NATIONAL EMISSION STANDARDS FOR HAZARDOUS AIR
POLLUTANTS
ETHYLENE PRODUCTION
BACKGROUND INFORMATION DOCUMENT FOR FINAL STANDARDS
SUMMARY OF PUBLIC COMMENTS AND RESPONSES
'
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Emission Standards Division
Research Triangle Park, North Carolina 277-11
l
May 2002
-------
NATIONAL EMISSION STANDARDS FOR HAZARDOUS AIR
POLLUTANTS
ETHYLENE PRODUCTION
BACKGROUND INFORMATION DOCUMENT FOR FINAL STANDARDS
SUMMARY OF PUBLIC COMMENTS AND RESPONSES
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Emission Standards Division
Research Triangle Park, North Carolina 27711
May 2002
-------
TABLE OF CONSENTS
Section Page
LO INTRODUCTION 1
Comment Letters Received ., 2
2.0 SUMMARY OF COMMENTS AND RESPONSES 3
2.1 APPLICABILITY 3
2.1.1'. : Units' That Produce Ethylene Incidentally : - 3
2.1.2 Ethylene Cracking Furnaces 5
2.1.3 Decoking Operations 6
2.1.4 Applicability of General Provisions : 7
2.1.5 Miscellaneous Applicability Issues 8
2.2 OVERLAP WITH OTHER RULES . '..:..: 9
2.3 SELECTION OF MACT FOR WASTE 14
2.3.1 Applicability Cutoff of the BWON 15
2.3.2 Compliance Options of the BWON : 18
2.3.3 Application of BWON Control Standards to All HAP j. 23
2.3.4 Monitoring Requirements -28
2.3.5 Certifications for Off-Site Waste Treatment 30
2.3.6 Waste Definitions 31
2.3.7 HON Point of Determination Concept 33
2.3.8 Waste Reporting Requirements 34
2.4 SELECTION OF MACT FOR HEAT EXCHANGE SYSTEMS 35
2.4.1 Monitoring Frequency 35
2.4.2 Monitoring Location 38
2.4.3 Repair Requirements 43
2.4.4 Suggested Revisions to the Standard 49
2.4.5 Approval of Alternate Analytical Methods 52
2.5 SELECTION OF MACT FOR EQUIPMENT LEAKS "-,- 53
2.6 SELECTION OF MACT FOR PROCESS VENTS 58
2.7 SELECTION OF MACT FORSTORAGE VESSELS, 63
2.8 SELECTION OF MACT FOR TRANSFER OPERATIONS _. 64
2.9 CLARIFICATIONS AND CORRECTIONS 65
2.9.1 Waste Provisions 65
2.9.2 Heat Exchanger Provisions 67
2.9.3 Process Vent Provisions 69
2.9.4 Closed Vent Systems 70
2.9.5 Miscellaneous Corrections 71
2.10 FORMAT OF RULE 71
-------
LO INTRODUCTION
On December 6, 200Q, the U.S. Environmental Protection Agency (EPA) proposed
regulations to establish National Emission Standards for Hazardous Air Pollutants (NESHAP) for
new and existing ethylene production sources. This document contains summaries of the public
comments that EPA received on the proposed standards and EPA's responses to those comments.
This summary of comments and responses serves as the basis for revisions made to the NESHAP
between proposal and promulgation. The EPA received 11 comment letters in response to the
proposed ethylene production NESHAP. The commenters are listed in the following table.
The proposed ethylene production NESHAP references the requirements of 40 CFR Part
63 Subpart SS National Emission Standards for Closed Vent Systems, Control Devices, Recovery
Devices, and Routing to a Fuel Gas System or a Process (subpart SS), Subpart SS was amended
when the ethylene production NESHAP was proposed. Many commenters provided comments
on the amendments to subpart SS in their comments on the ethylene production NESHAP,
Subpart SS applies to a variety of sources in addition to ethylene production facilities. The
amendments to subpart SS and therefor the comments on subpart SS are being addressed under a
separate rulemaking effort. Comments received on subpart SS are not summarized in this
document.
-------
Comment Letters Received Regarding the Proposed Ethylene Production NESHAP"
Item Number
Commenter and Affiliation
IV-D-01
IV-D-02
IV-D-03
IV-D-04
IV-D-05
IV-D-06
iv-D-07
IV-D-08
IV-D-09
IV-D-10
rv-D-ii
Stancy Simpson
Eastman Chemjeal Company
Longyiew, TX
Norman Morrow
ExxonMobil Chemical
Houston, TX
Stephen Knis, Susan Taylor, and John Ogle
The pow Chemical Company
Freeport.TX
John Soice
Union Carbide Corporation
South Charleston, WV
JackMcClure
Shell: Chemicals
Houston, TX
Courtney Price
American Chemistry Council
Arlington, VA
David Deal
American Petroleum Institute
Norbert Dee
National Petrochemical and Refiners Association
Robert Kelley
Formosa Plastics Corporation
Livingston, NJ
David Foerter
Institute of Clean Air Companies
Washington, DC
Stephen Knis, Susan Taylor, John Ogle
Dow Chemical Company
Freeport, TX
"Comment letters are available in Docket A-98-22 located at EPA's Air and Radiation Docket and
Information Center (MC-6102), 401 M Street SW, Washington, DC 20460, RoomM-1500,
Waterside Mall (ground floor, central mall). The docket is available for public inspection and
copying between 8:00 am and 5:00 pm Monday through Friday. A reasonable fee may be charged
for copying.
-------
2.0 SUMMARY OF COMMENTS AND RESPONSES
In addition to specific comments provided, several commenters (IV-D-01, IV-D-02, IV-
D-03, IV-D-04, IV-D-05, IV-D-08, IV-D-07) expressed overall support for the comments
submitted by the American Chemistry Council (IV-D-06).
2.1 APPLICABILITY
2.1.1 Units That Produce Ethvlene Incidentally
Comment: Five commenters (IV-D-02, IV-D-05, IV-D-06, IV-D-08, IV-D-07)
recommended that the applicability provisions be clarified to make sure that units that incidentally
produce streams containing ethylene and/or propylene are not inadvertently covered by the rule.
Four commenters (TWD-02, IV-D-05, IV-D-06, IV-D-08) noted that an ethylene manufacturing
process unit (EMPU) is defined as "a process unit that is specifically utilized for the production of
ethylene/propylene, including all separation and purification processes." The eommenters believe
that this definition could be construed to include process units that produce ethylene or propylene
incidentally. Two commenters (IV-D-02, IV-D-06) explained that most polyethylene plants,
ethylene glycol plants, and ethylene oxide plants have ethylene-containing recycle streams. The
commenters also noted that petroleum refinery process units that use ethylene or propylene as
feedstocks can also have such streams. One commenter (IV-D-08) explained that, as part of the
cracking process, petroleum refinery catalytic cracking units produce ethylene and propylene.
Ethylene is normally separated as a finished product and it is not uncommon for propylene to be
separated by distillation as a finished product or intermediate stream. Two commenters (IV-D-
02, IV-D-06) stated that units that produce ethylene and propylene were not included in the data
-------
on which the NESHAP is based and are subject or will be subject to other NESHAP and so
should not be subject to the ethylene production NESHAP. Three commenters (TVXD-02, IV-I>
06, IV-D-07) suggested the following revision to the definition of ethylene manufacturing process
unit: "Ethylene manufacturing process unit means a process unit thai is .speiifiuajly utilized for the
production of ethyltiae/propyleiie; whose,primary purpose is to produce ethvlene as a feedstock
for another process or for commercial sale, including all separation and purification processes."
One commenter (TV-D-05) suggested using the approach used in the hazardous organics
NESHAP (HON) of basing applicability on the unit's intended primary product, in this case,
ethylene.
Response: For consistency, EPA has decided to use the definition of an ethylene process
unit found in the HON and the Federal Register Notice that added the ethylene source category,
with one exception. The HON definition is for "ethylene process or ethylene process unit". The
EPA prefers to use the term "ethylene production" and "ethylene production unit", therefore,
"process" has been replaced with "production" in the HON definition. In the revised regulation,
the HON-based definition of ethylene production or ethylene production unit replaces the
definitions of ethylene production and ethylene manufacturing process unit in the proposed
NESHAP. The definition reads: "ethylene production or ethylene production unit means a
chemical manufacturing process unit in which ethylene and/or propylene are produced by
separation from petroleum refining process streams or by subjecting hydrocarbons to high
temperatures in the presence of steam. The ethylene production unit includes the separation of
ethylene and/or propylene from associated streams such as. a C4 product, pyrolysis gasoline, and
pyrolysis fuel oil. Ethylene production does not include the manufacture of SOCMI chemicals
such as the production of butadiene from the C4 stream and aromatics from pyrqlysis gasoline."
To ensure that refinery process units and units covered by the HON that produce ethylene
and propylene incidentally are not subject to the requirements of the ethylene production,
NESHAP, the EPA has added an.exclusion in §63.1103(e)(l)(iii) that states that the provisions of
the ethylene production NESHAP do not apply to process units and emission points subject to
subparts F, G, H, I, and CC of part 63.
-------
2.1,2 Ethvlene Cracking Furnaces
Comment: Three commenters (IV-D-Q2, IV-D-04, IV-D-06) expressed strong support for
EPA's determination that ethylene cracking (pyrolysis) furnaces are.part of the ethylene
production scSjai^e'icategoi^ and that no regulatory requirements should apply. One cowjraenter,
(IV-D-Op^) aj|reefi[ that ethylene cracking furnaces are unique and fundamental components of
ti " -. b'ft^vk. ' .. ^1'i-V '":.: " ' - "'': ' -''-L-1' "':'
ethyleie prooucfion units arid are more appropriately considered in the context of the ethylens
; .> '£; :.' ' " -'- '-.' ;'" ;< '." ; ' ''';,; 'r':'.-L:'
production NESHAP than |n-:tha> context of a broader, generic standard
-------
category. For this reason, "(G) All ethylene cracking furnaces and associated decoking
operations" was added to the list of emission points that make up an affected source in
' . ' ' ' "' > ';'': ! -H'" v''*> - '' v. .':<-., . "
§63.1103(e)(l)(i). The EPA believes that the addition of this language makes it clear that
. -- . . .' ' , . - .':.:,- N, ('' ''''~'; J'-h '" '£<
.-,. , .- . . - i. -: - ' ' " ' "*'S
cracking furnaces are part of the ethylene production source category and it is npt necessary to
-. -.:Vv- -' ' '.>%4, '-;>M.> EW>i;v&,:, ;;^:.V^.^vto;<'!,ff /^*V*'r£ VJi^'A'
further stale that they are not part of the process heater source: category, Add$i#fta|y* qPA
"=:» :?*;.;*; -'? v --* .-',,','.': :;g,^fu^^ .";..::/! f' ;'tt iMv 'V--- ^-^i|1 -*-,'*"' :
believes tbat, if there is any doubt aboiit1 which category cracking fufnaces are in.?t|je*!|»|ile wjll refer
to the preamble .of the proposed rule and. this document for fuirther clarification,
Ta.clariry that there are no requirements:for cracking furriaces, "(J) Air emissions, from all
:|i{ :, ; v:; . : ,.-':-'.; & ". (., ;. ' 1 ;"r : '. t ' * '!.l " '' ' , '. "n '" '"':
ethylene cracking furnaces, including furnace stack emissions during decoking operations" was
added to. the list of emission points in §63.1103 (e')(!.)(») that are in the ethylene production
source category but are not subject to the requirements of the etjiylene production NESHAP
The exception does not include non-furnace emissions during decoking operations. This is
because there are requirements in the NESHAP for decoking operations. Decoking operations
are considered a period; of shutdown in the NESHAP and non-ethylene cracking furnace
emissions must be addressed in the startup, shutdown, and malfunction plan.
I , : . > »
2.1.3 Decoking Operations
i . .'
Comment: Three commenters (TV-D-02, IV-D-04, IV-D-06) expressed support fpr
including decoking operations as part of the ethylene production source category. Two
t |
commenters (TV-D-04, IV-D-06) supported EPA's decision to regulate the decoking of ethylene
furnaces as a shutdown activity. One commenter (IV-D-06) stated that they are not aware of any
ethylene unit that employs air pollution control equipment on emissions from decoking and they
do not believe that the use of such equipment would be cost effective. The commenter added that
it is common practice for ethylene producers to establish standard operating procedures for
decoking, therefore, the requirement to include decoking in the startup, shutdown, and
malfunction plan is consistent with the practices employed by the best perforrjiing facilities. One
commenter (IV-D-02) suggested adding the following to §63.1103(e)(l)(i): ''(G) All ethylene
cracking furnaces and associated decoking operations" where emission points that are part of the
ethylene production source category are listed.
-------
Response: As suggested by the commenter, "(G) All ethylene cracking furnaces and
associated decoking operations" has been added to §63!1103(e)(l)(i) where the emission points
that make up 'the ethylene production source category are listed.
2,1.4 AppHcabflityafGferteral-Provisions.
dbrhrtietit: (bnb commuter (IV-D-Q9) asked whether 40 CFR Part 63 Subpart A - General
Provisions (-subpart A) apply to units affected by 40 CFR part 63 subpart XX. The commenter
noted that in §63.1(a)(4) iKis'stated that: 'The provision of this subpart apply to owners or
operators who are 'subject to subsequent subparts of this part, except when otherwise specified in
a particular subpart of in a relevant standard..." The commenter stated that owners and operators
of EMPUs with heat exchangers and waste streams are subject to subpart XX, which is
subsequent to subpart A. The commenter stated that this leads to the conclusion that the
provisions of subpart A apply to sources subject to subpart XX. The commenter does not believe
that this is EPA's intent and suggested that language similar to that found in 40 CFR part 63
subparts SS, TT, UU, and WW be added to subpart XX. The commenter suggested the following
language, from subpart SS, could be added: "The provisions of 40 CFR part 63, subpart A
(GeneraTProvisions) do not apply to this subpart except as specified in a referencing subpart."
The commenter recommended that a table of the subpart A general provisions that do and do not
apply be included in subpart XX and 40 CFR part 63 subpart YY for further clarification as was
done in other regulations.
Response: The EPA agrees that it was not their intent for all of the provisions of 40 CFR
part 63 subpart A to apply to sources subject to subpart XX. The sources subject to subpart XX
are subject to 40 CFR part 63 subpart YY. Subpart YY include requirements similar to those in
subpart A as well as references to subpart A. Language similar to that suggested by the
commenter has been added to subpart XX. The EPA believes that subpart YY clearly states
which parts of subpart A are applicable in §63.1100(b) and it is not necessary to provide a
separate table to repeat this information.
Comment: One commenter (IV-D-09) requested clarification on the general provisions of
40 CFR part 61 that apply. The commenter pointed out that, according to the proposed
NESHAP, compliance with 40 CFR part 63 subpart YY constitutes compliance with the BWON
-------
for waste streams that are subject to the- control requirements of both/. The commenter asked
whether .wast&,stfeams .that meet this criteria are therefore no longer subject to^he general
f
provisions of part 61. If the genera^ provisions of part 61 remain.applicable, the commenter
requested clarification.on which sections apply. The coi?;^atpE^u
applicable regulations!, such as^tables.2 and\7Arn; 40: CFR Part 6Q Si^bpart,
Wastewatei; New. Source Performance Standard wpuld^be helpful: Sirnilarly^ the qprijr
asked for clarificatjon on whether the;general provisions of part 60 apply, and if SQ> wh|ch, to a
* ', . ' -,£ » | - - r. ..... -. - . .
storage vessel, that is. subject to the.requirements of 40 CFR part 60 subpart Ka or Eb and is only
required to comply wj.th the requirements of; subpart YY. The cornmeater (IV-D-09)
recommended that only the .part 63 subpart A general provisions apply and that language be
added to §63.1100(g) to reflect this.
Response: The waste requirements of the ethylene production NESHAP reference the
requirements of the BWON, and therefore the general provisions of part 61. Regardless of
whether a facility was previously subject to the BWON, to comply with the waste requirements it
will have to comply with the general provisions of part 61. Because storage vessels subject to the
control requirements of both subpart YY and subpart Ka or Kb are required to comply only with
the provisions of subpart YY, the general provisions of part 60 would not be applicable.
2.1.5. Miscellaneous Applicability Issues
Comment: One commenter (IV-D-09) suggested that the following, emission points be
r
added to the list of equipment/activities that are in the ethylene production source category but
are not subject to applicability requirements at 40 CFR §63.1103(e)(l)(ii): "(J) Water from testing
of firefighting equipment"; and "(K) Equipment that is intended to operate in organic hazardous
air pollutant service, as defined in §63.1020 of subpart UU of .this part, for less than 300 hours
during the calendar year."
Response: The EPA agrees with the commenter that water from testing of firefighting
equipment is not subject to requirements of the NESHAP and has revised the list of exceptions to.
include it. Equipment that is intended to operate in organic hazardous air pollutant service for
less than 300 hours during the calender year can not be included in the list of equipment for which
there are no requirements because there are requirements for such equipment. In 40 CFR part 63
-------
subpart UU, §63.1022(b)(5) requires the identification of equipment in service less than 300 hours
per year in records. For this same reason^ the requirement to comply with subpart UU only for
equipment that is1 in service for 300 hours per year or more in table; 7 to §63V1 103 of subpart YY
-' 'i '.. ' . , - . %'--
was delete'ft IffisubpaTt^IiU, it is 'Clarified tfrat such1 equipment that' is\ifl-seryice "fes -Inan S'&O
;!, > - . 'i i ''.
hours per year is subject only to the identification requirements in §63.1
2.2 OVERLAP WITH OTHER RULES
Comment: Two commenters (IV-D-02, IV-D-06) recommended that 40 CFR
§63.1100(g)(4)(ii) be expanded to state that compliance with 40 CFR part 63 subpart UU (the
referenced subpart for equipment leak requirements) constitutes compliance with 40 CFR part 60
subpart W and 40 CFR part 61 subpart V (subpart V). One commenter (IV-D-06) justified the
change by noting that §63.1100(b)(4)(i) states that compliance with subpart TT constitutes
-'. ..} ' " :
compliance with subparts VV and V and subpart UU is more stringent than subpart TT. In a
similar comment, one commenter (IV-D-09) recommended that, for existing equipment that is
subject to both 40 CFR part 63 subpart YY and 40 CFR part 60 subpart J - National Emission
Standards for Equipment Leaks (Fugitive Emission Sources) of Benzene, only the requirements of
subpart YY should apply. The commenter suggested that §63.1100(g)(4)(i) could be revised to
accomplish this.
Response: The EPA agrees that subpart UU is more stringent than subparts VV and V.
In addition, since the requirements of 40 CFR part 60 subpart J (subpart J) mirror those of
subpart V, compliance with the equipment leak requirements of subpart UU also constitutes
compliance with subpart J. As a result, 40 CFR §63.1100(g)(4) has been revised to note that
compliance with the equipment leak requirements of suhpart UU constitutes compliance with
subparts W, V and J, and, as such, affected sources need only to comply with the requirements
of subpart UU.
Comment: Five commenters (IV-D-02, IV-D-05, IV-D-06, IV-D-07, IV-D-08) suggested
that operations included in other source categories should be expressly excluded from the affected
source. The commenters agreed with EPA's statement in the preamble that the affected source
does not include pieces of equipment currently included in other source categories and
9
-------
recommended that an express exclusion for operations subject to other tylACT standards be
included in the,list of exceptions to the affected spurce, One compenter (IV-D--P5) suggested :.
using the same, approach used in the hazardous organics iSfESHAP (HpNj) andjhe petroleum
grJe;i^|E;S^^
Response; As the commenter B.6te$V'tlje EPA states in the preamble that the affected
source for ethylene production does not include pieces of equipment currently included in other
source categories. To ensure that this is clear, §63.1 l:03(e)(l)(iii)has been added to subpart YY
to specifically exclude process unifs and emissjon points.subject to 40.CFR part 63 subpartS' F, G,
H,I, orCC.
Comment: One commenter (IV-D-09) recommended that language.be added to 40 CFR
§63.1101 clarifying that if terms defined within subpart YY are also defined in the Clean Air Act
or in the General Provisions at §63.2 or in other subpart;s (such as 40 CFR part 61 subpart FF at
§61.341), the definition in subpart YY applies. The commenter provided as an example the
definition of "Startup", which is different in subpart YY and the part 60 General Provisions. The
commenter contended that these different definitions could lead to inconsistent understanding and
enforcement of Startup requirements. The commenter suggested EPA clarify which definitions
are superceded.
Response: The EPA notes that there may be some discrepancies in definitions in subpart
YY, the general provisions part 63, and the general provisions of part 60.. However,, it is the
intent of the EPA that, in cases of overlap, the definitions that apply to the superseding subpart
take precedence over those applicable to the superceded subpart. For example, for equipment
lealcs, the definition of "Startup" found in subpart YY would supercede the definition presented in
40 CFR part 60 general provisions. Text has been added to §63.1100 of subpart YY to further
clarify overlap of subpart YY with other subparts.
Comment: One commenter (IV-D-09) pointed out that 40 CFR §63.1100(g)(l) through
§63.1100(g)(4) only address overlapping regulations for existing sources and does not include
new sources. The commenter stated that this leads to the conclusion that a new storage vessel,
for example, that meets the applicability requirements of 40 CFR part 60 subpart Kb (subpart Kb)
___ .. 10
-------
and the proposed NESHAP would have to comply with both. The.cpmmenter recommended that
new sources be regulated in the same way as existing sources with regard to overlapping,
requirements. The commentet suggested that the word "existing"vbe removed from.
§63.1 lQQ(g|(l) thiro^gh|<53f 1;l-i30^gXfXto accomplish this.
Respdnser T,he^P^,agrees with the comment, Therefore, we have revised subpart YY
by remqvirig:''ejqs^g'Vf|otn;;§6:3..riOO(g)(l) through (4). The storage vessel requkements of the
ethylene NESJIAP are.af jeast as stringent as. the requirements of subpart Kb.
Comment: One c^mm'enler CJjV-D-09) requested clarification as to which requirements
apply when there are overlapping requirements in 40 CFR part 63 subparts XX (subpart XX) and
SS (the Generic MACT)j., As; one example, the commenter pointed out that in §6l,349(a)(2)(iii)
of the Benzene Waste Operations NESHAP (BWON), flares must be in compliance with,§60.18>
However in §63.;987(a) pf. subpart SS, flares must be in compliance with §63.11(b). As another
example, the commenter, pointed out that §61.342(c)(l)(iii) of the BWQN applies to all waste
management units (WMU) downstream of the point of generation for waste streams while
§63.1106(a) applies to all WMU downstream of the point of determination. The BWON requires
WMU to be inspected quarterly, while the HON requires semi-annual inspections. The
commenter asked if WMU are to be inspected quarterly under the modified BWON and
semiannually under the requirements of 40 CFR part 63 subpart YY (subpart YY).
Response: The EPA notes that the comment is correct in that the BWON requires
compliance with §60.18 for flares while subpart SS requires compliance with §63.1 l(b).
However, the requirements of §60.18 and §63.1 l(b) are the same. The EPA believes that the
second example presented in the comment refers to situations where a ethylene production unit is
located at a facility with a waste management unit (WMU) that receives wastes from both
ethylene and non-ethylene processes, and the WMU is complying with, the requirements of the
HON. It is not our intent to require duplicate monitoring, .reporting, and recordkeeping for the
shared WMU. In cases where a WMU is currently complying with the requkements of the HON
and is receiving wastes from ethylene production, the requkements of the HON will apply to the
ethylene production waste stream once it ceases to be associated with the ethylene production
unit and has become associated with the WMU. While the waste stream is associated with the
-------
ethylene production unit, the waste requirements of subpart XX apply. It should be noted that the
waste requirements for ethylene production sources are contained in subpart XX, and affected
Comment: Oiie commenter (IV-D-09) asserted that the waste requirements of the
proposed NESHAP are over-burdensome and?ic0nifesiStg because the^ requite co;m|)%ii^e with the
revised .RWON for HAP-containihg Wa$tewzjter streams, the wastewater p^ch/isioMof^CFR
part 63 sqbpkrt YY and the HOHfor HAP-eor^taining wastewater streams: f he eprrpienter
stated that facilities will have to determine which waste streams are affected by at le'ast triree
different regulations, each of which essentially achieve the same HAP reduction. The conimenter
stated that it is difficult to determine which sections of the revised B WON and subpart YY Would
apply to a storage vessel subject to the proposed NESHAP and the HQN based on
§63.1100(g)(6).
The commenter predicted that even more overlap will occur when EPA finalizes the New
Source Performance Standards (NSPS) to control volatile organic compound (VOC) emissions
from wastewater streams in the Synthetic Organic Chemical Manufacturing Industry (SOCMI)
through the proposed 40 CFR part 60 subpart YYY.
The eommenter recommended that facilities should have to comply only with the
wastewater provisions of the HON rather than the revised BWON, proposed subpart YYY, and
the HON. The commenter stated that this would be consistent with EPA's current thinking as
demonstrated in the Consolidated Federal Air Rule (CAR). The commenter noted that the CAR
consolidates several air rules to eliminate confusing and overlapping issues, fulfilling a Presidential
initiative. Alternately, the commenter suggested that if §:63. 1 100(g)(6) is meant to clarify that
only the HON applies, language should be added to clarify that §63.1 106 does not apply.
Response: The EPA believes that the commenter may have misinterpreted the proposed
waste provisions. It was not our intent to require duplicative monitoring, recordkeeping, and
reporting for waste streams. The waste requirements for ethylene production units are contained
in 40 CFR part 63 subpart XX (subpart XX). While 40 CFR part 63 subpart YY contains
requirements for ethylene production processes, the waste requirements of §63.1106 of subpart
YY do not apply to ethylene production (see previous response). Ethylene waste is not subject to
12
-------
the requirements of the HON. If a wastewater stream, waste management unit, or waste
treatment unit is subject to both subpart YY and the HON, there are provisions in §61.1100(g)(5)
and (g)(6^) to address this issue.
Comment: One commenter (IV-D-02) expressed concern over apparent overlap of the
monitoring, reporting, and recordkeeping requirements for closed vent systems and contircil
devices in §61.349 of 40 CFRpart 61 subpart FF (subpart FF) (applies'to'waste) arid par£6.3
subpart SS (applies to the remainder of the ethylene process). The commenter noted thatlhis may
lead to confusion and compliance difficulties, especially for shared systems because differing
requirements could apply to the same systems. The commenter suggested that compliance with
'' l '. 1 '? '
the subpart SS closed vent system and control device requirements be allowed as an alternate to
complying with §61.349 6f subpart FF and its associated monitoring and recordkeeping
requirements. The commenter claimed that this would promote compliance, simplify training, and
provide common requirements for shared systems. The commenter suggested the following
addition to the rule:
"§63.1099 What alternates to Table 2 and 3 of this subpart are allowed?
(a) A source complying with the requirements of this subpart for wastes may, as an alternate to
the closed vent system and control requirements of §61.349 of subpart FF of part 61, comply with
the requirements of subpart SS of this part as specified in §63.982(b), (c)(2) or (e).
(b) A source using this alternate must maintain a record showing that they are complying with
control efficiency and language substitution requirements of Table 2 relative to each such
requirement."
Response: The EPA generally agrees with the comment. The EPA recognizes that
facilities may use a common closed vent system and control device for a combination of process
vents, storage vessels, transfer operations, and/or equipment leaks (which would be subject to
subpart SS) and vents from waste management units (which would be subject to §61.349 of
subpart FF). Therefore, there exists a possibility that two sets of requirements could be applicable
to the shared system. To remedy this situation, the EPA has inserted provisions into subpart YY
that will allow facilities to meet the requirements of subpart SS instead of subpart FF for closed
vent systems and control devices for waste vents (see §63.1100(g)(2)(iii)). In these overlap cases,
13
-------
subpart SS is more representative for determining compliance with multiple MACT regulations,
since subpart SS applies to^all HAP instead.pf benzene only. For consistency witti the subpart SS
requirements, the control device is required to reduce organic HAP by 98 weight-percent or
reduce organic HAP or TOC to a concentration of 20 ppmv.
. - S'!'-: " -n-^T
2.3 SELECTION OF MACT FOR WA3TB
:'r;;it, "-l&^l, - >i" '' ' 'v' - J.' '" -1' - ''-'_!
Comment: Seven commenteis (W-B;-Qt IV-D-02, IV-D-03, IV-D-04, IVfp-05, IV-D-
' p '. '"'»{("' '<.": '' "' ^ :'-,- / ' ' '; '':! " ' '- '
06, rV-D-Q8^< expressed strong support J>r using 40 CFR Part 6.1 Subpart FF Benzene Waste
OperatioriSjN^SHAP .(BWON) to regulate wastes under the ethylene production NESHAP.., Two
'' ^>J. '.- ; :V- -.'-.'.'. '' -i ''. . - -
commenters (JV-D-03, IV-D-05) stated that the BWON ensures protection of human health and
' : ' P;i' : ' : '.,_:. j1 ;
the environment, one commenter (IV-P-05) citing risk analyses and adoption of the,BWON as
:','( -s. ' >. -. ' '- ' ' ' - '.- ' '
MACT for the.petrpleumrefineries NESHAP,!as proof. Several commenters (IV-D-01, IV-p-02,
IV-D-03, IV-D-04, IV-D-05, rV-D-06? rV-0^-08) expressed extreme concern with the modified
version of the;BWON requirements in the proposed ethylene production NESHAP. One
commenter (IV-D-02) asserted that the modified BWON will require major revisions to Title V
permits and permit applications for all sites with EMPUs or that handle waste streams from
EMPUs and that this burden is not addressed in the proposal or in the information collection
request submitted to the Office of Management and Budget (OMB). The commenter (IV-D-02)
contended that the EPA is attempting to revise the BWON for waste streams shared by ethylene
and non-ethylene production units. The commenter stated that the EPA is doing so without
meeting the statutory requirements of § 112(d)(2), noting that § 112(q).requires the EPA to revise
pre-Amendment standards in accordance with § 112(d). Additional comments on specific
provisions are provided in detail in the following sections.
Response: In the proposed ethylene production NESHAP, EPA attempted to use the
BWON as the basis for the ethylene waste requirements. However, there are many differences in
the BWON requirements and those that are appropriate for ethylene waste. For example, the
BWON requires control of benzene, whereas the ethylene production NESHAP requires control
of HAP. These differences required many changes to be made to theBWON requirements,
resulting in a complicated and possibly confusing rule. Based on comments received and
additional information obtained, EPA has revised the waste requirements. They are now more
14
-------
similar to the BWON requirements and less complicated. Specific changes made are discussed in
v . ...--.; -,.;'
the following responses.
2.3,1 Applicability Cutoff of the BWON
Comment: Two commenters (IV-D-05, IV-D-06) contended that not including the
applicability cutoff of 1Q iriegagrarn (Mg) of total annual benzene (TAB) that is currently in the
:^U. ? .-.. '.: -v.'... ' . .-.-. ;.,;:..;:, :: }:-:;, ,./;- :
BWON jntt^e ethylerie production NESHAP is inconsistent with, apd an unjustified departure
i'. V' .': - ti'i,:,, "< '-.-- ' ' ." ' v ;. jf. ' ' ' '"'""
from, EfA's earlier conclusion thaj the 10 Mg TAB is a part of dje lyJACT floor for the
...: <:.& . - . ' ' .- :*'VJ : ' '" '" ' '*. " ' -.
petroleum refineries NESHAP. One'commenter (IV-D-06) cited National Emission Standards
'-.. '! . , <,-,, '. - ' ' ' '.:, '; ;... ' '; ,,.'., ' '
for Hazardous Air Pollutants, Petroleum Refineries - Background Information for final
',. , ' >-i ': ' ' "-'.','
Standards, Summary ofpublte Comments and Responses, in which it is explained why the 10 Mg
TAB applicability cut off V/^s-jnciu^ed in the requirements for refinery waste. The commenter
also cited thte preamble ft>rithe promulgated petroleum refineries NESHAP in which EPA states
that the 10 Mg applicability cutoff was not eliminated because the emission reduction achieved
i1 .'''' ': '
was not significant, given the associated costs. The commenter argued that there is no basis for
treating the ethylene industry differently than the refining industry with regard to waste
requirements. The commenter noted that, as with the refining industry, there are facilities in the
ethylene production industry that are not subject to the requirements of the BWON because they
have a TAB of less than 10 Mg and the same conclusions reached for the refining industry should
apply to the ethylene production, industry. The commenter noted that one of the facilities
determined to be among the five best-performing (Equistar, Louisiana) has a TAB of less than 10
Mg and therefore is not subject to the control requirements of the BWON.
Response: Based on the MACT floor analysis, four out of the five best performing
facilities are controlling HAP emissions from waste. At a minimum, the ethylene production
NESHAP must require all major sources to achieve the level of control achieved by these sources,
which requires the management and treatment of HAP-containing waste streams. Not requiring
facilities with a TAB less than 10 Mg to control to an equivalent level would be less stringent than
the MACT floor. The EPA does not regard the fact that the wastewater requirements for the
petroleum refineries NESHAP are different from the ethylene production waste requirements as
15
-------
an inconsistency even though both industries are subject to the BWON. It is not uncommon for
i . - ''",-'.-*'' ' . i .
different MACT floors to be determined for similar sources.
Comment: Three commenters (IV-D-04, IV-D-05, IV-D-06) argued that because, in their
opinion, the 10 Mg TAB applicability cutoff is part of the MACT flpor, not including it in the
NESHAP is la option more stringent than thfe floor (atipye-the-tioor option.) Grip ckamjnenter
.-;;" HJJ. ., , '' J-.'<\ h , :-'^;^..;' :1l ! v: ^x,;;--vv:.r .; ^->-.&: -v^'/ . -,
(IV-I>06) co'iitihded that not including the 1Q Mg TAB applicability cutoff is net a justifiable
- .;-<,;ss:-.i ". :*>' -. ':..'<'. if v'^t-i;^-: 'X. ']''-' i" ':" 7 .
above-the-floor\>ptk>n. The commenter referred to EPA's1 estimate that it wii coist faculties not
K.r'V'. '5 'V: .
-------
thaft 6 Mg/yr;- If the; TABsfor the facility is already "6 Mg7yr or lessi lho streams would have to be
managed and treated, which is not consistent with the MACT floor level of control. Requiring
facilities to^comply With the standard requirements of the BWON would al^o net be 'appropriate
becatlS^it'^y^fairet-h^facilitjeisto treat intermittent strearrisiwhich 'at^jgeneYaty noitorltrblled
by the b^stperfotrrtfHgi facilities that form the basis of the MAGT lloor deterrriuLi3itibrii
f- . ' '''''..
We haveideterrhined t-hattte most appropriate way to require facilities witiiaTAB
quantity feSs'than 10 Mg/yr to achieve the level of control achieved by the1' best perfblrQing
facilities is to specify the streams that must be controlled. Data received since pro;p6$ai indicates
that the best performing :ethylene facilities control two types of streams as part of thek BWON
compliance strategy: (I)1 spent caustic streams (wastes from the caustic washing process to
remove sulfur compounds and other contaminants from the process stream), and "(2) dilution
steam bloWdoWn streams (condensed steam used to quench the cracked gas condensates). We
have determined that it is appropriate to apply the flow rate and concentration control'
applicability cut-offs in the standard requirements of the BWON to these streams. The best
performing facilities are generally not'controlling intermittent streams.
Based on this information, the ethylene production NESHAP has been revised to require
that facilities with a TAB less than 10 Mg/yr manage and treat, according to the requirements of
the BWON, each spent caustic and dilution steam blow down waste streams with a flow rate
greater than or equal to 0.021pm, and annual wastewater quantity greater than or equal to 10
Mg/yr, and a benzene concentration greater than or equal to 10 ppmw. The control requirements
for these streams applu at all times except during periods of startup, shutdown, and malfunction
(SSM), if the SSM precludes the ability to comply and the facility follows the provisions of their
SSM plan.
Comment: One commenter (IV-D-06) disagreed with EPA's assumption that eliminating
the 10 Mg TAB applicability cutoff would simplify compliance with the proposed standard. The
commenter agreed that calculating a TAB is complicated and time-consuming but argued that
k
eliminating the cutoff does not reduce administrative burden and can not be justified through
1 burden reduction. The commenter stated that the new burden items that are being added by
17
-------
applying the BWON requirements to all HAP (see section,2,33) far exceed any, burden reduction
associated with eliminating requirements to estimate a TAB.
.Response: The EPA wqukl like to clarify that we have determined that the MACT; floor
i [ -
foi; .^a$j&^|^ ilAI-co.ntairiln^ st]ream&4t;ja]a[ fi^G|tttie^ that are
major sources. T||e;^sbn the-.pthylene; pr^di.icti0p:NEiS3E|AP d&e^npt include..^: 10:Mg;TAB'
applicability;cut-p^ is, to(accur,ately representtthe jyiACT floor, not.to reduce bujpde,rj. .The, EPA
would also;ljke; to: point put that, in the reYtised^ESHAPvbenzene, can be used as a surrogate for
HAP and therefore ..there is no new burden associated with applying the BWON requirements to
all HAP (see section 2.3.3).
Cornment: One commenter (IV-D-04) disagreed with EPA's assertion that the BWON
only applies,tp facffities with a TAB quantity of 10 Mg or greater. The commenter explained that
the BWOIS( applies to all chemical manufacturing plants, coke byproduct recovery plants, and
r ~ '' - *
petroleum refineries and includes recordkeeping and reporting requirements for facilities with a
TAB quantity lessithan 10 Mg. The commenter stated that EPA incorrectly determined that, the
BWON does not apply to several ethylene production.facilities, including one of their own, when
in fact, the BWON applies to all ethylene production facilities.
Response: The EPA clarifies that, in.previous statements, it was EPA's intent to state that
the requirements of the BWON to manage and treat waste streams only applies to facilities with a
TAB quantity of 10 Mg or greater. The EPA agrees that facilities with a,TAB quantity less than
10 Mg are subject to the BWON and must comply with associated recordkeeping and reporting
requirements.
2.3.2 Compliance Options of the BWON
Comment: Two commenters (IV-D-05, IV-D-06) contended that not including the
compliance options that are currently in the BWON (referred to as the 1, 2, and 6 Mg/yr
compliance options) in the ethylene production NESHAP is inconsistent with, and an unjustified
departure from EPA's earlier conclusion that the compliance options are a part of the MACT
floor for the petroleum refineries NESHAP. Another commenter (IV-D-02) noted that, by not
including the 1, 2, and 6 Mg/yr compliance options, the EPA is reversing a previous determination
that these compliance options meet the requirements of CAA § 112. The commenter contended
18
-------
that the EPA provided no legal or environmental justification fcir this;change. Two commenters
(IV-p-01, rV-D-,06) noted that the compliance options are different from the 10 Mg TAB
applicability cutoff in that they are equivalent compliance alternatives, not exemptions from the
BWON. ..Therefore*, according to-one commenter (IV-D-06), the rationale med to gjimcaEtte the
10 MgTiAfi app)icabipity.icutdffcan not be applied to the compliance options. Theveoinirnenter
argued that elinrinatito^of the compliance options is inconsistent 2 Mg/yr, alternative
(ExxonMobil, Texas and>Equistar, Texas) and one uses the & Mg/yr alternative (BP, Texas). The
commenter stated that only one of the facilities identified as being among the five best performing
complies with the control requirements of the BWON and does not use a compliance alternative.
One commenter (IV-D-Ol) claimed that facilities utilizing the 6 Mg/yr alternative may be
controlling HAP to a greater degree than facilities not using any compliance alternative. The
commenter, explained that a facility with a large wastewater flow may install a more efficient
steam stripper to control benzene and HAP to concentrations much lower than the required 10
ppm to avoid more costly control of smaller wastewater flows from maintenance operations. The
commenter contended that the compliance alternatives result in win-win situations for industry
. and the environment and encouraged EPA to adopt the BWON without changes.
Response: Since proposal of the ethylene production NESHAP, EPA has obtained
information regarding which .facilities are using compliance options and what streams they are
controlling. EPA's.general finding is that, regardless of how they are complying with the BWON,
facilities typically control continuous streams and tend not to control intermittent streams.
Examples of streams that are typically not controlled are samples and maintenance waste (both
during normal operations and turn-arounds). The fact that the same types of streams are typically
controlled, regardless of whether a facility is complying with the standard requirements or a
compliance option, supports the finding that the three compliance options are equivalent to the
19
-------
standard BWON requirements and to each other in the level of control achieved. EPA has
determined that it is appropriate to include the compliance options in the ethylene production
NESHAP for facilities with aTAB greater-tha^i 10 Mg.
Comment: Four c^tranenters (IV4XD2, IV-EWj IV-D-05, IV-D-06) claimed tfiat
because, in their.o|inionj the 1,2, and 6 Mg/yr compliance, alternatives ate part (M^the'M&CT
floor> not including the compliance alternatives in the NESilAP is an above-the-flbor option. One
commenter. (IVhD-Q6) argued that as ap above-the-fldor option, :not including the compliance
alternatives is not cost effective and therefore the alternatives should be included in the NESHAP.
The commenter disagreed with EPA's assumption that facilities already complying with the
BWON would have no additional compliance costs. The eommenter reported that, based on a
study they conducted, 20 ethylene manufacturing process units (EMPU) that currently comply
with the BWON using compliance alternatives would have to add hard pipe collection systems
and 4 would need additional wastewater storage capacity and/or treatment units. The commenter
added that additional vent control or closed loop sampling systems would need to be added to
other units. Overall, the commenter estimated that facilities currently using BWON compliance
alternatives would need to spend several million dollars to comply with the NESHAP if the
alternatives are not included. The commenter explained that they do not have sufficient
information to quantify emission reductions, but suggested that $15,000 per Mg of HAP
reduction would be a conservatively low estimate of the cost effectiveness of not including the
compliance alternatives. The commenter cited EPA's nationwide cost estimate, floor
determination, and default parameters as supporting documents that are incorrect because they
are based on the assumption that facilities with a TAB greater than 10 Mg will not.need to
upgrade equipment or install additional equipment to comply with the proposed NESHAP.
Another commenter (IV-D-05) referred to the study, stating that preliminary estimates of capital -
costs for complying with the BWON without compliance alternatives exceed $100 million
One commenter (IV-D-05) estimated that it would require capital expenditures of $15
million for two facilities to achieve compliance if the compliance options in the BWON are not
considered to be part of the MACT floor. The commenter added that this estimate .does not
include operating and maintenance expenses. Another commenter (IV-D-02) noted that the 2
20
-------
Mg/yr compliance'option is currently emplo'yed at one Of their ethylene production facilities so
that they will nOt have to control the very low volume, low HAP concentration waste streams
(usually maintenance; related); Tine commenter argued that control of these streams would be
tremendously costly, arid, difficult and result in minimal impact on HAP emissions (and that the
exiting 1, 2 and;6 Mg/yr cornpliahce options included in the BWQN sriould be retained.
One commenter (tV-D-^M,), stated that, according to the preamble, EPA considered and
rejected above-the-flboi::option;s for waste and therefore, the waste requketnepts in the proposed
NESHAP are inappropriate.
One commenter (IV-p^05J stated that an additional cost of not including compliance
options that EPA did not consider is the disruption of compliance that will occur. The commenter
explained that most EMPUs are collocated with other manufacturing units and facilities comply
on a source-wide basis and some are part of a BWON compliance program managed between
different companies. The commenter asserted that significant time and resources will be required
to extract the EMPU from existing monitoring and compliance assurance programs and reevaluate
compliance. The commenter added that sources will be required to modify permits fo'r EMPU
and other units at the facility and States will be burdened with the review and approval Of these
modifications.
Response: The EPA appreciates the detailed information that the commenters provided.
As discussed iii the previous response, the EPA has determined that it is appropriate to include
the compliance options in the ethylene production NESHAP. Therefore, it was not necessary to
add costs for facilities that are currently using the 1, 2, or 6 Mg/yr compliance options because it
is assumed that they will continue to do so.
Comment: Two corriirienters (IV-D-02, PV-D-06) disagreed withEPA's assumption that
eliminating the compliance options would simplify compliance with the proposed standard. One
commenter (PV-D-06) agreed that performing the calculation required to use a compliance option
is complicated and time-consuming but argued that eliminating the compliance options does not
reduce administrative burden and can not-be justified through burden reduction. One commenter
(TV-D-05) Speculated that EPA is implying that the reduction in burden associated with not
performing calculations for a compliance option offsets the increased burden due to the alterations
21
-------
to the rule, and argued that the burdens can not be compared. One commenter (IV-D-06) stated
that the new burden items that are;being.-added: by applying the BWQN requirements.tp all HAP
(see section 2.3.3) far exceed any burden reduction associated with eliminating the compliance
options. T^wo commenters (IV-D-05, iyjE)-02) noted that tabje 2 of the
references §6L3^7(:d)(2) which willjequire; pwners and operators to
calculations they,woaildhave.performed'if^sing a compliance option.'
Response: As discussed in, a previous response,, the EPA has.deterrnined.that it is
appropriate to include the compliance options in the ethylene production NESHAP,; The
compliance options are included because they were determined to reflect the leyel of control
achieved by the bestperforming facilities :and to be equivalent to the standard BWON ,
requkements. TheEPA would also like tp point out that, in the revised NESHAP, benzene can
be used as a surrogate for HAP and therefore there is no new burden associated with applying the
BWON requirements to all HAP.
Comment: Onp commenter (IV-D-02) contended that the proposal does not address .
implementation and compliance of the waste requkements at facilities with botrrethylene and non-
ethylene units subject to the BWON. The commenter questioned how EPA and State, agencies
will enforce these requirements since, as the commenter contended, the requirements seem
overlapping, unclear and contradictory. The commenter explained that the proposal will cause
portions of some streams to have 1, 2, and 6 Mg/yr compliance options, while the rest of the
stream will have no.compliance alternatives, claiming that this will make compliance and
compliance monitoring virtually impossible for these streams. The commenter added that if these
compliance obligations for shared systems can be identified, there will be a tremendous burden
imposed on the industry, States, and EPA for retraining personnel and inspectors^ OH the
requirements. The commenter (TV-Dr02) provided that they have an EMPU collocated on a site
with an aromatics operation, the BWON applies to both and the site uses the 2 Mg/yr compliance
option. The commenter stated that, having two different compliance systems at this site would-
increase burdens, cause confusion, and generally make compliance difficult. The commenter
added that removing ethylene streams from the 2 Mg/yr exemption would provide more room for
aromatics streams under the exemption. As a result, there would be no reduction in benzene
22
-------
emissions due to the elimination of the 2 Mg/y'r Compliance optidti for the ethylene streams. The
commenter recommends maintaining the existing 1, 2 and 6 Mg/yr compliance options as found in
the BWON.
Responses As discussed in a previous response, the EPA ha^s determined that it is
appropriate to include the compliance options in the ethylene production NESfeAP. Tne inclusion
of the compliance'options; eliminates the commenter's concerns regarding how ttb apply fee
, s , . i;, i .;
ethylene production NESHAP and' the BWON at a single facility!if a compliance'b-p'tion is being
[ ' , - ,. ; > . ' , |.
used. Under the current ethylene production NESHAP, such a facility can continue to use the
compliance option^ including waste streams associated with the ethylene process in the 2
Mg/yr exemption.
2.3.3 Application of BWON Control Standards to All HAP
Comment: Three commenters (IV-P-02, IV-D-04, IV-D-06) disagreed with EPA's
assumption that controls and treatment systems that achieve the necessary benzene reductions to
comply With the BWON will achieve the same reductions for all HAP listed in table 1 of the
proposed NESHAP (table 1 HA$). Two commenters (IV-D-02, IV-D-06) noted that no data or
information were provided to substantiate the assumption. The commenters stated that the
proposed NESHAP exceeds the MACT floor by imposing the same emission limitations and
treatment requirements for table 1 HAP as currently apply to benzene under the BWON, an 11-
fold increase in stringency (based on there being 11 table 1 HAP), according to one commenter
(IV-D-02.) Another commenter (IV-D-04) estimated that the HAP concentration in wastewater
streams is about three times the benzene concentration, according to survey data submitted to
EPA. The commenter concluded that changing the 10 ppmw benzene limit to a HAP limit is
inappropriate because the limits are not equivalent.
One commenter (IV-D-02) expressed concern over the replacement of "organics or
benzene" to only "organics" in table 2 of 40 CFR part 63 subpart XX. The commenter stated that
the reason benzene was included as an option to organics in the original rule (BWON) is that
measurement of total organics or achieving the required removal efficiency may be extremely
difficult or infeasible in many cases. The commenter asserted that the change enacted by table 2
23
-------
will impose unreasonable requirements for some systems. The commenter recommended deleting
entries in table 2 where "qrganics" is substituted for, "organips or benzene. "
j -,..- ^ . ->.' - - ! ."-" i. - ' '''-' :.<
One commenter (IV-D-02) claimed, that, because EPA provided little time to review and
comment; on the proposed NpS|JAE, ,%y,wer.|j ngt .able, to, fujly evaluate the:.i(rjpa.cts of-the
requirementsiP^heic:^treatment systems^^., p^fri^^rflmenfer p/oyi|pd|h;at j)re|(rnin>ry sampling on,
one of thfi% tEeaJnieiiit system^ (a stearji s^i|)p^ii) found that a reduction to; I 0 PPiny could ,110* be-
achieved for nattrrtMene, demonstrating that control of total HA? to lO.ppnav cannot be achieved
;--; 'i V'juis*",! ;-"'! :" ' '> . ''<;f?-* :«!. rf ».', - *,,'. *;:-; *
' '- ''- ".- ~; . '
with existing syjsteips. Another comment^r.(I^rp-03, iy-P-ll)iagreed;that ejcisting strippers
meetjing the ben^er^e removal requkernenfs.of tjie BWON are; also,removing.ftAPs listed in table
1. However, the commenter stated concentrations of other HAP in the unstripped streams is
often much lower than benzene, and although tiie stripped streams typically contain much less
than 10 ppm HAP, 99 percent removal of all HAP is not always practical The commenter
provided additional data in an attachment that is considered to be. confidential business
information. Another commenter (TV-D-04) provided that, using survey data^submitted to EPA,
the WaterS program showed that some of the HAP present in wastewater from ethylene
production facilities has a greater affinity for water than benzene and will therefore be more
difficult to remove. The commenter cited styrene and naphthalene as examples of such HAP.
The commenter suggested that cutoffs of; 1000 ppmw of table 9 HAP in the HON or 500 ppmw
total VOC in 40 CFR part 60 subpart YYY are more reasonable cutoffs for control applicability
when considering overall HAP content. The commenter noted that the cited regulations also
include low flow and concentration exemptions, similar to the BWON.
One commenter (IVrD-02) asserted that the carbon absorption systems and condensers
and other control systems they use to meet the benzene removal requirements of the BWON may
not achieve the same removal for other HAP. The commenter claimed that only a cryogenic
condenser system can achieve 98 percent removal for butadiene or hexane.
. . \
Response: The EPA's original intent in proposing stream applicability and treatment
requkements based on total HAP content rather than benzene content was to ensure that streams
containing HAP other than benzene are treated and controlled. The EPA maintains that because
compliance with the BWON represents the'MACT floor and results in control of HAP other than
24
-------
berizefte, the Mi&CT floor includes control of HAP other than benzene. However, EPA has
determined that it is not necessary to base stream applicability and treatment requirements on total
HAP to ensure that all HAP are managed and treated. The EPA has revised the proposed
NESHAf'so tftiit/'brg'ialliie's"' is noTKmger substituted for 'toganics or benzene^ Ififen^fion
obtatied'^Krotgl survey rfespohseVarid comments shows that, with few exceptions, an of;4he
.'. : " ; '....in. v. ; '^ ."' < . . . .: = ..-a. >*'. ,.': ; . :
waste steams fromethylMe pfbfcess units that contain HAP contain benzene. According to
commeriters;'of ail the waste sfrSams generated by 33 EPUs, only two do not contain benzene but
-.'"- i ' . - . - -' ' - . -f< ' '!\ ' if r :
contain other HAP- One streain isjvgenerated from a reflux drum oa a debutanizer column. The
stream contains l,3-butadiene#nd has a flow rate of 2 gallons per minute, the other stream is an
intermittent stream that is; generated during turn-arounds that contains naphthalene. - Applying the
' ' '"' l' ', '''ill' r ' I '' ' ' V "
finding that the best-perforrriiijg facilities generally control continuous streams but not intermittent
streams, either due to flow rate and concentration cut-offs or use of a compliance option* EPA
has determined that controlling the' continuous 1,3-butadiene stream, but not the naphthalene
turnaround stream, is consistent with the MACT floor. To ensure that continuous streams that
contain HAP other than benzene are controlled, while at the same time minimizing the burden of
'j i*1 '
identifying these streams, the ethylene production NESHAP specifically requires management and
treatment of waste streams that contain greater than or equal to lO.ppmw of 1,3-butadiehe. To
ensure that this requirement does hot result in the control of intermittent streams that are
generally not controlled, the flpw rate applicability cutoffs for benzene-containing streams (0.02
1pm or 10 Mg/yr of waste) have also been applied to the butadiene streams.
Comment: Five cornmenters
-------
--
compliance must still be demonstrated since the majority of their treatment systems handle waste
streams from ethylene and ppn-ethylene production. As a result^. two compliance dempnstr,atjon,s
would;be required for the. shared treatment systems.
Two bp;mm^nters (IV-P-Q3, iy^Q6) cited the petroleum refineries .NESHAP as
ifei/l5!' '":^-"V^ ";;'' ''-''>':' l*r- " ' -.: 'i;- '-''' :'a ' ' ^\ '"'V' *'': %--'-"^''Sp>;' "-'T'V .;'-" ^ .
providing a Tenant, precedent fox usi]nf4^zene as a surr<^gate for HAJ?. Ohe;ep^mmen]ter (IY
".' M' '''i'^'S?: ' ': 'ffY1-'"'''*1 '" ^""*;l ""'' ''''' '" ' t? "'" ": * }-' ' -T'^'' *'''"' '":'r' V; '"*5 "!.!>;''-' !; '.,..-'
06) qupted tfeiifeainble to the proposed :peta:01eu|a refineries .jSfESIJAP in, which; i^is stated feat
v. ^.^Vll" '.,'-( :^f4 '- -' -' - !" " -'V '"' '""- -' '';;?'- ; 'Llii k.if^ V'Sf' ''V'-'t1 'r.J'.t. ',:":?
benzene! wjas^tod to be an appropriate .surrogate lor HAP in refjnery,>va^t^^ater, streams
/> ,;" "'^tt-' "-/ '1- --" '/'*'-'' 'r "*' ' ' '' ' r"^r) * ^-^''i; '!:' ' -*!t- i'r ' t>
because, benz«n| ^similar to other comppunjds found in refmery ^as|e>yatier and jvolatite| HAP are
, .. 5> '.I f t .. i i J . 'J. - <* . - . > > i
present in a faJrlV constant ratio to benzene. The commenter stated1 thtat the BWON is
;, ; r'ff : r ': -. .' "i\, : : . t *!:,''' ! ;;<- ' ,;: .
incorporated byi'reference into the petraieuiFn refineries NESHAP and therefor ^pA concluded
that benzeiie is arj adequate surrogate. for applicability and treatment purposes and the. same
conclusipn ,is justified for the ethylene produptioh NESHAP. The other commenter quoted the
' / ; r '\ ' ''"'', i ; l- .-;'. < , , ' '
preamble to the^promulgated petroleum refineries NESHAP which states that ".. The BWON
. 'i '!"«!'' ' v .t ' : i ' ' i * . \ - '' : :
controls 75 percent of the benzene in refinery wastewater nationwide and 76 percent of the
volatile organic.HAP in refinery wastewater.." and noted that EPA did not provide. a
rationalization for why this judgement is invalid for EMPUs. One commenter (TV-D-06) explained
r i ' - ' ' - ' ;.-'' -r
that waste streapis in the ethylene units originate near the beginning of the process, before water
is separated and before liquid hydrocarbons are separated from each other by distillation.
According to the commenter, this results in the ratio of benzene to other HAP being consistent
throughout the process. The commenter states that available information indicates that the HAP
to benzene ratio in ethylene waste streams should be relatively consistent. The commenter
referred to a study they conducted that showed that only two wastewater streams at the facilities
covered by the study would be subject to the ethylene production NESHAP based on the content
of HAP other than benzene. The commenter added that because the table 1 HAP have
approximately the same volatility as benzene and have already been determined by EPA to have
the same strippability (based on the fraction removed (Fr) value assigned in the HON), they
should be adequately treated by air pollution control devices and treatment systems installed to
comply with the BWON. The commenter agrees with EPA's finding that table 1 HAP are
controlled to a similar level as benzene by management and treatment of the waste streams. The
26
-------
commenter stated thatrequiretnents to control all table I HAP would not result in greater
emission reductions but would impose substantial burdens and costs on affected sources. The
cornrnenteij listetj burden items including documenting the content of table 1 HAP in each waste
Stream;, performing newtests to show the HAP are controlled to the prescribed level; and
recording anil reporting Qther administrative details differently than tBey'^e for compliance with
the BWON./lThe commenter asserted that these burdens are not justified because the BWON,
without modification, adequately controls all relevant HAP. The commenter also n6ted;tbat
facilities wfll; have -to -update their permits and other approvals and State and local regulators will
have to manage.two versions of,the BWON. Another commenter (IV-D-02) noted that the
proposed standard increases the burden of complying with the BWON by requiring: initial and
annual calculations for HAPs other than benzene.
Response: The EPA.has determined that it is not necessary to express the treatment
requirements in terms of total HAP. The EPA agrees with commenters that treatment and control
devices.used to remove or destroy benzene will remove and destroy the other HAP> regulated by
this rule to approximately the same level. Benzene can be used as a surrogate to determine
treatment and control efficiencies. If no benzene is present in a regulated stream, another HAP
(such as 1,3-butadiene) must be used to show that treatment and control efficiencies required for
benzene are achieved for that HAP. In such cases, compliance can also be demonstrated by
routing the stream to a control device that is being used to comply with the BWON.
Comment: One commenter (IV-D-02) noted that table 2 imposes a potential increase in
control device efficiency over that required by 40 CFR §63.349(a) of the BWON, The BWON
requires removal of 95% of the organics or 98% of the benzene vented to the device. As
proposed, table 2 requires 98% of the HAPs or organics vented to the control device to be
removed. The commenter argued that EPA did not provide adequate justification for this above-
the-floor change of demonstrate that this is achievable by control devices that make up the floor.
The commenter noted that the proposed requirements would only apply to the ethylene portion of
the waste streams in shared systems, while the rest of the streams would be. subject to the BWON
requirements. The commenter argued that it may not be possible to apply two different
requirements, but it is clearly duplicative. Additionally, the commenter noted that to meet these
27
-------
requirements, many of the MACT floor control systems would require upgrading or replacement,
which would cost industry: multiple millions of dollars. The comrnenter stated that there is no
legal or environmental basis for this change, since;the current systems, represent the MACT floor.
The commenter Eecorijmended not changing the control, mpaitoringj recordkeepiag, and reporting^
requirements d^the BiWON. The commekter suggested tkotyisions; to specify rraonltpjihg of '':
another HAP (Jbf example, butadiene) where; benzene is not present, should-such an above-the-
floor requirement; be needed,
Response: The EPA's original intent in requiring a control device used in conjunction
with a closed vent system to achieve, a 98 percent destruction efficiency for all HAP was to reflect
the level of control achieved by control devices inf use. According to survey resporisesy facilities
that use a control device on a closed vent system are achieving a 9.8 percent destruction efficiency.
Although the survey requested the destruction efficiency to be reported in terms of tptal HAP, it
is likely that the facilities reported the destruction efficiency for benzene, since the control devices
in question are in place to comply with the BWON. Based on comments received and further
review of the BWON requirements, EPA has determined that the level of control currently being
achieved by the best performing facilities is best reflected by the control device requirements of
the BWON. Additionally, the EPA is no longer requiring compliance to be demonstrated for all
HAP and is instead allowing benzene to be used as a surrogate. The required destruction
)
efficiency for benzene in the BWON is 98 percent.
2.3.4 Monitoring Requirements
Comment: One commenter (IV-D-06) stated that the requirement to continuously
monitor the benzene concentration of streams exiting treatment processes is an unjustified
departure from prior EPA determinations. EPA selected continuous monitoring rather than
monthly monitoring as required by the BWON to meet enhanced monitoring requirements in
§ 114(a)(3) of the Clean Air Act. The comrnenter argued that the monthly monitoring
requirements of the BWON have been investigated by the EPA and found to be reasonable and
appropriate. The commenter referred to clarifying amendments proposed by EPA in 1992 that
included a proposal to replace the monthly monitoring requirement in the BWON with a
continuous monitoring requirement. The commenter noted that the EPA decided not to change
28
-------
the requirements h the final rule amendments in 1993 and Stated that monthly monitoring was
found to be adequate. The commenter also noted that the petroleum refineries NEiSHAi*,
promulgated in 1995, adopted the monitoring requirements of the BWON. two pomrnenters
(lV^02;iV-t)^6>s1;k^d th$*fie Compliance Assurance Monitoring (GAM) R|^ prcSmulgateii
pursuant to §114Ca)(3)o!f the'CIeMi Air Act, provides that standards' -prQp^sed#|ll:t; November
15, 1990 are exeinpt frbjftt the fciAM'rule because they already presumptively constitute
"continuous compliance detertiifiiafeohs methods." The commeritefs contended that, because the
. I ^. . i. ,' .':i tf.- - . - , ,j .".
monthly monitoring requirements* of the BWQN were ratified afterNoverhle! 15,1996 through
the clarifying amendment to tftirfi'WON and inclusion in the petroleum retteries- NESltAP, they
constitute a "continuous compliance determination method."
One commenter (IV-D-Q2) argued that continuous monitoring of treated effluent is an
extremely costly above-the-floor change to the BWON requirements that, in many eases, is
technically infeasible. The eornmenter stated that it is not possible to continuously monitor for
the 11 HAP covered by the proposed rule at the low concentrations found in.wastewater. The
commenter provided that they tried to monitor for benzene with an online GC but were unable to
keep the analyzer operating due to fouling of the system. The commenter noted that no
continuous monitoring systems exist that can be used in high-pH caustic streams, which precludes
continuous monitoring for at least two of the commenter's treatment systems. The commenter
contended that the EPA has not provided any technical analysis that demonstrates the technical
feasibility of continuous monitoring as required by § 112(d)(2). the commenter stated that the
*. '
installation and operation of continuous monitoring systems for seven treatment systems would
cost well over $1 million initially, and hundreds of thousands of dollars per year for apparently no
environmental benefit. The commenter noted that costs for continuous monitors were not
included in the development of the proposed standard. The commenter recommended that the
changes to the compliance monitoring requirements of the BWON be deleted.
Response: Based on the information provided regarding the feasibility of using continuous
monitors and the potential cost, the EPA has determined that requiring continuous monitoring is
not appropriate for treatment systems. Based on previous determinations made by EPA, the
monitoring requirements of the BWON adequately ensure compliance and are the basis for the
29
-------
requirements in the revised NESHAP, The revised NESHAP requires monthly monitoring of the
benzene concentration of the waste stream exiting the treatment process or continuous monitpring
"'-'--- . ' ,,-i-:.;1-' --vV1 .- ^r.'T/ ,../. i'-v :r.f;<' ^v --.-:-..::;^ * ''' ;"": < '
of a parameter that indicates the treatment system is operating properly.
Comment- One commenter (|V-|E)'K)9> reapested clarifle^tioij.pntlie terrr^^conjitxuwsly"
~T^ .-i*.:;'^ i'~>L"* :*l; >; 'fisixiijt;'' ir* Y <:%;}* ^^fc^.f'L^ffc-fc. '*& s^.'-avrV''^ \"^-!> . xv-'-;- .
with respect tO-teftBitoringmp tienzeitie ©Mcenttatipn ojfjfjhe waste: streajj^exitiijg treatment, as
... f\. ^YUJ^H; Tg^, ">;!-. ;.' h'J'd% \~ £-' !$*'<;' '-'N*'^ - '';V^ "*'* '" '"' ' "; T''~ ' *:'.-l~ : ' : ' 'lj " '- ! '
found in tabfc 2 of ^CER: sobpart 63, subB^JSXK;. SpBclficaUy, the comjjie'nter asksi if tJhp.w
Jii^Vti'-'^'r: :^.''₯. >--|m'f^: '- ' |:'|(':"'";' '.' 4':' ^'-^^:- 'M-1. "':"! ^' '
"continuously^'" is ^^'defiDijed. as a{ least one ^f .^intevffi' |5 minutes.
,:. .:--:3,-i^4''--' W'.'^. '''. '' '*>-''^ll-!v "?" ii-;rr ^^1^'-;- ' -;!
Res|?prise: {The,BPA has deterfni|^l|;fhat itjs n,ot appropriate to require Continuous
V "W::j)r v't' f^f.1;'"-'-'-:. '-: ; "^r'1''^'"''' '' '' ' "; ' ' ' ' '''' ' "' /"
monitoring qf-b:eni;ene or HAP concentration. ijh the. waste stream exiting a treatment unit,
' ''.^rHK-i* -V :vfi^4 : ' - i- : Ni!i*^i -«;'. - -: * ' ' '-';i -
therefore, it is aot necessary to farther d^fj|e ''bontinuously".
2.3.5 Certifications for Off-Site Waste; Treatment
. -.: - - -i _ ;>'' c::?t\f.'" ;'
' Comment: In 40 CFR § 63. 10&6|b), 'facilities that accept waste subject to control
^:.\' ;:, !, i i K? '''!:> " .,'
requkements of the ethylene produption;NESHAP are requked to certify to EPA and the
; : a. '': '-: > , *'-"' ; r ;-.. :.:
generator of the waste that the waste will be treated in accordance with the ethylene production
NESHAP. One commenter (IV-D-06) stated that the certification is unnecessary and should be
deleted because the off-site facilities will be subject to 40 CFR part 63 subpart DD, a MACT
standard for off-site. waste and recovery operations. Another commenter (TV-D-02) noted that
.: '"'K ] . ' : ' -t : " ' ' .,/;.
any receiver of an ethylene waste. stream containing 10 ppm benzene is subject to the BWON
requirements through §61.342(f) and that any receiver waste with greater than 0.5 ppm benzene
waste is covered by R.CRA. The commenter stated that receivers of non-benzene waste streams
have HAP control requkements through RCRA, other MACT rules, the POTW MACT, or the
Offsite Waste MACT. The commenter predicted that recyclers and other receivers will stop
accepting ethylene wastes because 1) ethylene wastes are generally a small. portion of the waste
sent to offsite treatment facilities, 2) the proposed requkements go, well beyond the BWON and
MACT requkements to which these facilities are akeady subject, and 3) the proposed
requkements are difficult to understand. The commenter stated that this will result in more, waste
streams being sent to incineration or major investments by the ethylene industry to install
treatment systems. The commenter noted that the negative envkonmental impacts and/or the
cost of new waste treatment systems were not considered by the EPA. The commenter suggested
30
-------
deleting the' proposed §63.1096 and relying on listing notice arid compliance' mechanisms' of the
BWON.
Response: The EPA has retained the certification requirements originally included in the
proposed NBSftAPfoVc^sitl; waste transfer, Which are basedro
The EPA contends that tfeesjsprovlJionsare^bt significantly different
BWON in that both requjf<§ the'tfeceiyer of off-site waste to comply with the provisio&s oi the
NESHAP- 'In enforcing this BWONT, EPA has experienced problems in hofeirig' wasteir^ceivets
responsible for the waste' oftce'it is" sent off-site. The certification requirements Wifi iaifefess .this
problem and help ensure that 'the' receiver of the waste will comply with the elnyletie $JBSHAP.
The provisions ^so addres'S'the actions to be taken if a receiver decides to ncnlonger agcept waste
transfers.
The ethylene waste requirements have been revised since proposal and are, with a few
, exceptions, the same as those of the BWON. This addresses the cornmenter's concern regarding
recyclers and receivers not accepting waste because the requirements are more stringent than the
BWON or difficult to understand.
2.3.6 Waste Definitions
Comment: One commenter (IV-D-04) suggested that EPA may have created confusion by
including direct-contact'cooling water in the definition of "process wastewater" and by not
explaining when a material that is recycled is to be treated as a process material, rather than a
waste material. The commenter expressed concern that the definition of "process wastewater"
could be interpreted to require treatment of direct-contact cooling water even when it is recycled.
The commenter noted that the BWON, the basis for the waste provisions, does not include direct-
contact cooling water in the definition of "process wastewater". The commenter stated that, in
the BWON as originally promulgated and as amended, certain recycled materials are not subject
to regulation and the determination of whether they are subject is made based on whether the
material ever leaves the production process (55 FR 8292 and 58 FR 3072, 3076-7.) The
commenter concluded that, if direct-contact cooling water never leaves the process, it is not
\
subject to the BWON. The commenter suggested that EPA should, in the preambleto the final
rule, clarify that materials that are recycled and never leave the process are to be. treated as
31
-------
process materials. The cpmmente.r also suggested .removing, the reference, to direclf-cpn|act.
cooling water from the definition of "process wastewater" The conto^ter recommended that,, ifr
V , _
EPA is concerned that proposed controls .fpf prQ^ss.equipmeht^are inadequate tp address
ftff'£|^..^!.* ;" '" > ^^i >.v ^ " : ^ ". '- " ::'-
cooling w^tfJF.'iHpw&yerj, El*A contends, t6at,^ased ^n the definition ofjprocess waste^ater,
. r>~- ' J4a,J^''i: . >^|i-j.i '.'. '- . i . i cl . - - ''*" .;P'| ' ' " " '. ' ' ,
dkec|:co|It|af definition Pf "waste managerrient unit." The
commenter stated that, based on the definition pf waste, any material, regardless pf HAP
concentration would have to be considered under the revised BWON. Additionally, the
commenter. stated,, any stream containing greater than 10 ppm HAP. would have to be-in
compliance with control, treatment, recordkeeping, and reporting requirements even if the stream
is being recycled and it contains minimal amounts of HAP. The commenter noted that HON
triggers for control and treatment are 1,000 ppm HAP at 10. liters per minute and. 10,000 ppm
HAP at any flow rate. The commenter suggested that, if the revised BWON requirements are
retained, EPA should reconsider the definition of waste.
Response: Based on the MACT floor analysis, the four of the five best perfprrning
sources are treating and managing waste streams to comply with the BWON. Therefore, they are
32
-------
treating aH vva^te .streatos liikt are subject Mthe'BWON, both'According to the BWON definition
of waste and the BWON control applicability cutoffs. Only one of the best performing facilities is
subject td fife HON.' Therefore, EPA contends that it is appropriate to base the waste '
req!^:&ents dsrf ^ stream^cbritroi^plifebllity.. cutoffs rather
thatt tKo?se: of tfe;BfeK:. TJ^Ep&wttuld like to point out that the BWOTf'io.es entourage
reeyclifig in that waste*'stife:|rns'tfedt Tare recycled to a process are notiis$bje:ct to. cdttrol
< .«" ' ti '.-,.. . :.':> ' i *. ; . . "' >- -
requirements, ft'is imporfant to hold that the EPA'has decided to allow berfzene to be used as a
surrogkte, therefore the 10 ppiriv c'ontrol applicability cutoff is no Ititfger bein^ expressed in terms
of HAP: s ; ;
* i .
2.3.7 HON Point of Determination Concept
Comment: One commenter (IV-D-09) strongly recommended that EPA consider allowing
facilities to comply with the process wastewater provisions of the HON mainly based on the
advantages of the "point of determination." (POD) concept of the HON over the "point of
generation" (POG) concept of the BWON. The commenter explained that the POG concept has
facilities determine applicability of requirements at the location where the waste stream exits the
process unit component or storage tank prior to handling or treatment in an operation that is not
an integral part of the production process" whereas the POD concept requires' applicability
determinations "where the process wastewater exits the chemical manufacturing process unit."
The commenter noted that proposed 40 CFR part 63 subpart YY applies the POD concept. The
commenter stated that the POG is more difficult to determine due to different understandings of
what is "integral" to a prociess and noted sources have been instructed to contact EPA for case-
by-case determinations. Therefore, the commenter concluded using a POD approach would
reduce EPA's burden. The commenter also noted that EPA adopted a POD approach over a
POG approach when the HON wastewater provisions were revised in January, 1997. The
commenter quoted an excerpt from the preamble for the 1997 final rule (62 FR 2726) in which it
is stated that "The EPA's intent in developing the POD approach was to have a decision criterion
that is replicable and clearly specifies the location for evaluation of a wastewater stream for the
purposes of control."
33
-------
The cornmenter further argued that/acilities with ar> EMPTJ tliat also have HON. affected
units are already familiar withahe HON requirements. The cornmenter referfed-to; the preamble ,
r ', '-^ 'T ''*.' *i h * " - ' ' ' t- "- ' *"'-' "*"i ^ I " T ''.''' '. r A, ~. '-,' ~ ' Tr- "'" ,'-"* ' .'''""''If,"'' /i. i ' '.'- , "r1, , J'j ., ', « '
fot the propf^dJ^HAP^§;FR;7gl:2^}.in w^;|PA st^tHait. th&.toyde^p^prnpliaiice is
red^ced^e^^
w^lfpi^
facilities .aBs^ld^c^mplyii^ with, f^;Cpjri^^r|ter' c^nfi^e^; tha_t they;havets^<
, , .
HON units, anoare ateadv farpiliai: wftft^te JJp^:wast^atei: ptovisipBs artd^ne^ded.thsLt they
' i,\.'.-" .1 "!>;. ' §j'^jt 'r-'i'lft'.X " 'i ; .'"^. ' "v !" ' ','$ y& .^'r '*!'., '' !?'-' :i*-f'"!^- '> - ' -t^ f .' '' ' ' ":'' '.'.
would have tes^'burden'CDmplying witii tti^'HC^ provisions than those -of the revisie,d BWpN.
K'V. r'-:l .^rf^'i't'j, F* ?' JT ; ' ;' ^Vrit'rT.'. ;'Jir^;.'r> fi--M.-:'. ..... =, i -. »;.:' i. .:--;;
Response::As discussed in the previdus response, four out of five of the best peifoitriing
facilities are complying with the BWON Waste requirements. white only onp.is complying withj
those of the HONL Therefore, EPA has determined that the requirements .Of the BWON most
accurately reflect the MACT floor and; are ;the appropriate basis for the.ethylene NESHAP waste
requirements. While many ethylene production units are co-located .with HON units andimay
. - : -.' _-[:.,]-? \; -' - '._.-.-:. .- .-. .
share waste treatment systems with them, evien more ethylene production units are akeady subject
to the requirements of the BWON. Basing the ethylene production NESHAP waste requirements
on the BWON introduces less burden than basing, the requirements on those of the HON.
2.3.8 Waste. Reporting Requirements
Comment: One commenter (rV-D-02) expressed concerns about the proposed annual
reporting requirement; for treated or controlled waste streams. The commenter contended that the
annual detailed waste stream inventory requirement of the BWON was already wasteful and
unnecessary and is greatly expanded through table 2 of 40 CFR part^S subpart XX. The
commenter stated that the EPA has not demonstrated any environrnenjtal benefit for the increased
burden of this expansion in paperwork. The commenter recommended that the annual reporting
requirement for waste streams treated or controlled by the requirements, of the rule be eliminated.
Response: The EPA assumes that the expanded reporting requirements the commenter
refers to are those that resulted from basing the requirements on^HAP rather than benzene. For
example,, the BWON requkes the benzene content of each waste stream to be reported. In the
proposed NESHAP, facilities were required to report the concentration of HAP in each waste
stream. Similarly, the BWON requires the benzene concentration of waste streams exiting
34
-------
treatment systems to be reported while the proposed NESHAP required the concentration of
HAP. As discussed in section 2.3.3, EPA has determine that benzene' is an acceptable surrogate
for HAP both for determining the applicability of control requirements and the efficiency and
effectiveness; Of, control, systems. Therefore, it in no longer necessary to require HAP
concentrations to-be repo?ie'd; ,1The reporting requirements of the ethylene, prOdu^ti^to NESHAP ;
for waste are now'the safne as these for the BWON, with one exception.' For w&st'e-streafnVthat
contain 1,3-butadIene'anfLda ndt contain benzene, facilities must report the mfoitnation normally"
required by the BWON for benzene for butadiene.
2.4 SELECTION OF MACT FOR HEAT EXCHANGE SYSTEMS
Comment; Two commenters (IV-D-05, IV-D-06) asserted that the proposed requirements
for heat exchange systems are not adequately supported. One commenter (IV-D-06) contended
that little justification for the requirements is provided in the preamble and no additional
information is supplied in supporting documentation found in the docket. The commenter claimed
that the lack of information prevented them from providing meaningful comments. The
commenter requested that EPA publish additional supporting information on heat exchange
system requirements and allow additional time for public comment. Two. commenters (IV-D-05,
IV-D-06) disagreed with EPA's determination of the MACT floor, stating that EPA failed to take
into account the size of the unit» the size of heat exchangers, circulation rates, heat exchanger
design, and the design of the unit (for example, whether installed spares are present). The
commenters disagreed with EPA's assertion that the heat exchanger program will impose no
additional costs. The eommenters stated that required monitoring and shutdowns will result in
significant costs. Detailed comments are provided in sections 2.4.2 and 2.4.3.
Response: The proposed requirements for heat exchange system have been revised to
allow for flexibility in the selection of heat exchanger monitoring location, as well a$ a longer time
for repair. The comments on monitoring location, frequency, and repair requirements are
discussed in more detail in sections 2.4.1 through 2.4.3. The EPA would also like to note that a
discussion on selection of MACT level of control for heat exchangers is provided in the preamble.
2.4.1 Monitoring Frequency
35
-------
Comment: One cpmmenter (IV-D-06) generally agreed with EPA's finding that more
frequent mpnitQpngKpf;copling-.water probably decreases.the amount of rtime altafc goes.
undetected. Qnjj'cpmmenter (ly-D-Op) hotedyrKvweyer,, that: among the: five beSt petforrning
sources, the one^hyjjps urfit ^hat co^dugfe^e^y rnonjt.pring (Exxpn^obijj T^xasj); reported,
higher HAP eri»s|iic>nsfrpr^its;GPpJkg:,wate^y^tiBrri,thaH the th£eeTacfl?,ti^ that ci^n^pf ^fttbtithly
mpnitpringi The^ommenter as^erted^hatHhjarpl^tes^BPA^s justification for requiring new:
sources to monitor weekly. One cprnmentBr .(I₯-D:-02^^greed- with the:EPA:that! mpnthty.,
monitoring for existing sources and weekly monitpring f©r new sources is representative of the-:
MACT floor.
Response: The EPA wpuld like to acknowledge :the commenters' support of the
monitoring frequency for heat exchange systems at new and existing sources., The EPA contends
that more frequent monitoring will generally result in lower HAP emissions from heat, exchange
systems because, once a leak is detected, actions are taken to repair the leak or take the.leaking .
equipment out of service, thereby minimizing emissions, from copUng water. Survey responses
generally supported this relationship between HAP emissions and how often a facility monitors
cooling water for the presence of compounds that would indicate a leak. While it is true .that the
ExxonMobil facility reported higher HAP emissions than some facilities with less frequent
monitoring programs, this appears to be an outlier to the survey data set in. general. It is also
important to note that variability in the reported data is not altogether unexpected, as there are
other factors (for example, size of heat exchange system or quantity of process material leaked
into cooling water) that could affect the quantity of emissions reported from a facility.
Comment: One commenter (IV-D-02) expressed concern that there are no allowances for
i.
reduced monitoring frequency fpr sustained good performance in the proposed regulation, which
is allowed in the heat exchange monitoring requirements of the HON. As a recommendation, the
cpmmenter suggested that the HON heat exchanger requirements be adopted in their entirety.
Another commenter (P/-D-Q5) questioned EPA's commitment to consistency because skip
periods for good performance are not provided fpr heat exchangers but are included in traditional
leak detection and repair programs.
36
-------
se: The:EPA generally disagrees with the; cofnmenter's assertion that the HON
allows for reduced heat exchange system monitoring for sustained good performance. The HON
. ,: , ' \ ' !jj;'- V ',' ... '.. " ''' I
allows quarterly monitoring after the first six months, regardless 6f performance. However, the
'. ."'"- . .. . '.,.; , : :*. '.'£ tj. , > . .V' ; i.v
EPA agfeesthat Sh'aljowarie^ should be available for sustained' gppd pprforrn^nce, Stock
aHowancersir e p'r6vide.d by the HON and etjiylene NESHAP for otfie^tj^e5;o'r' equipment' leak
.. . . , .. ' . . ' i '. ! " .''>[; . ';, ', 'M' " ' £": V ', ''i- .,'''*.-,
detection ^rid' repair programs. "We believe that providing an irieenfive ii} the rufe for reduced
monitoring will encourage facilities to undertake measures to diagnose tne causes of leaks. and
reduces the frequency of thefr occurrence. Accordingly, the final ^te" includes: a'provisiori for
- - - - - " . ' ' . " i1 ; . ' , : .J ' , . ' !
reduced monitoring for units:with sustained good performance in preventing leaks. An initial
monthly monitoring frequency for six months followed by a quarterly rriortitoring frequency,
provided there are no leaks, has been included in the heat exchange requirements in §63.1086. If
a leak is detected, the cooling water will be monitored monthly for six months following the
repair. Quarterly monitoring: may then resume if no leaks have been detected. This performance-
based approach is both consistent with the floor and reduces significant burdens imposed on
cooling systems that are also covered by the HON.
Comment: One corrimenter (IV-D-02) argued that the proposed regulation does not
provide flexibility to use continuous on-line monitoring. As a result, facilities currently using or
planning to install a continuous monitor wiH be penalized by having to perform monthly
monitoring in addition to the continuous measurements. The commenter noted that the HON heat
exchanger requirements provide flexibility that allows the use of continuous monitoring systems,
and recommended that the HON requirements be adopted instead of the proposed requirements.
Response: The EPA generally agrees with the comment, and has included provisions in
§63. 1086(c) .of subpart XX that allow for the monitoring of surrogate parameters that would
indicate a leak. This language, like that of the HON, allows for the use of continuous monitoring
systems, provided that certain monitoring and recordkeeping requirements are met
(§63.1086(c)(l)-(c)(3)).
Comment: One .commenter (IV-D-09) noted that previous regulatory determinations
included a less frequent monitoring schedule than the proposed regulation. The commenter
37
-------
suggested that the EPA consider using the approach of the BON* which requires mpnthly
monitoring' for the first six mpnttjs and qiiarterly thereafter^
jfoesponse: Tfee EP4 recognizes ^a5 the HON: alQWS; for a less frequent monitoring
^S^sixaM*^'^
'^^^|l- iVS^S^fe-^^^P^iiifc-V-'v^^-^'-^'"^ .HStv^ft- 1^/i:^A .& »
..- r . /, -; - .... , -
The sufv^yidafeSaili&red I0r the etMe^Apdiaction MMfiAP^iiTidicates that f he best perfbrrrang
Crifc^rfer-*#$ ^I'^^^'^^V'''^!^^^-:^^..;;!.;.1',^.-.- .?;'-.;-.:. M.;p .,:?; . *,-.;, ,
«|0nit.or;fieat e^d^bgfjj^ EP^.has
'"'--!" » " ^ ''-*'-ifet;^*- - ''i'ijt^T; ,;!:... r;S, '-fj i;^,!:', -1 hi,;..; :-.. '.;;' , t; ' . \. ' :-
iicoiis regulatory ddter^rimatlpns, &lid anaj^ed data prpvide$M;sapflprt,-pf the
^y -\\,\'- -f^*"'- ;. «1 !-..'!'-:"-.^: .^ivj: ?'$&. :' ^-;.'',. '""I;-1' \>'y"-
^ Ija^ determined t>bat a p^rft«7nai|ce-basedrflprnt<3ring freduency is appropriate. As
H '.Jiiiti '- '"':- ' K ' ' "i;- 'i:' '' '":'*.'' -i l 3, '('^':' :" i";"' '"' ' . '_
discussed in;iatl ear|ipr response, the heat exchang^ moi^toring requirements have been revised, to
| 'i;:;| ,jj?r-' -: "-1 .,';'-".- ^i--;::.:! «.*;. . : : ! .\ '- ' -.
allow for a reduced'monitoring frequency Uipon'dejmpnskatipn of sustained. goo4 perforrnance.
1 }|' ' . i1 fj '- '!;£ ' ' : !"' i " ' ! ' ">':I-L> " ' * *-' ' ' ' ' ' ' " '' '' '
The perfortnarice-based requirements have-been dfetermjped to be representative, of the MACT
tK ::'.- , ,.; .- !'.:' '_ ^ij -ii'.'' ' " i;i. ' '''-;'''"' '"'; "
floor and prev^de tjie additional benefit of Deducing additional burdens to owners and operators of
facilities with cooling systems which are shared between HON and ethylene production units.
The perfbnnarice-based monitoring frequency for. heat exchangers and heat exchange systems also
provides some, consistency in approach, to the requirements for equipment leaks.
2.4.2 Monitoring Location.
Comment: One commenter (IV-D-06) expressed several concerns regarding the
Fi ' . '. ''.-:'' ': ..''- '
requirement in, the proposed NESHAP to sample the cooling water at the inlet and outlet of each
heat exchanger. While the commenter agreed that circulation rates through ethylene
manufacturing process units are generally higher than those through synthetic-organic chemical
T . ., . '"'.' *. ' ' .' ""' i -
manufacturing (SOGMI). units (the current heat exchange system requirements, are based on
requirements for SOCMI units), the commenter did not .agree that the higher circulation rate
provides adequate justification for requiring sampling at the inlet and outlet of each heat
i "' : ' '.-'.''.
exchanger. The commenter stated that many heat exchangers are relatively small and have
relatively low cooling water flow rates and that sampling around each of these exchangers is not
necessary to prevent leaks from not being detected. Another eommenter (IV-D-03) agreed that
although the total circulation rate may be greater, the flows to individual heat exchangers in
EMPUs are commonly much less than total flows through SOCMI units. The commenter
38
-------
provided that the proposed NESHAP would apply to exchangers with flowiates as low as 2000.
gallons per minute (gpm) and a 1 ppm leak from such an exchanger would result in an emission
rate of 1 pound .per hour (Ib7hr). The cornmenter asserted that requiring a shutdown for this level
of emissioins is inc'dnsiste&t with Other MAGT;standards and is wMjbiit'rJasjs, ' ' '*
One ^ornrndntet (IV-D-06) pointed out that EFA's analysisMoes,not consider that the heat
exchange systems are branched1 systems and that a fraction of the tooling water goes: through each
branchv- Two comnienters (IV-D-09, IV-D-06) suggested requiring testing1 at each branch of heat
exchange system rather than at each heat exchanger. One cornrnehter (IV-b-09); noted that in
past regulatory determinations (specifically, the HON), the EPA has allowed sources to monitor
vthe heat exchange system as an option to each individual heat exchanger, The cdmmenter
suggested using the following language in subpart XX, §63:lQ86f"You must collect samples at
the entrance and exit of each heat exchange system or at locations where the cooling water enters
and exits each heat exchanger or any combination of heat exchangers." The cbmmenter cited
compliance costs, no value added with the proposed requirement, and previous regulatory
decisions as reasons for this recommendation; Another cornmenter (IV-D-06) asserted that
testing each branch rather than each heat exchanger in each branch is a viable option that EPA did
not acknowledge or assess. Furthermore, the cornmenter faulted EPA for not considering
alternative analytical methods or other approaches that could compensate for the1 relatively high
cooling water flow rates in ethylene manufacturing process units. Another cornmenter (IV-D-02)
suggested adopting the. HON heat exchanger requirements in their entirety in lieu of the proposed
requirements. The commenter recognized, however, that the EPA is apparently concerned about
the size of ethylene cboling systems, and suggested a requirement for inlet and outlet monitoring
for every 100;000 gprri of cooling water circulation flow rate if necessary.
Response: The EPA generally agrees with the comment, and has revised the requirements
:of §63.1086(a) so that a facility may decide the location for sampling and the test method
(provided it is approved by the Administrator) to be used as long as a leak of 3.06 kilograms per
hour (kg/hr) will be detected. The example given in the final rule shows how, for a test with a
sensitivity of 1 ppmv total HAP, each heat exchange system needs to be defined so that the
-cooling water flowrate is 51,031 liters per minute or less in order to detect a leak of 3.06 kg/hr.
39
-------
As discussed in the promulgation preamble, the 3,06 kg/hr^S ppund per. hpur (lb/h/)) value.
was calculated basedfjon,the assumption, that fc^ in ^n adequate,
lev^lof Jtea^, dite^tian and.that the circulation rafe pjfcpoling: water topugh^aft e
' ~ - '' ." f.' ' " ? " ' ' '
produciiojj unit'isteiglit ^^.^^a^i^^^^^^l^^^
assumptions, the 4; ppnj;|eak defjnitjpn ofihffljQipI, an4tr^ay^age;pf^^
for .ethylene viwfeto.Sflryey resp9nse§^it..wa&^tiriaat?d''tha;t,a 3.;06^g/hr:tea^;;j:ai:&LW;0;uld be
- '" " It **",' ' '' -'.-'' ';-'*': '-".":' ' " " '';"' '
de|:ects4, at a-BQH tt^it. :IJhe ethylene filler jsfended, to regujre tfee same^^M*^ detection.
The EPA .realizes; that there rnay^ be certain, situations where a facility would jather-monitor aa .
individual heat, exchanger (such as the t^stiqg ^required following a heat exchanger leak repair),
and has allowe^for this ia§63.1086(b).
Comment: Two cpmmenters (IV-D-04, TV-D-Q6) contended that sampling at the inlet and
outlet of every heat exchanger does not reflect|the practices employed by the best performing
ethylene production facilities. One commenter: (IVrD-O^) gave, as an example, a permit condition
for a facility in Texas in which the cooling water is.monitored monthly at the cooling tower. One
commenler (IV-D-06) determined that none of the five best performing facilities .are required to
sample.around individual heat exchangers. The commenter provided excerpts from permits to
support this determination. Two commenters (rVrD-02, IV-D-06) stated that,.because the five
best performing facilities are not sampling around each heat exchanger, it is an aboye.-the-flpor
option. One commenter (IV-D-06) noted that the EPA did not perform an analysis to support the
sampling requirement as an above-the-floor option,
One commenter estimated that the (IV-P-02) requirement to sample around each heat
exchanger results in a more stringent leak detection level than currently required at the, MACT.
floor facilities. The commenter estimated the smallest heat exchange system in an ethylene plant
to have a circulating flow rate of around 50,000 gallons per minute (gpm). According to the
commenter, if the assumption is made that the minimum flow rate of a MACT floor heat exchange
system is 10,000 gpm, than the current proposal imposes at least a 1000-fold decrease in
allowable leakage rates, when applied to an individual 10 gpm sample cooler. The commenter
noted that sample cooler emissions at 1 ppnvvtould.equate to approximately 30 pounds per year
of leakage. The commenter argued that the.potential shutdown,of an entire E^PU for small leaks
40
-------
is uTidasbtiableilnd thatEPAvshouldConsider alternative approaches tftat may pass the
reasonableness test required in § 112(d)(2:) of the Aet for above-the-floor-.actions.
Response: Intfie MACT floor analysis, it was determined that the best performing
facifities .aTe'mb'ttitoring heat exchanj^p systems monthly to detect leaks, ^fie location Qf sampling
was riot considered in ihe;'h4at exchange system floor determuiati6n> The EPA originally'
determined thai sampBngiat the inlet and outlet of each heat exchanger woiild be required to
J , . , , -L,
detect a leak'dfte to the relatively high circulation rate of cooling water. Eterwbver, b^sed on
additional cornments, data from survey responses, and previous ruleinakihgs, the El*A determined
that selecting a monitoring location and test method so that a leak of 3.06 kg/hr can be detected
wiH also adequately detect leaks (see previous response). This flexibility in selectiori'of
monitoring location, along with the revised repair provisions (see section 2.4.3) should address
the commenter's concerns regarding the possibility of ethylene production unit shutdowns caused
by very small leaks.
Comment: Two commenters (IV-D-02, IV-D-06) disagreed with EPA's finding that the
heat exchange system requirements would have no cost impact on the industry. The commenters
provided their own cost estimates based on the following assumptions: ethylene units subject to
the proposed NESHAP have an average of 25 heat exchangers; 1800 samples would be taken per
year per unit (2i5 exchangers x 6 samples per exchanger per month x 12 months per year = 1800);
$225/sample for a Method 624 test; and $425/sample for a Method 625 test. One commenter
(IV-D-02), using these assumptions and including one hour of labor per exchanger per month at
$35/hr and amortized costs for sample systems, ladders, platforms, and other associated costs
estimated the cost for their six. facilities to be greater than $8 million per year cost to their
Company alone. This commenter noted that even the sources comprising the floor would incur
additional costs under the proposed requirements, and that the HON and Texas cooling water
programs already achieve the benefits claimed for this portion of the proposal at a fraction of the
\
cost.
One commenter (IV-D-06) used the same basic assumptions and provided additional
information to support the assumptions in appendices to their comments. The commenter
estimated the cost-for one unit to be $1,170,000 per year and the cost for the industry (based on
41
-------
25 units) to be $29.2 millionipsrjear,. the cornmenter contended. %& affected spurges-would,
incur costs in addition; to the anajytical,costs. The connnenterstatedvthat many facilities w,$;
to add or retrofit,sample 'points;, (whieh may. require th^ consti^ptioja of-J^dde^s, p|^tforp^:and
other %c^ii^|;|^:a|i^^^t^|ps| of j$14PC^J?ea: sa^iple,j?oi^|.,.^^^i^jj^.5$at^ftli^t. %
conserva^e^^o^rsss4^nptic>art^ lOsajinpjte pointe;will nee-d, tp tje.a^ded'xa-ja^Apiijital recover
factor o£,(?£&^i]|i&;i^ industry of .$jl;.i millipn. ;1j^,eo,i^e|iter also
provided an: esti|i|ite of SO-SjUiiffipn^er, year JjOf the industry-for, the febor cost&,a$spc(at;ed; v^ith
collecting ,and,|£a|isporfing san^tes and niaijfainiiig records. Ttje cpRBtuenteT-npteti that theff^cost
estimates add.up to almost $3-3 million per, year. By conservatively assuming that all HAP
-.'-' -.,..,- ' . . ,-. i
emissions from heat exchange, systems' would eliminated, the commenter estimated..the-cost
effectiveness of sampling around each heat exchanger to be $1,400,000 per ton .of HAl* reduced.
Another cprnmenter (IV-D-04) stated that the proposed cpoling water monitoring
requirements would likely result in an additional $1.5 to 2.0 million per year in analytical casts for
their company, with apparently no environmental benefit. The commenter provided that the costs
dp not include the cost of installing safe access to heat exchange systems, and sampling
connections, labor costs associated with sampling, or premature shutdowns resulting from the 15-
day repair period- ,
One commenter (TV-D-09) contended that the EPA did not fully consider the costs
impacts of testing at the inlet and outlet of each heat exchanger, and that, the EPA^ only addresses
sampling frequency, and not location, in its discussion pf the MACT f|opr. The commenter
estimated that it would cost $60,000 per year to monitor an EMPU, which is, well above the
$7,600 per year per existing.facility average cost proposed by the EPA. The commenter noted
that currently one of their units must sample each at the inlet and .outlet of the cooling tower on a
weekly basis as part of a permit condition and, at $50 per sample, the direct cost to monitor the
unit is $5,200 per year. Using the same cost assumptions, the commenter asserted that for the
same unit (with 15 heat exchangers) it would cost $54,000 per year to mpnitor as a result of the
NESHAP, and the facility would still have the $5,200 per year cost to monitor for permit
conditions. As a result, the commenter expects a 91% increase in direct monitoring costs for a
minimal reduction in HAP emissions because the unit is already being monitored. The .commenter
42
-------
suspected 'that other EMPU in Texas also Have existing heat exchange system ttibniforihg
..requirement s' and1 'requiririg additional monitoring is over-bufdensorne. The cpraSienter asserted
'..'>. '.VS. ,.1 " f- '. . '.:'''',,'' ''.', [ ' '. ' :,', '
that aehievyig'a^QlTMg/^ feduetidn in emissions, representing 2.2 percent of the emission
feduetiW^tBBMa%li'td i|reprop6se^lteSf^, wbuld come at ^g^.^'craldsult in
mo '
ias determined the jMACT floor is represented by monitoring the heat
v .. ". :- .>!.< t ., , , ' '»." -. .')* " '' $L '' '.??.,'
exchan|e systea^tnbTithl^ (for easting sources) or weekly (for new sources) 'to detect a.leak. In
"' 'i-' \ ' ' I* 4- ''.'""; ''l - , i'ri . i " ' '- ' > i ' h ' 'ri
response td'OTrrteentsr^ve'added. art option to allow a facility to either monitor a heat exchange
system such triatta1leai call ie detected by sampling at a location and using' a test method that
~- ' ' ' . -i '
would detect a leak of 3. Op fcg/hr or by sampling at the inlet and outlet of each heat exchanger.
This change means that faciKtids would generally riot need to sample at the inlet and outlet of each
heat exchanger. In addition, there*"are simplifying assumptions provided in the final rule that allow
facilities the flexibility to either monitor at a single upstream point or assume ah entrance
concentration of zero (§63. 1086(d)). The final rule also has been 'revised to alow for reduced
'. . i
monitoring frequencies for existing facilities that have demonstrated good performance. We have
determined that the final N^SHAP reflects the MACT floor and, therefore, the EPA is not
required to consider the cost per ton of emission reduction in determining the appropriate level of
control. Considering the changes made to the heat exchanger monitoring requirements, the
simplyfiying assumptions that can be used for inlet concentrations, and the reduced frequency for
good performance, and that all but one of the facilities in the industry are already monitoring
cooling water, we do not believe that there will be any significant cost impacts imposed by the
rule. As a result, the cost impact analysis for cooling water monitoring presented in the proposal
preamble remains valid.
2.4.3 Repair Requirements
Comment: Six commenters (IV-D-01, IV-D-02, IV-D-04, W-D-05, W-D-06, IV-D-09)
disagreed with EPA's decision to require repair of leaking heat exchange systems within 15 days
of detection of the leak. Three commenters (IV-D-01, IV-D-02, IV-D-06) argued that the EPA's
analysis did not consider the repair periods employed by the five best performing facilities.
According to the commenters, each of the five' best performing facilities is required to repair leaks
43
-------
as soon as practicable, but no later than the next schedule shutdjown. Another cqrrjmenter (I.Y-p-
04) stated that a permit for one of their facilitiesreads,. "Faulty-ewipjnenc shaft be repaired at the ,
.%;-,: i ", "'Up "ylx "-.' w'''>'?!v Vf -.-{" -' ' H"'s;i"^': ;::$;-i'*;;'! v'V%;^:;;vx'.; i.- r^- ?-'-
earliest opportunity, &ut no later tfyan the next S!^hed^ed^shllt4a;^n," and re^oppnerids ,
'''-l'?'&-.'!'""''- ^i '.
;oi
coiflrfteiatta'fl^|5*-06) explained 'that the: time-required] to r^ajfca fe$k ci^rM^'vsries according
:/** ^uj}-^i.^ s:f^t^ :* :'?>. 'V;^-:!?. ' "J:; !^'--^^^^^-i.?,v*:. v^^i^.i
to cifcarrjStiM^S|.-i Thb coitmienter expfelheB the sorne,excha!ngfni;are ri&piaire^Mffi^biecatise
1 ''7^''.1li;U^-j"''? V1 f" "' ' \"^?l.':'-'-i;'-1i ' ?"* '-1'-' ; ''Jf-S'«?'-& '« ,-'f . - i;-.^rrtf«--J.-;...'i.-^i;
they are riot -ift^al |o operation or c^te, replaced by a spar%«^d,a unit shmtdj^nag not
\ i1-,;^;1 s. !';ii|i;!"-i ;';|i':' ^'T:.: -.'../ I'T-tl'v'1 ^.- .-* - .''''. -ft. i'=P';''^' v :'*"''' '*'*^«-:'>-,f. ' '
requiredfor repair. The commenter cpLntinted?that sonietii^es.excrhangers^ejTepajred quickly
i.S:'.:i ".ij,i-;"!!Jv: - -'':.' " -,. ; ^-" :'WH ^' -J-. :' ^ \" '^t:'<'-:' "J" -'"'u;''"': ' -1''-1;.
because a leak i* significant enough to tlau^e a jsafety hazard, or unacceptable enyjbronrnental
impacts-or to ensure compliance with'escisting permit conditions: Two cornrnenters (TV-D-Ol,
I .. -I?" ' ' ,1..'.' i' :':'!..* i' ';.; J^ . ' ' ' : 3 ' <*" ! : '
IV-D-06) asserted that because the fiye^best peiforrning facilities.are not required to repair
leaking heat exchange systems within 1;5 days, this is an aboveT the-floor option. One commenter
i ; |;,ji -. - -. ; '.; i' .': i ' i . '' ' ''.''.. .* ;i-- ' '
(TV-D-02) agrreed that this is an above- the- floor option and that the 4.34 tons per year HAP
reduction claimed by the EPA does not adequately justify going beyond the MACT floor.
Response; The EPA agrees that it'did pot consider the repair practices of the five best
performing-facilities when it selected the 15 day repair time, mainly because the EPA did not have
that particular, information at the time. As; a result, the 15 day repair time was selected for
consistency with the equipment leak repair time. The EPA has reviewed the information provided
in the comments and the requirements for heat exchange system repair fourfd in the HON. The
EPA has determined that the 45 day repair time for heat exchange systems is more consistent with
what the five best performing facilities are doing. As a result, EPA has revised the requirements
of §63.1087(a) to allow 45 days for leak repair. This is consistent wittf the,time allowed for
repair in the HON.
Comment: Three cpmmenters (IV-D-02, IV-D-05, IV-D-06) asserted that providing
consistency with the repair period allowed for other leaking components (valves, connectors, and
pump seals) is not an adequate justification for the 15-day repair requirement for heat exchange
systems. The commenters stated that heat exchange systems are.significantly different in design
than other components and are much more difficult to repair. Two cornrnenters (IV-D-02, IV-D-
44
-------
06) noted that valve packing and flange bolts,can often simply "be tightened externally-or, in
extreme cases; can; be externally pumped with a sealant or clamped to repair.*6ne cornmenter
(IV-D-06) explained that the other .components can be repaired quickly with minimum disruption
to surrounding eijf^prrimt d'r ©peration ofthe unit. Two cohimenters:(rv'"|D-02VlV-Di-(J^)
contrasted^he oth'er compdiiehts t'6 heat exchange systems, statingthatthe 'p^-'Way to-repair a
heat exefiahger i$ to shuffddwii the exchanger and isolate it from thte!'ptocess,*ele&i[i [and'operi the
exchanger, te-st to find the leak(s), and make the repair. The cominenfers added, thatreindving an
exchanger from service often requires a unit to be shutdown. One commented (tV-D-02) -noted
that equipment leak rules allow repiair of leaking components that cannot be repaired without a
shutdown totbe deferred:uijtil the next actual shutdown. One commented (IY-D-06) provided
detailed information on the.steps and costs involved in repairing a heat exchanger in an appendix
to the comments.
Response: As discussed in the previous response, EPA did not have information on the
repair periods employed by the best performing facilities prior to proposal of the NESHAP. The
\ . ' ;
15 day repair time was selected to be consistent with the requirements for equipment leak repair.
However, based on new information provided in the comments and a review of the requirements
found in the HON, §63.1087(a) has been revised to allow 45 days for repair. In addition, the
provisions in §63.1088 have been revised to allow delay of repair, if one of the following situations
apply: 1) a shutdown is expected within the next two months; 2) a shutdown for repair would
cause greater emissions than the potential emissions from delaying repair to the next shutdown; or
3) equipment, parts, or personnel necessary for repair are not available (in this case, repair may
delayed a maximum of 120 days). The time allowed for repair and the delay of repair provisions
are consistent with those found in the HON.
Comment: Four commenters (IV-D-01, IV-D-02, IV-D-05, IV-D-06) stated that the
repair provisions will require rapid, unplanned, and inefficient shutdowns of entire EMPUs. The
commenters contended that unplanned shutdowns are extremely costly, disruptive to the
operation of integrated facilities, and have environmental consequences that often outweigh the
benefits of repairing a leak. Three commenters (IV-D-02, IV-D-05, IV-D-06) explained that
unplanned shutdowns can result in material being burned in a flare, resulting in noh-HAP
45
-------
emissions such as NOX, paiticulate, non-HAP VOC, CO, and CQj as well as ng^se and light
which can cause CQflpeiTj in nearby eom|ii)i|4ti;es., One cornmeijtei; (T^.-£)-Q2) ftoted that
instromentblqwdpwn, drainage of;equi|>$ien!t apd,the,faulty heal: exchanger Aviljlesulkin increased
emissions,and \Y,as.te-,geiiierati<3{i^ Furthegimo^ , ,
- y; itv \-^i V v^'-x^1 > ' V ''*' 'is₯cxi?'H7.7^ .'O.^*<;£ >t.,E"^ ri%r..-'M.< :?&, -\f;«-**-~ *ty: y s.-.- . .". fj-V-'^f^--;-'^'^ -'is *
'- -i ~* . :- ;;' .V"' '. . "-' ';,-v£:v '-V ' . .-"-" " '.-- "" "' """" "'' ;':' :
shutdown ipaay.csi^sj|^^^
'. , .:. *.*^ '. '? \iyAhi ^"ib' -^(' ^tjU.; ': * "'^j,'1!11!_ I i>ljl-_"1[^ '" N ^ " ^ / ' r ' '"" - " ' ' \>'- '
shutdown, wtftoKt^adegiiateptapniijg^ tj^a^ig their ;^>fii--^§.pectjve. en^ssjs^, atn^jfripact^
One comHlfenter (rV-E>^06) asserted 'that ffc'teto&pnly on, IMsP' coiild.nesuJt ^dgcisiops that
V.q .' i^^^t '3\' * - "^ t i",^;'-,'^ y- -'^.'v 'J' .a f--f </ j.;^ ''S-;|fl.| '^ ' , '; V- '. - i ' ' ''J*'
minnnized'IMp ^iW^9&s- but are.ultinM^.Jeiss environnientally. benefiGial whea.pt^er
environmentalifactprs are considered. Xwp^coinmenters (IV^D.-02, IV-D-06) stated.tfaai the delay.
;.' ']."'-;; - ''-.' '.'*.!? . - ,, '"I; j;.v ' ,..,...,... .,
of repain p^pvigipns, which allow repairJtpbie Delayed if HAP emissions associated with the repair .
would be greater than emissions shottld^hje.repair be delayed until the next unit shutdown, should,
take into account total environmental irnr)&ct. One commenter (IV-P-01) noted that EPA did not
consider the cost of shutdowns in their Dialysis and suggested that, if the 15-da.y repair
requirement is retained, EPA provide a the justification required by §112(d) of the. Clean Air Act.
Response: The EPA has reviewed the information provided in the comments and the
requirements fjor heat exchange system repair found in the HON. Based on this review, the EPA
i * i. ' r , -| '.'.. -
has revised §63.1087(a) to allow 45 days for repair of heat exchange system leaks. The 45 day
repair time is a reasonable time to plan a shutdown so that non-HAP emission impacts will be,
minimized or eliminated. Additionally, repairs may be delayed in certain situations as provided in
§63.1088 (see previous response for repair delay conditions and requirements), The;revised
requirements are consistent with the time allowed for repair and the delay ;of repair provisions iii
the HON.
Comment: Three commenters (F/-D-01, IV-D-02, IV-D-04) suggested revising the
regulation to require repair "as soon as practical or by the next process unit shutdown." One
commenter (IV-D-04) suggested allowing up to 60 days. One commenter (IV-D-02) suggested a
reporting requirement with a trigger HAP release quantity (for example, 5 tons per year.due to the
leak) could be included to ensure that the enforcing agency is involved with the decision of when
the repair would be made. The commenter stated that, at a minimum, the HON language and
requirements should be used in place of the proposed requirements. Another commenter (IV-D-
46
-------
06) supported using the repair period allowed in the fiOJI (45 days.) The commenter believes
that this is, a reasonable period for making repairs and is justified by the similarities between
ethylene manufacturing' process units and SOCMI units. The commenter stated that shutting
i&Tirut i$'^ctpaUy more difficult than shutting down a7 typical SOCl^ uiiit^because
sizd anAthroughplit orlthyfene manufacturing process units.
Orie:commentef%V-t>-ti2) noted that a permit limit excee&ance discovered by routine
... . , _J. i . . K\ .. . , , -. V , '['''..
sampling wduld" drive' ^y regtijatory '"agency and action. The cofrimenter contended tfial the
permit limit 'action levels Ife^gMstantially above those required in this proposal. Tne 'commenter
explained that the State programs that form the floor accommodate the trade-off betweeri
controlling small leaks and th&risk of large emissions due to unplanned shutdowns atnd
-. ' ' i --"' ' ' -M- v
malfunctions that can occiir during such events. As a result, the State programs haitdfe the
decision on leak repair on a'caSe-by-case basis. The commenter suggested that the fePA should
",-:,'. " 3 '
defer to State experience.
Response: The EPA has considered the comments and the requirements for heat
exchange system repair of the HON. The EPA contends that the decision to extend the repair
time from 15 days to 45 days is consistent with the suggestions in the comments and the HON
requirements. Although other programs may not specify a repair time, the 45 day repair time is
necessary to provide consistency and eliminates a need for case-by-case review. It is also
important to note that the "...trade-off between controlling small leaks and the risk of large
emissions due to unplanned shutdowns" should be reasonably accommodated between the 45 day
h
repair allowance and the provisions for delay of repair found in §63.1088, particularly the
provision to allow delay of repair if a shutdown for repair would cause greater emissions than the
potential emissions from delaying repair until the next shutdown.
Comment: One commenter (IV-D-04) expressed concerns over the three conditions
proposed that would allow a delay of repair for heat exchangers. The commenter stated that the
first condition for allowing delay of repair if a shutdown is scheduled within the next 15 days does
not differ enough from the original requirement to repair the leaking exchanger within 15 days to
be of any use. The commentej: argued that the emission calculation used in the second condition
(i.e. shutdown emissions are greater than the potential emissions from delaying repair) is flawed,
47
-------
since the assumption that all HAP in,recircu}ating cooling water will be emitted to the
environment is not necessarily valid. The .commenter stated that the ,30-day de}ajr.pei;jpd-.p;royicied
by the thjrd^onditipn fe-'too- short to provide .much: benefit, above the original:
One conitient.e|;|^tp,tO?5) fouf4 tb.a^t^^-^^_d^J^provis|pj^-are-v
addresVli^"^^mpti;'pf^tfems dfie^air; sh.^p space pr. larg&cl&ne' ^v
recommend that iftheiifteett day,ie|afr rl(|»|rfeiftent;is,|etained, theji a|>|-d;£
period should be allowed. Anotbfer cpij$i$S^gij (iyXD-Q9| ex|tfessed^Qnjcem-;rej
feasibility of replacing an EMPU:heat e:xc|^an^r ^iihin 30, (Jla^s if a,:repjapernent e^cjiailgierimust
be obtained. r .
Response: The EPA has considered the information in the comments and has reviewed
the heat exchange system repair requireinents in the HON and revised the time allowed for repair
to 45 days. In addition, provisions for delay of repair have been significantly revised. As found id
§63.1088(a), a facility is allowed to delay repair if a shutdown is planned within the next two
months instead of within the next 15 days, as proposed. For cases where a shutdown is not
planned within the next two months, a facility may delay repair if it can dempnstrate that a
shutdown for repair would cause greater emissions than potential emissions from delaying the
repair to the next shutdown. The EPA recognizes that the method for calculating emissions in
this provision may result in a conservatively hkh estimate, but contends that the estimate would
v ... , |" - .;- '., '. : ,-,, ,, .-. , ..
not be unreasonable. Considering that heated water is exposed to the atmosphere in a turbulent
environment (the cooling tower) where the air/water interface is maximized, a relatively high HAP
volatilization rate to the atmosphere is expected. To address the concerns regarding repair shop
space or large crane availability, the provisions of §63.1088(c) have been revised to allow delay of
- , " -"" t': - ' ' , -
repair if "equipment" (which would include cranes, replacement exchangers, and repaired
exchangers) is not available.
Comment: Two commenters (TV-D-02, IV-D-06) contended that the requirement to test a
"heat exchange system" for leaks following a repair (40 CFR §63.1087(b)) is unnecessary and
wasteful. One commenter (IV-D-02) pointed out that the heat exchanger will be tested for teaks
-- - -; '.'.- .'i.
again in no more than 30 days. The commenter also argued that it is not necessary to test the
entire heat exchange system to determine if one exchanger has been repaired. The commenter
-__, 48
-------
recommended deleting the requirement to -tesi after 'a repair 6f requiring testing of the individual
* ,'.*,.-. " '. "'''.,' ^ ' ' ' i1 ' "' " '
repaired Heat exchanger instead of requiring monitoring of the entire, heat exchange, system.
Additi&na^y, the comrrienter noted that the testing methods specified in §63.1085 and §6$.1:086
.. - -'.; [:.- '..':... . f. > . --. . {.= ' . ;. '.', . ', ' .<,?,,. l\ ..'. %' .'-:; . ' ,.-A ' '' -., '<
cou'ld^si^£M^ restarting: 'tafe liru't. To retfleay thisi the
eornmeaiter !suggested aH6wing an| available method to test repairel^at fexphiprtgers, ifi^dihg
hydrtf testing. One eommenter (IV-D-06) recommended that the ter^%ek exc^%ie s^steM"
i .'.-'''" ! ' J - '^ < ' - F , "^ ' ''"?V.> '' '"'' ' ' '- V
replaced with "heat exchange system (or relevant circuit or individual he'at excSangfer)" in
§63.1087.
Response: The provisions of §63.1087(b) require use of the monitoring requirements of
§63. 108^ within 7 days of t&e repair or startup to confirm that the heat exchange system has been
\-
repairedr. The EPA has revised the mtinitoring requirements so that the facility has the option of
:> {.,;
monitoring the individual heat exchanger or multiple heat exchangers (§63.1086(b)) or using a
surrogate parameter (§63.l086(c)). These provisions are adequate to ensure that any heat
exchange system leaks are repaired completely while providing flexibility for the method used to
detect a leak.
2.4.41 Suggested Revisions to the Standard
Comment: Three commenters (TV-D-05, IV-D-06, IV-D-08) suggested that EPA develop
a performance-based standard for heat exchange systems. One commenter (IV-D-08) stated that
the requirements, as proposed, are inflexible and costly. One commenter 's (IV-D-06) suggestion
for a performance-based standard is based on EPA's conclusion that relatively high circulation
rates could mask a leak and EPA's estimate that the cooling water circulation rate through
ethylene manufacturing process units is approximately eight times the ckculation rate through
SOCMI units. The commenter assumed that the EPA's goal is to achieve the same level of
protection as the HON. The commenter noted that the definition of a leak in the HON is one
ppm. Using the assumption that ethylene unit cooling water circulation rates are 8 times SOCMI
unit circulation rates, an average ethylene unit cooling water ckculation rate of 102,000 gallons
per minute, and the 1 ppm leak definition, the commenter estimated that a 6.35 pound per hour
(Ib/hr) leak at a SOCMI unit would be detected and repaired. The commenter suggested that the
affected ethylene production facilities be requked to establish a site-specific monitoring plan that
49
-------
ensure a
would detect a leak of 6.35 Ib/hr. The commenter seated that such, a requirement would ens
''*-;; -ur-i -'',* '!'-^ .; V;- '-\ '':v;v ./''' i >;^IT;. \.;-.-. v " "' ''"'"
level of perlprmanee comparable to the,HON aij|;?wotild ! providfe. facptje^ fiemMJty to; t$ilpr a .
monjtorl&^grbgragi'to their unique circuin$anjces.^^
?^!t^
'' *
Two
i.,i.-*» - j»a,jt ' -'- -; - .... .. ._ . ^ . ^ .
facilifcie&.fe d0wb;fi Site-specific. alterftativid-njQiitoVing.'apfflpmches. Oije:
'" ''. iU'"v'vn'E^;:i-'.' ':^:' ' !'''"
-------
Response: The EPA conburs v^ftft'tlie! cotnmenter. After considering the comment and
reviewing: the heat exchange systettx monitoring provisions in the HON, the EPlA has include^
provisions for the monitoring of suJrogate parameters to indicate the presence of a leak in
§63.l086(p).
Comment: One oommenter (TV-D^03) suggested permitting a single analysis of supply
'.'': . . '>% . V% ' .<,.,".''..'" ;; :', . .
cooling water to. be performed fdr a set of exchangers. The commenter proposed that a sjngle
''.-.. . ':' ' - ; ' 'I', |Y;'' ' ' ' ' ''i ' ;- ' '
analysis could be used for the entire cooling water system in cases where there is not ^reasonable
opportunity for the concentration; of the monitored component(s) to change.
Response: The EPA agrees with this comment, noting that this appr6$ch is an acceptable
alternative to determining the entrance mean concentration as required in §6~3 J086(a) and (b).
Therefore, the EPA has included §63.1086(d)(2), which allows for a single upstream monitoring
location for determining the entrance mean concentration for a heat exchange 'system or individual
heat exchanger.
Comment: One commenter (IV-D-03) recommended allowing facilities to assume the
concentration of the monitored component(s) in the inlet water to an exchanger or group of
exchangers is zero.
Response: The EPA agrees with this comment, noting that this approach would result in
the most conservatively high estimate of the difference between entrance and exit concentrations.
Therefore, the EPA has included §63.1086(d)(l), which allows a facility to assume an entrance
mean concentration for the monitored substance of zero for a heat exchange system or individual
heat exchanger.
Comment: One commenter (IV-D-02) stated that the requirements in 40 CFR §63.1085(a)
appear to apply to all 188 HAP by requiring total HAP to be measured. The commenter asserted
that there is no logical, legal, or technical reason to require monitoring of all HAP that are not
volatile or are not present in ethylene process. The commenter suggested that it may be
technically infeasible to monitor for all 188 HAP. The commenter stated that the floor facilities
do not have such a requirement and EPA did not consider the feasibility and burden associated
with such a requirement. In a related comment, the commenter stated that the HON allows for
the measurement of surrogates, unlike the proposed NESHAP. The commenter recommended
51
-------
that EPA define HAP for the purposes of the heat exchange system requirements to include only
. < "' ", --. "^- '''' .'''' ' v~ '''- ''. ,' " *; .- "K, '.V,-,.1-'
those HAP listed in. table 1 to jpart 63 sobpart XX and clarify §63.1085 to allpw monitoring of an
'\-':--. X^'iK ,t'f/ ."?.i ^ " v ;'*" v : :.a .. ' :;--' "
appropriate HAP or HAPs from that table as a surrogate for all HAP.
' Y',. 'v - y& -:4-r .^| '** '-V ? / ' -, VA-iv^ <, -"» '*..:'
Response:. : tfie IP A gerjerally ; agrees, noting that the; monitoring provisions, oi
§63. 1086(a> 344 (b) apply to ":..the HAiP in table 1 of th&.subpart (either total or speeiated) or
:-nrt. .^;->.,|f^f^- ! ;::, >=V.- ;;, -rTfv '"..' ' ' ^:> '.."-:,;'- !A,S ^'V- ' ',.|7^1.sf '.
other, representafciv;6-;.sy.bstances {e.g., total oiEgaiR.iic carbon or volatile prganj
£-l^,'^l-''^'^.'^' :ir-- 'r^;iv't ;i;^r y,:i-, , ^^>5-?'; -;-
"'
. . .
that indicafe the presence of a lealc...""' Tha EPA contends that this language sufficiently
-.- r»a:>!;-:'-^i^;>it^-v .. ' -. . ; ^^ -1 'T-v : ' i*"':- -i'"i > ;1 .x -w.-.-
facilities to monitoif heat exchange systefras^foi: appropriate surrogates for all HAP,
; '"'.-; . v . :..>-.. '.:-.; i1 ; v."- ' '- ' -'-:;"i ^v- -:'i;-;y -vv ; S-.
2.4.5 Approval of Aftemate Analyticail Methods ' .
.; -. $-.:-- : ..' ! '. '"..i .."h " .- .;-
Comment: TWO commentei's (1V-D-02, IV-D-06) supported revisions to the requirement
'- '"'>, "';[ . :::'.l , '»... / .(. . ' ' .-'' .-;. . :
that a test method must be approved by; the EPA to be used as an alternate method for analyzing
-,.''.-: 'V -.':': ' '"'.'-. " " "'"..'
cooling water. Both commenters supported allowing any test method to be used;. one eommenter
:;. '. ) ' ' ..!'. '' .= " ' "' . ." --- ...
(IV-D-02) suggested that, if approval is required, it should be automatic within 90 days; One
eommenter (IV-D-06) explained that a variety of analytical methods are being used under heat
exchange standards that currently apply to the ethylene industry and many of them are not
' included in 40 CFR part 136 and would therefore require EPA approval. Two commenters '(IV-
D-02, IV-D-06) contended that requiring EPA to approve these methods, when they have already
been approved under existing State requirements or, in some cases, when they are approved EPA
test methods, is an unnecessary step that would impose substantial burdens on affected, sources.
Response: The EPA generally disagrees with the comment. Historically, the. EPA has only
allowed use of EPA-approved test methods to demonstrate compliance with NESHAPs. This
provides consistency and ensures that compliance is being achieved. The EPA has provided
procedures for acquiring EPA approval if a facility wants to a use a method not currently
-^i ' ' ' ^ :
approved. The EPA does not find adequate justification for adopting an alternative approach for
the ethylene industry. The EPA suggests that facilities wishing to use currently employed
alternative monitoring methods apply for approval of these methods prior to the compliance date
of the ethylene NESHAP.
Comment: One eommenter (TV-D-02) asserted that requiring facilities to use a method
listed in 40 CFR part 136 sensitive to concentrations as low as 10 ppm to test cooling water is ati
52
-------
unjustified above-the-ffeor option because it prevents the-use of methods that formed, the,flobr
and necessitates monitoring of individual heat exchangers. The commenter explained'that each of
their sites have developed successful methods customized to the individual cooling tower.. These
detection rJ2quire^n;tS^:
monitoring '
.
of m^igrtftadeless than
to certify under Title V that the|'are
ifc^ vyhich provides an enforceable mechanisif itp a
Response:
determined: to be
the presence of a leak. 1
EPA-approved test met'
,. The floor for heat exchangjb systems was
(at existing sources) or weekly (at ne^v, sources) to detect
determined that, to adequately detect the presence of a leak, an
t be used, and notes that Appendix D to part 63 icontajns a
--- - . . . -
procedure that would ajlejw fiicalitifcs to validate 40 CFR part 136 test methods for the HAP found
'.' i :V ' '- -': .
in table 1. Facilities interested jn using an alternative test method (e.g., not found in 40 CFR part
; . !'.. \ '". I'' ''
136) may request approv&l to do so from the Administrator.
2.5 SELECTION OF MACT FOR EQUIPMENT LEAKS
Comment: Two cbmmenters (IV-D-05, F/-D-06) stated that the proposed standard for
equipment leaks is ^fundamentally flawed because EPA significantly overestimated uncontrolled
emission rates from connectors. The commenters pointed out that the SOCMI emission factor
. used by EPA is based on data collected through an EPA study in the late 1970s. The study
obtained data frorn a variety of >SOCMI facilities, including but not limited to ethylene production
facilities. One commenter (IV-D-06) asserted that the data reflect the technology, construction,
and maintenance practices of the time and that experts generally accept that the SOCMI factors
!
are outdated and significantly overestimate current emissions. The commenter reported that they
performed a study to provide more current, accurate, and industry-specific information. The
commenter noted that their connector data for the study were gathered according to
recommendations from-BPA's staff. The commenter included the connector monitoring study and
related material as an attachment to their comments. The commenter reported that, based on their
study, the emission rate for uncontrolled connectors is more than 100 times less than the rate
53
-------
predicted using the average SOCMI emission factor. The commenter, suggested that the
difference in emissions factors is doe tp\he fact that SQCMi;e33|§sipti factors are based on; 1)
datatfiat are,.6ver2Q yeats old;.and 2.);da{a from a variety oFjMtfstrife'. t;he Gornrneijter
' x ' ' -'lat'theffkudyi w,^
iaj§iai' '^s^S^fy^-
. .... ;dlriss-Wil..«iifi W.?r ;": fo >S?J'4:'" U*1&^/tt^$k;l|Pil1-1"
of the pi
acco:
obe
causing them ltd disagree with EPA's conclusion that the data w$ra,npt .received intirjae:to
evaluated prior to the proposal. Another, commenter (TV-D-OS^ agreed, statink.that EPA/
adequate time to review and assess the informationjprovided by industry and urged EPA to
update their scientific database. Another commenter (TV-D^03) :supportes} th^ findings of the
study conducted by commenter IV-D-06 and recommended that EPA use the information in
setting the MACT floor for equipment leaks,
Response: The EPA understands that there are uncertainties in the SOCMI emission
factors used for the proposed LDAR MACT floor selection. Since the primary question seems to
be over the emission factor used for connectors and whether it would change the MACT floor
results, the EPA performed a MACT floor analysis using the emission factors supplied by the
commenter 7s (IV-D-06) study. The EPA's use of these factors does not indicate that they are
being endorsed by the EPA- Using these assumed emission factors to estimate each source's
emissions, the best-performing five facilities were identified and the current equipment'leak
detection and repair programs that represent the average of these best-performing five sources
were determined. This analysis showed that three of the five best-performing facilities are
' ' . p :
monitoring connectors, so connector monitoring is part of the MACT floor. Additionally, in a
separate analysis, a review of existing permits showed that certain facilities had adopted 100
percent connector monitoring annually in exchange for emission credits to be used for operational
flexibility. The permit information was compelling in that it establishes'emissions benefits are
realized due to connector monitoring, and that the monitoring is relevant in establishing a MACT
floor. As a result, the requirement to monitor connectors is retained. Details of the emission
factor and permit analyses are contained in docket A-98-22.
54
-------
Comment: Two cornmenters (JV-DvipS, TV-D-06) contended that, when EPA's MACT
floor analysis is performed using an emission: factor for uncontrolled connectors derived from the
study performed by the (Befihs Environmental Issues Task Group (OEItG), the fiv^best-
performing facilities are different;' One eommenter (IV-D-06) observed that, in Effr's MACT
floor analy/sis, the five1 Bers;t perforating facilities were found to be Dow, T^; rAmdcbi 'TXV '
Equistar, TX;. Equistaiv LA; arid Exxon, TX. The commenter claimed that thfe same analysis using
their emission:factor results M; Equistar, LA and Exxon, TX being replaced b^"Formosa^ TX and
Phillips, TX in the five;best perlorrning facilities. Two cornmenters (IV-D-05, IV-D-06) stated
that, according to survey datia,; only two of the five best performing facilities, as determined using
the OETTG connector emission factor, use comprehensive connector monitoring programs. One
commenter (IV-D-06) stated that the median facility of their five best performing employs audio,
visual, olfactory (AVO) monitoring and inspections for connectors. The commenter summarized
that, according to their analysis, when an industry-specific emission factor is used for connectors,
the MACT floor for equipment leaks for existing sources does not include connector monitoring
using Method 21.
Response: The EPA performed an additional LDAR program analysis using OEITG-
supplied emission factors for connectors. The EPA's use of these factors does not indicate that
they are being endorsed by the EPA. The five best performing facilities determined by the EPA
analysis did not entirely match those determined by the commenters. The EPA could not
determine, and therefore reproduce, the exact approach used by the commenters to determine the
five best performing facilities: However, the EPA's analysis resulted in a floor level of control
that includes connector monitoring because at least three of the five best-performing facilities in
the analysis using OEITG's emission factors have connector monitoring programs. As stated in
the previous response, the EPA also performed a review of existing permits to see what facilities
were required to do as a part of their permit, and this analysis showed that certain facilities had
adopted 100 percent annual connector monitoring in exchange for emission credits to be used for
operational flexibility. Again, the permit information is compelling in,establishing emissions
benefits'from monitoring and, as such, is relevant in establishing a MACT floor. As a result of
these analyses, the EPA maintains that connector monitoring is the MACT level of control for
55
-------
LDAR programs at ethylene manufacturing facilities. Details of these analyses are contained in
docket A-98-22.
Comment: Two- commenters (ly-D-OS, iy-P-06) asserted that based on their analysis,,
connector monitoring is not partpfthe MAGT.floo.r and is therefore it is^anabpye-theTflppr %
option, the; cpinmenters argue^thfi.!, as.Miibo^-tee^.or option, connector mpnjtpnng is.npt
cost .effective. The cornmentjers; p^spd.t^$.f|^iing, on a^study they conducjedin vi/hjch tljey found.:
that connector monitoring .progr.arns.do ,i|o^,f esultvin dgnificait^reduetion to emissions from,
connectors* One commenter.(JV~D-06) explained that np statistically signjficant.difference xas;
found, bet ween, the average emission rates,fpr.connectors included in an ongoing LDAR program
and those.being monitored for the first time,. .The commenter did not provide .an estimate.bf the
costs associated with connector; monitoring,but .stated that, because connector mpnitPring would
result in no statistical emission reduction, the cost effectiveness would far exceed any reasonable
benchmark.
Another commenter (IV-D-09) asserted that the previous cost estimates used by EPA and,
industry underestimate the initial cost to monitor connectors. The commenter estimated the cost
of monitoring and tagging to be $0.45 and $3.10 per connector, respectively. The commenter
estimated the direct cost to monitor 25,000 connectors to be $88,750. The commenter estimated
i
the cost to monitor one of its olefihs units to be greater than $100,000. The commenter also
noted that a four year study conducted by the industry concluded that there is an extremely small
occurrence, of leaking connectors in the ethylene industry. Furthermore, the; commenter noted*
current permit conditions require connector monitoring, which achieve HA!5 emission reductions.
Therefore, the commenter concluded that the proposed connector monitoring requirements
impose additional cost burdens with no substantial HAP emission reductions. Another .
commenter (IV-D-08) agreed that the potential emission reductions from connector monitoring
are small, and the requirement to monitor them is not justified. The commenter recommended
deleting this provision.
Response: The EPA generally disagrees with the comment. "As discussed previously, the
EPA's analyses have demonstrated that connector monitoring is part of the MACT floor level of
control. Since it is not an above-the-floor option, cost is not a consideration in determining
56
-------
whether tequiririg connector monitoring is appropriate. However, the EPA has estimated costs
associated wtft1 complying with the rule to evaluate-the impact ofthe rule and appreciates the
additional cost information submitted by the commenter. The EPA used the. costs'supplied by the
cornrrifehtei- to validate their 'estimates aritfourid'that the EPA estimates are actually Highli^ when
considered on aft aiimiSl basis* The primary reason for this is that the tiiaprity ofthe'^orriitiehter's
annual-costs areVfot initial tagging, so that annual costs drop oft signifiBan,tly aftesr th^%st $fear.
The EPA's anhiiaieosts Include' the amortized capital costs for monitoffiig equiprherf'aW initial
'".'--< '; ' . ' . . k ' . ' t -'t ' ' .
tagging, along wittt the bii-goyig rttonitoring\and, repair, maintenance, aathinistratidn/and pump
seal replacement costs. The EPA's estimated annual costs, while slightly less than thfe
comrnenter's for year one, are Japproximately three times those proyidedfty the comrhentef. It is
also important to note that EPA, in. selecting 40 CFR part 63, subpart UIJ for equipment leaks,
has allowed ethylene facilities to reduce their monitoring costs through good perfornjiance instead
of annual monitoring. This allows facilities to reduce monitoring (and therefore monitoring costs)
from 100 percent annual monitoring to once every eight years in cases where a low proportion of
connectors are leaking.
Comment: Two commenters (TV-D-05, IV-D-06) stated that, because the equipment leak
provisions of the NESHAP should not include connector monitoring requirements, 40 CFR part
63 subpart TT is an appropriate and justifiable standard for the ethylene industry. Ofle commenter
(IV-D-06) explained that subpart TT does not include requirements for connector monitoring
while the subpart referenced iri the proposed NESHAP, 40 CFR part 63 subpart UU, does. The
commenter noted that part 63 subpart W is incorporated by reference into the petroleum
refineries NESHAP and subpart TT is modeled after subpart W.
Response: The EPA generally disagrees with the comment. The EPA recognizes that the
MACT level of control selected for petroleum refineries does not include connector monitoring,
and as such, subpart W is appropriate. However, the EPA has concluded that connector
monitoring is a component of the MACT leVel of control for ethylene production units.
Therefore, 40 CFR 63 subpart UU is the appropriate subpart for use in the ethylene production
NESHAP-
'57
-------
Comment: One commenter (TV-D-01) requested clarification on the boundaries of the
'','" ' " " ."!.'
required fugitive emission monitoring. The, coirjmenler specifically questioned whether wastes or
^ .-,-- .^>.,: M; - - '. 'J: . '".rr . V> - .". - . -.: r,v; .: .;;.,.
wastewaterS;]3iped or transported to an integrated .waste: manjagemertt facility fgr-incineratipti are
covere4by thfreftiyleBe NESHAl? where pjpirjs-aad equjprijeat cpiaitv.mafe^lal ^ith,jgreat.gr than,
^^Ti'ir Y? f-|;^1'(;iSi£S^tV->"! '"'"-!> <,; V'vV* J : *''': ';:-' , :"- -M; * '. ,''>?/>'> .-.^-'< ' '«--* (.«?\* «- 5 : !?'"' 'W ,-' -'«'.'..'.
jt|er Gdjranejjtpr staged t^at.tJh(e|e ;tyr|es: ^
aft of fatties thatare ntot direptly jnyplyed, wit%ethylene.|!rpd}iGtiQn> The\ '
l"T^; 'i"&':-'- '' ' '-W '! ' ;' v. '"''A ."':' ;i ;:';'- -> -:; ,>.!'-^ * '. '" v
if the,,rule is intended., torapply to titiis equiprnent, svhether JjPfy gathered ,
- " ' '/"?' '..:' *'.'.'. 'Jt1 -\v i'tpti^ T:J ri,. ;: fr'*, ^K ; ri-\ - .::. ::;;,;. f3, -f .
associated cost data. . ; .
.;-.V' ' ' " ; .:
Response: The EPA recognizes that e^hylene units may often, send wastes to a waste.
i-.-jH-t -.,; :' . . ' --: '...'.... " : '.;:* .,.'.... );.. :
management facility on-site that also, handles wastewater streams from npn-ethylene processes. It
is important to notice, that the requirements of the ethylene NESHAP apply only to the affepted
source (Le., the ethylene production unit), as defined in 40 CFR §63.1 103(e)(3). The EPA would
also like to point out that wastewater streams associated with an ethylene production unit that
require control are subject to the BWON requirements for wastewater collection, transport, and
treatment systems rather than to the equipment leak requirements of the rule.
2.6 SELECTION OF MACT FOR PROCESS VENTS
Comment: One commenter (IV-D-06) responded to EPA's request for comments on
setting the applicability cutoffs for process vents at new and existing sources at the same level
The commenter does not believe that having consistent cutoffs is adequate justification for making
the requirements for existing sources more stringent than the MACT floor. The commenter
argued that, because EPA does not have adequate data to assess the costs associated with
expanding the applicability of control requirements, the proposal to set a standard more stringent
than the floor is legally unsupportable and should not be promulgated: In response to EPA's
suggestion that the more stringeint standard may be promulgated unless data .to refute the
assumption that few additional vents would have to be controlled, the commenter stated that it is
legally inappropriate for EPA to justify a regulation on the basis of unsupported assumptions.
Response: In the preamble to the proposed NESHAP, EPA explained that the decision to
set the applicability cutoff for process vents at existing sources at the same level as for process
58
-------
vents at new sources -was based on the assumptions that this would result/in negligible additional
costs, because few if any additional vents would have to be controlled, and that there would be a
benefit to presiding consistency between the cutoffs for vents at new and existing,sources, The
,?V\ -V- ' V'V. >!-'--':. *'.' " ' ':'-.' '. . ''': :\ >/',. '; ;. -"t. A '' ''"i:' '"'' '''" '
EPA soficited cxDrrirnents; attd'infpi&nation to refute or support EPA's aSsumptiofiiiS. Ko data were
' , , ^ii"'.ti"i->:A'U:; >'L ' '''/» '' " ' ;"'' *,'?v-' ''^:'0;r-,-.'.;";' '" '<.i-''
received on the nwhibec pf YeftitS: that would have to be controlled. Alfn0u&fyri0:t discussed
; /-/:/: .x. !'.<£. ?!:;- \?|. \s-.. ;;; .:.'' ' ' ':'.-; ..' <:;. ''*'$q:^'. .;/..' *: vx..
specifically, it is;apt)ar$ftit frpirL:4omments received that the comrn£nter$;.d<> not see a benefit.to
; -'i'-'. . V." ;-ki;./V ;. . I '.; . . ;.'' '". 'V^r'* . llfl».
having consistent applicabjility eiatoff for process vents at new and*existing Sources: Due to lack
:' -;.,--':;; i'-ft :u \V-';.; . .- ; " :'*:: ;!/-;_ < _ ;
of data to suppoirttfjie-o^giiiial'ai&siiirnptions regarding the number of iyents and cost of control and
"' .'' '- *:'- '-!'' i- '"'. " -"' ":-,-(, \ ^ :: .
given the fact that there do£s not appear to be support for providing; consistent cutoffs, the EPA
. -jts; .- :i i;!',1!,^:;! ' ' .:-."J ' ' ;- .: f ' . "' /
has determined that it .is not appropriate to require, an above-the-iloor option for process? .vents at
- .''';'- l >'". j !, 4. '' - ' ..>':' . . - ' '
existing sources. Thereforei, the.'applicability cutoffs for process yen,ts at existing sources are set
*c / .^ ' 'Six '''< ' ' ' ' '
at the floor level; (^Oppriry HA^ and 0.011 scmm).
- ';l .'' '- '' ' 'f\':\ ' ' ' - . .
Comment: On£ cornm^nter (IV-D-Ol) noted that the definition of ethylene manufacturing
process vent says that it is "a gas stream containing greater than 0.005 weight-percent and 20
parts per million by volume HAP..." The commenter stated that the use of "and" in this definition
makes it as impossible requirement because the relationship between weight and volume in a gas
stream is a function of molecular weight and is different for every compound, The commenter
also pointed out that the MACT floor for process vents is control of vents containing 50 ppmv
' i ' ' * ' '
HAP or greater and requiring control of vents containing 20 ppmv or greater is an above-the-floor
action.
" "\
Response: The 0.005 weight-percent and 20 pprny cutoffs are meant to be mutually
1 . i
exclusive. The gas stream is a vent if it contains both greater than 0.005 weight-percent HAP and
greater than 20 ppmv HAP. If either condition is not met, the gas stream is not a vent. The EPA
agrees with the commenter that the MACT floor for process vents at existing sources is control of
process vents containing greater than or equal to 50 ppmv HAP. Although a gas stream is
considered a process vent if it contains greater than 20 ppmv (and greater than 0.005 weight
percent) HAP, it is not required to be controlled unless it contains greater than or equal to 50
ppmv BLAP and has a flow rate greater than or equal to 0.011 scmm, as specified in table 7 to
§63.1103.
59
-------
Comment: One commenter (IV-D-04) noted that TNRCC Regulation V, the basis for the
MACT floor for process vents, requires vents with a VOC concentration above 500 pprnv to be
controlled. Thecommeriter asserted that it was inappropriate for EPA to reduce this limit to 50
ppmv HAP basedVn one surveyed process. Tb,e commenter stated that, unlessvE^ has some
additionalsupjportj^g-Woianation, ^4'IMMMpT fpermust be'se&at'^QO'ppniy.
Response: The Conversion ojf 500 piprxiVVoC to 5Q ppmv HAJ* is base^on the average of
vent composition data provided in survey responses for five process vents,. The\finding that
approximately, 10 percent of the VOC in process vents is HAP is also supportedi-by data provided
on emissions from equipment leaks. The arithmetic mean of HAP fractions reported1 for
equipment leaks for |7 ethylene manufacturing process units at 11 facilities is aljso 10 percent.
The EPA contends that this is a sufficient basis for developing the 50 ppmv HAj applicability
cutoff, which is included in the revised NESHAP for process vents at existing sources.
Comment: One commenter (IV-D-02) stated that, because the MACT floor for process
vents is based on Texas regulation 30 TAG chapter 115 subchapter B, it includes a 0.011 standard
cubic meters per minute (scmm) flow rate cutoff and failing to include this cutoff in the definition
of process vent is an above-the-floor action. The commenter stated that running pipe to collect
small streams would be extremely wasteful and unnecessary.
Response: The flow rate cutoff of 0.011 scmm has not been disregarded; it is included in
':
table 7 to §63.1103. The definition of a process vent does not indicate which process vents must
be controlled. The HAP concentration and flow rate cutoffs provided in table 7 indicate which
vents must be controlled and these, are consistent with the MACT floor level of control.
Comment: One commenter (IV-D-02) suggested that it should be clarified" in the
definition of ethylene manufacturing process vent that the concentration and flow limits are annual
averages. The commenter asserted that annual averages are consistent with the floor and other
similar limits in the rule (e.g. waste and equipment leaks).
Response: The EPA assumes the commenter is referring to the concentration and flow
rate cutoffs for control applicability, which are based on the MACT floor, rather than the
concentration cutoff in the process vent definition. The floor level of control for process vents at
existing sources is based on the Texas regulation in 30 TAC Chapter 115 for process vents. In
60
-------
the exemptions froni control, the cutoffs of 0.011 scmm and 500 ppmv VOC are not specified a£
being annual averages. The floorlevd of control for process vents atnew sources is based, on 40
CFR part 60 subpart NNM The 0.008 scmm flow rate cutoff for cbrititfl 'in subpart NNN ,is:;nj>t
an annual average. Subpart NNN. does not include a specific concentration cutoff, but uses l
instead a total resource. effectiveness (TRE) jndeix to determine Mich' vent s'trealns muk'be
controlled. The TRE index value- is partially based on vent strearfl ₯C)C concentration! The!
concentration used to Calculate the TRE index value is not an annual Average. The* EPA disagrees
with the commenter's statement that Using annual average is consistent with the floor level of
control and therefore, has not 'revised the applicability cutoffs to be stated as annual averages.
Comment: One commenter (IV-D-02) recommended that §63.1104(a) be revised to
exclude process vents routed to a flare from the applicability assessment in §63.1104. The!
commenter supported the exemption of vents that meet the HAP "or TQC removal requirements of
§63.1103 from the applicability assessment.
Response: The last sentence of §63.1104(a) was revised to read 'The owner or operator
of a process vent is not required to determine the criteria specified for a process vent that is being
- - - V-
controlled in accordance with the applicable weight-percent, TOC concentration, or organic HAP
concentration requirements in §63.1103". The control requirements in §63.1103 are to "reduce
emissions of organic HAP by 98 weight-percent; or to reduce organic HAP or TOC to a
concentration of 20 parts per million by volume, whichever is less stringent, by venting emissions
through a closed vent system to any combination of control devices and meet the requirements
specified in §63.982(b) and (c)(2) of subpart SS of this part." Provisions are provided in
§63.982(b) for routing process vents through a closed vent system to a flare. Therefore, it is not
necessary to revise the language in §63.1104(a) to exclude process vents routed to a flare.
Comment: One commenter (IV-D-04) contended that EPA should consider a total
resource effectiveness (TRE) index approach for determining which process vents must be
controlled in addition to flow and concentration Cutoffs. The commenter stated that the current
flow and concentration cutoffs are extremely low and most other MACT standards employ a TRE
index to determine if control is Warranted for low emission rates. The commenter asserted that
not including a TRE index means that controls must be installed even if the cost of controls
61
-------
relative to the emission reduction achieved is prohibitive. The commenter expressed concern that
NSPS NNN and RRR are rnentipsed in the preamble and in the l^ACX floor document asr.being,
applicable to the; sources considered, for t% MACT floor and bojh have TRE index compliance.
options. . , ( s .,.;
Response: The EPA consideredfaqluding using alfRE uidejc approach fipr ppcess vents.,
However, adequate data to develop a>TR||jmde*.were not available.. As discussed in the
preamble,teethe proposed .NESHAP/, ordy^wo of the five best performing facilities, reported that
the have process yent$ and vplujnptric^p.vtf'r.a);e and HAP cpncenfratipns of the .vents,, needed to
develop a TRE index, were not provided.. Jpow rate and conpentration cutoffs were developed
... L-^ , L. . . -..,' i , . , j
for control applicability based on the regulations the vents are subject to. These cutoffs serve a
similar purpose as a TRE index by not requiring control of very low flow rate or concentration
vents while reflecting the current level of control that is being achieved.
Comment: One commenter .(IV-D-10) argued that emission reductions achieved by the
ethylene prpductipn NESHAP are modest and fall short of levels that can be.achieved with
currently available control technologies. The commenter pointed out that EPA stated that the
proposed standard would be achieved .at minimal cost to the industry and would in fact result in a
cost sayings for seven facilities. The commenter noted that benzene has been tagged as one of
five "national driver." hazardous air pollutants based on public health impacts, The commenter
claimed that HAP emissions from ethylene production can be destroyed with either catalytic or
thermal oxidizers and that the air pollution control industry has regularly achieved 95 to 99
percent destruction efficiencies of all compounds on other processes. The commenter added that
'*& ' ''
catalytic oxidizers can easily control solvents.including toluene, hexane, and xyfenes (all major
emissions in ethylene prpductipn) typical of surface coating industry. The commenter encouraged
EPA to further assess the capabilities of control devices, relying on the more recent control and
cost information and note that improvements in technology and their costs have occurred over the
last, few decades. As. an example, the commenter noted that 15 to 20 years ago, catalytic
oxidizers achieved destruction efficiencies of 90 percent while today's.oxidizers are capable of
achieving greater than 95 percent. The eommenter provided that a temperature of 440°F and a
residence time of 0.24 seconds is required to achieve a destruction efficiency of 99% for benzene
62
-------
in a .catalytic oxidized while a teniperatiire of 1460°F'and,residence time of 1,0 second is Required
' * ' ' - -j *;J ' < ' '' ' '* ''''''-' 'tj'' f '* ' ''
to achieve the same effjcieticy in ^thermal oxidizer. The coifimehter also provided that greater
reduction.efficiencies are acffie^
^- -! -. - ' . /> .:', .. -,v , . f\ r '';.- .I''->';; ,-t.,y, .'/.)'.
typii6a|f^he oxidize* is s-jzM! td achfc-ve the'desired efficiency.' :The '
percenPidestruc'Eion ellcifcrfef ^'typicalfor'V6c: while 99 perceHt^tnote t
The commfentefreonciide(i tmatfealkig theMACT limit ofl
reflects Tdqbirernents'fbr the V^'C Me^ of control.
Response: Tfie fiPA appreciates the information provided by me eommentet. However,
the information received does not compel EPA to make any changes to the requirements of the
NESHAP. The EPA assumes that the commenter is mainly concerned with the process vent
provisions. The MACT floor for process vents is based oh the control achieved by the best
performing facilities, which are controlling process vents with either a boiler or a flare. None of
the best performing facilities are using a catalytic oxidizer to control vents, however a catalytic or
thermal oxidizer can be'used to'comply with the rule if it achieves 9&: percent reduction or 20
ppmv. Previous EPA analyses for several rules have concluded that 98 percent control can be
achieved on a long-term, continuous basis by the best-performing combustion controls including
boilers, process heaters, flares, and incinerators. We appreciate the commenters information that
catalytic oxidizers have improved over recent years from 90 percent destruction to levels of
performance that would meet the ethylene rule. The EPA did nof find any justification for
requiring a level of control more stringent than the MACT floor. There is insufficient data to
show that any higher level of control would be continuously achievable for ethylene process vents
or that the costs would be reasonable,
2.7 SELECTION OF MACT FOR STORAGE VESSELS
Comment: Two commenters (FV-D-02, IV-D-06) noted that air emission standards that
regulate emissions from storage vessels generally include an exemption for pressure vessels
because, by design, such vessels operate without emissions to the atmosphere. One commenter
(IV-D-06) cited the HON, 40 CFR part 60 siibpart Kb, and 40 CFR part 63 subpart Y as a few
examples of rules with a pressure vessel exemption. The commenter requested that the exemption
63
-------
be included in the ethylene production NESHAP based on precedents and the fact that pressure
vessels do not have HAP-emittiiig vents. Two commenter^(jy-I>02 and Jty-D-06) provided a
suggestion for the exemption which could be added to tlie ttst of exceptions in §63. 1 lOS^e.^l^ii)
or the definition of storage vesseiin §63,1101: "Pressure viesjsej.s designed to oo@|^te in jsxepss of
"**: .i;^ "vjyl** . '--' "ws.-- "--'»' ''V^^'v^* ^;?i'~'; ^^:>V^;':.*i cc> %>. v |f $8)*'"-:,'' ^J; .'H1^ -i^1
204.91 ikifetasciais, and without ermssiajis^o the atjn05phere," One conirnerrter t3tV-D-029:also
^.^<^%;if''-^ , uk,~\: ^'n,f^ "^V' ^jf ;t::«i^'f:^:/;".^'"fei^::;>'.'.; ^F^ =
noted .^ll^tc^age^essels that are part qfta^fc; iriipks pp t^ c^rf ;were.jcjot: e^5lu&0d. Th*
v -: ^l.'^iiiM'K-^"-"-' 1>':- '" :%'r^*ff.. ''';' ^l-'-!''J- ": fr^-S!Wv;:';/-f'i ;
commenfei' predietid that, at a rnSnimum, tnese^oVersisitts.will' impose .confusion and conflicits
' ' .' . ' ;r-^:-'^ -. i:';fr;y . -; ^ if^fe^^J^^: '«^'M!
with transfer and pther requirements. The,cqmnferiter suggested.the fplfov/ingjjpe incorporated
'.,';: '^iV .-'( ."'.'. r-'V-' " W' '-^ ^t>-'' ' --:." : * %
into the list of exceptions in §63.1103(e)(r)(ii): "Vessels permanently attached to motor vehicles
such as trucks, railcars, barges, or ships."
Response: The EPA agrees that pressure vessels and storage vessels that are part of tank
trucks or tank cars are not subject to the storage vessel control requirements of the. ethylene
; ' - . ' ' I ' . ...
production NESHAP. Pressure vessels operating in excess of 204.9 kilopascals and without
emissions to the atmosphere and vessels permanently attached to motor vehicles such as trucks,
railcars, barges, or ships have been exempted from the control requirements of the ethylene
production NESHAP These vessels, however, may be regulated by other.existing or future rules.
M' . i ' ' ' i ' '
2.8 SELECTION OF MACT FOR TRANSFER OPERATIONS
Comment: One commenter (TV-D-06) asserted that the MACT floor for transfer
t - '. -. ' r
operations is "no control", based on the finding that the transfer operations at the best controlled
facilities are not controlled. Based on this finding, the commenter contended that the transfer
operation requirements in the proposed NESHAP are an above-the-floor option. The commenter
asserted that the EPA has not justified the above-the-floor option in the record and concluded that
the transfer requirements do not meet the requirements for setting standards in §112(d) of the
Clean Air Act.
Response: Determining the MACT floor for transfer operations was complicated by the
fact that only one of the best performing facilities has transfer operations. As with other emission
types (process vents, storage vessels, wastewater), EPA augmented data on control devices and
techniques in place with the requirements of applicable regulations. The regulation applicable to
64
-------
the best per&nniftg famlities with,transfer operations is the Texas regulation; 30 TAG
' '-I', & ' '-;.''
Chapter 115 SubchiaptefC Volatile Organic Compound Transfer Operations jLoadjtng and
Unloading of Volatjiie'OrganicCompounds, which requires conrjolrof loading.ofVOQ witha
! "l . , ' . ' ',' -L ', 1 "" ' ' - J .I. "",' ' ' '* 4>lVl ' I '
' '- ' -I. ' ' *; f'-'t '- -'I;. '\ i . ' ! :-}i> ""A - ' . i * "'" ^ . 1 ' -.. ' ' , . !, ' jj*. * ' fiJ
ta$(&^edsd$«^ to t).S psia i£ 20,000 galops ol^Cfeor motfe are
,:_ . .;'"'"' ' ' *
and quantifies jtransferred met or exceeded the cutoffs. As explained in the floor analysis
documentation, this State; regulation is the basis of the MACT floor for transfer operations. The
EPA disagrees, that the transfer requirements are an above-the-floor option,. The same cutoffs in
the State rule have been incorporated into the ethylene production NESHAP. Therefore, the
-\ i. . - ' - ''''' rl-| - ':.:, :
NESHAP does not impose any additional requirements or control costs on the majority of existing
facilities.
2.9 CLARIFICATIONS AND CORRECTIONS
2.9.1 Waste Provisions
.. ^ .
Comment: One coramenter (IV-D-06) stated that the. last sentence of 40 CFR
§63J092(a)'is potentially misleading because it refers to the 10 Mg TAB and "any similar cutoff'
but. does not provide a citation to the relevant provision of the BWON. The corpmenter
suggested,the following change: "...The requirements do not include 40 CFR 61.342(aX which is
a provision to exempt sources with a total annual benzene, quantity less than 10 megagrams per
year, or any similai1 cutoff, from control requirements." The cdmmenter noted that the section
will not be needed if, as they suggested, the 10 Mg TAB js included in the ethylene production
NESHAP.
Response: The EPA agrees that the original language was misleading and has deleted the
phrase "or any similar cutoff' from §63.1092(a). However, as explained in 2.3. i , the 10 Mg
TAB is not included in the ethylene production NESHAP.
65
-------
Comment: One commenter (F/-D-06) suggested that 40 CFR §63.1092(b) should be
clarified. The commenter noted that the section states that; the ethylene: production NESHAP
waste requirements apply to screams containing table 1 HAP arid, fails to mention that there are
.-:-.,' , :":: ' .. K ,. > , '. - ': '.,"""."> '.;-\ ^:'H.Rv' > v; :
additional flow and loading' thresholds foj- applicability. The concenter &]a|^j^Ehe. following
-'""-' -'{.'[*- ''lii;*'"' '-'.'. .^.'k'i -'V"r": <»->'- r '": ,'^ti" 'fi- '-^^.^^^r^'^^4^ E.P?J;<' ''''>'''"
' for ethylete prodiBiitJon sources apply to waste sitei^'
anv of the at least 10 ppmw of HAP listed in Table 1 of this subpart, not only waste ''stress
~ " l - '\| . - ' -"..- t '.. . , ; , ,'-, '1 ,;, til?
; . ..-.;, .;,- JJK; .-".-: ^ - "' - : -"f ' ' ":(2) does
not define the terms "process wastewater" or "maintenance wastewater", o'nly the terms "waste"
and "waste stream." One commenter (IV-D-06) contended that the description in table 7 could
be confusing because 40 CFR part 63 subpart XX does not regulate "process wastewater" or
"maintenance wastewater"; it regulates "waste." The.commenter suggested that the description in
the first column of row 7 of table 7 be revised as follows: "Processes that generate waste (as
( ,
defined in paragraph (e)(2) of this section)." The other commenter (IV-D-09) noted that "process
wastewater" and "maintenance wastewater" are defined in §63.1101 of subpart YY and in the
66
-------
pmposed'revised;§6l .341 of the BWON, and "process wastewater* is defined in §63. 1037 of
subpart XX and requested clarification on which deftnitioitjs apply.
Response; The first and third columns of row 7 of .table 7 have been revised to address the
eonamenteis' concerns. - H^te /that the terms1' defined in subpart XX.apply,rQ the waste stream
":. ';vV*i v"..-c..,>. ;T^fc T'4-'$ *Ai * '''''' :': ' .'':.'-' vvi!1 .. '
requirements for fetHytenjs fjrodufctfon units,.
C|iHot|ignt 0ji£ cofnm^ter (rV-C>-6i5) suggested that the cross reference in 40 CFR
§63.110Q(g)<6f(ii) sh^u'tbe cdjrebtedfrom §63.1103(e)(2) to § 63.1103(e)(3).
Response: This correction has been made in the revised regulation,
FTfT T~r" ' ~'$ "' '.'..'. '"'."'. ' " ' '"'
2.9.2 HeatExchanger Provisions
:,-.} ;. ' . ;.. :-. .;A. .
Comment: One doinmenter (P/-D-03) asserted that a definition is needed for "heat
exchanger" and suggested; the following: "For the purpose of defining water sampling locations, a
cooling water heat exchanger is one or more heat exchanger units that are (1) used for a single
process service (such as ithe debutanizer column overhead condenser, or a cracked gas
compressor discharge cooler), and (2) where a single water sample can be collected and analyzed
. , t ^ , ^ ^ . \
to .determine the leak performance of all heat exchange units making up the cooling water heat
exchanger." The commenter proposed that the definition would eliminate confusion and
unnecessary samples. The commenter suggested that a unit-specific monitoring plan could use
this definition or a similar one and specified that the definition is not intended, to limit the number
i
of heat exchangers grouped together for sampling purposes.
Response: The EPA contends that the definition of heat exchange system, which can
include a single heat exchanger or multiple; heat exchangers, adequately describes the equipment
subject to the heat exchange system requirements. Additionally, the monitoring requirements for
heat exchange systems (§63.1086) clearly explain where samples must be taken, and have been
revised to allow flexibility to develop site-specific plans for sampling locations as long as the
specified leak detection criteria are met.
Comment: One commenter (r/-D-06) recommended that §63.1085(e) should be revised
to clarify that a specified increase in concentration across a heat exchanger constitutes a leak..
The commenter suggested the following revision: "...if the exit mean concentration is at least 1
ppmw or 10 percent of greater than the entrance mean,..."
67
-------
Response: The revision suggested by the commenter has been made. Note that §63.1085
of the proposed NESHAP is now §63. 10.86.
Comment: Two commenters (tV-tM)6, IV-D-09) suggested that the following text in 40
CFR §63.1086 is unnecessary,1 as it is inferred in §63.1083(e): "...in the ethylene rnan.ufacturuig
-' -,i'~ ,;; --i. ..".?:;: -;T ;: ,,/>: '. *''? > -J. < T':xv ..xl' i; ^# ;^O'V,.^4 '
process' uiaitsusea to 'cool fluids containtig 3 percent: by weight organic HAiP (or ottaet m
" '
>V r1. ' i--'1, "<- ' . ;*' ; ' ^ ' ;> ', ; \
substances) or greater." One commenter {IV-D-06) contended, that thVproviskKV as proposed, is
; , '' 'v -, ; ,.'.': ' ; , - "fc >- ;'' '.>!.' : /;., ^;( ;r ^\t :'^s '- i
redundant with, the exceptions listed in §^3.iOS6(f) and the term "other mentioned sjubsfanees" is
ambiguous and potentially misleading.
Response: The EPA agrees that tfie including the phrase in §63. 1086 is redundant and has
revised the section so that it does not include the phrase. Note that.paragfaph §63.1083(e) from
the proposed NESHAP is now §63. 1 084(e).
Comment: One commenter (IC-D-06) recommended that 40 CFR §63.1089 be corrected
to refer to the recordkeeping requirements of paragraphs (a) through (d). The section currently
refers to only paragraphs (a) through (c).
Response: The EPA has made the correction suggested by the commenter.
Comment: One commenter (IV-D-02) stated that the phrase "heat exchange system(s)" in
40 CFR. §63. 1083(a), (b), (e) and (f) should be replaced with "individual heat exchangers." The
commenter stated that, as written, it appears the entire heat exchange system (i.e. all heat
exchangers in the system) must meet the exemption criteria. The commenter believes that these
exemptions were provided to eliminate burden of monitoring heat exchangers that cannot leak
HAP to the environment but the proposed language doe.s not achieve this objective.
Response: The definition of "heat exchange system" has been clarified to reflect that a heat
exchange system can be defined as a single heat exchanger, multiple heat exchangers, or the entire
cooling system. Using this definition allows a single heat exchanger to be exempted from, the heat
exchange system requirements. Note that the exemptions in §63.1083(a), (b), (e) and (f) in the
proposed NESHAP are now in §63.1084(a), (b) and (e).
Comment: One commenter (IV-D-06) suggested that §§61.355(b) and (c) should be
revised to remove the reference to §61.355(a). The commenter stated that, according to table 3
of part 63 subpart XX, §61.355(a) is not included in the proposed NESHAP.
68
-------
Response: The provisions of §61.355(a) require the calculation1 of a TAB "for the purpose
pf.detenniniiig whether the BWON management and treatment requirements^are applicable. The
proposed NESHAP would have required all ethylene produc^bjnr^ti'ioc^t.'^:i^, , ^
and 20 parts per million by volume HAP." The commenter noted that the preamble and
69
-------
supporting documentation clearly indicate that the term "weight-percent" was intended to be
"standard cubic meters per minute."
Response: The revision suggested by the commenter has been made.
.". -';:. .', ~ '': ; > '.. ;.' -- "«?:' <<'.*
Comment: OBe-comrnenter (TV-D-Q6) suggested that the cross reference to 40 CFR part
T'f1-'" 'if.'i'k':', \ ,\ *&$£*' *K -, '. "\ . ' :> ":*' --V-", & '- " ' '"". ''"*-' ''n" - ' ;)' ~'<3'" ' -'<"'" '-'
63 subpait S^lifi row 4 of table 7 to §63.1103 shouMbe cjatifiiad. The GQmmenter explained.that;
VvjL^;«y-V ,/^r>-W,.udv: ''' f-'- :' t?-:^^, ^?>i, -f^ "Li ' *< '" ';*'%'^ 'V"r '^.f;;.T ',;: ?:' r-~
the third cptomn ifij row 4'remiirgs process yenjt cbntrol' devices to meet requiretaeiits^specified in
':'-.' '':',', -'^\i:' <^-£-' v1 '". -:fs-,.!"--' .' ;t LiX; <=- i/^d"^,-" ,:""' -*!'' v;v.t|' v> "" '.;>/...' - -.
§63.982(a)(2)L' Tti^relerenced r^quirernentp include those specified; in §^3;982(ei which apply to
- '.\/ ,;;>.':; ^>1,?'i j;.. : ! ' " \$$ . v :^ c- M ":..'.:.';. '.>;.,.' - if. '^ «'-;-. IT- --";.'?.v - "
vent streams subject to a total resource .effectiveness. (TRE^ased applic;abili|y;caatoff. Because
Y- .,'"! }'. I '>'ii--; '',<:';,''*' :''~, ' ''-< \.'''}^ ',';. "' *' '(- ' ' ' '; < '' '' ''*" "._ t it1. ' ' ' ! S^'">' ' i '
TRE appficab|ity ciltofFs do not,; appl^ to the,ethylene production NESH^P, thbicommenter
suggested-that the co.mment should be amended to state that §63.982(e) does nqt^apply.
' ; t*- ' !
Response: The requirements in table 7 for process vents have been revised to refer to
§63.982(b) and (c)(2). This eliminates the reference to §63.982(e), which the E^A agrees does
not apply to ethylene production process vents.'
Comment: One commenter (IV-D-06) suggested that §63.1103(e)(l)(i)(B) needs to be
corrected. The commenter noted that the section uses the term "process vents" and refers to
paragraph-(e)(2) for a definition, however, the definition in (e)(2) is for "ethylene manufacturing
j i ' '
process vents." The commenter suggested that the term "process vents" be replaced with
i ' '
"ethylene manufacturing process vents" to avoid confusion.
Response: The EPA has decided to refer to process vents from ethylene production units
as "ethylene production process vents" (or "ethylene process vents," in short) to be consistent
with the source category for this NESHAP, which is ethylene production. The NESHAP was
/ - .
revised to consistently use the terms "ethylene production process vent"or "ethylene process
vent."
2.9.4 Closed Vent Systems
Comment: One commenter (TV-D-02) recommended that the wording in column 3 of
Table 7 for items 2b, 3, and 4 be revised The commenter noted that the current wording requires
reduction of HAP emissions by 98 percent by routing through a closed vent system and any
combination of control devices meeting the requirements of 40 CFR part 63 subpart SS of this
part, as specified in §63.982(a)(l). The commenter pointed out that §63.982(a)(l) does not
70
-------
include a percentage; removal requirement .and "control device" lias a diflfef tent meaning in .stibpart
tbm,jn subpart YY. The commenter suggested revising the .Wording in column 3 of items' 2b
'.._ , ;_!.' ;- I, ' A ,;; ,~\, ' ';."'' . ': '' :- |k , . i'.'' [ ;,*' ,'."' , '!'»'.''l"V^. ' ' -ly. V-
3
of ite'fti 4 ispH^si ^Rillfce'e^iioiis of total-organic 'lfi\P Ijf^vMlgh^
"'.'' ' := "H*' ; <;
organic IIAP of TOC to a concenitjration of 20 parts per million by^ volume,; wfoefreVei'lJis jess
stringent; by ventuig-fri3.iisjsj.cms though a closed vent system to aay cdiiibuiation ^coilb'^l
.'-;." .'.-' - ': t',-.:: 'vL ' Y'v . Sj-' ils'^fv
devieiss meeting;the requllbi^ntg |f subpart 55 of tliis part, as specifiedin |63'.9f§2(a)iii) gf:
'.!.,,'/'! .'.' V ^~~ ' \k;'"'-'V '"''."? i:** f- :. ... -.
is to reduce emissions of totalrorgaaic HAP by 98 weight percent or rfeduce; organic HAP or TQC
-.' :--. -, ':." . :-T';;i '' i : <* . : J ' . 4'- M'./.-'^^.r't " - '
to a concentration of 20 plarts/per tnilb'on by volume, whichever is lessiistrinaeftl"
: - "-.;- i:1.. v^:;vr':!-: ' . :' --' ' 'fc, ,.;-': v..
Resp.orise: Table 71 hMbeem revised to address the commein};er'Js,concqrn.
!- - * '.: ; v : ( "' ?. '
2.9.5 Miscellaneous Corrections
Cornmeijit:' One copisrienter (IV-D-04) noted that, the list of 11 HAP in table 1 to 40 CFR
» . - i '
part 63 subpart XX is different than the list of 13 HAP in table 7 to §63;. 1103. The commenter
suggested revising the HAP listed in table 7 to match the HAP in table 1.
Response:, the list of HAP in table 7 has been revised to match .the HAP listed in table 1.
2.1.0 FORMAT OF RULE
Comment: One commenter (IV-D-04), in response to EPA's request for comments on the
use of the alternative generic methodology for determining MACT compliance, stated that the
Generic MACT is not appropriate for the ethylene production source category. The commenter
explained that in the solicitation for comments and in the preamble for the Generic MACT that
EPA indicated that the Generic MACT would be reserved for source categories with five of fewer
sources. The commenter noted that EPA has identified 37 ethylene production plants or which 30
or more are major sources and concluded that the Generic MACT is therefore not appropriate.
71
-------
Response: While the original intention may have been for GMACT to be used for source
categories with five or fewer somces, nothing precludes EPA from using it for categories with
greater than five sources. Trie EPA h,as fcurid that the GMACT provides an appropriate format
!th'emissjo;n tptis, astfcifite with ethyte)i#prQdu<
cons^^
'X r- -:i'i11 Hjj;[- ..; ^-feiivi- 'V'1'*:1- -! »''''
j-i'J- **"> ffflC". f-!i "'^ji'Vt^S t'f ;i-.'t i"1'
other rules.
Corntnent:. bne commenter (IV-D-OS) commended EPA's attempt to improve
coifflnunicajioris ^ifh the regulated community by using a common form of English to convey
, ' : ' ' ^ 1 ' : ' : ';rl : 1 ' : ' '. '- ' :-i:t ' ' ! M ' . '."
requirements^ Th,e cpmmenter suggested that the rule could be made eveii more readable if the
' '' . '''ilt i vi '' .: ' : .,..-. . v; ; ;.' :, !{';i ;
appUcability, aiid dejfinitidn sections are p|af ed at the beginning of the rule. The cbiininenter
explained that :tp;'J8rst understanding a source heeds to have is whether arjd how thejrule applies..
' '-; ' " I ' i ''{
''I : . '. ' ' :\ "' ' -: . i - . >
Response: E>0finifi6ns are how presented in the second section of subpart'XX- The EPA
' ' ''"
.,
believes that including separate applicability sections for heat exchange systems and waste directly
\ ' '...':.' ''ij/. '
prior to their respective requirements is preferable to providing a single applicability Section at the
beginning of the rule, with respect to readability.
Comment: One commenter (IV-D-02) contended that the reference approach used in 40
CFR part 63 subpart XX is difficult to understand. The commenter suggested that, if the
unmodified BWON is not adopted, then EPA should revise subpart XX so that it includes all of
the requirements and does not reference 40 CFR part 61 subpart FF.
Response: The waste requirements of the ethylene production NESHAP have been
significantly revised since proposal. Major revisions include allowing the use of benzene as a
surrogate and allowing the use of compliance options. Due to these revisions, the ethylene
production waste requirements are now much more similar to the BWON. The modifications to
the BWON required to produce the ethylene production NESHAP are now minimal. On this
basis, the EPA has determined that the most effective use of resources is to cross-reference the
BWON, and that this approach provides sufficient clarity and ease of use.
72
-------
Entered 20100506
ELvl K Srce d
Form Conf 0
Cont GPub f
Ills Fest 0
Replaced 20 100506
' Audn dtrl
Biog MRec
LitF 0 Indx 0
DtSt s Dates 2002 ,
Lang eng
Ctry ncu
OC ._ Page 1 of 1
Local File C:\Documents and Settings\emcnearn\Application Data\OCLC\Connex\Db\DefaultBib.bib.db
OCLC 611982700 Held by EOB - no other holdings
Save File 3
Rec stat n
Type a
BLvl m
Desc a
040 EOB *c EOB
099 EPA450-R-02-002
049 EOBD
245 0 0 National emission standards for hazardous air pollutants : =t=b ethylene production : background information
document for final standards : summary of public comments and responses / +C U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards, Emission Standards Division.
260 Research Triangle Park, N.C.: *b U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards, Emission Standards Division, =t=c 2002.
300 72 p.; =f=c 28 cm.
500 "May 2002."
650 0 Air quality =l=x Standards =t=z United States.
650 0 Air *x Pollution =t=x Standards *z United States.
650 0 Air *x Pollution tz United States.
650 0 Air quality management =t=x Standards ta United States.
710 1 United States. *b Environmental Protection Agency. =t=b Office of Air Quality Planning and Standards. =t=b
Emission Standards Division (Research Triangle Park, NC).
Delete Holdings- Export- Label- Produce- Submit- Replace- Report Error- Update Holdings-C Validate-C
Source-Derived Workflow-In Process
abo 5/6/2010
------- |