (32) pointed out problems related to the WCUS
data base. The data base lists a facility in Alaska (No. 1666)
as the second largest U.S. crude oil terminal. No such terminal
exists in the State of Alaska or anywhere else in the United
States.
The data base lists a facility located in the "Transshipment
Area" as the third largest crude oil terminal. This terminal is
reported to have a throughput of 179 million barrels. This
"terminal" is apparently the terminals associated with the Panama
Canal. Because it is not located in any U.S. State or territory,
it should not be included in the database and used in setting
MACT or RACT standards.
The database reports HAP emissions by terminal in tons and
megagrams. These figures should directly correlate for each
terminal, differing by the conversion factor between U.S. and
metric tons. However, some of the entries have emissions listed
in both columns but the values are different than the proper
l.l ratio (1.1 ton = 1 Mg).
One commenter (32) stated the WCUS data base underestimates
the number of terminals covered by the MACT standard. Based upon
specific product throughput, EPA also estimated that 1,231 of the
1,667 terminals analyzed emitted less than 0.5 tons of HAP
emissions in 1988. Again, this estimate ignores total product
throughput figures reported by the Corps of Engineers for these
terminals. Use of total throughput information included in the
data base would have led EPA to conclude that at least 128, and
possibly as many as 514, of those terminals would have exceeded a
1-ton HAP emission threshold in 1988. Further, the Corps
personnel responsible for developing the database indicate that
the total throughput figures themselves under-report actual
throughput by 10 to 20 percent and that they use 15 percent as
the best estimate of under reporting. These facts reinforce the
2-202
-------
conclusion that EPA has significantly underestimated the number
of terminals that will be subject to controls under the proposed
MACT floor.
Three commenters (32, 34, 89) stated that EPA has based its
HAP emission estimates on specific product throughputs and that
this has a major impact on MACT floor determination. The
commenters stated that specific product throughputs are
under-reported in the data base. Total throughputs, also
reported in the data base, exceed the sum of all individual
product throughputs by more than 20 percent.
For the 1,648 terminals EPA considered in developing the
MACT standard, the total volume of individual product throughputs
reported equals 2,649,129,893 barrels per year, while the total
throughput volume equals 3,205,444,509 barrels per year--a
difference of more than 20 percent. Looking at the larger set of
1,800 terminals actually reported in the database, the 20 percent
difference continues to hold true. The EPA ignored the
implications of this discrepancy in calculating HAP emissions.
The effect of EPA's reliance on under-reported throughput
figures is to understate the number of terminals impacted by the
Proposed Regulations. For example, due to reliance upon the
flawed data, EPA identified 444 terminals, collectively
accounting for 65 million barrels of throughput, as having zero
HAP emissions during 1988. It is more appropriate to assume that
between 30 and 71 of these terminals had in excess of l ton of
HAP emissions during 1988.
One commenter (32) stated that EPA appears to have made no
effort to gather real-world information to verify or supplement
its 1992 analysis. The record shows a very limited number of
phone contacts with terminals as well as few site visits. The
group sampled was so small, however, that the only benefit
appears to have been a background understanding of control
technology for EPA staff and consultants. Absolutely no
indication in the Administrative Record implies that the
2-203
-------
resulting information was used to either validate or modify the
database that supports EPA's analysis in this rulemaking.
Response; The Agency recognizes that there are limitations
to the data base generated during the development of these
standards but disagrees with the commenters that these limits
jeopardize the conclusions reached by the Agency. The Agency
believes the most likely effect of the data base limitations are
either an overstatement or understatement of the national impacts
of the final rule.
With the exception of requirements for VMT (see section
2.3.4 of this document), the Section 112 requirements are set at
the MACT floor (i.e., the minimum requirements allowed under the
Act). Therefore, cost impact data do not play a role in the
establishment of these MACT standards. The data on numbers of
affected sources, cost impacts, and control status submitted by
commenters (including more recent data from the Waterborne
Commerce in the United States data base) were used by the Agency
to calculate the MACT floors for the final rule (see Docket
Number A-90-44, Docket Entry IV-B-2 for additional information on
the determination of the MACT floors).
The Section 183(f) requirements in the final rule were
changed to reflect additional data received from commenters
(including Commenters 133 and 139). These data resulted in
higher cost effectiveness for VOC control. Again, national
impacts may be either overestimated or underestimated, but the
decision of the applicability of the final standards was based on
the best available data.
Comment: Two commenters (32, 34) cited the following
example of misstated documents in the record: Table 1 in the
Preamble purports to present the cost of the different RACT
regulatory alternatives. The EPA cites two documents from the
administrative docket as the sources of the information presented
in the table. (Item Nos. II-A-23 and 32.) From a review of the
docket, it is not possible to determine how EPA constructed the
table. The information contained in the documents does not
2-204
-------
correspond to the information in Table 1. The TSD does contain
information similar to that found in Table 1, but the numbers are
different. See TSD Table 4-1 at Page 4-4.
Two commenters (32, 34) cited the following example of
misstated documents in the record: In the discussion of the
determination of the MACT floor, EPA refers the reader to Docket
A-90-44, Item II-A-44 for further information. 59 Fed. Reg.
25014. Yet, this document does not exist in the Docket according
to the April 28, 1994, index obtained in May 1994.
Response: A typographical error was made in referencing the
docket entry for the MACT floor; the correct reference should be
Docket No. A-90-44, Item No. II-A-34. This memorandum was
contained in the docket prior to the proposal of these standards.
The TSD prepared by the Agency during the development of the
proposed standards was published in 1992. In the course of the
rulemaking activities leading up to proposal, there were a number
of changes and revisions to some of the information contained in
the TSD. These changes were included in the docket. The Agency
also notes that these changes did not affect the decisionmaking
process that led to the proposal of the standards or the legality
of the proposed standards.
Comment: One commenter (32) stated that EPA's assumptions
regarding controls on existing terminals are unsupported. The
Administrative Record does not indicate that EPA has taken any
actions to update the information on existing terminals contained
in the 1992 TSD. The 1992 TSD's information on terminals that
are subject to State regulations appears to have been derived
from information provided by Midwest Research Institute (MRI) in
1991, which is contained in the Administrative Record
(Documents II-A-22, ll-A-25 and II-A-26). The only other
substantive reference to State regulations in this portion of the
Administrative Record is an August 27, 1993, letter from MRI
(Document II-A-032) that summarizes the emissions controls
required by the various State regulations. Some comments
2-205
-------
received from terminal operators summarizing the regulatory
requirements in California are also included.
It is not accurate to assume that simply because a State
adopts a control efficiency standard, all sources in that State
have installed vapor-control technology that meets the standard.
Some sources will shut down rather than comply. Some may be
given extensions or will not comply for other reasons. The
assumption that all sources will comply is particularly
questionable when it is being made before the State standard
takes effect, as has occurred in the documents supporting EPA's
Proposed Regulations.
The commenter stated that there is a difference between what
a State requires and what sources achieve, and EPA has ignored
this. The MACT standard for existing terminals must be based on
the level of control achieved by the best performing 12 percent
of the sources in that category or subcategory. The document
supporting EPA's MACT floor determination, Document II-A-034,
states, for example: "Facilities in California and Louisiana are
required to control all VOC emissions from loading operations.
Therefore, the control levels of these regulations ... were used
as the actual HAP reductions for all regulated terminals in these
two States." Docket No. II-A-034 at p. 2 (emphasis added). The
fact is, however, that information in EPA's own Administrative
Record demonstrates the difference between what is required and
what is actually being achieved.
Docket No. II-A-022, a summary of information on terminals
in California, Louisiana and New Jersey, dated July 9, 1991,
includes a table of 33 terminals, only 4 of which are shown as
having on-line and operating controls. The text indicates that
in Louisiana only one terminal had a control system in place, and
four others planned to have systems operational before a
December 1991 deadline. Apparently none of the terminals
discussed handled crude oil, for which Louisiana had a May 1992
deadline. The EPA assumed 22 terminals in Louisiana were meeting
that State's standards.
2-206
-------
Two site visits in New Jersey, both on September 20, 1991,
3 months after the compliance date for that State's regulations,
are reported in Documents II-A-025 and -026. Both facilities
were operating under administrative consent orders and hoped to
conduct compliance tests in October of 1991.2 EPA assumed
24 terminals in New Jersey were meeting that State's standards.
This assumption is surprising, in light of the fact that the
Administrative Record establishes certain New Jersey terminals
had not achieved compliance by the required date, and the record
contains no later information on these facilities' compliance
statuses.
In fact, no indications exist in the record of any effort by
EPA to verify the compliance status of any of the terminals EPA
used to calculate the MACT floor. Other than the MRI documents
referenced above, there are notes from some phone contacts
regarding possible terminal site visits. These contacts are all
from 1991, and none of them references regulatory status.
If EPA had tried to validate its assumption that 181
terminals in States with regulations have installed control
equipment capable of meeting the State standards, it would have
learned that fewer than 60 of these terminals have actually
installed or are installing such equipment. This figure was
provided by API following a canvas of members.
Looking at the States used to establish the MACT floor,
based on contacts with industry sources, regulatory agencies and
equipment suppliers, it appears that only 15 terminals in
California, 17 terminals in New Jersey and between 10 and
16 terminals in Louisiana have installed control equipment. It
is not clear that all of these terminals - particularly the ones
in Louisiana - would be subject to the MACT regulation.
20ne facility later reported that its operational problems
had been corrected and that it was in compliance with New
Jersey's regulations.
2-207
-------
There are several reasons why terminals subject to State
regulations have not installed controls. For example, in
California many terminals used emission offsets or purchased
emission reduction credits rather than installing controls.
Regardless of the reasons, however, it is apparent that EPA's
assumption regarding the level of control achieved by sources
subject to State regulation is incorrect.
Based upon the analysis done for one commenter (89), the
Agency appears to have underestimated the number of terminals
with more than 1 ton of HAP emissions per year by at least
50 percent, if not by more than 100 percent. As a result, the
number of terminals making up the best-performing 12 percent is
between 1-1/2 and 2 times larger than the number used by EPA in
its MACT determination. At the same time, the number of
terminals with controls in the States that have adopted
regulations is much smaller than EPA assumed. As a result, the
level of control achieved by the best-performing 12 percent is
lower than EPA concluded, whether the number of terminals making
up that 12 percent is increased or not. Further, it is
impossible for the public to evaluate whether the MACT standard
was correctly set based on the Administrative Record.
Two commenters (89, 135) stated that their analysis of the
1992 WCUS data base indicates that there has been change in
marine tank vessel loading and unloading activity since 1988.
First of all, fewer total terminals were reported in 1992 than in
1988, with the reduction primarily in fewer small terminals. In
addition, there is a shift towards more marine tank vessel
activity with barges compared to 1988. The Corp of Engineers has
indicated that they believe that the 1992 data base has a more
complete and accurate, overall reporting for waterborne commerce
than the 1988 data base, especially for specific product
throughputs. Because the 1992 data base promises to be more
complete, and because more specific products are outlined in the
data base, we believe that the 1992 data is a more reliable
estimate of the potential number of facilities impacted by the
2-208
-------
proposed EPA requirements. The commenter was able to estimate
the potential number of facilities impacted in the enclosed
reporting using fewer assumptions than was necessary in analyzing
the 1988 data base.
Response; The MACT floors established in the proposed rule
were based on the information available to the Agency at the time
of proposal. The WCUS data base was used primarily as a broad
tool to establish national impacts and the MACT floors for the
standards. The commenter believes that the difference between
the 1988 and 1992 data represents some underlying shift in the
industry rather than short-term market changes or changes in
reporting to the WCUS itself. The commenter did not provide
sufficient data to support their comment.
The Agency has used data submitted by commenters in the
determination of the MACT floor for the final standards. This
determination includes information on the number of controlled
terminals (at least 119 controlled terminals) submitted by one
Commenter 135. However, the commenters did not submit
information to allow the Agency to readily determine the emission
reduction of these "controlled" terminals. The Agency,
therefore, relied on State regulations and emissions estimates to
calculate the emissions reduction of the controlled terminals,
and, in turn, determine the MACT floors for the final standards
(see Docket No. A-90-44, Item No. IV-B-2).
Comment; One commenter (32) stated that the administrative
record does not support EPA's determination of subcategories.
This is a key issue because the selection of subcategories in the
proposed regulations has effectively dictated the outcome of the
MACT floor determination.
Response: The Agency disagrees with the commenter regarding
the ability of the administrative record to support EPA's
determination of subcategories. The commenter did not supply the
necessary data to contradict the Agency's determination of these
subcategories. The Agency's reevaluation of the subcategories
carried out during the development the final rule to respond to
2-209
-------
the public comments has resulted in additional subcategories and
redefinitions of some of the proposed subcategories. The Agency
believes that the decisions for establishing these subcategories
are fully documented in the Administrative record for these
standards (Docket Number A-90-44, Docket Entry IV-B-2).
2-210
-------
2.13 COST EFFECTIVENESS/IMPACTS
2.13.1 Cost Effectiveness
Comment; Several commenters (04, 28, 30, 32, 34, 37, 41,
45, 46, 89) expressed dissatisfaction with EPA's cost-to-industry
estimate of approximately $99,000 per ton of HAP removed. Most
of these comments contend that incorrect or flawed data were used
as the basis for the cost-effectiveness calculations. One of the
most discussed issues involves EPA's estimate of the number of
facilities (terminals) that will be impacted by the MACT
standard.
One commenter (89) estimated (based on the 1992 data base
analysis and using EPA assumptions) that the number of terminals
exceeding the 1-ton HAP threshold, not accounting for current
State/regional controls, would increase from 294 predicted by EPA
in the proposed rule using 1988 data, to 375 in 1992. If
State/regional controls are assumed to be in place, the best
estimate of the number of terminals exceeding the l-ton
threshold, and therefore potentially required to install or
expand controls (i.e., control of all vapor pressure products
loaded) pursuant to the Section 112 proposed rule, is 343 for
1992.
One commenter (34) stated that, based on the 1988 data, EPA
established emission estimates on specific product throughput at
1,667 marine tank vessel loading terminals, assuming
444 terminals had zero HAP emissions. However, when combined,
those 444 terminals account for over 65 million barrels of total
throughput and, therefore, are likely to have emissions greater
than zero. Based on their analysis of average emissions from
such facilities, commenter 34 estimates that between 30 and 71 of
those 444 terminals exceeded 1 metric ton of emissions in 1988.
In addition, EPA estimates that 1,231 out of the
1,667 terminals emitted less than 0.5 metric ton of HAP. This
estimate is also based on specific product throughput data,
rather than total throughput data. Use of the total throughput
data would result in an estimated 128 to 514 terminals exceeding
2-211
-------
the 1 metric ton HAP emission threshold in 1988, depending on
product assumptions. For the smaller facilities emitting
approximately l metric ton of HAP per year, the
cost-effectiveness is likely to exceed $500,000 per metric ton
removed.
One commenter (45) believed the actual number of facilities
that will be required to comply with this regulation will be
nearly 900.
One commenter (34) stated that the result of EPA not being
able to consider the factors listed in Section 183(f) (because of
the listing of marine tank vessel loading and unloading sources
under Section 112(d)) is an extremely expensive, cost-ineffective
regulation for hundreds of marine loading facilities.
One commenter (90) provided data based on EPA's proposed
RACT Alternative II for VOC control that resulted in a cost
effectiveness of $145,000 per Mg of HAP. The commenter stated
that such cost effectiveness is extraordinarily poor and
recommended that the HAP control portion of the proposal not be
promulgated. The commenter suggested that EPA reexamine the
selection of Alternative II over Alternative I for VOC reduction
based on cost effectiveness. The basis for the commenter's
suggestion was the estimate by the Office of Technology
Assessment that the value of VOC control in nonattainment areas
ranges from $25 to $1,600 per Mg.
One commenter (30) presented data estimating the maximum HAP
emissions (excluding oxygenates added to gasoline) from an
offshore terminal platform to be 22.5 Mg total HAP/yr for
conventional gasoline. Applying the minimum 93 percent MACT
control, the resulting total HAP emissions would be 1.6 Mg/yr.
Combined with the annual control costs of $3.6 million for this
facility (calculated previously), the cost effectiveness would be
approximately $170,000 per Mg. The commenter added that the
economics of offshore terminal platforms are quite different from
offshore production platforms. Consequently, the imposition of
2-212
-------
costly marine vapor recovery controls will have significant and
potentially prohibitive economic impacts on these terminals.
One commenter (34), with regards to the cost effectiveness
for controlling low vapor pressure products, presented data (and
a test report) on the calculated cost per ton of HAP removed for
the control of various product types. The commenter stated that
these data showed that the control of products below
approximately 2.0 psia vapor pressure becomes exceedingly
cost-ineffective. The cost for marine vapor control ranges from
$70,000 to over $400,000 per ton for JP-4 (1.3 psia) and
astronomical figures exist for heavy gas oils (up to $36 billion
per ton of HAP removed). The commenter believed only the higher
vapor pressure products above 2.0 psia have cost-effectiveness
values in the range of reasonableness.
One commenter (34) estimated that, for a typical loading
station with a minimum product throughput of 10,000 barrels per
hour, a complete cost for the installation of controls is
$3 million at a minimum. When applied to small operations
emitting in the range of 1 metric ton per year of HAP, the cost-
effectiveness for the proposed requirements would easily exceed
half a million dollars per metric ton of HAP removed.
One commenter (28) stated that a preliminary capital
estimate for controls at a remote facility would be $10 million.
The annual operating cost and indirect expenses are estimated at
$1 million per year. The facility would recover about 5 tons/yr
of HAP at a cost of about $525,000 per ton. The VOC recovery
would cost about $9,300 per ton. These figures far exceed the
guidelines cited in the preamble to the standards and clearly
show that vapor recovery in this case is not cost effective.
One commenter (34) estimated costs between $3 million and
$25 million for nlower-48n systems and installation exceeding
$100 million for marine vapor controls at Alyeska's VMT. This
rule would require that the petroleum industry expend well over
$1.5 billion in capital for marine tank vessel loading terminals.
2-213
-------
An additional $100 million for marine tank vessel retrofits would
also be required.
One commenter (37) currently has marine vapor control
systems at two refineries. At one refinery, a marine vapor
recovery system for a single berth facility was completed in 1992
at a cost of $5 million. At the second refinery, a marine vapor
combustion system for a multiple berth facility was completed in
1993 at a cost of $16 million. The commenter has initiated
design work on a second multiple dock, multiple berth combustion
system at a third refinery; the estimated cost of this system is
in excess of $16 million. The two existing facilities, which
were built to State marine vapor control standards, will not meet
the proposed federal RACT or MACT standards; therefore,
modifications at additional costs to these facilities will be
necessary.
One commenter (133) supplied data that supported an
expansion of the RACT limits based on the "reasonability" of the
control requirements for the Title I standards.
Response: As a result of the Agency's responding to these
and other public comments, the Agency has reduced the number of
terminals required to control emissions to approximately
29 terminals (an additional 32 terminals have already installed
the required control technologies to meet the final standards)3.
The Agency has addressed the commenters' concerns regarding the
costs associated with controlling emissions by including cost
ranges to represent the nationwide impacts. These ranges
represent the Agency's cost estimates (lower range) and
commenters' estimates (upper range). The Agency has also raised
the limits for the applicability of RACT to terminals loading
3Refer to the Regulatory Impacts Analysis prepared for this
final rulemaking for additional information concerning the
impacts associated with this rulemaking; the impacts included in
this document include information submitted by commenters,
including comments concerning the number of controlled terminals
and the costs of installing emissions control equipment.
2-214
-------
10 million barrels or more of gasoline per year or 200 million
barrels or more of crude oil per year. The Agency notes,
however, that with exception of the standards for the Alyeska
Pipeline Service Company's Valdez Marine Terminal, the final
standards represent the MACT floor level of control. In cases
where the MACT floor is the selected MACT for a standard, costs
are not used as a basis for selecting the control technology to
be applied to a group of sources.
The Agency has subcategorized off-shore terminals (see
Section 2.3.2) and agrees with the commenter that the
cost-effectiveness is too high to warrant controls beyond the
MACT floor of no control or for terminals regulated under
Title I. Additionally, the MACT floor for sources emitting
10 tons/yr of any one HAP or 25 tons/yr of total HAP was
reformatted to provide for a vapor pressure limit of 1.5 psia.
Comment; One commenter (32) stated that the cost of a vapor
control system for the VMT would be approximately $120 million
(not including ship modification costs) to comply with the
proposed rule. One commenter (125) stated that the projected
costs of controlling two berths at VMT would be approximately
$92 million. In another submittal, this commenter (126) provided
a cost/benefit analysis for controlling two and three berths at
VMT.
Two commenters (07, 32) stated that the vapor control
process at the VMT will be the most expensive one ever installed
anywhere. According to one commenter (32), the capital expense
to install controls at two berths is $92 million. The additional
cost to install controls to a third berth is $28 million in
capital expense alone. Because 90 percent of throughput will be
loaded from the two controlled berths when averaged over the
remaining life of the terminal, it would cost considerably more
on a per megagram basis to install controls on the third berth.
In fact, it would cost more than three times as much per megagram
to control emissions on the third berth as it would be on the
other two.
2-215
-------
One commenter's (32) analysis showed that the average cost
to remove a megagram of pollutant between compliance and the year
2015 is $239,225 at the two controlled berths and $766,060 at the
third berth. Given its limited use, it would not be cost-
effective to install controls at the third berth. The cost-
effectiveness of controlling the third berth versus the other two
berths should be a major factor in the regulatory equation.
One commenter (32) calculated the cost-effectiveness of a
dollar per megagram ($/Mg) basis for installing controls on VMT's
Berth 3 using the State of Alaska's throughput forecast and the
berth usage contemplated by the proposed vapor control strategy.
Based on these assumptions, the incremental cost effectiveness
for controlling Berth 3 in addition to Berths 4 and 5 equals
$5,995 per Mg of VOC controlled. By comparison, EPA determined
that RACT is not cost-effective at an amount greater than $2,500
per Mg of VOC controlled (see 59 FR 25012). The EPA should reach
the same conclusion for the VMT's Berth 3.
Moreover, EPA should weigh the cost effectiveness in areas
where VOC emissions do not form ozone (for example, Valdez);
there are no benefits from ozone reduction from controlling VOC
emissions from Berth 3. The cost-effectiveness analysis for the
VMT is, therefore, quite different from the cost-effectiveness
analysis for a marine tank vessel loading terminal located in a
nonattainment area where a high cost of control may be justified
given the potential for ozone reduction.
One commenter (80) stated that the review of costs, outputs,
and employment impacts for affected products did not consider
companies that offer portable vapor recovery units for use on
tankers or barges that are not equipped with vapor recovery
systems. Such companies require no capital investment on the
part of the terminal and may be feasible for terminals that load
sporadically. They also may help delay the capital outlay
associated with purchasing equipment.
Response; The Agency has established a subcategory for the
VMT (see Section 2.3.4 of this document). The Agency has
2-216
-------
incorporated the commenter's information into the cost analysis
prepared for this rulemaking. These cost data were used in the
Agency's decisionmaking for the final standards for the VMT
subcategory (i.e., control of the majority of throughput loaded
via at least two of the VMT berths provided that the projected
throughput decline continues).
2-217
-------
2.13.2 Other Cost-Related Impacts
Comment; Two commenters (23, 30) disagreed with EPA's
assertion that many vessel owners will be able to pass forward
retrofit costs in the form of higher transport prices. One
commenter (23) noted that this assertion completely ignores
today's free market economic realities. Today's transport prices
are not representative of any kind of cost pass-through. Margins
have simply narrowed because the vessel owners have had to absorb
the cost impacts in the long term. Therefore, EPA should not
expect the cost impacts of the proposed rule to be simply passed
through as higher transport prices. One commenter (30) stated
that current charter rates often do not cover base operational
costs let alone the costs of financing larger newer tankers and
certainly not the additional costs imposed by this rulemaking.
The EPA should undertake a much more specific and valid
examination of the potential economic impact of this rulemaking
on vessel owners before adopting a final rule.
Response: An economic impact and regulatory flexibility
analysis for this final rule was performed and is included within
the regulatory impact analysis that has been submitted to the
public docket (Docket Number A-90-44, Item Number IV-A-2). The
regulatory flexibility analysis identified two types of
businesses that could incur adverse economic impacts from this
standard, marine terminal operations and marine vessel
operations. With regard to marine vessel operations, the
economic impact analysis considered the majority of these
operations to be small businesses. However, the number of vessel
operations significantly impacted from the proposed standard is
not expected to be substantial. Only a relatively small
percentage of U.S. marine transported throughput will be impacted
by the standard. Excluding crude oil volume shipped by large
tankers from the VMT, no more than one-third of the remaining
U.S. marine transported throughput is expected be impacted by the
standard. It is expected that an even smaller percentage of U.S.
vessels will need to be retrofitted to accommodate the volume of
2-218
-------
affected products. Only the largest and newest vessels (i.e.,
those that will cost least to retrofit) will therefore need to be
retrofitted. Moreover, it is expected that vessel owners will be
able to pass forward most retrofit costs in the form of higher
prices. Vessels that cannot retrofit cost effectively and that
cannot pass through costs can be dedicated to transporting
unregulated products.
Comment: One commenter (28) estimated the additional dock
time required for marine loading vapor control range from
90 minutes to 4 hours. The economic value of a vessel can be
approximated by the demurrage cost currently incurred when there
are unexpected delays. This value can amount to $6,000 per hour.
For a facility that loads 100 ships per year, the economic value
of the lost tanker time could be well over $1 million per year.
Further, for a dock facility currently operating at full
capacity, the delays due to vapor control would cause either lost
throughput or trigger the need to build additional dock
facilities.
Response: The Agency is unable to agree with the commenter.
The commenter did not provide any data to support their claim of
additional impacts due to emissions control.
Comment: In addition to installation at the marine tank
vessel loading terminal, one commenter (34) estimated that
approximately 2,000 marine tank vessels will require retrofit so
that they can connect to marine vapor controls. This possibility
represents a challenge and concern to the industry because, to
the commenter's knowledge, there are currently fewer than
60 marine tank vessel loading terminals with working marine vapor
control systems in the United States and, furthermore, fewer than
20 contractors that have experience with such installations.
Response; The Agency has extended the compliance schedule
for the RACT and MACT standards to afford additional time for the
sources subject to the emission reduction requirements in the
standards and vessels to install control and/or collection
equipment. However, the Agency notes that additional comments
2-219
-------
submitted by commenters show that there are at least 119
controlled terminals currently in operation and that these data
were used in determining the MACT floors for the final standards.
Comment: Three commenters (36, 41, 54) stated that the
proposed requirement relating to MACT are too broad and should
not apply to the numerous terminals owned and/or operated by
independent marketers. The rule would impose significant costs
and operational restrictions on these terminals and yield only
very limited HAP reductions. Such costs, in turn, would place
these facilities at a serious competitive disadvantage, weakening
their financial position and in some instances prohibiting their
operation.
One of the commenters (36) also stated that EPA incorrectly
assumed that many of the terminals that are placed at risk by the
regulations are owned and controlled by major integrated oil
companies with substantial financial resources. Independents are
small and medium-sized companies that are not affiliated with
major integrated oil companies and bring incremental supply to
the market at very competitive prices.
The proposed rule would apply to many of these independent
terminals. It would impose substantial costs and competitive
pressure on them at a time when the reformulated gasoline program
already places them at a severe competitive disadvantage. These
companies cannot afford to incur millions of dollars of
additional costs imposed by this regulation. In many instances,
the terminals will become unprofitable and will not be in a
position to continue operating, thereby reducing competition and
eventually leading to unnecessarily higher petroleum product
prices. The commenters also suggested that EPA narrow the scope
of the rule to minimize these problems.
Response: The Agency notes that the changes in the
subcategories based on HAP emissions (see Section 2.3.1 of this
document) have eliminated control requirements for several of the
smaller terminals. An economic impact and regulatory flexibility
analysis was performed during the development of this final
2-220
-------
rulemaking and is included within the regulatory impact analysis
that has been submitted to the public docket (Docket
Number A-90-44, Item Number IV-A-2). The regulatory impact
analysis noted that, with regard to marine terminal operations,
only the very largest terminal operations are expected to be
affected by this standard. The decision not to require controls
at existing smaller operations greatly reduces the potential for
adverse economic impacts on small terminal operations.
Nevertheless, some of the smaller terminal operations that will
be affected by this regulation could be put under increased
competitive pressure as a result of this rule. Of these
terminals, however, it is expected that few or none are
independently owned. The rest are part of large integrated
petroleum operations. The number of small business terminal
operations affected by this regulation is expected to be minimal.
Comment: One commenter (46) thought the proposed rule would
act as a subsidy for foreign gasoline imports which is in direct
opposition to the Clean Air Act Amendments of 1990. Without
controls on foreign marine loading of product gasoline this
proposal clearly acts as a subsidy to foreign gasoline, refiners
and exporters. Foreign gasoline producers/exporters will not be
making the same expenditures, thereby making their product less
costly since the costs associated with installation/operation of
control equipment will be factored into the cost of the product.
Response: The Act requires EPA to reduce air emissions from
marine vessel loading operations. The Agency is complying with
these requirements by setting national standards for reducing air
emissions at major sources that engage in these operations.
However, the Act does not provide the Agency with the authority
to require controls on similar foreign operations. Any domestic
environmental regulation may cause disparities between domestic
and foreign production costs; however, the commenter did not
supply any data showing that foreign competition will be a
problem in this specific instance.
2-221
-------
Comment; One coiranenter (46) pointed out that past
experience has shown EPA greatly underestimates control costs.
For example, preceding the Benzene Waste NESHAP, the EPA
estimated capital costs to be $250 million, while actual costs
were approximately $2 billion! The actual costs were almost
10 times the EPA estimate! The costs estimated for this proposal
includes the VMT, which accounts for over half of all the annual
HAP emissions from marine tank vessel loading. The EPA
acknowledges that it is not compelled to control marine tank
vessel loading under Section 112. Since Section 183(f)
specifically requires consideration of cost-effectiveness and
since the demonstrated benefits are extremely low compared to
costs, EPA should regulate solely under the Section 183(f)
parameters.
Response; The Agency has addressed comments regarding its
decision to regulate under Title III in Section 2.1.3 of this
document. The Agency has included commenter-supplied cost data
in its determination of the nationwide cost impacts attributable
to this final rulemaking.
Comment; One commenter (58) was concerned about the effect
that air quality regulations will have on the future economic
viability of future oil production from Alaska's North Slope.
The North Slope's future is influenced by the continued economic
viability of the Trans Alaska Pipeline System (TAPS). The effect
of environmental regulations on transportation costs could have
chilling effect on new exploration in North Slope fields. The
continued economic viability of TAPS is important to the Alaska
Natives living in the North Slope Region and to the State of
Alaska as a whole. The rural villages of the North Slope Region
depend upon State funding for sewer and sanitation facilities,
education, and housing. About 85 percent of Alaska's budget
derives from North Slope royalties and tax revenues.
One commenter (67) believed that North Slope operations are
so profitable--even at current oil prices—that there are no
financial barriers to requiring full vapor recovery.
2-222
-------
Response; The Agency has not received any information from
the Alyeska Pipeline Service Company indicating that vapor
control costs threaten the viability of the pipeline.
2-223
-------
2.14 MISCELLANEOUS
2.14.1 Private Citizens (Groups)
Comment: Nineteen commenters (01, 09, 11, 12, 13, 14, 15,
16, 17, 18, 19, 61, 62, 63, 64, 65, 66, 67, 81) submitted letters
in support of EPA's proposed standards for marine tank vessel
loading and unloading operations. All of the commenters live or
work near Valdez, Alaska and Alyeska's VMT. The comments ranged
from "concerned" to "furious and outraged" regarding Alyeska and
its impact on the health of nearby residents, workers (including
Alyeska employees), and the environment.
Response: The Agency is grateful for the commenters' input
and support of the regulations. Specific suggestions made
regarding needed changes in the proposed rule were thoroughly
considered. Where appropriate, changes have been made to the
final rule.
2-224
-------
2.14.2 Publication of Regulation in the Federal Register
Comment: Twelve commenters (04, 10, 23, 29, 34, 50, 68, 71,
73, 75, 78, 79) indicated that the proposed regulation should be
published in the Federal Register.
One commenter (04) questioned the legality of proposing
regulations without inclusion of the regulatory language in the
Federal Register. Two commenters (04, 73) stated that the Clean
Air Act as amended on November 15, 1990, clearly requires EPA to
publish "...notices of proposed rulemakings... in the Federal
Register as provided under section 553(b) of title 5 [of the
United States Code]." Two commenters (73, 78) indicated that
EPA's use of the TTN to communicate proposed regulations is
inconsistent with the spirit, if not the letter, of the
Administrative Procedures Act (APA). One commenter (73) cited
several court cases regarding the APA; while the APA does not
require the notice of proposed rulemaking to contain actual
wording, the notice should allow all parties reasonable
opportunity to participate in the rulemaking process. The
commenter cited additional court cases where questions arose
regarding whether or not the final rule was a logical outgrowth
of the proposed rule. Without the published regulatory language,
it is difficult to differentiate between changes that result from
public comment and those that reflect a change in EPA's intent.
Four commenters (34, 50, 68, 75) stated Federal Register
publication assures complete public access to the proposed rules
and provides a consistent means to reference particular pages of
the proposed rule. Another commenter (71) stated that failure to
publish proposals interferes with industry's right of review and
comment. Two commenters (73, 78) stated that including the
entire regulatory language in the Federal Register allows
effective public comment.
Three commenters (23, 73, 75) are concerned that EPA's
practice of making changes to the electronic version on the
bulletin board after proposal does not provide adequate notice.
Two of these commenters (73, 75) mentioned that EPA made changes
2-225
-------
after proposal to the regulatory text on the bulletin board for
the proposed version of Section 112(g). One commenter (73)
indicated that changes made after the proposal date result in
parties commenting on different versions of the regulatory
language.
Five commenters (04, 29, 34, 50, 73) noted that the
regulatory language is only available upon request (i.e., from
the docket or EPA contact) or through the EPA electronic bulletin
board. One commenter (79) stated that requesting a copy of the
regulation from EPA through the mail takes time out of the
already limited comment period. Two commenters (04, 73) pointed
out that since the only docket is located in Washington D.C., not
all parties subject to this proposal who have need of materials
from the docket have ready access to it.
Five commenters (04, 34, 73, 75, 79) pointed out that all
parties do not have the capability (i.e., necessary hardware and
software) of accessing the electronic bulletin board. One
commenter (75) stated that significant phone charges are incurred
from downloading large files associated with major rules. Six
commenters (04, 23, 29, 71, 73, 78) stated that formatting
problems and difficulties occur with the text and graphics
downloaded from the electronic bulletin board. Three commenters
(71, 73, 78) indicated that formulas in § 63.565 of the proposed
rule (the electronic version) were scrambled or were not
received.
Two commenters (23, 71) stated EPA should both publish
rulemakings in the Federal Register and post the rulemaking
package electronically on the TTN. One commenter (79) stated if
something must be omitted from the Federal Register, the actual
regulation should be printed and the preamble left out; those who
need detailed information on the development of the regulation,
as contained in the preamble, can then make efforts to obtain the
preamble.
One commenter (75) noted that for air rules appearing in
both the Federal Register and the TTN, the TTN version is
2-226
-------
typically available in advance of the version published in the
Federal Register; early availability is helpful and of
significant benefit.
Response; The Agency has reviewed its responsibility to
adequately inform the affected public of proposed actions. The
decision to reduce the amount of printed material in the
Federal Register and to assure that the material, including the
proposed regulatory text of the proposed rule, is accessible for
public comment and judicial review does not conflict with the
statutory requirements of the Administrative Procedures Act
(APA), the Federal Register Act (FRA), or the Clean Air Act
Amendments of 1990. In the preamble to the proposed rule, the
Agency noted that the proposed regulatory text was available and
would be sent free of charge by the contact person upon request.
Access to material that is used as the basis of the proposed rule
(officially located in the Air Docket created by the CAAA) is
identified in the preamble to the proposals and promulgations of
rules. Specifically, the Agency clearly established and will
continue to look for additional connections and will include
directions for obtaining the text of information not printed in
the Federal Register. Currently, this information may be
obtained through one of the following sources: (1) the TTN's
"Recently Signed Rule" bulletin board; (2) directly from the Air
and Radiation Docket and Information Center; (3) distribution to
trade associations; (4) plaintiffs in court-ordered regulatory
actions; (5) contact with small business ombudsman system in each
State; and (6) if necessary, through the contact person at the
Agency. The response to this approach has been positive as the
process has aged.
The Agency believes that all information that is developed
in the course of the development of a proposed and final rule is
important; however, EPA believes they have realistically and
responsibly addressed the need to publish information in the
Federal Register. The Agency will continue to review the issue
of extensive publishing in the Federal Register along with its
2-227
-------
responsibility to adequately inform affected parties of our
proposed and final actions.
2-228
-------
2.14.3 National Security
Comment; One commenter (03) did not believe the regulation
would impact Department of Defense (DOD) bulk petroleum
terminals; however, the commenter expressed concern about any
vapor control technology requirements that would preclude DOD
ships (owned, operated, or chartered) from obtaining fuel at
refineries and commercial terminals. The DOD access to
commercial petroleum facilities is critical to national security.
DOD petroleum tankers and military vessels do not incorporate
vapor control technologies. Backfitting of these ships is
inconsistent with the cargoes carried and, especially for combat
support ships, incompatible with ship missions and safety
criteria. The commenter is evaluating the impact of this rule on
DOD fueling activities. The commenter asked for the opportunity
to readdress these issues if an assessment demonstrates that RACT
or MACT are applicable to DOD bulk petroleum facilities or that
the regulation restricts DOD access to commercial petroleum
facilities.
Response: The final rule imposes requirements for vapor
control equipment primarily upon loading terminals. Marine tank
vessels are only required to demonstrate vapor tightness and
compatibility with shore-side control equipment. This
requirement should require little backfitting of DOD ships and
should enhance shipboard safety.
Comment: One commenter (74) requested that its Nederland
Terminal be excluded from the provisions of the NESHAP in the
case of loading and unloading and emissions resulting from an
emergency drawdown of the Strategic Petroleum Reserve (SPR).
Under normal operating conditions, loading, unloading, and
emissions thresholds at the terminal would not trigger the
provision of this rule. Congress passed the Energy Policy and
Conservation Act (Public Law 94-163), which authorizes the
development of crude oil and petroleum products storage
facilities, to diminish the United States' vulnerability in the
event of a severe energy supply interruption and to carry out
2-229
-------
obligations of the United States under the International/Energy
Program (IEP). To assist in meeting this objective, the act
created the SPR administered by the U.S. Department of Energy
(DOE). Current SPR capacity is 750 MMBBL. The commenter has
contracted with the DOE to deliver crude oil from the SPR to
markets on the Texas and Louisiana Gulf Coast if an emergency
drawdown is ordered by the President of the United States.
During emergency drawdown conditions, the commenter's terminal
would be required to load up to 250 million barrels of crude oil
across its docks onto marine tank vessels over a 240 day period.
Although it is unlikely that a drawdown on the SPR will occur, it
is likely that a severe energy supply interruption necessitating
a drawdown of the SPR would be a one time occurrence.
Response: The Agency agrees with the commenter that the
referenced terminal should not be required to control emissions
under the final standards. The Agency believes that this
terminal would not be subject to the emission reduction
requirements found in the final rule because its HAP emissions
are less than 10 tons/yr of all individual HAP and less than
25 tons/yr of total HAP; furthermore, its throughput is less than
the throughput thresholds of the RACT standards. If at some
future time, the facility would exceed these emissions levels
(because of an emergency drawdown of the SPR), the President
could provide a waiver based on the national security nature of
the drawdown per Section 112(i)(4) of the Act.
2-230
-------
2.14.4 Other Miscellaneous Comments
Comment: One commenter (38) stated that many boilers and
heaters employ staged combustion burners that may lower the
overall temperature and lengthen the flame zone of the combustion
section. Therefore, the rule should require that the vented
VOHAP stream be introduced into the combustion flame zone as
close to the burner fuel inlet as possible and still maintain
good mixing.
Response; The commenter did not submit any data that would
allow the Agency to determine the validity of the commenter's
claim regarding the lowering of the overall temperature and
lengthening of the flame zone of the combustion section. The
Agency has therefore not included the commenter's suggestion in
the final rule.
Comment; One commenter (21) recommended that EPA promulgate
the final rule for this source category by November 15, 1994.
Promulgation by the scheduled date is important to those who will
have the obligation of implementing and enforcing the NESHAP
standards and requirements for affected sources. Conversely, one
commenter (34) stated that EPA should postpone promulgation of
the Section 112 rule until the year 2000 or at least until the
Agency has reason to believe that facilities will be able to
achieve compliance within 3 years from a new promulgation date.
Response: As a result of a Clean Air Act litigation suit,
Sierra Club v. Browner, the proposal and promulgation dates for
several standards were agreed upon in a consent decree. The
federal rule and NESHAP for marine tank vessel loading operations
was included in this consent decree, and the court-ordered
deadline for promulgation of this NESHAP is July 28, 1995. The
EPA will promulgate this rulemaking on schedule.
Comment; One commenter (59) supported EPA's approach to
processing of the vapors; "... it is impractical for marine tank
vessels to carry their own vapor processing systems given the
limited space on individual vessels."
Response: The Agency agrees with this comment.
2-231
-------
Comment: One commenter (06) stated that vapor-tightness
requirements similar to § 63.563(a) should be made applicable to
storage tanks that operate as part of a vapor balance system to
control emissions from marine tank vessel loading. The maximum
pressure setting requirements and leak checking requirements will
be equally as important for storage tanks in vapor balance
systems as for the ships themselves. Alternatively, EPA should
consider requiring pressure monitoring devices on remote storage
tanks that cannot be easily checked for leaks.
Response; Such storage tanks are covered by other
regulations and are not addressed in this regulation. Sources
with vapor balancing systems are not subject to the emission
reduction requirements found in the final rule, but are required
to maintain the closed vent system (i.e., vapor collection
system) consistent with Sections 63.562(b)(1), (c)(2), and
(d) (1) .
Comment; Two commenters (76, 80) stated that the scope of
the rule should be expanded to limit potential emissions during
transport. One of these commenters (76) stated that weather
changes from night to day causes in-breathing and out-breathing,
which causes pressure to develop on the barge tanks. With no
control of the pressure, one can expect 3 to 10 percent of an
empty barge's volume (approximately 56,000 cubic feet) to
exchange with the atmosphere. Typical daytime to nighttime
temperature changes can cause additional emissions. Improving
barge tightness could reduce these emissions significantly.
Certification of barge tightness would be best performed by
testing for leaks at each opening as opposed to monitoring
pressure loss over time.
One commenter (80) did not support the provision to allow
loading of nonvapor-tight vessels under the conditions specified
in Section 63.563(a)(2)(iii)(B). Vapor tightness should be
required for all vessels, independent of the control level at the
marine tank vessel loading terminal (even if the terminal is
subject to control requirements). Fugitive leaks from non vapor-
2-232
-------
tight vessels are not only a problem during loading and
unloading, but also result in emissions throughout the transport
trip of the vessel. If EPA does not feel it is appropriate to
include such a standard under this rulemaking, organic
liquid/vapor cargo tanks (marine or otherwise) should be listed
as a source category under 112(k), and vapor-tightness standards
should be established.
Response: The regulation addresses vapor tightness only for
loading operations. The suggested regulation of vessels in
transit would be covered under mobile sources.
Comment: One commenter (77) provided a copy of a document
that was referenced in a previously submitted comment but not
included with that comment submittal.
Response; The Agency appreciated the commenter's submittal
of this information. The referenced document was reviewed by the
Agency when responding to all of the commenter's submittals.
Comment; One commenter (122) requested that the Agency
extend the comment period to allow the commenter to conduct a
more extensive comment preparation and data base analysis than is
currently allowed.
Response; The Agency did not elect to extend the comment
period but has accepted late comments and additional data
(including the analysis performed by the commenter) and has
considered these submittals during the development of the final
rulemaking. In addition, the Agency has reopened the comment
period on two occasions to request additional comment on specific
issues.
Comment: One commenter (124) encouraged the control of
emissions from VMT so that economical electrical energy from
vapor burning operations would be possible.
Response: The Agency encourages the recovery of energy as a
byproduct from any air pollution control device but does not
favor the selection of one control technology over another, given
that these standards are based on efficiency and not specific
equipment.
2-233
-------
Comment; One commenter (128) reviewed the applicability of
the New Source Performance requirements to VMT.
Response; The Agency has no response to this comment since
the requirements specified in the comment are implemented by a
State government.
2-234
-------
2.15 COMMENTS ON PROPOSED APPENDICES TO 40 CFR PART 64
Note: The Agency has superseded performance
specification 101 (PS-101) and performance specification 102
(PS-102) with performance specifications 8 and 9 respectively.
This action was promulgated along with the Magnetic Tape
Manufacturing NESHAP on December 15, 1994 (59 FR 64580). A copy
of this notice is available in the docket (Docket No. 1-90-44,
Item No. IV-1-4.
Comment; One commenter (31) indicated that the requirement
in PS-101 stating that at least 90 percent of the organic
components must be determined will sometimes be unnecessary and
result in noncompliance. In practice (due to the complex
compositions in some process vent streams) it may be impossible
to identify 90 percent of the organic components in the stream
using Method 18. The owner or operator must know all of the
components of the effluent stream in order to determine when
90 percent of them have been identified. This requirement is a
real problem, for example, in complex hydrocarbon streams that
may contain 20, 30 or 40+ individual chemical components. It
would be difficult to prove to enforcement personnel that this
requirement has been achieved, especially when the majority of
the components will only be present in low concentration levels.
Other problems will be associated with emissions of
components that have either high molecular weights or low vapor
pressures (at either stack or instrument conditions) or compounds
that may polymerize before analysis. These materials cannot be
measured by Method 18, yet they could be a significant portion of
the VOC components in the effluent stream.
Other information may exist (such as general information on
the chemistry of the process or previous studies conducted by
methods other than Method 18) that will help identify organic
components. Therefore, the relevant portion of PS 101 (4.1)
should be amended by the following:
Determine all of the organic components in the effluent
stream which can be reasonably identified using
2-235
-------
Method 18 (40 CFR Part 60, Appendix A), process
chemistry, or previous studies.
Response: The Agency agrees with the commenter. The
referenced requirements are not contained in PS-9.
Comment; One commenter (31) stated that Method 18 may not
be adequate in all situations and that the reference to § 4.1 of
PS 101 (40 CFR Part 64, Appendix A) should be amended to allow
some flexibility in characterizing the make-up of the process
vent stream. The complex compositions in some process vent
streams may make it impossible, in practice, to identify
90 percent of the organic components in the stream using
Method 18. The owner or operator must know all of the components
of the effluent stream in order to confirm when 90 percent of
them have been identified. This requirement is a real problem,
for example, in complex hydrocarbon streams that may contain 20,
30 or 40+ individual chemical components. It would be difficult
to prove to enforcement personnel that this requirement has been
achieved, especially when the majority of the components will
only be present in low concentration levels. Other problems will
be associated with emissions of components that have high
molecular weights, low vapor pressures (at either stack or
instrument conditions), or compounds that may polymerize before
analysis. Such materials cannot be measured by Method 18, yet
they could be a significant portion of the VOC components in the
effluent stream. Additionally, there may be other information
(such as general information on the chemistry of the process, or
previous studies conducted by methods other than Method 18) that
will help identify organic components. Therefore, the relevant
portion of PS 101 (4.1) should be amended by the following:
Determine all of the organic components in the effluent
stream which can be reasonably identified using Method 18
(40 CFR Part 60, Appendix A), process chemistry, or previous
studies.
Response; The Agency agrees with the commenter.
Performance Specification 8 cites the method required in the
applicable regulation. Additionally, the language requiring the
2-236
-------
determination of all organic components has been removed from
PS-8.
Comment; One commenter (71) noted that PS-101 and PS-102 of
Part 64 Appendix A for VOC Monitors are technically faulty
because these unapproved and unproven methodologies are not
applicable to every case and contain technical problems.
Recommendations:
1. Incorporate the specific recommendations from
Attachment A (CMA's proposed Implementation of MACT/Incorporation
of Section 112(d) MACT into Title V Permits as presented in a
letter dated May 26, 1994, from T. Ted Cromwell, CMA, to
L. Wegman, USEPA/OAQPS) into PS-101 and PS-102;
2. Revise 63.563(b)(3) as follows: (revisions in bold)
... In cases were VOC concentration at the outlet of a
combustion control system is the monitored operating
parameter during the performance test, the source shall
follow PS-101 and PS-102, of Appendix A of Part 64, ....
3. Revise last sentence of 63.564(b) as follows:
...The CEM's must be in operation whenever the carbon
absorber is in operation, (delete "and shall meet PS-101 and
PS-102, as appropriate, of Appendix A of Part 64 of this
chapter").
Response; The Agency disagrees with the commenter that PS-8
and PS-9 are faulty. These performance specifications were
developed based on technology currently being used and following
criteria achievable with currently available instrumentation.
Comment; One commenter (31) proposed sampling at two
concentrations (high and low) for the daily tests to determine
the 7-day calibration error (CE), rather than the three
concentrations daily tests required by PS-101 and PS-102. The
commenter stated that the requirement for three concentrations
increases the cost of the test procedure without adding
proportionate value. Conducting the test at high and low
concentrations would provide an estimate of calibration error at
the portions of the calibration curve that are most likely to
deviate from the actual value. Thus, it is highly unlikely that
2-237
-------
the true CE (even if sampled at an infinite number of
concentrations) would ever exceed the value determined by
sampling high and low concentrations during the 7 days of the
test. Three level tests are more appropriate for the quarterly
or annual tests like those specified for gas OEM's in 40 CFR
Part 60, Appendix B and Appendix F.
The EPA should use the procedure described in Appendix C
Part 4.1, which is referred to in these sections. That procedure
requires only two (high and low) concentration points to be
tested. Therefore, the relevant portion of PS-101 (4.8) should
be amended by the following:
... at the low- and high-calibration levels....
The relevant portion of PS-101 (6.1.1) should be amended by the
following:
... following section 4.1 of Appendix C of this part,
The relevant portion of performance specification 102 (4.6)
should be amended by the following:
... at the low- and high-calibration levels....
The relevant portion of performance specification 102 (6.1.1)
should be amended by the following:
... following section 4.1 of Appendix C of this part ....
Response: The Agency disagrees that two concentration
measurements are sufficient for calibration purposes. For VOC
instrumental techniques, three point comparisons are necessary to
demonstrate instrument linearity over the applicable range. This
practice is common QA/QC criteria. Furthermore, the commenter
did not submit any data to allow the Agency to validate the
adequacy of using only two measurements.
Comment; One commenter (31) stated the specified level of
calibration error (5 percent) in performance specification 102 is
too stringent considering the best available technology today.
The ability to measure low concentrations of organics in process
vents will depend upon the analyte, its concentration,
interferences (matrix effects), and the capability of the best
2-238
-------
available on-line analytical instruments. In many cases
5 percent will be unachievable with present technology.
For example, the best technology available may not be able
to completely resolve two components in the emission stream.
Poor resolution can easily degrade the performance of the monitor
to a level well above the 5 percent level. Also, although a
component may be a major portion of the volatile organics in the
effluent stream, it will most likely only be a very minor
component of the total stream, therefore, the component would be
close to the detection limit of even the best available
technology.
The acceptable calibration error should be determined on a
case-by-case basis. Therefore, the relevant portion of PS 102
(4.6) should be amended by the following:
For the initial 7-day CE test, the CEM's calibration
response must not differ by more than the relevant performance
specification agreed to by the owner/operator and the permitting
authority from the calibration gas value at each level after each
24-hour period.
Response; The calibration determination procedure listed in
PS-9 produces a value that indicates reasonable performance for
instruments currently in use for monitoring purposes.
Comment: One commenter (31) stated that the reporting
requirements specified in the last sentence of PS 102, add too
much paperwork to the permitting process. The specification
would require the commenter to send the permitting authority box
after box of calculations, data records, etc., regardless of
whether the permitting authority wanted them. A summary of the
test results of the test should be sufficient for the permitting
process. Supporting data should be available upon request.
Therefore, the relevant portion of PS 102 should be amended by
the following:
All data sheets, calculations, CEM's data records, and
cylinder gas or reference material certifications shall
be made available to the permitting authority upon
request.
2-239
-------
Response: Performance Specification 9 defers to the
reporting requirements of the applicable subpart regarding the
submission of calculations, datasheets, CEM's data records, and
cylinder gas or reference material certifications. The reporting
requirements of § 63.567 do not require the submission of this
material unless requested by the Administrator or permitting
authority.
Comment: One commenter (31) stated that requiring a linear
regression r2>0.995 in PS-102 section 4.7 eliminates the
possibility of other types of acceptable calibration techniques;
therefore, this section should be modified to allow other
techniques.
In some cases, linear regression analysis is not adequate
for the calibration. Other calibration techniques would be
preferred [e.g., if using the three levels, an rSo.995 cannot be
achieved; using more standard levels will allow the use of
multi-variate calibration techniques such as partial least
squares (PLS)] to fit the data adequately. Some of these
nontraditional calibration techniques (e.g., PLS) do not generate
an r2 value. Performance requirements should not limit the
technique used to obtain acceptable performance. Other
techniques do produce various "goodness of fit" indicators
(e.g., P15 uses a Standard Error of Prediction), which can be
used in the same way that the r2 is used. Therefore, the
relevant portion of Appendix A, PS 102, Section 4.7 should be
modified as follows:
... The calibration model for each organic compound shall
have a "goodness of fit" parameter value equivalent to an
r2>20.995.
Response: The Agency believes that the analysis
requirements found in PS-9 are reasonable and assure data quality
as these data relate to instrument linearity.
Comment; One commenter (31) stated the last sentence in
Appendix B (3) is unclear and should be clarified. It says:
2-240
-------
These specifications are guidelines, except for those cases
where reference method (RM) tests are not required.
To the commenter, this statement means that if RM tests are
required, the specifications in Appendix B(3) are "guidelines,"
rather than legally binding requirements. The question is, what
are these specifications in cases where RM tests are not
required? One commenter (31) recommended the following change to
clarify the meaning:
These specifications are guidelines in those cases where a
reference method (RM) test is required. In cases where no
RM test is required these specifications do not apply.
Response: The requirements of PS-8 and PS-9 clarify that
the methods are to be used consistent with the requirements of a
specific subpart (in this case, Subpart Y). Provisions in
Subpart Y and in the incorporated Subpart A exist for owners or
operators of affected sources to use alternative methods
following application and approval by the Administrator.
Comment: One commenter (31) stated that the EPA should not
specify temperature programmable gas chromatographs for gas
chromatographic continuous emission monitoring systems.
In proposed PS-102, EPA specifies that the gas chromatograph
(GC) must be "temperature programmable." "Temperature
programmable" is a term used to denote the ability of the GC oven
to accurately raise the oven temperature during an analysis to
achieve adequate separation of otherwise unresolved components in
the sample. The resolution of components in a sample is highly
dependent on the nature of the components and the sample matrix
and, therefore, is dependent on the particular emission unit. In
many cases, complete separation can be achieved simply by holding
the oven temperature at a constant value. This is called
"isothermal operation" as opposed to "temperature programmed."
A significant cost increase exists for a temperature
programmable GC over an isothermal GC. If the analytical method
can achieve the enhanced monitoring protocol requirements of
representative, reliable, frequent, and timely results with
isothermal operation, then there is no need for a more expensive
2-241
-------
"temperature programmable" GC. Therefore, the relevant portion
of Appendix A performance specification 102, Section 2.1
(58 PR 54694, col. 2) should be amended as follows:
... the gas concentration. (deletes the last sentence: "The
GC must be temperature programmable.")
The relevant portion of Appendix A performance
specification 102, Section 4.1 (58 FR 54694, col. 2) should be
amended by the following:
... separation column, oven, and detector. (deletes
"temperature programmable" requirement for the oven)
Response: The Agency agrees with the commenter.
Performance Specification 9 requires that the gas chromatograph
(GC) must be "temperature controlled" which means possessing the
ability to maintain a given temperature. A temperature-
programmable GC is not required for this performance
specification as long as all other requirements for precision,
linearity, and accuracy listed in PS-9 are met. The Agency notes
in Section 2.1 of PS-9 that a temperature programming GC will
typically speed up peak elution, thus allowing increased sampling
frequency.
2-242
-------
I. REPORT NO.
453/R-95-014
4. TITLE AND SUBTITLE
Federal Standards for Marin
and National Emission Stan<
for Marine Tank Vessel Loa
Document for Final Standar
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
2.
e Tank Vessel Loading Operations
lards for Hazardous Air Pollutants
iding Operations-Technical Support
ds
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME Al
Midwest Research Institute
401 Harrison Oaks Bouleva
Gary, NC 27513
>1D ADDRESS
ird
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
July 1995
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
6503-93-02
10. PROGRAM ELEMENT NO.
1 1 . CONTRACT/GRANT NO.
68-D1-0015, WA 93
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
Project Officer is David Markwordt, Mail Drop 13, (919) 541-0837
16. ABSTRACT
Federal standards and national emissions standards to control emissions of VOC and HAP from new and
existing marine tank vessel loading operations are being promulgated under Section 183(f) and
Section 112 of the Clean Air Act. This document contains information on the background and authority,
a summary of public comments and Agency responses, and a summary of changes to the regulation
following proposal and the resulting environmental and economic impacts.
17.
KEY WORDS AND DOCUMENT ANALYSIS
». DESCRIPTORS
Air pollution
Air pollution control
National emissions standards
Federal standards
Volatile organic compounds
Hazardous air pollutants
Marine tank vessel loading industry
Marine terminals
18. DISTRIBUTION STATEMENT
Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group
Air pollution control 13B
Marine Tank Vessels
Stationary sources
19. SECURITY CLASS (Report) 21. NO. OF PACES
Unclassified 260
20. SECURITY CLASS (Page) 22. PRICE
Unclassified
EPA Form 2220-1 (Rev. 4-77)PREVlOUS EDITION IS OBSOLETE
------- |