United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-453/R-99-006
September 2000
&EPA
CONSOLIDATED FEDERAL AIR RULE FOR THE
SYNTHETIC ORGANIC CHEMICAL
MANUFACTURING INDUSTRY - BACKGROUND
INFORMATION FOR PROMULGATED
STANDARDS
-------
EPA-453/R-99-006
Consolidated Federal Air Rule
for Synthetic Organic Chemical
Manufacturing Industry
Background Information for
Promulgated Standards
Emission Standards Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
U.S. Environmental Protection
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Floor
Chicuo, IL 60604-3590
September 2000
-------
-------
CONSOLIDATED FEDERAL AIR RULE FOR
SYNTHETIC ORGANIC CHEMICAL
MANUFACTURING INDUSTRY
BACKGROUND INFORMATION FOR
PROMULGATED STANDARDS
Disclaimer: This report is issued by the Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency.
Mention of trade names or commercial products is not intended to
constitute or endorsement or recommendation for use. Copies of
this report are available free of charge as supplies permit from
the Library Services Office (MD-35), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711, (919) 541-
2777, from the National Technical Information Service, 5285 Port
Royal road, Springfield, Virginia 22161, (703) 487-4650, or from
the internet (http://wvrw.epa.gov/ttn/).
11
-------
Environmental Protection Agency
CONSOLIDATED FEDERAL AIR RULE
FOR SYNTHETIC ORGANIC CHEMICAL
MANUFACTURING INDUSTRY
BACKGROUND INFORMATION FOR
1 PROMULGATED STANDARDS
Prepared by:
Rick Colyer^ '/ Date
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
1. The promulgated consolidated Federal air rule provides a
compliance option to several standards previously promulgated
under Sections 111 and 112 of the Clean Air Act. It acts to
consolidate similar rules that apply to the same types of
emission points at an industrial site, thereby eliminating
differences among these rules and reducing burden. One of
President Clinton's and Vice President Gore's reinventing
environmental regulation initiatives directs EPA to consolidate
Federal air rules, so that Federal air rules for a single
industry would be incorporated into a single rule. This rule
would consist of "...one set of emission limitations, monitoring,
and recordkeeping and reporting requirements."
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; State and Territorial Air Pollution Program
Administrators,-' EPA regional Administrators; Local Air Pollution
Control Officials; Office of Management and Budget; and other
interested parties.
3. For additional information contact:
Mr. Rick Colyer
Office of Air Quality Planning and Standards (MD-13)
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
Telephone: (919) 541-5262
111
-------
4. Copies of this document may be obtained from:
U.S. Environmental Protection Agency Library (MD-35)
Research Triangle Park, N.C. 27711
Telephone: (919) 541-2777
National Technical Information Service
5285 Port Royal Road
Springfield, Virginia 22161
Telephone: (703) 487-4650
IV
-------
LIST OF ACRONYMS
Act Clean Air Act
CAR Consolidated Federal Air Rule
CEMS Continuous emissions monitoring systems
CFR Code of Federal Regulations
CMPU Chemical manufacturing process unit
CPMS Continuous parameter monitoring system
EFR External floating roof
EPA Environmental Protection Agency
FR Federal Register
HAP Hazardous air pollutant
HON Hazardous Organic NESHAP
IFR Internal floating roof
L/G Liquid-to-gas
MACT Maximum achievable control technology
NESHAP National emission standard(s) for hazardous air
pollutants
NSPS New source performance standards
PPM Parts per million
PPMV Parts per million by volume
PSIG Pound per square inch gauge
QIP Quality improvement plan
RACT Reasonably available control technology
RCRA • Resource Conservation and Recovery Act
SCU SOCMI CAR unit
SIP State implementation plan
SOCMI Synthetic organic chemical manufacturing industry
SSM Startup, shutdown, and malfunction
TRE Total resource effectiveness
v
-------
TABLE OF CONTENTS
1.0 INTRODUCTION 1-1
2 .0 GENERAL 2-1
2.1 CAR OBJECTIVES AND BURDEN REDUCTION 2-2
2.2 SCOPE 2-7
2.2.1 Incorporating or Allowing Other Rules
in the CAR 2-7
2.2.2 Miscellaneous Scope Comments 2-12
2.3 APPLICABILITY 2-14
2.3.1 SOCMI CAR Unit 2-14
2.3.2 Pointer Paragraphs in Referencing Subparts 2-18
2.4 STRUCTURE AND FORMAT 2-22
3.0 SPECIFIC COMMENTS BY SUBPART 3-1
3.1 GENERAL PROVISIONS '.'.... 3-1
3.1.1 Definitions 3-2
3.1.2 Startup, Shutdown, and Malfunction .... 3-13
3.1.3 Reporting 3-21
3.1.4 Compliance 3-26
3.2 STORAGE VESSELS 3-29
3.3 PROCESS VENTS 3-32
3.4 TRANSFER RACKS 3-40
3.5 EQUIPMENT LEAKS 3-43
3.5.1 Leak Detection 3-43
3.5.2 Leak Repair 3-45
3.5.3 Delay of Repair 3-48
3.5.4 Valves--Difficult- and Unsafe-to-Monitor . 3-48
3.5.5 Valves--Subgrouping and Monitoring
Frequency 3-50
3.5.6 Valves--Other Comments 3-52
3.5.7 Pumps--Percent Leaking Pumps Calculation . 3-54v
3.5.8 Pumps--Visual Inspections 3-56
3.5.9 Pumps--Other Comments 3-58
3.5.10 Connectors--Exemptions to the Connector"
Standards 3-61
3.5.11 Connectors--Other Comments 3-64
3.5.12 Monitoring Instrument Procedures 3-65
3.5.13 Compressors 3-67
3.5.14 Sampling Connection Systems 3-68
3.5.15 Enclosed-Vented Process Units 3-69
3.5.16 Batch Product--Processes 3-70
3.5.17 Periodic Report Contents 3-71
3.5.18 Alternatives and Exemptions 3-73
3.5.19 Other Equipment Leak Comments 3-73
VI
-------
TABLE OF CONTENTS
3.6 CLOSED VENT SYSTEMS AND CONTROL DEVICES 3-75
3.6.1 Performance Tests 3-75
3.6.2 Control Requirements 3-77
3.6.3 Monitoring, Recordkeeping, and Reporting . 3-81
4.0 IMPLEMENTATION AND TITLE V 4-1
4.1 IMPLEMENTATION 4-1
4.2 TITLE V 4-4
5.0 CHANGES TO THE EQUIPMENT LEAKS REFERENCING SUBPARTS . .5-1
5.1 CLARIFICATION OF INTENT 5-1
5.2 CLARIFYING OR TYPOGRAPHICAL EDITS 5-6
6.0 MISCELLANEOUS 6-1
VII
-------
1.0 INTRODUCTION
On October 28, 1998, the U.S. Environmental Protection
Agency (EPA) proposed the "Consolidated Federal Air Rule (CAR):
Synthetic Organic Chemical Manufacturing Industry" (63 FR 57748) .
The Consolidated Federal Air Rule (CAR) is a consolidation of
Federal air rules affecting the synthetic chemical organic
manufacturing industry (SOCMI). We selected the Federal air
rules applying to the SOCMI for a pilot project to study the
feasibility and practical implications of consolidating and
streamlining existing rules, and to establish a workable process
for consolidation that can be applied to other consolidation
efforts in the future. The CAR is one of the initiatives
announced by President Clinton and Vice President Al Gore on
March 16, 1995 to reinvent environmental regulation.
We solicited public comments on the success of this pilot
project as measured against the 10 principles for reinventing
environmental regulation, which are listed in the proposal
preamble (63 FR 57792) . We received comments on these issues as
well as other issues raised by the commenters. We also solicited
public comments on whether the CAR meets specific goals,, which
are discussed in sections VI and XII of the proposal preamble
(63 FR 57759 and 57792) . A public hearing was offered, but none
was requested. At the request of several parties, we reopened
the comment period for 30 days so that all parties could have
additional time to prepare comments on the proposed rule. The
EPA received written public comments from the commenters listed
in table 1.
This document summarizes the comments received and presents
EPA's responses. The comments have been grouped into sections by
1-1
-------
subject. Section 2.0 contains comments of a general nature, such
as those addressing the scope, applicability, structure, and
format of the rule. Section 3.0 presents comments specific to
each of the subparts of the proposed rule (general provisions,
storage vessels, process vents, transfer racks, equipment leaks,
and closed vent systems and control devices). Section 4.0 lists
comments on implementation issues and title V interaction.
Section 5.0 hosts the comments EPA received on the proposed
changes to the equipment leaks referencing subparts. Finally,
section 6.0 concludes by listing miscellaneous comments.
1-2
-------
TABLE 1. LIST OF COMMENTERS
Document Number
IV-D-01
IV-D-02
IV-D-03
IV-G-01
IV-G-02
VI-D-01
VI-D-02
VI-D-03
VI-D-04
VI-D-05
Commenter
B.M. Higgins, Chair, STAPPA Air Toxics
Committee, and R.H. Colby, Chair,
ALAPCO Air Toxics Committee, State and
Territorial Air Pollution Program
Administrators/Association of Local
Air Pollution Control Officials
(Request to extend comment period.)
N.L. Morrow, Safety and Environmental
Affairs Division, Exxon Chemical
Americas (EGA) , Houston, TX
(Request to extend comment period.)
Comments of Millennium Specialty
Chemicals, Inc. (MSC) and Bush Boake
Allen, Inc. (BBA) (collectively, the
Terpene Processors), submitted by H.T.
Vinyard, Jr., Counsel, Smith, Hulsey &
Busey, Jacksonville, FL
L. Platt, EH&S Regulatory Management
Expertise Center, Dow Chemical
Company, Freeport, TX
Supplemental comments by L. Platt,
EH&S Regulatory Management Expertise
Center, Dow Chemical Company,
Freeport, TX (duplicate)
B. Mathur, Chief, Bureau of Air,
Illinois Environmental Protection
Agency, Springfield, IL
N.L. Morrow, Exxon Chemical Americas
(EGA) , Houston, TX
R.I. Zvaners, Senior Manager,
Environmental Policy, Chemical
Manufacturers Association (CMA) ,
Arlington, VA
J.L. Woolbert, Technical Associate,
Eastman Chemical Company, Longview, TX
N. Dee, Director, Environmental
Affairs, National Petrochemical &
Refiners Association (NPRA) ,
Washington, DC
1-3
-------
Document Number
VI-D-06
VI-G-01
VI-G-02
VI-G-03
VI-G-04
VI-G-05
Commenter
B.M. Higgins, Chair, STAPPA Air Toxics
Committee, and R.H. Colby, Chair,
ALAPCO Air Toxics Committee, State and
Territorial Air Pollution Program
Administrators/Association of Local
Air Pollution Control Officials
Comments of Stepan Company, submitted
by P.P. Sharkey, Attorney, Mayer,
Brown, & Platt, Chicago, IL
J.S. Pew, Attorney, Earthjustice Legal
Defense Fund, Washington, DC
J. Marsh, Regulatory Affairs Director,
Chemical Industry Council of Illinois
(CICI), Rosemont, IL
Supplemental comments of Stepan
Company, submitted by P.P. Sharkey,
Attorney, Mayer, Brown, & Platt,
Chicago, IL (duplicate)
N. Carlson Manager, Environmental
Programs, Elf Atochem North America,
Inc., Philadelphia, PA
1-4
-------
2 . 0 GENERAL
Comment: Several commenters (VI-D-02, VI-D-03, VI-D-04,
VI-D-06) commended our efforts in developing the CAR. One
commenter (VI-D-03) praised the process we used to achieve this
proposed rule and encouraged using the CAR development process as
a model in other rulemaking activities.
Response: We appreciate this supporting comment and agree
that lessons learned from the CAR development process will be
applied in any future consolidation efforts.
Comment: Several commenters (IV-G-01, VI-G-03, VI-D-03,
VI-D-06, VI-D-01) support our proposal of making the CAR an
optional compliance method. One commenter (IV-G-01) cited
several examples, some from the preamble of the proposed rule, of
why the CAR should be optional. Another commenter (VI-D-03)
indicated that the CAR will neither benefit all facilities nor
benefit all facilities equally, therefore facilities will need to
determine the benefit of the CAR on a case-by-case basis. One
commenter (VI-D-06) pointed out that by making the CAR optional,
V
the CAR affords industry a reasonable amount of flexibility while
maintaining current stringency levels and avoiding an excessively
complex rule.
Response: We appreciate this supporting comment and note
that the final rule retains the CAR as an optional compliance
method. We also agree that the CAR would not necessarily benefit
every source and that your decisions will be made on a case-by-
case basis.
Comment: One commenter (VI-D-06) encouraged us to develop
consolidated air rules for other industry sectors in the future.
The commenter (VI-D-06) suggested that rules applying to surface
2-1
-------
coating operations would be amendable to this more efficient
compliance mechanism.
Response: We will monitor how this pilot consolidated rule
for the SOCMI is received. Depending on its popularity and
effectiveness, we may expand the current rule or develop other
consolidated rules; given available resources.
2.1 CAR OBJECTIVES AND BURDEN REDUCTION
Comment: Two commenters (VI-D-02, VI-D-03) agree that the
CAR contains explicit solutions for many troublesome overlap
issues among existing regulations. Specifically, the commenter
(VI-D-03) believes the CAR: (1) facilitates implementation and
compliance by clarifying requirements, (2) explicitly describes
how to determine applicability and rule overlap using a detailed
level of analysis rather than a generic level, and (3) addresses
many overlap issues that had not been resolved before this
proposal using a logical, consistent, and rational approach.
Response: We thank the commenters for their support of the
approach taken by the CAR to reduce confusion regarding overlap
of existing regulations.
Comment: One commenter (VI-D-03) agrees that the CAR meets
the President's objectives for rule consolidation. Specifically,
the commenter stated that the CAR incorporates the burden
reduction options allowed in earlier rules and extends the
V
benefits to other compliance areas. The commenter stated that
the CAR also accomplishes the reinvention objective by providing
a more user-friendly structure and format and through
clarification and simplification. Another commenter (IV-G-01)
stated that the CAR goes a long way in reducing conflicting and
overlapping requirements, but does not eliminate them which is
one of the President's objectives: to "eliminate conflicting and
overlapping Federal air compliance requirements."
Response: We appreciate feedback from the commenter
regarding meeting the President's objectives for rule
consolidation. We maintain, however, that all overlap issues
2-2
-------
have been eliminated for the rules that are consolidated in the
CAR. Not all rules have been included and further discussion on
the decisions made regarding the rules to include in the CAR is
in section 2.2, Scope.
Comment: One commenter (VI-G-03) agreed that the CAR
provides a single set of recordkeeping, reporting, and monitoring
requirements by consolidating the requirements of several rules
into one rule. This consolidation provides a significant
reduction in burden; for example, the consolidated provisions
require periodic reporting only twice annually. Two commenters
(VI-D-02 and VI-D-06) asserted that the consolidation of
monitoring, reporting and recordkeeping requirements will result
in simplifying regulatory requirements, encouraging source
compliance with the regulations, and providing a clearer
oversight role for the implementing agencies.
Response: We thank the commenters for this feedback
regarding the CAR'S consolidated recordkeeping, reporting, and
monitoring requirements.
Comment: One commenter (VI-G-03) indicated that, as
currently proposed, the CAR does not eliminate all instances of
overlapping requirements. The commenter (VI-G-03) gave an
example when a product accumulator vessel (either distillate or
bottoms) is also used to store material, it is subject to both
subparts V and FF of 40 CFR part 61. Subpart V requires product
V
accumulator vessels to be equipped with closed vents and control
devices, but subpart FF does not limit controls to closed vents
and control devices. The commenter further stated that subpart V
required leaks in the closed vent system to be repaired within
5 days, whereas subpart FF allows these units to leak for up to
45 days before repairs are required. The commenter (VI-G-03)
recommended that the CAR should directly address issues such as
the multiple uses of a product accumulator vessel and should
eliminate persistent conflicts under existing regulations.
2-3
-------
Response: The provisions of 40 CFR part 61, subpart FF apply
to waste management units. The provisions of 40 CFR part 61,
subpart V apply to surge control vessels and bottoms receivers.
The decision as to which rule applies would have already been
made prior to using the CAR. The commenter did not elaborate on
the situation where one vessel could be subject to both rules and
we can not identify a situation where this might happen. Also,
subpart FF is not within the scope of the CAR; all overlapping
provisions have been eliminated among the rules consolidated in
the CAR.
Comment: One commenter (IV-G-01) expressed that the degree
of burden reduction from using the CAR has not been
substantiated. The commenter (IV-G-01) further asserted that
burden reductions from the CAR will vary from facility to
facility and that, even if the overall average burden reduction
is significant, the reduction to any one facility may be so small
that it would not offset the additional time, effort, and expense
necessary to train personnel to use the CAR. Such facilities may
choose not to use the CAR.
Response: We agree with the comment that burden reductions
from using the CAR will vary among facilities and the burden
reduction to any one facility may not offset the time, effort,
and expense of training personnel to use the CAR. Such
facilities may choose not to use the CAR. However, we point out
V
that not all facilities will fall into this category. Many
facilities will incur a benefit from using the CAR, especially
those subject to several of the referencing subparts. While the
true burden reduction of any proposed rule cannot be known prior
to the rule's promulgation and implementation, we have made every
effort to reasonably estimate the burden reduction attributable
to the CAR. We maintain that several benefits can be obtained by
sources complying with the CAR as described in the proposal
preamble at 63 FR 57790. Each source can assess its benefits and
decide whether to use the CAR as a compliance option. These
benefits were not questioned by the commenter.
2-4
-------
Comment; One commenter (IV-G-01) indicated that switching
from compliance with one rule to compliance with another rule
requires a steep learning curve on the part of company and
facility personnel. The commenter (IV-G-01) suggested that any
time such a switch occurs there is increased potential for
noncompliance until the facility becomes sufficiently experienced
with the new rule. The commenter (IV-G-01) suggested that this
would be a particular concern with the CAR, because it is more
detailed in some respects than the rules it would replace.
Response: We concur with the commenter's assertion that the
CAR will require a learning curve on the part of company and
facility personnel. However, the provisions in the CAR are based
on provisions in the referencing subparts with which sources must
already comply. Therefore, company and facility personnel should
find that they are familiar with most of the provisions in the
CAR. Where these provisions differ is usually a reduction in
burden or an increase in flexibility, or a limited increase in
burden, such as with the requirements for halogen scrubbers. Any
source believing that the learning curve is too steep can
continue to comply with the existing subparts with which they are
familiar. Additionally, a source considering using the CAR can
take any amount of time necessary before making the decision to
do so. Therefore, we do not concur that there will be an
increase in misunderstanding of compliance obligations. It is
V
important to remember that, at all times, facilities must be in
compliance either with the referencing subparts or the CAR. When
choosing to comply with the CAR, facilities are obligated to
comply with the applicable referencing subparts until the CAR is
implemented.
Comment: One commenter (VI-G-03) alleged that the CAR'S
imposition of more stringent control requirements is a
substantial disincentive to using the CAR. The commenter
(Vl-G-03) contended that we should reconsider whether the
consolidation of procedural requirements such as monitoring and
2-5
-------
reporting necessarily requires application of the most stringent
level of control at any given SOCMI CAR unit(SCU).
One commenter (VI-D-03) accepts the imposition of a halogen
scrubber requirement for some new source performance standard
(NSPS) process vents, because it is acceptable to have occasional
increased stringency as a trade-off for the rule simplicity
provided by the CAR, particularly since the CAR is an optional
rule.
Response: The commenter (VI-G-03) did not specify the
increases in stringency that are substantial disincentives to
using the CAR. However, we assume the commenter is referring to
the proposed increase in control stringency due to controlled
fittings being applied to 40 CFR part 60, subparts Ka and Kb
tanks; halogen vent stream control applied to 40 CFR part 60,
subparts III, NNN, and RRR process vents and 40 CFR part 61,
subpart BB transfer racks; and instrument monitoring for
connectors applied to connectors subject to 40 CFR part 60,
subpart W and 40 CFR part 61, subpart V. In these cases, we
proposed an increase in control stringency because maintaining
the differences in stringency would have made for a very
complicated rule that was not actually a consolidation. Also,
the stringency could not be decreased for the most stringent rule
because this would cause less emissions to be controlled than
required by the most stringent rule.
V
We agree with the commenter (VI-D-03) that the optional
nature of the CAR alleviates concern with the rare increases in
stringency. We also maintain that the level of burden associated
with the control of additional fittings and control of
halogenated vent streams is small when considering that the
number of SOCMI tanks and halogenated vent streams that would not
be subject to the hazardous organic NESHAP (HON) is small. There
are few SOCMI sources that would experience this increase in
stringency. Also, if an individual source would not experience a
net benefit from using the CAR, there is no requirement forcing
the source into the CAR. With the change in the final rule
2-6
-------
eliminating the SCU concept (see section 2.3.1) and allowing any
source or affected facility subject to a referencing subpart to
opt into the CAR, there is even more flexibility for a facility
to use the CAR only where it benefits the facility.
For connectors, however, we have been persuaded that the
increase in burden of instrument connector monitoring for sources
not subject to the HON but subject to 40 CFR part 60, subpart W
or 40 CFR part 61, subpart V could be a substantial disincentive
to using the CAR for these sources. We have provided alternative
provisions to the proposed instrument connector monitoring that
follow the inspection requirements for connectors in subparts V
and W. These alternative provisions are an option for
connectors subject to subparts V or W but not the .HON. See
section 3.5.10 for further discussion on the change to the
connector provisions.
2.2 SCOPE
2.2.1 Incorporating or Allowing Other Rules in the CAR
Comment: Several commenters (IV-G-01, VI-D-02, VI-D-03,
VI-D-04, VI-D-05) support increasing the scope of the CAR so that
it completely incorporates other rules applicable to SOCMI
sources. All of these commenters (IV-G-01, VI-D-02, VI-D-03,
VI-D-04, VI-D-05) specifically mentioned the HON wastewater
provisions (40 CFR part 63, subpart G); four of the commenters
(VI-D-02, VI-D-03, VI-D-04, VI-D-05) specifically mentioned the
Benzene Waste NESHAP provisions (40 CFR part 61, subpart. FF); and
three commenters (IV-G-01, VI-D-02, VI-D-04) specifically
mentioned the SOCMI wastewater NSPS (40 CFR part 60,
subpart YYY). One commenter (IV-G-01) requested consolidation of
several other rules affecting SOCMI including HON Group 2
transfer racks and storage vessels, emission points not requiring
control under the non-HON referencing subparts, marine loading
under 40 CFR part 61, subpart BE, and equipment leak provisions
under resource conservation and recovery act (RCRA) rules
subpart BB. One commenter (IV-G-01) argued that, without
2-7
-------
including additional regulations in the consolidation, the CAR
cannot achieve its goals of (1) having a single set of
consolidated requirements for the SOCMI, and (2) reducing the
complexities of overlapping regulations among different Federal
air programs. The commenter (IV-G-01) alleges that sources, in
most cases, have no incentive to use the CAR and concludes that
the CAR must consolidate several additional rules in order to
provide this incentive.
Four commenters (VI-D-01, VI-D-02, VI-D-03, VI-D-05) were
specifically interested in consolidating all rules that may apply
to a facility that has a SOCMI process unit on site. All four
commenters requested that sources subject to the Petroleum
Refinery NESHAP, 40 CFR part 63, subpart CC be allowed to use the
CAR to comply with subpart CC. Two commenters (VI-D-02, VI-D-03)
also referred to the following 40 CFR part 63 subparts as rules
that should use the CAR as a compliance option: I, U, W, DD, TT,
OO, UU, WW, and JJJ. The commenter (VI-D-03) encouraged us to
define CAR requirements as acceptable for requirements in non-
consolidated rules that are likely to overlap with the CAR at
SOCMI sites. The commenter (VI-D-03) noted that one goal of the
CAR is to allow similar and shared systems at SOCMI facilities to
use one set of compliance requirements. Therefore, the commenter
(VI-D-03) concluded it is important to assure that all rules that
could impact those facilities are considered in the
V
consolidation. Until these relevant rules are added to the CAR,
the commenter (VI-D-03) argued, the critical overlaps muŁt be
addressed because this is, perhaps, the single largest impediment
to facilities adopting the CAR. The commenter (VI-D-03) asserted
that maximizing the use of the CAR will provide the most burden
reduction and compliance assurance for States and facilities.
The commenter (VI-D-03) provided examples of language from other
rules that address overlap as well as recommended changes to the
language of the CAR at §65.1(c) and addition of new language at
§65.1(n).
2-8
-------
Response: We agree that the CAR does not consolidate all
rules applicable to the SOCMI or to sources with SOCMI processes
on site. However, as stated in the preamble (63 FR 57750), the
scope of the CAR, as a pilot project, was limited to Federal
Clean Air Act (Act) rules that apply to SOCMI, because these
rules would provide benefit to affected sources, yet the scope
was defined well enough to ensure a reasonable chance of success.
Other rules were considered in this consolidation, but their
incorporation would have made the process unmanageable in a
reasonable amount of time. Some rules considered for
consolidation were subject to substantial changes in litigation
when the CAR process started and others are currently in
litigation. To include additional rules in the consolidation
effort at this point would require a supplemental proposal. We
consider our efforts better spent finishing this proposal. We
maintain that there can be significant burden reduction with the
rules that are currently consolidated and that reduction will
persuade sources to use the CAR. We do not agree that sources
have no incentive to use the CAR.
Comment: Four commenters (VI-D-01, VI-D-02, VI-D-06,
VT-G-03) requested that we consider using the CAR as a compliance
option for new regulations. Two commenters (VI-D-01, VI-D-02)
specifically mentioned the Miscellaneous Organic NESHAP; one
commenter (VI-D-06) mentioned the Generic Maximum Achievable
\
Control Technology (MACT); and one commenter (VI-D-02) mentioned
the Ethylene MACT. One commenter (VI-D-01) recommended that any
new regulations applicable to the SOCMI that may be promulgated
should be incorporated into the CAR for use by affected sources
that have opted to use the CAR. The commenter (VI-D-01) stated
that in this case, additional incorporated rules would follow
part 70 on opting new rules into a title V permit as they are
promulgated.
Response: We may consider using the CAR in future
rulemakings. Because of the timing of the generic MACT's
promulgation, it was not possible to consolidate that rule into
2-9
-------
the CAR. However, the generic MACT employs similar structure,
concept, and provisions as the CAR.
Comment: One commenter (VT-D-03) noted that the proposed
language in §60.560(j) and (k) would not allow polystyrene
process vents subject to 40 CFR part 60, subpart DDD, to use the
CAR. The commenter pointed out that similar polyethylene process
vents subject to subpart DDD are allowed to use the CAR. The
commenter suggested that this exclusion resulted from a drafting
problem and provided revised language to allow polyethylene and
polystyrene process vents the option of complying with the CAR.
Response: We agree that the polystyrene process vents
subject to subpart DDD that choose the control device or flare
compliance option should be allowed to opt into the CAR. We have
edited the final rule to allow this option.
Comment: Some commenters (IV-G-01, VI-D-02, VI-D-03,
VI-D-04, VI-D-06) requested that rules that refer to referencing
subparts should be allowed to use the CAR as a compliance option.
Three commenters (IV-G-01, • VI-D-02, VI-D-03) specifically ''
mentioned 40 CFR part 63, subpart I and the polymers and resins
MACT standards. Two commenters (VI-D-04, VI-D-06) requested any
MACT standard that points to the HON be allowed the option to
comply with the CAR. Two commenters (IV-D-02, VI-D-03) stated
that subpart DDD requires compliance with subpart W for
equipment leaks. The commenters (VI-D-02, VI-D-03) indicated
V
that a change is needed in §60.560 of subpart DDD to make clear
that compliance with the CAR subpart F for equipment leaks is
allowed as an alternate to complying with 40 CFR part 60,
subpart W. The commenter (VI-D-02) stated that the CAR, in
general, is unclear on the situation in which a process unit is
subject to a referencing subpart (in this case subpart W) by
reference from another rule (subpart DDD). The commenter
suggested that we amend the applicability of the CAR to clearly
state the options for these types of units.
The commenter (VI-D-03) stated that additionally, because
40 CFR part 60, subpart DDD is applicable only on a process
2-10
-------
section by process section basis, a preamble discussion is needed
relative to §60.560 (j) and (k) to clarify that process sections
not triggering the applicability of 40 CFR part 60, subpart ODD
will not become subject to the CAR if the CAR is adopted. The
commenter (VI-D-03) provided suggested language for these
recommended revisions.
Response: We have not expanded the scope of the final rule
to include other regulations that refer to referencing subparts.
The details, approach, and ramifications of allowing the CAR for
these other rules have not been investigated. Many of the rules
that point to the CAR'S referencing subparts for requirements
generally have complex references, with conditions and exceptions
to the referencing subparts. To allow these rules to comply with
the CAR would require us to study the conditions and exceptions
and possibly develop detailed references for compliance with the
CAR. We consider our efforts better spent finishing this rule so
that the burden reductions associated with it can be used as soon
as possible. Furthermore, expanding the scope of the CAR to
other rules at this point would entail additional proposals.
The proposal preamble contained discussion that no equipment
would become subject to a rule just because the CAR was used for
compliance [63 FR 57751].
Comment: One commenter (VI-D-05) requested allowing
equipment at petroleum refineries subject to a referencing
V
subpart to opt into the CAR on the basis of ease of compliance
determination by all parties, a reduction in paperwork and
administrative cost, and flexibility.
Response: Under the proposed CAR, equipment at refinery
processes subject to referencing subparts that were on the same
plant site as a SCU complying with the CAR could opt into the
CAR. However, with the increase in flexibility in the
applicability of the final CAR any equipment subject to a
referencing subpart can use the CAR as a compliance option. See
section 2.3 for further discussion on the change in
applicability.
2-11
-------
2.2.2 Miscellaneous Scope Comments
Comment: One commenter (VI-G-03) asked that the CAR allow
owners or operators to obtain emission reduction credits for any
additional controls required under the CAR that are more
stringent than the controls required by the referencing subparts.
The commenter stated that these credits would be in addition to
the credits the unit has already earned according to a voluntary
program. The commenter asserted that this would encourage the
use of the CAR.
The commenter (VI-G-03) noted that SOCMI sources in many
States voluntarily apply controls in excess of existing
regulations to obtain usable and marketable emission reduction
credits. The commenter (VI-G-03) wanted to know if participants
in the State programs or the Federal Early Reduction Program will
still get credit for these programs if they opt to use the CAR.
The commenter was concerned that the CAR would require an owner
or operator to give up existing credits or the ability to obtain
future credits when opting into the CAR.
One commenter (VI-G-03) noted that the monitoring and
recordkeeping requirements for units subject to the Early
Reduction Program in part 63, subpart D are specified in the
title V permit rather than in the rules themselves. The
commenter stated that this makes it unclear whether facilities
with Early Reduction Units can opt into the CAR. The commenter
V
(VI-G-03) supports allowing owners and operators with Early
Reduction Units to opt into the CAR for the benefit of consistent
regulation of storage vessels and other units. In such cases,
the commenter stated that the owner or operator would retain the
benefit of early reduction.
Response: The CAR does not have an emissions averaging or
trading program specifically associated with it. Therefore,
there is no basis to establish credits or debits for compliance
with the CAR. If a State would like to incorporate the CAR into
their voluntary emissions credit program and the program does not
violate the state implementation plan (SIP) this is within the
2-12
-------
States purview to establish. Whether a State gives credits for
compliance with the CAR depends on the State's plan and the
specific site. Whether a State allows sources to retain existing
credits or generate additional credits for compliance with the
CAR may affect the sources' decisions to use the CAR. It should
be noted, however, that the added flexibility of opting into the
CAR on an equipment basis rather than an SCU (see section 2.3,
Applicability, for more discussion on this) has given the
facility the option to choose the equipment complying with the
CAR that will give the most benefit.
There is nothing in the CAR that prohibits all or part of an
Early Reduction Source to opt into the CAR. However, the early
reduction limits for the source would still have to be met unless
the source chooses to give up its compliance extension.
Comment: One commenter (VI-D-04) asserted that one of the
substantial transition costs to implement the CAR at existing
facilities is the cost related to incorporating the CAR as an
alternative compliance approach in Part 70 Operating Permits.
The commenter pointed out that if the rule is promulgated very
soon, there could be an opportunity in the State of Texas for
some SOCMI sources to minimize this transition cost by being able
to implement the CAR into their initial part 70 permit
applications. The commenter recommended that the rule be
finalized quickly.
V
Response: We agree with the commenter that getting the CAR
out in a reasonably timely manner is preferred, not only, so that
the implementation is coordinated with the Texas Title V permit
program, but also to allow the burden reduction opportunities
provided to the Industry to be used. We have processed the
comments received, responded to the comments and made the
appropriate changes to the regulation in a reasonably timely
manner given the complexity of the rule.
Comment: One commenter (VI-D-03) praised the adoption of
the CAR styles, structures, and language models into the generic
MACT and other recent rulemaking actions and encouraged us to
2-13
-------
continue this approach because it provides significant
opportunity for burden reductions in regulatory development.
Response: We thank the commenter for their support.
2.3 APPLICABILITY
2.3.1 SOCMI CAR Unit
Comment: Several commenters (VI-D-01, VI-D-03, VI-D-06,
VI-G-03) commented on the complexity and the confusion of either
the SCU definition or the assignment procedures or both. Three
commenters (VI-D-01, VI-D-06, VI-G-03) stated that SCU and what
constitutes an SCU are not clearly defined in the rule. Two of
the commenters (VI-D-01, VI-D-06) suggested that the CAR include
some "real world" examples, as we did in the Early Reduction
rules.
One commenter (VI-G-03) supports the proposal to allow the
option of using the CAR on the basis of individual SCU's but said
further clarification was needed. The commenter (VI-D-03)
suggested describing an SCU'determination in the following way:
(1) as per the SCU definition, identify all equipment associated
with the process that is impacted or could be impacted by the
HON, subpart III, NNN, RRR, and/or W rule requirements,
(2) subject the identified equipment to a review of the
applicable underlying rules to determine the "affected"
equipment, and (3) apply the CAR only to the "affected"
V
equipment.
Another commenter (VI-D-01) noted that the flexibility of
the CAR is manifested in the ability of an affected source to
define the affected chemical manufacturing process unit (CMPU)
and to be able to subgroup the CMPU, which will clearly define
the SCU's affected by the CAR. The commenter (VI-D-01)
recommended that the CAR should encourage this subcategorization.
Another commenter (VI-D-06) believes that the applicability
procedures may prohibit implementation at the State and local
levels. The commenter (VI-D-06) specifically cited the
assignment procedures in §65.1 (j) and (k) of the CAR (emission
2-14
-------
points commonly shared between process units) as excessively
complicated. The commenter (VI-D-06) suggested that we consider
a provision that would allow groups of like equipment subject to
one of the referencing subparts (such as transfer racks or
storage tanks) to implement the CAR, while other portions of the
SCU may continue to comply with applicable referencing subparts.
Another commenter (VI-D-01) expressed concern that the CAR
may be applicable to only a fraction of the tanks at a facility.
The commenter (VI-D-01) recommended that if more than 50 percent
of the storage vessels at a source are subject to the CAR, then
the CAR should cover all storage vessels at that source. The
commenter (VI-D-01) contends that this approach will simplify
applicability, recordkeeping, and reporting.
Response: The proposed CAR required facilities to opt into
the CAR on a SCU basis because we thought this would reduce
potential complexity of implementing the CAR for regulatory
authorities. During development of the proposed CAR, State
representatives expressed a desire to allow larger portions of
sources to opt in as opposed to a more piecemeal approach,
indicating that it would be easier for them to enforce the rule.
They felt that opting in on an SCU basis would provide a small
enough collection of emission points and equipment to provide
operational flexibility to the facility, but a large enough
collection to avoid possible confusion and additional burden for
i
regulatory authorities.
However, after reviewing the comments regarding the. SCU and
assignment procedures, we have concluded that the perceived
confusion and complexity added by the SCU assignment procedures
outweigh the reduction in burden and complexity to State
inspectors by requiring facilities to opt in on a SCU (large
collection of equipment) basis. Keeping track of which equipment
is in or out of a SCU and which SCU is complying with the CAR
appears to be more burdensome than keeping track of which
emission point is complying with what rule which must be done
anyway. We have simplified the applicability provisions of the
2-15
-------
CAR by allowing in the final CAR any affected source subject to a
referencing subpart to use the CAR as a compliance option with
two exceptions described below. This means that a facility may
choose to opt in, for example, one subpart Kb tank or all
equipment at the facility that is subject to a referencing
subpart. For both regulator .and industry personnel, this
eliminates the assignment procedures that determine what
equipment constitutes a SCU. With this change, it is not
necessary to keep track of new regulated sources and whether they
are part of a SCU or not.
There are two situations where the regulated source in the
CAR does not match the affected source of the referencing
subpart. In one situation, the affected source for 40 CFR part
61, subpart V is an individual piece of equipment like a pump or
a valve. We determined that allowing owners or operators to opt
in to the CAR on an individual piece of equipment would not be
workable. Therefore, owners or operators must opt in the group
of affected equipment at a process unit. This does not alter the
applicability of subpart V to a facility; it only affects the set
of equipment that can comply with the CAR.
The second situation where the regulated source in the CAR
does not match the affected source of the referencing subpart is
in the HON. Under the HON, the affected source is the total of
all applicable emission points at the plant site that are subject
V
to the HON. Thus, a HON facility that contains more than one
CMPU, would consist of only one affected source, which would be
the collection of all subject CMPU's. However, under the CAR the
regulated source is collection of emission points within each
CMPU (as proposed under the original concept in the CAR of the
SCU). Thus, a HON facility can choose to opt into the CAR on a
CMPU basis, and not the entire collection of CMPU's that comprise
the HON affected source.
Although we believe that in most cases facilities will opt
in larger groups of equipment (e.g., most or all subpart Kb
tanks), the States and owners or operators have the opportunity
2-16
-------
to work together to determine the basis on which facilities can
opt in their equipment that will provide the "best fit" for both
regulators and industry.
Comment: One commenter (IV-G-01) suggested that the
proposal preamble discussion at 63 FR 57757 and 57758 regarding
the SCU's and the definition of regulated source carries a
misleading implication. The commenter (IV-G-01) stated that the
discussion implies that if one SCU within a plant site elects to
comply with the CAR, then all new and existing sources,
regardless of their relation to the SCU, must comply with the CAR
if they are part of the same plant site and subject to one of the
referencing subparts. The commenter (IV-G-01) suggested an edit
to clarify this issue.
Response: Because of the change in the applicability
provisions, the clarification suggested by the commenter is no
longer necessary.
Comment: The commenter (VI-D-03) asserted that
polyethylene, polypropylene, and polystyrene units should be
defined as SCU's, by adding 40 CFR part 60, subpart DDD to the
list of rules in the SCU definition. The commenter argued that
this would clarify the pointers in subpart DDD and the
requirements that apply to these units. The commenter stated
that it would also allow stand-alone polyethylene, polypropylene,
and polystyrene units to take advantage of the CAR'S burden
v
reductions. The commenter pointed out that additionally, some
facilities operate a number of stand-alone polyethylene,-
polypropylene, and polystyrene sites at sites where SCU's are
present. The commenter asserted that if subpart DDD units were
included in the SCU definition, facilities with these units could
use the CAR for the entire site. The commenter stated that the
facility would not have to maintain compliance systems for both
rules, thereby removing an 'impediment to using the CAR. While
this would expand the scope of the CAR, the commenter reasoned it
would allow more of the SOCMI industry to take advantage of the
CAR.
2-17
-------
Response: Because of the change in the applicability
provisions there is no longer a need to define the SCU. A
facility can choose to opt any affected source or facility
subject to a referencing subpart into the CAR regardless of its
affiliation to a SCU. The SCU definition has been deleted in the
final CAR.
Comment: One commenter (VT-D-02) questioned why §65.1(1) (2)
refers to subparts W, III, NNN, and RRR. The commenter stated
that the purpose of §65.1(1)(2) is to identify process units
where equipment assignment procedures are unnecessary, because
the referencing rule has already defined the SCU. The commenter
stated that the definitions of process units contained in
subparts W, III, NNN, and RRR are not specific enough to
identify all of the equipment that is part of the process unit.
The commenter (VT-D-02) requested a revision of §65.1(1) to
require that the equipment included in an SCU is the equipment
defined as part of the process unit in the referencing subparts,
as applicable, as well as any other equipment that is part of the
SCU as determined by the assignment rules in §65.1 (j) through
(m) .
Another commenter (IV-G-01) stated that the proposal
preamble (63 FR 57756) contains a discussion about assignment
procedures for assigning transfer operations to process units.
The commenter (IV-G-01) advised that this discussion should not
V
refer to thermoplastic product process units (TPPUs) and
petroleum refinery process units (PRPUs) because subparts U and
JJJ do not regulate transfer operations.
Response: Because of the changes in the applicability
provisions, the assignment procedures have been deleted.
Therefore, the suggestions made by the commenters no longer need
to be addressed.
2.3.2 Pointer Paragraphs in Referencing Subparts
Comment: One commenter (IV-G-01) suggested that
§§60.110a(e) and 60.110b(g) should not state that §60.16 of
subpart A still applies. The commenter pointed out that §60.16
2-18
-------
of subpart A is a list of prioritized major source categories and
is not necessary for sources complying with the CAR.
Response: We have left the reference to §60.16 of 40 CFR
part 60, subpart A, in the list of provisions that still apply to
sources complying with the CAR. Although it is just the list of
source categories, we consider it important to provide a complete
list of applicability provisions that have not been incorporated
into the CAR, but that still apply to sources complying with the
CAR.
Comment: One commenter (IV-G-01) advised that table 4 in
40 CFR part 63, subpart H is duplicative to table 1A being added
to 40 CFR part 63, subpart G. The commenter also noted that
table 1A of subpart G is missing a reference to 63..5(f) (1) that
is in table 4 of subpart H.
Response: Table 1A and table 4 both list the 40 CFR part 63
general provisions (subpart A) requirements that apply to owners
or operators choosing to comply with the CAR. Both tables are
necessary, one for subpart G of the HON (which applies to process
vents, storage vessels, and transfer operations) and one for
subpart H of the HON (which applies to equipment leaks). We have
added the missing reference to §63.5(f)(1) to table 1A of
subpart G. Also, there are some additional general provisions
that apply to sources referenced to the CAR from 40 CFR part 63,
subpart G that were inadvertently left off of table 1A of
V
subpart G in the proposed rule. These were identified in table 1
of subpart A of part 65 but were not included in the subpart G
table. These provisions have been added in the final CAR.
Comment: One commenter (VI-D-03) indicated that the
proposed §§60.110a(c) and 60.110b(e) pointer paragraph wording
suggests that all subject storage vessels in all SCU's at a site
must opt into the CAR together, rather than all subject storage
vessels within an individual SCU opting in together. The
commenter (VI-D-03) suggested the following language as a change:
"...for storage vessels that...and that is part of an individual
SCU. "
2-19
-------
The commenter noted that §§60.llOa(f) and 60.110b(h) make
clear that all subject storage vessels within an SCU must opt
into the CAR together, so leaving the word "all" out of these
paragraphs does not change the requirement, but does clarify that
all subpart Ka and Kb storage vessels in all SCU's at a site need
not opt into the CAR at once.
Response: Because of the changes in applicability, the
pointer paragraphs in subparts Ka and Kb have been revised to not
include the reference to SCU. We believe that this
simplification has taken care of the clarification suggested by
the commenter.
Comment: One commenter (VI-D-03) noted that for subpart DDD
process vents opting into the CAR, the proposed language at
§60.560(m) requires the vents to comply with the CAR subpart D.
The commenter pointed out that the total resource effectiveness
(TRE) approach in the CAR subpart D, however, is not the approach
used under subpart DDD to define covered vents. Therefore, the
commenter concluded only subpart G of the CAR is appropriate for
subpart DDD process vent provisions. The commenter (VI-D-03)
proposed changes to the language of §60.560(m) to address this
problem. The commenter (VI-D-03) also suggested that
clarification may be needed in all referencing subparts that
subpart DDD vents are not subject to the CAR subpart D.
Response: For owners and operators choosing to comply with
the CAR, §60.560 (j) requires subpart DDD process vents to comply
with subpart G of the CAR (not subpart D of the CAR). The
proposed §60.560(m) requires other process vents subject to other
referencing subparts (40 CFR part 60, subparts III, NNN, RRR, and
the HON) to also comply with the CAR if the process vents are
located in the same SCU. However, because of the changes to
applicability in the final CAR, §60.560(m) has been deleted.
Comment: One commenter (VI-D-02) requested that the ongoing
recordkeeping requirements of §60.116 be duplicated in the CAR
subpart C, so the owner/operator of a storage vessel complying
with the CAR need not refer back to part 60, subpart Kb.
2-20
-------
Response: We have not included the recordkeeping
requirements of §60.116b in the final CAR. This record is
required in both subpart Ka and Kb of part 60. It is an
applicability record of the material stored and its vapor
pressure. Because this record is required in only two of the
rules consolidated, and it is an applicability record, we decided
not to include it in the CAR. In general, all applicability'
requirements remain in the referencing subparts.
Comment: One commenter (IV-G-01) recommends changes to the
proposed text of 40 CFR part 60, subparts III, NNN, and RRR to
clarify specifically which affected facilities may elect to
comply with the CAR. The commenter (IV-G-01) asserts that the
proposed text does not clearly indicate which affected facilities
may choose to comply with the CAR, but rather implies by sequence
of the wording that only specific exempt facilities may choose to
comply with the CAR. Using subpart III at §60.610(d) as an
example, the commenter (IV-G-01) stated that, as it was written,
only sources subject to paragraph (c) could use the CAR: "Owners
or operators of process vents that are subject to this subpart
may choose to comply with the provisions of 40 CFR part 65,
subpart D to satisfy the requirements of paragraph (c) of this
section and §§60.612 through 60.615 of this subpart, except
§60.615(a), as provided in paragraphs (d)(1), (d)(2) and (e) of
this section." The commenter (IV-G-01) proposed language for
V
clarification, with an example from subpart III as follows:
"...Owners and operators of process vents that are subject to
60.610(a), (b) and (c) may choose to comply with the provisions
of 40 CFR Part 65, subpart D to satisfy the requirements of
60.612 through 60.615 of this subpart, except 60.615(a), as
provided...".
Response: We have clarified which facilities subject to
subparts III, NNN, or RRR may choose to comply with the CAR, but
have used different wording than what was suggested. The CAR
contains the exemptions from control for process vents with low
flow, low concentrations, and TRE values greater than 4.0.
2-21
-------
Therefore it is appropriate to state that the CAR satisfies the
requirements for process vents exempt from control. The text of
subparts III, NNN, and RRR has been modified to clarify this.
For example, the following text from subpart III at §60.610(d)
was modified: "Owners or operators of process vents that are
subject to this subpart may choose to comply with the provisions
of 40 CFR part 65, subpart D to satisfy the requirements of
§§60.612 through 60.615 of this subpart, except §60.615(a), as
provided in paragraphs (d)(1), (d)(2) and (e) of this section.
The provisions of 40 CFR part 65 also satisfy the requirements of
paragraph (c) of this section. Other provisions..."
2.4 STRUCTURE AND FORMAT
Comment: One commenter (IV-G-01) asserts that the CAR does
not provide a single set of provisions by which sources may meet
all regulatory requirements, because process units will have to
comply with the CAR and with certain paragraphs of the
referencing subparts. Two commenters (IV-G-01, VI-D-01) claimed
that the CAR does not completely simplify the set of regulations
being consolidated and thereby may cause more instances of
confusion and noncompliance. As an example, the commenter
(IV-G-01) pointed out that each of the referencing subparts
require the owner or operator to follow the general provisions of
subpart A of the CAR and some specific general provision
\
requirements from 40 CFR parts 60, 61, or 63.
Another commenter (VI-D-01) noted that the attempt to
provide a modular, stand-alone applicability section for the CAR
has resulted in a very complex structure and still requires the
implementing agencies and affected sources to rely upon the
individual rules that are consolidated in the CAR.
One commenter (IV-G-01) claimed that the CAR will have no
benefit to sources that have any group 2 or non-regulated vent
streams, or are subject to any State requirements. The commenter
(IV-G-01) reasoned that most sources eligible to use the CAR will
have streams subject to rules, or portions of rules, that are not
2-22
-------
being consolidated. The commenter (IV-G-01) maintained that the
CAR will be beneficial only to sources that have only group 1 or
regulated process vents, storage tanks, transfer operations,
and/or equipment subject to federal leak standards.
Response: We agree with the commenter that there are
applicable provisions in the referencing subparts and referencing
general provisions that require familiarity and awareness on the
part of sources complying with the CAR. However, these
provisions pertain to applicability and are used to determine
whether the rule applies to the source or equipment and whether
control is required. We maintain that once an owner or operator
of a facility complying with the CAR determines the applicability
of control, then the owner or operator will rarely have to refer
to the referencing subparts and referencing general provisions.
A change affecting applicability will require referring to the
referencing subpart.
We disagree with the commenter's assertion that the CAR will
not benefit facilities that have group 1 and group 2 storage
vessels and transfer racks. Group 1 storage vessels and transfer
racks are emission points which must be controlled. Group 2
includes emission points where control is not required. In
general, requirements for group 2 emission points include records
and reports that confirm the group status of the emission points.
Once the group 2 reports are submitted, there are few additional
V
requirements in the referencing subparts with which a facility
must comply, and the CAR requirements are not relevant to group 2
storage vessel and transfer rack emission points. We believe
significant benefits can accrue to sources subject to multiple
referencing subparts with group 1 points.
Comment: Two commenters (IV-G-01, VI-D-03) recommended that
we incorporate tables listing the provisions of the referencing
subparts that are still applicable when using the CAR, in order
to reduce the amount of regulatory information that stakeholders
must review to determine applicable requirements. One commenter
(IV-G-01) asserted that such a table would not only clarify the
2-23
-------
more subtle requirements of the CAR, but would also highlight 'the
remaining requirements from the referencing subparts that
otherwise would be easy to overlook. Another commenter (VI-G-03)
suggested adding a "road map" table which gives a section-
by-section comparison of the existing requirements in the
referencing subparts and the alternative CAR provisions.
Response: As suggested by the commenters, we have
incorporated a table that lists the provisions of the referencing
subparts that still apply to sources complying with the CAR as an
aid to the user. We have also developed tables that give a
paragraph-by-paragraph comparison of each referencing subpart and
its corresponding CAR paragraph in the proposed rule. These are
referred to as "correlation tables" and can be found on the EPA
Technology Transfer Internet site
(www.epa.gov/ttn/uatw/car/car_rdpg.html).
Comment: Another commenter suggested that the CAR should be
revised as a "stand alone" part so that sources subject to the
referencing subparts could simply opt to comply with the CAR,
thereby eliminating the need for cross-referencing the
referencing subparts.
Response: We have made every effort to make the CAR a stand
alone part that sources can be referred to with very little
looking back at the original referencing subpart. It has been
necessary to maintain the provisions in the referencing subparts
v
pertaining to applicability, however. Significant complexity
would have been added to the CAR if incorporation of the.
referencing subpart applicability had been attempted. Also,
there are a few recordkeeping or reporting provisions of some
referencing subparts that needed to remain applicable. These are
cases where we have determined that the provisions are necessary
for implementing and enforcing the referencing subpart.
By adding the tables of referencing subpart provisions that
still apply to sources complying with the CAR, we contend that
confusion has been eliminated and the final CAR and referencing
subparts are clear and easily implemented.
2-24
-------
Comment: One commenter (VI-D-03) expressed appreciation for
the CAR's new formatting features. The commenter stated that the
significant reorganization of regulatory requirements by end-user
need, by implementation step, by specific regulatory topic, and
by applicability is a key component to making the regulation less
confusing. The commenter stated that the frequent use of labels
to facilitate quick recognition of the organizational structure
provides clarity. The commenter (VI-D-03) indicated that the use
of direct language and translation of appropriate regulatory
requirements in step-wise procedures will help reduce the
learning time and disputes regarding desired regulatory
objectives.
Response: We appreciate the commenters support for the
formatting features found in the CAR.
2-25
-------
3.0 SPECIFIC COMMENTS BY SUBPART
3.1 GENERAL PROVISIONS
Comment: One commenter (VI-D-02) noted that §65.1(f)
requires an implementation schedule for sources opting to use the
CAR. The commenter stated that it is not clear whether this
schedule must be approved. The commenter (VI-D-02) argues that
the schedule should not be subject to review and approval,
because the CAR is voluntary and at least as stringent as the
referencing subparts. The commenter (VI-D-02) requests that we
clarify, through rewording of the regulatory text and the
preamble, that the implementation plan is submitted for planning
purposes only and that review and approval by the permitting
authority are not required.
Response: The implementation date is to be established in
coordination with the Administrator or delegated authority. We
consider this step to be important so that the regulator can be
aware of the process being used to convert the facility from
complying with the referencing subparts to complying with the
CAR. It provides an opportunity to the regulator to ensure that
»
there will be no compliance gaps, and the benefits of CAR
implementation for the regulator can be balanced with the
benefits to the source of using the CAR. The regulator also has
workload and resource constraints to consider and may prefer to
modify permits at a more opportune time than envisioned by the
source. In addition, the regulator must have some comfort level
with respect to implementation and enforcement and the emission
sources the facility wants to opt in. We have clarified in the
final CAR that the implementation date must be established by
mutual agreement between the facility and regulator.
3-1
-------
3.1.1 Definitions
Comment: One coramenter (IV-G-01) indicated that many
definitions in the underlying subparts were not included in the
CAR and that this forces the user to rely on the definition
sections in the referencing subparts, which is not explicitly
allowed in the CAR. The commenter (IV-G-01) recommends that the
CAR clearly state that definitions in the referencing subparts
should be used for terms that are used in the CAR and that are
not defined in the Act or in subpart A of the CAR.
Response: We have clarified in the CAR that the definitions
in the referencing subparts should be used for terms used in the
CAR and not defined in either the Act or subpart A of the CAR.
Comment: One commenter (IV-G-01) expressed that §65.2 of
the CAR needs to define the term "excess emissions." The
commenter (IV-G-01) pointed out that "excess emissions" is
explained in §65.3(a)(4) of the CAR but is not defined in §65.2.
The commenter (IV-G-01) requested that §65.2 should be consistent
with §65.3(a)(4), if this section actually contains the • •
definition of excess emissions.
Response: The term "excess emissions" is defined for
purposes of the CAR where it is first used, in §65.3(a)(4). The
term is used in the following locations in the CAR:
§65.6(b) (1) (i),
§65.6(b) (4) (iv) ,
§65.6(c), v
§65.162(a)(2)(iv), and
• §65.163(c) .
We have edited each of these sections to include a reference to
§65.3(a)(4), where the term "excess emissions" is defined. A
definition has not been added for "excess emissions" in §65.2
because the term has different connotations throughout the
parts 60, 61, and 63 regulations. Since the CAR affects rules in
all three parts, we have decided not to put a definition in the
definition section to avoid confusion. Instead the term will be
clarified when it is used in the CAR by cross references to
§65.3(a) (4) .
3-2
-------
Comment: One commenter (VI-D-06) requested clarification of
terms that contain subjective adjectives either in the
definitions section or in the sections where the terms occur.
Two commenters (VI-D-01, VI-D-06) specifically mentioned the
definitions of "excused excursion" and "reasonable intervals."
Another commenter (VI-G-03) pointed out that "excused excursion"
is not defined.
Response: It is inevitable that some terms will be used in
a regulation that are not immediately obvious to everyone. For
technical terms and terms where confusion is possible, we provide
definitions. For many other terms, we rely on the meaning given
the terms in common language. The term "reasonable intervals" is
used one time in the CAR at §65.3(d)(3). The language used in
this paragraph is consistent with the language of
§65.100(k) (9) (iii) . We consider this term sufficiently clear.
The term excused excursions has been clarified in the sections
where it is used.
Comment: One commenter (IV-G-01) suggested that the
definition of "repaired" be modified to require monitoring "as
appropriate" to verify repair because visual leakers may not
require remonitoring.
The commenter(IV-G-01) also concluded that the phrase
"definition of repair" found in §65.106(d)(2) should be changed
to "definition of repaired or first attempt at repair" in order
to parallel the definition of "repaired" in §65.2 of subpart A
and because monitoring is also required after the first .attempt
at repair.
Response: "Visual leakers" requirements in the CAR
typically include language to the effect of "repaired in this
instance means that the visual indications of a leak have been
eliminated." Therefore, the standard definition of "repaired" in
§65.2 would not apply, and adding the phrase "as appropriate" to
the definition of "repaired" is confusing and is not necessary.
We maintain that it is not appropriate to add the suggested
language to §65.106(d)(2).
3-3
-------
Also, in §65.106(d)(2), the CAR requires follow-up
monitoring after a leaking valve has been repaired. The
requirement is equally applicable whether "repair" was successful
on the first, second, third, or any subsequent attempt at repair.
If the required "first attempt at repair" is successful (and
proven through monitoring to be "repaired"), then §65.106(d)(2)
requires a follow-up monitoring within 3 months.
Comment: One commenter (IV-G-01) expressed the opinion that
the term "high throughput transfer rack" should be defined in the
CAR subpart A in §65.2. The commenter also noted that
§65.145(b) (1) (iii) has a reference to a "non low-throughput
transfer rack" and recommended that the term "high throughput
transfer rack" be used for consistency.
Response: We agree that using the term "high throughput
transfer rack" would be beneficial. We have edited the CAR to
incorporate the term where appropriate and have added a
definition for high throughput transfer rack.
Comment: One commenter (IV-G-01) reasoned that the
definition of "relief device or valve" found in the CAR
subpart A, §65.2 should be revised to include relief valves that
allow passage of nitrogen to prevent vacuums. The commenter
suggested that without such a change, the CAR definition of
"relief device or valve" would be inconsistent with the change
that is proposed to the HON definition of "pressure relief
V
valves" as discussed at 63 FR 57762.
Response: The definition in the CAR for "pressure .relief
device or valve" specifically notes that "devices activated. . . by
a vacuum are not pressure relief devices." A "relief device or
valve" under the CAR means a "device or valve used only to
release an unplanned, nonroutine discharge," not necessarily
relieving pressure. Pressure relief devices are a subset of
relief devices. There are specific equipment leak requirements
for pressure relief devices in subpart F of the CAR, but the CAR
does not have requirements for relief devices in general.
3-4
-------
The change noted by the commenter (and discussed at 63 FR
57762) is regarding the definition of "open-ended valve or line."
This term is defined differently in the referencing subparts, but
the intent is the same: relief devices, the broader category of
devices needed for safety purposes or equipment protection, are
not considered open-ended valves. The CAR language therefore
specifically exempts "relief valves" in the definition of
"open-ended valve or line" instead of exempting "pressure relief
valves," as was done in the HON and 40 CFR part 61, subpart V.
(No change has been proposed to the HON definition of "pressure
relief device or valve" or "open-ended valve or line.")
Comment: One commenter (IV-G-01) recommended that the terms
"alternative test method" and "equivalent test method" found in
the CAR subpart A, §65.2 should be combined into one term and
used consistently, because they seem to mean the same thing.
Another commenter (VI-D-02) asserted that there is no reason
for the CAR to require a Method 301 demonstration for alternate
test methods, if the Administrator has approved an alternate as a
general matter and thereby determined the alternate to be
generally applicable. The commenter (VI-D-02) recommended
deleting this condition from the definition of "alternative test
methods."
Response: The definitions and requirements for alternative
test methods and for equivalent methods are not the same. The
V
two terms are used in different circumstances. Equivalent
methods are demonstrated to show "a consistent and quantitatively
known relationship" to the required test method for the specific
condition under which that test method would be run. An
alternative method provides results adequate for the
Administrator's determination of compliance.
We have provided more detail on the types of changes that
constitute alternatives to test methods by adding definitions and
examples for minor, intermediate, and major changes. These
definitions have obviated the need for the proposed definition of
"alternative test method." As indicated in these definitions,
3-5
-------
only major and intermediate changes to test methods must undergo
a Method 301 demonstration. Further, in those limited situations
where we determine that a Method 301 demonstration is
representative of an entire source category, we may approve a
major alternative for application to the entire source category,
such that only the -initial application of the change would
undergo a Method 301 demonstration.
There is a need to maintain the requirement to perform
Method 301 for changes to test methods that have not been
approved by the Administrator. By specifying that Method 301 be
used, we ensure that we will receive the necessary data to
evaluate an alternative to a test method. Also, companies
submitting alternative test methods for approval will know what
information is expected.
Comment: Two commenters (VI-G-03, VI-G-04) commented on
confusion in the definitions for recapture, recovery, and control
devices. One commenter (VI-G-03) supported the comments of
VI-G-04 on this issue. One 'commenter (VI-G-04) contended that
the goals of "regulatory reinvention" mandate that we use the CAR
as an opportunity to eliminate the artificial distinction between
recapture, recovery, and control devices and articulate a simple,
straightforward principle applicable to any device that controls
emissions. The commenter (VI-G-04) stated that, as proposed, the
CAR definitions are overly complex and penalize the use of
i
"recovery devices" to achieve emission control. The commenter
(VI-G-04) further stated that this approach not only results in
confusion in permitting, compliance, and enforcement of the CAR,
but also creates a regulatory disincentive for the recycling of
recovered chemical materials. The commenter (VI-G-04) added that
these results are contrary to the goals of the White House paper
"Reinventing Environmental Regulation" which emphasizes deleting
"conflicting", "overlapping", "duplicative", and "confusing"
requirements in favor of regulations that are "understandable to
those who are affected by them."
3-6
-------
The commenter (VI-G-04) referenced the definitions for
"control, recovery, and recapture devices" as confusing and
specifically called attention to the exclusion of recovery
devices as control devices (in the control device definition) and
the exception to this exclusion in §65.63(a)(2)(ii). The
commenter (VI-G-04) stated that this exception was confusing,
difficult to apply, and unjustified from an environmental
standpoint. The commenter (VI-G-04) elaborated on why the
conditions of the exception are unjustified from an environmental
standpoint. The commenter (VI-G-04) also contended that the
proposed CAR creates a recycling penalty and encourages the use
of recapture devices where a waste is formed. The commenter
(VI-G-04) stated that this is a practice that should be
discouraged. The commenter (VI-G-04) stated that if we are
concerned that a recovery device may actually be a part of the
process, that this concern should be addressed by a simple,
straightforward statement in the rule (i.e., that a recovery
device may not be relied upon as a control device if it is
necessary to the operation of the emission source). The
commenters (VI-G-03, VI-G-04) suggested that recapture and
recovery devices should be considered control devices to the
extent that they are relied upon by the owner or operator to
reduce emissions. The commenter (VI-G-04) stated that this can
be accomplished by shortening the definition of "control device"
i
to just the first two sentences in the proposed definition and
the exclusion and exceptions in §65.63(a)(2) should be deleted.
Response: The recapture device concept was added to the HON
in recent changes to that rule. This change was discussed in the
HON amendment proposal preamble on August 26, 1996 (61 FR 43704);
the change was promulgated in January 17, 1997. In the proposal
preamble, we explain that this change fills a "regulatory void
for non-combustion/non-recovery devices while preserving the
approach used in this rule (and earlier rules) to differentiate
between process and control in this industry." We point out in
this discussion, the NSPS process vent provisions (40 CFR 60,
3-7
-------
subparts III, NNN, and RRR) and the HON provisions, prior to this
change, treated all carbon absorbers, condensers, adsorbers, and
scrubbers as "recovery devices" and never considered these
devices to be used in situations where the material was captured
and disposed of. Therefore, the recapture provisions were added
to the HON to include these devices as a compliance option in
situations when they are not used as recovery devices (i.e., when
they are used as recapture devices).
In the final CAR, an additional change was made to provide
clarity and reduce the complexity of these provisions. The CAR
has eliminated the term "recapture device" and now relies on the
term "control device" in the situation where this term has been
used. Even under the proposed CAR all recapture devices were
control devices. We have used this fact in the final rule to
simplify the language. Elements of the recapture device
definition have been incorporated into the control device
definition. This is in order to make it clear that, for process
vents, control devices include non-combustion/non-recovery' •
devices (devices that capture material that is not used, reused,
or sold), but do not include recovery devices (devices that
recover material that is used, reused, or sold).
The CAR does not penalize the use of recovery devices as the
commenter suggests. This change has not changed the fact that
recovery devices are allowed to be used to comply with the rule
V
by raising the TRE index value. Under the NSPS, HON, and CAR
process vent provisions, equipment is considered to be part of
the process if the recovered materials are used, reused, or sold
for use as the chemical or for fuel. This provision is stated in
the rule by specifying that all applicability determinations take •
place after the final recovery device. If the process vent meets
the criteria for TRE, flow, and concentration then control is
required. In the process vent rules, three options are given for
compliance: (1) a flare can be installed; (2) a control device
(not a recovery device) meeting 98 percent emissions reduction or
a 20 ppmv outlet concentration can be installed; or (3) a
3-8
-------
recovery device can be installed that increases the TRE to
greater than 1.0 -- this option, therefore extends the process
out to the end of the new recovery device. Although the rule
does not consider a recovery device to be a control device for
process vents (because recovery devices are part of the process),
this third option encourages the use of recovery devices (or
process changes) to prevent pollution. By recovering and reusing
chemicals, emissions can be reduced such that the TRE is
increased to greater than 1.0 and a control device does not need
to be applied. The recovery device does not need to meet the
98 percent reduction or 20 ppmv criteria, it only needs to
increase the TRE to greater than 1.0. (In many cases, less than
98 percent reduction can increase the TRE to greater than 1.0.)
By making the distinction between recovery devices and non-
combustion/non-recovery devices that capture material not used,
reused, or sold, we have expanded the list of control devices
that can be used to meet option (2) (the 98 percent emissions
reduction or 20 ppmv outlet concentration) to include non-
combustion/non-recovery devices as well as combustion control
devices.
Recovery devices are part of the process and are therefore
not control devices. Non-combustion/non-recovery devices that
capture material that is not used, reused, or sold are not part
of the process and are therefore control devices. The
V
distinction between these devices is necessary to maintain the
distinction between process and control equipment. The .
distinction between process and control equipment was an integral
part of the data analysis used to support the process vent rules.
We do not consider this an arbitrary distinction.
It would not provide adequate guidance to simply state in
the rule that a recovery device can not be relied upon as a
control device if it is necessary to the operation of the
emission source, as suggested by the commenter. There would be
several possible interpretations on whether a recovery device is
integral to a process, for instance:
3-9
-------
• is a recovery device that provides an increase in
production efficiency greater than 10 percent integral
to a process? what about a 5 percent increase?;
• is a recovery device that decreases the use of raw
materials by 5 percent integral? what about a
2 percent decrease?
The convention that the last recovery device is considered part
of a process has been used since 40 CFR 60, subpart III was
proposed in October 1983 and in three other rules since. We
consider this convention fair, understandable, and easily
applied.
The exception to the exclusion of recovery devices used to
meet the 98 percent control requirement specified in
§65.63(a)(2)(ii) was added to the HON and incorporated into the
CAR to provide for a specific situation. Industry brought to our
attention the situation where a control device exists after the
final recovery device and the control device can not meet the
98 percent control requirement and can not meet the 20 ppmv
outlet concentration requirement consistently. In this situation
the facility is faced with the decision to stop reusing the
material recovered in the recovery device or add an additional
device. (By no longer using the material recovered in the
recovery device, the recovery device can be considered a control
device and be used to help meet the 98 percent control
requirement.) Because of the costs, the facility would be more
\
likely to stop reusing the recovery material, thereby creating an
additional solid waste stream and not reducing emissions, any
further. We decided to avoid the creation of additional solid
waste streams, especially with no further reduction of emissions,
and to allow these recovery devices to be considered control
devices. There are several conditions required to be met before
this exception can be used. These conditions are necessary to
make sure that the exception is only available for these specific
situations.
Comment: One commenter (VI-G-03) requested a deletion of
the sentence within the definition of "malfunction" that reads,
3-10
-------
"Failures that are caused in part by poor maintenance or careless
operation are not malfunctions." The comraenter noted that the
terms "poor maintenance" and "careless operation" are not
defined. The commenter believes the standard for identifying a
malfunction is already articulated in the definition by the terms
"sudden, infrequent and not reasonably preventable," and the
commenter therefore maintains that the terms "poor maintenance"
and "careless operation" are ambiguous, unenforceable,
unnecessary, and likely to lead to litigation.
Response: The language in the CAR definition of
"malfunction" regarding "poor maintenance or careless operation"
is identical to the language in the definition sections of the
40 CFR part 60 and part 63 general provisions. This language
exists in the definition to add clarity to the "sudden,
infrequent and not reasonably preventable" language. It
specifies that any event that was caused in part by poor
maintenance, for instance, is not a malfunction. We contend this
clarification is necessary and that its meaning is clear.
Comment: One commenter (VI-G-03) suggested that the
definition of "permit program" should include State permitting
programs that are not part of the title V program and Federally
Enforceable State Operating Permits programs for minor and
synthetic minor sources.
Response: In the context of the CAR, "permit program"
V
refers to title V permits established pursuant to part 70 or 71;
it does not refer to other types of permit programs. Changing
the definition as the commenter has suggested would only add
confusion where this term is used.
Comment: A commenter (IV-G-01) pointed out that the
definition of regulated source in subpart A of the CAR does not
match the definition in the proposal preamble.
Response: There is no definitions section in the proposal
preamble, therefore, we assume the commenter is referring to the
text in section V.A, Amendments to the Referencing Subparts:
General Concepts (63 FR 57758). In this section, "regulated
3-11
-------
source" is described. The words used in this description are
different than those used to define it. However, the meaning is
the same.
Comment: One commenter (IV-G-01) indicated that "maximum
true vapor pressure" as used in subpart C, §65.41(b) of the CAR
should be defined. •
Response: Maximum true vapor pressure is a term necessary
for determining the applicability of control for storage vessel
rules. Therefore, it is defined with the applicability
provisions in the referencing subparts. The definition from the
applicable referencing subpart should be used when that term is
encountered in the CAR.
Comment: One commenter (IV-G-01) noted that the term
"referencing subpart" does not include HON subpart F.
Response: The commenter is correct. 40 CFR part 63,
subpart F is not listed as a referencing subpart in the
definition of referencing subpart. Subpart F is not a
referencing subpart because 'there is no reference to the CAR from
subpart F. The references from the HON to the CAR are in 40 CFR
part 63, subpart G for storage vessels, process vents, and
transfer racks, and in 40 CFR part 63, subpart H for equipment
leaks.
Comment: Two commenters (IV-G-01, VI-D-03) pointed out
that, in subpart A of the CAR, the definition of a closed loop
system contradicts the definition of a closed vent system. One
commenter (IV-G-01) stated that the definition of a closed vent
system intentionally excludes piping going back to the process
from the scope of a closed vent system. The commenters (IV-G-01
and VI-D-03) recommended that "except through a closed vent
system" be removed from the definition of a closed loop system.
Response: We agree with the commenter that there is
potential for confusion regarding the definitions of a closed
vent system and a closed loop system. . Therefore, we have revised
the definition of a closed loop system in subpart A of the CAR to
3-12
-------
state that a "closed loop system means an enclosed system that
returns process fluid to a process."
Comment: One commenter (VI-D-03) asserts that the first
sentence of the CAR control device definition implies that a
combustion device fueled from a fuel gas system is a control
device. The proposed definition is inconsistent with the
treatment of fuel gas systems by the CAR. The commenter
suggested a revision of the control device definition to clarify
that combustion devices that only burn fuel as a primary fuel are
excluded from the definition, as follows: "Control device means
any combustion device combusting a regulated material as a
secondary fuel, recovery device, recapture device, or any
combination of these devices used to comply with this part...."
(Suggested language is underlined.) The commenter (VI-D-03)
asserted that this change would allow simplification of the
boiler and process heater language in the CAR subpart G, because
the exceptions for primary fuel could be eliminated. For
example, the commenter said §65.149(b)(2)(ii) could be deleted
and paragraph (c)(1) of the same section could be greatly
simplified.
Response: The commenter is correct when pointing out that
the first sentence of the CAR control device definition does seem
to imply that a fuel gas system could be considered a control
device. However, the fourth sentence clearly states "A fuel gas
i
system is not a control device." We consider this clarification
to be sufficient to avoid the confusion suggested by the.
commenter. Also, the change suggested by the commenter would not
be appropriate in situations where the regulated stream is the
primary fuel, such as in the case of some flares.
3.1.2 Startup. Shutdown, and Malfunction
Comment: Three commenters (IV-G-01, VI-D-02, VI-D-03)
pointed out that the requirement in §65.6(c)(2) of the CAR is
inconsistent with the preamble of the CAR which states that
provisions in the part 63 general provisions regarding immediate
reporting of periods of startup, shutdown, and malfunction have
3-13
-------
not been included in the CAR. One commenter (IV-G-01) advised
that the startup, shutdown, and malfunction immediate reporting
requirement in §65.6 (c) (2) of the CAR should be removed because
it is inconsistent with the provisions from subpart G of the HON.
The commenter noted that the HON requires sources to report such
actions in the periodic report instead'of an immediate report.
Response: We intended to be consistent with HON
requirements and therefore did not intend to require immediate
reporting of periods of startup, shutdown, and malfunction under
the CAR. This inadvertent error has been corrected by removing
this requirement from §65.6 (c). As noted in the proposal
preamble, requiring these reports with the periodic reports
instead of as immediate reports not only sufficiently ensures
compliance but also provides for report consolidation.
Comment: One commenter (IV-G-01) recommended that we remove
the criteria under §65.6 (c) (1) (ii) of the CAR, which require that
the periodic startup, shutdown, and malfunction reports include
the number of startup, shutdown, and malfunction events and- the
total duration of all periods of startup, shutdown, and
malfunction for the reporting period. The commenter claimed that
this section only requires reporting if the total duration
exceeds either of the durations specified in §65.6 (c) (1) (ii) (A)
or (c) (1) (ii) (B) and the commenter (IV-G-01) objects to the
criteria because they add calculation burdens that are not
*
required by the referencing subparts.
The commenter (IV-G-01) suggested that, if we retain the
provisions in §65.6(c) (1) (ii) (A) and (c) (1) (ii) (B) of the CAR, we
should clarify specifically what startup, shutdown, and
malfunction information sources must include in their periodic
startup, shutdown, and malfunction reports. The commenter
(IV-G-01) indicated that it is often easier to report all
startup, shutdown, and malfunction periods that caused excess
emissions rather than determining the percentage of time a
Continuous Parameter Monitoring System (CPMS) is not operating or
3-14
-------
is malfunctioning, or the percentage of time in which startup,
shutdown, and malfunction events caused excess emissions.
Regarding §65.6(c) (1) (ii) (A) , the commenter (IV-G-01)
specifically requested clarification of the assumption that a
separate downtime percentage should be calculated for each
individual CPMS and that only startup, shutdown, and malfunction
event durations need to be reported on CPMS for which the period
of inoperation or malfunction is equal to or greater than
5 percent. The commenter (IV-G-01) also requested clarification
of whether to include or exclude periods of CPMS inoperation or
malfunction when the actions taken by an owner or operator are
consistent with the procedures specified in the source's startup,
shutdown, and malfunction (SSM) plan. The commenter (IV-G-01)
also asked that we clarify which periods of CPMS inoperation or
malfunction should be included in the startup, shutdown, and
malfunction report and requested that only the periods of CPMS
inoperation or malfunction for which actions are inconsistent
with the SSM plan be included.
Regarding §65.6(c) (1) (ii) (B) of the CAR, the commenter
(IV-G-01) requested that we clarify how to calculate the
percentage of startup, shutdown, and malfunction time during
which excess emissions occur because of inconsistency in the
definition of "regulated source." The commenter (IV-G-01)
requests that the percentage should be calculated on an emission
point by emission point basis considering the operating time of
each emission point. The commenter (IV-G-01) assumes that when
one emission point exceeds 1 percent of that regulated source's
operating time, the information in §65.6(c)(1)(ii) of the CAR
needs to be reported only for that emission point. The commenter
(IV-G-01) requested clarification regarding which periods of
startup, shutdown, and malfunction should be included in the
percentage calculation. The commenter (IV-G-01) requested that
only the periods of startup, shutdown, and malfunction for which
actions are inconsistent with the SSM plan be included in the
3-15
-------
percentage calculation and the periodic startup, shutdown, and
malfunction report.
Response: A semi-annual summary report of the occurrences
and durations of each startup, shutdown, and malfunction during
which excess emissions occur is required by the CAR general
provisions. We consider the semi-annual summary report an
important addition to the startup, shutdown, and malfunction
provisions, because it would highlight when startup, shutdown,
and malfunction conditions exist for a significant amount of
time, and it would indicate conditions that happen frequently
during a semi-annual period. The purpose of reporting only above
certain thresholds was to reduce burden, primarily if occurrences
are rare or infrequent.
Nevertheless, we agree with the commenter's assertion that
it may be easier to report all startup, shutdown, and malfunction
periods that caused excess emissions, rather than reporting the
percentage of time a CPMS is malfunctioning or not operating, or
the percentage of operating time in which regulated source
startup, shutdown, and malfunction events occurred that caused
excess emissions. We consider the determination of the operating
time for the various regulated sources to be burdensome and
difficult to keep track of, making the calculation of the
percentages difficult. Therefore, we have eliminated the
requirement to calculate the percentage of time an SSM event
V
occurs and the reporting exemption associated with it. The
facilities that would have benefitted from this exemption, those
with infrequent SSM events, will have very little to report and,
therefore, will only see a small increase in burden. The final
rule requires the owner or operator to report all startup,
shutdown, and malfunction periods.
We have clarified the information that must be included in
the periodic startup, shutdown, and malfunction report as
requested by the commenter. The report must include the number
of discrete startup, shutdown, and malfunction events and how
3-16
-------
long these events lasted for each regulated source and CPMS
regardless of whether the SSM plan was followed or not.
Comment: One commenter (IV-G-01) noted that the last
sentence of the reporting discussion in the proposal preamble
(63 FR 57767) should use the terms "greater than or equal to" in
reference to total duration of excess emissions and CPMS downtime
instead of "less than."
Response: The commenter is correct in that the preamble
should have read "greater than or equal to" in the sentence
referenced by the commenter. However, the paragraphs that this
proposed preamble text refers to (§65.6 (c) (1) (ii) (A) and
(c) (1) (ii) (B)) have been deleted in the final CAR for the reasons
mentioned in the previous response.
Comment: One commenter (IV-G-01) expressed that it is not
accurate for the CAR preamble to state that "[SSM] plans are
considered inadequate under the HON if they fail to provide for
the operation of the regulated source during startup, shutdown,
and malfunction to minimize emissions to at least the levels
required by all relevant standards." The commenter points out
that the HON was revised to override the provision in
§63.6(e) (1) (i) of 40 CFR part 63, subpart A and addresses'the
issue in §63.102(a) (4) of 40 CFR part 63, subpart F. The
commenter (IV-G-01) states that §63.102(a)(4) provides that
"during startups, shutdowns, and malfunctions when the
requirements of this subpart [subpart F], subparts G and/or H of
this part do not apply..., the owner or operator shall implement,
to the extent reasonably available, measures to prevent or
minimize excess emissions to the extent practical...." The
commenter (IV-G-01) asserted that the preamble to the CAR should
imply that HON sources are subject to a different standard. The
commenter (IV-G-01) also requested that future rulemaking should
reflect the reality that we recognize that it is "impracticable,
as well as contradictory... to expect sources to continually meet
applicable emissions standards while experiencing a startup,
shutdown, or malfunction."
3-17
-------
Response: We acknowledge that the CAR preamble had an
incorrect reference to the HON. The reference should have been
to the general provisions of part 63, as the commenter (IV-G-01)
points out by referencing the general provisions §63.6 (e) (1) (i)
of 40 CFR part 63, subpart A. The CAR has the same wording used
by the HON on this issue. The paragraph referenced by the
commenter, §63.102(a)(4), is incorporated in the CAR as
§65.3(a)(4).
In regard to the commenter's request that the Agency's
future rulemaking should reflect the reality that the Agency
realizes that it is "impracticable, as well as contradictory... to
expect sources to continually meet applicable emissions standards
while experiencing a startup, shutdown, or malfunction," we will
determine the provisions that are appropriate for future
rulemakings when these rules are developed. The commenter is
free to make these comments for future rulemakings during the
public comment period for those rules.
Comment: One commenter (IV-G-01) suggested that the
proposal preamble should be amended to indicate that excursions
will not count against the number of excused excursions when an
SSM plan applies and is followed.
Response: The commenter is correct that an excursion that
occurs during a startup, shutdown, or malfunction event in which
the SSM plan is followed does not count against the number of
V
excused excursions. The CAR is clear on this point.
Comment: Two commenters (VI-G-03, VI-D-06) support, our
decision not to require the incorporation of the SSM plan into a
facility's title V permit as a requirement in the CAR. One
commenter (VI-G-03) stated that the CAR SSM plan and procedures
provide useful flexibility. The other commenter (VI-D-06) agrees
with the statements in the preamble that incorporating the plan
into the title V permit would be counter-productive due to the
fact the SSM plans will need to be modified periodically. One
commenter (IV-G-03) elaborated on their support by comparing
these provisions in the CAR with those in the HON. The commenter
3-18
-------
(VI-G-03) stated that the HON requires, unlike the CAR, that the
SSM plan be incorporated by reference into the title V permit.
Because of this, the commenter (VI-G-03) inferred that a title V
permit modification would be necessary each time the SSM plan is
changed. The commenter (VI-G-03) believes the SSM plan should be
updated as often as necessary to accommodate current process
conditions and scenarios. The commenter (VI-G-03) asserted that
a facility should not have to apply for a permit modification to
maintain its SSM plan. The commenter (VI-D-06) suggested that
the CAR'S approach for SSM plans should be used for other rules,
such as MACT standards, where a source is required to develop SSM
plans.
Conversely, another commenter (VI-D-01) recommended that the
CAR adopt similar requirements to 40 CFR part 63 general
provisions. The commenter (VI-D-01) noted that the
40 CFR part 63 general provisions require a reference to a
facility's SSM plan in that facility's title V permit. The SSM
plan is then required to be filed on site.
Response: As stated by some of the commenters and in the
proposal preamble (63 FR 57766), the CAR does not require the SSM
plan itself to be incorporated by reference into a source's
title V permit because of the frequency at which SSM plans may be
updated. The general provisions of part 63 were used as the
basis for the SSM provisions of the CAR, and they appear to
V
require incorporation by reference of the plan itself by stating
in §63.6(e) (3) (i) :
"The plan shall be incorporated by reference into the
sources title V permit."
However, a clarification to this language has been released in a
memorandum from John Seitz, Director of OAQPS to Regional Air
Directors (January 18, 1996) entitled "Incorporation of Startup,
Shutdown, and Malfunction Plans into Source's Title V Permits."
This memorandum clarifies the language by stating:
"The language in §63.6(e)(3)(i) is to ensure that the
requirement to prepare and implement a SSM plan is
explicitly stated within a source's operating permit. Our
3-19
-------
intention is not for the contents of the plan to be actually
written into the permit."
Therefore, the commenter (IV-G-03) is incorrect in stating that
the HON requires the plan to be incorporated by reference into
the title V permit. The HON does not specifically state that the
plan must be incorporated by reference.but relies on the
requirements of §63.6(e)(3)(i). As just explained, the
provisions of §63.6(e)(3)(i) have been clarified in the
memorandum to mean that the requirement to prepare and implement
a SSM plan must be in the source's operating permit. The SSM
requirements under the HON are the same as those in the CAR but
the CAR is clearer in that it incorporates the clarification we
released in the before mentioned memorandum.
In response to the other commenter (VI-D-01), the CAR does
adopt the requirements of the general provisions of part 63 but
it incorporates this clarification.
In reviewing these provisions, we have determined that the
requirement to keep the SSM plan onsite needs to be stated
explicitly. Although it is obvious that the plan must be kept
on-site based on the general record retention requirements in
§65.4(c), it is not completely clear that the plan must be
retained on-site after 2 years (or 6 months as specified in
§65.4(c)(2)). Therefore, we have clarified in the SSM provisions
that the SSM plan must be retained on-site. This is necessary
because of the frequency in which this document may change and v
the need to have it available for review. Also, among other
reasons, the document may need to be revised if found to be
inadequate.
Comment: One commenter (VT-D-02) requested a revision to
§65.3(a)(3) to recognize that shutdown of control devices is
allowed during periods of planned routine maintenance for control
devices used on storage vessels, as specified elsewhere in the
CAR.
Response: In the language at §65.3(a)(3), shutdown of
control devices is prohibited "during times when emissions are
3-20
-------
being routed to such items of equipment if the shutdown would
contravene requirements of" the CAR. During planned routine
maintenance of a control device on a storage vessel, the
requirement to operate the control device does not apply [see
§65.42(b)(4) and (b)(5)(iii) of the CAR]. Therefore, a shutdown
of the control device would not contravene the requirements of
the CAR and would be allowed under §65.3(a)(3), as drafted in the
proposed rule. In order to make this exception clear, we have
edited the final language of §65.3 (a) (3) to reference
§65.42(b)(4) and (b)(5)(iii).
3.1.3 Reporting
Comment: Three commenters (VI-D-02, IV-G-01, VI-G-03)
supported limiting the amount of time for approval or disapproval
of a schedule change for reporting. Two commenters (IV-G-01,
VI-G-03) advocated that, under §65.5(h)(3) of the CAR, an owner
or operator should be able to assume approval of a request for an
adjustment to a time period or postmark deadline if written
disapproval is not received within 15 days. Another commenter
(VI-D-02) suggested modifying §65.5(h) to provide automatic
approval of requests in 90 days, unless the Administrator
indicates disapproval or a desire to review the request, in which
case the change would not be allowed until the review is
completed. One commenter (IV-G-01) stated that many requests go
unanswered and that the commenter's company has been waiting on
V
some approvals for more than 1 year.
Response: We acknowledge that it is desirable for the
Administrator to approve or deny requests within the designated
15-day period. However, it is important that the changes in
schedule be made upon mutual agreement between the facilities and
the States, because both parties must change their respective
schedules for handling the reports. Therefore, it is not
appropriate to grant a blanket approval for all requests that go
unanswered in that period of time. It is suggested that the
facilities consider their experience with typical turn-around
3-21
-------
times on requests and factor that experience into their schedules
when submitting requests.
Comment: Similarly, one commenter (VI-G-03) advocated that
a time frame be provided for the Administrator's action on
requests for a determination of equivalence for alternative
emission limits. The commenter (VI-G-03) argued that alternative
emission limits that provide equivalent protection should not
need to be published in the Federal Register. The commenter
stated that in the standard for determination of equivalency, the
use of the term "performance standards" is confusing and should
be deleted.
Response: In regards to requests for alternative means of
emission limitation, there are several steps required for the
Administrator to approve or deny these requests. For this
reason, it is difficult to require a specific time frame for this
review. It is suggested that the facilities consider their
experience with typical turn-around times on requests and factor
that experience into their schedules when submitting requests.
We assume that the commenter is referring to the alternative
means of emission limitation provisions of §65.8 when referring
to "alternative emission limits." The alternative means of
emission limitation provisions allow a source to propose an
alternative to any design, equipment, work practice, or operation
standard. A source can not propose an alternative to a
performance standard which might be in the form of an emission
limit or percent emission reduction. In the case of performance
standards, it is not necessary to propose alternatives because
performance standards allow any means of reduction a source
chooses as long as it meets the performance level, the emission
limit, or emission reduction. The inclusion of the phrase "(but
not performance standards)" is to clarify that alternative means
of emission limitation requests are not appropriate for
performance tests. We consider this clarification helpful.
We maintain that public review through publication in the
Federal Register is appropriate because while the requirements of
3-22
-------
the CAR have undergone public review, the requested alternatives
have not.
Comment: One commenter (IV-G-01) suggested that, for the
purpose of using an alternative means of emission limitation, it
may be unreasonable and costly to require an owner or operator to
demonstrate the emission reduction achieved by a required work
practice for 12 months. The commenter (IV-G-01) reasoned that we
should already know the emission reduction achieved by the
required work practice, and that the owner or operator should
only have to demonstrate the emission reduction achieved by the
proposed alternative work practice. The commenter (IV-G-01) also
noted that in §65.102(d) (2) (ii), the term "equipment" is used
rather than "work practice."
Response: We agree with the commenter and the 12-month
requirement, §65.102(d) (2) (ii), was removed from the final
regulation. It is the operator's burden to provide sufficient
data to support an alternative work practice which could be more
or less than 12 months.
The commenter is incorrect in that the term "work practice"
should have been used in §65.102(d)(2)(ii) instead of
"equipment." This paragraph refers to each equipment type not
each work practice.
Comment: One commenter (VI-D-02) recommended revising
§65.4(a) to refer to the regulated source or site, rather than to
V
the owner or operator. The commenter contended that record
retention requirements should be specified in terms of sites or
sources that are or are not subject to title V, because owners or
operators may have to comply with title V at one site but not at
another.
Response: We have made changes in the final rule in
§65.4(a)(1) and (a)(2) to make the clarification suggested by the
commenter.
Comment: One commenter (IV-G-01) requested that we clarify
which months the first periodic report should cover for sources
electing to comply with the CAR upon initial startup. The
3-23
-------
commenter suggested that the first periodic report should cover
the 6-month period beginning on the date the Initial Compliance
Status Report is due, because parameters may not be established
for 240 days after the applicable compliance date specified in
the referencing subparts, or 60 days after the performance test.
Another commenter (VI-G-03) suggested that the language at
§65.5 (e) (2) for periodic report due dates is too complicated.
The commenter (VI-G-03) suggested alternative language as
follows: "The report is due on or before February 28 for the
period of July 1 through December 31 and August 31 for the period
of January 1 through June 30."
Response: The commenter (IV-G-01) is correct that the
proposed CAR is not clear on when the first periodic report is
due for sources electing to comply with the CAR upon initial
startup. The final rule has been clarified to state that the
first periodic report covers the 6-month period after the Initial
Compliance Stats Report is due. As the commenter noted, the
parameter ranges will be included in the Initial Compliance
Status Report and before that time there will be no basis for
judging performance.
We acknowledge that the language would be simpler if the
periodic reporting 6-month periods were required to be January 1
through June 30 and July 1 through December 31. However, the
compliance dates do not often occur at the beginning or middle of
the calendar year. In order to ensure that all periods are being
reported on, the due date must be triggered by the compliance
date. Also, the report is due within 2 months after the 6-month
period. It is difficult to explicitly state these requirements,-
some sources may find this paragraph confusing. However, the
source has the opportunity, through the title V process and by
less formal means, to revise or adjust the semiannual report
schedule with the delegated authority. Also, through §65.5(h)
the source and the delegated authority can agree on a different
schedule. We consider the clarity and flexibility sufficient to
make this paragraph workable.
3-24
-------
Comment: One commenter (VI-G-03) asserted that to maintain
certainty, consistency, and the intent of the CAR, that record
retention periods should only be specified in §65.4(a)(1) and
(a)(2). The commenter (VI-G-03) objected to the regulated
community having to search throughout the CAR for longer
retention times than those specified in §65.4(a)(1) and (a)(2).
The commenter (VI-G-0'3) asked that all other requirements or
references to retention periods other than those found in
§65.4(a)(l) and (a)(2) be removed.
Response: It is necessary to keep some records longer than
the retention times listed in §65.4(a)(l) and (a)(2). These are
situations where the information does not change often but it is
still necessary to have the information available, such as
monitoring data for connectors monitored every 8 years. To add
clarity and to avoid the regulated community having to search
throughout the CAR for longer retention times, we have listed in
§65.4(a)(1) and (a)(2) all paragraphs where longer retention
times are required.
Comment: One commenter (VI-D-02) noted that the current
language in §65.5(d)(2) appears to require submission of the
initial compliance status report on the 240th day after the
compliance date or on the 60th day after the performance test.
The commenter (VI-D-02) requested a change of the wording to
§65.5(d)(2) to specify that the report is required within
240 days "or within 60 days after the completion of the
performance test."
Response: We agree that the initial compliance status
report does not have to be submitted on the 240th day after the
compliance date or on the 60th day after the performance date.
We have made the suggested edits so that §65.5(d)(2) requires
submission within the 240 or 60 day time periods.
Comment: One commenter (VI-D-02) noted that §65.5(i)
requires some information to be submitted via title V or
otherwise, but does not specify when or for what purpose. The
commenter (VI-D-02) asked if this is the content requirement for
3-25
-------
the compliance status report and, if so, if the requirement
should be moved to paragraph (d).
Response: The items listed in §65.5 (i) are information that
we assume would be included in the source's title V permit
application. However, in the case where the owner or operator
may not think to include this- information, we are explicitly
listing these items as needing to be included in the title V
permit. This was also done in the HON at §63.152(e).
3.1.4 Compliance
Comment: One commenter (VI-G-03) asserted that §65.3(b) and
(c) fail to provide standards for determination of compliance.
The commenter contended that the terms "acceptable operation and
maintenance procedures" and "proper operation and maintenance
practices" are subjective and do not provide plant personnel with
guidance on what is actually required. The commenter asserted
that this will lead to unnecessary litigation. The commenter
stated that these subjective standards should be replaced by
objective standards such as performance tests, emission ' •
standards, specified operating parameters and other
specifications in the rules, permits, and applicable startup,
shutdown, and malfunction plans.
Response: The provisions of these paragraphs [§65.3(b) and
(c) ] are general provisions giving the Administrator the
authority to use a variety of sources of information to make
findings of compliance. These paragraphs specify several types
of information that can be used to determine if "acceptable
operation and maintenance procedures" or "proper operation and
maintenance practices" have been followed. These provisions also
refer to the more specific provisions for performance tests,
monitoring, and SSM plans.
Also, the language of these paragraphs is consistent with
the 40 CFR part 63 general provisions and provisions of 40 CFR
63.152 of the HON. We consider this language to be sufficiently
clear.
3-26
-------
Comment: One commenter (VI-G-03) wanted to know what the
consequence of having been "deemed to have failed to have applied
the control in a manner that achieves the required operating
conditions" as set forth in §65.3(b)(l). The commenter (VI-G-03)
noted that §65.3(b)(2) states that excursions are not violations.
Response: If a source has been "deemed to have failed to
have applied the control in a manner that achieves the required
operating conditions," then the source has violated the operating
standard. The commenter is incorrectly reading §65.3(b)(2).
This paragraph reads "An excursion is not a violation..., if the
conditions of paragraphs (b)(2)(i) or (b)(2)(ii) of this section
are met." We have edited this sentence to make it easier to read
and less likely to be read incorrectly.
Comment: One commenter (IV-G-01) stated that the
information in §65.3(b)(2) was a subset of the information in
§65.156(d). The commenter asserted that all of the information
should be in one place in subpart G.
Response: The provisions in §65.3(b)(2) are of a general
nature regarding excursions of parameter monitoring ranges.
These provisions are a part of the general compliance
determination procedures when parameter monitoring is used. The
provisions in §65.156(d) pertain to more specific requirements on
how to determine excursions, excursions that are not violations,
and how many excused excursions are allowed. We maintain that
the segregation of the general provisions and the more detailed
ones is important for understanding.
Comment: One commenter (VI-G-03) requested that the
following sentence be removed from §65.3(c): "The Administrator
will make findings of compliance with the standards of this part
using metric units." The commenter questioned if this would
require facilities to replace all English unit monitoring
equipment with metric unit equipment. If so, the commenter
contended this would be an unreasonably excessive cost. The
commenter stated that if facilities are allowed to keep English
3-27
-------
unit monitoring equipment, the compliance determination may infer
non-compliance due to rounding and conversion factors. The
commenter asserted that this would result in disputes that would
be burdensome to facilities, EPA, State enforcement, and
permitting agencies.
Response: Unlike the HON and many other MACT standards, we
have provided both English and metric units for all values in the
CAR. The HON and other rules only provide values in metric units
because the Administrator does make all findings of compliance
based on metric units. In order to help readers of the rule, we
have added the English units to the CAR. To clarify that finding
of compliance will only be made using metric units, we added the
sentence the commenter objects to. If this sentence were deleted
then all of the values in English units would also have to be
deleted. This would still not change the fact that findings of
compliance are based on using metric units. This has not meant
in the past nor does it mean now that sources must change their
monitoring equipment. It only means that any monitoring data
collected in English units will be converted to metric and
compared to the metric parameter monitoring ranges or emission
limits of the standard. The use of conversion factors are
unavoidable. See the US Code, title 15, section 205a for a
statement of the government's policy of metric units.
Comment: One commenter (VI-G-03) requested that the phrase
v
"or to protect personnel safety" be added to the following
language in §65.3(a)(3) regarding when it is allowable to shut
down items of equipment required by the CAR: "Paragraph (a)(3)
of this section does not apply if the item of equipment or CPMS
is malfunctioning or if the owner or operator must shut down the
equipment to avoid damage due to a contemporaneous startup,
shutdown, or malfunction from the regulated source or portion
thereof."
Response: The proposed paragraph §65.3(a)(3) has been
deleted in the final CAR. We decided that the language of this
paragraph was becoming complicated in an attempt to enumerate all
3-28
-------
the instances where equipment should not be operated during
startup, shutdown, or malfunction. Instead we determined that
the general duty language throughout the CAR [the new §65.3(a)(3)
which was the proposed §§65.3(a) (4) and 65.156(c) (2) and (c) (5)]
regarding SSM is sufficient to communicate the requirements
during an SSM.
Comment: One commenter (VI-D-02) asserted that it' was
intended for the CAR to allow one excused excursion of each
operating parameter per reporting period. However, the commenter
pointed out that the paragraphs covering excused excursions,
presumably §65.3(b)(4)(iii), are missing although they are
referenced from §65.3(b)(1) and (b)(4). The commenter suggested
that the missing paragraphs should be added. Another commenter
(VI-G-03) also noted that, there is no §65.3(b) (4) (iii) .
Response -. The reference to §65 .3 (b) (4) (iii) in §65.3(b)(l)
should have been §65.3(b)(2); and the reference to
§65.3(b)(4)(iii) in §65.3(b)(4) should have been §65.3(b)(4)(ii).
These cross-references have been changed in the final CAR. The
provision that the commenter refers to regarding the CAR allowing
one excused excursion is in §65.156(d)(2).
3.2 STORAGE VESSELS
Comment: One commenter (IV-G-01) suggested that the CAR
include an explanation that surge control vessels and bottoms
receivers are treated as storage vessels under the CAR, because
the CAR currently does not have a definition of storage vessels
in §65.2 of subpart A. Because this definition is absent from
§65.2, the commenter (IV-G-01) claims that the user is forced to
rely on the storage vessel definition in the HON (40 CFR part 63,
subpart F, §63.101), which specifically excludes surge control
vessels and bottoms receivers from the definition of storage
vessels. The commenter (IV-G-01) stated that, in this case, the
user would need to consult the amendments to the HON
(40 CFR part 63, subpart H) or 40 CFR part 61, subpart V to
3-29
-------
remember that surge control vessels and bottoms receivers should
be treated as storage vessels under subpart C of the CAR.
Response: We have clarified subpart C of the CAR to
indicate that surge control vessels and bottoms receivers will be
treated as storage vessels when referenced to subpart C. We have
also added a definition of bottoms receiver to 40 CFR part 61,
subpart V for completeness.
Comment: One commenter (IV-G-01) objected to the use of the
term "noncontact floating roof" found in §65.43(a)(4)(i) of the
CAR, as well as subpart Kb and the HON. The commenter points out
that subpart Ka only states "...each opening in the roof...."
The commenter maintains that the term "noncontact" is confusing
because the roof is required to float on the liquid.
Response: The term "noncontact" is defined in relation to
storage vessel floating roofs within the definition of external
floating roof (EFR). The definition for internal floating roof
(IFR) also alludes to a noncontact floating roof. To clarify
what is meant by the term, we have revised the IFR definition to
be as explicit as the EFR definition.
Comment: One commenter (VI-D-02) noted that §65.48(c)(1)
requires notification at least 30 days prior to refilling an IFR,
EFR, or EFR converted to an IFR, while §65.48(c)(2) requires
notification 30 days prior to an EFR seal gap inspection. The
commenter (VI-D-02) requested combining the notification
requirements of §65.48(c)(1) and (c)(2), where applicable, and
clarifying the proposed language.
The commenter (VI-D-02) asserted that §65.43(c)(5) and
§65.48(c)(2) should also be clarified that the seal gap
notification need not be submitted exactly 30 days before the
measurement, as the current language seems to suggest, but "at
least" 30 days before. The commenter (VI-D-02) also pointed out
that both seal inspections and refilling can occur with less than
30 days warning, when a storage vessel outage is unexpected.
Thus, the commenter (VI-D-02) pointed out, provision for shorter
3-30
-------
notification periods is needed in §65.48(c)(2) to parallel the
option provided in §65.48(c) (1) (ii) .
Response: There is no prohibition in the CAR from including
both notifications [§65.48(c)(1) and (c)(2)] in a single
submission. We maintain, however, that these provisions cover
different situations and that separate requirements for refilling
notifications and for seal gap measurements are appropriate.
We agree with the commenter that the language in
§65.48(c) (2) should be modified to clarify that the report is due
at least 30 days prior to the seal gap measurement. We have also
edited the CAR to provide an option for notification when
unplanned seal gap measurement occurs. We have edited the CAR to
include provisions in §65.48(c) (2) that parallel the options at
§65.48(c) (1) .
Comment: One commenter (VI-D-02) requested additional
wording as follows for §65.48(d) to clarify intent: "...shall be
based on the annual inspections required by §65.43(c)(1)(i) and
(c)(2)(ii)(A) and any observations made at other times when the
roof is viewed."
Response: We agree with the commenter and have made the
suggested edit to help clarify the intent of §65.48(d).
Comment: One commenter (VI-D-02) requested that the
requirements for EFR's converted to IFR's be spelled out in the
text of §65.45 rather than referring to paragraphs from the EFR
V
and IFR sections of subpart C. The commenter stated that this
will reduce confusion and enhance compliance for sources in this
situation.
Response: We have not edited the final CAR to include text
that spells out the requirements for EFR's converted to IFR's.
The proposed and final text clearly spells out the paragraphs
that apply.
Comment: One commenter (VI-D-03) requested a clarification
to §65.47(e) to indicate that the new record requirement for
landing a floating roof on its legs does not apply during periods
3-31
-------
of startup, shutdown, and malfunction. The commenter provided
suggested language for the clarification.
Response: It is not often that a storage vessel will be the
subject of a startup, shutdown, or malfunction but in these cases
the SSM plan must specify the actions that should be taken.
Depending on the effect of the SSM on the storage vessel, these
records may still be significant.
Comment: One commenter (VI-G-03) cautioned that
§65.43(a) (4) (iii) of the proposed CAR is confusing when compared
to §63.119(b) (5) (iii) because it does not include the phrase,
"...for the purpose of sampling."
Response: The phrase "...for the purpose of sampling" was
deleted in the CAR because it implied that you could have
penetrations for purposes other than sampling. The only
allowable penetrations are those specified in one of the
paragraphs under §65.43(a)(4).
3.3 PROCESS VENTS ' '
Comment: One commenter (IV-G-01) stated that the text of
the applicability section for process vents (§65.60) implies that
subpart D applies to only process vents that require control.
The commenter stated that this is not the case for Group 2A and
2B process vents. The commenter suggested the following changes
to the text:
i
" The provisions of this subpart and of subpart A of
this part apply to regulated material emissions from
process vents ."
Response: We agree that subpart D of the CAR is also
applicable to Group 2A and 2B process vents and that control is
not required for these process vents. However, we contend that
the language regarding being referenced from a referencing
subpart is important to ensure that only sources referred to
subpart D can use subpart D. We have modified the language in
the final CAR to read as follows:
" The provisions of this subpart and of subpart A of
this part apply to regulated material emissions from
3-32
-------
process vents where a referencing subpart references
the use of this subpart."
Comment: One commenter (IV-G-01) asserted that an owner or
operator should be allowed to designate a group 1 vent stream as
halogenated without having to perform any testing or
calculations. The commenter (IV-G-01) pointed out that this
allowance would assume a "worst case" and provide the
corresponding level of control. The commenter contended that the
allowance also would be similar to designating a process vent as
group 1 in accordance with §65.62(b)(1) of the CAR.
Response: We agree that it is reasonable to allow an owner
or operator to designate a group 1 vent stream as halogenated
without having to perform any testing or calculations. This
clarification has been made in the rule.
Comment: Commenter (IV-G-01) stated that the paragraph
heading in §65.63(a) (2) should be revised to add "by weight"
after 98 percent.
Response: The paragraph headings are meant to be short
descriptors to aid in reading the document. We do not consider
adding "by weight" an improvement in clarity and consider shorter
headings that still convey the point better than longer ones.
For this reason, we have revised this heading to remove "by
volume" after 20 parts per million. The heading now reads: "98
percent or 20 parts per million standard." The text of the rule
specifies the 98 percent be "by weight" and the 20 parts per v
million be "by volume."
Comment: One commenter (IV-G-01) suggested that
§65.63(f) (4) through (f) (6) refer to the corresponding sections
in recordkeeping §65.66(d) and reporting §65.67(b).
Response: The proposed and final CAR at §65.63(f) contain a
reference to §65.66(d). We have added a reference to §65.67(b)
in §65.63(f).
Comment: One commenter (IV-G-01) urged that §§65.64(c) and
S5.158(c)(2) of the CAR should be amended to include both Method
3-33
-------
18 and Method 26 or 26A in order to properly determine the
hydrogen halide concentrations in any applicable vent stream.
Response: The method cited for use in determining the
halogen status of a vent stream is correct in both §§65.64(c) and
65.158(c)(2). In these sections of the rule, procedures are
given to determine whether a vent stream is considered
halogenated and Method 18 is cited. Under the CAR, as under the
HON, if "the mass emission rate of halogen atoms contained in
organic compounds" is equal to or greater than 0.45 kg/hr the
vent stream is considered halogenated. Therefore, to determine
if a vent stream is halogenated, the organic and inorganic
halogens must be differentiated. Method 18 speciates the
halogens so that the "halogen atoms contained in organic
compounds" can be determined. Methods 26 and 26A do not
differentiate between organic and inorganic halogenated
compounds, so they cannot be used to determine whether a vent
stream is halogenated.
Methods 26 or 26A are required when determining if a
scrubber installed after a combustion device has reduced the
halogens by 99 percent during a performance test [see
§65.158 (c)] . Because many of the organic halogens would be
converted to inorganic halogens in the combustor, the distinction
between inorganic and organic halogens is not relevant.
Comment: One commenter (IV-G-01) asserted that footnote (a)
\
to table 2 in the CAR preamble (63 FR 57770 "CAR Process Vent
Group Determinations") is unclear, confusing, and does not
capture the points of the explanation given in the preamble at
63 FR 57770. The commenter (IV-G-01) suggested that the footnote
be revised as follows:
"The 50 ppm HAP concentration cutoff only applies to
40 CFR part 63, subpart G sources. Process vents
subject to only 40 CFR part 60, subparts RRR or NNN are
eligible for the 300 ppm TOC cutoff. There is no
concentration cutoff for subpart III sources. The
process vent provisions of subpart DDD are not
consolidated under subpart D of the CAR."
3-34
-------
Response: We agree and have revised the footnote to table 2
(63 FR 57770) to incorporate the recommendation of the commenter
in the final CAR.
Comment: One commenter (VI-D-03) counseled that the CAR
should be revised and the proposal preamble clarified to ensure
that vent streams considered non-halogenated under a referencing
subpart will be considered non-halogenated under the CAR. The
commenter (VI-D-03) stated that it is possible that a process
vent stream that is halogenated under the part 60 referencing
subparts (III, NNN, RRR) would be considered non-halogenated
under the CAR. The commenter (VI-D-03) asserted that this would
result in a change in which the TRE equation must be used to
determine whether the stream requires control. Furthermore, the
commenter (VI-D-03) added that a benzene loading operation
subject to 40 CFR part 61, subpart BB would have to demonstrate
that it is not halogenated if it opts to use the CAR, subjecting
these sources to unnecessary burden. The commenter (VI-D-03)
reasoned that this will result in some group 2 process vents
moving to group 1 because of the change in the TRE equation. The
commenter (VI-D-03) suggested changes to the language of the CAR
at §§65.2, 65.64(g), and 65.85(c) to provide clarification.
Response: As explained in the preamble, at 63 FR 57772, we
acknowledge that because the CAR adopts the HON halogenated vent
stream determination procedures, the halogenated status of
process vents under 40 CFR part 60 subparts III, NNN, and RRR may
change under the CAR. We maintain that the number of vents
affected will not be significant. To be affected, the process
vent would have to include halogenated components, be subject to
a process vent NSPS, and not be subject to the HON. We believe
that this is a small subset of vents. And for the fraction of
this subset that would experience a change in halogenated status
under the CAR, we remind the owner or operator that compliance
with the CAR is an optional replacement for continuing to comply
with the referencing subparts. In the specific case of 40 CFR
part 61, subpart BB, we maintain that there are very few (if any)
3-35
-------
loading operations that are both subject to subpart BB and also
halogenated. The applicability of subpart BB is such that it
would be nearly impossible for the liquid to be halogenated.
Comment: One commenter (VI-D-03) supports the uniformity
achieved by consolidating the performance test procedures from
the SOCMI NSPS rules. These rules require combustion devices
that do not use supplemental air to correct effluent
concentrations values to a 3 percent oxygen basis. The commenter
acknowledges that sources subject to 40 CFR part 60, subpart DDD
do not have this requirement. The commenter stated that
subpart DDD sources will, therefore, incur increased stringency
if opting to use the CAR, but this would probably affect very few
of those sources.
Response: We thank the commenter for this support, and
agree with the commenter that a consolidated approach to
performance testing is simple and desirable even though it may
increase the stringency of the rules for some.
Comment: One commenter (VI-D-01) indicated that the CAR'S
requirement specifying a correction to 3 percent oxygen for all
combustion device concentration measurements could result in a 30
percent increase in reported concentration values. The commenter
(VI-D-01) stated that this will be problematic since, typically,
a 7 percent oxygen correction is required in current regulations
or in Federally enforceable permits. The commenter (VI-D-01)
noted that the CAR should not be more stringent than the
underlying rules.
Response: The commenter is incorrect that current Federal
regulations affecting the SOCMI use a 7 percent oxygen
correction. All rules consolidated that have an oxygen
correction use a 3 percent oxygen correction. The CAR has not
changed this requirement from what is in the referencing
subparts.
Comment: Two commenters (VI-D-02, VI-D-03) request that the
proposed CAR subpart D be revised to provide a return to a
process or fuel gas system as a compliance option for group 1
3-36
-------
process vents. One commenter (VI-D-03) stated that this will
make subpart D of the CAR consistent with other CAR sections and
will provide incentive for shared return systems to use the CAR.
This commenter (VI-D-03) asserts that such gas streams can be
"process vents" (1) when the gas stream was, for whatever reason,
identified as a process vent when applicability was determined,
(2) if the stream is split, with a portion going to the
atmosphere or to a control device, or (3) if the stream is
sometimes returned to a process or fuel gas system and at other
times sent to a control device.
Response: The CAR, as well as the HON, does not have a
provision specifically allowing process vents to be routed to a
process or fuel gas system as a compliance option. These rules
do not provide this option because it would be confusing or
inconsistent with definitions in some rules.
Route to process. Specifically, the NSPS and HON define
applicability of requirements for vent streams at the equipment
that originates the vent stream and determine the need for
control at the exit of the last recovery device in the process.
Since routing to a process would most likely require either
routing to an existing recovery device in the process line or
adding a recovery device to recycle the desired stream
components, allowing a compliance option of routing to a process
would merely move the point where the need for control is
evaluated. This option already exists in all of these rules; it
is the option to maintain the TRE index value greater than 1.0.
(The typical way to achieve and maintain a TRE greater than 1.0
is to use product recovery to reduce organic compound emissions.)
Based on past experience with these rules, we believe that it
would be confusing to many readers if in addition to maintaining
the TRE index value greater than 1.0 option, we also allowed a
compliance option of routing to process. People would not
understand the distinction between the two cases and therefore,
would be confused as to the requirements of the rule.
3-37
-------
Route to fuel gas system. We believe that adding route to
fuel gas system as a compliance option would also be a source of
confusion. Under the HON, gas streams that are routed to a fuel
gas system are not process vents based on the definitions in
40 CFR part 63, subpart G. Therefore, this option would create
confusion about the classification of the gas stream. Under the
definitions in the three NSPS for SOCMI, the streams would still
be regulated vent streams, but the rule already provides
compatible compliance options for combustion devices. If we were
to add route to fuel gas system as a compliance option for SOCMI
NSPS vent streams, there would be questions as to which
compliance option was being used in such cases and why the rule
retained provisions for use as part of the primary fuel sent to a
boiler or process heater. It would be a problem for people
implementing the rule, because there would be no clear
distinction between the cases.
Comment: One commenter (VT-G-03) pointed out several small
burden reductions resulting from the engineering assessmerit
provisions, and noted that the ability to use engineering
assessments is not useful for units that have been previously
tested. However, the commenter (IV-G-03) recognized that some
relief is also afforded from the extensive recalculation and
reporting if a process change does not affect the group 2B status
of a vent. The commenter (VI-G-03) pointed out that only a
V
statement to that effect is required under the CAR, not a
detailed report.
Response: The commenter is correct. If a source has
already determined the applicability it would not be necessary to
redo the applicability when coming under the CAR. The
engineering assessment provisions provide several burden
reduction opportunities for units not previously tested and that
undergo process changes.
Comment: One commenter (VI-G-03) asserted that the CAR
provides some relief from monitoring process vents regulated
under subparts NNN and RRR. The commenter (VI-G-03) stated that
3-38
-------
no group 2B vents (TRE greater than 4.0 and less than 8.0) are
defined by these rules, therefore monitoring is required by these
referencing subparts.
Response: We assume that the commenter is referring to the
TRE index value criteria for monitoring that was changed from 8.0
in 40 CFR part 60, subparts NNN and RRR to 4.0 in the CAR.
Therefore, in the CAR, process vents subject to subpart NNN or
RRR that have a TRE between 4.0 and 8.0 do not have to monitor
but would have to monitor under subpart NNN or RRR.
Comment: One commenter (IV-G-01) claimed that the last
sentence of paragraph (b)(3) of §65.142 should state that no
other provisions of the subpart apply to Group 2 process vents,
because the paragraph applies to Group 2A and 2B vents, not just
Group 2A vents.
Response -. The provisions of subpart G of the CAR are not
applicable to Group 2B process vents. These vents are not
referenced to subpart G from subpart D. Therefore, it is not
appropriate to change the reference from Group 2A to Group 2.
However, the earlier reference to Group 2 in §65.142(b)(3) should
also be Group 2A, this has been corrected in the final CAR.
Comment: One commenter (IV-G-01) suggested that references
made to TRE index value greater than 1.0 in the CAR §§65.63(c)
and (d), 65.65(b), and 65.66(e) should also say "and less than or
equal to 4.0" to clarify that Group 2A process vents have the
V
same specifications as Group 2A vents defined in §65.62(c).
Response; The reason for defining Group 2 A and Gro.up 2B
process vents in the CAR is because several words can be
eliminated each time a Group 2 vent is referred to. These long
descriptions that follow the term Group 2 in the HON can
sometimes make for long confusing sentences. We consider the
reliance on the terms Group 2A and Group 2B to be a clarification
and simplification of the HON language.
Also, in the paragraphs referenced by the commenter the
language described situations when "maintaining a TRE above 1.0."
It would not be appropriate, even without the Group 2A and 2B
3-39
-------
terms, to add "and less than or equal to 4.0" because we do not
require that the TRE be maintained at all times between 1.0 and
4.0, only that it be maintained above 1.0 at all times.
Comment: One commenter (VI-D-03) suggested that the
provisions of §60.665(1) (6) of subpart NNN and §60.705(1) (5) of
subpart RRR should be revised so that the units opting to comply
with the CAR can follow the provisions of the CAR for determining
stream parameters and conducting performance tests. The
commenter pointed out that the way these provisions are currently
drafted that the source would have to follow the applicable
provisions of subpart NTSTN or RRR for determining stream
parameters and conducting the performance test before going to
the CAR to comply.
Response: The provisions of §§60.665(1) (6) and 60.705(1) (5)
have been edited to allow sources choosing to comply with the CAR
to use the provisions of the CAR to determine stream parameters
and to conduct the performance tests.
3.4 TRANSFER RACKS
Comment: One commenter (IV-G-01) advised that the CAR
should be revised to allow most of the exemptions for transfer
racks that are routed to vapor balance systems to apply to
transfer racks routed to process or fuel gas systems. The
commenter (IV-G-01) reasoned that the piping leading to the
V
process or fuel gas system should not be considered a closed vent
system and noted that the fuel gas systems are excluded .from the
definition of control devices. The commenter (IV-G-01) provided
suggested language for this change.
Response: We agree that this is a warranted clarification
of the intent of these requirements. The CAR has been revised to
allow exemptions for transfer racks routed to processes or fuel
gas systems.
Comment: One commenter (IV-G-01) noted that the term
"transfer rack" is used in §65.80 of the CAR, but it is not
defined in subpart A of the CAR. The commenter requested that
3-40
-------
the term "transfer rack" be defined in subpart A of the CAR, or
else subpart A should reference the transfer rack definition in
HON subpart F and the loading rack definition in the benzene
transfer operations NESHAP subpart BB.
Response: Because the term is not defined in the CAR,
"transfer rack" has the same meaning in the CAR as it does for
HON sources and the same meaning as "loading rack" for'subpart BB
sources. The language proposed to be added to subpart BB in
§61.300 (f) makes it clear that "loading racks" and "transfer
racks" are the same.
Comment: One commenter (VI-G-03) commended the
consolidation of the transfer rack rules from the HON into the
CAR subpart E and noted that the consolidation should help
facilities to demonstrate full compliance. The commenter
(VI-OG-03) also noted that HON group 1 transfer racks are
relieved of an unnecessary recordkeeping burden by using the CAR.
The commenter (VI-G-03) stated that once a transfer rack has been
designated group 1 (requiring controls), records on throughput,
HAP concentration, and partial pressure should not be required.
Response: We thank the commenter for this support.
Comment: One commenter (IV-G-01) requested clarification of
several points from the proposal preamble at 63 FR 57779. The
commenter (IV-G-01) first requested we specify how the CAR
consolidation of the HON storage vessel provisions clarify the
v
HON transfer monitoring provisions. The commenter (IV-G-01)
stated secondly that we clarify that continuous monitoring is not
required for low-throughput transfer racks as opposed to transfer
racks as stated in the preamble.
Response: By consolidating on the HON storage vessel
provisions for low-throughput transfer racks, the monitoring
provisions for storage vessels are extended to low-throughput
transfer racks. Under the CAR, owners or operators of low-
throughput transfer racks can now monitor according to a
monitoring plan they develop instead of monitoring specified by
the rule.
3-41
-------
The commenter is correct that the continuous monitoring
requirements are not required for low-throughput transfer racks
unless continuous monitoring is specifically included in the
monitoring plan.
Comment: One commenter (IV-G-01) noted that the option to
compress and route, regulated material vapors to a process is
given in §65.83(a)(3). This option appears to be the same as the
"routed to a process" option given in §65.83(a)(4). The
commenter (IV-G-03) suggested that the option should be removed
from §65.83(a)(3), since it appears redundant.
Response: The provisions of §65.83(a)(3) specify what
constitutes vapor balancing. In the HON, vapor balancing
included compressing the regulated material and routing it to the
process. The CAR has also included this as a form of vapor
balancing. There are very few provisions that pertain to vapor
balancing. The provisions of §65.83(a)(4) are for routing
emissions to a fuel gas system or process. If a source uses this
option to comply, some provisions of subpart G of the CAR must
also be followed.
Although both of these provisions pertain to routing the
emission to a process, the provisions are different depending on
whether the vent stream is compressed or not. We maintain that
these are two separate activities that require two separate
paragraphs.
Comment: One commenter (IV-G-01) noted that §65.84(a)
requires the owner or operator of a transfer rack to ope.rate the
equipment in the manner specified in paragraphs (a)(1) or (a)(2).
However, the commenter noted, paragraph (a)(1) is a closed vent
system which routes the regulated material vapors to a control
device and paragraph (a)(2) is process piping that routes the
regulated material vapors to a process or fuel gas system or to a
vapor balance system. The commenter (IV-G-01) suggested a
revision to §65.84(a) because the current wording suggests that
one can operate a CVS or process piping.
3-42
-------
Response: We have added to the text of §65.84(a) in order
not to imply that one can operate a CVS or process piping. The
final CAR states:
" An owner or operator of a transfer rack shall
operate it in such a manner that emissions are routed
through the equipment specified in paragraph (a)(1) or
(a)(2) of this section."
3.5 EQUIPMENT LEAKS
3.5.1 Leak Detection
Comment: One commenter (VI-G-03) expressed support for the
streamlined leak detection and repair requirement provided in the
CAR. The commenter agreed with our decision to eliminate the
quality improvement program (QIP) for leaking valves. The
commenter noted that the provisions of subpart F of the CAR that
require increased monitoring frequency for consistently leaking
valves are a sufficient incentive toward quality improvement.
The commenter supported our decision to extend the maximum period
for valve monitoring from 1 year to 2 years, and the maximum
period for connector monitoring from 4 years to 8 years. The
commenter noted that these provisions will eliminate unnecessary
monitoring and give participating facilities an even better
incentive to install and maintain "leak free" components.
However, the commenter also stated that while most facilities
will view extended monitoring periods for valves and connectors
t
as a significant incentive to opt into the CAR from the HON, the
incentive is not expected to be as compelling for smaller
facilities. According to the commenter, extended monitoring
periods are expected to save a facility with 1,000 valves and
connectors only approximately $325 per year.
Response: We thank the commenter for this support and agree
that in general, the CAR offers more benefit to larger, complex
sources.
Comment: One commenter (IV-G-01) noted that §65.104(a)(2)
of the CAR provides a list of cases in which sensory monitoring
for leaks is required. The commenter stated that §65.117(b)(6)
3-43
-------
requires sensory monitoring and suggested that we include that
section in the list of cases in which sensory monitoring is
required.
Response: The CAR contains the procedures for conducting a
pressure test of a batch process in §65.117(b)(6). The
procedures call for the use of a test liquid and the visual
indications of liquids dripping in this circumstance are
occurring under controlled conditions at scheduled times. The
list in §65.104(a) (2) contains references to standards where
sensory monitoring is required to detect infrequent,
unanticipated leaks of regulated material. It is not appropriate
to include §65.117(b) (6) in the list provided by §65.104(a) (2) .
Comment: One commenter (IV-G-01) noted that §65.104(a)(1)
of the CAR provides a list of some of the cases in which
instrument monitoring for leaks is required. The commenter
suggested that we expand §65.104(a)(1) to be a complete list of
cases in which instrument monitoring is required.
Response: We note that the intent of §65.104(a) (1) is'to
provide a comprehensive list of all routine instrument monitoring
requirements. This section does not list the non-routine
instrument monitoring requirements, such as cases in which
instrument monitoring is required only when certain criteria are
triggered or where instrument monitoring is offered as a
compliance option.
\
Comment: One commenter (IV-G-01) requested the removal of
§65.104(e) (2) (i) and (e) (2) (ii), claiming that these two.
paragraphs do not provide a complete list of what needs to be
recorded upon finding a leak. The commenter noted that
§65.105(f) provides a complete list.
Response: There are two separate recordkeeping
requirements. The first is triggered upon detection of a leak.
The information that must be recorded is specified in
§65.104(e)(2), and consists of the instrument and the equipment
operator's name along with the date the leak was detected and the
leaking equipment identification. The records specified in
3-44
-------
§65.105 (f) are the second set of records, and they document the
repair of the leak (not the discovery of the leak). For example,
§65.105 (f) (1) and (f) (2) require that the date of first attempt
at repair and the date of successful repair be recorded. We
maintain that both sets of records are unique and necessary.
Comment: One commenter (IV-G-01) recommended that we use
the term "observed" rather than "monitored" in
§65.104(a) (2) (iii) . The commenter also suggested that the
reference in §65.104(a) (2) (iii) to §65.104(e) (1) (i) should be to
(e) (1) (iv) .
Response: We agree with the commenter and have replaced the
term "monitored" with the term "observed" in §65.104(a) (2) (iii) .
We have also changed the reference in §65.104(a)(2)(iii) to
§65.104(e)(1)(iv).
3.5.2 Leak Repair
Comment: One commenter (IV-G-01) suggested that subpart F
of the CAR should include language that addresses what is and
what is not considered a violation when attempting to repair a
leak, similar to language in §63.162(h) of subpart H (HON
equipment leaks). The commenter (IV-G-01) pointed out that
§63.162(h) of the HON states that if an attempt to repair a leak
is made within the specified time, but the attempt is
unsuccessful, the owner or operator is not in violation of the
HON. The commenter (IV-G-01) expressed concern that successful
V
repair of a leak may require multiple attempts and that
applicable State regulations and permit conditions may p.revent an
owner or operator from designating the leaking piece of equipment
for delay of repair. The commenter (IV-G-01) suggested that
similar language be included in subpart F of the CAR which would
clarify that failing to take action upon discovering a leak is
violation, but that good-faith, unsuccessful attempts at repair
are not violations.
Response: The CAR contains language that clarifies that
leaks, in and of themselves, are not considered violations of the
standard. The standards require action upon detecting leaks,
3-45
-------
such as repair and recordkeeping requirements. Failing to take
the required actions are violations of the standards; detecting a
leak is not a violation of the standards. Therefore, it is not
necessary to add language from the HON to the CAR to clarify this
issue. If it is necessary to delay repair beyond the required
repair time, the source can employ the delay of repair
provisions. A source that neither repairs a leak nor uses the
delay of repair provisions is in violation.
Comment: One commenter (IV-G-01) recommended that we revise
the leak detection sections in subpart F of the CAR to allow
consistent exceptions for different types of equipment. The
commenter noted that in §65.106(b) (the leak detection section
for valves), the owner or operator is required to monitor valves
unless otherwise specified in §§65.102(b), 65.117, 65.118, or
paragraph (e) of this section. The commenter questioned whether
other requirements, such as monitoring for pumps or connectors,
qualified for the same exceptions. For example, the commenter
specifically called out §65.107(b), pump leak detection. This
section does not reference §65.117, so the commenter questioned
whether §65.117 is an allowable alternative (even though §65.117
references §65.107). The commenter suggested that we make the
exceptions to monitoring consistent with the valve exceptions by
revising §§65.107 through 65.114.
Response: We acknowledge that the proposed CAR did not
v
explicitly list the exceptions to every requirement, especially
where the exception itself specified the sections to whi.ch it was
available. To clarify the issue while keeping the language
simple, we have edited the CAR to remove the individual
references to the exceptions while adding language to the
exceptions to more clearly state where they are applicable.
We have clarified in the rule the four "paths" through the
standards of the subpart.
1. Comply with the standards as they are written. This
includes some specific exceptions within a standard
(for example, the exception to pump monitoring for
pumps equipped with dual mechanical seals).
3-46
-------
2. Comply with §65.117 (batch processes) instead of the
regular standards of §§65.106 through 65.114 and
65.116.
3. Comply with §65.118 (enclosed-vented process units)
instead of the regular standards of §§65.106 through
65.116.
4. Comply with any of the above options as modified by an
approved alternative means of emission limitation
pursuant to §65.102(b).
We recognize that there is some confusion that results from
specifically listing the exceptions granted by §§65.116 and
65.117. The exception is explicitly mentioned in some standards,
but not in others. We have edited the CAR to remove all the
individual references to §§65.116 and 65.117 from the standards
in §§65.106 through 65.114. We also edited §§65.116 and 65.117
to clarify that those two sections are each considered to be
alternatives to the entire set of standards set out in §§65.106
through 65.114.
Comment: One commenter (IV-G-01) suggested that the leak
identification removal requirement in §65.105(c)(1) of the CAR
for valves and connectors be revised to apply only to valves and
connectors in gas/vapor or light liquid service. The commenter
(IV-G-01) pointed out that, according to §65.110(b) of the CAR,
no monitoring is required for valves and connectors in heavy
liquid service if a leak detected by auditory, visual, or
olfactory inspection is eliminated within 5 days. v
Response: We agree with the commenter and have revised
§65.105(c)(1) to apply only to valves and connectors in gas/vapor
or light liquid service. It was our intent, however, to have all
leaking equipment identified. Therefore, we revised
§65.110(b)(2) to require that if instrument monitoring identifies
a leak, the equipment must be identified.
Comment: One commenter (IV-G-01) noted that
§65.107 (e) (1) (viii) specifies that "when a leak is detected
pursuant to paragraph (e)(1)(vi) of this section, it shall be
repaired as specified in 65.105(a)." The commenter suggested
3-47
-------
that we change the reference from 65.105(a) to 65.105 to be
consistent with other sections of subpart F and to allow for
delay of repair.
Response: We agree that the reference to §65.105(a) should
be made more general. The reference has been edited to §65.105
to be consistent with the other sections of subpart F of the CAR.
3.5.3 Delay of Repair
Comment: One commenter (IV-G-01) noted that the delay of
repair requirements for valves, connectors, and agitators in
§65.105(d)(3) require purged material to be collected and
destroyed or recovered in a control device complying with
§65.115. The commenter suggested that we add the option of
routing the purged material to a process or fuel gas system.
Response: We agree that adding the option of routing to a
process or fuel gas system would increase operational
flexibility. We have edited the CAR at §65.105(d)(3)(ii) to
incorporate this addition. We note that it will not always be
feasible to route to a fuel' gas system or to a process, but where
it can be done we believe it is a useful option.
Comment: One commenter (VI-D-03) suggested that we revise
§65.105(d) to make the delay of repair language in §65.105(d)
consistent with §65.105(d)(1). The commenter (VI-D-03)
recommended the following language for §65.105(d): "...The owner
or operator shall maintain a record of the facts that explain any
delay of repairs and, where appropriate, why repair within
15 days was technically infeasible without a process shutdown."
Response: We agree with the commenter and have edited the
language in §65.105(d) to be consistent with that at
§65.105(d)(1).
3.5.4 Valves--Difficult- and Unsafe-to-Monitor
Comment: One commenter (IV-G-01) recommended that unsafe-
to-monitor and difficult-to-monitor valves be exempt from the
provisions in §65.106(e) of the CAR which require follow-up
monitoring 3 months after a leak is repaired. The commenter
(IV-G-01) stressed that it would be impractical and costly to
3-48
-------
conduct follow-up monitoring within 3 months for
unsafe-to-monitor and difficult-to-monitor valves. The commenter
(IV-G-01) also pointed out that a similar exemption is provided
in §65.108(e) of the CAR which exempts unsafe-to-monitor
connectors from the requirement for follow-up monitoring 90 days
after repair.
Response: In order to remain consistent with the provisions
of §63.168(h) of the HON, §65.106 (e) (1) of the CAR has been
revised to state that unsafe-to-monitor valves are exempt from
the 3 month follow-up monitoring provisions of §65.106(d)(2) of
the CAR. To remain consistent with §63.168(i) of the HON,
however, §65.106(e)(2) of the CAR has not been revised and
difficult-to-monitor valves continue to be subject to the 3 month
follow-up monitoring provisions of §65.106(d)(2).
Difficult-to-monitor valves have some obstacle to overcome
before they can be monitored, but monitoring does not pose a
safety hazard. The written plan required by §65.103(c) (4) (ii)
for difficult-to-monitor valves specifies annual monitoring at a
minimum. Because personnel are not put at risk and the valves
must be monitored at least annually, it is not appropriate to
exempt difficult-to-monitor valves from the 3 month follow-up
monitoring upon repair of a leak.
Comment: One commenter (IV-G-01) suggested that
difficult-to-monitor valves should be limited to 3 percent at new
or reconstructed sources, but not limited at existing sources.
The commenter (IV-G-01) also suggested that the criteria, for
designating difficult-to-inspect components on a closed vent
system should not include a limitation on the number of
components. The commenter (IV-G-01) proposed revised language
for the CAR, 40 CFR part 61, subpart V and 40 CFR part 60,
subpart W that would reflect the above suggestion and also
create consistency with subpart H of the HON.
Response: We intend to model the difficult-to-monitor
allowance in the CAR and in the proposed revisions to subparts V
and W after the provisions in the HON. To correct drafting
3-49
-------
errors at proposal, we revised the language in §65.103(c)(2) of
the CAR to include a 3 percent limit on the number of valves that
can be designated as difficult-to-monitor at new or reconstructed
sources, and to impose no limit on the number of difficult-to-
monitor valves at existing sources. It is appropriate to limit
new sources because designers, can consider equipment layout when
designing a new source. Also, the proposed amendments to
40 CFR part 61, subpart V were revised at §61.242-11(1)(2) to
remove the 3 percent limit on the number of difficult-to-inspect
components included in closed-vent-systems.
3.5.5 Valves--Subgrouping and Monitoring Frequency
Comment: One commenter (IV-G-01) expressed concern with the
wording in §65.106(c)(1)(i) of the CAR which states that "the
owner or operator shall decide no later than the implementation
date of this part or upon revision of an operating permit whether
to calculate percent leaking valves on a process unit or group of
process units basis." The commenter (IV-G-01) said that the
phrase, "group of process units basis" in §65.106(c)(1)(i)' is
confusing if it is intended to refer to a subgroup of valves
within a process unit. The commenter (IV-G-01) suggested that
the language be revised to read, "...on a process unit or a valve
subgroup basis."
Response: The intent of §§65.106(b) and 65.106(c) is to
provide the owner or operator with maximum flexibility for
»
managing the monitoring of valves. To be eligible for valve
subgrouping provisions, the owner or operator must first.
demonstrate that less than 2 percent of valves are leaking either
within a process unit or within a group of process units. The
decision at this first step is setting up the collection of
valves (either the valves in a process unit or in a group of
process units) that may be eligible for subgrouping. This
collection of valves must perform better than a 2 percent leak
rate before subgrouping of the collection is allowed.
If the owner or operator decides to calculate the percentage
of leaking valves on a process unit basis, and less than
3-50
-------
2 percent of the valves are leaking within that process unit,
then the valve subgrouping provisions of §65.106(b) apply to
valves within the process unit. If the owner or operator decides
to calculate the percentage of leaking valves on a group of
process units basis (more than one process unit), and less than
2 percent of the valves are leaking within that group of process
units, then subgroups of valves may be designated within the
group of process units (both within and across individual process
units) . The owner or operator may decide whether or not to group
several process units together for the purpose of calculating the
overall percentage of leaking valves. Section 65.106(c) (1) (i)
specifies that this decision must be made no later than the
implementation date of the CAR or upon revision of an operating
permit.
Comment: One commenter (IV-G-Q1) requested clarification
regarding the appropriate time to notify the Administrator of a
decision to begin or end subgrouping valves. The commenter noted
that §65.106(b) (4) (v) requires the owner or operator to "notify
the Administrator no later than 30 days prior to the beginning of
the next monitoring period of the decision to begin or end
subgrouping valves." The commenter also noted that according to
the same section, notification may be included in the next
periodic report. The commenter requested clarification on
whether notification can be included in the next periodic report
V
regardless of when the next monitoring period begins.
Response: The notification to begin or end subgrouping of
valves must be submitted at least 30 days prior to the beginning
of the next monitoring period. This notification can be included
with a periodic report; it does not have to be a separate notice.
If you choose to submit the notice as part of a periodic report,
then the periodic report must be submitted at least 30 days prior
to the beginning of the next monitoring period.
Also note that only a single notice is required. If you
submit the notice as part of your periodic report, then a
separate notification is not required.
3-51
-------
We have revised §65.106(b)(4)(v) to clarify these two
options for notifications.
Comment: One commenter (VI-G-03) noted that §§65.106(b)(3)
and 65.108 (b) (3) in subpart F of the CAR determine the monitoring
frequency for valves and connectors, respectively. The commenter'
observed that for each doubling of the monitoring period (for
example, from 1 to 2 years), the percent of leaking components
allowed is cut in half (for example, from 0.5 to 0.25 percent for
valves). The commenter suggested that this pattern creates a
disincentive to strive for longer monitoring periods, noting that
frequent changes in the leak detection and repair program are
burdensome. The commenter questioned how we determined the
percent leaking valves that are used to determine the required
monitoring frequencies.
Response: Regarding the perceived disincentive to
establishing longer monitoring periods, we recognize that there
may be some burden involved with changing a leak detection and
repair program. This burden, however, may be more than offset by
the reduction in monitoring events. Going from annual monitoring
to monitoring once every 2 years effectively cuts the number of
monitoring events in half, with the only increase in burden being
procedural changes in the monitoring program.
Regarding the rationale for determining the percent leaking
valves used to determine the required monitoring frequencies, we
V
note that we maintained the same emission rate. Doubling the
monitoring period while halving the allowable percent le_akers
maintains the effective overall emission rate.
3.5.6 Valves--Other Comments
Comment: One commenter (IV-G-01) recommended that subpart F
of the CAR and subpart H of 40 CFR part 63 both allow the owner
or operator to designate sealless valves as operating with no
detectable emissions, as is allowed in 40 CFR part 60, subpart W
and 40 CFR part 61, subpart V. The commenter (IV-G-01) suggested
that allowing this designation would provide an incentive for
facilities to install sealless valve technology. The commenter
3-52
-------
(IV-G-01) also noted that facilities subject to subpart W and
subpart V would incur an increased monitoring burden under the
CAR if they had previously designated some sealless valves as
operating with no detectable emissions.
Response: The provisions in 40 CFR part 60, subpart W and
40 CFR part 61, subpart V for designating valves as operating
with no detectable emissions require that the owner or operator
monitor these valves annually to verify that these valves
continue to operate with no detectable emissions. The extended
monitoring periods and valve subgrouping provisions of the CAR
allow an owner or operator to monitor valves even less
frequently. We expect that an owner or operator would continue
to have incentive to install advanced valve technology, because
these valves could be designated as part of a subgroup, and could
potentially be monitored as infrequently as once every 2 years if
the technology proves effective. Therefore, we did not find it
necessary to revise subpart H of the HON and subpart F of the CAR
as suggested by the commenter.
Comment: One commenter (IV-G-01) recommended that the VL
term in the percent leaking valves calculation in
§65.106 (c) (1) (ii) should include valves found leaking pursuant to
§65.106(d) (2) (iii) (A) and (d) (2) (iii) (B), as applicable. The
commenter expressed concern that without explicitly including
these valves in §65.106 (c) (1) (ii), an owner or operator may
\
overlook the requirement to include valves found leaking during
the 3 month follow-up monitoring required by §65.106 (d) (.2) .
Response: We agree that the VL term needs to be clarified,
and we have added a reference to §65.106(d) (2) (iii) (A) and
(d) (2) (iii) (B) to the VL term in the percent leaking valves
equation at §65.106(c) (1) (ii) .
Comment: Regarding plant sites with less than 250 valves,
one commenter (IV-G-01) advised that §65.106 (e) (3) should be
revised to read as follows: "Instead, the owner or operator
shall monitor each valve in regulated material service for leaks
once each quarter except as provided in paragraphs (e)(1) and
3-53
-------
(e)(2) of this section." The commenter noted that this revision
will drop the reference to §§65.106 (b) (4) (iii) through (b) (4) (v) ,
which the commenter contends is not needed.
Response: We acknowledge that the provisions for plant
sites with fewer than 250 valves [see §65.106(e)(3)] contained an
incorrect reference at proposal. We have edited the section to
specify that at plant sites with fewer than 250 valves,
monitoring will be required quarterly or at a frequency specified
through the optional subgrouping procedure.
3.5.7 Pumps--Percent Leaking Pumps Calculation
Comment: One commenter (IV-G-01) suggested that, in
§65.107 (c) (4) of the CAR, the definition of the PT term (the
total number of pumps in regulated material service) in the
percent leaking calculation for pumps be changed to explicitly
include pumps routed to a process or fuel gas system or equipped
with a closed vent system routed to a control device. The
commenter (IV-G-01) reasoned that if sealless pumps and dual
mechanical seal pumps can be included in the PT term, then a
plant should be given credit for pumps vented to a closed vent
system routed to a control device or pumps routed to the process
or a fuel gas system.
Response: We agree with the commenter that the P-p term of
the percent leaking equation in §65.107(c)(4) should include
pumps routed to a process or fuel gas system or equipped with a
\
closed vent system. The definition of the PT term in this
section has been clarified to include pumps meeting the .criteria
in §65.107(e)(3), in addition those meeting the criteria in
paragraphs (e)(1) and (e)(2).
Comment: One commenter (IV-G-01) requested clarification on
how unsafe-to-monitor pumps should be accounted for in the
percent leaking pump calculation in §65.107 (c) (4) of the CAR.
The commenter (IV-G-01) inquired whether unsafe-to-monitor pumps
should be excluded from the PL term (the number of pumps found
leaking as determined through monthly monitoring) because they
are not considered part of the "monthly monitoring" of pumps.
3-54
-------
Also, the commenter (IV-G-01) inquired whether unsafe-to-monitor
pumps should be included in the P-p term (the total number of
pumps in regulated material service). The commenter (IV-G-01)
reasoned that if the PL term excludes unsafe-to-monitor pumps
then the PT term should exclude them as well. Similarly, the
commenter (IV-G-01) noted that if the PL term includes
unsafe-to-monitor pumps, then they should be included in the PT
term.
Response: The PL term in §65.107(c)(4) of the CAR is
defined as the number of pumps found leaking as determined
through monthly monitoring as required in §65.107(b)(1). We
agree with the commenter that the PL term does not include
unsafe-to-monitor pumps because they are not included in the
monthly monitoring required by §65.107(b)(1). Note that
§65.107 (e) (6) specifically excludes these pumps from monthly
monitoring pursuant to §65.107 (b) . The P-p term is defined as the
total number of pumps in regulated material service, and
therefore includes unsafe-to-monitor pumps in regulated material
service.
Comment: One commenter (VI-D-03) requested clarification of
which pumps are included in the divisor of the calculation that
determines the rate of leaking pumps. The commenter (VI-D-03)
recommended that the divisor of this calculation should be the
SCU for any designated SCU's. The commenter (VI-D-03)
V
recommended that the divisor should be based on process units
only for equipment that is not part of a SCU. The commenter
(VI-D-03) provided that the term "process unit" in the proposed
rule is confusing in the issue of percent leaking pumps
calculation. The commenter pointed out that the definition of
"process unit" in the proposed CAR refers to the referencing
subpart definition, but several referencing subparts are
frequently involved in a single SCU. The commenter stated that
it is also confusing how to handle the unit grouping decision as
equipment comes under the CAR.
3-55
-------
The commenter (VI-D-03) recommended that the choice of a
SCU, or group of SCU's, as the basis for the calculation should
be allowed when the first SCU comes under the CAR, regardless of
the choice of divisor under the HON or for units continuing to
comply with the HON. The commenter stated that such a choice
would be binding for future equipment coming into the CAR. The
commenter (VI-D-03) asserted that a "new election" should be
allowed for HON units, because the number of pumps in an SCU will
be different than the number of pumps in a HON unit (because only
HAP-containing pumps are covered by the HON). The commenter
(VI-D-03) provided language to make this change in the
requirements at §65.107(c).
Response: The final CAR does not include the SCU concept.
Therefore, the process unit referred to by the divisor of the
equation is clearly referring to the process unit that has opted
to comply with the CAR.
With regards to "new election" for HON units, we have
clarified that when a facility opts into the CAR, the decision to
base the calculation on a process unit or group of process units
can be made. Because the facility's permit will be open for
modification to opt into the CAR, it is also an opportunity to
re-evaluate and modify the selection of process units used as the
basis of the calculation for the process units remaining under
the HON.
3.5.8 Pumps--Visual Inspections
Comment: One commenter (IV-G-01) suggested that we.revise
§65.107(e) (1) (viii) to state that a leak detected pursuant to
paragraph (e)(1)(v) or (e)(1)(vi) must be repaired as specified
in §65.105. The commenter cautioned that if paragraph (e) (1) (v)
were not included, then there would be no repair time limit for
leaks detected under that paragraph.
Response: Under the CAR at §65.107(e) (1) (v), weekly visual
inspections are required for dual mechanical seal pumps. If
there are visual indications of liquids dripping, then the owner
or operator has a choice to instrument monitor the pump or take
3-56
-------
action to eliminate the drip. At this point, a leak has not been
detected. A leak is detected only if the owner or operator
chooses to instrument monitor and gets a reading in excess of
1,000 ppm. If the owner or operator chooses to take action to
eliminate the drip, then no leak is detected.
The distinction is important because, as the commenter
pointed out, there is no time limit in the proposed CAR for
owners or operators to take action to eliminate the drip. There
is also no time limit in the proposed CAR for performing the
instrument monitoring if the owner or operator chooses that
option. But there is a time limit (15 days) to repair a leak
after its presence has been confirmed through instrument
monitoring.
To remedy this inconsistency, we edited the language in
§65.107 (e) (1) (v) to specify that one of the two procedures
(perform instrument monitoring or eliminate the drip) must be
completed "prior to the next required inspection." Similar
language was also added to parallel requirements at §65.109(b)(3)
(agitator seals), §65.109 (e) (1) (iv) (dual mechanical agitator
seals), and §65.110(b)(1) (heavy liquid service equipment;
pressure relief devices in liquid service; and instrumentation
systems). Note that §65.110 does not specify required visual
inspection frequency. In this section, there is a 15 day time
limit to perform instrument monitoring or to take action to
V
eliminate the indications of a leak.
Comment: One commenter (VI-D-02) suggested that we. revise
the proposed language for §§65.107(b)(4) and 65.107(e)(1)(v) of
the CAR to eliminate a potential requirement for weekly
monitoring of pumps. The commenter pointed out that pumps must
be monitored monthly by EPA Method 21, and must also be monitored
weekly by visual inspection. If weekly visual inspection
identifies "liquids dripping," the owner or operator is required
either to monitor the pump by Method 21, or to eliminate the
visual indications of liquids dripping. The commenter noted that
if the owner or operator chooses to monitor the pump by
3-57
-------
Method 21, and no leak is found, then nothing is required.
However, the commenter pointed out that visual indication of
liquids dripping is likely to still exist each week and therefore
require monitoring by Method 21 each week. The commenter pointed
out that a small amount of liquid drip from a pump is not
unusual, particularly where a heavy barrier fluid is present.
The commenter suggested that once an owner or operator has shown
that a drip from a pump is not a leak, subsequent weekly
monitoring should not be required. The commenter recommended
that additional paragraphs be added to §§65.107(b)(4) and
65.107(e)(1)(v) to clarify that additional instrument monitoring
of a pump due to visual indication of liquids dripping is only
required once between routine monthly monitoring.
Response: We thank the commenter for pointing out this
potential problem, and we have edited the language in the pumps
standards at §§65.107(b) (4) and 65 .107(e) (1) (v) . The revised
standards specify that if weekly visual inspections indicate
"liquids dripping" and if instrument monitoring shows that there
is not a leak, then no additional instrument monitoring is
required until the next regularly scheduled (monthly) instrument
monitoring.
3.5.9 Pumps--Other Comments
Comment: One commenter (IV-G-01) suggested that language in
§65.107 (e) (5) of the CAR be revised to state that if more than
V
90 percent of the pumps at a process unit are equipped with dual
mechanical seals or have no externally actuated shaft {i..e.
sealless pumps), then the process unit should be exempt from the
percent leaking calculation in §65.107(c) of the CAR rather than
being exempt from the leak detection requirements in §65.107(b)
of the CAR. The commenter (IV-G-01) stated that this revision
would make the CAR consistent with §63.163 (i) of the HON.
Response: We agree with the commenter, and we have revised
the language in §65.107(e)(5) of the CAR to state that if more
than 90 percent of the pumps at a process unit are either
equipped with dual mechanical seals or have no externally
3-58
-------
actuated shaft then, the process unit is.exempt from the percent
leaking calculation in §65.107 (c) and not exempt from the leak
detection requirements in §65.107(b).
Comment: One commenter (IV-G-01) requested clarification on
portions of the QIP requirements for pumps in §65.116 of the CAR.
It is the understanding of the commenter (IV-G-01) that the data
analysis of pumps in the QIP (specified by §65.116(d)(5) of the
CAR to be completed within 18 months of beginning the QIP) is not
required if the facility meets the criteria to exit the QIP in
less than 18 months. The commenter (IV-G-01) requested that this
point be clarified. The commenter (IV-G-01) also requested
clarification on whether a facility in the QIP for the first time
would be required to comply with the requirements of the trial
evaluation in §65.116(d)(6) of the CAR if a data analysis has
already identified a superior pump design, technology or
operating and maintenance practice. The commenter (IV-G-01) also
pointed out that the quality assurance program in §65.116(d)(7)
and the pump replacement program in §65.116(d)(8) require that a
facility implement these programs after having been in the QIP
for 3 or 4 years, depending on the number of employees and number
of pumps at the facility. The commenter (IV-G-01) requested
clarification on how to determine the length of time a facility
has been in the QIP program if the facility has exited and
reentered the QIP program one or more times.
V
Response: In response to clarifying what to do for a
facility that exits a QIP in less than 18 months, we agree with
the commenter that the first data analysis would not be required.
In response to the issue of facilities implementing a QIP
for the first time, we agree that a trial evaluation program
would not be necessary if a data analysis specific to the
individual situation at the facility had previously been
conducted. This pre-existing data analysis would have already
identified the services, operating or maintenance practices, and
pump or pump seal design technologies with better than average
emission performance. The requirement under the QIP would then
3-59
-------
be to begin implementation of the superior technology through the
replacement program.
In response to the question regarding the time period
requirements of the QIP, the 3 and 4 year requirements would
refer to the time passed since the first triggering of the QIP.
The QIP was developed for poorly performing facilities and was
not envisioned as an additional burden to facilities operating on
the edge of triggering a QIP. We recognize that, in the absence
of data identifying a superior technology, a facility entering
and exiting a QIP must re-enter the QIP at the performance trial
step.
We note that the intent of the CAR is to create an incentive
to improve performance such that the QIP is not triggered. We do
not anticipate many facilities needing to comply with the QIP,
and we expect very few sources to be operating "on the bubble,"
constantly entering and exiting the QIP.
Comment: Two commenters (IV-G-01 and VI-D-03) recommended
changing the leak definition for heavy liquid pumps that are not
in polymerizing monomer or food/medical service from 1,000 ppm to
2,000 ppm in §65.110. The commenters (IV-G-01 and VI-D-03)
pointed out that, although the leak definition for heavy liquid
pumps is 1,000 ppm, they are not required to be repaired unless
they are detected to be leaking at or above 2,000 ppm. One
commenter (IV-G-01) also stated that a 2,000 ppm leak definition
i
for heavy liquid pumps would be consistent with the requirements
of 40 CFR part 63, subpart H (HON equipment leaks).
Response: We agree with the commenter; for all heavy liquid
pumps that are not in polymerizing monomer service, we intended
§63.169 of the HON and §65.110 of the CAR to state that an
instrument reading of 2,000 parts per million indicates a leak.
We have revised the portions of §65.110(b) (2) of the CAR and
§63.169(b) of the HON relating to pumps in heavy liquid service
to read: "If an instrument reading of ... 5,000 parts per million
or greater for pumps handling polymerized monomers, 2,000 parts
per million or greater for all other pumps (including pumps in
3-60
-------
food/medical service) ... is measured, ... a leak is
detected...."
3.5.10 Connectors--Exemptions to the Connector Standards
Comment: Two commenters (IV-G-01 and VI-D-03) suggested
that owners or operators electing to use the CAR should not be
required to monitor connectors that are subject only to the
provisions of 40 CFR' part 61, subpart V or 40 CFR part 60,
subpart W. Both commenters noted that these referencing
subparts require only sensory inspection of connectors. The
commenters (IV-G-01 and VI-D-03) stated that instrument
monitoring of connectors in gas/vapor or light liquid service
would represent a substantial burden increase over sensory
inspection and may be a disincentive for owners or operators to
opt into the CAR. One commenter (IV-G-01) provided an example
from the CAR preamble which states that agitators subject only to
the requirements of subpart W would not be subject to the
provisions for agitators in the CAR because there are no
provisions that apply to agitators in subpart W.
One commenter (VI-D-03) suggested that an overall
environmental benefit would be achieved if facilities currently
using sensory inspection for connectors were exempt from
monitoring using EPA Method 21. The benefit, according to the
commenter (VI-D-03), would be that facilities subject to
subpart W of part 60 and subpart V of part 61 would be more
likely to opt into the CAR and would therefore be subject to a
more stringent leak detection and repair program for valves. The
commenter (VI-D-03) stated that leak definition for valves is
10,000 ppm in subparts V and W compared to 500 ppm in subpart F
of the CAR. The commenter (VI-D-03) pointed out that although
sensory inspection of connectors instead of instrument monitoring
appears to be a relaxation of the proposed requirements of the
CAR, sensory inspection is not a change from existing
requirements. The commenter (VI-D-03) also alleged that
historical and recent connector emission data indicate that
3-61
-------
little benefit would be gained from instrument monitoring versus
sensory inspection of connectors.
Response: We have considered the commenters request. We
have determined that at a facility currently performing only
sensory monitoring for connectors, initiating instrumental
monitoring may present a significant disincentive to using the
CAR. Because we believe that the more facilities that use the
CAR, the more burden that will be reduced for both industry and
regulators, we have provided a sensory monitoring option for
sources subject only to subparts V and W to eliminate this
disincentive for opting into the CAR. No degradation of
environmental protection will result from the CAR requiring
sensory monitoring for connectors coming into the CAR from
subparts V and W because that is what those two referencing
subparts require. The final CAR has been modified so that it
contains two connector monitoring programs. The first will
consist of sensory monitoring and will be available as an
alternative to connector monitoring for equipment referenced to
the CAR from subparts V and W. The second will consist of the
instrument monitoring procedures as proposed; this program will
be applicable to equipment coming to the CAR from the HON and
will be available to equipment coming to the CAR from subparts V
and W.
Note, however, that when sensory monitoring indicates a
potential leak and the owner or operator performs instrument
monitoring, a leak definition of 500 ppm is used for consistency
with the other connector monitoring provisions in the CAR.
Comment: One commenter (IV-G-01) noted that although
§65.104(a)(2)(ii) states that "inaccessible, ceramic, or ceramic-
lined connectors... shall be observed pursuant to 65.108(e) (2),"
there is no actual requirement to observe connectors in
§65.108(e) (2). The commenter also noted that although
§65.104(a)(2)(iv) requires several types of equipment to be
observed pursuant to §65.110(b)(1), there is no actual
requirement to observe this equipment in §65.110(b) (1) . The
3-62
-------
commenter recommended that we remove §§65.104(a) (2) (ii) and
65.104(a)(2)(iv).
Response: We concur with the commenter and have deleted
§§65.104 (a) (2) (ii) and (a)(2)(iv). We clarify that
§§65.108 (e) (2) and 65.110(b) (1) do not require the owner or
operator to perform regularly scheduled inspections or
observations. They do require that action be taken, however, if
evidence of a leak is observed.
Comment: One commenter (IV-G-01) suggested that we revise
§65.108(e)(1) to make unsafe-to-monitor connectors exempt from
the connectors compliance schedule in §65.108(a) and leak
detection requirements in §65.108(b). The commenter pointed out
that unsafe-to-monitor connectors should not be included in the
initial screening required by §65.108(a) because they may not be
safe to monitor at any time during the first 12 months.
Response: We agree with the commenter and have made edits
to the CAR at §65.108 (e) to specify that unsafe-to-monitor
connectors are neither included in the initial screening required
by §65.108(a) nor in the leak detection provisions of §65.108(b).
This does not relieve you from having to inspect during safe-to-
monitor periods.
Comment: One commenter (IV-G-01) recommended that
inaccessible connectors mentioned in §65.108(e)(2)(i) should be
exempt from the leak repair requirements in paragraph (d) because
V
§65.108 (e) (2) (ii) specifies the timing for repair for
inaccessible connectors.
Response: We clarify that §65.108(e) does not reference
§65.108(d) for leak repair. In §65.108(e) (2) (ii), the
requirement is to eliminate the "visual, audible, olfactory, or
other indication of a leak...as soon as practical." The standard
leak detection procedures in §65.105 do not apply to these types
of connectors (inaccessible, ceramic, or ceramic-lined), so no
reference is provided to §65.108(d) because §65.108(d) specifies
the use of §65.105. Also, note that §65.108(d) explicitly states
3-63
-------
that §65.105 is to be used only for leaks "detected pursuant to
paragraphs (a) and (b)."
3.5.11 Connectors--Other Comments
Comment: One commenter (IV-G-01) requested that the
definition of the %CL term in the percent leaking connectors
calculation in §65.108 (c) of the CAR be revised to clarify that
connectors found leaking during the 90 day follow-up monitoring
are excluded from the calculation. The commenter (IV-G-01)
pointed out that including connectors found leaking during the
90 day follow-up monitoring would lead to double counting of
leaking connectors. The commenter (IV-G-01) recommended revising
the definition of %CL to read, "%CL = Percent leaking connectors
as determined through monitoring required through periodic
monitoring required in paragraphs (a) and (b)(3)(i) through
(b) (3) (iii) ."
Response: We agree with the commenter that including
connectors found leaking during the 90 day follow-up monitoring
in the %CL term of the percent leaking connector calculation
would lead to double counting of leaking connectors. We have
revised the language in §65.108(c), as suggested by the
commenter, in order to clarify that connectors found leaking
during the 90 day follow-up monitoring are not included in the
percent leaking connector calculation.
Comment: One commenter (VI-G-05) suggested that we include
\
a subgrouping program for connectors. The commenter (VI-G-05)
stated that it is often practical for monitoring personnel to
monitor valves and their associated connectors at the same time
because the monitoring route followed is typically ordered by
location rather than by equipment type. The commenter (VI-G-05)
reasoned that it would make sense to allow similar subgrouping
programs for both valves and connectors. The commenter (VI-G-05)
also suggested that the process stream contained within the
equipment plays a role in causing leaks. The commenter (VI-G-05)
reasoned that if a certain area or subgroup is experiencing high
3-64
-------
leak rate problems for valves, it is likely that the connectors
in that area are experiencing similar problems.
Another commenter (VI-D-03) suggested that a subgrouping
program for connectors in the CAR was not justified. The
commenter (VI-D-03) stated that the increased complexity of a
connector subgrouping program for owners, operators, and agency
inspectors is not justified for the minimal environmental
benefit. The commenter (VI-D-03) provided several reasons for
not creating a subgrouping program for connectors. One reason
the commenter gave was that the leak frequency for connectors has
not been shown to be a function of the process, fluid temperature
or operating pressure. A second reason the commenter gave is
that 40 CFR part 63, subpart H and the proposed CAR provide for
connectors in gas/vapor service to be monitored once per year and
allows longer monitoring periods for process units with connector
leak rates less than 0.5 percent. A third reason the commenter
gave was that historical industry data indicate that, in general,
emissions from connectors is already extremely low and a recent
study showed insignificant differences in connector leakage
emissions at facilities with and without leak detection and
repair programs.
Response: We do not believe it is appropriate to add a
subgrouping program for connectors, given the added complexity of
such a program for minimal environmental benefit or burden
i
reduction.
3.5.12 Monitoring Instrument Procedures
Comment: One commenter (IV-G-01) recommended that water be
given as an example of an inert in the phrase "For process
streams that contain nitrogen, air, or other inerts that are not
organic HAPs or VOC ..." found in §65.104(b)(2) and
§65.143 (c) (1) (ii) .
Response: We agree with the commenter and have added
"water" to the lists of example inert compounds in §§65.104(b)(2)
and §65.143 (c) (1) (ii) .
3-65
-------
Comment: One commenter (VI-D-03) recognized that in
§§65.104(b)(2) and 65.143(c)(1) of the CAR, we have reduced the
burden of determining an instrument response factor compared to
the requirements of subpart H of the HON. The commenter
(VI-D-03) requested that we clarify what is required and how to
demonstrate that the requirements of §§65.143 (c) (1)
and 65.104(b)(2) have been met. The commenter (VI-D-03) noted
that the HON requires that an instrument response factor be based
on the mathematical average response factor for a given process
fluid. The commenter (VI-D-03) also noted that we recognized the
difficulty in calculating individual stream response factors,
particularly for complex streams and that we specified in the CAR
that response factors could be based on a representative response
factor. The commenter (VI-D-03) agreed with the simplification
but expressed concern that the proposed language regarding the
requirements and demonstrations of compliance were unclear. The
commenter (VI-D-03) provided suggested revisions to the language
in §§65.143 (c) (1) (ii) and 65'.104 (b) (2) (i) and also provided
language for an additional recordkeeping requirement to be added
to §§65.119(b) and 65.163(a).
Response: We have adopted some of the changes requested by
the commenter in §§65.104(b)(2) and 65.143(c)(1), and we have
added the explicit records suggested by the commenter to
§§65.119(b) and 65.163(a). Both changes help clarify the rule by
spelling out exactly what the procedures are and what records
must be kept when modifying the instrument response factor under
Method 21 in cases where a representative composition of the
process fluid is used.
Comment: Two commenters (VI-D-03 and VI-D-02) suggested
that we revise §§65.104(b)(4) and 65.143(c)(1)(v) to allow gases
other than methane to be used as a Method 21 calibration gas.
Both commenters expressed concern that an owner or operator of a
source subject to subpart W and using hexane as a calibration
gas would face the unnecessary burden of switching to methane in
order to opt into the CAR. One commenter (VI-D-02) suggested
3-66
-------
that there are many materials, in addition to methane and hexane,
that have response factors in the proper range to assure good
measurements. The commenter (VT-D-02) requested that we allow
non-methane calibration gases, particularly hexane, as long as
the instrument performance criteria are met.
Response: We clarify that hexane is an allowable substitute
for methane as a calibration gas in cases where methane cannot be
used because the monitoring instrument does not respond to
methane. This allowance is spelled out at §65.104(b) (4) (ii) , "A
calibration gas other than methane in air may be used if the
instrument does not respond to methane..." This approach is
consistent with the HON, provides for a single consolidated
calibration procedure, and still allows n-hexane to be used in
cases where methane is not able to be used.
Comment: One commenter (IV-G-01) suggested that, in
subpart F of the CAR, we revise all of the references to
monitoring methods. The commenter provided a list of paragraphs
from subpart F of the CAR that require the owner or operator to
detect leaks by the method specified in §§65.104(b), (c), and
(e). The commenter pointed out that §65.104(b) contains the
monitoring method information, §65.104(c) contains information
about adjusting the instrument readings, and §65.104(e) is about
identification and records of leaking equipment. The commenter
recommended that we change the references to either just
§§65.104(b) and (c), or to just §65.104(b).
Response: We agree that referencing §65.104(e) is not
necessary, so long as §65.104(b) and (c) are referenced. We have
edited the language in subpart F of the CAR accordingly.
3.5.13 Compressors
Comment: One commenter (VI-D-03) expressed concern that the
language in §65.112 (f) of the CAR differs from the HON in the
criteria for designating a compressor as operating with "no
detectable emission." The commenter explained that the last
sentence of §65.112(f) is not in the HON and implies that if a
compressor has ever had an instrument reading above 500 parts per
3-67
-------
million (ppm), then it cannot be designated as having no
detectable emissions. The commenter (VI-D-03) recommended that
the last sentence of §65.112 (f) be removed. The commenter
cautioned that the current language in this section eliminates
any incentive for an owner or operator to upgrade a compressor
seal that was expected to have no detectable emissions, but has
had a problem. The commenter stated that an upgraded compressor
seal would not allow the owner or operator to redesignate the
compressor as having no detectable emissions if an instrument
reading above 500 ppm has been detected at any time. The
commenter pointed out that according to the current language of
the HON, an instrument reading above 500 ppm for a compressor
seal designated as having no detectable emissions constitutes a
violation of the standard. The commenter (VI-D-03) stated that
this is adequate incentive for an owner or operator to avoid
designating a compressor seal as having no detectable emissions
if the seal is likely to have a problem.
Response: To clarify the intent of the provisions, we•have
revised §65.112(f)(1) by removing the last sentence. This change
was suggested by the commenter, and we agree that the language is
clearer without this sentence.
The proposed language implied that once a compressor has an
instrument reading greater than 500 ppm, it can no longer qualify
for the alternative compressor standard. This was not our
intent. The intent of the CAR is that if an instrument reading
of greater than 500 ppm is observed, the standard has been
violated. The newly revised language is consistent with similar
language in the HON and in the pressure relief device standards
in the CAR.
3.5.14 Sampling Connection Systems
Comment: One commenter (IV-G-01) pointed out that
§65.113(c) of the CAR requires compliance with §§65.113(c) (1)
through 65.113(c)(5), but that §65.113(c)(2), which requires the
owner or operator to collect and recycle purged process fluid to
a process, is a subset of §65.113(c)(1), which requires that a
3-68
-------
closed purge, closed loop, or closed vent system return the
purged process fluid directly to a process line or to a fuel gas
system. The commenter (IV-G-01) recommended that §65.113 (c) (2)
be deleted, and that the paragraphs in §65.113(c) be renumbered
accordingly. The commenter (IV-G-01) also suggested that
§65.113 (c) (1) should specify that if the purged process fluid is
returned to a fuel gas system or process, it should meet the
requirements of §65.115, which specifies the standards for
emissions routed to a fuel gas system or process. The commenter
(IV-G-01) pointed out that this would be consistent with other
sections in subpart F of the CAR which reference §65.115 when
allowing the owner or operator to route emissions to a fuel gas
system or back to the process.
The commenter (IV-G-01) then suggested that language in
40 CFR part 63, subpart H, 40 CFR part 61, subpart V and
40 CFR part 60, subpart W be revised to be consistent with the
above comments relating to sampling connection systems.
Response: We maintain that §65.113(c) (2) is not a subset of
§65.113(c) (1) , and therefore did not remove §65.113 (c) (2) .
Separate provisions are necessary to clarify that a sampling
connection system may consist of a direct connection such as a
closed purge, closed loop or closed vent system as described in
paragraph (c)(1). Alternatively, a sampling connection system
may consist of plant personnel using a bucket or drum to manually
return a purged sample to the process as allowed by
paragraph (c) (2) . Also in §65.113 (c) (1) , we did not add. a
reference to comply with the closed vent system and control
device requirements of §65.115 because we determined that the
associated compliance and recordkeeping burden was not necessary
for sampling connections.
3.5.15 Enclosed-Vented Process Units
Comment: One commenter (IV-G-01) recommended that the
enclosed-vented process unit provisions in §65.118 of the CAR be
revised to allow for either a process unit or a portion of a
process unit to be designated as an enclosed-vented process. The
3-69
-------
commenter (IV-G-01) also requested that enclosed-vented process
units be given the option to be routed to a fuel gas system or to
a process. In addition, the commenter (IV-G-01) suggested that
enclosed-vented process units should not be exempt from the
closed vent system requirements of §65.115 of the CAR.
Response: We note that the enclosed-vented process unit
alternative is intended for process units entirely contained
within large buildings, where all emissions will vent through a
limited number of exhaust ports. Many of these process units are
unmanned. Pharmaceutical process units are typical examples of
this type of operating scenario.
We maintain that it is inappropriate to allow the
enclosed-vented alternative for portions-of process units. Doing
so creates confusing compliance situations and stretches the
scope of the allowance beyond what was originally intended.
We also note that it is not appropriate to allow these units
to vent to a process or fuel gas system. Vents coming off an
enclosed-vented process unit are typically very dilute with
negligible heating values.
3.5.16 Batch Product--Processes
Comment: One commenter (IV-G-01) requested that we clarify
the requirements of §65.117(b) regarding pressure testing and
retesting of batch product-processes. The commenter noted that
§65.117(b)(4)(i) requires a batch product-process to be retested
if a leak has been detected and that §65.117(b)(4)(ii) states
that "if a batch product-process fails the retest or the. second
of two consecutive pressure tests," it must be repaired within
30 days of the second test. The commenter requested
clarification on whether or not §65.117 allows for two
consecutive pressure tests, whether the retest is considered the
second consecutive test, and whether repair is required within
30 days of the retest if the first test failed.
Response: The intent of the requirements in §65.117(b) is
as follows. The retest is the second of two consecutive pressure
tests. The "or" in the rule has been replaced by parenthesis to
3-70
-------
help make the issue clearer. The rule now reads, "...if a batch
product-process fails the retest (the second of two consecutive
pressure tests)..."
This should help clarify that you have 30 days after failing
the retest to repair the leak. This is the same as specifying
30 days after failing the second of two consecutive pressure
tests.
Comment: One commenter (IV-G-01) asked if a pressure test
for a batch process must continue indefinitely if no pressure
loss or gain equal to a rate of 1 pound per square inch gauge
(psig) per hour is seen when using a pressure measurement device
of ±10 percent. The commenter (IV-G-01) noted that, according to
§65.117(b)(5)(iv), if a more accurate measurement device is not
available, and an owner or operator elects to use a pressure
measurement device with a precision of at least 10 percent, then
the duration for the test must extend for the time necessary to
detect a pressure loss or rise that equals a rate of 1 psig per
hour. The commenter questioned if the test is required to
continue indefinitely if a pressure loss or rise does not equal
1 psig per hour.
Response: To clarify how long the test must be extended,
consider the following example. A process operating at 200 psig
is tested, and you elect to use a pressure measurement device
with a precision of 20 psig (± 10 percent of the test pressure).
V
Such a device would not be able to detect a pressure drop of
1 psig/hour in 1 hour because it could only detect a change of
+20 psig. The test must be extended to 20 hours. After
20 hours, if the process is losing pressure at a rate greater
than 1 psig/hour, then the instrument will be able to detect the
change because the change would be greater than the precision of
the device (+ 20 psig).
3.5.17 Periodic Report Contents
Comment: One commenter (IV-G-01) recommended that the
requirement to report nonrepairable components as part of the
periodic reporting requirements of §65.120(b)(1) of the CAR be
3-71
-------
removed. This paragraph requires the owner or operator to
"include the number of leaking components that were not repaired
as required by §65.105(a), and for valves and connectors identify
the number of components that are determined by §65.106(c) (3) to
be nonrepairable." The commenter (IV-G-01) pointed out that this
requirement is redundant with the requirement in §65.120(b)(2) to
report occurrences of delay of repair. The commenter (IV-G-01)
also pointed out that §65.106(c)(3) is referenced in the valve
section, but there is no parallel reference for nonrepairable
connectors in the connector section. The commenter (IV-G-01)
requested clarification on reporting the number of leaking
components that were not repaired as required by §65.105(a) of
the CAR. The commenter (IV-G-01) asked if the intent was for the
owner or operator to report the number of components which missed
either the 5 day first attempt and/or the 15 day final repair for
reasons other than delay of repair.
Response: With regard to the apparent redundancy in
periodic reporting requirements, §65.120(b)(1) requires reporting
of the number of leaking components that were not repaired. This
number refers to the components not repaired within the 15 day
time period. It does not include the number of components that
are not repaired pursuant to the requirement to perform a first
attempt at repair within 5 days.
In addition, this number may not be the same number as the
instances of delay of repair, which is required to be reported
under §65.120 (b) (2) . For example, one component may lea_k
multiple times during a reporting period. This may necessitate
more than one instance of delay of repair, but the number of
components that leaked is still one.
With regard to reporting the number of nonrepairable
connectors, we recognize that the CAR does not provide for
designating connectors as nonrepairable. To correct this
oversight, we edited §65.120(b)(1) so that the section does not
refer to connectors.
3.5.18 Alternatives and Exemptions
3-72
-------
Comment: One commenter (IV-G-01) suggested edits to
§65.103 (c) (3) so that it is clear that the planned schedule for
monitoring required by the paragraph includes an explanation of
why the equipment is unsafe-to-monitor or difficult-to-monitor.
The commenter pointed out that the proposed language only
addresses explanations of why the equipment is difficult-to-
monitor.
Response: We agree with the commenter and have revised
§65.103(c) (3) to require explanation not only for difficult-to-
monitor equipment but also for unsafe-to-monitor equipment.
Comment: One commenter (IV-G-01) suggested that whenever a
reference is made to routing emissions to a process or fuel gas
system or routing emissions to a closed vent system and control
device meeting the requirements of §65.115, we should also
provide an alternative reference to §65.102(b) (request for
alternative means of emission limitation). The commenter noted
that §§65.Ill(d) and 65.118(a) both refer to a control device
meeting the requirements of either §65.115 or §65.102(b). The
commenter suggested that we use §§65.Ill(d) and 65.118(a) as
examples for how other sections should be revised.
Response: The original text was not clear, as it explicitly
referenced alternative means of emission limitation in only two
places. We have clarified the rule by removing the references to
§65.102(b) from §§65.Ill(d) and 65.118(a). The intent of the CAR
V
is to provide for alternative means of emission limitation for
all control devices. Also, just referencing §65.115 for. control
device requirements is sufficient, because §65.115(b) references
§65.102(b) where applicable.
3.5.19 Other Equipment Leak Comments
Comment: One commenter (IV-G-01) suggested that the
language in §65.103(b)(5) regarding the requirement to identify
"instrumentation systems subject to the provisions of this
subpart" would be clearer if it referenced §65.110 (the
instrumentation system standards) instead of "this subpart."
3-73
-------
Response: While the instrumentation systems standard is
contained within §65.110, many other sections of subpart F of the
CAR are also potentially applicable. For example, instrument
monitoring provisions of §65.104, and alternative means of
emission limitation requested under §65.102(b) are all sections
within subpart F of the CAR that may apply to instrumentation
systems. The intent of §65.103(b)(5) is to identify all
instrumentation systems subject to subpart F of the CAR {"this
subpart") and not just those complying with §65.110.
Comment: One commenter (VI-G-05) suggested that we include
language in the CAR to clarify that a performance test is not
required for control devices used only to control emissions from
equipment leaks. The commenter (VI-G-05) recommended that, aside
from annual visual inspection, the only requirement for such
control devices should be "operation of the control device at all
times when emissions are vented to them." The commenter
(VI-G-05) suggested that startup, shutdown, and malfunction plan
requirements would be unnecessarily burdensome for these minor
control devices which are infrequently used.
Response: In §65.146 (nonflare control devices used for
equipment leaks only), paragraph (b) specifically states that, "A
performance test is not required for any control device used only
to control emissions from equipment leaks." Regarding SSM plan
requirements, we maintain that the SSM plan applies to all
"equipment equipped with a closed vent system and control device
subject to subpart G" of the CAR (see §65.6). Excepting,certain
control equipment from the S"SM plan would not only create
confusion but also potentially result in increased emissions.
Comment: One commenter (IV-G-01) suggested we include
visual inspection records for agitators in §65.119(c)(4). The
commenter reasoned that if weekly inspection for pumps must be
documented, then weekly inspection for agitators should be
documented as well.
Response: We agree that it is reasonable and consistent to
require documentation of the weekly inspection for agitators.
3-74
-------
We added this record to §§65.109 and 65.119.
Comment: One commenter (VI-D-02) suggested that the current
requirements for open ended lines are unrealistic and that we
should consider a work practice approach for handling emissions
from open ended lines and open ended valves. The commenter noted
that the current requirement states that open ended lines and
valves must either be equipped with a second valve or be plugged
or capped when not in use. The commenter pointed out that at a
large facility, it is virtually impossible to achieve 100 percent
compliance with this requirement because of various operational
and maintenance situations. The commenter suggested that we
consider a work practice that would confirm the status or correct
the status of a cap or plug each time an open valve is monitored
under the valve monitoring provisions, without creating a
violation. The commenter suggested that the CAR would be a good
opportunity to introduce such a program, and said their company
would welcome the opportunity to work with EPA to develop a low
burden work practice system for open ended lines and valves.
Response: We did not include a work practice approach for
handling emissions from open ended valves and lines.
Incorporating such a standard would have required additional
study and analysis. Performing this analysis was not within the
scope of the CAR, so we did not incorporate the work practice
approach suggested by the commenter.
V
3.6 CLOSED VENT SYSTEMS AND CONTROL DEVICES
3.6.1 Performance Tests
Comment: One commenter (IV-G-01) asserted that performance
tests should not be required if a source has previously conducted
one for any referencing subpart, even if the methods or
conditions during the test were different than those specified in
the CAR. The commenter (IV-G-01) reasoned that the CAR'S
requirement to conduct a performance test would be an
unaffordable burden to any company bringing a significant number
of sources and control devices under the CAR and would likely
3-75
-------
keep companies from opting to use the CAR. The commenter
(IV-G-01) contends that the requirement to conduct a new
performance test would not create any new environmental benefit.
Response: The CAR does not impose a new burden by requiring
new performance tests when test conditions have changed. This
requirement already exists in the referencing subparts.
Performance tests are only needed if a source has previously
conducted one under a referencing subpart but the conditions were
different than those specified in the CAR. Re-testing is
required even by the referencing subparts when a process change
has occurred. This is appropriate because a process change could
change the conditions such that the control device no longer
meets the rule requirements. The CAR allows an option for the
facility to demonstrate that the performance test demonstrates
compliance despite the process change [see §65.157(b)(1) of the
CAR] .
We stress that, with very few exceptions, the referencing
subparts use the same test methods as the CAR for their
performance tests. We therefore contend that there will be very
few, if any, instances where initial performance tests will need
to be repeated, if they have been previously conducted under the
requirements of a referencing subpart. The only differences in
test methods where new performance tests may be required between
the CAR and the referencing subparts are: subpart BB requires
Method 25A or 25B while the CAR requires Methods 18 or 25A for
concentration of regulated material; subpart DDD require.s Method
3 while the CAR requires Method 3B for oxygen concentration.
Comment: One commenter (VI-D-03) requested clarification of
the CAR preamble language requiring the preservation of past
compliance obligations. Specifically, the commenter (VI-D-03)
cited the proposal preamble at 63 FR 57761: "In addition, owners
and operators who choose to comply with the CAR are still
obligated to fulfill requirements that applied while they were
complying with a referencing subpart. For example, if a facility
is required by a referencing subpart to complete a performance
3-76
-------
test, opting to comply with the CAR does not remove the
requirement to conduct a performance test or protect the source
from enforcement actions for not completing the test." The
commenter stated that the last sentence of that paragraph implies
that the CAR performance test requirements are overridden by
referencing subpart requirements, even after a source has come
under the CAR.
Response: A source is obligated to fulfill the requirements
of the original referencing subpart. The CAR does not, for
example, require performance tests where performance tests have
previously been conducted and no process changes have been made.
But if a referencing subpart requires a performance test and the
source has not yet completed the test, this obligation.remains.
The source cannot opt into the CAR and avoid the requirement.
This is similar to the "there shall be no gaps in
compliance" language of §65.1(f) (2). Compliance with periodic
reporting and all of the other requirements under the referencing
subparts must continue while the source is implementing the CAR.
Then, on the implementation date [see §65.1(f)(1)] the CAR
requirements are followed in place of the specified requirements
of the referencing subparts.
3.6.2 Control Requirements
Comment: One commenter (IV-G-01) requested that we allow
routine maintenance of control and recovery devices without
requiring a facility to shut down its process, as long as the
facility could demonstrate that the process could not be. shut
down. The commenter (IV-G-01) cited that this type of option has
been available to some facilities subject to MACT standards,
listed the HON storage vessel provisions as an example, and
suggested that this option could be made available only for a
limited number of hours per year. The commenter (IV-G-01)
maintained that it is often safer and more environmentally
beneficial to perform maintenance on control and recovery devices
without shutting down the entire process, because many continuous
processes can require days to start up and shut down. The
3-77
-------
commenter (IV-G-01) indicated that the current provisions create
an atmosphere among owners and operators that they must install
"redundant" emission control equipment at "considerable expense"
in order to minimize excess emissions. The commenter (IV-G-01)
provided three examples of situations in which it was claimed
that the emissions•from a shutdown and subsequent startup of the
process would be greater than if the control devices had been
serviced while the processing continued.
Another commenter (VI-D-03) requested that the allowance of
240 hours per year for planned routine maintenance be extended to
control devices that control process vents, transfer operations,
and equipment leak emissions. The commenter stated that it is a
common industry practice to combine vent streams for an emission
control device. The commenter asserted that allowing planned
routine maintenance for all control devices on emission sources
would increase the incentive to use the CAR. The commenter
contended that this request is appropriate because the same types
of control devices are presently used to control emission from
process vents, transfer operations, and equipment leak systems
and are being operated and maintained in the same manner. The
commenter stated that allowing planned maintenance of these
systems will reduce malfunctions and thus provide environmental
benefit. The commenter (VI-D-03) provided suggested edits to the
CAR for this provision.
Response: We do not find the industry examples compelling
and point out that two of the examples should be considered
malfunctions in the source's startup, shutdown, and malfunction
plan rather than routine maintenance.
The commenter's (IV-G-01) request is addressed through the
provisions for startup, shutdown and malfunction in §65.6 of the
CAR. In the commenter's (IV-G-01) first example, a compressor
that normally collects vent streams, compresses them, and then
returns them to the process sustains mechanical problems. This
is not an example of routine maintenance, but rather an example
3-78
-------
of malfunctioning process equipment and the situation should be
addressed through the SSM plan as specified in §65.6 of the CAR.
The commenter's (IV-G-01) second example is not an example
of routine maintenance either. The commenter refers to a
situation where a carbon bed is due for replacement and that this
replacement would normally occur during a planned process unit
shutdown. Due to abnormal fouling or plugging the replacement
cannot occur during a planned shutdown. Such an event would
qualify as malfunction, not as routine maintenance. As with the
commenter's first example, this situation should be handled
through the startup, shutdown, and malfunction provisions of
§65.6 of the CAR.
For the commenter's (IV-G-01) third example, we remind
owners and operators that when they determine a control strategy
for complying with regulations, they need to consider issues of
routine maintenance of control and recovery devices. If an owner
or operator chooses to route several processes to a single
control or recovery device, this would obviously necessitate
shutting down all the processes routed to that device in order to
perform maintenance on the device. This is a choice that the
owner or operator makes. The owner or operator just as easily
could choose to route those processes through more than one
control or recovery device. In fact, owners or operators could
choose, in the example provided by the commenter (IV-G-01), to
V
route emissions to a backup recovery device or control device,
such as a flare, rather than vent uncontrolled emissions, from
several processes to the atmosphere while maintenance is
performed on the primary control or recovery device. The owner
or operator has flexibility in the choice of how to design the
emissions control system, and that flexibility precludes
uncontrolled emissions during situations of routine maintenance.
With regard to the HON storage vessel provisions that
contain this provision, we point out that when an owner or
operator has a malfunction in a control device for a storage
vessel, it is not possible to simply "shut down" the storage
3-79
-------
vessel, that is, one cannot always put the stored liquid
somewhere else. Also, more emissions would result from emptying
and degassing the storage vessel then would occur from allowing
the emissions to bypass the control device for up to 240 hours
per year. Therefore, there had to be some provision for shutting
down control devices for storage vessels to allow for their
maintenance and repair. We do not believe that it is appropriate
to provide this allowance for process vents.
Comment: One commenter (IV-G-01) requested that we
incorporate changes in §65.147 of the CAR that were recently
promulgated in amendments to 40 CFR parts 60 and 63 [see §§60.18
and 63.11(b)] regarding the operation of flares that burn
hydrogen as fuel or that could be modified to burn hydrogen as
fuel.
Response: We concur that the requested changes would be
beneficial and have incorporated the changes into §65.147 of the
CAR.
Comment: One commenter (IV-G-01) suggested revising the CAR
at §65.144(a), which currently states "except as provided in
65.3(b)(l) of subpart A...." The commenter (IV-G-01) suggested
that the paragraph should be revised to state "except during
periods of startup, shutdown, and malfunction...." This would
make the paragraph consistent with the CAR at §65.145(a).
Response: We have revised the final rule at §§65.144(a) and
65.145(a) to state "...except during periods of startup,
shutdown, and malfunction as specified in §65 .3 (a) . . . . " '.We agree
that the two paragraphs should be consistent. However, the
reference to the appropriate general provisions paragraph was
also needed so that the details of compliance during startup,
shutdown, and malfunction were clear.
Comment: One commenter (IV-G-01) claimed that drafting
errors in §65.155 (other control devices) should be corrected so
as to make §65.155 applicable to §65.145 (storage vessels and
low-throughput transfer racks) and to §65.146 (equipment leaks).
3-80
-------
Response: We contend that the citations are correct as
proposed, noting that §65.155 provides for "other" control
devices on process vents and high-throughput transfer racks only.
Comment: One commenter (IV-G-01) requested clarification to
the proposal preamble (63 FR 57781) on which referencing subparts
lack specific halogen vent stream requirements.
Response: The only referencing subpart with halogen vent
stream requirements is the HON. No other referencing subpart had
specific halogen vent stream requirements. However, the halogen
vent stream requirements of the CAR do not pertain to equipment
leaks or storage vessels. Therefore, the CAR halogen vent stream
requirements have been extended to sources subject to 40 CFR
part 60, subpart III, NNN, RRR and 40 CFR part 61, subpart BB.
Comment: One commenter (IV-G-01) suggested that
§65.144(b)(2) should apply to equipment leaks routed to a
process, and that leaving "equipment leaks" out of the first
sentence was a drafting error.
Response: We clarify that while §65.144 pertains in general
to fuel gas systems and processes to which storage vessel,
transfer rack, or equipment leak regulated material emissions are
routed, not all paragraphs of §65.144 apply to all types of
emission points. In particular, §65.144(b)(2) was not intended
to apply to equipment leak emissions routed to fuel gas systems
or processes. That is why §65.144(b)(2) explicitly states, "For
storage vessels and transfer racks...."
3.6.3 Monitoring, Recordkeeping, and Reporting
Comment: One commenter (IV-G-01) requested clarification of
the monitoring requirements for a halogen scrubber system where
the first scrubber is the acid scrubber and the second is the
caustic scrubber. The commenter (IV-G-01) points out that
personnel who are expert with scrubbers indicate that pH should
be measured on the first (acid) scrubber and the liquid-to-gas
(L/G) ratio should be measured on the final (caustic) scrubber.
Response: We have not added specific monitoring parameter
instructions for monitoring a halogen scrubber system because
3-81
-------
these systems can vary in design. Specific parameter monitoring
requirements may not fit all cases. However, monitoring
procedures for scrubber systems (or any alternative monitoring)
can be approved under the CAR provisions in §65.7(b). If an
owner or operator feels that pH should be measured on the acid
(first) scrubber and the L/G on the final (caustic) scrubber, it
is possible to have such a monitoring plan approved under the
alternative monitoring provisions.
Comment: One commenter (VI-D-01) expressed that the CAR
effectively consolidates the monitoring, recordkeeping, and
reporting requirements of the applicable rules. The commenter
(VI-D-01) recommends enhancing this concept by establishing one
stringent requirement for all applicable reporting frequencies,
one date for periodic reporting, and the same retention
requirements for all records. Another commenter (VT-G-03) stated
that there is benefit to stating specific dates for reports
rather than gearing report dates to other unspecified dates.
Response: We note that §65.5(h) provides for establishing a
common schedule for all reporting frequencies at a facility. We
maintain, however, that it is beneficial and not unduly
burdensome to allow owners or operators to adjust the reporting
dates.
Comment: One commenter (IV-G-01) suggested removing the
reference to continuous emissions monitoring systems (CEMS) from
the proposal preamble (63 FR 57760) because it incorrectly states
that "...no CEMS...provisions are included in the CAR since they
are not applicable,...." The commenter noted that CEMS is one of
the monitoring provisions under the HON and the CAR.
Response: We would like to clarify what was meant in the
preamble. The commenter is correct that, in certain situations,
a continuous organic concentration monitoring device is an
alternative to parameter monitoring for recovery devices.
However, the general provisions regarding CEMS are not applicable
to these devices. Therefore, the proposal preamble specified
3-82
-------
that "...no CEMS...provisions are included...since they are not
applicable."
Comment: One commenter (VI-G-03) suggested that all CAR
monitoring, recordkeeping, and reporting should use uniform
standards and procedures that are reflected in the relevant
title V permit. The commenter (VI-G-03) gave an example: cost
savings can be achieved if the frequency and timing of
monitoring, recordkeeping, and reporting for all CAR units is
consistent. The commenter (VI-G-03) stated that there is also a
benefit to making these requirements consistent with the most
recent rulemaking approaches, such as the HON. The commenter
(VI-G-03) suggested that future regulatory amendments to part 65
should agree with these more uniform requirements, rather than
create new, duplicative requirements. The commenter (VI-G-01)
also suggested that the amount of required monitoring,
recordkeeping, and reporting should be weighed against the
benefit to the environment. The commenter (VI-G-03) gave an
example: for some equipment, there may be no net benefit to
equipment leak monitoring semi-annually versus monitoring
annually.
Response: The commenter did not specify what they believe
is inconsistent between the CAR and title V records and reports.
We believe the CAR is consistent with title V. With respect to
equipment leak monitoring, the CAR retains the "good performance"
rewards of the HON by allowing longer periods between monitoring.
Indeed, the CAR has taken this a step further by allowing even
longer periods than the HON for even better performance.
Comment: One commenter (IV-G-01) pointed out an apparent
contradiction in the proposal preamble at 63 FR 57782 in the
paragraph beginning with "The CAR adopts the requirements of
40 CFR parts 61 and 63..." The commenter asked how the immediate
repair or replacement of CPMS parts requirements are not in the
General Provisions of Part 60, but these requirements are spelled
out in sections 40 CFR 60.11(d), and 60.13(e) and (f) .
3-83
-------
Response: In the proposal preamble, we made the point that
this provision is not explicitly in the 40 CFR part 60 general
provisions, but it is implicit in the part 60 general provisions,
at §§60.11(d) and 60.13(e) and (f). These provisions spell out
good control practices.
Comment: One -commenter (IV-G-01) suggested a restructuring
of the proposal preamble discussion at 63 FR 57782 for the
paragraph beginning with "The CAR provisions are different from
the non-HON referencing subparts..." The commenter stated that
the rest of this paragraph about 3-hour averages seems to be a
different topic and the first sentence seems to go with the next
paragraph about the CAR allowing the use of ranges from the
non-HON referencing subparts instead if site-specific ranges.
The commenter (IV-G-01) expressed confusion regarding the
discussion in these paragraphs.
Response: The discussion in the paragraph cited by the
commenter was meant to be compared with the information in the
previous paragraph. The previous paragraph describes the
provisions of the CAR and the paragraph cited by the commenter
describes the provisions of the referencing subparts. By reading
both paragraphs, the differences between the two are apparent.
In order to clarify the differences, we will describe them
here. The CAR and HON require daily averages and the
40 CFR parts 60 and 61 (non-HON) referencing subparts require
3-hour averages. The non-HON referencing subparts provide the
parameter range that will be used; the CAR and HON allow
site-specific ranges to be developed. Under the CAR and HON, if
the daily average is outside the range, it is a violation of the
operating conditions (after the one excused excursion). The
Administrator may then require a performance test to determine if
there is a violation of the standard, but may also use any
credible evidence to determine noncompliance. Under the non-HON
referencing subparts, if the 3-hour averages are outside the
range, there is no violation of the operating conditions, but the
Administrator can require a performance test to determine if a
3-84
-------
violation of the standard has occurred (again, the Administrator
may also use any credible evidence to determine a violation).
Comment: One commenter (IV-G-01) stated that the proposal
preamble (63 FR 57784) , is not accurate in saying that 40 CFR
parts 60 and 63 general provision for flare requirements require
hourly records of pilot flame monitoring results. The commenter
gave the following reasons for this contention:
1. There is no mention of hourly recordkeeping for pilot
flame monitoring in the flare requirements of §§60.18
or 63.11;
2. The only use of the term "hourly" in the part 60
general provisions is in §60.14(h) through (j) where
"maximum hourly emissions" are referred to; and
3. The only use of the term "hourly" in the part 63
general provisions is in §63.8(g)(2) where "valid
hourly average" is referred to.
Response: The commenter is correct that the general
provisions of parts 60 and 63 do not explicitly state that hourly
records of pilot flame monitoring results are required. However,
both general provisions require that monitoring be carried out to
ensure that the flares are operated and maintained properly. It
is implicit in these provisions that records of the monitoring
results are required. The CAR and the HON explicitly require
hourly records indicating whether there has been outage of all
pilot flames any time during the hour. This is a burden
reduction for 40 CFR part 60, subpart DDD, RRR, and NNN sources v
where continuous records are required.
Comment: One commenter (IV-G-01) claimed that upon sensory
indications of a leak occurring during closed vent system annual
inspections performed pursuant to §65.143(d)(1), the language
should require "elimination of the leak" and not "elimination of
the indications of a leak."
Response: We clarify that a "leak" is not detected until
confirmed through instrument monitoring. If, at the time of the
annual visual inspections, there are indications of potential
leaks (visible, audible, or olfactory), then the owner or
3-85
-------
operator has a choice as to what action to take, but a "leak" has
not yet been detected. One option is to eliminate the
indications of the leak. The other option is to instrument
monitor the closed vent system. If the monitor indicates that
there is a leak, then the presence of a leak is confirmed and the
leak must be repaired.
Comment One commenter (IV-G-01) questioned why closed vent
system leak detection instruments should be calibrated at
10,000 ppm, which seems to be required by §65.143(c) (1) (v) . The
commenter pointed out that the requirement should be to calibrate
at 500 ppm (the leak definition for closed vent systems) or at a
concentration that does not exceed the leak definition by more
than 2000 ppm. The commenter also questioned the need for
§65.143(c)(1)(v)(C).
Response: We clarify that §65.153(c)(1)(v)(B) requires the
calibration gas to be less than 10,000 ppm. Sources opting into
the CAR from 40 CFR part 61, subpart V, will likely have
calibration procedures in place based on 10,000 ppm. By requiring
the calibration gas to be less than 10,000 ppm, the CAR does not
impose additional burden on these sources, yet it is flexible
enough to accommodate the other subparts.
We maintain that §65.143(c) (1) (v) (C) is necessary, as this
paragraph provides for devices with multiple calibration scales.
Note that §65.143(c)(1)(v)(C) specifies a maximum concentration
i
of 2,500 ppm for the lower scale's calibration gas; this is the
leak definition for closed vent systems (500 ppm) plus 2y000 ppm.
3-86
-------
4.0 IMPLEMENTATION AND TITLE V
4.1 IMPLEMENTATION
Comment: Two commenters (VI-D-01 and VI-D-06) support our
proposal to recognize the CAR as an alternative compliance
approach to the individual subparts being consolidated. One
commenter (VI-D-01) pointed out that this will allow States that
provide for alternatives in the implementation of the
consolidated rules to implement the CAR upon promulgation.
Response: We thank the commenters for supporting the CAR as
an alternative compliance approach.
Comment: Four commenters (IV-D-01, VI-D-03, VI-D-06,
VI-G-03)' support our proposal to waive formal delegation of the
CAR where States already have delegated authority to implement
the underlying NSPS and NESHAP. One commenter (VI-G-03) noted
that States need to continue to include the regulated community
as an active participant in the process of "tailoring" the
Federal rule to the States' needs; States need to provide the
regulated community with due process, appropriate standards, and
opportunity for appeals. One commenter (VI-D-03) stated that the
implementation approaches outlined in the preamble remove an
obstacle to CAR implementation and foster further consolidation
with State rules. The commenter (VI-D-03) suggested that we
consider using direct final rules to expedite non-controversial
changes in State implementation plans.
Response: We thank the commenters for supporting the
approaches outlined in the proposal preamble and will consider
using direct final rules in the future, when appropriate.
Comment: One commenter (VI-D-6) also supports an approach
that would provide for implementation of the CAR in States with
4-1
-------
an approved title V program, regardless of whether the State has
received formal delegation of the underlying rules.
Response: At proposal, we stated that delegation of the CAR
could occur if the State has delegation for all the referencing
subparts. However, we agree with the commenter that delegation
of the CAR could also occur when States have an approved title V
program. We recognize that fewer States have accepted delegation
of the part 63 rules than the parts 60 and 61 rules. By
incorporating the part 63 rules into the title V permit as
applicable requirements, the terms and conditions of the part 63
rules become enforceable by the permitting authority through the
permit, as if the part 63 rules themselves were delegated. We
agree that the CAR could be delegated to permitting authorities
with approved title V programs in place, however there are
advantages to obtaining formal delegation of the CAR by the
permitting authority. Delegation should be conditioned to ensure
the CAR is substantively incorporated unchanged into the permit.
As stated above, there are advantages to accepting formal
delegation of the CAR. Permitting authorities that accept formal
delegation of the CAR through delegation of the referencing
subparts, i.e., the HON (or accept formal delegation of any
section 112 requirement), are the clear enforcement authority.
In other words, if the permitting authority does not accept
formal delegation of the referencing subparts, then the EPA
V
Regional Office remains the enforcement authority and sources
must submit duplicate reports to both the EPA Regional Qffice and
the permitting authority. Additionally, if the permitting
authority accepts formal delegation of the referencing subparts
then the permitting authority can make the discretionary
decisions regarding the general provisions authorities. For
example, if a source wants to change some facet of its monitoring
program, then, in some cases, a permitting authority that has
accepted delegation of the CAR can approve this change. See the
40 CFR part 63, subpart E preamble dated January 12, 1999 (64 FR
1879) for more information.
4-2
-------
Comment: Four commenters (VI-D-01, VI-D-03, VI-D-06,
VI-G-03) agreed that we should pre-approve the CAR for reasonably
available control technology (RACT) equivalency for monitoring,
recordkeeping, and reporting. Two commenters (VI-D-01, VI-D-06)
agreed with our decision to allow States to amend SIP-based and
other Federally-based rules so that they point to the CAR as a
compliance alternative. One commenter (VI-D-01) noted 'that this
could help prevent some SCUs from being subject to a separate set
of state regulations, thereby subverting the CAR'S purpose. The
commenter (VI-D-01) stated that it is unnecessary to propose a
streamlined process for approval of SIP submittals that
incorporate the CAR, if the State agency has delegation for
implementation of the subject regulations and has been delegated
the CAR. The commenter (VI-D-01) indicated that the consolidated
rules in the CAR should cover any RACT or air quality standards
considerations and that other alternatives can be approved
through section 112(1) of the Act. One commenter (VI-D-06) also
supports our approach which recognizes that the permitting
authority is logically the best entity to determine whether the
CAR is an appropriate alternative to the SIP.
Response: We thank the commenters for their support and
appreciate the feedback regarding RACT interaction.
Comment: One commenter (IV-G-01) asserts that sources
subject to State air regulations will still have to comply with
State rules, at least for some amount of time. The commenter
(IV-G-01) supports that the proposed CAR includes an accelerated
method of incorporating the CAR so that it supersedes the State
rules in the hierarchy of rule applicability. The commenter
(IV-G-01) cautions that a State may also opt not to incorporate
the CAR, particularly in States that have more stringent
requirements than the CAR. One commenter (VI-D-01) supported the
CAR as a voluntary option for the States.
Response: It is not clear whether the commenter is
referring to State regulations, or to Federal regulations that
are implemented by the State and therefore have been incorporated
4-3
-------
into the State regulations. If the commenter is referring to
State regulations, facilities will continue to be subject to
State regulations regardless of complying with the CAR. Some
States have developed additional regulations that apply to SOCMI
emission points that apply under State law. Some of these
regulations are more stringent than the Federal regulations upon
which the CAR is based. The CAR does not affect these rules.
If the commenter is referring to Federal regulations
implemented by the States, some States can exercise the CAR as an
alternative compliance approach to expedite the implementation of
the CAR, as described in the proposal preamble at 63 FR 57785.
Here, the only delay in implementing the CAR should be the time
necessary to agree on the schedule for the source to come into
compliance with the CAR. If the State cannot immediately allow
use of the CAR as an alternative compliance option because of
procedural or legislative hurdles, there may be a delay in the
implementation of the CAR at that State while the necessary
changes are made and ratified in the State's regulations. The
States also have the option to consolidate the CAR requirements
with the SIP requirements.
The commenter is correct in noting that a State can decide
not to allow the CAR as a compliance alternative. We have made
the CAR optional for industry at each State's discretion by
requiring that the implementation schedule be established by
mutual agreement. Whatever the State decides, the more stringent
State rules will still apply. A more stringent State equipment
leak program, for example, would not be affected by
implementation of the CAR. State rules still apply even when the
CAR is implemented, unless the State removes the obligation.
4.2 TITLE V
Comment: Two commenters (VI-D-01, VI-D-06) support our
approach with regard to incorporating the CAR into title V
permits. One commenter (VI-D-01) stated that a facility's
title V permit is the final consolidation document for the
4-4
-------
monitoring, recordkeeping, and reporting required by all
applicable rules. The commenter (VI-D-01) stated that the title
V permit also provides a checklist for determining compliance and
pursuing enforcement.
Response: We thank the commenters for this support.
Comment: Two commenters (VI-D-01, VI-D-06) pointed out that
some permitting authorities will find that the part 70 permit
renewal is the most reasonable time to implement the CAR. One
commenter (VI-D-01) asserted that this would allow time for the
final issuance of the first phase of title V permits and would
provide guidance for consolidation in the second phase of title V
permit issuance. The commenter (IV-D-01) also indicated that
this would reduce burden to States and sources. The commenter
(IV-D-01) stated that title V permits that have already been
issued would not have to be re-opened, so attention can continue
to be focused on first-round issuance.
Two commenters (VI-D-01, VI-D-06) asked that we give notice
that implementing agencies can use the title V renewal period as
an appropriate implementation timing for the CAR. The commenter
(VI-D-01) notes that because the CAR is not mandatory, it is not
necessary to reopen a permit. The commenter (VI-D-01) pointed
out that States can apply the CAR under title V at renewal rather
than either re-opening a previously issued permit before renewal
or delaying approval of an initial permit.
Response: We have no objections with the approaches
outlined by the commenter for dealing with title V and CAR
interaction at renewal time. We recognize that the States have
the authority to decide how and when to allow sources to modify
their permits and begin compliance with CAR, and would use the
timing as a means to better manage permitting resources.
Comment: One commenter (VI-D-01) indicated that, if the
circumstances change and the CAR were to become mandatory, the
commenter would support our position outlined in section VIII of
the preamble to the proposed rule. Under this position, the
commenter contended, incorporation of the CAR into issued title V
4-5
-------
permits would be allowed under a "notice-only" provision that
would not require EPA and public review. The commenter stated
that this would be applicable only if the permit is incorporating
previously-adopted requirements and if source-specific
requirements are not being established through the permit.
Response: The CAR has been finalised as an optional rule.
Comment: One commenter (VI-D-06) supports our position that
incorporation of the CAR should not require EPA and public review
under a revised part 70. One commenter (VI-D-06) agrees with our
distinction between the types of changes that trigger a
significant permit revision and those that will qualify for a
minor revision. The commenter concludes, as we proposed, that
adopting the CAR into a title V permit is a minor revision. The
commenter (VI-D-06) stated that a "notice only" approach may be
appropriate if such an approach is provided for in the part 70
rules.
Response: We thank the commenter for the support on this
issue. We do note, however, that interaction with the State is a
crucial step in CAR implementation. Some States will have
different timing requirements for CAR transition and permit
modification, so working with your State is very important.
4-6
-------
5.0 CHANGES TO THE EQUIPMENT LEAKS REFERENCING SUBPARTS
5.1 CLARIFICATION OF INTENT
Comment: Two commenters (IV-G-01, VI-D-03) agreed with the
definitions in the CAR for a closed vent system and for a control
device. The commenters suggested revisions to language in the
referencing subparts in order to be consistent with the CAR. The
commenters (IV-G-01, VI-D-03) recommended revisions to
definitions of closed vent systems and control device in §63.161
of 40 CFR part 63 subpart H, §61.241 of 40 CFR part 61,
subpart V, and §60.481 of 40 CFR part 63 subpart W. The
definitions in these referencing subparts, according to the
commenter (VI-D-03), contradict the definition of a closed vent
system in the CAR. Three commenters (IV-G-01, VI-D-03, VI-D-04)
recommended that process piping that routes vapors to a process
or fuel gas system should not be considered a closed vent system
in any rule. The commenter (VI-D-04) suggested that this change
would clarify that hard-piping, ductwork, and connections are
"process" components and not part of a closed vent system. One
commenter (IV-G-01) also agreed with the definition of a control
device and its exclusion of a fuel gas system as a type of
control device.
The commenter (VI-D-03) also suggested revisions to other
sections of the referencing subparts in order to maintain
consistency with the closed vent system definition in the CAR and
to clarify that piping routed to the process or to a fuel gas
system in not part of a closed vent system. The commenter
recommended changes to language in §§63.163(g), 63.164(h),
63.170, and 63.173(f) of the HON which refer to routing vapors to
a control device from pumps, compressors, surge control vessels,
and agitators, respectively. The commenter recommended changes
5-1
-------
to several paragraphs in §60.482 of subpart W. To be consistent
with the suggested definitions of closed vent systems and control
devices, the commenter also recommended changes to several
paragraphs in §61.242 of subpart V.
One commenter (IV-G-01) notes that under the CAR
§65.143(b)(2) and (b)(3), parts of a closed vent system that are
unsafe-to-inspect or difficult-to-inspect are exempt from all
monitoring and inspections requirements as long as certain
conditions are met. The commenter (IV-G-01) claimed that
part 60, subpart W, the HON subpart H, and part 61, subpart V
should be changed to be consistent with the CAR on this issue.
One commenter (VI-D-04) suggested that we revise the
proposed language in §61.242-4(d)(2) of 40 CFR part 60, subpart V
to allow pressure relief devices to be isolated from process
service after a pressure release. The commenter recommended
revising the language to accommodate a specific rupture disk
design arrangement.
One commenter (IV-G-01) requested a revision of the
definition of "connector" in 40 CFR part 61, subpart V and
part 60, subpart W to match the definition in 40 CFR part 63,
subpart H. The commenter pointed to the reasoning in the CAR
preamble at 63 FR 57764 to support this request.
Response: With respect to revising the referencing
subparts, the intent of the CAR is to add references that allow
an owner or operator to choose to comply with the CAR and to make
minor clarifications to the referencing subparts where
appropriate for drafting errors, typographical mistakes, and
internal inconsistencies. It is not the intent of the CAR to
make revisions to the original language or intent of the
referencing subparts to match the CAR improvements. However, We
have made revisions to the equipment leak referencing subparts
where these revisions conform to the safety provisions contained
within the HON and the CAR. We have not generally revised
referencing subparts to conform to new language in the CAR,
5-2
-------
although there are some exceptions in which revisions to the
referencing subparts were a practical solution.
In part, the reason for not making significant revisions to
the referencing subparts is that such changes could have
unintended implications or consequences for owners or operators
who choose to continue to comply with a referencing subpart. The
CAR could then be disrupting existing interpretation or
implementation of the referencing subparts. Therefore, we did
not generally make substantive revisions to the referencing
subparts as suggested by the commenter, and specifically we did
not make the changes suggested by the commenters in this comment.
Comment: One commenter (IV-G-01) pointed out that the delay
of repair provisions of 40 CFR part 63, subpart H,
40 CFR part 61, subpart V, and 40 CFR part 60, subpart W create
a disincentive for an owner or operator to attempt a repair
between the 15th day after detection and the next process unit
shutdown. The commenter (IV-G-01) suggested that similar
revisions in language to the delay of repair provisions found in
the CAR at §65.105(d) should also be made in §63.171 (a) of the
HON, §61.242-10 (a) of subpart V, and §60.482-9 (a) of subpart W.
Response: In order to clarify the intent of the referencing
subparts, we made the suggested revisions to the referencing
subparts. Like the CAR, the other equipment leak rules now make
it clear that the owner or operator can successfully repair a
leaking piece of equipment within the time period without the
equipment being disqualified from the original delay of repair
classification. It should be noted that these revisions are
considered within the scope of the CAR because they represent
clarifications and not revisions to the original intent of the
referencing subparts.
Comment: One commenter (IV-G-01) requested that an option
be provided in subpart V for surge control vessels and bottoms
receivers to comply with the EFR or IFR requirements of 40 CFR
part 63, subpart G. The commenter (IV-G-01) also requested that
the option to route vapors to a fuel gas system or a process
5-3
-------
should be provided in both HON subpart H and part 61, subpart V
and that the option to route emissions to a process be added to
subpart V. Two commenters (IV-G-01 and VI-D-04) noted that there
was no reference to the new tables 1 and 2 added to subpart V.
The commenter (IV-G-01) also pointed out that the reference to
tables 1 and 2 should be added to §61.242-9 of subpart V.
Furthermore, the commenter (IV-G-01) noted that table 1 and
table 2 are for existing and new sources, respectively. The
commenter suggested that "new source" and "existing source" are
MACT terminology and are not appropriate terminology for
subpart V.
One commenter (VI-D-04) requested that we clarify the
purpose of adding a definition for "maximum true vapor pressure"
to §61.241 of 40 CFR part 61, subpart V and the purpose of adding
table 1 and table 2 to subpart V. The commenter stated that they
had not located any occurrences of the term "maximum true vapor
pressure" in subpart V except in the newly added tables 1 and 2.
Response: We would like to clarify that some provisions
were inadvertently left out of the proposed rule. We have
corrected these by doing the following:
• Adding a reference to tables 1 and 2 in §61.242-9 of 40
CFR part 61, subpart V; and
• Adding an allowance to subpart V so that surge control
vessels and bottoms receives can comply with the
floating roof storage vessel provisions of the HON or
by routing to a process or control device.
We did not add the option to route storage vessel emissions to a
fuel gas system to the HON or to subpart V because it is not
within the scope or intent of the CAR to change the compliance
options for sources that choose to continue using the referencing
subparts. The changes we made in this case to subpart V were
necessary to clarify the control requirements for surge control
vessels and bottom receivers. We did not analyze the impacts of
expanding the control options under the HON and subpart V.
5-4
-------
We used "new" and "existing" in the newly added tables 1 and
2 to subpart V to be consistent with the corresponding tables and
concept contained in the HON. The terms "new source" and
"existing source" are not new to subpart V. The definitions are
provided in the part 61 general provisions. See §61.02.
We would also like to clarify that we added the definition
for "maximum true vapor pressure" to subpart V because that term
is used in the newly added tables 1 and 2. Without adding this
definition the rule would not provide guidance on how to
determine the maximum true vapor pressure.
Comment: One commenter (VI-D-04) noted that the proposed
language to §§60.482-2(g) of 40 CFR part 60, subpart W and
61.242-2(g) of 40 CFR part 61, subpart V are significantly
different than §63.163(j) of the HON regarding the designation of
unsafe-to-monitor pumps. The commenter pointed out that the HON
provides more exemptions than subparts W and V for pumps
designated as unsafe-to-monitor. The commenter specifically
pointed out that the leak repair requirements and the barrier
fluid demonstrations are not required by the HON for unsafe-to-
monitor pumps. The commenter suggested that we revise the
proposed language for §60.482-2(g) of subpart W and §61.242-2(g)
of subpart V to match the corresponding language in the
§63.163 (j) of the HON. To achieve consistency with the HON, the
commenter recommended that we add "and (d)" to the list of exempt
paragraphs in §§60.482-2(g) and 61.242-2(g).
Response: We would like to clarify that we intended to
adopt the HON unsafe-to-monitor provisions into subparts W and
V. We inadvertently created some inconsistencies, and we have
edited the CAR as suggested by the commenter to fix the problem.
Comment: Two commenters (VI-D-03, VI-D-04) recommended
revisions to 40 CFR part 60, subpart W and 40 CFR part 61,
subpart V to limit monitoring frequency of unsafe-to-monitor
pumps to monthly during safe-to-monitor times. The commenters
explained that the proposed language for §§60.482-2(g)(2) and
61.242-2(g)(2) require monitoring of unsafe-to-monitor pumps as
5-5
-------
frequently as practicable during safe-to-monitor times. The
commenters stated that during safe-to-monitor times, the
frequency of monitoring for unsafe-to-monitor pumps should be
limited to the monitoring frequency for other pumps. One
commenter suggested adding "but not more frequently than the
periodic monitoring schedule otherwise "applicable" to the end of
§§60.482-2(g) (2) and 61. 242-2(g) (2) .
Response: We have adopted the commenter's suggested
language and have edited the language to be consistent with that
in the CAR at §65.103 (c) (4) (i) . The rule now specifies that
monitoring must occur "as frequently as practical during safe-to-
monitor times, but not more frequently than the periodic
monitoring schedule otherwise applicable."
Comment: One commenter (VI-D-04) suggested that EPA revise
the language in §61.242-8(a) to clarify the requirements for
repair of pressure relief devices in liquid service and
connectors. The commenter suggested that if a potential leak, as
indicated by visual, audible, or olfactory evidence, were • •
repaired, then it is not necessary to monitor the pressure relief
device or connector. The commenter also suggested that EPA add a
paragraph (a)(3) to §61.242-8 to clarify the repair criteria for
potential leaks indicated by visual, auditory, or olfactory
evidence. The commenter pointed out that these revisions would
be consistent with §63.169(a) of the HON.
Response: We agree with the commenter and have made the
requirements for pressure relief devices in liquid servi_ce and
connectors consistent with other, similar provisions in the HON
and in the CAR. Our edits give the owner or operator the choice
of either eliminating the indications of a leak or performing
instrument monitoring to confirm the presence of a leak.
5.2 CLARIFYING OR TYPOGRAPHICAL EDITS
Comment: One commenter (VI-D-04) pointed out that in
several places in the proposed amendments to subparts W and V,
we included the following language, "...routed to a process or
5-6
-------
fuel gas system connected by a closed vent system to a control
device..." The commenter also noted that we used some
inconsistent language to get the point of this requirement
across, and that common, consistent language would improve the
clarity of the rule. The commenter suggested that the language
should read, "...routed to a process or fuel gas system or
connected by a closed vent system to a control device..." The
commenter listed the following affected sections:
§60.482-2 (d) (1) (ii) of 40 CFR part 60, subpart W;
§60.482-4(c) of 40 CFR part 60, subpart W;
§61.242-3(h) of 40 CFR part 61, subpart V; and
§61.242-4(c) of 40 CFR part 61, subpart V.
Response: We agree with the commenter that the suggested
changes would clarify the intent of the rule, and have made the
language consistent in the above mentioned affected sections.
Comment: One commenter (VI-D-04) pointed out that the
language regarding the monitoring of pumps located at an unmanned
site is slightly different in §§60.482-2(h) of 40 CFR part 60,
subpart VV and 63.163(h) of the HON. The commenter suggested
revising the list of exempt requirements in §60.482-2(h) to
include the daily requirements of §60.482-2(d)(5).
Response: We agree with the commenter and note that we
intended to include the daily requirements of §60.482-2(d)(5) to
the list of exempt requirements at unmanned plant sites. We have
edited §60.482-2(h) to reflect this.
Comment: Two commenters (IV-G-01, VI-D-04) advised that the*
sentence found at §61.242-2 (d) (6) (iv) that reads "If thejre are
indications of liquids dripping from the pump seal..." is not
needed because it is redundant to §61.242-2(d)(6)(ii). The
commenter recommended that the sentence at §61.242-2(d)(iv) be
removed.
Response: We agree with the commenter that revising
§61.242-2(d)(6)(iv) as proposed is unnecessary. The language
proposed to be added is redundant to existing language at
§61.242-2(d)(6)(ii). The final rule does not contain this
sentence.
5-7
-------
Comment: One commenter (VI-D-04) pointed out that in the
proposed language in §61.242-11(k) of 40 CFR part 61, subpart V,
the reference to paragraph (1)(2) may be in error. The commenter
also pointed out that the reference in §61.242-11(1)(3) of
subpart V to §60.486(c) of 40 CFR part 60, subpart W may be in
error. The commenter suggested that the appropriate reference
within subpart V would be §61.246(c).
Response: We maintain that paragraph (1)(2) should be
referenced from §61.242-11(k) because paragraph (1)(2) provides
for the designation of the parts of the closed vent system that
are difficult-to-inspect. We agree, however, that the reference
in §61.242-11 (1) (3) should be to §61.246(c), not §60.486(c). We
have edited §61.242-11(1) (3) to correct the error.
5-8
-------
6.0 MISCELLANEOUS
Comment: One commenter (IV-D-03) requested that we include
a provision in the CAR confirming that the SOCMI rules and the
CAR apply only to manufacture of materials produced from the
11 basic SOCMI chemical building blocks and not to extraction or
derivation of chemicals from natural products. The commenter's
request (IV-D-03) is specifically made in regard to turpentine
and turpene alcohols. The commenter cited a body of evidence
from past EPA regulations and background information documents to
support that, currently, EPA does not intend to include chemicals
derived from naturally occurring substances in the SOCMI
regulations. The commenter (IV-D-03) noted that such an
exclusion is not explicit in any SOCMI regulations except the
SOCMI wastewater rule, and suggested adding an explicit exclusion
to the CAR.
Response: We are not addressing issues of applicability
within the referencing subparts under this rulemaking. The
commenter's request is outside the scope of the CAR.
Comment: One commenter (VI-OG-02) alleged that our SOCMI
regulations under 40 CFR part 60, subparts W, III, NNN _and RRR
and 40 CFR part 63, subparts F and G do not include emission
standards for hexachlorobenzene as required under
section 112(d)(2) of the Act. The commenter (VI-OG-02) provided
information from a January 25, 1999 petition from the Sierra Club
to EPA for review of the parts 60 and 63 rules. The commenter
(VI-OG-02) concludes that the aforementioned regulations are in
violation of section 112(c)(6) of the Act and that the CAR
likewise will be in violation of section 112 (c) (6) of the Act
because it consolidates these rules.
6-1
-------
Response: As stated in the preamble to the CAR
(63 FR 57749), it is not the purpose of the CAR to change the
scope of the requirements or applicability of the referencing
subparts. If future changes are made to the emission standards
in the referencing subparts, then those changes will be
incorporated into the CAR at that time if appropriate.
Comment: One commenter (VI-G-05) requested that we use the
CAR to clarify the referencing subparts regarding a source
becoming "no longer subject" to the rules. As an example, the
commenter noted that the HON does not specifically provide for a
facility that has decommissioned equipment, shut down
high-emitting processes, or implemented pollution prevention
process changes. Such facilities may no longer meet the HON's
major source definition and are therefore no longer subject to
the HON. The commenter contends that the referencing subparts do
not provide a clear procedure for how to handle this situation,
and the commenter would like the CAR to not only provide the
procedure but also encourage its use.
Response: It is not the intent of the CAR to fundamentally
or significantly change the referencing subparts. Each
referencing subpart is different with regard to how and when to
determine or re-determine applicability. The CAR is a compliance
alternative available, for example, at facilities subject to the
HON; the CAR does not attempt to alter the initial or any
subsequent determinations of HON applicability to a source.
Comment: One commenter (VI-G-03) noted that the proposed
CAR does not address the mechanism by which future updates for
regulatory programs, such as the Miscellaneous Organics NESHAP,
will be added to the CAR. The CAR should also address how the
requirements for units subject to case-by-case MACT "hammer"
requirements will be addressed under the CAR.
Response: This proposal cannot address mechanisms for
opting to use the CAR in future rules, because these rules do not
yet exist. However, the CAR would not necessarily require
significant modification for use through a future rule. Future
6-2
-------
rules could be written to contain pointers to the CAR. Any
appropriate edits to the CAR to consolidate new referencing
subparts would be made at the time the new referencing subparts
are proposed and promulgated.
6-3
-------
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO 2.
EPA-453/R-99-006
4 TITLE AND SUBTITLE
Consolidated Federal Air Rule for the Synthetic Organic
Chemical
Manufacturing Industry-Background Information for Promulgated
Standards
7. AUTHOR(S)
U.S. Environmental Protection Agency
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Emission Standards Division (Mail Drop 13)
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
12. SPONSORING AGENCY NAME AND ADDRESS
Director
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
September 2000
6. PERFORMING ORGANIZATION CODE
8 PERFORMING ORGANIZATION REPORT NO
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This document contains a summary of public comments received on the Consolidated Federal Air Rule for
the Synthetic Organic Chemical Manufacturing Industry (40 CFR 65, subparts A through G). This
document also provides the EPA's response to comments, and outlines the changes made to the proposed
regulation in response to comments received.
17 KEY WORDS AND DOCUMENT ANALYSIS
a DESCRIPTORS
Environmental Protection, Air Pollution Control, Air
Emissions Control, Synthetic Organic Chemical
Manufacturing Industry, Reporting and Recordkeeping
Requirements
18. DISTRIBUTION STATEMENT
Release Unlimited
b IDENTIFIERS/OPEN ENDED TERMS c. COSATI V
Field/Group
Hazardous air pollutants
Volatile Organic Compounds
19. SECURITY CLASS (Keporr; 21. NO OF PAGES
Unclassified 138
20. SECURITY CLASS (Page) 22 PRICE
Unclassified
-------
U.S. Environments! Protection Agency
Region 5, Library (PL-12J)
71 West Jackson Boulevard, 12lh
Chicago, IL 60604-3590
------- |