United States
 Environmental Protection
 Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-453/R-96-001b
May 1996
Air
Hazardous Air Pollutant
Emissions from Process
Units in the Thermoplastics
Manufacturing Industry--

Basis and Purpose Document for
Final Standards, Summary of Public
Comments and Responses

-------

-------
                                            EPA-453/R-96-001b
    Hazardous Air Pollutant Emissions
         From Process Units in the
              Thermoplastics
         Manufacturing Industry--

      Basis and Purpose Document for
       Final Standards, Summary of
      Public Comments and Responses
        Emission Standards Division
   U.S. Environmental Protection Agency
        Office of Air And Radiation
 Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
                May 1996

-------
                          DISCLAIMER

This Report has been reviewed by the Emission Standards
Division of the Office of Air Quality Planning and Standards,
EPA, and approved for publication.  Mention of trade names or
commercial products is not intended to constitute endorsement
or recommendation for use.
                               ii

-------
               ENVIRONMENTAL PROTECTION AGENCY
 Hazardous Air Pollutant Emissions from Process Units in the
 Thermoplastics Manufacturing Industry -- Basis and  Purpose
         Background Information for Final Standards


      The  standards regulate organic hazardous air pollutant
      (HAP) emissions from the production of acrylonitrile
      butadiene styrene  (ABS) resin, styrene acrylonitrile
      (SAN) resin, methyl methacrylate acrylonitrile
      butadiene styrene  (MABS) resin, methyl methacrylate
      butadiene styrene  (MBS) resin, polystyrene resin,
      poly(ethylene terephthalate)  (PET) resin, and  nitrile
      resin.  Only those thermoplastic product process units
      that are part of major sources under section 112 (d) of
      the  Clean Air Act  (Act) will be regulated.

      For  additional information  contact:

      Mr.  Robert'  Rosensteel
      Organic Chemicals Group
      U.S. Environmental Protection Agency
      Research Triangle Park, NC   27711
      Telephone:  (919)  541-5608

      Paper copies of this document may obtained from:

      U.S. Environmental Protection Agency Library  (MD-36)
      Research Triangle Park, NC   27711
      Telephone:   (919)  541-2777

      National Technical Information Service  (NTIS)
      5285 Port Royal Road
      Springfield, VA   22161
      Telephone:  (703)  487-4650

t.     Electronic  copies  of this document may be obtained from
      the EPA Technology Transfer Network  (TTN) .  The TTN is
      an electronic bulletin board system which is free,
      except  for  the normal  long  distance charges.   To access
      the Basis and Purpose Document:

      •   Set  software communication setting to 8 bits,  no
          parity,  and 1  stop bit
      •   Set  a terminal emulation of either VT100, VT102,  or
          ANSI
      •   Baud rates of  1200,  2400,  9600, and  14,400  are
          accepted
      •   Use  access number  (919)  541-5742; access problems
          should be directed  to  the system operator at
           (919)  541-5384
                             ill

-------
Register online by providing a personal name,
password, and company name, address, and phone
number
Specify TTN Bulletin Board:  CAAA (Clean Air Act
Amendments)
Select menu item:  Title III:  Hazardous Air Poll
Select menu item:  Policy Guidance Documents
To download, type filename: GRIVB&P.ZIP
                     IV

-------
                       TABLE OF CONTENTS
Section                                                   £SSe
1.0  SUMMARY	•	1"1
     1.1  INTRODUCTION	1~1
     1.2  SIGNIFICANT COMMENTS AND CHANGES SINCE
          PROPOSAL   . ,	1~1
          1.2.1  Applicability Provisions and
                 Definitions	1~6
          1.2.2  Continuous  Process Vent Provisions  .  .  .  1-12
          1.2.3  Batch Process Vent Provisions   	  1-14
          1.2.4  Wastewater  Provisions	•  1-23
          1.2.5  Process Contact  Cooling Tower
                 Provisions	1-24
          1.2.6  Equipment Leak Provisions	1-28
          1.2.7  Emissions Averaging  Provisions 	  1-32
          1.2.8  Compliance  and Performance  Test
                 Provisions  and Monitoring Requirements .  1-33
          1.2.9  Recordkeeping and Reporting  	  1-37
2.0  APPLICABILITY	2"1
     2.1 SELECTION OF  SOURCE CATEGORY	•   2-1
     2.2 SELECTION OF  SOURCE	2-2
          2.2.1  Area Source	2-2
          2.2.2  Major Source ......  	   2-3
           2.2.3  Affected Source	   2-5
           2.2.4   Exemptions	• •  2-20
      2^3  SELECTION OF POLLUTANTS	2-24
      2.4  SELECTION OF THE STANDARDS (GENERAL)   	2-27
      2.5  COORDINATION WITH OTHER CLEAN AIR ACT
           REQUIREMENTS	2-33
           2.5.1  Subpart A	2-33
           2^5.2  SubpartS F, G,  H, I  .	2-37
           2.5.3  Other NSPS	•  2-41
      2.6  SUPPORT FOR OTHER SUBMITTED COMMENTS  . .  .  . . 2-42
      2.7  DEFINITIONS	2-42
      2.8  BURDEN	2-48

-------
                 TABLE OF  CONTENTS,  CONTINUED
Section
2 . 9 COMMENT EXTENSION 	

3.0
4.0
5.0





6.0







7.0


8.0

2.10 COMPLIANCE ....... 	
STORAGE VESSELS 	
CONTINUOUS PROCESS VENTS 	 ......
BATCH PROCESS VENTS 	
5.1 SELECTION OF STANDARDS 	
5.2 APPLICABILITY 	
5.3 GROUP DETERMINATION 	
5 . 4 DEMONSTRATION OF COMPLIANCE 	 	
5.5 RECORDKEEPING AND REPORTING 	
POLY (ETHYLENE TEREPHTHALATE ) (PET) VACUUM SYSTEMS
6.1 EXISTING AFFECTED SOURCES 	
6.1.1 Ethylene Glycol Jets as Retrofit
Technology 	
6.1.2 Alternative Technologies .......
6.1.3 Vacuum System Wastewater Requirements
6.1.4 Miscellaneous 	
6.2 NEW AFFECTED SOURCES 	
WASTEWATER OPERATIONS 	
7.1 SELECTION OF STANDARDS 	
7.2 MISCELLANEOUS 	 	 	
EQUIPMENT LEAKS 	
o 1 fl-RT.F.CTION OF THE STANDARDS/APPLICABILITY . .
Page
. 2-51
. 2-53
3-1
4-1
. 5-1
. 5-1
. 5-4
. 5-7
. 5-12
. 5-18
. €-1
. 6-3

. 6-3
. 6-11
. 6-13
. 6-15
6-19
7-1
7-1
7-10
. 8-1
. 8-1
9.0
          8.1.1  Polystyrene and Methyl Methacrylate
                 Butadiene Styrene	8-1
          8.1.2  Poly(ethylene terephthalate)	8-12
     EMISSIONS AVERAGING	9-1
10.0 PERFORMANCE TESTING	10-1
11.0 MONITORING    	H-l
     11.1 OPERATING PARAMETER MONITORING	.11-1
     11.2 SELECTION OF MONITORED PARAMETERS	11-6
     11.3 EXCURSIONS    	11-10
     11.4 DEMONSTRATION OF COMPLIANCE	11-13
12.0 RECORDKEEPING AND REPORTING	12-1

                              vi

-------
                 TABLE OF CONTENTS, CONTINUED
Section
     12.1 GENERAL BURDEN	12-1
     12.2 RECORDKEEPING	  12-3
     12.3 REPORTING	I2"9
13.0 POLYMERS NSPS BASED REQUIREMENTS  FOR
     PROCESS VENTS AT PET AND  POLYSTYRENE AFFECTED
     SOURCES USING CONTINUOUS  PROCESSES  	  13-1
     13.1 SELECTION  OF STANDARDS	13-1
     13.2 APPLICABILITY .	13-9
     13.3 DEMONSTRATION OF  COMPLIANCE	  13-19
     13.4 MISCELLANEOUS	13-21
14.0 OTHER	14~1
                               VI1

-------

-------
                 ACRONYM AND ABBREVIATION LIST
ACRONYM

ABS

ACT

AMSAN/ASA


CAA

CAS

CBI

CFR

CMA

COD

CRC

CTG

DMT

EPCRA


EPS

FR

HAP

HON

ICR

LDAR

MABS


MACT

MBS

NDE
TERM

acrylonitrile butadiene styrene

Alternative Control Technology

alphamethyl styrene acrylonitrile/
acrylonitrile styrene acrylate

Clean Air Act

Chemical Abstracts Service

confidential business information

Code of Federal Regulations

Chemical Manufacturer's Association

chemical oxygen demand

Chemical Rubber Company

control techniques guidelines

dimethyl terephthalate

Emergency Planning and Community
Right-to-Know Act

expandable polystyrene

FEDERAL REGISTER

hazardous air pollutant

hazardous organic NESHAP

information collection request

leak detection and repair

methyl methacrylate acrylonitrile butadiene
styrene

maximum achievable control technology

methyl methacrylate butadiene styrene

no  detectable emissions
                             viii

-------
NESHAP

NSPS
OEL
PET
PRO
pH
POM
PQL
RACT
RCRA
RCT
RDI
RIA
SAN
SID
SOCMI

SPI
SSP
TOC
TPA
TPPU
 TRE
 TRI
 TTN
 VOC-
 VOHAP
national emission standards for hazardous air
pollutants
new source performance standards
open-ended lines
poly(ethylene) terephthalate
pressure relief devices
hydrogen-ion  concentration
polycyclic organic matter
practical quantitation limit
reasonably available  control technology
Resource Conservation and Recovery Act
reference control technology
rupture disks
regulatory  impact analysis
styrene acrylonitrile
supplementary information  document
synthetic  organic chemical  manufacturing
industry
Society of the Plastics  Industry,  Inc.
 solid state polymerization
 total organic compound
 terephthalic acid
 thermoplastic product process unit
 total resource effectiveness
 toxic release inventory
 Technology Transfer Network
 volatile organic compound
 volatile organic hazardous air pollutant
                                IX

-------
WTU
                 wastewater treatment unit
ABBREVIATION



Btu/hr



C



cfm



F



gpm



kg/yr



kPa



minHg



MW



ppm



ppmv



psia



tpy
UNIT OF MEASURE



British thermal units per hour



degrees Celsius



cubic feet per minute



degrees Fahrenheit



gallons per minute



kilograms per year



kilopascals



millimeters of mercury



megawatt



parts per million



parts per million by volume



pounds per  square inch, absolute



tons per year
                                x

-------

-------
                          1.0   SUMMARY

 1.1   INTRODUCTION
      On March 29,  1995,  the U.S.  EPA proposed NESHAP  for HAP
 emissions from the production of  ABS resin, SAN resin,  MABS
'resin,  MBS resin,  polystyrene resin, PET  resin, and nitrile
 resin under Section 112(d) of the Act.  Public comments were
 requested on the proposed standard and  comment letters  were
 received from industry representatives  and governmental
 entities.  A total of 27 comment  letters  were received.
 Table 1-1 presents a listing  of all persons submitting  written
 comments, their affiliation,  and their  docket number.  A
 public hearing was not requested.
      The written comments that were submitted on  the  proposed
 rule have been summarized and responses have been provided,
 and both are included in the  following  chapters.   This  summary
 of  comments and responses serves as the basis for revisions
 made to the NESHAP between proposal and promulgation.
 1.2  SIGNIFICANT COMMENTS AND CHANGES SINCE PROPOSAL
      In response to comments  received on  the proposed
 standards, changes have been made to the  final  standards.
 While several of these changes are clarifications designed to
 make the EPA's intent clearer, a number of them are
 significant changes to the requirements of the  proposed
 standards.  A summary of the substantive  comments and/or
 changes made since the proposal are described in the following
 sections.  The rationale for these changes and detailed
 responses  to public comments are  included in the Basis and
 Purpose  Document  for the final standards.  Additional
 information  on the final standards  is contained in the docket
 for  these  standards  (see ADDRESSES  section of this preamble).
                               1-1

-------
TABLE 1-1.
LIST OF. COMMENTERS ON PROPOSED NATIONAL EMISSION
    STANDARDS  FOR HAZARDOUS AIR POLLUTANTS
 Docket number3-
                Commenter and affiliation
     IV-D-01       Pamela  S.  Dolbee
                   Regulatory Compliance Engineer
                   Dart  Container Corporation
                   Mason,  Michigan 48854

     IV-D-02       Dale  L.  McKinnon
                   Technical  Director
                   Manufacturers  of Emission Controls
                     Association
                   1707  L  Street,  N.W., Suite 570
                   Washington,  D.C.  20036-4201

     IV-D-03       James W. Walpole
                   Plant Manager
                   BP Chemicals Inc.
                   Ft. Amanda Road
                   P.O.  Box 628
                   Lima, Ohio  45802-0628

     IV-D-04       Linda J. Liszewski
                   Manager, Environmental  Issues
                   Eastman Kodak Company
                   Rochester, New York  14652-6263

     IV-D-05       Doyle R. Pendleton
                   Acting  Deputy Director,  Office of Air Quality
                 '  Texas Natural Resource  Conservation
                     Commission
                   P.O.  Box 13087
                   Austin, Texas  78711-3087

     IV-D-06       Timothy J. Hebrink
                   Senior Environmental Engineer
                   3M Environmental  Engineering and Pollution
                     Control
                   Building 41-01-05
                   P.O.  Box 33331
                   St. Paul,  Minnesota  55133-3331

     IV-D-07       Neal E. Thurber
                   Director,  Environmental Planning &
                     Regulations
                   Amoco Oil Company
                   200 East Randolph Drive
                   Chicago, Illinois  60601-7125	
                              1-2

-------
TABLE 1-1.  LIST OF COMMENTERS ON PROPOSED NATIONAL EMISSION
      STANDARDS FOR HAZARDOUS AIR POLLUTANTS  (CONTINUED)
  Docket number3
          Commenter and affiliation
      IV-D-08
      IV-D-09
      IV-D-10
      IV-D-11
      IV-D-12
      IV-D-13
      IV-D-14
Maureen A. Healey
Director, Federal Environment and
  Transportation Issues
The Society of the Plastics, Inc.
1275 K Street, N.W., Suite 400
Washington, D.C.  20005-4006

Charles W. Keffer, Jr.
Director, Regulatory Management
Monsanto Company
800 N. Lindbergh Boulevard
St. Louis, Missouri  63167

Matthew O. Tanzer/Bernadine D. Flood
Environmental Counsel/Senior Environmental
  Engineer
GE Plastics
Legal Operation
General Electric Company
One Plastics Avenue
Pittsfield, Massachusetts  01201

M.L. Mullins
Vice President-Regulatory Affairs
2501 M Street, NW
Washington, DC  20037

Michael J. Wax
Deputy Director
Institute of Clean Air Companies
1707 L Street, N.W., Suite 570
Washington, DC  20036-4201

Thomas X. White
Associate Vice President
Pharmaceutical Research and Manufacturers of
  America
1100 Fifteenth Street, NW
Washington, DC  20005

Thomas R. Herman, P.E.
Corporate Environmental Manager
Zeon Chemicals Incorporated
P.O. Box  34320
Louisville, Kentucky  40232-4320	
                              1-3

-------
TABLE 1-1   LIST OF COMMENTERS ON PROPOSED NATIONAL EMISSION
      STANDARDS FOR HAZARDOUS AIR POLLUTANTS  (CONTINUED)
  Docket nutnbera
          Commenter and affiliation
      IV-D-15
      IV-D-16
      IV-D-17
      IV-D-18
      IV-D-19
      IV-D-20
       IV-D-21
M.A. Piersante
Senior Environmental Engineer
Chevron Chemical Co.
Environmental & Health Protection
6001 Bellinger Canyon Road
San Ramon, California  94583

Rand N. Shulman
Vice President
Health, Safety and Environment
Shell Chemical Company
One Shell Plaza
P.O. Box 2463
Houston, Texas  77252-2463

Joe L. Ritchie
President
Hoechst Diafoil Co.
Hood Road, P.O. Box  1400
Greer, South Carolina  29652

David W Gustafson/Toby A. Threet
Env. Health & Regulatory Affairs/Legal
   Department
Dow Chemical Company
2030 Dow  Center
Midland,  Michigan  48674

J.E. Cooper
Director  Environmental Quality
AlliedSignal Inc.
P.O. Box  1139
Morristown, New Jersey   07962-1139

Susan  P.  Engelman
Vice President
Environmental, Health &  Safety  Affairs
Hoechst Celanese Corporation
Route  202-206
P.O. Box  2500
Somerville, New Jersey   08876-1258

Richard C.  Phelps
Environmental Associate
Eastman Chemical Company
P.O. Box  1993, FANB-4
Kingsport,  Tennessee  37662	
                               1-4

-------
TABLE 1-1.  LIST OF COMMENTERS ON PROPOSED NATIONAL EMISSION
      STANDARDS FOR HAZARDOUS AIR POLLUTANTS  (CONTINUED)
  Docket numbera
          Commenter and affiliation
      IV-D-22
      IV-D-23
      IV-D-24
      IV-D-25
      IV-D-26
      IV-D-27
Jim Burt
Wellman, Inc.
P.O. Drawer 188
Highway 41/51 North
Johnsonville,  South Carolina
                                                  29555-0188
Richard A. DiMenna
Manager, Air and SARA Programs
Environmental Regulatory Affairs Department
Rohm and Haas Company
Independence Mall West
Philadelphia, Pennsylvania  19105

Edward J. Sweeney
Plant Manager
BASF Corporation
P.O. Box 550
Joliet, Illinois
                                      60434-0550
John A Dege Jr.
Manager-Air Programs
DuPont SHE Excellence Center
1007 Market Street
Wilmington, Delaware  19898

James Frick
Senior Environmental Engineer
Firestone
Fibers & Textile
105 Winston Churchill Drive
Hopewell, Virginia  23860

Michael  J. Wax
Deputy Director
Institute of  Clean  Air  Companies
1707  L Street,  N.W., Suite 570
Washington, DC 20036-4201	
 a The docket number for the Group IV thermoplastics docket is
   A-92-45.1
                               1-5

-------
1.2.1  Applicability Provisions and Definitions
1.2 .1." 1  Designation of Affected Source and the Definition of
Thermoplastic Product Process Unit
     Commenters expressed confusion about the definitions of
"affected source" and "TPPU" in the proposed standards.  The
EPA reviewed both definitions and agreed that they needed
clarification,  in response, the EPA has revised the language
describing "affected source" and "TPPU" for the final
standards.  The definition of "affected source" included in
§63.502 of the proposed standards was revised and the
definition now references §63.500(a) [§63.1310(a)3,  and
§63.500(a) [§63.1310(a)] describes the affected source.  The
provisions in §63.500(a), which at proposal defined
applicability in terms of the existence of one or more TPPUs,
have been revised to define applicability in terms of the
affected source.  As part of this revision, the provisions in
proposed §63.500(b), which described the affected source, were
removed.   [Note:  In the proposed standards, the definition of
TPPU attempted to describe all the equipment and operations
that would be included in an affected source.  In the final
standards, the description of what the affected source
includes is contained in §63.500(a)  [§63.1310(a) ] . ]
     The existing affected source is defined as each group of
one or more TPPUs that manufacture the same thermoplastic
product as their primary product, and  (1) are located at a
major source plant site,  (2) are not exempt, and  (3) are not
part of a new affected source.  This means that each plant
site will have only one existing affected source in any given
subcategory.
     New affected sources are created under various
circumstances.   If a plant site with an existing affected
source producing thermoplastic A as its primary product
constructs a new TPPU also producing thermoplastic A as its
primary product, the new TPPU is a new affected source if the
new TPPU has the potential to emit more than 10 tons per year
of a single HAP  or 25 tons per year of all HAP.  In this
situation, the plant  site would have an existing affected
                              1-6

-------
source producing thermoplastic A and a new affected source
producing thermoplastic A.  Each subsequent new TPPU with   .
potential HAP emissions above major source levels  (i.e., 10/25
tons per year) would be a  separate new affected source.  New
affected sources are also  created when a TPPU  is constructed
at  a major  source plant site where the thermoplastic product
was not previously produced, with no regard to the potential
HAP emissions from the TPPU.  This approach to defining new
affected source was selected in order to make  this subpart
more consistent with the HON.
     Another instance  where a new affected source  is created
is. if a new TPPU is constructed at a new plant site  (i.e.,
green field site) that will be a major source. The final
manner in which a new  affected source is created is when an
existing affected source undergoes reconstruction, thus making
the previously existing  affected source subject to new source
standards.
     This  standard  differs from the HON, however,  in that  it
applies to  multiple source categories.  Thus,  unlike the HON,
a newly added TPPU  at  a  facility is covered by this rule even
if that TPPU is  in  a  different source category from the
existing TPPUs at the  facility.  It is the EPA's position  that
the addition of  a process  unit  in  a different  source category
' is a new  source  and must meet  the  requirements for new sources
even though the  TPPU has the potential to  emit less than 1-0
tons per year of  a  single HAP or 25 tons per year  of all HAP.
 Indeed,  if a source covered by another MACT  standard  (i.e., a
different  source  category) were built at a HON facility, that
source would be  subject to new source requirements under that
MACT standard.
     Also,  each affected source  includes the  following
 emission points  and equipment that are  associated with each
 group of  TPPU:   (1) each wastewater stream;
  (2) each wastewater operation;  (3)  each heat  exchange  system;
  (4) each process contact cooling tower used in the manufacture
 of PET that is associated with a new affected source;  and   (5)
 each process contact  cooling tower used in the manufacture of
                              1-7

-------
PET using a continuous terephthalic acid high viscosity
multiple end finisher process that is associated with an
existing affected source.
     For the final standards, the number of existing affected
sources present at a plant site will equal the number of
thermoplastic products manufactured at that plant site.  A
plant site manufacturing 3 different thermoplastic products
has 3 existing affected sources.   [Note:  Each different
thermoplastic product represents a different subcategory, and
each subcategory comprises a separate existing affected
source.]  The number of existing affected sources at a plant
site could range from 1 to 19.
     The definition of TPPU was revised and now includes a
list of the collection of equipment that comprises a TPPU.
This equipment includes process vents from process vessels,
storage vessels, and equipment subject to the equipment leaks
provisions.  Because wastewater streams, wastewater
operations, heat exchange systems, and process contact cooling
towers are equipment that are often used by more than one
TPPU/ these items are not included as part of the definition
of TPPU.   Instead, said items are  included in the definition
of affected source.  Because each  wastewater stream can be
traced back to the TPPU from which it originated, each
wastewater stream can be associated with an affected source.
On the other hand, wastewater operations may service
wastewater streams associated with more than one affected
source,  just as heat exchange systems and PET process contact
cooling  towers could service multiple affected sources.
Therefore, for wastewater operations, heat exchange systems,
and  PET  process contact cooling towers, the final rule does
not  require  an owner or operator to designate said emission
points  and equipment to be associated with a particular
affected source.   Instead, said emission points and equipment
are  subject  to all applicable requirements associated with
each affected source that said emission points and equipment
may  service.   In  a simple example, a heat exchange system is
associated with two  affected sources that are both subject to
                              1-8

-------
the final rule.  The owner or operator must comply with the
provisions for heat exchange systems contained in the final
rule.  In a more complex example, a piece of wastewater
operations equipment services wastewater streams from two
affected sources subject to the final rule and from one source
subject to the Hazardous Organic NESHAP (HON).  This piece of
wastewater operations equipment must comply with both the
final rule and the HON.
1.2.1.2  Definition of Organic HAP
     Numerous commenters recommended that the EPA restrict the
list of organic HAP in the final standards to those that are
used or are present in significant quantities at TPPUs or
those that are listed in the HON, subpart F, table 2.  The EPA
agreed with the commenters suggestion that a table providing a
listing of the specific organic HAP expected to be regulated
for each subcategory covered by the standards should be added
to the final standards.  Therefore, the definition of organic
HAP was revised to  specify those organic HAP that are known to
be used or present  in significant quantities for each
subcategory, thereby restricting the organic HAP regulated by
the  final standards.  This list  is provided  in Table 2 of the
final standards.
     The revised  definition of organic HAP was developed using
available process description information received from
industry and gathered  from available literature.  Because
there may be additional organic  HAP present  at an affected
source, the final standards require owners or operators to
identify the presence  of  any additional organic HAP based on
the  following  criteria:   (1) the organic HAP is knowingly
introduced  into  the manufacturing process other than as an
impurity, or has  been  or  will be reported under any Federal or
State program,  such as EPCRA Section  311, 312, or 313 or
Title V;  and  (2)  the organic HAP is  listed  in Table 2 of
subpart F.                                 •
1.2.1.3  Determining New  Source  Status
      The  EPA received  comments  regarding the procedure  for
determining if new or  existing  source  requirements would apply
                              1-9

-------
to a particular  TPPU.   In response to those comments, the EPA
has revised the  provisions  in the final  standards.
     Under the final standards, new affected sources are
created under each of  the following four situations:   (1) if a
plant  site with  an existing affected source producing a
thermoplastic product  as its primary product constructs a new
TPPU also producing the same thermoplastic product as its
primary product, the new TPPU is a new affected source if the
new TPPU  has the potential  to emit more  than 10 tons per year
of a single HAP  or 25  tons  per  year of all HAP;  (2) when a
TPPU is constructed at a major  source plant site where the
thermoplastic product  was not previously produced as the
primary product  of an  existing  affected  source;  (3) if a new
TPPU is constructed at a new plant site  (i.e., green field
site)  that will  be a major  source; and  (4) when an existing
affected  source  undergoes reconstruction, thus making the
previously existing affected source subject to new source
standards.
     These revisions  reflect the EPA's  intent that new source
•requirements apply if  the added TPPU has the potential to emit
major  quantities, as  in the HON, or the  added TPPU is a new
affected  source.  The  HON applied  to only one source category,
and it was not  possible to  add  a process unit subject to the
HON that  was in a new source category.   Therefore, the only
differentiation to be  made  was  between process units emitting
major  quantities of organic HAP and those not emitting major
quantities of organic HAP.   'On  the other hand, the
thermoplastics  standards apply  to  multiple source
categories/subcategories,  and it  is possible to add a TPPU
subject  to the thermoplastics standards  that is in a new
source category/subcategory. For  this  reason, if a TPPU is
added  to an existing plant  site and  said TPPU manufactures a
thermoplastic product as its primary product not previously
produced at the plant site as  the  primary product of an
existing affected source,  that  TPPU,  regardless of emissions,
is a new affected source at that  plant  site.
                              1-10

-------
1.2.1.4  Solid State PET Processes
     Commenters contended that all PET solid state
polymerization units, including collocated units, should be
exempted from regulation.  They stated that PET solid state
polymerization units are a vastly different technology than
dimethyl terephthalate  (DMT) and terephthalic acid (TPA)
processes and have different emission characteristics.   The
EPA has concluded that PET solid state processes are distinct
from DMT or TPA processes.  The EPA did not collect data on
PET solid state processes, and it was not possible to conduct:
the required analyses for regulating PET solid state
processes.  Therefore, the final standards do not regulate
these processes.  However, these processes may be regulated in
a future standard.
1.2.1.5  Flexible Operation Units
     The flexible operation unit provisions included in the
proposed standards, which were modeled after the HON, have
been retained in the final standards.  Under these provisions,
an owner or operator of a process unit that is designed and
operated as a flexible  operation unit will commit to being
subject to this rule or not being subject to this rule based
on a five-year projection of products to be manufactured and
production quantities.
     These provisions were modified to provide clarification
of the EPA's intent and flexibility in complying with the
provisions.  Under the  final rule, once an owner or operator
commits to being subject to this rule, there are two options
for complying.  Under the first option, an owner or operator
shall determine the group status  (i.e., Group 1 or Group 2) of
each emission point based on the production of the expected
primary product  (i.e.,  the thermoplastic product that
convinced the owner or  operator to commit to being subject to
this rule) .  Once the group status of each emission point is
determined, the owner or operator shall comply with the
applicable emission  standards at all times, regardless of what
product is being manufactured.  Under the second option, an
owner or operator shall determine the group status of each
                             1-11

-------
emission point each time a different product is being
manufactured, regardless of whether or not said product is a
thermoplastic product.  The EPA recognizes that neither option
is an ideal situation.  Under the first option, an owner or
operator may find themselves operating a control device to
control a Group 1 emission point that has none to negligible
emissions when a; different product is being manufactured.
Under the second option, an owner or operator may find
themselves performing multiple group determinations.  Again,
the EPA recognizes that neither option is an ideal situation,
but believes the tradeoff between these inconveniences and
flipping in and out of separate MACT standards is worthwhile.
     As part of demonstrating compliance with the rule, an
owner or operator required to operate a control device must
establish parameter monitoring levels and conduct monitoring.
Under either option of the flexible operation units
provisions, an owner or operator must establish multiple sets
of parameter monitoring levels to reflect the manufacture of
different products.  These provisions allow an owner or
operator to demonstrate that the parameter monitoring levels
established  for the primary product are appropriate for the
manufacture  of other products.  If this is not the case, the
provisions require that unique parameter monitoring levels be
established.
1.2.1.6  Coordination with Other Clean Air Act Requirements
     At proposal, the EPA had proposed to amend subpart DDD of
40 CFR part  60 by removing all references to polystyrene and
PET  facilities.  This action was being taken because the
proposed thermoplastics standards would supersede the
requirements in subpart DDD for polystyrene and PET affected
sources after the compliance date of the thermoplastics
standards.   Commenters  also suggested that subpart I of 40 CFR
part 63 be amended by removing all references  to MBS and MASS
affected  sources after  the  compliance date of  the
thermoplastics standards.   Other commenters requested that the
EPA further  clarify  that  after the compliance  date of the
                              1-12

-------
thermoplastics standards, affected sources will no longer be
subject to certain NSPS.
     The EPA clarified the relationship between the
thermoplastics standards and existing applicable standards in
§63.501 [§63.1311] of the final rule.  The final rule was
revised to state that affected sources subject to both the
thermoplastics standards and another subpart are to comply
with the provisions of the thermoplastic standards only after
the compliance date for the thermoplastic standards, for those
standards listed in §63.501(g) through (1) [§63.1311(g)
through (1)].  Further, after the compliance date for these
standards, these affected sources will no longer be subject to
the other subparts.  The EPA determined that a clear
understanding can be provided in these standards without
making modifications to other subparts.  Thus, the proposed
amendments to subpart DDD were not made as part of the final
rulemaking, nor were the suggested amendments made to subpart
I.
1.2.2  Continuous Process Vent Provisions
1.2.2.1  Reorganization of the Standards to Distinguish
Between Continuous Process Vents Subject to Provisions from
the HON and Continuous Process Vents Subject to Provisions
Adapted from the Polymer Manufacturing NSPS
     To better distinguish between the various requirements
for continuous process vents, the proposed standards were
reorganized.  In the final standards, separate sections apply
to the following subcategories:  those required to comply with
subpart G of the HON and those producing PET or polystyrene
using a continuous process.  In the final standards, §63.505-1
[§63.1315] references subpart G; §63.505-2 [§63.1316]  through
§63.505-6  [§63.1320] apply to select continuous process vents
at affected sources producing polystyrene and PET using a
continuous process.  Further, because not all process vents at
affected sources producing polystyrene and PET using a
continuous process are subject to §63.505-2 [§63.1316] through
§63.505-6  [§63.1320], the provisions of §63.505-2  [§63.1316]
                             1-13

-------
designate which process vents are subject to §63.505-2
[§63.1316] through §63.505-6
[§63.1320], which are subject to §63.505-1 [§63.1315]
(i.e., the HON), and which are subject to §63.506 [§63.1321]
(i.e., the batch process vent provisions).  This
reorganization is one way the EPA has clarified the standards
between proposal and promulgation to reduce complexity and
eliminate potential confusion.
1.2.2.2  Applicability of Emission Limits for Process Sections
from §63.505-2  [S63.13161 to the Collection of Process
Sections at an Affected Source
     Commenters stated that the regulatory construction of the
proposed standards implied that the process vent emission
limits adapted from the Polymer Manufacturing NSPS  [proposed
§63.505(b) and  (c) ] apply to each collection of material
recovery-sections, raw material preparation sections, and
polymerization sections, respectively, within an affected
source and not to each individual process section
(e.g., material recovery section), as under the Polymer
Manufacturing NSPS.
     At proposal, the EPA had intended for each individual
process section to meet the emission limits in proposed
§63.505(b) and  (c), as applicable.  However, since proposal,
the EPA has determined that revising the proposed standards to
allow each collection of process sections within an affected
source to meet the applicable emission limit would simplify
compliance while  achieving the same emission reductions.
Therefore, the  final standards apply the emission limits
adapted from the  Polymer Manufacturing NSPS to each collection
of material recovery sections, raw material preparation
sections, or polymerization reaction sections, as appropriate,
within an affected source.
                              1-14

-------
1.2.2.3  Clarification of Compliance Demonstration Provisions
for Final Condenser Temperature Limits
     Commenters suggested modifying the provisions adapted
from the Polymer Manufacturing NSPS that provide for a
demonstration of compliance by limiting the final condenser
outlet temperature.  Commenters explained that the reporting
provisions in the Polymer Manufacturing NSPS state that the
temperature limit is only exceeded when the average condenser
outlet temperature for a 3-hour period is more than 6 °C above
(i.e., warmer) the average operating temperature established
during the most recent performance test at which compliance
was demonstrated.  Commenters requested that the final
standards incorporate those monitoring, test method, and
recordkeeping and reporting requirements from the Polymer
Manufacturing NSPS that provide this flexibility (i.e, the six
degree window).
     The EPA intended for the proposed standards to be
equivalent to the Polymer Manufacturing NSPS in this regard
and have revised the final standards to provide the desired
flexibility  (i.e., the six degree window).  In addition, the
EPA has disassociated the six degree window from the results
of the performance test and has instead associated it with the
applicable temperature limit in the standard.  The final
standards allow all owners or operators complying with the
final condenser operating temperature limits to be 6 °C warmer
than the applicable temperature limit for the 3-hour averages.
The EPA considered that the proposed provisions did not
achieve an even-handed implementation of the requirements
because some affected sources would be allowed to have 3-hour
averages at warmer temperatures than others because their
performance test results indicated a temperature closer to the
applicable temperature limit.
1.2.3  Batch  Process Vent Provisions
1.2.3.1  Exemption of Certain Batch Process Vents
     Commenters supported the use of cut-offs for the group
status determination for batch process vents as found in
proposed §63.506-2(d).  Specifically, Commenters agreed that
                             1-15

-------
low annual organic HAP emissions and low flow rate cut-offs
are suitable.  Commenters explained that batch processes are,
by nature, suited to low volume production and the manufacture
of specialty products, and as such, low flow, low emitting
process vents are likely in batch operations.
     These provisions were retained in the final rule with one
exception.  The EPA removed the requirement to determine the
volatility class  (i.e., low, medium, or high) for batch
process vents.  As a result, there is a single minimum
emission  level cut-off in the final provisions of §63.506-2(d)
[§63.1323 (d)] .  In addition, the EPA chose to add a minimum
emission  level of 225 kg/yr to the definition of batch process
vent.  This modification made the batch process vent
provisions more consistent with the continuous process vent
provisions which have a minimum organic HAP concentration
level  as  part of the definition of continuous process vent.
An emission point with emissions equal to or less than
225 kg/yr is not considered a batch process vent.  At
proposal, the 225 kg/yr level was part of the batch process
vent group determination procedures; Group 2 batch process
vents  with annual emissions less than 225 kg/yr were subject
to reporting  requirements related to process changes.
1.2.3.2   Revisions to Group Determination Procedures
     Commenters suggested changing the group determination
provisions to only utilize emissions data from a TPPU' s
primary product.  In  addition, it was requested that batch
process vent  group determinations be performed on an annual
basis  instead of  for  every process change.   Commenters stated
that the  proposed group determination provisions were
considerably more complex than the continuous process vent
group  determination provisions.  Commenters  felt that, not
only did the batch process vent group determination provisions
require  an owner or operator  to obtain emissions data for
every  product that  it manufactured, but  even the most minor
process  changes (i.e.,  lengthening cycle times, altering
process  temperatures  and pressures, etc.) triggered the need
for  a  new group determination to be performed.  Given the
                              1-16

-------
inherent process variability associated with batch operations,
commenters contended that it would be very difficult to
perform a group determination.  Furthermore, because batch
units often need to implement sudden process changes in
response to customer demands, the proposed provisions could
potentially require repeating the group determination exercise
several times in a single year.  Commenters explained that
such a situation would not only serve to complicate a batch
unit's compliance status, but could also adversely impact its
ability to remain competitive in the marketplace.
     Four issues related to the group determination procedures
were reviewed by the EPA:   (1) a request to perform the group
determination on the primary product,  (2) a request to perform
the group determination on an annual basis,  (3) an objection
that the group determination procedures require a new group
determination to be made whenever minor process changes occur,
and  (4) an objection to the requirement to perform the group
determination when a sudden process change is required.
     The EPA has considered the request to perform the group
determination on the primary product and agrees that this
would provide acceptable results from an environmental
perspective while simplifying the compliance requirements for
and improving the enforceability of the batch process vent
standards.  The final rule  contains provisions allowing the
owner or operator of an affected source to perform the group
determination for batch process vents based on annualized
production of a single product.  To ensure protection to the
environment, the final rule specifies that the highest organic
HAP emitting product must be used when determining the group
status based on a single product.
     In addressing the request that the group determination be
required on an annual basis  instead of for every process
change, the EPA believes the proposed rule was clear on this
point.  The proposed rule required that emissions and average
flow rate be determined  on  an annual basis and describe how to
account for the production  of different products throughout
the year.  In this way,  the group determination  is done on an
                              1-17

-------
annual basis and can account for expected changes in the
product being produced.  The final rule does not reflect any
changes related to this specific issue.
     The third issue raised was an objection to the
requirement that a new group determination be performed
whenever minor process changes occur (e.g., lengthening cycle
times, altering process temperatures and pressures, etc.).
The proposed rule addressed the issue of minor process changes
as they affect Group 2 batch process vents.  If a process
change affecting a Group 2 batch process vent occurs, a group
determination must be made.  However, the group determination
provisions state that "changes that are within the range on
which the original group determination was based" are not
considered process changes.  This allows an owner or operator
to perform the initial group determination considering the
potential for minor process changes.  The EPA believes that
the proposed provisions were clear that minor process changes
 (e.i., variations in operating conditions) do not require that
a new group determination be performed.  Addressing this
concern as it relates to Group 1 batch process vents, the
proposed provisions do not require a redetermination of.group
status for Group 1 batch process vents under any
circumstances.  Therefore, if minor process changes were to
occur, the owner or operator would not be required to perform
another group determination.  The final rule does not reflect
any changes related to this specific issue.
     The fourth issue raised was an objection to the
requirement to perform a new group determination when a sudden
process change is required.  In light of the third issue
raised, EPA interpreted "sudden process change" to potentially
mean  (1) that a new product is being made, (2) that the same
product is being made in a fundamentally different way
 (e.g., with different raw materials), or  (3)  that the same mix
of products is being made but in a different proportion.  In
the first two cases, the EPA desires and intends that a new
group determination be made.  In the third case, the owner or
operator has the flexibility to consider this situation when
                             1-18

-------
performing the initial group determination.  If this situation
was not considered, then a new group determination would be
required.  The EPA feels that these types of process changes
warrant a new group determination to ensure that the emission
standards are being met.  The final rule does not reflect any
changes related to this specific issue.
1.2.3.3  Emissions Testing and Performance Testing
     Commenters requested that more flexibility should be
allowed in designing an emissions testing scheme for batch
process vents.  Commenters cited an example, provided as part
of the proposed definition of batch emission episode, where
the charging of a vessel and the heating of the same vessel
are considered two distinct batch emission episodes.  In this
example, the definition of batch emission episode would
necessitate that separate emissions measurements be made for
the charging and the heating of the vessel.  This would
require that a large number of samples be taken to
characterize processes  that have multiple, short duration
process steps.  Commenters felt that the flexibility to test
the emissions from several steps as a  single batch emission
episode would reduce testing costs without jeopardizing the
quality of the emissions  data.  It was suggested that three or
more batch cycles  could be tested to obtain a  representative
average  emission rate  for the batch cycle.
     After consideration  of this comment,  the  EPA chose to
leave  the provisions  related to emissions  testing of batch
process  vents unchanged as they relate to  this specific
comment.  The EPA  felt adequate data were  not  presented to
warrant  changing these provisions.  However, the emissions
testing  provisions in §63.506-2(b)  and §63.506-4(c)
 [§63.1323(b)  and  §63.1325(c)]  have  been  modified to provide
flexibility  and reduce the  burden of testing,  while continuing
to ensure  that  results are  satisfactory  for applicability
determinations  and performance tests.  The final provisions
allow an owner or operator to test  just  a  portion of the batch
emission episode  selected to be controlled when  the owner or
operator can demonstrate that emissions  during the period to
                              1-19

-------
be tested represent emissions for the entire batch emission
episode or are greater than the average emission rate for the
batch emission episode.
1.2.3.4  Flow Rate Estimation Procedures
     Commenters asserted that the equations and test methods
for calculating annual average flow rate in the proposed rule
were not warranted.  Commenters felt that the volumetric flow
rate testing methods and the requirement to measure flow every
15 minutes specified in proposed §63.506-2(e) were overly
burdensome and would not always provide representative
measurements.  It was suggested that average flow rates for a
batch emission episode are better defined by calculations of
displacement volumes with respect to the durations of the
displacement episodes or by other more simplified methods.
     The EPA agrees that there are more simplified and
potentially more accurate techniques for estimating flow rate
for batch process vents.  The final rule contains provisions
that allow engineering assessment, as well as testing, to be
used for estimating flow rate.  Also, for certain types of
batch emission episodes, equations for estimating flow rate -
have been added to the final rule.
1.2.3.5  Emissions Estimation Procedures
     Commenters recommended removing the emissions estimation
equations in proposed §63.506-2(b) from the rule.  Commenters
recommended that measurements or engineering estimates be
allowed in place of the equations.  It was felt the emissions
estimation equations would not allow the flexibility necessary
to account for differences in process technologies and
operating methods.
     Commenters also supported the provisions that allowed
owners or operators to use direct measurement or engineering
assessment to estimate emissions in cases where the emissions
estimation equations are inappropriate for a particular type
of operation or where, speaking to direct measurement, a more
refined estimate of emissions is necessary.  However,
commenters objected to the requirement to demonstrate that the
emissions estimation equations and direct measurement methods
                             1-20

-------
 are not appropriate before engineering assessment can be used.
      In response to the first issue raised, the emissions
 estimation equations have been retained in the final rule.
 The EPA found nothing in the public comments that would
 warrant removing these procedures.
      In response to the second issue, the EPA believes the
 data required to use the emissions estimation equations should
 be obtainable with reasonable effort.  Further, specific
 comments regarding the inaccuracy or inappropriateness of the
 equations were not made.  Given this, the EPA favors a more
 consistent estimation technique which is provided by the use
 of the emissions estimation equations, and the final rule
 requires the owner or operator to demonstrate that the
" emissions estimation equations are inappropriate before the
 use of engineering assessment is allowed.
      However, independent of the comments provided, the EPA
 has concluded that direct measurement of emissions may prove
 to be difficult and may or may not provide an increased
 assurance of accuracy over the use of engineering assessment.
 Therefore, if an owner or operator can demonstrate that the
 emissions estimation equations are not appropriate, the final
 provisions allow the use of either direct measurement or
 engineering assessment.
 1.2.3.6  Other Changes Resulting from EPA Review
      In  addition to changes made to  the proposed rule as a
 result of public comment, changes were made as a result of EPA
 independently reviewing the rule between proposal and
 promulgation.  Because the batch process vent provisions
 included in the rule are among the EPA's first attempts to
 regulate batch process vents, the EPA felt an ongoing,
 independent review of these provisions after proposal was
 warranted.  Changes resulting from this review are listed
 below:
       (1)  Allow applicability determinations and compliance
      demonstrations  (i.e., performance tests) to be based on
      total  organic compounds  (TOO or organic HAP.  Allow the
      use of Method 25A  to  compliment the use of TOG as a
                              1-21

-------
      potential basis for applicability determinations and
      compliance demonstrations.
      (2)  Allow the establishment of parameter monitoring
      levels to be based on performance tests or a combination
      of performance tests and engineering assessment
      (discussed in more depth in Section H,  Monitoring).   To
      accommodate this change, modifications to the batch
      process vent testing provisions were required.
      (3)  Add provisions specifying how the batch cycle
      limitation is to be determined.
      (4)  Change the reporting requirement for batch cycle
      limitation records from quarterly to annually.
 These changes are discussed in the paragraphs below.
      In the final rule, the EPA has allowed the use of TOG as
 the basis for applicability and compliance demonstrations
 (i.e., performance tests) as an alternative to organic HAP.
 The EPA has done this to provide flexibility to the regulated
 community and to reduce the overall burden of the rule.  The
 EPA considered the impacts of allowing TOG to serve as a
 surrogate for organic HAP in applicability and compliance
 demonstrations and did not find any negative impacts.
 Further, allowing the use of Method 25A as a complement to the
• use of TOG as a surrogate to organic HAP reduces the burden of
 implementing the final rule with little to no adverse impact
 on the measurement of pollutants in the regulated batch
 process vents.  To the best of the EPA's knowledge, the batch
 process vents regulated by this rule are predominantly organic
 HAP.  Also, with one exception  (i.e., ethylene glycol), the
 regulated organic HAP, which are listed in the definition of
 organic HAP found in the final rule, have response factors to
 Method  25A adequate to ensure  satisfactory measurement of TOG
 in the batch process vents.  For certain emission points where
 the EPA considered the presence, of ethylene glycol and its
 corresponding poor response  factor to call into question the
 results that would be obtained using Method 25A, the use of
 Method  25A is not allowed.   For all other emission points, the
 EPA has allowed the use of TOG for applicability and
                              1-22

-------
compliance demonstrations as an alternative to organic HAP in
the final rule.
     In the final rule, the EPA allows the establishment of
parameter monitoring levels to be based on either performance
tests, as in the proposed rule, or a combination of
performance tests, engineering assessment, and manufacturer's
recommendations.  This change affects all emission points
which are required to establish parameter monitoring levels,
including batch process vents.  The rationale for this change
is discussed in detail in Section H, monitoring, of this
document.  For batch process vents, this change in the
procedures for establishing parameter monitoring levels
necessitated changes to the performance test provisions.  When
an owner or operator chooses to establish parameter monitoring
levels based exclusively on performance tests, the final rule
directs that the performance test must include the entire
batch emission episode selected to be controlled.  As
discussed earlier, an owner or operator may choose to control
just a portion of a batch emission episode; in such a
scenario, the performance test must include the entire portion
of the batch emission episode selected to be controlled.
Alternatively, when an owner or operator chooses to establish
parameter monitoring levels based on a combination of
performance tests, engineering assessment, and manufacturer's
recommendations, the final rule allows an owner or operator to
test either the entire batch emission episode, or portion
thereof, selected to be controlled or to test only a period of
the batch emission episode that is less than the entire batch
emission episode, or portion thereof, selected to be
controlled.   [Note:  The flexibility to test a period of the
batch emission episode that is less than the entire batch
emission episode, or portion thereof, selected to be
controlled is discussed earlier in this section.]
     The final rule includes provisions specifying how the
batch cycle limitation, required for Group 2 batch process
vents, is to be established.  The EPA felt that the proposed
rule was ambiguous concerning the establishment of the batch
                             1-23

-------
cycle limitation and added these provisions to make the rule
more complete.  The added provisions provide additional
description of the purpose of the batch cycle limitation and
describe what documentation is required as part of
establishing the batch cycle limitation.
     In the final rule, the EPA changed the requirement for
reporting the number and type of batch cycles accomplished for
a Group 2 batch process vent from quarterly to annually.  The
EPA felt that quarterly reporting was unwarranted given that
the compliance requirement (i.e., the batch cycle limitation)
was on an annual basis.
1.2.4  Wastewater Provisions
1.2.4.1  Steam Stripping Stvrene-Containing Wastewater Streams
     Numerous commenters claimed that the selection of steam
stripping as the basis of the standards for the treatment of
styrene-containing wastewater streams was inappropriate due to
polymerization problems.  The EPA acknowledges that steam
stripping styrene-containing wastewaters may prove to be
impractical in some cases due to issues raised by the
commenters.  However, steam stripping is not required by the
standards.  Both the proposed and final wastewater provisions
provide several options for complying with the standards..  If
the owner or operator judges steam stripping to be impractical
for their process, one of the other wastewater compliance
options may be used.  The EPA considers one of these
compliance options, the use of enclosed sewers to a biological
treatment operation unit, a favorable option because styrene
is highly biodegradable.  Further, because the organic HAP
emitted from the subcategories regulated by this rule are
highly biodegradable, the EPA has determined that it is not
necessary to require affected sources to demonstrate that
95 percent of the mass of organic HAP listed in Table 9 of the
HON are removed when using a biological treatment unit, as
required by the HON.  Therefore, the final thermoplastic rule
does not require an owner or operator to make this
demonstration.
                             1-24

-------
1.2.4.2  Elimination of Regulations Pertaining to Wastewater
from Polystyrene Affected Sources
     In addition to considering the comments made concerning
the impracticality of steam stripping styrene-containing
wastewaters, the EPA evaluated the need to regulate wastewater
from polystyrene affected sources.  Because the water
solubility of styrene is limited to approximately 300 ppm and
styrene is the only known organic HAP emitted during the
production of polystyrene, it is not possible for a wastewater
stream that only contains styrene to meet the Group 1
applicability criteria  (i.e., 1,000 ppm minimum
concentration).  Therefore, because, to the best of EPA's
knowledge, there can be no Group 1 wastewater streams at a
polystyrene affected source, the final standards do not
regulate wastewater from this subcategory.
1.2.5  Process Contact Cooling Tower Provisions
1.2.5.1  Ethvlene Glvcol Jet Retrofit for PET Existing
Affected Sources
     Commenters disagreed with the EPA's position that
ethylene glycol jets are the vacuum system technology of
choice for a retrofit application and with the EPA
determination that ethylene glycol jets are cost effective,
Commenters contended that the EPA had failed to consider
numerous cost factors and design considerations.   [Note:
While the proposed standards did not require the use of
ethylene glycol jets, they were the technology on which the
prohibition of process  contact cooling towers was based.]
     Based on the data  available at proposal, the costs and
emission reductions achievable through the use of ethylene
glycol jets in retrofit situations were acceptable.  The EPA
knew that the proposed  standards for ethylene glycol jets were
based on limited data and limited knowledge obtained from one
manufacturer of PET.  The  EPA took  special effort to make this
clear and to solicit comments on the installation and
operation of ethylene glycol jets in the preamble to the
proposed standards  (see pages 16104 and 16107 of the
preamble).  Based on the  comments provided, the EPA agrees
                             1-25

-------
 that ethylene glycol jets are not a suitable  retrofit
 technology for existing affected sources,  either  technically
 or economically, for meeting the provisions of  the  proposed
 standards that prohibit the use of process contact  cooling
•towers.  These comments are presented and discussed in detail
 in the Basis and Purpose Document for the final standards.
      As explained in chapter 6.0 of the Basis and Purpose
 Document for the final standards, steam jet vacuum  systems can
 be used to create a vacuum on the process vessels,  contact
 condensers can be used to condense steam,  and the contact
 condenser effluent can be recirculated to the process  contact
 cooling tower.  In the process contact cooling  tower,
 stripping and drift may occur, resulting in organic HAP
 emissions to the atmosphere.  Volatilization  of organic HAP
 from the vacuum system may also occur.  The  following
 paragraphs describe some of the public comments on  this issue
 and the EPA's responses.
      Given that the EPA was convinced by the  commenters'  •
 arguments that ethylene glycol jets are not  a suitable
 retrofit technology for existing affected sources,  the EPA
 considered alternate options for controlling emissions from
 the vacuum system.  Both VOHAP and ethylene  glycol  are emitted
 from the vacuum system, and the EPA chose to approach  each of
 these  emissions separately.
      To address VOHAP emissions from the vacuum system at •
 existing sources, the final standards treat  the contact
 condenser effluent as wastewater and apply the  same provisions
 to it  as are  applied to process wastewater.   Contact condenser
 effluent is considered process wastewater based on  the
 proposed wastewater provisions, and without  any special
 provisions or specific mention, the wastewater provisions will
 apply.  The EPA judged that treating the contact  condenser
 effluent prior to any significant opportunity for
 volatilization protects  the environment.  Further,  the HON
 wastewater provisions, which are the basis for the  wastewater
 provisions  in the proposed and final standards, have been
 judged to be  environmentally effective and cost effective
                               1-26

-------
overall.  Therefore, if the wastewater provisions deem a
wastewater stream to be Group 2 (i.e., not requiring control),
that means it is not cost effective to control any VOHAP that
may be contained in that wastewater stream.
     Addressing emissions from the process contact cooling
tower was more complex.   [Note:  The HON did not specifically
address process contact cooling towers because they are not
used extensively in the synthetic organic chemical
manufacturing industry  (SOCMI).]  Given that the emissions of
VOHAP would be dealt with through the use of the wastewater
provisions at existing affected sources, the key issue related
to cooling towers became the emissions of ethylene glycol.
Between proposal and promulgation, the EPA spent considerable
effort gathering information on PET vacuum systems and their
emissions.  Much of this effort focused on emissions from the
cooling towers, specifically emissions of ethylene glycol.  In
addition to gathering information on emissions, the EPA
investigated control options aimed at reducing emissions of
ethylene glycol from cooling towers.  Possible control
options, not considering ethylene glycol jets, included
treating a slipstream of the cooling tower water to reduce the
concentration of ethylene glycol or installing a large heat
exchanger to isolate the cooling tower from the process.  None
of these control options were shown to be cost effective.  All
of the factors discussed above have led the EPA to conclude
that, with one exception discussed below, specific provisions
for controlling emissions from process contact cooling towers
are not warranted for existing affected sources.  Instead, the
EPA is requiring that the wastewater provisions be applied to
all vacuum system wastewater.
     The one exception where it was found to be cost effective
to control emissions from existing process contact cooling
towers was for affected sources manufacturing PET using a
continuous TPA high viscosity multiple end finisher process.
Based on industry reported emissions, total ethylene glycol
emissions for all PET subcategories are approximately
340 Mg/yr, and approximately 230 Mg/yr are being emitted from.
                             1-27

-------
the single plant site that is part of this subcategory.   This
subcategory is required to keep the concentration of ethylene
glycol in the process contact cooling tower water to 4 percent
by weight or less.
     As at proposal, the final standards require that owners
or operators of new affected sources manufacturing PET not
send contact condenser effluent associated with a vacuum
system to a process contact cooling tower.
1.2.5.2  Vacuum System Wastewater
     Many commenters objected to the proposed provision
designating all contact condenser effluent as Group 1
wastewater streams.  Commenters stated that the EPA had not
provided adequate rationale explaining why the standards for
contact condenser effluent were more stringent than the
standards for other wastewater streams.  Because the automatic
designation of contact condenser effluent as Group 1
wastewater was done in an effort to make the use of noncontact
condensers within steam-based vacuum systems equivalent to
ethylene glycol jets, and that need no longer exists for
existing affected sources, there is no longer a need or
justification to designate all contact condenser effluent from
existing PET affected sources as Group 1 wastewater.  As a
result, the EPA has changed the final standards to implement
the group determination procedure  (i.e., Group 1 or Group 2)
for contact condenser effluent from existing PET affected
sources.
     On the other hand, the final standards continue to
prohibit the use of process contact cooling towers for new PET
affected sources.   [Note:  This requirement is equivalent to
the MACT floor.]  However, the provisions designating all
contact condenser effluent streams from new PET affected
sources as Group  1  wastewater have been dropped from the final
standards.  Like  existing PET affected sources, the final
standards  implement the group determination procedure for
contact condenser effluent from new PET affected sources.  As
described  earlier,  the purpose for designating all contact
condenser  effluent  streams as Group 1 wastewater was to ensure
                              1-28

-------
equivalency between the use of ethylene glycol jets and the
use of noncontact condensers.  At the time of proposal,  the
EPA considered ethylene glycol jets to be pollution free
because there was no wastewater stream produced.   The
automatic designation of all contact condenser effluent as
Group 1 wastewater was meant to address the absence of
wastewater when ethylene glycol jet systems are used.  Since
proposal, the EPA has come to understand that the use of
ethylene glycol jets does not eliminate emissions because the
additional loading to the glycol recovery unit can create
additional emissions from process vents or wastewater streams
or both.
     Realizing that ethylene glycol jets are not pollution
free, the EPA considered the additional emissions at the
glycol recovery unit when ethylene glycol jets are used and
the emissions from the contact condenser effluent when
noncontact condensers are used.   [Note:  The use of ethylene
glycol jets would still achieve more control than using steam
jets and subjecting the contact condenser effluent to the
wastewater provisions.]  In  either case, an equivalent
quantity of VOHAP is introduced to the vacuum system from the
process.  Process vents and  wastewater streams at the glycol
recovery unit are subject to the appropriate provisions of the
standards.  For example, a wastewater stream at the glycol
recovery unit would be subject to a group determination and,
based on the results, would  be controlled if required.  The
EPA judged that to require the same action for the contact
condenser effluent from noncontact condensers would ensure
equivalency between a new affected source using ethylene
glycol jets and one using noncontact condensers.
1.2.6  Equipment Leak  Provisions
1.2.6.1  Polystyrene
     A number of commenters  stated that the EPA had
overestimated the emissions  from  equipment leaks at
polystyrene affected  sources in general and, in particular,
from components containing  styrene as the organic HAP.  The
commenters claimed that the  use of the average SOCMI emission
                              1-29

-------
factors to estimate emissions was inappropriate because the
vapor pressure of styrene at typical operating conditions in
polystyrene affected sources equates to a concentration of
less than 10,000 ppm and no leaks would be detected (using a
leak definition of 10,000 ppm).  The commenters also stated
that the effect of using the average SOCMI emission factors
was to justify the HON equipment leak program (i.e., subpart
H) and that, if the EPA used a more realistic estimate of
emissions  (e.g., the emission factor for components with
concentrations less than 10,000 ppm), a re-evaluation of the
costs and emission reductions would likely result in a
conclusion that the HON is not cost effective.  The commenters
recommended that if any program was to be implemented for
polystyrene affected sources and especially for components in
styrene service, it should either be a visual-only program or,
at most, a program based on a State RACT program.
     In response to these comments, the EPA re-evaluated the
analysis that served as the basis for proposing the. equipment
leak provisions from the HON for polystyrene and other
styrene-based resin affected sources.   (The results of this
re_evaluation are contained in Item IV-B-2, Docket A-92-45.)
In re-evaluating that analysis, the EPA determined the MACT
floor for each subcategory.  Because a MACT floor was
determined to exist and the Clean Air Act does not allow the
EPA to set a standard less stringent than the MACT floor, the
EPA determined that some standard must be set for these
subcategories.
     The EPA agrees with the commenters that the average SOCMI
emission factors are likely to overestimate emissions from
components containing or contacting styrene.  Therefore, the
EPA lowered the emission factors used to estimate the
emissions  from these components.  The EPA also adjusted leak
rate factors based on additional data and comments from the
industry.  Adjustments to the emission factors and leak rates
had the effect of reducing both the overall emission estimates
and the emission reductions associated with the current
                              1-30

-------
industry programs, the MACT floor programs, and the HON
program.
     In addition, the EPA re-evaluated the costing program
used to estimate the costs for the various equipment leak
programs.  Errors were discovered in the costing program.  The
net effect of correcting these errors was to lower the
estimated costs for all of the various equipment leak
programs.  The incremental differences between the various
programs decreased as well, but not by as much as the overall
costs.
     To determine whether or not the subpart H provisions were
cost effective, the EPA examined the incremental costs and
emission reductions between the HON program and the MACT floor
program for each of the subcategories.   [Note:  In the
original analysis, the EPA examined the cost between the HON
program and each facility's specific current program.]  Based
on the incremental differences between the HON and the MACT
floor programs, the EPA determined that the HON requirements
are cost effective for each of the styrene-based resin
subcategories, including the three polystyrene subcategories.
Therefore, the EPA has retained the HON requirements in the
final rule.
1.2.6.2  PET
     Many commenters objected to the imposition of the HON
equipment leak requirements on PET affected sources,
especially those using the TPA process.  These commenters
stated that, due to the preponderance of components in heavy-
liquid service, no program should be imposed or, if a program
is required, it  should simply be a visual-only program.  Some
of the commenters referred to the rationale in the development
of the Polymer Manufacturing NSPS  (40 CFR part 60, subpart
DDD), which exempted PET affected sources using a continuous
TPA process from equipment leak regulation.
     As  was done  for the styrene-based resin subcategories,
the EPA re-evaluated the emission estimates and cost estimates
for all  of the PET subcategories.  The results of this
re_evaluation  are found in Item IV-B-3, Docket A-92-45.  The
                             1-31

-------
following actions were taken as part of the re-evaluation.
First, the EPA lowered the emission factors used to estimate
emissions from components in heavy liquid service to take into
account the properties of ethylene glycol.  Second, the EPA
limited the re-evaluation of costs and emission reductions to
only those facilities that provided specific component
profiles  (i.e., component counts and types of service).  The
results of the re-evaluation showed that, except for one PET
subcategory, the HON requirements are cost effective.  Thus,
the final rule retains, for the most part, the HON
requirements for the PET subcategories.   [Note:  Modifications
to the HON requirements are discussed below.]
     The re-evaluation of the proposed equipment leak .
provisions showed that the provisions were not cost effective
for PET affected sources using a continuous TPA high viscosity
multiple end finisher process.  Thus, this subcategory, which
currently contains only one affected source, is exempt from
the equipment leak requirements in the final rule.
     With regard to exempting PET affected sources altogether
based on the exemption contained in the Polymer Manufacturing
NSPS, the EPA does not agree with the commenters.  During the
development of the Polymer Manufacturing NSPS, information
available to the EPA indicated that all components at PET TPA
continuous facilities were in heavy liquid service.
Information provided by the industry during the development of
this rulemaking, however, shows the presence of components in
gas/vapor service and in light liquid service at some PET TPA
facilities.  The decision to require an equipment leak program
for this rulemaking is based on this newer information.
However, the EPA continues to agree that, if a PET TPPU
consists solely of components in heavy liquid service (or in
vacuum service) , then said TPPU should be exempt from the
equipment leak standards.  The final rule contains this
provision.
     Many commenters were concerned about the recordkeeping
and reporting requirements, especially for PET TPA affected
sources where the vast majority of components are in heavy
                             1-32

-------
liquid service.  While the EPA believes the original
requirements were the minimum required to ensure compliance
with the overall program, the EPA has reduced the burden of
the recordkeeping and reporting requirements for these
subcategories in light of the preponderance of components in
heavy liquid service.  The most significant change is the
elimination of the requirement to initially list the
identification numbers of components in heavy liquid service.
     Finally, several commenters expressed concern with the
use of Method 21 to detect leaks from components in ethylene
glycol  (i.e., heavy liquid) service, stating in part that
Method 21 would not detect an ethylene glycol leak based on
the properties of ethylene glycol and the operating conditions
of the process.  The EPA agrees that Method 21 is
inappropriate for determining ethylene glycol leaks where the
leaking component is in  heavy liquid service.  However,
§63.169 of subpart H does not require the use of Method 21; it
is one of two alternatives for determining the presence of
leaks from components in heavy liquid service.  The other
alternative  is to use a  sensory-based detection method.  The
final rule has been revised to clarify that leaks are to be
determined exclusively through the  use of visual, audible,
olfactory, and any other detection  methods; Method 21 is not
to be used to determine  leaks from  components in heavy liquid
service at PET facilities.
1.2.7  Emissions Averaging Provisions
1.2.7.1  Number of Emission Points  Allowed in Emissions
Averaging
     Numerous  commenters requested  that  the number of emission
points  allowed in an  emissions average be increased to
20 points  from the 5  points allowed at proposal.  If pollution
prevention measures  are  used, commenters requested that an
additional  5 emission points be  allowed  from the 3 emission
points  allowed at proposal.  The commenters stated that this
would be  consistent  with the HON.
     In response  to  comments and to be more consistent with
the  HON and the  Group I  Polymers and Resins  (Elastomers)
                             1-33

-------
NESHAP, the number of emission points allowed in emissions
averaging at a plant site has been increased to 20 points,
with an additional 5 points allowed if pollution prevention
measures are used.  These values (i.e., 20/25) are the maximum
allowed for all emissions averages at a plant site, regardless
of the number of affected sources present at a plant site or
the number of emissions averaging programs implemented.
1.2.7.2  Including Batch Process Vents in Emissions Averaging
     Several commenters requested that batch process vents be
included in the emissions averaging provisions.  The EPA had
not allowed emissions averaging of batch process vents at
proposal.  The proposal preamble stated that the accuracy and
consistency of emissions estimates needed for emissions
averaging were considered to be greater than those needed for
applicability determinations.  However, upon review, the EPA
has determined that having the same procedures to estimate
emissions for applicability determinations and emissions
averaging is reasonable.  The final standards allow emissions
averaging of existing batch process vents as well as aggregate
batch vent streams.
1.2.8  Compliance and Performance Test Provisions and
Monitoring- Requirements
1.2.8.1  Excused Excursions
     Many commenters requested that the proposed standards
allow excused excursions in the same way that the HON allows
excused excursions.  In the final standards, the EPA included
provisions to excuse a certain number of excursions for each
reporting period.  These provisions are identical to the HON
provisions.  The decision to include excused excursion
provisions was based on data and information presented in
public comments received on the proposed standards and
received during industry meetings held after proposal.  The
commenters contended that by not allowing excused excursions
in these standards, the EPA has made these standards more
stringent than the HON, which the proposed standards and the
costs of the proposed standards were modelled after.  The
commenters requested that the EPA justify the increased cost
                             1-34

-------
of imposing the more stringent "no excused excursion"
provisions.  The EPA agreed with the commenters that not
allowing excused excursions could impose significant
additional capital and operating costs on the affected source
for only negligible corresponding reductions in air emissions.
As is always the case, a State has the discretion to impose
more stringent requirements than the requirements of NESHAP
and other federal requirements and could choose not tb allow
the excused excursion provisions contained in these standards.
     The EPA considered the number of excused excursions that
would be most appropriate for these standards and determined
that the number of excursions allowed in the HON is
reasonable.  Therefore, the final standards allow a maximum of
6 excused excursions for the first semiannual reporting
period, decreasing by 1 excursion each semiannual reporting
period.  Starting with the sixth semiannual reporting period
(i.e., the end of the third year of compliance) and
thereafter, affected sources are allowed one excused excursion
per semiannual reporting period.
1.2.8.2  Parameter Monitoring Levels
     Commenters requested that more flexibility be permitted
for establishing compliance levels for parameter monitoring.
The commenters asked the EPA to allow the use of the HON range
concept, which recognizes that a process or control device
operates properly over a range of conditions.
     The EPA revised the requirements for establishing
parameter monitoring levels to incorporate the concepts
included in the HON range concept.   [Note:  The final
standards  continue to use the term level.]  Under the final
rule,  the  owner or operator can choose between three
procedures for establishing parameter monitoring levels.  By
providing  the flexibility to establish parameter monitoring
levels based on one of these three procedures, site-specific
levels can be chosen which best accommodate variation in
emission point characteristics and control device designs.
These  three procedures for establishing parameter monitoring
levels are based  on:   (1) performance tests;  (2) performance
                             1-35

-------
tests, engineering assessments, and/or manufacturer's
recommendations; and (3) engineering assessments and/or
manufacturer's recommendations.  The establishment of a
parameter monitoring level based solely on performance tests
is preapproved by the Administrator; however, parameter
monitoring values determined using the last two procedures,
which may or may not use the results of performance tests,
must be approved by the Administrator for each individual
case.
1.2.8.3. Other Changes Resulting from EPA Review
     In addition to changes made to the proposed rule as a
result of public comment, changes were made as a result of EPA
independently reviewing the rule between proposal and
promulgation.  In the final rule, the EPA has allowed the use
of TOG, minus methane and ethane, as the basis for compliance
demonstrations  (i.e., performance tests) as an alternative to
organic HAP for continuous process vents, batch process vents,
storage vessels, and wastewater streams.   [Note:  The term
"TOG," as used in the remainder of this discussion and as
defined in the final rule, denotes "TOG, minus methane and
ethane."]  The final rule also allows the use of TOG as an
alternate basis for applicability determinations for batch
process vents; TOG is not allowed as an alternate basis for
applicability determinations for continuous process vents.  In
addition, the final rule allows the use of Method 25A to
measure TOG in those instances where TOG serves as an
alternate basis for compliance demonstrations or applicability
determinations.
     These changes were made to provide flexibility to the
regulated community and to reduce the overall burden of the
rule.  The EPA considered the impacts of allowing TOG to serve
as a surrogate for organic HAP in compliance demonstrations
and did not find any adverse impacts.  Further, in allowing
the use of Method 25A as a complement to the use of TOG as a
surrogate to organic HAP, the EPA judged that the burden of
implementing the final rule would be reduced with no adverse
impacts.
                             1-36

-------
     As proposed, the provisions concerning applicability
determinations and compliance demonstrations allowed an owner
or operator to measure total organic HAP or TOG;  however,  the
proposed rule required Method 18 for measurement  of organic
compounds in both cases.  In light of this,  the major
difference between the proposed and final rule is the
flexibility to use Method 25A in measuring TOC.  In deciding
to allow the use of Method 25A, the EPA considered the
composition of the emission streams and considered provisions
that could be implemented to safeguard against inappropriate
uses of Method 25A.  To the best of the EPA's knowledge, the
emission streams regulated by this rule are predominantly
organic HAP.  Also, with one exception (i.e., ethylene
glycol), the regulated organic HAP, which are listed in the
definition of organic HAP found in the final rule, have
response factors to Method 25A adequate to ensure satisfactory
emissions measurements.  To safeguard against inadequate
emissions measurements that might result from the
inappropriate use of Method 25A, the final rule specifies the
calibration gas to be used  (i.e., the organic HAP representing
the largest percent by volume) and what is an acceptable
response for the calibration gas  (i.e., 20 times  the standard
deviation of the response from the zero calibration gas) .
     In considering the use of TOC as an alternate basis for
applicability determinations, and the complementary use of
Method 25A to measure TOC, the EPA evaluated applicability
determinations for continuous process vents and batch process
vents.   [Note:  The applicability determination procedures for
storage vessels and wastewater streams do not use airborne
organic HAP concentration values.  Therefore, the use of
Method 25A for these emission points is not applicable.]  The
applicability determination procedure for continuous process
vents  (i.e., the total resource effectiveness {THE} equation)
requires an owner or operator to estimate the concentration
and emission rate for both organic HAP and TOC.  Because both
of these values are required and were the basis for the
original TRE analysis, the EPA judged that it was
                             1-37

-------
inappropriate to use TOG as an alternate basis to organic HAP
for the applicability determinations for continuous process
vents.  On the other hand, the original analysis for batch
process vents  (i.e., the Batch Processes ACT) was based on the
control of TOG; thus, the EPA considered the use of TOG as an
alternate basis for the applicability determinations for batch
process vents to be satisfactory.
1.2.9  Recordkeeping and Reporting
     Several commenters stated that the recordkeeping and
reporting requirements of the proposed standards were
extremely burdensome.  These commenters requested that the EPA
reduce unnecessary recordkeeping and reporting requirements in
the final standards.  Commenters requested that the frequency
with which records must be retained be reduced.  The
commenters also contended that records should only be required
if an excursion has occurred.  The commenters contended that
records showing compliance with the standards were
unnecessary.
     The EPA has made every effort to reduce the recordkeeping
requirements of the final thermoplastics rule.  The EPA
recognizes that unnecessary recordkeeping and reporting
requirements would burden both the affected source and EPA
enforcement agencies.
       The EPA reviewed the recordkeeping required by the
proposed rule and has made reductions in the amount of
information that is required to be recorded.  The final rule
has been changed to require recording and retention of hourly
average values of continuously monitored values.   The proposal
required that 15-minute averages be calculated and recorded.
Under the proposal, if the daily average value, was above the
minimum or below the maximum established parameter monitoring
levels  (i.e., excess emissions occurred), the 15-minute values
had to be retained; if the daily average value did not exceed
the established parameter monitoring level,  the 15-minute
values could be converted to hourly averages, and the hourly
averages could be retained instead of the 15-minute averages.
Upon reconsideration, the EPA found that hourly average values
                             1-38

-------
provide a sufficient record to support the calculation of the
daily average value.  Therefore, to reduce the recordkeeping
burden, the rule has been changed to specify that only hourly
averages must be calculated and recorded.  The rule no longer
requires calculating or recording of 15-minute average values.
     For emission points where continuous parameter monitoring
is required, the value of the parameter must still be measured
at least once every 15 minutes, but only the hourly average
must be calculated and recorded.  Many facilities already have
computerized systems and monitor parameters more frequently
than once every 15 minutes for process control purposes.  The
15-minute monitoring frequency is consistent with the General
Provisions and previous NSPS and NESHAP for emission points
from similar industries.
     In addition, the EPA added provisions  [§63.513(h)
[§63.1335(h)]] that allow an owner or operator to implement a
reduced recordkeeping program provided that certain criteria
related to the monitoring system and the performance of the
process, as it relates to maintaining compliance with the
monitoring provisions, are met.  Under these provisions it is
possible for an owner or operator to retain only daily average
values or, after a period of 6 consecutive months without an
excursion, to retain no daily records.
     The EPA believes that the recordkeeping requirements of
the final rule are necessary to show compliance.  The EPA will
continue to require owners or operators to keep records,
regardless or whether there was an excursion or not.   These
records are necessary to prove compliance when no excursion
has occurred and are used to determine the severity of a
violation, and, thus, how much of a penalty should be assessed
once an excursion has occurred.
     The EPA has made every effort to reduce the reporting
requirements of the final rule.  The EPA reviewed each report
required at proposal, and determined that two of these
reports, the Initial Notification and the Implementation Plan,
contained many requirements that were duplicative with the
existing operating permit program.  For this reason,  the EPA
                             1-39

-------
has removed the requirements for the Initial Notification and
Implementation Plan from the final rule.
     The EPA considers the recordkeeping and reporting
requirements of the final rule to be the minimum necessary to
ensure compliance.
     Upon further review, the EPA decided to add a new report,
the Precompliance Report, to the final rule to allow the owner
or operator of an affected source to request an extension of
compliance or to request approval to use alternative
monitoring parameters, alternative continuous monitoring or
recordkeeping, or alternative controls.  At proposal, these
items were submitted in the Implementation Plan.  Overall,
these changes, deleting the Initial Notification and the
Implementation Plan and adding the Precompliance Report,
result in a reduction of the reporting burden for the affected
source.
                              1-40

-------
                      2.0  APPLICABILITY

2.1  SELECTION OF SOURCE CATEGORY  .
     Comment;  One commenter  (IV-D-03) supported the addition
of a separate source category for nitrile resins production
with a promulgation date no later than November 15, 2000.
     Response:  A nitrile resins source category covering the
production of nitrile resins has been added to the source
category list under section 112 (c) of the Act and to the
source category schedule under section 112(e) of the Act.
Nitrile resins production has been included as one of the
seven source categories regulated under the final rule.
     Comment:  One commenter  (IV-D-25) recommended that the
final rule exempt processes which manufacture resin-containing
binder systems for paints, coatings, or adhesives.  The
commenter  (IV-D-25) stated that production of some resins
subject to the proposed rule, such as SAN, are incidental to a
number of paint manufacturing processes.  The commenter
(IV-D-25) stated that if the definitions and applicability
provisions of the proposed rule are read too narrowly, there
may be ambiguity as to whether  such paint manufacturing
facilities are subject to the rule.  The commenter  (IV-D-25)
noted that the paint, coating,  and adhesive manufacturing
source category will be subject to a NESHAP scheduled for
promulgation by November 15,  2000.
     Response:  Processes which manufacture resin-containing
binder systems for paints, coatings, or adhesives are not
subject to the final rule.  To  clarify this, the EPA has added
a  specific exemption to the applicability provisions of the
final rule.   The exemption, which is  in §63.500 (d) (2)
 [§63.1310 (d) (2)],  states that processes which manufacture
binder systems containing a thermoplastic product  for paints,
coatings,  or adhesives are exempted from the definition of
                              2-1

-------
affected sources; 'therefore, these processes are not subject
to the rule.
     Comment;  One commenter (IV-D-04) requested that an
exemption be added to §63.500 (d) of the rule to exempt plastic
parts molding processes which manufacture a thermoplastic
product.  The commenter  (IV-D-04) offered an example of a
plastic parts molding process as one in which two chemicals
are combined in a mold, a chemical reaction (typically
polymerization) occurs, and the materials harden in the mold
to form the intended plastic part.  The commenter (IV-D-04)
stated that they do not currently manufacture thermoplastic
products by molding operations.  However, the commenter
 (IV-D-04) was concerned that if styrene were used in such a
process, there may be the possibility that all the
applicability criteria in the proposed rule could be met.
Further, although it is unlikely that any emission limitations
in the proposed rule would  apply to such a process, the
commenter  (IV-D-04) was  concerned that recordkeeping and
reporting requirements could apply.  The commenter  (IV-D-04)
stated that these requirements  would provide no environmental
benefit and would impose  a  burden on a source for a process
that the commenter presumed the EPA had not intended to
regulate.
     Response;   The EPA  did not intend to regulate plastic
parts molding process  operations under the thermoplastics
rule.  To clarify this,  the EPA has added a specific exemption
to the applicability provisions of the final rule.  The
exemption appears  in §63 .500 (d) (2)  [§63 .1310 (d) (2) ] and states
that polymerization processes  that occur in a mold  are
exempted from the'affected source.
 2.2  SELECTION OF SOURCE
 2.2.1  Area source
     Comment;   One commenter  (IV-D-05) requested that the  EPA
 defer permitting of area sources in  the  final rule.  The
 commenter (IV-D-05)  contended that State agencies have a
 significant burden in the permitting of  major sources at this
 time.   Another commenter (IV-D-10)  contended that the EPA  is
                              2-2

-------
illegally attempting to regulate area sources located at a
plant site, one or the other which is a major source.  The
commenter stated that subjecting small TPPUs located at a
plant site that is a major source to this rule is unjustified
and contrary to the Congressional intent that only major
sources be regulated by NESHAP.  The commenter (IV-D-10)
contended that the EPA does not have the statutory authority
to regulate area source process units in a source category
unless it first makes the necessary finding that these area
sources present a threat of adverse effects to human health or
the environment.
     Response:  The EPA is not regulating area sources under
this rulemaking, and therefore, the rule does not trigger the
requirement in §63.1 (c) (2) of subpart A to determine whether
area sources will require permits.
     The final rule only regulates emission points at major
sources.  [Note:  The area source/major source determination is
made for the plant site, not for each individual TPPU at the
plant site, and the term major source is defined in §63.2 of
subpart A.]
     The addition of process units to a major source is
addressed in the next section of this document
 (i.e., Section 2.2.2).
2.2.2  Man or source
     Comment;  Three commenters  (IV-D-09, IV-D-10, IV-D-18)
urged the EPA not to apply new source requirements to a TPPU
that is added to a major source when the potential emissions
of the added TPPU are below major source levels.  The
commenters  (IV-D-09, IV-D-10, IV-D-18) stated that according
to §63.500(g)(1) of the proposed rule, the new source
requirements will apply to any TPPU that is added to a major
source, regardless of the potential emissions of the TPPU.
Two of the  three commenters  (IV-D-09, IV-D-18) specifically
stated that the proposed rule does not specify the potential
emissions  for the added TPPU.  The two commenters  (IV-D-09,
IV-D-18)  stated that the proposed rule is more stringent than
the HON, which applies new source requirements only if the
                              2-3

-------
added process unit has emissions above the major source
levels.  The comtnenters  (IV-D-09, IV-D-10, IV-D-18) urged the
EPA to revise the provisions in the thermoplastics rule to
match the requirements of the HON.
     Response:  The EPA reviewed and modified the provisions
of the proposed rule for applying new and existing source
requirements for changes and/or additions made at existing
plant sites.  These provisions were in §63.500(g) and (h) of
the proposed rule; they are now addressed in §63.500{i)
[§63.1310(i)] of the final rule.
     The EPA agrees with the commenters that a TPPU that is
added to an existing plant site should be subject to new
source requirements only when the added TPPU has emissions
above the major source levels, as applied to the HON.   [Note:
Major source levels refers to the potential to emit 10 tons
per year or more of any HAP or 25 tons per year or more of any
combination of HAP.]  However, as applied to the
thermoplastics rule, determining whether to apply existing or
new source requirements to a TPPU added to an existing plant
site requires an examination of the thermoplastic product
being manufactured.  The EPA finds that a TPPU that is added
to an existing plant site should be subject to new source
requirements when said TPPU has emissions above the major
source levels and the TPPU manufactures a thermoplastic
product currently manufactured at that plant site.  If the new
TPPU manufactures a thermoplastic product not previously
manufactured at that plant site, said TPPU is subject to new
source requirements regardless of the emission levels.  The
second scenario differs from the HON because the HON covers •
only one source category.  The thermoplastics rule covers
multiple source categories (7 source categories containing a
total of 19 subcategories).  Because the affected source, as
defined for this rule, is limited to a single subcategory, the
addition of a TPPU producing a thermoplastic product not
currently manufactured at the plant site equates to adding an
affected source to the plant site.  All new affected sources
are subject to new source requirements, regardless of their
                              2-4

-------
potential to emit.  The final provisions for §63.500 (i) (1) (i)
[§63.1310(i) (1)(i)l are as follows:

           (i)   If a TPPU is added to a plant site, said
     addition shall be subject to the requirements for a new
     affected source in this subpart upon initial start-up or
     by  [insert publication date], whichever is later, if said
     addition meets the criteria specified in paragraphs
      (i) (1) (i) (A) through  (i).(l) (i) (B) and either  (i) (1) (i) (C)
     or  (i)(1)(i)(D) of this section:
           (A)   Said addition meets the definition of
     construction in §63.2 of subpart A;
           (B)   Such construction commenced after
     March 29,  1995; and
           (C)   Said addition has the potential to emit  10  tons
     per year  or more of any HAP or 25 tons per year or more
     of  any combination of HAP, and said addition manufactures
     a thermoplastic product currently produced at the  plant
     site;  or                                           .
           (D)   Said addition manufactures a thermoplastic
     product not previously produced at the plant  site.

The  intention  of  the other provisions contained  in proposed
§63.500 (g)  and (h) were not changed subs t ant ively, however,
they were substantially revised to state the  intended
requirements more clearly.
2.2.3  Affected Source
      Many commenters were  confused about the  meaning of
 "affected source"  in  the proposed rule.  The  following
comments and responses present  their  concerns and explain how
the  EPA has clarified the  meaning of  "affected source" and
related terms for the  final rule.
      Comment:   Three  commenters (IV-D-11,  IV-D-17,  IV-D-20)
 supported the use of  "affected source"  in  the proposed rule.
 Three commenters (IV-D-10, IV-D-11,  IV-D-18)  suggested ways
 the  term "affected source" could be clarified in the final
 rule.   One of  these commenters (IV-D-10)  asserted that the
 definition of  "affected source" in §63.502 of the proposed
 rule was unnecessary,  unclear,  and internally inconsistent.
 The commenter  (IV-D-10)  stated that "affected source" was
 defined  in  subpart A and was clarified in §63.500 (b) of the
 proposed rule; therefore,  given these clear designations in
                               2-5

-------
the applicability section of the rule, the definition of
"affected source" in §63.502 of the proposed rule caused
confusion and should be deleted.
     The comraenter  (IV-D-10) stated that the definition was
also unclear in that it defined "affected source" to be
"... all TPPUs at a plant site  . . . ", but also stated that
"... if a process unit produces more than one thermoplastic
product type in the same equipment, then that process unit
shall be a single affected source."  The commenter (IV-D-10)
stated that it was unclear whether the affected source was the
group of all TPPUs at a site or individual TPPUs when
different thermoplastic products are produced.  The commenter
(IV-D-10) stated that this inconsistent language was
confusing.
     Two commenters  (IV-D-11, IV-D-18) expressed confusion
about the definitions of "source  (affected source)" and
"TPPU."  The commenters  (IV-D-11, IV-D-18) stated that the HON
carefully differentiated between "source" and "process unit"
(i.e., TPPU) , but the proposed  rule has confused the two
terms.
     Response;  The meaning of  "affected source" was not clear
in the proposed rule and it has been clarified in the final
rule.  The definition of "affected source" included in §63.502
of the proposed rule has been revised to remove all
descriptive language and now refers to §63.500(a)
[§63.1310(a)].  At proposal, the provisions in §63.500(a) tied
applicability of the rule to the existence of one or more
TPPUs.  In the final rule, these provisions define
applicability of the rule in terms of the affected source.
Also, §63.500(b)  [§63.1310(b)]  of the final rule does not
contain the provisions previously found in §63.500 (b) of the
proposed rule that  described the affected source.
      In the final.rule, under §63.500(a)  [§63.1310(a) ],
"affected source" is defined as the following:
                              2-6

-------
          (a)  Definition of affected source.   The provisions
     of this  subpart  apply to each affected source.  An
     affected source  is either an existing affected source or
     a new affected source.   Existing affected source is
     defined  in  paragraph (a)(6)  of this section,  and new
     affected source  is defined in paragraph (a)(7) of this
     section.  The affected source also includes the emission
     points and  equipment specified in paragraphs  (a)(1)
     through  (a)(5) of this section that are associated with
     each group  of TPPU.
          (1)  Each wastewater stream;
          (2)  Each wastewater operation;
          (3)  Each heat exchange system;
          (4)  Each process contact cooling tower used in the
     manufacture of PET that is associated with a new affected
     source;  and
          (5)  Each process contact cooling tower used in the
     manufacture of PET using a continuous terephthalic acid
     high viscosity multiple end finisher process that is
     associated  with  an existing affected source.
          (6)  Except as specified in paragraphs  (b) through
     (d) of this section, an existing affected source is
     defined as  each  group of one or more thermoplastic
     product  process  units  (TPPUs) that is not part of a new
     affected source  as defined in paragraph (a)(7) of this
     section, that is manufacturing the same primary product,
     where each  TPPU uses as a reactant, or uses as a process
     solvent, or produces as a by-product or co-product any
     organic hazardous air pollutant  (organic HAP), and that
     is located  at a  plant site that is a major source.
          (7) Except as specified in paragraphs  (b) through
     (d) of this section,- a new affected source is defined as
     a source meeting the criteria of paragraph (a)(7)(i),
     (a) (7) (ii), or  (a) (7) (iii) of this section.
          (i) At a plant site previously without HAP
     emissions points, each group of one or more TPPUs
     manufacturing the same primary product that is part of a
     major source on which construction commenced after March
     29, 1995,
          (ii)  A TPPU meeting the criteria in paragraph
     (i)(1)(i) of this section, or
           (iii)   A reconstructed affected source meeting the
     criteria in paragraph  (i)(2)(i) of this section.


     The definition of  "affected source" states that each
affected source will include all TPPUs at the plant  site that

produce the same thermoplastic product as their primary

product that are not included as part of a new affected

source.  Also, each affected source will include each

wastewater stream and operation originating from the TPPUs

included in the affected source, each heat exchange  system

                              2-7

-------
associated with the TPPUs included in the affected source,  and
each process contact cooling tower used in the manufacture of
PET associated with the TPPUs included in the affected source.
If only one TPPU produces a given thermoplastic product,  then
the affected source will be comprised of the one TPPU and any
wastewater streams and operations, heat exchange systems, and
process contact cooling towers used in the manufacture of PET
associated with that one TPPU, as applicable.  If six TPPUs
all produce the same thermoplastic product as their primary
product, then the affected source will be comprised of the six
TPPUs and any wastewater streams and operations, heat exchange
systems, and process contact cooling towers used in the
manufacture of PET associated with those six TPPUs, as
applicable.
     For the final rule, the number of affected sources
present at an existing plant site that is limited to existing
TPPUs will equal the number of thermoplastic products
manufactured at that plant site.  If a plant site has three
affected sources, then three thermoplastic products are
manufactured at said plant site.  Each thermoplastic product
is  in a different subcategory; there are 19 subcategories
within the 7 source categories covered by the rule.  The
number of affected sources could be as few as 1 affected
source and as many as 19 affected sources.   [Note:  A plant
site with new and existing TPPUs may not follow the above
example because a new TPPU  (i.e., an added TPPU) may or may
not be a new affected source.]
     This scenario differs from the HON rule which covers a
single source  category and defines affected source such that
there is one existing affected source per plant site.  This
significant difference between these two rules  is a result of
the requirement  in the Act that states that the scope of an
affected source  cannot extend across source categories or
subcategories.
     The revisions to  the definition of  affected  source  and
TPPU have  addressed commenter concerns  that  the proposed rule
confused the two terms  of  "affected  source"  and "TPPU."  In
                              2-8

-------
the final rule, a TPPU is defined as a distinct entity and an
affected source is made up of one or more TPPU and other
associated equipment.
     Comment:  Six commenters (IV-D-07, IV-D-08, IV-D-10,
IV-D-11, IV-D-15, IV-D-18) urged the EPA to clarify the
proposed definition of "TPPU. "  Some of the commenters
(IV-D-08, IV-D-li, IV-D-15, IV-D-18) asked that the proposed
definition of  "TPPU" be clarified and revised to closely match
the definition of "chemical manufacturing process unit," as
used in the HON.  These commenters  (IV-D-08, IV-D-11, IV-D-15,
IV-D-18) believed this clarification/revision would avoid
applicability determination problems.
     Pour commenters  (IV-D-07, IV-D-10, IV-D-11, IV-D-15)
stated that the definition of "TPPU" should include process
equipment, such as raw material storage tanks and product
storage vessels; however, ancillary equipment such as the
wastewater treatment facility, cooling towers, and heat
exchangers should not be  included in the definition of "TPPU."
Four commenters  (IV-D-07, IV-D-11,  IV-D-15, IV-D-18) felt
ancillary equipment  should be removed from the equipment list
in §63.500(b)  of the proposed rule because ancillary equipment
is not directly used to produce thermoplastics.  Specifically
addressing wastewater, two commenters  (IV-D-07, IV-D-08)
requested that the EPA clarify that wastewater and associated
treatment residuals  are part of the affected source
(i.e., subject to the rule) , but are not part of the TPPU.
The commenters (IV-D-07,  IV-D-08) stated that a stream becomes
wastewater when  it exits  the process unit, on its way to
treatment or disposal, and that the process unit is supposed
to consist of  equipment that produces a product.
     Four commenters  (IV-D-07, IV-D-10, IV-D-11, IV-D-15)
contended that the EPA must provide a clearly defined boundary
for the TPPU.  One commenter  (IV-D-10) stated specific
concerns with  the definition of TPPU as it relates to shared
equipment.  This  commenter  (IV-D-10) described  a very complex
system of process lines that pass product back  and forth
across process lines and  that share reactors, dryers, and
                              2-9

-------
other unit operations.  The commenter (IV-D-10) expressed
concern with how to determine the boundary of the TPPU.  In
their described collection of process lines, the commenter
(IV-D-10) was not certain which process lines would be
"connected" and, therefore, part of a single TPPU.
     To illustrate the uncertainty and confusion in the
proposed definition of TPPU, the commenter  (IV-D-10) provided
the following discussion:  consider, for example, two or more
separate process trains which share one or more manufacturing
components  (e.g., a shared dryer or reactor).  Such a
collection of equipment would arguably be a single TPPU.
Viewing the separate process trains as individual TPPUs could
lead to conflicting compliance requirements, especially for
the shared components.  Therefore, designating this collection
of equipment as a single TPPU would be logical.  However, the
commenter  (IV-D-10) stated that the proposed definitions of
TPPU, affected source, and process unit do not clearly
indicate that two or more process trains that share some
common process equipment would be considered a single TPPU.
The commenter  (IV-D-10) stated that designating the entire
collection of equipment as a single process unit, identified
by its primary product, would lead to a consistent and uniform
set of compliance requirements that would also be consistent
with interpretations provided by the EPA under various NSPS
(subpart III and NNN), where affected sources which are
connected to common equipment  (such as recovery devices) are
considered together as a single affected source.
     Next, three commenters  (IV-D-07, IV-D-11, IV-D-15)
explained that the definition of "TPPU" states that "a
compounding unit is not part of the TPPU."  However, in the
proposed rule,  §63.505(b)(2) and §63.505(b)(3) require that
continuous  and batch  vents  from "other parts of the process"
comply with §63.505 (a).  In addition to compounding units, the
commenters  (IV-D-07,  IV-D-11, IV-D-15) requested that the EPA
exclude  other post-polymerization units such as spinning
units, drawing units, and  extruding units from the definition
of "TPPU"  to avoid inconsistent interpretations of the rule.
                             2-10

-------
     Response:   The definition of "TPPU" in §63.502 [§63.1312]
has been clarified for the final rule.  The EPA agreed with
many of the suggestions for clarification submitted by the
commenters, and the EPA has revised the final rule.  The
revised definition of "TPPU" is included below, along with a
discussion of the major changes made to the definition.

     Thermoplastic product process unit (TPPU) means a
     collection of equipment assembled and connected by pipes
     or ducts,  excluding gas, sanitary sewage, water
      (i.e., not wastewater), and steam connections, used to
     process raw materials and to manufacture a thermoplastic
     product as its primary product.  This collection of
     equipment includes process vents from process vessels;
     storage vessels, as determined in §63.500(g)
      [§63.1310(g)]; and the equipment (i.e., pumps,
     compressors,  agitators, pressure relief devices, sampling
     connection systems, open-ended valves or lines, valves,
     connectors, instrumentation systems, surge control
     vessels, and  bottoms receivers that are associated with
     the thermoplastic product process unit) that are subject
     to the equipment leak provisions as specified in §63.509
      [§63.1331].

     First, the EPA agreed with the commenters that the
definition of "TPPU" should include a list of the collection
of equipment that  comprises a TPPU.  Such a change makes the
definition of TPPU very similar to the definition of "chemical
manufacturing process unit" from the HON.
     Second, the EPA agreed with the commenters that emission
points such as wastewater  streams and operations, heat
exchange systems,  and process contact cooling towers used  in
the manufacture of PET are ancillary equipment and should  not
be defined as part of the  TPPU.  Therefore, said emissions
points were removed from the list of equipment that was
included in the revised definition of "TPPU."  In the final
rule, these emission points are  included in the definition of
"affected  source"  in  §63.500(a)  [§63.1310(a)].
     Third, commenters requested clarification on the
boundaries of the  TPPU, and one  commenter was  specifically
concerned  when  multiple process  units share equipment
 (i.e., beyond raw  material or product storage vessels).  As
                             2-11

-------
discussed above,  the revised definition of TPPU does not
include  ancillary equipment; the EPA believes this change
significantly improves the description of the TPPU boundary.
Speaking to comments regarding shared equipment, the proposed
rule was unclear  as to at what level  (i.e., affected source or
TPPU)  shared equipment was to be considered.  Because the
provisions discussing the affected source in the proposed rule
included the list of equipment to be regulated, it was
possible to interpret that shared equipment was to be
considered in relation to the affected source.  On the other
hand,  §63.500)(f) of the proposed rule clearly assigned shared
storage  vessels to an individual TPPU.  In the final rule, the
EPA considers it  most appropriate to link the concept of
shared equipment  to the TPPU.  By removing the list of
equipment from the definition of affected source and putting
it in  the definition of TPPU and by disassociating ancillary
equipment and emission points  (e.g., heat exchange systems)
from the TPPU list of equipment, the revised final definitions
of "affected source" and "TPPU" clarify the boundary of the
TPPU and clarify  the concept of shared equipment.
     One commenter was particularly concerned with a complex
scenario of interrelated process lines.  The sharing of raw
material or product storage vessels by two or more TPPUs does
not "connect" the TPPUs.  The provisions in §63.500 (g).
[§63.1310 (g)]  of  the final rule assign storage vessels that
are shared among  two or more TPPUs to a specific TPPU for
purposes of regulation.  If two TPPUs share a different type
of equipment,  for example a dryer, and said TPPUs meet the
criteria of the definition of TPPU describing "connected",
then said TPPUs are considered to be a single TPPU.  If two
TPPUs  share only  storage vessels, then each TPPU is considered
a separate entity.  The storage vessels would be assigned to
one of the TPPUs  as discussed earlier.
     Fourth,  the  EPA agreed with commenters that, in addition
to compounding units, other post-polymerization units should
not be included as part of the TPPU.  In order to clearly
exempt these processes in the final rule,  the EPA has listed
                             2-12

-------
them in §63.500(d)  [§63.1310(d)], which excludes certain
processes from the  affected source.
     Comment;  One  commenter  (IV-D-10) stated that the paired
definitions of "process unit" and "TPPU" in the proposed rule
were confusing.  The commenter  (IV-D-10) suggested
consolidating the concept of  "process unit" or "TPPU"  (i.e., a
collection of equipment assembled and connected by pipes or
ducts to process raw material and manufacture an intended
thermoplastics product) into a single definition which
includes all the intended attributes of a TPPU.
     Response;  The EPA has clarified the difference between
the terms "process  unit" and  "TPPU" in the final rule.  A
discussion of the changes made to the definition of "TPPU" is
presented in the previous comment and response.  At proposal,
the term "process unit" meant equipment used to manufacture an
intended thermoplastic product.  This definition overlapped
with the definition of "TPPU" and did not account for
equipment used to manufacture a product which may or may not
be a thermoplastic  product.  Therefore, the EPA revised the
definition of "process unit" for the final rule to mean a
collection of equipment assembled and connected by pipes or
ducts to process raw materials and to manufacture a product.
This definition is  broad and includes TPPUs as well as process
units used to manufacture nonthermoplastic products.
     It is necessary to have a definition for process unit and
TPPU because the provisions for determining primary product,
which are fundamental to determining applicability of the
rule, require the use of both terms.  Until the primary
product is determined, "a collection of equipment assembled
and connected by pipes or ducts to process raw materials and
to manufacture a product" is known as a process unit.   Only
once the primary product is determined to be a thermoplastic
product, as defined by the final rule (i.e.,  1 of the
19 subcategories) ,  is this same entity known as a TPPU.  The
EPA considered revising the provisions for determining primary
product to exclude  the term process unit but found the
                             2-13

-------
regulatory language to be obtuse and less precise than the
final provisions.
     Comment:  Two commenters (IV-D-09, IV-D-18)  supported the
use of the primary product of a TPPU to determine
applicability for the proposed rule.
     One commenter (IV-D-25), however, contended that TPPUs
which produce only thermoplastic products should be exempt
from determining their primary product.  The commenter
(IV-D-25) explained that many TPPUs which produce PET are
capable of using either TPA or DMT as raw materials.  The
commenter (IV-D-25) stated that such TPPUs would be considered
flexible operation units under the proposed rule and would
have to fulfill the requirements of §63.500 (e) (3) of the
proposed rule.  Further, the commenter  (IV-D-25)  stated,
because both products are thermoplastic products, such process
units are unquestionably subject to the proposed rule and
determining which product is the primary product is wasteful
and unnecessary.  .
     Response;  The use of the primary product of a TPPU to
determine applicability has been retained in the final rule.
The primary product of a process unit is determined to see if
said process unit meets the definition of a TPPU.  If a
process unit meets the definition of a TPPU and meets the
other applicability criteria of §63.500(a)  [§63.1310(a)] in
the final rule, the owner or operator can determine the
primary product to know which standards the TPPU must meet.
 [Note:  The  standards for process vents and storage vessels
are different  for  some of the subcategories covered by the
thermoplastics rule.]
     The EPA understands that in the case of TPPUs which
produce PET  and  are capable  of using either TPA or DMT as raw
materials, determining the primary product may seem bothersome
and unnecessary.   However,  the EPA continues to find it
necessary for  the  owners or operator to determine the primary
product of a process  unit.   The EPA considered modifying the
provisions related to demonstrating compliance for flexible
operation units  to incorporate flexibility  for a process unit
                              2-14

-------
producing only thermoplastic products to avoid determining its
primary product.  However,  the EPA judged that the added
complexity to the rule was not justified.  Therefore,  the
final rule continues to require the owner or operator to
determine the primary product of each process unit.
     The final provisions for determining the primary product
have been expanded and clarified.  Language has been added to
the final rule specifying that the affected source is required
to comply only with the standards that pertain to the primary
product of the TPPU.  This provision will prevent owners and
operators from having to comply with two sets of standards if
two or more thermoplastic products are produced by a single
affected source.
     In the proposed rule,  the EPA stated that for flexible
operation units, the primary product should be determined
based on the expected operations for the 5 years following the
promulgation date for existing sources and for the first
5 years after the initial start-up for new affected sources.
The purpose of these provisions is to preclude the owner or
operator of the TPPU from having to redetermine their primary
product as a result of a small process change.  The 5-year
time frame was provided to coordinate with the operating
permits program.  The final rule continues to base the primary
product determination for flexible operation units on a 5-year
time frame.
     These provisions were further modified to provide
clarification of the EPA's intent and flexibility in complying
with the provisions.  Under the final rule, once an owner or
operator commits to being subject to this rule, there are two
options for complying.  Under the first option, an owner or
operator shall determine the group status  (i.e., Group 1 or
Group 2) of each emission point based on the production of the
expected primary product (i.e., the thermoplastic product that
convinced the owner or operator to commit to being subject to
this rule) .  Once the group status of each emission point is
determined, the owner or operator shall comply with the
applicable emission standards at all times, regardless of what
                             2-15

-------
product is being manufactured.  Under the second option,  an
owner or operator shall determine the group status of each
emission point each time a different product is being
manufactured, regardless of whether or not said product is a
thermoplastic product.  The EPA recognizes that neither option
is an ideal situation.  Under the first option, an owner or
operator may find themselves operating a control device to
control a Group 1 emission point that has no or negligible
emissions when a different product is being manufactured.
Under the second option, an owner or operator may find
themselves performing multiple group determinations.  Again,
the EPA recognizes that neither option is an ideal situation,
but believes the tradeoff between these inconveniences and
moving in and out of separate MACT standards is worthwhile.
     Provisions have been added to the final rule for
demonstrating compliance with the rule during those periods
when a flexible operation unit is producing a product other
than a thermoplastic product or is producing a thermoplastic
product that is not the primary product.  These provisions
require the owner or operator to establish new parameter
monitoring levels or to demonstrate that the parameter
monitoring levels established for the primary product are
appropriate for those periods when products other than the
primary product are "being produced.
     Comment:  One commenter  (IV-D-18) recommended that
§63.500 (f) (2) (ii) (B) of the proposed rule be revised to
address situations where multiple on-site and off-site process
units have input to or output from a storage vessel.  The
commenter  (IV-D-18) explained that the proposed provisions are
adequate if there is only one on-site process unit using the
storage vessel.  However, if there are two or more on-site
process units using the storage vessel, the proposed
provisions do not provide guidance.  The commenter  (IV-D-18)
stated that there may be many situations where this situation
occurs so it is important to provide clarification in the
rule.
                              2-16

-------
     The comtnenter  (IV-D-18) also recommended that
§63.500(f)(2)(ii) of the proposed rule be amended to address
situations where a  storage vessel serves both a TPPU and a
different process unit that is subject to another MACT
standard.  The commenter  (IV-D-18) stated that assigning the
storage vessel to both process units is duplicative and will
not provide additional corresponding benefit.  The commenter
{IV-D-18) continued that even if the two rules allow exactly
the same control technology, there will still be duplicative
recordkeeping and reporting burdens.  The commenter (IV-D-18)
suggested that the  rule be revised to avoid this situation by
allowing the owner  or operator to choose to follow the first
MACT standard to become effective for the storage vessel.
     Response:  The proposed rule was changed to address the
commenter's concerns.  In the final rule, situations where
off-site and on-site process units use the same storage vessel
are addressed by §63.500(g) (3) [§63.1310(g)(3) ] .  Under these
provisions, if a storage vessel is shared among process units,
then the storage vessel belongs to the process unit located on
the same plant site that has the greatest input into or output
from the storage vessel.
     In the final rule, situations where multiple on-site
process units use the same storage vessel and a predominant
use cannot be determined are addressed by §63.500(g)(4) and
(g) (5)   [§63.1310(g) (4) and  (g)(5)].  Under §63.500(g) (4)
[§63.1310 (g) (4) ], if predominant use cannot be determined for
a storage vessel that is shared among process units,  and if
one of the process  units is a TPPU subject to this subpart,
the storage vessel  shall belong to that TPPU.  Under
§63.500(g)(5) [§63.1310(g)(5)], if predominant use cannot be
determined for a storage vessel that is shared among process
units and, if more  than one of the process units are. TPPUs
that have different primary products and that are subject to
this subpart, then  the owner or operator shall assign the
storage vessel to any one of the TPPUs.
     Finally, situations where a storage vessel may already be
subject to another  MACT standard are addressed by
                             2-17

-------
§63.500{g)(1)  [§63.1310(g)(1)].  If a storage vessel is
already subject to another MACT standard, the storage vessel
must continue to comply with that standard and would not be
subject to the thermoplastics rule.
     Comment;  One commenter  (IV-D-10) stated that some of the
plant sites that produce the thermoplastic products covered by
the proposed rule also contain equipment used in the
production of elastomer products expected to be subject to the
elastomers rule  (subpart U).  For example, the commenter
 (IV-D-10) explained, ABS production plant sites may also
produce polybutadiene latex at the same  location.  The
commenter (IV-D-10)  recommended that  the EPA definitively
establish the  limits of the process operations affected by
each of these  future rules.
     Response;   The  EPA has added language to the final rule
in  §63.500 (e)  [§63.1310(e)] to clarify when equipment that is
used to produce  a non-thermoplastic product that is
subsequently used to produce  a thermoplastic product is
subject to this  subpart.   The provisions state that if a
polymer that is  not  subject to this subpart is produced within
the equipment  (i.e., collocated) making  up a TPPU and that at
 least  50  percent of  the polymer  is used  in the production of a
thermoplastic  product,  the unit  operations involved in the
production of  the polymer are considered part of the TPPU and
 are potentially subject to this  rule. These provisions do not
 apply if  the unit  operations  are already subject to another
 MACT  standard regulating the  same  emission points.
      Comment:   One  commenter  (IV-D-09) urged the EPA to remove
 the references to  by-product  and co-product from the
 applicability section in §63.500(a)(2) of the proposed rule.
 The commenter (IV-D-09)  noted that the preamble  (60 FR 16093)
 to the proposed rule states that "applicability will be based
 on the primary product ... by-products,  co-products, and
 isolated intermediates would not be considered  in  determining
 applicability."  The commenter (IV-D-09) also noted that
 according to HON BID Volume 2D,  by-product/co-product  language
 is not used in the HON rule.
                              2-18

-------
     Response:  The commenter is correct that the preamble to
the proposed rule stated that the applicability of the
proposed rule is based on the primary product for a process
unit, and this concept was retained in the final rule.
Further, the preamble also stated that by-products,
co-products, and isolated intermediates will not be considered
in determining applicability, however, this language is
misleading.  While by-products, co-products, and intermediates
are not considered products of a process, they are considered
in determining whether the rule applies to a TPPU.
     The proposed and final rule applies to TPPUs that emit
organic HAP in the manufacture of a primary product.  If the
primary product does not use as a reactant, or use as a
process solvent, or produce as a by-product or co-product any
organic HAP, then it is not subject to the final
thermoplastics rule.  Similar by-product/co-product criteria
are used in the HON rule.
     Comment:  One commenter  (IV-D-04) recommended that the
concept of  "predominant use" for applicability determinations
be expanded to include recovery operations.  The commenter
 (IV-D-04) was concerned that recovery operations which are
ancillary to the polymer manufacturing process  (i.e., recovery
operations whose predominant use is to serve a process other
than a polymer manufacturing process) may be considered part
of the TPPU unless clarification is made in the final rule.
For example, the commenter  (IV-D-04) explained, methanol and
ethylene glycol recovery operations, whose predominant use is
to serve a  PET manufacturing process, may have multiple
 "product streams" depending on the height and temperature
profile of  the distillation column serving the polymer
manufacturing process.  While the predominant use of these
recovery operations  is to purify material that is returned
directly to the PET  TPPU, minor side streams may be sent for
further purification to  another recovery operation that is
distinct from the  PET manufacturing process.  The primary
purpose or  predominant use of this other,  "non-PET" recovery
                              2-19

-------
operation is the purification of material from nonpolymer
manufacturing processes.
     The commenter  (IV-D-04) encouraged the EPA to establish a
process for assigning recovery operations to process units to
ensure that recovery operations integral to the polymer
manufacturing processes are covered by the rule, and that
recovery operations ancillary to the polymer manufacturing
process are not subject to the rule.  The commenter (IV-D-04)
requested that a new paragraph be added to §63.500, and that
this paragraph be worded the same as §63.500 (f) except that
the term "recovery operation" be substituted for "storage
vessel" in the new paragraph.
     Response:  Provisions to assign recovery operations
equipment to an individual process unit have been added to the
final rule  [§63.500(h)  [§63.1310(h)]].  These added provisions
are very similar to the provisions used to assign storage
vessels to process units  [§63.500(g) [§63.1310(g)]] and they
borrow principles from the provisions used to designate the
primary product of a flexible operation unit [§63.500(f)
[§63.1310(f)]].  Specifically, determining the predominant use
of recovery operations equipment for which operations vary
from year to year is based on a 5-year projection.  And, once
recovery operations equipment have been assigned to a process
unit, the owner or operator is never required to redetermine
the assignment.  If desired, an owner or operator may reassign
recovery operations equipment to another process unit at any
time.  However, such an action would likely require amending
the facility's operating permit or permit application because
the applicable requirements for the facility would change.
[Note:  Recovery operations equipment has been defined in the
final rule.]
2.2.4  Exemptions
     Comment;  Four commenters  (IV-D-03, IV-D-04,*IV-D-10,
IV-D-11) supported  §63.500 (c) of the proposed rule which
exempts vessels storing organic liquids that contain organic
HAP only as impurities from the proposed rule.   In addition,
three of these commenters  (IV-D-03, IV-D-04, IV-D-10)
                             2-20

-------
suggested the provisions be expanded to also exempt vessels
and equipment containing nonliquid material  (i.e., activities
such as storing, moving, and handling solid material) that
contain organic HAP only as impurities.  The commenters
(IV-D-03, IV-D-04, IV-D-10) also requested that the exclusion
apply to vessels that contain no organic HAP.  One commenter
(IV-D-10) also requested that this same change be made in item
(4) of the definition of storage vessel in §63.502 of the
proposed rule.
     Response;  The EPA has addressed the commenter's concerns
in §63.500(c)(5)  [§63.1310(c)(5)] of the final rule and item
(4) in the definition of storage vessel in §63.502  [§'63.1312].
The provisions now read as follows:
     Vessels and equipment storing and/or handling material
     that contains no organic HAP and/or organic HAP as
     impurities only.
     Comment;  Six commenters  (IV-D-04, IV-D-06, IV-D-11,
IV-D-17, IV-D-20, IV-D-21) discussed the provisions for solid
state processes collocated with PET processes in §63.500(d)(4)
of the proposed rule.  The commenters  (IV-D-04, IV-D-06,
IV-D-11, IV-D-17, IV-D-20, IV-D-21) agreed with the EPA's
conclusion that emissions  from independently located solid
state processes are insignificant and should be exempted from
regulation.  In addition,  four of the commenters  (IV-D-04,
IV-D-11, IV-D-17, IV-D-20) contended that all solid state
polymerization units, including collocated units, should be
exempted from regulation.  The commenters  (IV-D-04, IV-D-11,
IV-D-17, IV-D-20) stated that solid state polymerization units
are a vastly different technology than DMT and TPA processes
and have different emission  characteristics.
     One commenter  (IV-D-06) believed the EPA intended to
regulate solid  state PET processes which are directly
connected to the  manufacturing process.  The commenter
 (IV-D-06) stated  that the  proposed rule includes all solid
state PET processes located  at the same plant site.  The
commenter  (IV-D-06) suggested the provisions be reworded to
exclude  solid state PET processes not directly connected to
                              2-21

-------
the virgin resin manufacturing process.  The commenter
(IV-D-06) explained that solid state PET units not directly
connected to the virgin resin process are distinctly different
processes which should not be regulated under this rule.
     Three commenters (IV-D-11, IV-D-17, IV-D-20)  stated that
solid state processes should not be controlled unless the EPA
first performs a MACT floor determination and cost analysis.
One commenter  (IV-D-04)  stated that the solid state processes
should not be controlled unless the EPA establishes them as a
separate source category and conducts emission surveys from
all such sources as part of the rulemaking process.  Two
commenters  (IV-D-17, IV-D-20) recommended that the EPA
consider solid state processes to be a class or type of source
within each subcategory  (as allowed by Section 112 (d) (1) of
the Act) and determine if the requirement to control existing
sources beyond the floor is reasonable.  The commenters
(IV-D-17, IV-D-20) stated that if the EPA determines that
control of solid state processes is warranted, then all solid
state processes should be controlled, not just those that are
collocated with PET production processes.  The commenters
(IV-D-17, IV-D-20) contended that the EPA has provided no cost
justification  for controlling collocated solid state
processes, and likewise, has provided no cost justification
for excluding  stand-alone units.  Four commenters  (IV-D-04,
IV-D-il, IV-D-17, IV-D-20) stated that the proposed rule
places collocated units at an unwarranted competitive
disadvantage.
     Response;  The information gathering surveys used by the
EPA  (i.e.. Section 114 questionnaires and generic information
collection requests) did not distinguish solid state processes
from DMT or TPA processes.  At the time the surveys were
prepared, the  EPA did not have the specific knowledge required
to adequately  and accurately make a distinction between solid
state and DMT/TPA processes.  Second, to avoid precluding the
industry from  providing  information on a process unknown to
the EPA  or on  a process  that varied from the  "typical," the
EPA chose not  to distinguish PET processes in any great level
                             2-22

-------
of detail.  Therefore, those industry members who did provide
data on solid state processes, did not identify the
distinctions between solid state and DMT/TPA processes that
have become apparent during the comment period.  As a result,
the EPA had minimal data concerning solid state processes, and
the available data were incomplete because they did not
accurately portray the differences between solid state and
DMT/TPA processes.
     Given this, and based on the comments received since the
proposal, the EPA has concluded that solid state processes are
distinct from DMT/TPA processes.  Furthermore, the EPA has
little data concerning solid state processes, and the required
analysis for regulating solid state processes in the final
rule have not been performed at this time.  Therefore, solid
state processes will not be regulated as part of the final
rule.
     Comment;  One commenter  (IV-D-21) contended that as
written, the proposed rule implies that emissions from solid
state processes are part of the polymerization reaction
section.  The commenter  (IV-D-21) stated that where solid
state processes exist, they are distinctly separate processes
from melt-phase  (i.e., DMT/TPA)  PET production facilities.
The commenter (IV-D-21) also stated that the inference that
solid state processes are part of the polymerization reaction
section causes difficulties in determining compliance.  The
commenter (IV-D-21) explained that at their facilities, solid
state processes are operated and managed separately from the
production of low molecular weight PET.  The commenter
(IV-D-21) requested that solid state processes not be
considered part of the polymerization reaction section, and
that the definition of "polymerization reaction section"
specifically state this.
     Response;  As explained in response to the previous
comment, the final rule does not regulate solid state
processes.  The EPA did not intend for unit operations
involved in the solid state process to be included in the
polymerization reaction section.  In the final rule, solid
                             2-23

-------
state processes have been excluded from the affected source in
§63.500(d) (4)  [§63.1310 (d) (4)] .
     Comment:  One commenter  (IV-D-04) disagreed with the
EPA's rationale that, because HAP emissions from independently
Located solid state processes are not significant,
independently located solid state processes are likely to be
"nontnajor  sources," and thus, they would not be subject to the
proposed rule.  The commenter  {IV-D-04) explained that the
determination as to whether a source is "major" or "area"
(i.e., nonmajor) is based on the collective HAP emissions
potential  of all sources located at the contiguous plant site,
not solely those emission points associated with a specific
process.   The  commenter  (IV-D-04) was concerned that the
proposed rule  establishes an artificial distinction between
collocated and independently  located solid state processes
that will  arbitrarily assign  a competitive disadvantage to
collocated processes, irrespective of HAP emission potential.
     Response:  The EPA agrees with the commenter in principle
(i.e., a  source is major or area based on the collective HAP
emissions  from all activities at the contiguous plant site).
The EPA made the statements concerning independently located
solid  state  processes based on limited information indicating
that these processes would be the only activity at these plant
sites.  As explained in response to an earlier comment, the
EPA did not  collect or receive data sufficient to analyze
solid  state  processes.  As a  result, the final rule does not
regulate  solid state processes.
2.3  SELECTION OF  POLLUTANTS
     Comment:   Numerous commenters  (IV-D-07, IV-D-08, IV-D-09,
IV-D-10,  IV-D-11,  IV-D-18, IV-D-21) recommended that the EPA
restrict  the list  of organic  HAP in the final rule to those
that are  used or are present  in significant quantities at
TPPUs  (IV-D-11) or to those  that are  listed in the HON,
subpart F, Table 2.  One  of  these commenters  (IV-D-09)
requested that once  the major organic HAP of concern have been
identified,  the owner or  operator only be required to identify
any other organic  HAP contained on the subpart F, Table 2 list
                              2-24

-------
which are:  (1) specifically used in the TPPU, and  (2) present
in any material introduced into the TPPU provided the used or
introduced material contains organic HAP above significant
levels  (1 percent for noncarcinogens or 0.1 percent for
carcinogens).
     Some commenters  (IV-D-07, IV-D-08, IV-D-09, IV-D-11)
stated that  the definition of organic HAP in the proposed rule
is confusing and will be difficult for the regulated community
to interpret and for the regulators to enforce.  Three
commenters  (IV-D-08, IV-D-09, IV-D-11) noted that the proposed
definition of organic HAP would include all 189 HAP "commonly
considered to be organic" and "contain(ing) one or more carbon
atoms."  Two commenters  (IV-D-09, IV-D-11) explained that as
defined, the owner or operator would be required to determine
which HAP listed in the Act are organic and if any are present
at the TPPU.  Some commenters (IV-D-08, IV-D-09, IV-D-11)
stated that  the proposed approach will require needless and
burdensome analysis for a large number of chemicals not
typically present at thermoplastic manufacturing facilities.
Also, the commenters  (IV-D-07, IV-D-08, IV-D-11) claimed that
the definition is so imprecise that it would include certain
HAP, such as polycyclic organic matter (POM), that are defined
so broadly in the Act that they encompass whole classes of
materials, such as polymers  (for instance polystyrene and
PET) .
     One commenter  (IV-D-08) questioned the applicability of
the required monitoring methods to organic HAP which are not
listed in Table 2 of the HON.  The commenter (IV-D-08)  stated
that hydrogen cyanide is one such HAP for which monitoring by
required methods would be problematic.  The commenter
(IV-D-08) stated that limiting organic HAP to those in Table 2
of the HON would cover all HAP expected from thermoplastic
facilities and eliminate any potential problems with
monitoring.
     One commenter  (IV-D-18) indicated five specific concerns
with the definition as proposed; they are as follows:
                             2-25

-------
      (1) If Congress or the EPA identifies more HAP,  this
definition will not acknowledge them; it addresses only the
original 189 HAP.
      (2) The definition uses two different tests to identify
an organic HAP.  The two tests might produce different
answers.  First, the definition says an organic HAP is any HAP
"commonly considered" to be organic.  Then, the definition
says  a HAP is considered organic if it contains one or more
carbon atoms.   It seems quite possible that a HAP might have a
carbon atom, but not be commonly considered organic.   For
example, the cyanide grouping  (in substances such as hydrogen
cyanide) includes a carbon atom, but is commonly considered
inorganic.
      (3)  The EPA should not base the definition on whether a
substance is commonly considered organic.  This technique
seems to require some sort of  a poll or survey.  Furthermore,
it does not identify the group of people whose opinion
matters.  In the absence of further clarification, the
commenter supposed that the definition must mean  "commonly
considered by  the general public."   The commenter doubted
that  most members of the general public have any opinion of
whether specific substances are organic or not.  And, in cases
where opinions  differed, what  would they do.  For example, the
CRC Handbook of Chemistry and  Physics lists carbon
tetrachloride  in its table of  physical constants  for  inorganic
compounds.  We could interpret this to mean that  carbon
tetrachloride  is "commonly" considered inorganic; however, at
least in  the regulatory field, we believe  that carbon
tetrachloride  is "more  commonly" considered organic.  The HON
 (Table  2  to subpart F)  lists  carbon tetrachloride as  an
organic HAP.   As you can see,  considerable confusion  will
result  if the  EPA bases its definition on whether a substance
is  "commonly considered"  organic.
      (4)  The EPA should not base this definition  on whether  a
substance contains a carbon atom.   Inorganic carbon compounds
 (such as  cyanide compounds)  exist,  and  some are currently HAP
                              2-26

-------
or may be listed as such in the future.   They are still
inorganic, even though they have carbon in them.
     (5) The proposed definition would let the applicability
of this r.ule expand in unforeseeable ways.  For example,
Congress has identified "cadmium compounds," "mercury
compounds," and "nickel compounds" as HAP.  There must be
literally thousands (perhaps tens of thousands, or even more)
of individual substances that contain cadmium, mercury, or
nickel somewhere in their molecular structure.  Some of those
compounds probably contain carbon and are commonly considered
organic.  These substances may exist in nature and, may
possibly be trace contaminants in incoming raw materials, or
even be present in metallic process equipment.
     To avoid these five difficulties, the commenter  (IV-D-18)
suggested that the EPA borrow, by cross-reference, the
definition of "organic hazardous air pollutant" from
subpart F.  The commenter  {IV-D-18) explained that the
definition refers to Table 2, where the organic HAP are listed
by name and by the CAS registry number.  Alternatively, if
there is something about Table 2 that is inappropriate for the
proposed rule, the commenter  (IV-D-18) suggested that the EPA
make up a new table that lists each of the organic HAP of
concern.  Either approach would eliminate all ambiguity, and
it would also allow the EPA to incorporate any newly listed
organic HAP by adding them to the table.
     Response:  In response to requests from commenters, the
EPA has developed a list of known organic HAP for each
subcategory that will be regulated by the thermoplastics rule.
The list is presented in Table 6 of the final rule and is
referred to by the revised definition of organic HAP found in
§63.502  [§63.1312].  This table appears below as part of this
response.
     The revised definition was developed using available
process description information received from industry and
gathered from available literature.  Because there may be
additional organic HAP present at an affected source that were
not included in Table 6, the  final rule requires owners or
                              2-27

-------
 operators to notify the EPA of the presence of  any additional
 organic HAP based on the following criteria:   (1)  the  organic
 HAP is knowingly introduced into the manufacturing process
 other than as an impurity,  or the organic HAP has  been or will
 be reported under any Federal or State program,  such as EPCRA
 or Title V; and (2)  the organic HAP is presented in Table 2  of
 subpart F.
 2.4  SELECTION OF THE STANDARDS (GENERAL)
      Comment;   One commenter (IV-D-02)  supported the use of
 performance standards in the proposed rule.  The commenter
 (IV-D-02)  contended that allowing flexibility in complying
 with the rule ensures that  the most practical and
 cost-effective options are  available.
      Response:   The thermoplastics rule is  written as  a
 performance standard to allow the use of any control device
 that meets  the performance  requirements.  Nothing  in the rule
 precludes any technology from competing in  the marketplace.
 The EPA even provides provisions  for alternative control
 technologies and promotes innovative control strategies.
      Comment:   Four commenters (IV-D-11,  IV-D-17,  IV-D-20,
 IV-D-23)  believed the EPA had not complied with  the
 requirements of section 112 (d)  of the Act in developing the
 proposed thermoplastics rule.   The  commenters (IV-D-ll,
 IV-D-17,  IV-D-20,  IV-D-23)  contended that the EPA had not
 determined  the  MACT  floor,  as  required  by the Act.
      Response:   The  EPA disagrees with  the commenters
 assertion that  the requirements of  the Act were not met  in
 developing  the  thermoplastics  rule.  A  description of the
methods  used by the  EPA to conduct  the MACT floor analysis for
 the rule were  included in the  Basis  and Purpose Document for
 the proposed thermoplastics rule  (EPA-453/R-95-004a;
March 1995).  Specific comments about the MACT floor analysis
and selection of the standards for  different emission points
and subcategories  are  included in this document  in  Chapter 7.0
for wastewater  and Chapter 8.0 for  equipment leaks.
                             2-28

-------
 09
 4->
 U
 S
 u
-H
4J
 CQ
 (C
 O


 0)
 8
 CQ
 4J
 S
 rt
 O
 PU
-H
<:

 CQ
 (0
 N
 cC
td
 u
•rl
 S
 E1
o


 I
 a)
rH









i
ra
u
P ^*
|s
> »
|s
u
CD
O



01 tn
f >
S IM
u-f
IT
£>?
Is
"o
SJ3
«fe
Ul

c '^
x »L
.S »>
Q SQ
*^C

0)
1
JB
L. r~
o ^
II
4-*
«
•5
Thermoplastic Pro
Subcategory



S














-
-




a
CO
S


















-
s




ABS using a batch
|| emulsion process



s














s
-



u
«ABS using a batch
suspension proces


















-
-


CO
^
«ABS using a conti
emulsion process



s














s
s


3
£
1
u
in

0)
.52
'« a
tn
il


















s
-




II ASA/AHSAN



s




















£


















-
-




in
1


















-





i



















-




c
0)
u
u
'£!




^








s







-
01
a
a
IPET using a batch
dimethyl terephth
process













s.







-


PET using a batch
terephthalic acid
process




s








s







V
|S
c a
•- .e
If
e ai
4^
SI? a













s







-
3
^
IPET using a conti
terephthalic acid
process







s>





^







-
!«?
B* a
PET using a conti
terephthalic acid
viscosity multipl
finisher process
                                                                                                                                           c\
                                                                                                                                           N
                                                                                                                                            I

-------
 tn

 o



!
 CM

 U
•ri
 4J
 ra
 O




 0)
 8
 4->_^
 m


 O
 rt
 N


 W

 O
 •H
 a



 I
 o
                     8^

                             U JQ
                             c.
                             x a



                             £P



                             II
                                    ss
                                    C- •!-
|58
                             o
                             a


                             2
                             ai
                                  a>
                                  c
                                  01
                             CD    CO
                             r-    44


                         .55-  -a
                         I  U> CD      _
                         i  rtj *w    QJ C

                           t- C    —•-
                                                     D>
                                                            L. u) oi  « >- eo  a
                                                            (u   u  L. c. —•  eu

                                                               c 4J  v eg j:
                                                            4J 0> CO  C   +*  Ctt
                                                            o .X    5 cu -c —•
                                                            a TJ «»  i- *> Q.---

                                                            tSr  i^- £t
                         CO ••- —»   4J  0> C-

                         O t- X «l 9J—; O
                        — 4-" i. —• E  >• a
                                                                                                                                    o

                                                                                                                                    CO
                                                                                                                                     1

-------
     Comment;   Three commenters (IV-D-08, IV-D-11, IV-D-23)
contended that use of the HON, the Polymers Manufacturing
NSPS, and the Batch Processes ACT as regulatory alternatives
to the MACT floor is reasonable because they control similar
emission points and have been extensively analyzed and
demonstrated to be reasonable from a cost and impact
perspective.  However, the commenters  (IV-D-08, IV-D-11,
IV-D-23) argued that a regulatory impact analysis must be
performed for any deviations from these alternatives that
result in a significant increase in stringency, cost of
compliance, or regulatory complexity.  The commenters
(IV-D-08, IV-D-11, IV-D-23) stated that, for the proposed
rule, the underlying standards have been modified in
significant ways, but the necessary impact analyses were not
performed.  The primary modification cited by the commenters
(IV-D-08, IV-D-11, IV-D-23) was the lack of excused
excursions.   [Note:  Comments concerning excused excursions
are summarized and addressed in Chapter 11.0, Monitoring.]
     Response;  A formal regulatory impact analysis  (RIA)
document was not prepared for this rulemaking.  A Regulatory
Impact Analysis is only required for rules that are considered
to be significant  (i.e., a total annual cost of greater than
$100 million); the thermoplastics rule is not a significant
rule.  The Act clearly requires the EPA to justify selecting
standards beyond the MACT floor based on impact analyses, and
the costs of the regulatory alternatives and of the economic
impact were determined prior to proposal.  The EPA disagrees
with the commenters that any changes to the underlying
standards must be subject to an impact analysis.  As stated
previously,.the appropriate point at which impact analyses
should begin is the MACT floor.
     With the exception of the lack of excused excursions,
which is discussed in Chapter 11.0, Monitoring, the EPA
examined each instance where the selected standards were more
stringent than the MACT floor.  For those emission points and
subcategories where the selected standards are more stringent
than the MACT floor, impacts are presented in the Basis and
                             2-31

-------
Purpose Document  [EPA-453/R-95-004a] ,  Chapter 8.0, Section
8.3', Rationale for the Selection of the Proposed Standards.
For those emission points and subcategories where the selected
standards are equivalent to the MACT floor, the EPA is not
required to justify the selection of the standards based on
cost or other impacts; these impacts were not presented in the
Basis and Purpose Document  [EPA-453/R-95-004a] .  However,
impacts of the proposed rule were determined as part of the
economic impact analysis, Docket A-92-45, Item No. II-A-8;
these impacts are presented in a memorandum titled "Summary of
Cost, Emission Reduction, and Energy Impacts for Group IV
Resin Sources," dated March 24, 1995  (Docket A-92-45, Item No.
II-B-27).
     Comment:  Two commenters  (IV-D-10 and IV-D-18) asked for
clarification on  what is considered "batch" and what is
considered  "continuous" in  situations where a TPPU is made of
various unit operations, some of which are batch while others
are continuous.   One  commenter  (IV-D-18) supported the
following interpretation of the proposed rule, and asked for
confirmation that their understanding of the provisions was
accurate.
      (1) It is possible to  have a  "batch" unit operation
within  a  "continuous" TPPU.
      (2) It is possible to  have a  "continuous" unit operation
within  a  "batch"  TPPU.
      (3) Within  a TPPU, process vents from a  batch unit
operation are batch process vents.  Process vents from
continuous  unit  operations  within  the same TPPU are continuous
process vents.
      (4) The determination of whether a  TPPU  is  "batch"  or
 "continuous" does not depend on any single piece  of equipment,
 such as a  reactor.   Rather,  it is  based  on the overall nature
 of the  process  unit.
      Response;   The EPA agrees with the  commenter's
 interpretation and understanding of the  proposed  rule, with
 the exception of item (4)  in the above  comment  summary.  The
 EPA distinguishes a TPPU as using a continuous or batch
                              2-32

-------
process depending on how the polymerization reactors are
operated.  If the determination followed a less specific
criteria  (e.g., the overall nature of the process unit), there
would be too much room for interpretation and confusion.  In
the final rule, the definitions of continuous process and
batch process have been modified to indicate that a process
producing polymers is characterized as continuous or batch
based on the operation of the polymerization reactors.
2.5  COORDINATION WITH OTHER CLEAN AIR ACT REQUIREMENTS
2.5.1  Subpart A  (General Provisions)
     Comment:  One commenter (IV-D-18) commended the EPA for
providing Table 6 in the proposed rule, which specifies the
portions of the General Provisions that apply to affected
sources.  Another commenter (IV-D-10) suggested that the final
rule include language explaining the applicability of the
General Provisions to this subpart.  Another commenter
(IV-D-09) maintained that Table 6 of the proposed rule
referenced many marginal or nonapplicable paragraphs from the
General Provisions which would lead to confusion.  The
commenter  (IV-D-09) suggested that Table 6 be revised to
improve clarity and to correspond to other references in the
proposed rule.  The commenter  (IV-D-09) included nine
suggested clarifications.
     First, the commenter  (IV-D-09) recommended that the
references to  "reserved" paragraphs in the General Provisions
say "No" not "Yes" as they did at proposal.  The commenter
(IV-D-09) explained that the notation "No" should be shown so
that if these paragraphs are added to the General Provisions
at a later date,  they will not automatically apply until an
affirmative decision on applicability is made.
     Second, the  commenter  (IV-D-09) stated that the inclusion
of §63.1(a)(3) as applying to subpart V will not overcome the
stringency determination required of an owner or operator or
enforcement official.  The commenter  (IV-D-09) stated that
even though  §63.501(e) and  (f) of the proposed rule would
appear to provide that they are more stringent than 40 CFR
part 60,  subpart  Kb and subpart DDD, respectively, unless some
                             2-33

-------
reference to stringency is included in them, there may be a
possibility that a violation of subpart V might also be
considered a violation of a respective part 60 subpart as
well.  The commenter  (IV-D-09) stated that this could lead to
two violations in each case, not one if the subpart V
provisions, in fact,  totally override the part 60 subpart.
The commenter  (IV-D-09) included suggested language.
     Third, the commenter  (IV-D-09) recommended that Table 6
of the proposed rule  note that §63.Kb)(3) and §63.10(b)(3)
should not apply because the Enabling Document for the General
Provisions states that the EPA plans to delete §63.1(b)(3) in
upcoming amendments to the General Provisions.  The commenter
 (IV-D-09) explained that §63.10(b)(3) is the companion
recordkeeping provision for the negative determinations
required by  §63.Kb) (3) .
     Fourth, the commenter  (IV-D-09) stated that  the HON shows
§63.1(a) (4)  through  (a) (11),  (b) (1),  (b) (3) ,  (c) (1) through
 (c) (3),  (c) (5),  (d),  and  (e) as not  applying and  therefore,
the provisions  should not  apply to this rule for  the same
reasons  used in the HON.
     Fifth,  the commenter  (IV-D-09)  stated that the compliance
dates  referenced  in  §63.6 (b) (1) and  (2) should not apply
because  the  subpart V compliance  dates  are  included in
 §63.502.  The  commenter (IV-D-09)  requested that  the table
show  "No"  for  this reason.
     Sixth,  the commenter  (IV-D-09)  stated  that §63.6(b)(4)
through (c) (4)  should be  shown as "No"  for  the same reasons  as
 shown  in the HON table.
      Seventh,  the commenter (IV-D-09)  stated that not all of
 §63.6 (i) applies  because some of  the provisions apply to
 section 112(f)  compliance  extensions.   The  commenter  (IV-D-09)
 explained that the HON table does a good  job of indicating
 which subparagraphs  apply or do not apply and  should be used
 in Table 6 as well.
      Eighth, the commenter (IV-D-09)  stated that  §63.7(g)
 should not apply for the same reason it did not  in  the  HON
                              2-34

-------
table.  The HON table states that performance test reporting
is specified in the rule.
     Ninth, two commenters  (IV-D-09,  IV-D-10) stated that  the
entry for  §63.8 should say  "No" since the paragraph is broken
down into  subparagraphs in  the table.
     Response;  Table 6 of  the proposed rule was reviewed  and
revised between proposal and promulgation to ensure
consistency and clarity between the final thermoplastics rule
and the General Provisions.  The table has been renumbered
Table 1 to reflect the order in which the tables are mentioned
in the text of the final rule.  Language has also been added
to §63.501(f)  [§63.1311(f)] of the final rule directing the
reader to  Table 1 for guidance on which provisions of subpart
A apply to this rule.
     As requested by the commenter, the table now indicates
that General Provisions paragraphs which are reserved do not
apply to the thermoplastics rule.  If  these reserved
paragraphs are replaced with provisions at a later date, the
EPA will make a determination as to whether the provisions
apply to the thermoplastics rule.
     The second comment regarded stringency determinations and
the commenter's concern that there may be a possibility that a
violation  of this subpart could also be counted as a violation
of another subpart, thus resulting in  two violations for one
occurrence.  To ensure that "double counting" of violations
does not occur, §63.501 [§63.1311]  of  the final rule specifies
the subparts that do or do not apply to the thermoplastics
rule.  By  specifically stating in the  rule whether another
subpart applies or not, situations where two violations may be
assigned for one occurrence can be prevented.
     The third commenter requested that §63.1(b)(3)  and
§63.10(b) (-3) not apply to the thermoplastics rule.   The EPA
agrees and this decision is reflected  in Table 1 of the final
rule.  In  the final rule,  §63.500(b)   [§63.1310(b)]  provides
documentation requirements to show that a source is not an
affected source, and thus,  not subject to the thermoplastics
rule or subpart A requirements.
                             2-35

-------
     In response to the fourth comment, the EPA reviewed the
specific provisions the commenter discussed and the EPA's
responses are as follows.  Paragraph 63.1(a)(4) of subpart A
applies to the final thermoplastics rule; this paragraph
states that the relevant standard  (i.e., subpart JJJ, Table 1)
specifies the provisions of subpart A that apply.  Paragraphs
63.1(a) (5),  (a) (9),  (c) (3), and  (d) do not apply, as these
paragraphs are reserved.   Paragraphs 63.1(a)(6) through  (a)(8)
apply to this rule; they provide information about obtaining a
list of source categories  to be regulated under section 112 of
the Act and information about 40 CFR part 63,  subparts D and
E.  Paragraph 63.1(a)(10)  does not apply; the  final
thermoplastics rule specifies calendar or operating days.
Paragraph 63.1(a)(11), pertaining to postmarks, applies to  the
final rule.  Paragraphs 63.1(b)(l) and  (c)(1), applicability
determinations, apply  to the thermoplastics rule.  Paragraphs
63.Kb) (3) and  (c) (2)  do not apply to the thermoplastics rule.
Paragraph 63.1(c) (5),  which applies, states that if an  area
source  increases  its emissions so  that it becomes a major
source, then that  source is subject to major source
requirements and  emission  standards, and  also  notification
requirements; Table  1  states that  affected  sources are  not
subject to the  notification requirements  overridden by
Table 1.  Paragraph  63.1(e), applicability  of  the permit
program before  a  relevant  rule has been set under this  part,
applies to this final  rule.
     In response  to  the fifth comment,  §63.6 (b) (1) and  (b) (2)
apply to the thermoplastics rule.  These  paragraphs  state that
the  owner or operator must comply with the  applicable rule  by
the  effective  date of  the  rule or by start-up.
     In response  to  the sixth comment,  the  commenter is
referred to  Table 1  for guidance as to which of  the  paragraphs
are  applicable to the rule.
     In response  to  the seventh  comment,  all of  §63.6 (i)
applies, with  the exception of  §63.6(i)(15), which is
reserved.
                              2-36

-------
     In response to the  eighth  comment,  §63.7(g)  does  apply,
except that the references to the Notification of Compliance
Status report in §63.9(h) of subpart A are  replaced with the
requirements in §63.513 (e) (5)  [§63.1335(e) (5)]  of subpart JJJ.
     In response to the  ninth comment,  §63.8 is broken into
subparagraphs in Table 1.  The  general entry in the table at
proposal for §63.S has been removed from the final rule.
     Comment:  One commenter  (IV-D-09)  supported  the reference
to compliance extensions in §63.6 (i) referenced in §63.513(b)
of the proposed rule  if  language is added to incorporate
several provisions contained in the HON [§63.151(a) (6)],  as
amended by the upcoming  HON technical  correction  notice
scheduled for late 1995.  The commenter (IV-D-09)  stated that
the needed changes are as follows.  The information required
in the request for the extension under §63.6(i)(6)(i)  need
only include the elements of §63.6 (i) (6) (i) (A),  (B) , and (D) .
The element §63.6(i) (6) (i) (C) is only  required for extensions
of compliance connected  with section 112(f), not  section
112 (d)  standards.  The submission of the request  for the
compliance extension  should be  able to be made at any  one of
three times --as part of the operating permit application or
as part of the Initial Notification  [§63.513(b)(1) of  the
proposed rule] or as  a separate submittal.
     Response:  §63.513 (e) (3)  [§63.1335(e) (3)]  of the  final
rule contains the requirements  for submitting  a request for an
extension.  These provisions, which were contained in
§63.513(b)(1)  of the  proposed rule, were revised  to require
that a request for an extension of compliance  be  submitted in
the Precompliance Report if it  had not already been submitted
in an operating permit application.  The request  for a
compliance extension  must include the  information in
§63.6 (i) (6) (A) , (B) ,  and (D) of subpart A.   [Note:  The
Initial Notification  was removed from  the final rule.]
2.5.2  Subparts F. G. H.  I  (HON)
     Comment:  One commenter  (IV-D-03)  asked the  EPA to reopen
the comment period for this rule if the HON wastewater
provisions,  which were referenced by the proposed rule and
                             2-37

-------
which are currently under litigation, are revised.  Another
commenter  (IV-D-09) recommended that all of the wastewater
provisions in the final rule reference the HON  (subpart G)
wastewater provisions so that all final resolutions of the HON
litigation will automatically be included in this rulemaking.
     One commenter  (IV-D-10) was concerned that the EPA may
not have sufficient resources to fully evaluate the impact to
this rule of changes that are made to the HON as part of the
litigation.  Another commenter  (IV-D-24) asked how changes
that may be made to the HON will be incorporated into this
final rule.
     Response;  As part of the HON litigation proposal, the
EPA requested comments specific to the thermoplastics rule.
If comments specific to the thermoplastics rule are received,
they will be addressed as part of the HON rulemaking actions
or in actions specific to the thermoplastics rule, depending
on the  comments.   Therefore, the comment period for the
thermoplastics  rule will not specifically be reopened.
     The EPA agrees that the wastewater provisions and some
other HON  provisions should be referenced in the
thermoplastics  rule so that future revisions to the HON will
automatically be included in the thermoplastics rule.
However, all changes made to the HON will be evaluated by the
EPA to  determine applicability to this rule.  The  "automatic"
part refers to  the fact that text changes will not need to be
made to this rule  once the EPA  finds the HON changes to be
applicable.  If the EPA determines that any changes to the HON
are not applicable to this rule, the thermoplastics rule will
be revised as appropriate.   If necessary, a proposal to
revise  the thermoplastics rule would be issued.
      Comment:   Two commenters  (IV-D-11, IV-D-23) supported the
EPA's proposal  to  require MBS and MABS affected sources to
comply with the provisions  of this rule for equipment leaks
after the  compliance date,  and  no longer require these
affected sources to comply with subpart I.  The commenters
 (IV-D-11,  IV-D-23)  requested that the applicability listing of
MBS  and MABS resins in  §63.190 (b) of subpart I  and the
                             2-38

-------
definitions of the two resins in §63.191 of subpart I be
deleted, and that this action be effective on the same date as
the compliance date for the equipment leak provisions of this
rule.
     One commenter (IV-D-10) requested that proposed
§63.509(b) of the proposed rule be changed to allow sources
previously subject to subpart I to begin complying with the
thermoplastics rule prior to the compliance date, but after
the promulgation date.  The commenter (IV-D-10) included
suggested language.  In addition, the commenter  (IV-D-10)
requested that §63.509(b) of the proposed rule be expanded to
include language clarifying that an affected source which has
already satisfied the initial reporting requirements of
subpart I not have to repeat the Initial Notification or
Notification of Compliance Status reports specified in
§63.182(b) and (c) of subpart H.
     Response:  In §63.501(d) of the proposed rule, MBS and
MASS affected sources were required to comply with the
provisions of subpart I until the compliance date(s) for the
thermoplastics rule.  As specified in §63.501(d)(1) of the
proposed rule, after the compliance date(s),  these affected
sources would no longer be subj ect to subpart I.
     The EPA believes that this language,  which has been
modified slightly in the final rule [§63.501(g)(1)
[§63.1311(g) (1)]  of the final rule], is sufficient to clarify
the relationship between the two rules,  without the need to
revise subpart I as suggested by the commenters.  Therefore,
the EPA is not proposing to revise subpart I.
     In addition, as the result of this comment and as
discussed in the responses below to several other comments,
the EPA has modified the language for similar paragraphs in
§63.501  [§63.1311] of the final rule to clarify that affected
sources subject to this subpart that are also subject to a
specific other subpart are no longer subject to the other
subpart after the compliance dates specified in §63.501
[§63.1311] of the final rule.  However,  affected sources
subject to both this subpart and subpart Q of 40 CFR part 63
                             2-39

-------
are subject to both subparts and are required to comply with
both subparts.
     One commenter requested that affected sources be allowed
to begin complying with the thermoplastics rule prior to the
compliance date, but after the promulgation date.  The EPA
considered this possibility but did not change the final rule
to allow this.  Under subpart I, and subsequently under the
thermoplastics rule, affected sources can reduce their
monitoring and recordkeeping burden by implementing a quality
improvement program.  §63 .501 (g) (2) [§63 .1311 (g) (2) ]  of the
final rule ensures that an affected source previously subject
to subpart I is not penalized as compliance switches from
subpart I to the thermoplastics rule [§63.501 (g) (2)
[§63.1311(g) (2)]] in terms of the quality improvement
programs.  The EPA considers this provision to be sufficient
to provide any benefit that could be gained in complying with
the thermoplastics rule prior to the compliance date.
     The same commenter requested that an affected source not
be required to repeat submittal of information provided in the
Initial Notification or Notification of Compliance Status
provided as part of complying with subpart I.  The final rule
no longer requires an Initial Notification.  However, the EPA
still considers the Notification of Compliance Status an
essential'element in enforcing the rule.
     Comment;  One commenter (IV-D-18)  asked the EPA to
clarify that the proposed rule is not intended to amend the
HON.  The commenter  (IV-D-18) noted that the proposed rule
references provisions contained in the HON and specifies
necessary changes and adaptations.  The commenter  (IV-D-18)
was concerned that in instances where the proposed
thermoplastics rule states that the proposal and promulgation
dates of the proposed thermoplastics rule "replace" the
corresponding HON dates,  the FEDERAL REGISTER staff could
interpret this as an instruction to amend the HON.  The
commenter  (IV-D-18) recommended that the words  "shall apply"
be used in place of  "replaces."
                             2-40

-------
     Response;   The EPA does not intend for the HON rule to be
amended in any way as a result of the wording in the
thermoplastics rule.  To reduce any chance of confusion, the
EPA has used the words "shall apply" in place of the word
"replaces" when necessary in the final rule.
2.5.3  NSPS
     Comment:  Three commenters (IV-D-11, IV-D-17, IV-D-20)
recommended that the final rule specifically state that
compliance with the fugitive emission leaks NSPS  (subpart W)
and the distillation NSPS (subpart NNN) are not required for
sources that are subject to the thermoplastics rule. The
commenters (IV-D-11, IV-D-17, IV-D-20) referred to §63.501(e)
and (f) of the proposed rule which state that storage tanks
and process vents that are subject to both an NSPS and the
MACT standard would only be required to comply with the MACT
standard.  The commenters (IV-D-11, IV-D-17, IV-D-20) stated
that equipment leaks from new or modified production processes
or distillation columns that qualify as part of a SOCMI
process could be covered redundantly and create enforcement
and compliance difficulties with no corresponding
environmental benefit  (i.e., methanol recovery sections at PET
affected sources could be covered redundantly).
     Response:   The final rule has been clarified under
§63.501  [§63.1311] by adding new paragraphs (k) and  (1)  which
state that compliance with the provisions of the fugitive
emission leaks NSPS  (40 CFR part 60, subpart W) and the
distillation NSPS (40 CFR part 60, subpart NNN) are not
required for sources that are also subject to the
thermoplastics rule, as of the compliance date; such sources
are to comply with the provisions of the thermoplastic rule.
As noted in an earlier response, each paragraph also states
that affected sources subject to the thermoplastic rule that
are also subject to subpart W or NNN, as appropriate, are no
longer subject to the subpart W or NNN, as appropriate, after
the compliance dates specified in §63.501 [§63.1311].
                             2-41

-------
     Comment:  One commenter  (IV-D-18) supported the EPA's
decision that the proposed thermoplastics rule override the
NSPS for storage vessels  [40 CFR part 60, subpart Kb] .
     Response:  The commenter is correct that, after the
compliance date of the final thermoplastics rule, a storage
vessel at an affected source that is also subject to 40 CFR
part 60, subpart Kb, is only required to comply with the
provisions of the final thermoplastics rule.  This is stated
in §63.501 (h)  [§63.13'11 (h) ] of the final rule.  Also, as noted
previously, the final rule states that, after the compliance
dates specified in §63.501(d)  [§63.1311(d)], sources subject
to the thermoplastics rule that are also subject to subpart Kb
are no longer subject to  subpart Kb.
2 .6  SUPPORT FOR OTHER SUBMITTED COMMENTS
     Comment;  Five commenters  (IV-D-03, I-V-D-09, IV-D-15,
IV-D-16, IV-D-23) supported the comments submitted by the
Chemical Manufacturer's Association  (CMA)  (IV-D-11).  Three
commenters  (IV-D-15, IV-D-23,  IV-D-24) supported the comments
submitted by the Society  of the Plastics Industry, Inc.,  (SPI)
 (IV-D-08).
     Response;  In drafting the final rule, the EPA has taken
into consideration that the commenters listed above support
the comments of CMA and SPI.   The specific  comments of CMA and
SPI have been  considered  and  responded to in this document.
2.7  DEFINITIONS
     Comment;  One commenter  (IV-D-10) requested that all the
definitions  in the proposed rule be reviewed.  The commenter
 (IV-D-10)  stated that the proposed definitions are confusing
and need clarification.   The  commenter  (IV-D-10) suggested
that the same  definition  of a term be used  throughout part 63,
that it  appear once within part 63, and  that where the term
requires redefining, it be redefined  in  and for the specific
section  of the subpart  which  requires the redefinition.  Also,
the commenter  (IV-D-10) requested that definitions be
 "rebuilt"  so that  the meaning of phrases or terms are
understood from  the meaning of the words or terms from which
they are composed.
                              2-42

-------
     Response:  The EPA reviewed the definition section of the
proposed rule, §63.502, between proposal and promulgation.  As
a result of this review, many definitions were revised and
clarified for the final rule.  The EPA has referred to subpart
A, F, G, and H definitions in the final rule when applicable,
and has redefined terms that either do not share the same
definition as in the other subparts or were not defined in a
previous subpart.  For easier reference, the organization of
terms referenced in subparts A, F, G, and H was changed in the
final rule; referenced terms are listed in alphabetical order
with the subpart where they are defined specified in
parentheses after each term.
     The EPA has also reviewed and revised (or "rebuilt" as
the commenter suggested) many definitions.  Examples of
definitions that were revised are included in this section.
     Comment:  One commenter  (IV-D-10) contended that terms
should be used consistently, such as the term "storage vessel"
which is sometimes called a "storage tank."
     Response:  To ensure that terms are used consistently in
the final preamble and rule, the EPA has checked these
documents and uses the term "storage vessel."  [Note:  Other
changes made to the definition of storage vessel are discussed
in Section 2.2.4 of this chapter.]
     Comment:  Three commenters  (IV-D-09, IV-D-11, IV-D-18)
requested that the definition section of the proposed rule
cross reference the definition of "research and development
facility" contained in the HON.
     One commenter  (IV-D-11) requested that the definition
section of the proposed rule cross reference the definition of
"maintenance wastewater" contained in the HON.
     One commenter  (IV-D-09) requested that the definition
section of the proposed rule cross reference the definition of
"impurity" contained in the HON.
     One commenter  (IV-D-09) requested that the definition
section of the proposed rule cross reference the definition of
"maximum true vapor pressure" contained in the HON.
                             2-43

-------
     One commenter  (IV-D-18) requested that the cross
reference in the definition section of the proposed rule be
revised to borrow subpart H's definition of "in organic
hazardous air pollutant service."  The commenter (IV-D-18)
explained that the proposed rule incorrectly references the
term "in organic hazardous air pollutant" and that the word
"service" was omitted.
     Response:  The EPA intended for the following HON terms
to apply to the proposed rule and these terms have been
incorporated into the  final rule:  "research and development
facility," "maintenance wastewater," "impurity," "maximum true
vapor pressure," and  "in organic hazardous air pollutant
service."              .
     Comment:  One  commenter  (IV-D-10) recommended removing
the definitions of  "Group 1 and Group 2 wastewater streams"
from the definition section of the final rule and, instead,
cross reference the term as used in the HON.  The commenter
 (IV-D-10) stated that the referenced definition of
 "wastewater" from  §63.101 of  subpart F is too broad and is
inconsistent with  the Group 1 and Group 2 wastewater
definitions  in §63.111.  The  commenter  (IV-D-10) included
suggested language for clarifying the §63.101 definition of
wastewater  (see p.  15 of comment letter).  The commenter
 (IV-D-10) also suggested that the definition of waste
management unit in §63.111  be modified to clarify that a
wastewater management unit  does not include equipment or
component parts which are used  to recover any portion of
production/  co-product, by-product, or process solvent,
 catalyst or  inhibitor or other  material with applied value to
 the thermoplastic  product process unit.
     Response:  The EPA agrees  with the commenter that the
 definitions  of  "Group 1 wastewater streams" and  "Group 2
 wastewater streams" contained in the proposed rule are
 repetitive with the definitions contained  in the HON.  These
 terms have been removed from  the definition section, and  the
 final rule cross  references the terms as used in the HON.  The
 commenters statements concerning the  definitions of wastewater
                              2-44

-------
and waste management unit are being addressed as part of the
HON litigation.  As stated previously, changes made to the HON
as part of the litigation will be evaluated for applicability
to the thermoplastics rule.
     Comment;  One commenter  (IV-D-18) requested that the
definition of "continuous process" be reworded to state that
"continuous processes typically approach steady-state,"
instead of the proposed language that states that "continuous
processes are typically steady-state."  The commenter
(IV-D-18) explained that a true steady-state condition is not
truly the norm.
     Response:  The EPA agrees with the commenter and the
wording in the definition of  "continuous process" has been
revised for the final rule.
     Comment:  Two commenters (IV-D-08, IV-D-11) encouraged
the EPA to include the HON continuous process vent de minimis
criterion in the definitions of "continuous process vent" and
to include a similar criterion for "batch process vent."  The
commenters (IV-D-08, IV-D-11) stated that continuous process
vents with less than 0.005 weight percent HAP and batch
process vents with less than 225 kg/yr HAP emissions are
several orders of magnitude below the Group 1 criteria for
which emission control would be required.  The commenters
(IV-D-08, IV-D-11) stated that batch process vents with
emissions less than 225 kg/yr should be excluded from the
definition of batch process vents.  One commenter (IV-D-11)
stated that numerous small batch vessels are used by industry
for the preparation and feed of process additives or other
small batch process operations.   The commenter (IV-D-11)  said
that an emission level of 225 kg/yr,  which the EPA has
proposed as the threshold for the type of Group 2 batch
process vent with the minimum recordkeeping/reporting
requirements, would be an appropriate de minimis criterion.
     One commenter (IV-D-11)  contended that creating de
minimis levels for the two types of process vents would reduce
the regulatory burden associated with these insignificant
emission points.   The commenter (IV-D-11)  also explained that
                             2-45

-------
using the de minimis criterion in the definition would be
consistent with the HON and would simplify the rule.
     Response;  The EPA agrees with the commenters and has
added de minimis criteria to the definitions for continuous
and batch process vents in the final rule.  An emission point
otherwise meeting the definition of "continuous process vent"
must meet the de minimis criteria of having a flow rate
greater than or equal to 0.005 standard cubic meter per minute
and a total organic HAP concentration greater than or equal to
50 parts per million by volume.  An emission point otherwise
meeting the definition of "batch process vent" must meet the
de minimis criterion of having annual organic HAP emissions
greater than  225 kilograms per year.  Emission points not
meeting the de minimis criteria are not considered continuous
or batch process vents, as appropriate.
     Comment;  One commenter  (IV-D-10) stated that the
proposed definition of "batch cycle limitation" should be
revised.  The commenter  (IV-D-10) stated that the last two
sentences of  the proposed definition were a regulatory
requirement for Group 2 batch process vents and do not belong
in the definition text.  The commenter  (IV-D-10) suggested the
sentences be  removed from the definition and inserted in an
appropriate location in §63.506.
     Response:  The definition of batch cycle limitation was
revised to remove regulatory wording that did not belong in
the definition.  The definition now reads as follows:

     Batch cvcle limitation means an enforceable restriction
     on the number of batch cycles that can be performed in a
     year for an individual batch process vent.

The wording  that was removed  from the definition was moved to
§63.506-1  [§63.1322] of the final rule.
     Comment;   One commenter  (IV-D-18) reasoned that including
the phrase  "or  Administrator's designee, hereafter referred to
as Administrator, " after mention of the term Administrator, is
unnecessary  and should be deleted because the EPA already
                              2-46

-------
proposed to borrow subpart A's definition of "Administrator,"
which already provides for delegation.  The commenter
(IV-D-18) explained that the wording could be interpreted to
mean that a State with delegated authority could in turn have
a "designee" who would be considered the Administrator.  To
avoid this confusion, the commenter {IV-D-18) urged the EPA to
delete this wording from the final rule.
     Response;  The EPA agrees with the commenter and has
removed the phrase "or Administrator's designee, hereafter
referred to as Administrator," from the final rule.  The
definition of "Administrator" as defined under subpart A is
applicable to this rule.
     Comment:  One commenter  {IV-D-10) contended that the
definition of "continuous process vent or continuous vent
stream" in the proposed rule is confusing and appears to
refer to vents from continuous processes.  The commenter
(IV-D-10) maintained that "continuous vent stream" also has a
common usage as a stream that is venting continuously.  The
commenter  (IV-D-10) stated that continuous vent streams exist
at their facilities from batch processes.  The commenter
(IV-D-10) explained that by defining "process," "vent," "vent
stream," "batch process," "unit operation," and "continuous
process," the clearly defined terms become "continuous process
vent stream"  (not a continuous process vent) and "batch
process vent stream"  (not a batch process vent).  Such
definitions would eliminate the need for, and usage of, the
terms "continuous vent stream" and "batch vent stream."
     Another commenter  (IV-D-18) stated that the definition of
"process vent or vent stream" in the proposed rule is less
clear than in the HON.  The commenter  (IV-D-18) explained that
the proposed rule does not distinguish between solid, liquid,
and gas streams.  The commenter  (IV-D-18) stated that they are
assuming that the EPA intends to regulate only gas streams as
process vents.  Additionally, the commenter  (IV-D-18) asserted
that the proposed definition  for "process vent" or "vent
stream" is a bit confusing as to whether a process vent
                             2-47

-------
originates before or after any control or recovery devices.
The commenter  (IV-D-18) included suggested language.
     Response:  The EPA has revised the definition of
"continuous process vent or continuous vent stream".  The
revised definition combines the concepts of a physical opening
(i.e., the vent) and the emission stream.  The revised
definition of  "continuous process vent" is as follows:

     Continuous process vent means a point of emission from a
     continuous unit operation within an affected source
     having a  gaseous emission containing greater than
     0.005 weight percent total organic HAP.  Continuous
     process vents exclude relief valve discharges and leaks
     from equipment regulated under §63.509 [§63.1331].

This definition allows the use of one term throughout the
rule, which aids in clarifying the provisions.
     The term  "process vent" was also changed in the final
rule.  The definition is as follows:
     Process vent means a point of emission from a unit
     operation having a gaseous emission stream.  Typical
     process vents include condenser vents, dryer vents,
     vacuum pumps, steam ejectors, and atmospheric vents from
     reactors and other process vessels, but do not include
     pressure relief valves.
The revised definition distinguishes that the emission stream
is a gas stream.  Further, each time a gaseous emission stream
exits a unit operation, the emission stream is considered a
process vent.  For example, a process vent exists at each
point where the gaseous emission stream exits a unit
operation, such as a reactor, condenser, or incinerator.
2.8  BURDEN
     Comment;  Six commenters (IV-D-07, IV-D-08, IV-D-09,
IV-D-10, IV-D-15, IV-D-23) stated that the proposed rule is
complex, difficult to understand,  and confusing.  Commenters
(IV-D-08, IV-D-09, IV-D-10, IV-D-15)  stated that this
situation is partially the result of the inclusion of several
subcategories in one rule and partially from the numerous
cross references to other existing rules (i.e.,  the HON, the

                             2-48

-------
Polymer Manufacturing NSPS,  and the Batch Processes ACT).   The
commenters (IV-D-08, IV-D-10, IV-D-15, IV-D-23)  stated that
the rule is difficult to interpret and could result in
compliance violations strictly due to industry's inability to
understand the rule.
     Two commenters (IV-D-11, IV-D-23) urged the EPA to
streamline the proposed rule to conform with EPA and
Administration policies and directives, the most recent being
the March 16, 1.995, Regulatory Reinvention Initiative and the
resulting EPA Reinventing Environmental Protection
Implementation Strategy.
     Three commenters  (IV-D-07, IV-D-08, IV-D-11) recommended
improving the proposed rule in the following ways:   evaluating
cross references, especially in the process vent sections, and
eliminating unnecessary back and forth reviewing; using titles
or subheadings for  different subcategories to allow the reader
to make rapid applicability determinations; eliminating
distinctions without differences such as the one between surge
control vessels and storage vessels; considering referencing
some equations in an appendix, for example, the batch process
vent equations or some of the equations related to analytical
methods; make wider use of exclusions such as the one in HON
for process vents with less than 0.005 weight percent HAP;
standardizing requirements for the various regulated
subcategories, making them as similar as possible to each
other and to the standards from which they were derived; and
eliminating sections of the regulation with little or no
applicability to existing facilities in the selected category
or subcategory  (i.e., unlikely sources affected by this rule
have halogenated vent  streams) .  One commenter  (IV-D-11)
included a table showing all the areas of the proposed rule
that could be simplified but did not provide specific
suggestions.
     Response:  The EPA made numerous changes to the final
rule to further clarify the  requirements and to reduce the
complexity of the  rule.  These changes include general format
changes, rearranging text, revising wording, clarifying
                             2-49

-------
tables, providing additional headings, and revising equations.
In addition to responding to the specific commenter requests,
the EPA reviewed the rule and made additional clarifications
where necessary.  The EPA agrees with the commenters that
covering nineteen subcategories under one rule and cross
referencing other existing rules has resulted in a
comprehensive rule; however, the EPA considers the final rule
to be clear, concise, and understandable.  Although the
thermoplastics rule cross references the HON quite
extensively, it does not, as the commenter claimed, cross
reference the Polymer Manufacturing NSPS or the Batch
Processes ACT.
     As discussed in the next response, the EPA has cross
referenced portions of the HON, and changes made to the HON  as
a result of litigation will, in turn, apply to the
thermoplastics rule.  If any of the said changes are not
appropriate for the thermoplastics rule, the EPA will amend
the thermoplastics rule accordingly.
     In drafting the proposed and final rule, the EPA has
striven to adhere to EPA and Administration policies and
directives, including the March 16, 1995, Regulatory
Reinvention Initiative and  the resulting EPA Reinventing
Environmental  Protection Implementation Strategy.  For
instance, one  of the themes of the recent policies is focusing
on results, not  on how industry achieves those results.  Where
possible  in the  final rule, the EPA has specified a level of
control or a percent reduction that must be achieved and left
the decision  for how to meet that requirement  (e.g., what type
of technique  or control  device to use or what type of process
changes to make)  to  the  discretion of the owner or operator  of
the affected source,  so  that such decisions can be made based
on economic  and other  industry driven factors.
     Another theme of  the  recent initiative is to reduce the
paperwork requirements  of  standards by  25 percent.  The EPA
reviewed  the paperwork requirements of  these  standards
 (i.e.,  monitoring,  recordkeeping, and reporting) and
determined that there  were ways  to reduce the paperwork
                              2-50

-------
 requirements without diminishing the compliance  and
 enforcement aspects of the standards.   These  reductions  are
 discussed in detail in Chapter 2,  Recordkeeping  .and Reporting.
      Comment:  One commenter (IV-D-10)  supported the concept
 of building new rules from the efforts  of preceding rules;
 however,  the commenter believed that the "incorporation  by
 reference" approach used in the proposed rule is too complex
 and confusing in practice.   The commenter (IV-D-10)
 recommended that the EPA build the text of the proposed
 thermoplastics rule by transcribing desired text from other
 subparts,  refining the wording to be source category specific,
 and placing the refined wording in the  rule.
      Response:  The decision to cross reference  other rules
' instead of actually including the text  of those  rules in the
 thermoplastics rule was based on two factors.  First,  EPA
 budget constraints require that Federal Register notices be
 kept to a minimum to reduce printing costs.   Second,  because
 other rules on which these proposed rules are based,  such as
 the General Provisions and the HON,  are currently under
 litigation, the EPA prefers to cross reference these rules  so
 that any changes made to those rules as a result of  litigation
 will,  in turn, apply to the thermoplastics rule.  As discussed
 previously, changes made to the referenced rules will be
 evaluated for applicability to the thermoplastics rule.
 2.9  COMMENT EXTENSION
      Comment:  Two commenters (IV-D-01,  IV-D-22)  requested  an
 extension of the public comment period.   One  commenter
 (IV-D-22)  requested a 45-day extension  and the other commenter
 (IV-D-01)  requested a 60-day extension.   One  commenter
 (IV-D-22)  stated that the additional time was  needed to  assess
 the economic impact and potential  emission reductions  of the
 proposed rule, as well as review references to the HON and
 other promulgated regulations that require further reading,
 comprehending, and evaluating.   The other commenter  (IV-D-01)
 stated that the additional  time was needed because of  the
 complexity of the proposed rule and the fact  that it
 references the HON, with which the commenter was  unfamiliar.
                              2-51

-------
In addition, the commenter (IV-D-01) stated that  the proposed
rule was not available in the Federal Register, and they found
it burdensome and time consuming to obtain readable copies of
the proposed rule and associated documents from the EPA's TTN
bulletin board.
     Response;  The EPA was unable to extend the  public
comment period as the commenters requested.  Due  to the court
imposed signature date for the final rule, the EPA was only
able to provide 60 days for public comment on the proposed
rule.  However, the EPA continued to meet with commenters
after the official comment period was over to allow commenters
to express  their comments in person and to gather additional
clarifying  information.  The EPA considers the 60 days that
were provided and the additional meetings sufficient
opportunity to comment on the proposed rule.  The EPA met with
industry commenters in person after proposal on seven
occasions to discuss their specific comments on the proposed
rule.  These  seven meetings included visits to several plant
sites that  will be subject to the final rule.  Summaries of
these meetings are contained in docket A-92-45, Section IV-C.
In addition,  a joint meeting between the EPA, the Chemical
Manufacturer's Association, and the Society of the Plastics
Institute,  Inc. was held on September 6, 1995.  At this
meeting, the  EPA discussed preliminary EPA staff level
responses developed for the comments received on the proposed
rule.  The  EPA provided preliminary feedback and members of
industry provided additional comments.  A copy of the
discussion  topics and handouts for  this meeting is in docket
A-92-45, Item IV-E-1.  The EPA continued to accept comments
from external groups up until final preparation of the
promulgation  package, and the comments have been addressed in
the  final rule where appropriate.
                              2-52

-------
2.10  COMPLIANCE
     Comment:  One commenter (IV-D-18) supported the proposed
3-year timeframe for existing affected sources  (excluding
equipment leaks) to comply with the rule and requested that
these provisions be retained in the final rule.  The commenter
 (IV-D-18) explained that a shorter time period would be
inadequate because, for example, several provisions of the
rule require existing affected sources to install additional
controls.  Time would be required to perform the following
activities:  design controls that will be compatible with the
physical and chemical environment of the process; construct or
fabricate the controls; install controls, as well as design,
install and test monitoring devices; and perform tests or
engineering evaluations.
     Response:  The EPA thanks the commenter for their
support.  The final rule has retained the provisions allowing
3 years for existing affected sources to comply with the rule.
     Comment:  One commenter (IV-D-18) requested that the EPA
clarify the compliance date in §63.500(g)(3)(ii)(B) of the
proposed rule for changes that cause Group 2 emission points
to become Group 1 emission points.  The commenter  (IV-D-18)
stated that the proposed rule gives the compliance date of
initial start-up or by 3 years after the promulgation date.
The commenter (IV-D-18) emphasized there is nothing to
indicate which date applies for changes that may occur many
years in the future.  The commenter (IV-D-18) assumed the
intent is to require compliance by the later of the two dates.
 [Note:  §63.500(g) of the proposed rule addresses changes made
to existing plant sites.]
     Response;  The commenter is correct that the owner or
operator must be in compliance upon initial start-up or by
3 years after the promulgation date, whichever is later.   The
words "whichever is later" have been added to these
provisions, which are in §63.500(i)(3) [§63.1310(i)(3)]  of the
final rule.
                             2-53

-------

-------
                      3.0  STORAGE VESSELS

     Comment:  One commenter (IV-D-19)  requested that storage
vessels at PET continuous TPA affected sources be exempted
from the final rule.  The commenter (IV-D-19)  stated that PET
continuous TPA processes do not handle raw materials or
products that have a vapor pressure high enough to exceed the
threshold levels requiring control, and, therefore,  no
emission reduction will be achieved by the rule.  Instead, the
commenter (IV-D-19) stated that the rule would simply create
an unnecessary paperwork burden.
     Response:  The EPA decided not to exempt PET continuous
TPA affected sources from the final rule.   The EPA recognizes
that some affected sources will use materials or products with
vapor pressures that do not exceed the levels requiring
control.  These affected sources will be required to comply
with the requirements for Group 2 storage vessels.   The EPA
disagrees with the commenter that Group 2 requirements are "an
unnecessary paperwork burden."  Under the final rule,  the only
requirement for Group 2 storage vessels is to keep a record
providing the dimensions of the storage vessel and an analysis
showing the capacity of the storage vessel.  The EPA considers
these requirements to be minimal.
     In consideration of the commenter's concern, the final
rule does exempt certain chemicals from the storage vessel
provisions; these chemicals are ethylene glycol for existing
and new affected sources and styrene for existing affected
sources.  Storage vessels containing these chemicals are
completely exempt from the provisions of the rule and are not
required to meet any recordkeeping or reporting requirements.
The EPA judged that storage vessels containing these chemicals
                              3-1

-------
would not meet the applicability criteria requiring controls
under typical operating conditions.
     In addition, the EPA has added a provision to the final
rule that allows the owner or operator of an affected source
to identify those storage vessels containing no organic HAP,
other than as an impurity, which are part of the affected
source.  Once an owner or operator has made this
identification, no further action is required for those
storage vessels.  If a process change occurs, the owner or
operator must either stipulate that the process change did not
result in the establishment of one or more storage vessels
containing organic HAP, other than as an impurity, or the
owner or operator must comply with this subpart for that
storage vessel.
     Comment;  Two commenters  (IV-D-11, IV-D-25) stated that
medium storage vessel  (i.e., 20,000 to 40,000 gallon capacity)
vapor pressure applicability criteria for the HON were based
on the MACT  floor and not on cost effectiveness as stated in
the preamble to  the proposed thermoplastics rule.  Therefore,
the commenters  (IV-D-11,  IV-D-25) continued, the EPA should
have established vapor pressure  applicability criteria for
medium storage vessels for  the thermoplastics rule based on a
cost-effectiveness evaluation.   The commenters  (IV-D-11,
IV-D-25) recommended that,  based on a cost-effectiveness
evaluation using EPA methodology and investment estimates, the
correct vapor pressure applicability criteria for medium PET
storage vessels  should be 13 psia.
     Response;   While the EPA misstated the basis for the HON
storage vessel  applicability criteria for medium storage
vessels  in the  preamble  to the proposed thermoplastics rule,
this misstatement has no effect  on the proposed standards for
storage vessels. As discussed in  the technical memorandum
documenting  the MACT floor analysis  (docket  item II-B-28,
A-92-45) , the  MACT  floor for each  emission point was
determined and compared to the HON for each  subcategory.  In
                              3-2

-------
all cases, the MACT floor for storage vessels was determined
to be either equivalent to the HON or more stringent than the
HON.  In cases where the MACT floor was equivalent to the HON,
the HON was used to express the MACT floor in regulatory terms
(i.e, level of control and applicability criteria).   In cases
where the MACT floor was more stringent than the HON
(e.g., polystyrene continuous processes), the MACT floor was
defined by a unique expression of applicability and level of
control  (see docket item II-B-23, .A-92-45) .   In either case, a
cost-effectiveness evaluation was not done for medium storage
vessels because the EPA is not required to cost-justify
setting the standard at the MACT floor.  Therefore,  the final
rule has not been changed to specifically address this
comment.
     Comment:  One commenter (IV-D-25) contended that the EPA
erred in the cost-effectiveness evaluation for large storage
vessels  (i.e., greater than 40,000 gallon capacity)  under the
HON.  First, the commenter  (IV-D-25) claimed that the EPA had
failed to consider the molecular weight of lower vapor
pressure chemicals used by HON sources, and second,  the EPA
had based its floating roof estimates on obsolete technology.
The commenter (IV-D-25) stated that simple floating roofs that
were once used by petroleum refineries are unsafe for chemical
plant storage of HAP because they often fail and are unsafe to
walk on to conduct seal gap inspections.   Current modern roofs
cost about twice as much to install as the old simple floating
roofs, but they are today's minimum technology requirements
for safety and performance, claimed the commenter (IV-D-25).
     The commenter  (IV-D-25) included a cost-effectiveness
evaluation done for methanol, which is the primary chemical
stored at PET facilities using a DMT process that is
potentially subject to storage vessel controls.  The commenter
(IV-D-25) claimed this analysis supported using a vapor
pressure applicability criteria of 13 psia for both medium and
large storage vessels.  The commenter  (IV-D-25) used data from
                              3-3

-------
the HON background information document for refrigeration
system investment and operating costs.
     Response:   Through supplemental communication with the
commenter, the EPA learned that the commenter desires that
each chemical's vapor pressure and molecular weight be
considered in determining applicability criteria for
individual chemicals through a cost-effectiveness evaluation.
In summary, the commenter has raised two issues:
(1) applicability criteria should be developed for each
individual chemical covered by the rule and (2) the EPA's
costs for internal floating roofs are too low.
     Under the proposed and final rule, the EPA has developed
storage vessel standards using the approach from previous
storage vessel standards  (e.g., the HON and subpart Kb of
40 CFR part 60).  This approach considers the storage of all
liquids for a given source category or subcategory as a group,
and applicability criteria are developed to be used for all
storage vessels within the source category or subcategory.
For the EPA to determine storage vessel standards on a
chemical specific basis would' be a significant departure from
the approach used for many other existing rules.  The EPA does
receive comments, in fact, to consolidate rules for similar
equipment so they are more alike, easier to understand, and
easier to implement.  To depart from past practice based on
one public comment without more opportunity to gain input from
others on the pros and cons of such an approach did not seem
appropriate at this time.  Basing applicability criteria on
individual chemicals could create administrative and
compliance conflicts between existing and new rules.   The EPA
foresees that there could develop situations were one out of
two identical storage vessels containing identical materials,
both subject to NESHAP, is required to apply controls and one
is not.  For these reasons, the EPA did not undertake a
chemical specific analysis as requested by the commenter.
                              3-4

-------
     Based on analyzing all storage vessels within a
subcategory as a group, which considers vessels storing
ethylene glycol and methanol, the standards contained in the
proposed and final rule represent the MACT floor.   As a
result, the EPA was not required to conduct cost-effectiveness
evaluations as part of setting the standards.
     However, if the EPA had developed standards for
individual chemicals  (i.e., methanol) for the commenter's
subcategory, which is PET using a DMT process, the standards
would have come out to be the same based on the data available
to the EPA.  In the storage vessel capacity range of 20,000 to
40,000 gallons, data were provided by industry for 3 methanol
storage vessels and all are controlled; the HON would also
require control of these storage vessels.  Therefore, the MACT
floor is equivalent to the HON, and, as representative of the
MACT floor,  the HON would have been  selected as the standard.
In the storage vessel  range  of 40,000 gallons and greater,
there are  6  methanol  storage vessels with no control; the HON
would require control  of these storage vessels.  In this
hypothetical example  where the MACT  floor is no control,
selection  of the HON  as the  standard would need to be
justified  based on cost and  other  impacts.  Based on either
the commenter's cost  analysis for  internal floating roofs
provided  supplemental to the public  comment period  (see Docket
A-92-45,  Section  IV-D) or  on the EPA's  cost  analysis for the
known  methanol  storage vessels,  control of storage vessels in
the capacity range of 40,000 gallons or greater would be
judged to be in the  same range of  cost  effectiveness accepted
by the EPA as  reasonable  for the HON and  other parts of  the
thermoplastics  rules.  Therefore,  the HON would have been
selected as the standard.
      In the second point  raised,  the commenter challenged  the
validity of the costs for internal floating roofs  used for the
HON  storage vessel analysis.  The commenter stated that the
costs  used for the HON were low and that  current  modern roofs
                              3-5

-------
cost about twice as much to install.  However, data available
to the EPA show that the HON costs for internal floating roofs
were higher than costs included in the recent Alternative
Control Technology'(ACT) document entitled "Alternative
Control Technology Document:  Volatile Organic Liquid Storage
in Floating and Fixed Roof Tanks" (EPA-453/R-94-001; January
1994) prepared by the EPA subsequent to developing the HON
costing procedures.  In developing the ACT, the EPA obtained
vendor quotes for internal floating roofs and made every
effort to gather the latest costs for storage vessel controls.
For installation of a 40,000 gallon capacity storage vessel,
capital costs under the HON were approximately the same as
costs under the ACT, and for larger storage vessels, costs
based on the HON procedures were higher than costs based on
the ACT procedures.  This comparison is available in the
docket as part of a memorandum entitled "HON/CTG Cost
Comparison for Internal and External Floating Roof Tanks"  (see
docket A-92-45, section IV-B).
     During the public comment period, the commenter did
provide a cost analysis for control of one 40,000 gallon
storage vessel using a refrigerated condenser.  Based on this
one analysis, the commenter felt that the vapor pressure
applicability criteria for large storage vessels should be
13.0 psia.  Subsequent to the public comment period, the
commenter provided  a cost analysis for control of methanol
storage vessels using an internal floating roof.  Accompanying
this cost analysis, the commenter stated that refrigerated
condensers, which were the basis for the commenters public
comment letter, were more expensive than internal floating
roofs.  The EPA considers that the public comment letter
provided by this commenter presented a worst-case analysis in
terms of both the  cost of the control technology used
 (i.e., refrigerated condensers) and the size of the storage
vessels costed  (i.e., a 40,000 gallon storage vessel to
represent the 40,000 gallons and greater criteria range).  It
                              3-6

-------
needs to be reiterated that under the HON, the EPA selected
the vapor pressure applicability criteria to be representative
of the average cost effectiveness for controlling all storage
vessels within the greater than 40,000 gallon capacity size
range.  Finally, and most importantly, regardless of the
adequacy of the HON analysis, setting the standard at the MACT
floor, which is the case for this rule, does not require the
EPA to conduct a cost-effectiveness analysis.
     For all the reasons set forth in this response, the EPA
has not changed the final rule to specifically address this
comment.
     Comment:  One commenter (IV-D-24) questioned why the
combinations of vapor pressure and storage vessel capacity
differ for determining Group 1 storage vessels for existing
and new continuous polystyrene affected sources.
     Response:  Two factors contribute to the difference in
applicability criteria  (i.e., vapor pressure and storage
vessel capacity) for polystyrene Group 1 storage vessels at
existing and new affected sources.  First, these criteria are
based on the MACT floor.  Second, the Act defines the MACT
floor differently for existing and new affected sources, and
the population of best existing controls that define the MACT
floor for polystyrene existing and new affected sources are
different..  Paraphrasing the Act, the existing source MACT
floor is defined as the average of the level of control
achieved by the five best controlled existing sources for
subcategories with less than 30 sources, and the new source
MACT floor is defined as the level of control achieved by the
best controlled existing source.  Because the population of
"best controlled existing sources" that contributed to the
existing source and new source MACT floor are different for
polystyrene affected sources, the applicability criteria
reflecting the MACT floor are different.  It is conceivable
that the five best controlled existing sources for a
subcategory could be controlled to the same level.  In that
                              3-7

-------
case, the MACT floor for existing and new sources would be the
same.
     Comment:  One commenter  (IV-D-18) urged the EPA to
clarify what is meant in Tables la and 2a of the proposed rule
when the name of a HAP was used as an applicability criteria
instead of a vapor pressure.  The commenter (IV-D-18)
requested confirmation as to whether this demonstrated an
intent to address storage vessels which contain the listed HAP
regardless of the vapor pressure of that HAP at any specific
temperature.  In addition, the commenter  (IV-D-18) questioned
whether the EPA intended this approach to include only the
listed HAP by itself, or mixtures of compounds as well.
     Response;   The EPA is assuming that this commenter's
concern is based on the storage of liquids below ambient
temperature.  Data were not provided by industry that
indicated this situation, and the EPA did not anticipate it
when selecting a chemical name as an applicability criteria.
In the final rule, where a chemical name had been used as an
applicability criteria, a vapor pressure has been associated
with that chemical.
     For affected sources producing nitrile, there is only one
identified affected source, and all storage vessels located at
the facility are controlled.  For affected sources producing
ASA/AMSAN, there is also only one identified affected source,
and all storage vessels located at the facility are
controlled.  For these two subcategories, the associated vapor
pressure for both new and existing affected sources is the
lowest reported vaporr pressure for each chemical.  The Act
states that standards must be at least as stringent as the
MACT floor.  Because every storage vessel included in the MACT
floor analysis  (i.e., every storage vessel at the facility)
was controlled and controlled to the same level
(e.g., 95 percent), the lowest reported vapor pressure has
been used to ensure that the applicability criteria capture
each and every storage vessel.  If a vapor pressure
                              3-8

-------
 applicability criteria was used that did not capture each and
 every storage vessel, the standards would not be as stringent
 as  the MACT floor.
      For new affected sources producing ABS using a continuous
 mass  process, the proposed standards represent a regulatory
 alternative beyond  the MACT  floor  (see Chapter 8.0 of the
'proposal Basis  and  Purpose Document  [EPA-453/R-95-004a,
 March 1995]).   The  applicability criteria in the proposed rule
 was a mixture of capacity/vapor pressure and capacity/chemical
 name.  A vapor  pressure  was  also selected to accompany the
 chemical name as an applicability  criteria for styrene storage
 vessels.
      Comment:   Three commenters  (IV-D-08, IV-D-11, IV-D-15)
 stated that internal floating roofs were inappropriately
 selected as the reference control  technology for styrene
 storage vessels.  The commenters  (IV-D-08, IV-D-11, IV-D-15)
 stated that because styrene  has a  high tendency to polymerize
 on  the interior walls of a storage tank movement of the
 internal floating roof  is inhibited, and the roof is rendered
 ineffective for emissions control.  For these reasons,
 floating roofs  are  not  commonly used in the polystyrene
 industry.  Two  commenters  (IV-D-08, IV-D-11) stated that with
 the exception of a  closed vent system, the alternative
 reference control technologies contained in the proposed rule
 are not appropriate for styrene storage vessels.  The
 commenters  (IV-D-08, IV-D-15) suggested that condensers, which
 are the most commonly used method  of control for storage
 vessel emissions, be evaluated and selected as the reference
 control technology  for  styrene storage vessels.
      Response;  The storage  vessel provisions of the proposed
 rule  do not specify the controls that are to be applied
 (i.e., internal floating roofs are not required) when control
 is  required.  The reference  control technology for storage
 vessels is any  of the options listed in the storage vessel
 provisions.  Further, the impacts  analysis for the proposed
                              3-9

-------
rule was based on the use of a closed vent system and
refrigerated condenser which the commenters referred to as the
most -appropriate control technology.  Finally,  due to the very
low vapor pressure of styrene, typically reported as
0.07 psia, it is very unlikely that styrene storage vessels
would be required to apply controls.  The lowest vapor
pressure applicability criteria are 0.75 psia for existing
storage vessels and 0.1 psia for new storage vessels.  With a
vapor pressure of 0.07 psia, styrene storage vessels do not
meet the vapor pressure applicability criteria and would not
be required to apply controls.
                              3-10

-------
                 4.0  CONTINUOUS PROCESS VENTS

     Comment;   One commenter (IV-D-09)  requested credit for
treatment of Group 2 batch process vents at levels of control
beyond those stipulated in the proposed rule.  The commenter
(IV-D-09) stated that proposed §63.505(a)(11) requires Group 1
controls on any combined vent stream (i.e., combination of
continuous and batch process vents) containing at least one
Group 1 continuous process vent.  The commenter (IV-D-09)
contended that the control requirements for combined vent
streams should be dependent on the ratio of Group 1 and
Group 2 process vents and the need to maintain as high a level
of destruction as practicable.  The commenter (IV-D-09)
suggested that requiring Group 2 batch process vents to be
controlled by 85 percent when in a combined vent stream and
requiring the Group 1 continuous process vents to be
controlled by 98 percent would be a technically appropriate
level of control.  The commenter  (IV-D-09) stated that
calculating a mass weighted destruction efficiency requirement
based on the ratio of Group 1 and Group 2 process vents in the
mixture would be equitable and technically feasible.
     Response:  As in the proposed rule, the final rule
requires that if a batch process vent is combined with a
continuous process vent prior to being routed to a control or
recovery device, the combined vent stream will be evaluated
under the continuous process vent provisions.  More
specifically, if a batch process vent is combined with a
Group 1  continuous process vent or a Group 1 batch process
vent is  combined with a continuous process vent prior to being
routed to a control device, the combined vent stream shall
meet all the  requirements, including control requirements, of
a Group  1 continuous process vent.  If a batch process vent is
combined with a  Group 2 continuous process vent prior to being
                              4-1

-------
routed to a control device or combined with any continuous
process vent prior to being routed to a recovery device,  the
group status of the combined vent stream shall be determined
at maximum representative operating conditions (i.e.,  when the
batch process vent is emitting).   Alternatively,  the owner or
operator may choose to emissions average Group 2 batch process
vents if control was applied after November 15, 1990,  as
opposed to combining them with continuous process vents.
     The commenter requested that Group 2 batch process vents
be required to be controlled by only 85 percent when combined
with Group 1 continuous process vent streams.  The standards
developed under the MACT standards program are
technology-based standards.  It is contrary to the concept of
a technology-based standard to operate a control device such
that it provides a level of control below its capability or to
allow the use of technology that is not equivalent to the
maximum achievable control technology.  The maximum achievable
control technology identified by the EPA for continuous
process vents provides emission reductions of 98 percent.
Therefore, the final rule continues to require that combined
vent streams with one or more Group 1 continuous process vents
be controlled by 98 percent.  Therefore, the final rule has
not been changed to address this specific concern.
     Comment;  One commenter  (IV-D-10) stated that requiring
combined continuous and batch process vents to be tested
during maximum representative operating conditions may not be
reasonable because of the nature of batch processing.
Specifically, the commenter stated that it may be virtually
impossible and impractical to operate all vents at their
maximum conditions during testing.  The commenter requested
that the EPA allow for alternative procedures to document the
performance  of control equipment for combined vent streams.
The commenter  (IV-D-10)  requested that they be allowed to use
engineering  methods  to extrapolate test data, in cases where
maximum representative operating conditions could not be
achieved during testing.  The  commenter recommended the
addition of  the phrase  "...or  such other operating conditions
                              4-2

-------
as the Administrator may approve" wherever reference  is made
to testing  at  "maximum representative operating conditions."
     Response:  The EPA agrees with the commenter that it may
be impossible  or  impractical to operate all vents at  their
maximum  conditions during testing.  However, the EPA  does not
consider it necessary to modify the final rule to specifically
address  this comment.  The proposed and final rules state
"... maximum representative operating conditions shall be
during periods when batch emission episodes are venting to the
control  device, resulting in the highest concentration of
organic  HAP in the combined vent stream."  The EPA considers
that this language provides adequate flexibility to both the
Administrator  and the owner or operator to ensure testing is
done under  conditions that best represent the maximum
operating conditions.  This provision does not state  that "all
batch emission episodes" must be venting to the control
device,  nor does  it state that the absolute "highest
concentration  of  organic HAP" must be achieved during the
test.  The  final  rule has not been changed to specifically
address  this comment.
     Comment:  One commenter (IV-D-09) recommended that the
final rule  allow  an owner or operator to use continuous
process  vent gases as feed air (i.e., combustion air)  for
either a single boiler or a combination of boilers without
being required to conduct boiler performance testing  if:
(1) the  boiler heat input stipulated in the proposed  rule is
met by the  single boiler and the combination of boilers,
(2) the  thermal destruction for the single boiler and the
combination of boilers is equivalent, and (3)  the combination
of boilers  is  used as backup to the single boiler.   The
commenter (IV-D-09) explained that the proposed rule
[§63.505(c) (2) (i) (C) (2J ] provides for the use of an individual
boiler or process heater with a design heat input of
150 million Btu/hr to control emissions without performance
testing.  On the  other hand, the commenter (IV-D-09)  stated,
the proposed rule requires testing of smaller boilers  which
may be used to control the same emissions.
                             4-3

-------
     At one facility, the commenter (IV-D-09)  explained,
several smaller oil/gas fired boilers, each with individual
heat inputs less than 150 million Btu/hr, are used to backup a
larger coal fired boiler, which has a heat input greater than
150 million Btu/hr, when it is down for maintenance.  The
commenter  (IV-D-09) explained that normally a continuous
process vent goes to the larger boiler for control.  During
maintenance of the larger boiler, the continuous process vent
is split to the multiple smaller boilers.  The total heat
input of these smaller boilers is greater than
150 million Btu/hr.  Over the last 5 years the smaller boilers
have operated an average of 7 percent of the time, operating
for a minimum of 4 percent and a maximum of 11 percent of the
time for a given year.  As proposed, the rule would require
testing these smaller boilers to demonstrate compliance during
the maintenance outage period for the larger boiler.  The
impact on  emission control using the smaller boilers versus
the larger boiler  is negligible, if any, stated the commenter
 (IV-D-09) .  The commenter  (IV-D-09) requested that the final
rule allow multiple  smaller boilers with a combined total heat
input of  150 million Btu/hr to be used as backup control
devices without requiring performance testing.
     Response;  The  EPA  does not have data indicating whether
or not there will  be a negligible impact on emissions control
when using the  smaller boilers.  The EPA has documented the
basis for the large  boiler performance test exemption in
several past rules (e.g., Hazardous Air  Pollutant  Emissions
from Process Units in the Synthetic Organic Chemical
Manufacturing Industry--Background  Information for Final
Standards, Volume  2A, EPA-453/R-94-003a) .  The EPA has
determined that boilers  with heat input  capacities greater
than 44 MW would be expected to  achieve  emission reductions
greater than 98 percent,  and, based on this, the EPA has not
and does  not require a  performance  test  to establish emission
reduction efficiency.   However  for  boilers with heat input
 capacities less than 44 MW,  the  EPA does not have  the data to
 support  a similar determination.  Therefore, boilers with heat
                              4-4

-------
input capacities less than 44 MW are required to establish
emission reduction efficiency through a performance test.  The
performance test would be a one-time test.  Also, when a
continuous process vent is introduced with the primary fuel or
is used as the primary fuel in a boiler or process heater, a
performance test is not required.
     Comment:  One commenter  (IV-D-27) stated that controls
have been installed on thermoplastics manufacturing industry
process vents which provide 95 percent or better removal
efficiencies.  The commenter  (IV-D-27) stated that removal
efficiencies of 99 percent are possible using thermal
oxidation.  The commenter  (IV-D-27) stated that other
technologies, including but not limited to catalytic
oxidation, can provide comparable removal efficiencies to
thermal oxidation.
     Response;  The EPA thanks the commenter for their
information concerning the availability of control technology.
                              4-5

-------

-------
                   5.0  BATCH PROCESS VENTS

5.1  SELECTION OF STANDARDS
     Comment:   One commenter (IV-D-05) expressed concern about
short-term exposure that results from emissions from batch
processes.  The commenter  (IV-D-05) explained that while the
emissions may add little to annual organic HAP emissions, high
concentrations of organic HAP in populated areas near affected
sources could result from these emissions.   Because of the
unpredictable nature of emissions from batch processes and the
need to reduce acute health risks to the general public, the
commenter (IV-D-05) supported controlling these emissions.
     Response: The EPA appreciates the commenter's support,
and batch process vents continue to be regulated in the final
rule.
     Comment:   Commenters  (IV-D-06, IV-D-08, IV-D-11, IV-D-23)
objected to the fact that, under the proposed rule, condensers
would not be classified as recovery devices but would instead
be considered part of the batch unit operation.  Further, the
commenters  (IV-D-06, IV-D-08, IV-D-11, IV-D-23) stated that it
was their belief that, based on the proposed rule, only
absorbers, which are rarely used in monomer service, would be
considered recovery devices.
     One commenter  (IV-D-06) stated that condensers not
considered essential parts of the process should be considered
control or recovery devices.  The commenter (IV-D-06) stated
that after-condensers would be an example of a nonessential
part of the process; after-condensers are commonly used in
polymer production in order to remove and recover easily
condensible constituents which would otherwise be emitted to
the atmosphere.  Because after-condensers are not essential
parts of the process, the  commenter  (IV-D-06)  believed they
                              5-1

-------
should be considered control or recovery devices for the
purposes of calculating emission reductions.
     One commenter  (IV-D-10) believed that other types of
recovery devices in addition to the three types of condensers
included in §63.506-2(a)(1) of the proposed rule can operate
as part of the unit operation without increasing emissions to
the atmosphere.  The commenter  (IV-D-10) stated that, for
example, if a batch process vent goes to an absorber, which
recovers chemicals for fuel value, use, reuse or sale, then
the absorber is clearly part of the unit operation.  The
commenter  (IV-D-10) suggested that the phrase "In addition to
other equipment" be inserted at the front of the second
sentence of §63.506-2(a)(1) of the proposed rule as
clarification.
     Response;  Contrary to the commenters' statements, all
condensers are not  considered part of the batch unit
operation.  In §63.506-2(a)(1) of the proposed rule, the EPA
identified three types of  condensers that are considered an
essential part of the batch unit operation  -- reflux
condensers, batch stripping condensers, and batch distillation
condensers.   §63.506-2(a)(1) of the proposed rule is not
intended to mean that all  condensers are considered part of
the batch unit operation.
     Commenters did not identify any additional recovery
devices that  meet the EPA's interpretation of being an
essential part of the batch unit operation.  Therefore, the
final provisions continue  to list only  the  three types of
condensers mentioned above as essential parts of the batch
unit operation.  Other  types of condensers or condensers used
in other applications are  considered control devices.
     The significance of distinguishing a recovery device as
an essential  part  of the batch  unit operation is that
applicability is determined based on the emissions exiting the
batch unit operation  [§63 .506-2 (a) (1)  [§63 .1323 (a) (1) ] ] .  In
other words,  applicability is determined based on the
emissions  prior to any control  devices.   [Note:  As discussed
in another section of  this chapter, the EPA has defined all
                              5-2

-------
devices that reduce or oxidize organic HAP emissions as
control devices for the purposes of the batch process vent
provisions.]
     Comment;  Two commenters (IV-D-07, IV-D-08) supported the
use of cutoffs for the group status determination for batch
process vents  [§63.506-2(d) of the proposed rule].
Specifically, the commenters (IV-D-07, IV-D-08) agreed that
low annual organic HAP emissions and low flow rate cutoffs are
suitable.  The commenters  (IV-D-07, IV-D-08) explained that
batch processes are, by nature,  suited to low volume
production and the manufacture of specialty products.  As
such, the commenters  (IV-D-07, IV-D-08) stated that low flow,
low emitting process vents are likely in batch operations.
The commenters  (IV-D-07, IV-D-08) requested that the
provisions containing the discussed cutoffs be retained in the
final rule.
     Response:  The EPA has retained these provisions in the
final rule with one exception.  As discussed in response to
another comment  (see Section 5.2), the EPA has removed the
requirement to designate batch process vents as low, medium,
or high volatility.  As a result, there is a single minimum
emission level cutoff in the final batch process vent
provisions as opposed to three cutoffs.
     In addition, the EPA chose to add a minimum emission
level to the definition of batch process vent.  In the final
rule, a minimum emission level of 225 kg/yr is part of the
definition of batch process vent.  An emission point with
emissions equal to or less than 225 kg/yr is not considered a
batch process vent.
     Comment:  One commenter  (IV-D-13) agreed with the
principle of establishing batch cycle limitations for
individual batch process vents.
     Response:  The EPA  thanks the commenters for their
support.  The  final rule retains the provisions related to
batch cycle  limitations.
                              5-3

-------
5.2  APPLICABILITY
     Comment:   One commenter  (IV-D-18) stated that, according
to §63.506-1(e)(2) of the proposed rule, batch process vents
that are combined with continuous process vents are exempt
from the batch process vent provisions if, among other things,
there are no emissions to the atmosphere up to the point where
the vents are combined.  The commenter  (IV-D-18) stated that
if the phrase "no emissions" is read literally, it would make
this exemption meaningless.  If the piping or duct that
conveys the batch process vent has any flanges or valves,
there will be small, but perhaps measurable, emissions
 (i.e., equipment leaks) from these points.  The commenter
 (IV-D-18) requested that the EPA clarify that these types of
emissions do not disqualify a batch process vent from this
exemption.
     Response:  The EPA did not intend that negligible
emissions from well maintained piping or ductwork disqualify a
batch process vent from this  exemption.  The final-rule has
been clarified to represent the EPA's original intention.
     Comment;  One commenter  (IV-D-18) stated that
§63.506-2(h)(1)(i) of  the proposed rule allows the owner or
operator to use process knowledge that no halogens or hydrogen
halides are present in the process to determine whether a
batch emission episode is halogenated.  If taken literally,
the commenter  (IV-D-18) stated, this would not allow a
negative determination to be  made if a single molecule of a
halogen or hydrogen halide were present in the process; such
an interpretation would make  the exemption meaningless.  The
commenter  (IV-D-18) requested that the EPA clarify the final
rule to avoid  this problem.
     As an example, the commenter  (IV-D-18) described
situations for production  facilities  on the Gulf Coast and in
Michigan.  On  the Gulf Coast, the air contains salt from
seawater evaporation.  In  Michigan, the air contains salt from
road de-icing  in  the  winter.  In either case, the  salt
contains chlorine, which  is  a halogen,  and the chlorine
sometimes makes its way  into process  equipment.  For example,
                              5-4

-------
air enters the process during maintenance operations or during
the "breathing" of storage vessels.  This could result in
trace quantities of chlorine being present in the process as
an impurity.
     The commenter (IV-D-18) also requested that
§63.506-2(h)(1)(i) of the proposed rule be clarified so that
process knowledge that halogens or hydrogen halides cannot
become part of a batch process vent can be used to demonstrate
that a batch emission episode is not halogenated.  The literal
wording of  §63.506-2(h)(1)(i) of the proposed rule states
there cannot be halogens or hydrogen halides anywhere in the
entire process.  The commenter  (IV-D-18) felt that the EPA
should not  object to the presence of halogens or hydrogen
halides in  portions of  the process that are unrelated to the
batch process  vent.  For example, the commenter  (IV-D-18)
stated, examine the case where  a process vent comes from raw
material  purification  equipment at the  "front end" of the
process and hydrogen chloride is used to adjust the pH of an
intermediate stream near the  "back end" of the process.  The
commenter (IV-D-18) maintained  that the owner or operator
should be allowed to use process knowledge that the hydrogen
chloride  will not be able  to  enter the batch process vent at
the  "front  end" of the process  to  declare that batch emission
episodes  associated  with  this batch process vent are not
halogenated.  The commenter (IV-D-18) included suggested
language.
      Response: The  commenter requested that the EPA clarify
the exemption in  §63.506-2 (h) (1) (i) of  the proposed rule to
account  for halogens or hydrogen halides present in the
process  in negligible amounts.   In the  final rule,
 §63. 506-2 (h) (!•) (i)  [§63.1323 (h) (1) (i)]  has been  reworded to
 read as  follows:

      Process knowledge that no  halogens or hydrogen
      halides are present in the process may be used to
      demonstrate that a batch emission episode  is
      nonhalogenated.  Halogens  or hydrogen halides that
      are unintentionally introduced into the process
                               5-5

-------
     shall not be considered in making a finding that a
     batch emission episode is nonhalogenated.
     The same commenter requested that an owner or operator be
allowed to use process knowledge that halogens or hydrogen
halides cannot enter a specific batch process vent, even
though they are present in the process, to determine that a
batch emission episode is not halogenated.  The EPA believes
this type of determination falls under §63.506-2(h)(1)(ii)
[§63.1323(h)(1)(ii)], engineering assessment, and the final
rule has not been changed to address this specific comment.
     Comment;  One commenter  (IV-D-18) urged the EPA to allow
dehalogenated batch process vents, as described by
§63.506-1(c)(2) of the proposed rule, to comply with any of
the three compliance options provided under §63.506-1(a) of
the proposed rule, not just two of the options.  The commenter
{IV-D-18) inquired why the proposed standards stated that if
the owner or operator scrubs out the halogens from a batch
process vent, they must comply with §63.506-1(a)(1) or  (a)(2)
of the proposed rule but cannot use the third option
[§63.506-1(a)(3)  of the proposed rule] which requires.
achieving an overall emission reduction of 84 percent.  The
commenter (IV-D-18) requested that these provisions be revised
to include the 84 percent overall control option for all
dehalogenated batch process vents.
     Response;  The compliance requirement provided by
§63.506-l(a)(3)  [§63.1322(a)(3)] is only applicable to new
affected sources producing SAN using a batch process; it is
not a compliance option for any other type of affected source.
This compliance option represents the MACT floor for this
subcategory and is considered to be a more stringent
requirement.  These provisions have not been changed in
response to this comment.
     Comment;  One commenter  (IV-D-18) suggested that the EPA
pick one of the three annual emission threshold levels
presented in  §63.506-2(d) of the proposed rule.  The commenter
(IV-D-18) was confused by the three different threshold levels

                              5-6

-------
for low, moderate, and high volatility batch process vents and
felt the values of 11,800 kg/yr for low volatility streams;
7,300 kg/yr for moderate volatility streams; and 10,500 kg/yr
for high volatility streams did not follow any logical linear
pattern.  The commenter  (IV-D-18) suggested that because the
values are within an order of magnitude of each other, the EPA
pick a single value for all three volatility classes; they
suggested 11,800 kg/yr or an average value of 9,900 kg/yr.
The commenter  (IV-D-18) included suggested language.
     Response:  The Batch Processes ACT developed an annual
threshold emission level for each of the three volatility
classes.  The different values reflect the consideration of
two different control devices (i.e., condensers and thermal
incinerators) for batch process vents of differing volatility
in the development of control options.  Given this, the EPA
initially judged that selection of a single annual threshold
emission level would not be appropriate and included all three
levels in the proposed rule.  However, upon further review,
the EPA has found no adverse impact will result from the use
of a single annual threshold emission level.  Further,
removing the need to designate a volatility class for a batch
process vent has significantly simplified the final rule.
Besides removing the requirement to determine the volatility
class, the final rule now contains only one equation for
determining the cutoff flow rate [§63.506-2(f) [§63.1323(£)3],
which is the last step in the group determination process.
For all these reasons, the final rule contains a single annual
threshold emission level.
5.3  GROUP DETERMINATION
     Comment:  One commenter (IV-D-09) suggested changing the
group determination provisions in §63.506-2 to only utilize
emissions data from a TPPU's primary product.  In addition,
the commenter  (IV-D-09) requested that §63.506-2 be revised to
require that batch process vent group determinations be
performed on an annual basis - instead of for every process
change.
                              5-7

-------
     The commenter .(IV-D-09) stated that the proposed batch
process vent group determination provisions are considerably
more complex than the continuous process vent group
determination provisions.  Not only do the batch process vent
group determination provisions require an owner or operator to
obtain emissions data for every product that is manufactured,
the commenter  (IV-D-09) explained, but even the most minor
process changes  (i.e., lengthening cycle times, altering
process temperatures and pressures, etc.) would trigger the
need for a new group determination to be performed.  Given the
inherent process variability associated with batch operations,
the commenter  (IV-D-09) .continued, it would be difficult to
perform a group determination for a batch unit that
manufactures only one product, much less multiple products.
Furthermore, the commenter  (IV-D-09) stated, because batch
units often need to implement sudden process changes in
response to customer demands, the proposed provisions could
potentially require repeating the group determination exercise
several times  in a single year.  The commenter  (IV-D-09)
explained that  such a  situation would not only serve to
complicate a batch unit's compliance status, but could also
adversely impact its ability to remain competitive in the
marketplace.
     Response:   The commenter has raised four issues:
 (1) performance of the group determination on the primary
product,  (2) performance of the group determination on an
annual  basis,  (3)  objection that  the group determination
provisions  require a new group determination to be made
whenever minor process changes occur, and  (4) objection to the
requirement  to perform the  group  determination when a sudden
process change is  required.
      The EPA has considered the  request  to perform the group
determination on the  primary product and agrees that this
would provide acceptable results  from an environmental
perspective while simplifying  the compliance requirements  for
and improving the enforceability of the  batch process vent
provisions.   The final rule contains provisions allowing the
                              5-8

-------
owner or operator of an affected source to perform the group
determination for batch process vents based on the annualized
production of a single product (i.e., batch unit operation).
However, the EPA does not consider it to be equitable from an
environmental perspective to allow anything other than the
worst-case HAP emitting product to be considered when basing
applicability on a single product.  Therefore, the final rule
specifies that the worst-case HAP emitting product be used
when an owner or operator chooses to determine applicability
based on the annualized production of a single product.  The
final rule defines the worst-case HAP emitting product and
describes how emissions are to be annualized to represent
full-time production; full-time production does not
necessarily mean operating at the maximum production rate.  As
described above, the EPA has modified the final rule to
address the commenter's concern.
     The commenter requested that the group determination be
required on an annual basis instead  of for every process
change.  The EPA believes the proposed rule was clear on this
point.  The proposed provisions in §63.506-2 require that
emissions, average flow rate, and volatility class be
determined on an annual basis.   [Note:  As discussed earlier,
the  final rule does not require the  determination of
volatility class.]  Procedures contained within §63.506-2
 [§63.1323] of the  final rule describe how to account for the
production of different products'throughout the year.  In this
way,  the  group determination is done on an annual basis and
can  account  for expected changes  in  the types and quantities
of products  being  produced.  The  final rule does not reflect
any  changes  related to  this comment.
      The  commenter objected to the requirement that a new
group determination  is  required whenever minor process changes
occur (e.g.,  lengthening cycle times, altering process
temperatures and pressures, etc.).   The proposed and final
rule addresses  this  concern as it relates to  Group 2 batch
process vents.   The  provisions in §63.506-2(1)  [§63.1323(1)3
address process changes to Group  2 batch process vents and  the
                              5-9

-------
requirement to perform a new group determination.  These
provisions provide examples of what things are process changes
and what things are not process changes.  Also, these
provisions state that "changes that are within the range on
which the original group determination was based" are not
considered process changes.  This methodology allows an owner
or operator to perform the initial group determination with
the minor process changes cited by the commenter in mind.  The
EPA believes that the proposed rule was clear that truly minor
process changes  (variations in operating conditions) do not
require that a new group determination be performed.
Addressing this concern as it relates to Group 1 batch process
vents, the proposed and final rule do not require a
redetermination of group status for Group 1 batch process
vents.  Therefore, if minor process changes, as cited by the
commenter, were to occur, the owner or operator would not be
required to perform another group determination.  The final
rule does not reflect any changes related to this comment.
     Finally, the commenter objected to the requirement to
perform a new group determination when a sudden process change
is required.  The EPA has interpreted "sudden process change"
to potentially mean:   (1) that-a new product is being made,
(2) that the same product is being made in a fundamentally
different way  (e.g., with different raw materials), or
(3) that the same mix of products is being made but in a
different proportion.  In the first two cases, the EPA desires
and intends that a new group determination be made.  In the
third case, the owner or operator has the flexibility to
consider this situation when performing the initial group
determination.  If this situation was not considered, then a
new group determination would be required.  The EPA feels that
these types of process changes warrant a new group
determination to ensure that the standards are being met.  The
final rule does not reflect any changes related to this
comment.
     Comment;  One commenter  (IV-D-07) felt the proposed rule
excluded certain batch process vents from making a group
                             5-10

-------
determination.  The commenter  (IV-D-07) objected to the
exclusion and recommended that all batch process vents be
allowed to make a group determination.
     Response;  §63.506-2(a) of the proposed rule states,
"Except as provided in paragraph  (a)(2) of this section and in
§63.506-1(a)(3)(B) , the owner or operator ... shall determine
the group status of each batch process vent  ... "  The two
paragraphs cited in §63.506-2(a)  [§63.1323(a)] are situations
where compliance has already been achieved and requiring a
group determination is not necessary.  Because regulatory
language is typically oriented towards describing the required
actions of owners or operators, as opposed to describing
optional activities, the proposed provisions did not
specifically state that an owner or operator may perform a
group determination on any batch process vent if so desired.
However, the proposed provisions, by not prohibiting such an
activity, allow an owner or operator to perform a group
determination for any batch process vent if he or she chooses
to do so.  The EPA does not believe that any revisions are
required to address this comment and none have been made.
     Comment:  Four commenters  (IV-D-07, IV-D-08,  IV-D-11,
IV-D-23) objected to the requirement to determine group status
for batch process vents prior to control or recovery devices. '
Three of the four commenters  (IV-D-08, IV-D-11,  IV-D-23)
reported that this would be a source of confusion for the
regulated community and regulatory authorities.   They
requested that the HON approach, which requires group
determinations for continuous process vents to be made after
recovery devices, be used.  Two commenters (IV-D-08,  IV-D-23)
stated that making group determinations after control or
recovery devices would provide an incentive for adopting
pollution prevention or recovery measures,  including those
that might be based on innovative recovery methods.   One
commenter (IV-D-07) contended that the group determination
should be based on emissions to the atmosphere,  not on
uncontrolled emissions.  The commenter (IV-D-07)  stated that
the requirements in the proposed rule represent a significant
                             5-11

-------
difference and serious inconsistency with the requirements for
continuous process vents.
     Response;  The EPA is aware that the group determination
procedures for batch and continuous process vents are
different.  This difference is a direct result of the analyses
that developed the control requirements for batch process
vents versus continuous process vents.  Given the differences
between batch and continuous process vents and the fact that
control requirements were based on different analyses, the EPA
does not find differences in the group determination
procedures objectionable.  Further, the analysis that
developed the control requirements for batch process vents
 (i.e., the Batch Processes ACT) was based on applying controls
prior to recovery or control devices.  The EPA does not feel
required to revise 'this  analysis based on the comments brought
forth, nor did the EPA feel required to revise the analysis
based on their own observations prior to proposal.  The final
rule retains the requirement to make group determinations for
batch process vents prior to control devices.
     The significance of distinguishing a recovery device as
an essential part of the batch unit operation is that
applicability is determined based on the emissions exiting the
batch unit operation  [§63.506-2(a) (1)  [§63.1323(a) (1)]] .  In
other words, applicability is determined based on the
emissions prior to any control devices.   [Note:  As discussed
in another section of this chapter, the EPA has defined all
devices that reduce or oxidize organic HAP emissions as
control devices for the  purposes of the batch process vent
provisions. ]
5.4  DEMONSTRATION OF COMPLIANCE
     Comment;  One commenter  (IV-D-13) asserted that the
equations and test methods for calculating an annual average
 flow rate in  §63.506-2 (e) of the proposed rule are not
warranted.  The commenter  (IV-D-13) explained that the
volumetric  flow rate  testing methods prescribed in
 §63.506-2 (e) (1) of the proposed rule and the minimum 15-minute
measurement  frequency defined  in  §63.506-2 (e) (2) of the
                             5-12

-------
proposed rule are overly burdensome and will not always
provide representative measurements.  The commenter (IV-D-13)
maintained that average flow rates for a batch emission
episode are better defined by calculations of displacement
volumes with respect to the durations of the displacement
episodes or by other more simplified methods.
     Response:  The EPA agrees with the commenter that there
are more simple and potentially more accurate techniques for
estimating flow rate for batch process vents.  The final rule
contains provisions that allow engineering assessment, as well
as testing, to be used for estimating flow rate.  The final
provisions in §63.506-2(e)  t§63.1323(e)3 allow the owner or
operator to select the most appropriate technique for
determining flow rate from the two options of testing and
engineering assessment.
     Comment:  One commenter -(IV-D-07) recommended removing
the emissions estimation equations in §63.506-2(b) of the
proposed rule from the rule.  The commenter  (IV-D-07)
recommended that measurements or engineering estimates be
allowed instead of using the equations.  The commenter
 (IV-D-07) felt the emissions estimation equations would not
allow the flexibility necessary to account for differences in
process technologies and operating methods.  Reasons the
commenter  (IV-D-07) provided for removing the emissions
estimation equations from the rule were:   (1) although based
upon the ideal gas law, the format of the equations is
needlessly complex;  (2) the EPA offers no descriptions.of the
individual process steps;  (3) the equations do not consider
different or alternate operations  (i.e. reactor blanketing
with nitrogen); and  (4) it  is unclear where some process steps
end and where maintenance or shutdown practices begin.
     Six commenters  (IV-D-07, IV-D-08, IV-D-09, IV-D-13,
IV-D-18, IV-D-23) supported the use  of the options provided in
§63.506-2(b)(5) and  §63.506-2(b)(6)  of the proposed rule.  The
provisions contained in these paragraphs allow owners or
operators to use direct measurement  or engineering assessments
to estimate emissions  in  cases where the emissions estimation
                             5-13

-------
equations are inappropriate for a particular type of operation
or where a more refined estimate of emissions is necessary.
Four commenters  (IV-D-07, IV-D-09, IV-D-10, IV-D-18) objected
to the requirement to demonstrate that the emissions
estimation equations and direct measurement methods described
by §63.506-2(b)(1) through  (b)(5) of the proposed rule are not
appropriate before engineering assessments [§63.506-2(b)(6) of
the proposed rule] can be used.  One commenter (IV-D-18)
stated that methods may be equally appropriate but engineering
assessment data may already be available, whereas the other
methods listed often require costly performance testing.  Two
commenters  (IV-D-09, IV-D-18) requested that the EPA allow
owners or operators to use the most practical and
cost-effective method to estimate emissions.   One commenter
(IV-D-18) maintained that this request is not without
precedent because the HON allows engineering assessments to be
used provided their basis is documented.
     Response:  One commenter requested that the emissions
estimation equations in §63.506-2(b) of the proposed rule be
dropped from the final rule.  The commenter did not provide
specific comment that would lead the EPA to take this action,
and the emissions estimation equations have been retained in
the final rule.
     Several commenters requested that the final rule allow
the owner or operator to select among the three techniques for
estimating emissions provided in the proposed rule
(i.e., remove the hierarchial organization for using the
techniques included in the proposed rule).   The EPA believes
the data required to use the emissions estimation equations
should be obtainable with reasonable effort.   Further,
specific comments regarding the accuracy or inapproprlateness
of the emissions estimation equations were not made.  Given
this, the EPA favors a more consistent estimation technique
which is provided by the use of the emissions estimation
equations.  The final rule continues to require use of the
emissions estimation equations.
                             5-14

-------
     However, independent of the comments provided, the EPA
has concluded that direct measurement of emissions may prove
to be difficult and may or may not provide an increased
assurance of accuracy over the use of engineering assessment.
Therefore, if an owner or operator can demonstrate that the
emissions estimation equations in §63.506-2(b)(1)  through  (4)
[§63.1323(b)(1) through (4)] of the final rule are not
appropriate, the final rule allows the selection of either
direct measurement or engineering assessment. Further,
criteria for demonstrating that the emissions estimation
equations are not appropriate for a specific batch emissions
episode have been added to the final rule.  These criteria
require either:   (1) the availability of test data that
demonstrates a greater than 20 percent discrepancy between the
test value and the estimated value, or (2) that the owner or
operator demonstrate to. the Administrator that the emissions
estimation equations are not appropriate for a given batch
emissions episode.
     Comment;  One commenter  (IV-D-09)  stated that more
flexibility be allowed in designing a testing scheme for batch
process vent emissions.  The commenter (IV-D-09) stated that
the definition of "batch emission episode" in the proposed
rule presents an example where the charging of a vessel and
the heating of the same vessel are considered two batch
emission episodes.  The commenter  (IV-D-09) stated that this
definition of batch emission episode would require a separate
sample to be collected for the charging and the heating of the
vessel.  This requirement would require a large number of
samples to be taken to characterize processes that have
multiple, short duration process steps.  The flexibility to
test the emissions from several steps as a single emission
episode would reduce testing cost without jeopardizing the
quality of the emissions data.  The commenter (IV-D-09)
suggested that three batch cycles  (or more) could be tested to
obtain a representative average emission rate for the batch
cycle.  The  commenter  (IV-D-09) included suggested language
                             5-15

-------
but did not provide data to support the claim that emissions
data quality would not be compromised.
     Response:   After consideration of this comment,  the EPA
has chosen to leave the provisions related to emissions
testing of batch process vents unchanged as they relate to
this specific comment.  The EPA believes that testing even a
single batch emission episode would be challenging in the
conduct of the test and the review of the test results.  The
EPA believes that testing a batch cycle comprised of multiple
batch emission episodes would not be feasible.  Therefore, the
final rule has not been modified with regard to this specific
concern.  However, the emissions testing provisions in
§63.506-2(b) have been modified to provide flexibility and to
reduce the burden of testing, while continuing to ensure that
results are satisfactory for applicability determinations.
     In the final rule, an owner or operator may choose to
test just a portion of the batch emission episode selected to
be controlled.   [Note:  The final rule also allows an owner or
operator to control just a portion of a batch emission episode
as part of achieving  the requirement  of 90 percent emission
reduction for the batch cycle.]  The  final rule allows an
owner or operator to  choose to test a portion of the batch
emission episode only when the owner  or operator can
demonstrate that emissions during  the period to be tested
represent emissions  for the entire batch emission episode or
are greater than the  average  emission rate for the batch
emission episode.   Similar provisions appear  in the Polymers
and Resins Group  II  MACT  final rule,  and the EPA felt that
similar flexibility was warranted in  this  rule.  Therefore,
the final rule  allows an  owner or operator to test the entire
batch  emission  episode or to  test just  a portion of the batch
emission  episode  if certain criteria  are met.
     Comment;   One commenter (IV-D-09)  recommended that
guidance  be provided on the applicability  and the assumptions
involved  with using the emissions estimation equations  in
§63.506-2 (b) (1) ' through (4)  of  the proposed  rule.  The
commenter (IV-D-09)  wished to prevent an owner  or operator
                              5-16

-------
from errantly applying the equations to a process activity
that is beyond the scope of their applicability and, as a
result, generate inaccurate emissions data.  The commenter
(IV-D-09) suggested that because the equations were obtained
from the EPA's Batch Processes ACT document, the EPA should
include the appropriate sections of this document that address
the derivation of the emissions estimation equations as part
of the appendix to the final rule.
     Response;  The EPA appreciates the commenter's concern
and directs the commenter to the EPA's Technology Transfer
Network electronic bulletin board to obtain a copy of the
Batch Processes ACT.  Also, a copy can be obtained from the
Library Services Office of the EPA in Research Triangle Park,
North Carolina.
     Comment;   One commenter  (IV-D-09) requested that the EPA
better define how to calculate the weighted average molecular
weight of organic HAP in vapor  [Equation 7 in
§63.506-2(b)(4)(iii) of the proposed rule] when it is utilized
in the vessel heating equation  [Equation 4 in §63.506-2(b)(4)
of the proposed rule] .  The commenter (IV-D-09)  explained that
the vessel heating equation requires input parameters from
both the heating cycle's initial and final conditions.  The
commenter  (IV-D-09) stated that the equation that has been
provided to calculate the weighted average molecular weight of
organic HAP in vapor (Equation 7) can only calculate values at
a single condition  (i.e., temperature).   The commenter
(IV-D-09) stated that the EPA needs to define at which point
in the heating cycle this variable should be calculated for
use in Equation 4.  The commenter (IV-D-09) stated that the
molecular weight of a vapor stream exiting a process vessel
can vary considerably from the start to the end of a heating
cycle.
     Response:  The final rule has been changed so that
Equation 4 calculates an average of the weighted average
molecular weight of organic HAP in vapor at the initial and
final temperatures for the heating cycle.  In other words, an
owner or operator is asked to use Equation 7 twice, once to
                             5-17

-------
determine the weighted average molecular weight at the initial
temperature and once at the final temperature; these two .
values are then averaged in Equation 4.
5.5  RECORDKEEPING AND REPORTING
     Comment:  One commenter  (IV-D-18) urged the EPA to
clarify what records are required to be maintained by
§63.506-5 (a) of the proposed rule.  The commenter  (IV-D-18)
stated that the words "shall include" be deleted because they
imply that other, unspecified records may also be required.
     Response;  The EPA agrees with the commenters and the
recordkeeping provisions of the final rule have been clarified
to specify those records that are required to be maintained.
     Comment:  One commenter  (IV-D-09) requested clarification
of the recordkeeping requirements of  §63.506-5 (f) (2) (iv) of
the proposed rule.  This requirement, stated the commenter
 (IV-D-09), implies that operation of  batch cycles where
monitored parameter levels are outside of established values
is the norm, not the exception.  The  commenter  (IV-D-09)
explained that  §63.506-5(f)(2)(iv) of the proposed rule
requires  the owner or operator to keep "records of the batch
cycle daily  average value of  each continuously monitored
parameter" except when the operating  parameter values are at
proper levels.  The commenter (IV-D-09) stated that the
emphasis  of  this paragraph implies that these extra records
are more  routinely required than should be the case.  The
commenter (IV-D-09) included  suggested language.
     Response:  The EPA disagrees with the commenter's
interpretation of  §63.506-5 (f) (2) (iv) of the proposed rule.
 [Note:  These provisions are  in §63.506-5(e)  [§63.1326(e)] of
the final standards.]  The intent of  these provisions are to
provide flexibility to the owner or operator for those times
when all  recorded values for  a monitored parameter during an
operating day are  above the minimum or below the maximum
established level.   [Note:  The final standards no longer use
the two  step process  of determining batch cycle averages and
then determining a daily average, as  was required at proposal;
instead,  a batch cycle daily  average  is calculated from all
                              5-18

-------
parameter values measured during the operating day.]   The
intent of this paragraph is that an owner or operator may
either "record the fact that all values were above the minimum
or below the maximum" or "calculate and record the batch cycle
daily average."   For some owners or operators, the
calculation and recording of a batch cycle daily average will
be automated, and the EPA did not want to require something
additional of these owners or operators.  And, for other
owners or operators, the flexibility to not calculate a batch
cycle daily average would lessen the recordkeeping burden.  No
changes have been made to these provisions in the final
standards.
     Comment:  One commenter  (IV-D-09) recommended that
submission of the report required by §63.506-6(b) or  (c) of
the proposed rule should be within 180 operating days or as
part of the next Periodic Report, whichever is later.  The
commenter  (IV-D-09) explained that both §63.506-6(b)  and
(c) of the proposed rule require that in the event of a
process change for a Group 2 batch process vent that meets the
requirements of §63.506-2(i) of the proposed rule, the owner
or operator must "submit a report within 180 operating days
after the process change."  The commenter  (IV-D-09) stated
that only in §63.506-6 (c) of the proposed rule, is the option
given to submit the report as part of the next Periodic
Report.  The commenter  (IV-D-09) requested that this option be
made available in §63.506-6(b) of the proposed rule.   The
commenter  (IV-D-09) also requested that the report be filed by
the later of the two dates  (i.e., 180 days or the next
Periodic Report).  The commenter  (IV-D-09) included suggested
wording.
     Response':  The EPA agrees with the commenter, and the
final standards contain the suggested changes.
5.6  OTHER CHANGES RESULTING FROM EPA REVIEW
     In addition to changes made to the proposed rule as a
result of public comment, changes were made as a result of the
EPA independently reviewing the rule between proposal and
promulgation.  Because the batch process vent provisions
                              5-19

-------
included in the rule are among the EPA's first attempts to
regulate batch process vents, the EPA felt an ongoing,
independent review of these provisions after proposal was
warranted.  Changes resulting from this review are listed
below:
     (1)  allow applicability determinations and compliance
     demonstrations  (i.e., performance tests) to be based on
     total organic compounds  (TOG) or organic HAP and allow
     the use of Method 25A to compliment the use of TOC as a
     potential basis for applicability determinations and
     compliance demonstrations;
      (2)  allow the establishment of parameter monitoring
     levels to be based on performance tests or a combination
     of performance tests and engineering assessment
      (discussed in more depth in Section H, Monitoring) --to
     accommodate this change, modifications to the batch
     process vent testing provisions were required;
      (3)  add provisions specifying how the batch cycle
     limitation is to be determined; and
      (4)  change the reporting requirement for batch  cycle
     limitation records from quarterly to annually.
These  changes are discussed  in the paragraphs below.
     In the final rule, the  EPA has allowed the use of TOC as
the basis for applicability  and compliance demonstrations
 (i.e.,  performance tests) as an alternative to organic HAP.
The EPA has allowed  this  testing  to provide flexibility to the
regulated community  and to reduce the overall burden  of the
rule.   The EPA considered the impacts of allowing TOC to serve
as a  surrogate for organic HAP in applicability and compliance
demonstrations and did not find any negative impacts.
Further,  allowing the use of Method 25A as a complement to the
use of TOC as a surrogate to organic HAP reduces the  burden of
implementing the final  rule  with  little to no adverse impact
on the measurement of pollutants  in the regulated batch
process vents.  To the  best  of the EPA's knowledge, the batch
process vents  regulated by  this rule are predominantly organic
HAP.   Also, with one exception (i.e., ethylene glycol) , the
                              5-20

-------
regulated organic HAP, which are listed in the definition of
organic HAP found in the final rule, have response factors to
Method 25A adequate to ensure satisfactory measurement of TOC
in the batch process vents.  For certain emission points where
the EPA considered the presence of ethylene glycol and its
corresponding poor response factor to call into question the
results that would be obtained using Method 25A, the use of
Method 25A is not allowed.  For all other emission points, the
EPA has allowed the use of TOC for applicability and
compliance demonstrations as an alternative to organic HAP in
the final rule.
     In the final rule, the EPA allows the establishment of
parameter monitoring levels to be based on either performance
tests, as in the proposed rule, or a combination of
performance tests, engineering assessment, and manufacturer's
recommendations.  This change affects all emission points
which are required to establish parameter monitoring levels,
including batch process vents.  The rationale for this change
is discussed in detail in the preamble to the promulgated
rule.  For batch process vents, this change in the procedures
for establishing parameter monitoring levels necessitated
changes to the performance test provisions.   When an owner or
operator chooses to establish parameter monitoring levels
based exclusively on performance tests,  the final rule directs
that the performance test must include the entire batch
emission episode selected to be controlled.   As discussed
earlier, an owner or operator may choose to control just a
portion of a batch emission episode; in such a scenario, the
performance test must include the entire portion of the batch
emission episode selected to be controlled.   Alternatively,
when an owner or operator chooses to establish parameter
monitoring levels based on a combination of performance tests,
engineering assessment, and manufacturer's recommendations,
the final rule allows an owner or operator to test either the
entire batch emission episode, or portion thereof,  selected to
be controlled or to test only a period of the batch emission
episode that is less than the entire batch emission episode,
                             5-21

-------
or portion thereof, selected to be controlled.   [Note:  The
flexibility to test a specific period of the batch emission
episode that is less than the entire batch emission episode,
or portion thereof, selected to be controlled is discussed
earlier in this section.]
     The final rule includes provisions specifying how the
batch cycle limitation, required for Group 2 batch process
vents, is to be established.  The EPA felt that the proposed
rule was ambiguous concerning the establishment of the batch
cycle limitation and added these provisions to make the rule
more complete.  The added provisions provide additional
description of the purpose of the batch cycle limitation and
describe what documentation is required as part of
establishing the batch cycle limitation.
     In the final rule, the EPA changed the requirement for
reporting the number and type of batch cycles accomplished for
a Group 2 batch process vent from a, quarterly to an annual
reporting requirement.  The EPA felt that quarterly reporting
was unwarranted given that•the compliance requirement
 (i.e., the batch cycle limitation) was on an annual basis.
                              5-22

-------
    6.0  POLY(ETHYLENE TEREPHTHALATE)  (PET) VACUUM SYSTEMS

      In the preamble to  the proposed rule,  the EPA solicited
 comments and data on the technical  feasibility and costs  for
 emission control techniques applicable to  the vacuum system
 and associated process contact  cooling towers used in  PET
 production.   Control of  these emissions went beyond the MACT
 floor of no control.  In response to this  request  for
 information,  the EPA received data  and information used to
 develop the requirements for the final rule.  Because  of  the
 large volume of comments received and the  detailed and
 interrelated nature  of the  comments,  this  chapter  is organized
 somewhat differently than others in this document.  The
 primary difference is the amount of introductory or background
 text  preceding presentation of  specific comments.  Also,
 because so  many comments were received that made essentially
 the same point,  these comments  are  not presented in the same
 level of detail as for other chapters  in this document.
 Finally, the comment  summaries  do not  identify which
 commenters  made which exact  points  except for instances where
 only  one or two commenters made the particular comment.  The
 reader is referred to the actual comment letters in the docket
 for additional detail and to understand the specific comments
made  by each commenter.  The following commenters provided
public comments  related  to vacuum systems at PET facilities:
 IV-D-04, IV-D-05,  IV-D-06, IV-D-11,  IV-D-16, IV-D-17,  IV-D-19,
 IV-D-20, IV-D-22,  IV-D-24, IV-D-25,  IV-D-26.
      Issues  related to existing and new affected sources are
presented separately.  Within the discussion of existing
affected sources,  the use of ethylene glycol jets as a
retrofit technology and  the use of alternative technologies
are discussed separately.  Following this discussion,  comments
on the provision designating all vacuum system wastewater
                              6-1

-------
streams (i.e., contact condenser effluent) as Group 1 are
presented.  Industry recommended alternatives to the proposed
rule and miscellaneous comments complete the presentation of
comments on existing affected sources.  The discussion of new
affected sources is very short in comparison because
commenters did not object to the use of ethylene glycol jets
for new affected sources.
     To provide some background, a brief discussion of
emissions from PET vacuum systems completes this introduction.
Based on information received from the commenters, the
emissions from PET steam jet vacuum systems and the associated
process contact cooling towers occur as a result of
•(1) volatilization from the contact condenser effluent and
 (2) stripping and drift within the process contact cooling
tower.  The most typical vacuum  system used within the PET
industry uses steam jets to create the vacuum, condenses steam
from the  steam jets using  contact condensers, and recirculates
the contact condenser  effluent to the process contact cooling
tower.   [Note:  Some affected sources use once-through cooling
water which then goes  directly to wastewater treatment.]  Both
VOHAPs  (e.g., methanol,  acetaldehyde, 1,4-dioxane etc.) and
ethylene  glycol are present  in the contact condenser effluent.
 [Note:  Ethylene glycol  is a  HAP but  is not volatile enough to
be considered a VOHAP.]
     The  VOHAP are  emitted through volatilization from the
 contact condenser  effluent as it leaves the barometric leg  and
makes  its way to  the  process  contact  cooling tower.  Any VOHAP
 remaining in the  contact condenser effluent  is volatilized  in
 the process contact cooling tower.  Ethylene glycol  is emitted
 from the process contact cooling tower  and occurs as a
 combination of stripping (i.e. volatilization)  and drift.
 Stripping occurs when air forced through  the  liquid
 (i.e.,  cooling tower water)  accelerates the  volatilization  of
 compounds within the liquid.   Drift  occurs when very small
 water droplets are blown out of the  process  contact  cooling
 tower.   Drift emissions occur when pollutants are soluble  in
 water or are contained in the water (i.e.,  particulates)  and
                               6-2

-------
are carried out as part of the water droplet.   Further contact
with the commenters revealed that one commenter believed an
equal quantity of ethylene glycol emissions resulted from
stripping (i.e., volatilization) and drift.  The majority of
commenters did not .mention stripping as an ethylene glycol
emission mechanism.
6.1  EXISTING AFFECTED SOURCES
6.1.1  Ethvlene Glycdl Jets as Retrofit Technology
     Comment  (Technical Concerns);   Many commenters provided
comments on the proposed requirements for existing affected
sources.  The bulk of the comments discussed the
inappropriateness of ethylene glycol jets as a retrofit
technology.  While the proposed rule did not require the use
of ethylene glycol jets, they were the technology which was
used to develop the cost basis prohibiting process contact
cooling towers in the proposed rule.  Commenters provided
technical reasons why ethylene glycol jets would cost
substantially more to retrofit at existing processes than
estimated by the EPA at proposal.  Commenters also provided
comments on additional cost issues that the commenters felt
the EPA had not considered.  Commenters disagreed with the
EPA's determination that ethylene glycol jets are the vacuum
system technology of choice for a retrofit application and
with the EPA determination that ethylene glycol jets are cost
effective.  Commenters contended that the EPA had failed to
consider numerous cost factors and had not demonstrated that
the retrofit of ethylene glycol jets was cost effective.
     Commenters presented numerous technical challenges they
would face if required to retrofit their existing vacuum
systems to use ethylene glycol jets.  These technical
challenges have costs associated with them which are discussed
within this comment summary.  Additional cost considerations
are discussed  in the next section.
    . Commenters contended that the EPA had not appropriately
considered the  retrofit of ethylene glycol jets as a process
change.  Commenters stated that  the EPA misjudged the
significance  of the technical aspects of retrofitting the
                              6-3

-------
 vacuum system and this  contributed to  an underestimation  of
 the  costs.
      Commenters stated  that vacuum systems are an integral
 part of the manufacturing process,  and the EPA had failed to
 recognize that the cost impacts  and design considerations of a
 process change are substantially different from those
 associated with the installation of add-on pollution control.
 Commenters discussed the risk  of changing an existing
 performing system to an untried  new system.  They felt there
 could be lingering problems with product quality, a prolonged
 shakedown period for new equipment,  and increased maintenance
 costs.
      Commenters stated  that a  retrofit to the vacuum system is
 likely to affect product quality immediately following the
 retrofit and is likely  to affect it for some time afterwards.
 Commenters  stated that  producing unacceptable product is  a
 cost that was not considered by  the EPA.   [Note:  The impact
 of process  unit shutdown is discussed  later.]  Commenters also
 stated that design requirements  and cost estimates for process
 changes are different from those associated with add-on
 pollution control equipment.   As an example,  typical add-on
 pollution control equipment is placed at the end of the
 process and does not require a significant amount of
 additional  piping to make it functional.  However,  the
 retrofit of ethylene glycol jets would take place in the
 middle  of the process and a significant amount of additional
 piping  would be required.  Commenters stated that costing
 methodologies used by the EPA  are geared toward add-on
 pollution control equipment and  will always understate the
 costs for a process change.  In  addition,  the commenters
 reminded the EPA that in addition to the retrofit costs of
buying  and  installing the new  equipment,  additional costs and
 resources would be required to adapt the equipment  to the
 specific retrofit and to perform the necessary maintenance.
     Commenters  stated that the  EPA analysis  did not  consider
the need to manage the level of  solids going  to the vacuum
system.   Not  managed, solids carried over from the
                              6-4

-------
polymerization reactors will regularly buildup in the vacuum
jets and associated equipment (e.g., condensers)  and cause
system downtime.  The commenters explained that to manage this
problem it would be necessary to (1) design and install a
scrubber system to remove solids prior to the vacuum jets, or
(2) install a spare vacuum jet system.  Both approaches are
currently used by the industry.
     One commenter (IV-D-19) discussed licensing and patent
considerations that could either prevent acquisition of
ethylene glycol jets and associated equipment
(i.e., scrubbers) or make the acquisition more costly.  The
commenter  (IV-D-19) stated that several companies hold PET
process technology patents, and the development or acquisition
of vacuum system technology is complicated by the competitive
relationship between patent holders and other companies.  The
commenter  (IV-D-19) stated that some vendors they contacted
were hesitant to discuss the licensing of any vacuum system
technologies because the commenter  (IV-D-19) was in direct
competition with the license holder.
     Commenters felt that the EPA had not considered the
safety risks associated with the use of ethylene glycol jets.
Commenters expressed concern that pressurized ethylene glycol
is both flammable and toxic and presents a potential safety
threat in the event of a piping or equipment failure.
Commenters cited examples of fires that had occurred and
examples of overseas PET facilities converting from ethylene
glycol jets to steam jets as indications of the serious nature
of these safety concerns.  Because of this risk,  ethylene
glycol jet systems require careful equipment design and
special operating practices, including strict maintenance
procedures.  Commenters stated that piping codes and equipment
standards for flammable vapor handling would require replacing
all piping and equipment currently in steam service.
Introducing flammable ethylene glycol into areas presently
classified for general electrical service would necessitate
upgrading some electrical systems.
                              6-5

-------
     Comment  (Cost Concerns):  Many commenters stated that the
EPA had not considered the costs and other financial
ramifications associated with process unit shutdowns and
production curtailments that would occur as part of
retrofitting ethylene glycol jets.  [Note:  The production of
unacceptable product was discussed earlier.]  Commenters
explained that existing affected sources typically do not
experience these problems when installing add-on pollution
control equipment.  However, the retrofit of ethylene glycol
jets is a process change and process unit shutdowns and
production curtailment is a certainty.  Commenters explained
that the loss of production and potential loss of market
position would cause great economic injury.  Commenters
insisted that the EPA should have considered these costs in
their analysis.
     One commenter  (IV-D-21) stated that any benefit from the
decreased use of the cooling tower after retrofitting ethylene
glycol jets would be site specific.  The commenter  (IV-D-21)
stated that even though a cooling tower would no longer be
required to support the vacuum system, it may still be
required for  other parts of the process or for other processes
at the facility.
     Commenters stated that additional costs would be incurred
if existing structures required modification to accommodate
the retrofit  of ethylene glycol jets.  For example, commenters
stated that it was possible that the weight limit of building
floors could  be exceeded.  Another possibility would be that
building height limits may be exceeded if a new floor was
required or if retrofit equipment protruded from the existing
top floor.  The most likely problem would be that floor space
would be limited and substantial rearranging of existing
process equipment would be required.
     One commenter  (IV-D-19) stated that the EPA had not
considered the potential need for additional glycol recovery
equipment to  handle the recycling of ethylene glycol.from an
ethylene glycol jet vacuum  system.  The commenter  (IV-D-19)
stated that use of  an ethylene glycol jet vacuum system will
                              6-6

-------
increase the quantity and change the characteristics of the
recovered ethylene glycol streams.  These changes would likely
require additional recovery equipment.
     Commenters universally stated that it was not cost
effective to retrofit existing facilities with ethylene glycol
jets. Commenters provided independent cost estimates and
calculations of cost effectiveness values for the retrofit of
ethylene glycol jets.  Commenters provided cost-effectiveness
values for retrofit of $13,000 to $48,000/ton (IV-D-20);
$50,000/ton  (IV-D-11); and approximately $8,200/ton to
$17,000/ton  (including shutdown costs)   (IV-D-25).
     One commenter  (IV-D-19) felt that their facility would
experience a disproportionately large impact in comparison to
other facilities, largely because of the differences in their
process compared to other continuous processes.   This
commenter  (IV-D-19) requested a separate subcategory be
established for their process and proposed an alternate MACT.
[Note:  More comment on the request for a separate subcategory
by this commenter is presented in Section-6.1.4.]
     Many commenters contended that the EPA had not provided
adequate information on the expected compliance costs for PET
facilities to meet the cooling tower prohibition to enable the
public to evaluate the EPA's analysis.   The commenters
acknowledged that the EPA had stated that these costs were
confidential business information (CBI).  However,  the
commenters argued that they were not able to assess the EPA's
analysis and these costs should not be considered CBI.
     Response;  The EPA agrees that some technical and cost
aspects associated with the retrofit of ethylene  glycol jets
were not considered in the analysis supporting the proposed
rule.  The analysis at proposal was limited by two factors.
First, data concerning the retrofit of ethylene glycol jets
were limited to a single facility.   In the absence of
additional data, the EPA assumed the facility and data were
representative of the industry; that turned out not to be the
case.  Second, because some impacts (i.e.,  lost production and
variations on retrofit costs)  are very facility-specific,  the
                             6-7

-------
EPA does not attempt to account for them in an overall cost
analysis, as was done for the Group IV proposal.  Regardless,
based on the comments provided, the EPA agrees that ethylene
glycol jets are not a suitable retrofit technology, either
from a technical or cost-effective perspective, for meeting
the provision of the proposed rule that prohibits the use of
process contact cooling towers.
     The EPA knew that the proposed rule was based on limited
data and a limited knowledge obtained by the EPA from one
producer of PET concerning installation and operation of
ethylene glycol jets.  Therefore, the EPA took special effort
to make this clear and to solicit comments in the preamble to
the proposed rule  (see pages 16104 and 16107 of the preamble).
Based on the data available at proposal, the costs and
emission reductions achievable through the use of ethylene
glycol jets in retrofit situations were acceptable.  In
contrast, several commenters provided detailed cost estimates
for the retrofit of ethylene glycol jets, and these data
showed that the retrofit of ethylene glycol jets were not at
acceptable cost effective levels.  The EPA also received
comments from the company that provided the original ethylene
glycol jet retrofit data.  This commenter provided additional
information describing the special circumstances that made the
retrofit of ethylene glycol jets economical  (i.e., from a
business perspective) and cost effective  (i.e., dollars per
ton pollutant removed) for the specific facility.  In
addition, they pointed out how those same circumstances did
not exist at another of their  facilities and provided a cost
estimate for retrofitting the  second plant; this estimate
demonstrated that  such a retrofit would not be cost effective.
     The EPA understands why commenters were concerned by the
limited  amount of  cost analysis that was presented at
proposal.  However, when data  are requested to be treated as
CBI, the EPA cannot present the data without receiving
permission from  the data provider or making a finding that the
data is  not CBI.   The EPA did  not feel it was critical to
present  the ethylene glycol retrofit data in more detail
                              6-8

-------
 because the rulemaking action was a proposal,  and the EPA
 fully expected that the public would provide additional data.
      Given that the EPA was convinced by the comraenters'
 arguments that ethylene glycol jets are not a suitable
 retrofit technology, the EPA considered alternate options for
 controlling emissions from the vacuum system.   Both VOHAP and
. ethylene glycol are emitted from the vacuum system, and the
 EPA chose to approach each of these emissions separately.
      To address VOHAP emissions from the vacuum system, the
 final rule treats the contact condenser effluent as wastewater
 and applies the same provisions to it as are applied to
 process wastewater.  Contact condenser effluent is considered
 process wastewater based on the proposed wastewater
 provisions, and, without any special provisions or specific
 mention, the wastewater provisions will apply.  The EPA judged
 that treating the vacuum system wastewater prior to any
 significant opportunity for volatilization makes environmental
 sense.  Further, the HON wastewater provisions, which are the
 basis for the wastewater provisions in the proposed and final
 rule, have been judged to be environmentally effective and
 cost effective overall.  Therefore, if the HON wastewater
 provisions designate a wastewater stream as Group 1,  it is
 cost effective to control the VOHAP present in the wastewater
 stream.
      Addressing emissions from the process contact cooling
 tower was more complex.  [Note:  The HON did not specifically
 address process contact cooling towers because they are not
 used extensively in that industry.]  Given that the emissions
 of VOHAP would be dealt with through the use of the wastewater
 provisions, the key issue related to process contact cooling
 towers became the emission of ethylene glycol.  Between
 proposal and promulgation,  the EPA spent considerable effort
 gathering information on PET vacuum systems and their
 emissions.  Mudh of this effort focused on emissions from the
 process contact cooling towers, specifically emissions of
 ethylene glycol.  Based on industry reported emissions, total
 ethylene glycol emissions for the industry are approximately
                               6-9

-------
340 Mg/yr, and approximately 230 Mg/yr are being emitted from
a single facility.  That means the other eleven facilities
emit an average of 10 Mg/yr of ethylene glycol.  In addition
to gathering information on emissions, the EPA investigated
control options aimed at reducing emissions of ethylene glycol
from process contact cooling towers.  Possible control options
that were investigated, not considering ethylene glycol jets,
include treating a slipstream of the cooling tower water to
reduce the concentration of ethylene glycol or installing a
large heat exchanger to isolate the process contact cooling
tower from the process.  None of these control options were
shown to be cost effective.  All of the factors discussed
above have led the EPA to conclude that, with one exception,
specific provisions for process contact cooling towers are not
warranted.
     The one exception where it was found to be cost effective
to control emissions from existing process contact cooling
towers was for affected sources manufacturing PET using a
continuous TPA high viscosity multiple end finisher process.
There is a single facility using this process, and the
representatives of said facility (IV-D-19) requested that a
separate subcategory be established for their process and
proposed an alternate MACT.  This facility has become subject
to the Polymer Manufacturing NSPS during this rulemaking
(see docket A-45-92, Section IV-D).  Because this facility is
already required by the Polymer Manufacturing NSPS to reduce
the concentration of ethylene glycol in the cooling tower
water to 6 weight percent, it was shown to be cost effective
to reduce the concentration further to 4 weight percent.  The
analysis demonstrating costs and emission reductions was
considered CBI by the facility, and is, therefore, not
presented.  Discussion of establishing a separate subcategory
is presented in Section 6.1.4.  As part of establishing a
separate subcategory and setting a standard for process
contact cooling towers, the issue of compliance time had to be
considered, and it was the EPA's judgement that 3 years would
be required to achieve compliance.  The single facility in
                             6-10

-------
this subcategory has just become subject to the Polymer
Manufacturing NSPS and requested the full 3 years to achieve
compliance with this rule.  Representatives of the facility
provided information demonstrating that additional equipment
may be required and experience operating the system would be
needed before the facility could be assured of meeting the
standard in the thermoplastics rule.
6.1.2  Alternative Technologies
     Comment:  Many commenters provided alternatives to the
provisions of the proposed rule that prohibit the use of
process contact cooling towers.  These alternatives included:
(1) meeting the requirements of the Polymer Manufacturing
NSPS, (2) reducing emissions from process contact cooling
towers by 80 percent,  (3) establishing an emissions
performance level similar to the level established in the
NSPS, (4) treating the process contact cooling tower as a
process vent,  (5) stripping the contact condenser effluent on
the hot side of the condenser and then recirculating it to the
cooling tower, and  (6) treating vacuum system wastewater as
any other wastewater stream.
     In addition, comments were provided on the technologies
the EPA had discussed in the proposed rule as alternatives to
ethylene glycol jets.  Commenters stated that the alternate
technologies were not technically feasible and were not cost
effective options.  Detailed costs for these alternate
technologies were not provided by the commenters.  However,
approximate costs were provided by comtnenters to support their
claim that the alternate technologies were not cost effective.
One commenter  (IV-D-19) felt the EPA was required to conduct a
thorough analysis of these alternatives and failed to do so.
     Commenters stated that the use of noncontact condensers
was not feasible for high vacuum needs.  Commenters stated
that the high vapor flow rates would necessitate the use of
very large noncontact condensers.  Commenters went on to state
that large noncontact condensers would require significant
building modifications.  Commenters stated that solids from
the process would deposit inside the noncontact condenser,
                             6-11

-------
eventually causing plugging and significant downtime.
Commenters explained that each reactor would either need spare
noncontact condensers to be used when the primary noncontact
condenser was being cleaned or would need a scrubber to
prevent solids from entering the vacuum system.  One commenter
(IV-D-19) stated that the capital costs of installing
noncontact condensers would equal or exceed the cost of
retrofitting ethylene glycol jets.  Another commenter
(IV-D-21) stated that when annual operating costs are
considered, costs for noncontact condensers would be higher
than ethylene glycol jets.
     Commenters stated that mechanical vacuum pumps were
impractical because of the large volume of vapor and the high
vacuum requirements of PET polymerization reactors.  One
commenter  (IV-D-21) stated that mechanical vacuum pumps
suitable for high vacuum are quite expensive and more
complicated than ethylene glycol jets.  The commenter
(IV-D-21) went on to state that when increased maintenance
costs and lost production due to reliability problems were
considered, ethylene glycol jets were more economical.  One
commenter  (IV-D-19) stated that the very high vacuum
requirements of their process would necessitate a large number
of mechanical vacuum pumps for each reactor and more than
100 pumps overall.
     Response;  While commenters provided adequate data to
dissuade the EPA from prohibiting the use of process contact
cooling towers, adequate data were not provided to enable the
EPA to consider the recommended alternatives.
     The EPA was aware at proposal that these alternatives
would not be suitable 'in all cases,  and the provided comments
confirmed this understanding.  While comments provided in
writing described the circumstances under which the
alternatives would not work, information gathered through
industry meetings and plant trips prior to proposal showed
that these alternatives had been applied and were favorably
considered in certain situations.  Based on,the data gathered
                             6-12

-------
prior to proposal, the EPA included these technologies as
alternatives to ethylene glycol jets in the proposal.
     One commenter stated that the EPA was required to conduct
a thorough  analysis of the alternatives and had failed to do
so.  The EPA did not have the data at proposal to do such an
analysis and did not consider one to be required.  The EPA did
not portray the alternatives as the basis of the rule, but
offered them as "alternatives" and specifically solicited
comments on their applicability, effectiveness, and cost.
6.1.3  Vacuum System Wastewater Requirements
     Comment:  Many commenters objected to the proposed
provision designating all vacuum system wastewater streams as
Group 1.  Commenters stated that the EPA had not provided
rationale explaining why the standard for vacuum system
wastewater  was more stringent than the standard for other
wastewater  streams.
     Commenters contended that in order to require the control
of all vacuum system wastewater streams, the EPA would be
required to perform a cost and impacts analysis examining the.
incremental cost of controlling Group 2 wastewater streams.
Commenters  stated that the EPA had not provided the results of
such an analysis.  Commenters felt that requiring the control
of Group 2  wastewater streams would impose unreasonable costs
without providing additional emission reduction.
     In addition, commenters explained that designating all
vacuum system wastewater streams as Group 1 would create
confusion about how the wastewater standards should be
applied.  For instance,  the commenters asked how the standards
would be applied when a wastewater stream had a VOHAP
concentration that already met the VOHAP concentration limit
for some of the treatment options found in the wastewater
provisions.  Further,  commenters explained that by designating
all vacuum  system wastewater streams as Group 1,  material
removed from the streams would be considered residuals.
Commenters  stated this would impact their ability to recover
ethylene glycol and might also require costly treatment  of
                             6-13

-------
oligomers that are removed from the wastewater stream,  without
any corresponding environmental benefit.
     Response;  The purpose for designating all vacuum system
wastewater streams as Group 1 was explained in Chapter 6.0 of
the Basis and Purpose Document for the proposed rule
[EPA-453/R-95-004a; March 1995].  To summarize, at the time of
proposal, the EPA considered ethylene glycol jets to be
pollution free because there was no vacuum system wastewater
stream produced.   [Note:  Since proposal, the EPA has come to
understand that ethylene glycol jets are not a pollution free
technique because the additional loading to the glycol
recovery unit can create additional emissions from existing
process vents or wastewater streams or both.]  Therefore, in
order to ensure equivalency between the use of ethylene glycol
jets and noncontact condensers, the EPA determined that all
vacuum system wastewater produced by the noncontact condensers
must be controlled.
     As discussed above, ethylene glycol jets are not suitable
for retrofit, and the alternative technologies, including
noncontact condensers, were not pursued as viable alternatives
upon which to base the prohibition of process contact cooling
towers.  Because the automatic designation of vacuum system
wastewater streams as Group 1 was done in an effort to make
noncontact condensers equivalent to ethylene glycol jets and
that need no longer exists, there is no longer a need or
justification to designate all vacuum system wastewater
streams as Group 1.  As a result, the EPA has changed the
final rule to implement the group determination procedure for
vacuum system wastewater, consistent with the requirements for
other process wastewater.
     Commenters  felt that the EPA was required to conduct an
impacts analysis examining the incremental cost of controlling
Group 2 wastewater streams.  The EPA did not conduct such an
analysis  at  proposal because noncontact condensers, the use of
which required  an  automatic Group 1 designation, were offered
as an alternative  only, not as the basis of the standard.
                              6-14

-------
     Finally, because the EPA is no longer automatically
designating any wastewater streams as Group 1 and is instead
using the applicability criteria found in the proposed
wastewater provisions, the implementation problems described
by the commenters are no longer a concern.
6.1.4  Miscellaneous
     Comment;  One commenter (IV-D-16) stated that the EPA had
failed to establish "a credible floor that sets ethylene
glycol jets as the technology of choice from steam jet vacuum
systems."  The commenter (IV-D-16) continued that "while this
technology may be appropriate for new sources, we challenge
the Agency to make a credible demonstration that this
technology represents the floor for existing sources."
     Response:  The EPA did not intend to imply that the
retrofit of ethylene glycol jets, which served as the basis
for prohibiting the use of process contact cooling towers, was
representative of the MACT floor.  In Chapter 8.0 of the Basis
and Purpose Document  [EPA-453/R-95-004a, March 1995], the EPA
states that the MACT floor for process contact cooling towers
is no control.  Also in Chapter 8.0, the EPA states that "the
Polymer Manufacturing NSPS is more stringent than the MACT
floor.  However, as discussed earlier, ethylene glycol jets
were determined to be more cost effective in reducing
emissions than the requirements of the Polymer Manufacturing
NSPS."  The EPA intended this to be a clear statement that the
retrofit of ethylene glycol jets was considered to be more
stringent than the MACT floor.
     Comment;  One commenter (IV-D-19) requested that a
separate subcategory be established for the continuous TPA
high viscosity multiple end finisher process.  The commenter
(IV-D-19) contended that the subcategorization scheme for PET
continuous processes ignored fundamental differences between
the continuous TPA high viscosity multiple end finisher
process and other high viscosity processes and other
continuous processes.  The two most significant differences
are the higher vacuum required by the multiple end finisher
process and  the absence of scrubber technology.  The commenter
                             6-15

-------
(IV-D-19) stated that the EPA did not evaluate the specific
impacts on their process.  As a result,  the commenter
(IV-D-19) was expecting a disproportionately large impact when
compared to owners or operators that use other processes.
     Response:  The EPA has gained an increased understanding
of the TPA continuous high viscosity multiple end finisher
process since proposal and agrees with the commenter that this
process warrants a separate subcategory distinct from other
continuous PET processes.  In addition to the points raised by
the commenter, a significant difference between this process
and others is the higher level of ethylene glycol emissions
from the process contact.cooling tower.  This difference in
emissions is also a consideration in establishing a separate
subcategory for this process.
     As  discussed earlier in Section 6.1.1, this commenter
 (IV-D-19) proposed an  alternate MACT for controlling emissions
of ethylene glycol from the process contact cooling tower.  A
condition of  including the alternate MACT  for this unique
process  and facility was establishing a separate subcategory.
     The final  rule  has established  a distinct  subcategory for
this process, which  is used  at a  single facility within  the
PET  industry, with requirements for  controlling emissions  from
the process contact  cooling  tower.
      Comment:   In response to the EPA's request for
 information on  methodologies for estimating emissions  from
process  contact cooling towers,  two  commenters  provided
methods  they have used.  One commenter (IV-D-06)  estimated
 emissions  using a material balance methodology.  Emissions
 were calculated by measuring the flow and HAP concentration at
 specific locations;  inputs to the process contact cooling
 tower were restricted to those streams where HAP
 concentrations could be measured with confidence.   Another
 commenter (IV-D-05)  included a method for measuring process
 contact cooling tower VOC emissions which was developed as an
 inspection and maintenance program rather than as an emissions
 estimation technique.
                              6-16

-------
      Response:   The EPA appreciates the information submitted
 by the commenters.   The request for an emissions  estimation
 methodology was based on the need to consistently estimate
 emissions from process contact cooling towers  for the  purposes
 of emissions averaging.  In the final rule,  emissions  from
 process contact cooling towers at existing affected sources
 are not regulated for the majority of affected sources.
 [Note:   As discussed previously in this chapter,  the final
 rule will control VOHAP emissions from the contact condenser
 effluent using the wastewater provisions from  the proposed
 rule;  this action will control any significant VOHAP emissions
 from this emission point.]   For the single subcategory where
 process contact cooling tower emissions from existing  affected
 sources are regulated,  the EPA has judged that emissions from
 process contact cooling towers will not be eligible  for
 emissions averaging (see Chapter 9).   Therefore,  an  emissions
 estimation methodology for process contact cooling towers  is
 not required for the final rule.
      Comment:   Commenters  stated that the EPA's
 characterization of ethylene glycol jet vacuum systems as
 "pollution free" is inaccurate.   Commenters  explained that
 ethylene glycol jet vacuum systems send recovered  ethylene
 glycol,  which contains  VOHAP,  to the  glycol  recovery unit  in
 quantities • greater  than what would be sent  if  a steam jet
 vacuum  system were  being used.   At the  glycol  recovery unit,
 the VOHAP (e.g.,  acetaldehyde,  1,4-dioxane,  etc.)   contained  in
 the recovered ethylene  glycol stream  is  typically stripped and
 either  emitted  to the atmosphere or is  sent  to an air
 pollution control device.   [Note:   The proposed and final
 rules would regulate this  emission point  as either a
 continuous or batch process  vent.]  The water and glycol
 stream  is sent  to glycol recovery.  Eventually, a wastewater
 stream  is generated as  a result  of  the recovery process;  this
wastewater stream would be subject  to the proposed and final
 rules.   In addition, .there is  a process vent from the
mechanical vacuum pump, generally the last stage of an
                             6-17

-------
ethylene glycol vacuum jet system, that has the potential to
emit HAP.
     Response;  The EPA appreciates the provided comments and
relied on them in developing the requirements for the final
rule.
     Comment:  Commenters stated that the EPA had not provided
the basis used to determine that ethylene glycol jets are less
costly than the Polymer Manufacturing NSPS requirements for
controlling ethylene glycol emissions from process contact
cooling towers.  Commenters did not believe the retrofit of
ethylene glycol jets was less costly than applying the
requirements of the Polymer Manufacturing NSPS, but did not
provide data of their own.  Further, commenters stated that
the retrofit of ethylene glycol jets could not be less costly
for a facility that was already complying with the Polymer
Manufacturing NSPS requirements.   Commenters questioned why
the Polymer  Manufacturing NSPS requirements were not chosen
over the retrofit of  ethylene glycol jets.
     Response:  At proposal, the  EPA stated that the retrofit
of ethylene  glycol  jets was considered  to be both less
expensive  and more  effective than the Polymer Manufacturing
NSPS requirements.   Because of the threshold emission rate
applicability criteria in the Polymer Manufacturing NSPS,
which  requires  a facility's emissions to be above a certain
level  before being required to take action and meet the
emission limits of the rule, the  EPA had the understanding
that very few,  if any,  facilities had been required to take
positive action to meet the  requirements of the  Polymer
Manufacturing NSPS.   Data collected during the development of
 the proposed rule did not indicate that any  facilities had in
 fact become subject to the Polymer Manufacturing NSPS
 requirement.  Further, the EPA did not  believe that
 implementation of the Polymer Manufacturing NSPS requirements
 would achieve significant emission reductions.   For  these
 reasons, the EPA did not consider it  necessary at proposal  to
 evaluate the requirements of the Polymer Manufacturing NSPS  as
 a regulatory alternative.
                              6-18

-------
6.2  NEW AFFECTED SOURCES
     Comment:  Several commenters mentioned the prohibition of
process contact cooling towers for new affected sources,
however, none presented objections to it.  Commenters did
object to the provision that designated all vacuum system
wastewater streams as Group 1.
     Response:  The provision of the proposed rule for new
affected sources which prohibits the use of process contact
cooling towers has been retained in the final rule.  Owners or
operators have several options to meet this requirement --
ethylene glycol jets, once-through cooling water system,
noncontact condensers, or isolation of the contact condenser
cooling loop from the cooling tower to name a few.  It should
be noted that while many commenters objected to these
techniques for retrofit applications, very few objectionable
comments were made regarding their use in a new facility.  In
fact, some commenters supported the use of ethylene glycol
jets in new facilities.  The EPA has determined this
requirement to be the MACT floor for new affected sources.
However, the provision designating all vacuum system
wastewater streams as Group 1 has been dropped from the final
rule.  The final rule implements the group determination
procedure for vacuum system wastewater, consistent with the
requirements for other process wastewater.
     As described earlier, the purpose for designating all
vacuum system wastewater streams as Group 1 was to ensure
equivalency between the use of ethylene glycol jets and the
use of noncontact condensers.  At the time of proposal, the
EPA considered ethylene glycol jets to be pollution free
because there was no vacuum system wastewater stream produced.
The automatic designation of all vacuum system wastewater as
Group 1 was meant to address the absence of vacuum system
wastewater when ethylene glycol jet systems are used.  Since
proposal, the EPA has come to understand that ethylene glycol
jets are not pollution free because the additional loading to
the glycol recovery unit can create additional emissions from
process vents or wastewater streams or both.
                             6-19

-------
     Realizing that ethylene glycol jets are not pollution
free, the EPA considered the additional emissions at the
glycol recovery unit when ethylene glycol jets are used and
the emissions from the vacuum system wastewater when
noncontact condensers are used.  In either case, an equivalent
quantity of VOHAP is introduced to the vacuum system from the
process.  Process vents and wastewater streams at the glycol
recovery unit are subject to the provisions of the standard.
For example, a wastewater stream at the glycol recovery unit
would be the subject of a group determination and, based on
the results, would be controlled if required.  The EPA judged
that to require the same action for the vacuum system
wastewater from noncontact condensers would ensure equivalency
between a new affected source using ethylene glycol jets and
one using noncontact condensers.
                              6-20

-------
                   7.0  WASTEWATER OPERATIONS
 7.1  SELECTION OF STANDARDS
      Comment;   Two commenters  (IV-D-10 and IV-D-23)  felt
 several  differences between wastewater streams from the
 thermoplastics industry and the  HON industry served to refute
 the assumption that the HON is a ready-made regulatory
 alternative.   One commenter (IV-D-23)  cited the following
 differences  between the wastewater streams from the
 thermoplastics and HON industries:   (1). HAP emissions  per
 thermoplastic  source are an order of magnitude lower than for
 HON sources,  (2)  the cost per ton of HAP removed for
 thermoplastic  wastewater is an order of  magnitude higher than
 for HON  wastewater,  and (3)  the  HON reference control
 technology  (steam stripping) is  not a  viable option for  most
 thermoplastic  wastewaters.
      Three commenters (IV-D-08,  IV-D-11,  IV-D-23) explained
 that  the EPA should not make an  industry subject  to  standards
 which it knows will not result in any  emission reductions.
 The  commenters (IV-D-08,  IV-D-11,  IV-D-23)  stated that even if
 the  installation  of controls are  not required under  the
 proposed wastewater provisions, there  would still be
 considerable monitoring,  recordkeeping,  and reporting costs.
 Also, affected sources  would still  be  at  risk of noncompliance
 with monitoring,  recordkeeping, and reporting requirements.
 The commenters (IV-D-08,  IV-D-11,  IV-D-23)  stated that the
 wastewater provisions in the proposed  rule  contradict the
 Reinventing Environmental Regulation initiative
 (March 16,  1995).
     Pour commenters (IV-D-08,  IV-D-11,  IV-D-15, IV-D-23)
objected to the EPA's determination  that  the wastewater MACT
 floor, which is no  control for most  affected  sources, is
equivalent to  the HON level of control.   Commenters  (IV-D-08,

                             7-1

-------
IV-D-11, IV-D-23) stated that the EPA inappropriately
concluded that in cases in which affected sources have not
installed controls and no controls are required by the HON
wastewater provisions, the level of control installed by the
affected source  (i.e., none) is equivalent to the HON level of
control.  Commenters  (IV-D-08, IV-D-23) stated that only
affirmative actions, not a lack of control actions, should be
considered in determining whether or not an affected source is
controlled at a  level equivalent to that of another
regulation.  The commenters  (IV-D-08,  IV-D-11, IV-D-15,
IV-D-23) stated  that, based  on the data, the EPA should have
concluded that there  is no basis for regulating wastewater
emissions.  Two  commenters  (IV-D-08, IV-D-15) specifically
objected to the  decision to  regulate wastewater for the
polystyrene subcategories.
   •   Commenters  (IV-D-08,  IV-D-11,  IV-D-15,  IV-D-23) went on
to say that any controls beyond  the MACT floor of  no control
should be  justified on a  cost impact basis.   Three commenters
 (IV-D-08,  IV-D-11,  IV-D-23)  cited  cost estimates  from  the SID
for the proposed rule of  about $5,700  to $27,500  per ton of
emission reductions compared to  the $430 per ton  of emission
reduction determined for the HON.   [Note:  • The
cost-effectiveness values cited by the commenters include
 impacts experienced at the MACT floor.]   The commenters
 (IV-D-08,  IV-D-11, IV-D-23)  concluded that the  fact  that these
 costs differ so dramatically reveals that thermoplastic and
HON wastewater  streams are not similar.  The commenters
 (IV-D-08, IV-D-11, IV-D-15, IV-D-23)  contended that  any
 control requirement more stringent than the MACT floor of no
 control would not be justified based on cost.
      Finally, three commenters  (IV-D-08, IV-D-11, IV-D-15)
 stated that the selection of steam stripping as the reference
 control technology  (RCT) for treatment of wastewater from
 manufacturing processes using styrene is inappropriate due to
 polymerization  problems.  The commenters  (IV-D-08, IV-D-11
 IV-D-15) explained  that the styrene will polymerize
 heated and will cause  plugging, resulting in
       when
downtime
the
                               7-2

-------
steam stripper and lost production time.  One commenter
(IV-D-08) presented an example of this problem.  At a SOCMI
facility, where styrene concentrations in the wastewater were
relatively low (i.e., 136 ppm) when compared to what is
expected at a polymer facility, a steam stripper plugged in
approximately three months.  A gelatinous polymer formed in
the lower section of the column and in some of the associated
piping.  The steam stripper was cleaned and continued use was
possible with a two to three day shutdown for maintenance
occurring approximately every six months.  This commenter
(IV-D-08) was not aware of a single polystyrene facility that
satisfactorily employs steam stripping to reduce emissions
from process wastewaters.  Commenters (IV-D-08, IV-D-11,
IV-D-15) stated that the EPA should identify a technology
capable of achieving the standards, and the EPA must provide
data and a cost analysis to support the required technology.
Also, the commenters  (IV-D-08, IV-D-11,  IV-D-15) stated that
it is not acceptable for the EPA to transfer the RCT from an
entirely different industry  (i.e., the HON) to the polystyrene
subcategory without considering the specific technical
challenges that will be faced.
     Response;  Commenters stated that wastewater streams from
the HON and thermoplastics industries were different enough
that the HON wastewater provisions were not a ready-made
regulatory alternative.  In response, it must be considered
that the HON covers a very large source category.  The
characteristics of SOCMI wastewater streams quoted by the
commenters represent the average characteristics from this
very large industry.  In fact, there are wastewater streams
within the SOCMI that are very similar to thermoplastic
wastewater streams in that they have concentration values and
flow rates close to the limits of the applicability
requirements (i.e., 1,000 ppm and 10 liters/minute).  Also,
there are wastewater streams within the PET subcategories with
flow rates significantly above the applicability criteria of
10 liters/minute.  For example, during industry meetings and
plant trips, the EPA was commonly told that flow rates for the
                              7-3

-------
first contact condenser in a steam jet vacuum system were in
the range of 1,135 to 1,325 liters/minute.  Finally,  the
organic HAP regulated by the thermoplastics rule are also
regulated by the HON.  In summary, based on the limited data
made available to the EPA and because wastewater stream
characteristics overlap between the two industries, the EPA
considers the use of the HON as an applicable regulatory
alternative for the thermoplastics industry to be reasonable.
     Commenters stated that the EPA should not make an
industry subject to standards when no emission reductions are
expected.   [Note:  This comment applied to the polystyrene
subcategories.]  Commenters went  on to state that even if the
application of control equipment  is not required, affected
sources will be subject to monitoring, recordkeeping, and
reporting requirements.  The EPA  has determined that emission
reductions are expected for the PET subcategories, but, based
on the available data, emission reductions are not expected
for  the nitrile resin subcategory.  Regardless, for those
subcategories where emission reductions are not expected, data
are  not sufficient to be sure that emission reductions would
not  be achieved upon implementation of the rule.   If, in fact,
controls are  not required  for nitrile resin facilities, the
recordkeeping and reporting requirements  for Group 2
wastewater  streams are considered by  the  EPA to be essential
for  ensuring  compliance with the  wastewater provisions.  The
Group  2 requirements entail identifying Group 2 streams, their
'flow rate,  and  their concentration in the Notification of
Compliance  Status.   Further, monitoring is not required unless
control  equipment has been applied.
     Commenters objected  to the EPA's determination of the
wastewater MACT floor being equivalent to the HON based on the
fact that existing  affected sources do not control wastewater
and the  HON does not require  control.  The EPA considers the
MACT floor to be comprised of  two things:  (1) a  control
technology,  and (2)  a description of  the  emission points to
which it is applied.   Once a  control  technology  is selected,
the EPA has determined that comparing applicability  criteria
                              7-4

-------
from existing rules to the controls and emission point
characteristics at existing facilities is an appropriate way
to approximate a description of the MACT floor.  As described
in the Basis & Purpose Document for the proposed rule
[EPA-453/R-95-004a; March 1995], regulatory alternatives
equivalent to or more stringent than the MACT floor were
developed by comparing controls at existing facilities to a
selected set of rules/regulations  (i.e., the HON for
wastewater control).  When it was determined that the HON was
approximately equivalent to the MACT floor, the EPA chose to
rely on the applicability criteria of the HON to describe the
MACT floor for purposes of the rule.  The use of the HON
applicability criteria was based on the rationale that for
those subcategories where the HON was, in fact, slightly more
stringent than the MACT floor, the Administrator had
determined that the HON level of control had reasonable
impacts and was therefore acceptable.  When the EPA found the
HON to be clearly more stringent than the MACT floor
(i.e., the ASA/AMSAN subcategory), the HON was accepted or
rejected based on cost effectiveness and other nonair,
environmental, and energy impacts.  For the regulation of
wastewater from the production of ASA/AMSAN, the HON cost
effectiveness was approximately $25,000/Mg; the HON wastewater
provisions were rejected in this case.
     Commenters cited cost-effectiveness values for the
control of wastewater streams from the proposal.   The data
received by the EPA from the industry in response to the
Section 114 and ICR responses were minimal and were not
adequate to determine reliable cost-effectiveness values for
the range of wastewater streams covered by the rule.   Because
of the lack of reliable data supplied to the EPA, the EPA
analyzed the HON cost-effectiveness values and determined that
the HON cost-effectiveness numbers considered a wide range of
wastewater streams which appeared to be similar to the types
of wastewater streams present at thermoplastic facilities.
Therefore, because of the lack of more specific data, the EPA
determined that the HON cost-effectiveness values provided the
                              7-5

-------
most representative cost-effectiveness numbers available for
the thermoplastics industry lacking more industry specific
data for the regulated source categories.
     Commenters claimed that the selection of steam stripping
as the RCT for treatment of styrene-containing wastewater at
proposal was inappropriate due to polymerization problems.
The EPA acknowledges that steam stripping styrene-containing
wastewaters may prove to be impractical in some instances due
to issues raised by the commenters.  The proposed and final
wastewater provisions provide several options for complying
with the wastewater provisions.  If the owner or operator
judges steam stripping not to be practical for their process,
one of the other wastewater compliance options may be used.
The EPA considers one of these compliance options, biological
treatment, a favorable option because of the biodegradation
characteristics of the organic HAP of concern, such as  styrene
and acrylonitrile.  A well-operated and well-maintained
biological treatment system can achieve organic HAP emission
reductions as  high as  99 percent.  Because the organic  HAP
regulated by the thermoplastics rule are very biodegradable,
the EPA has chosen to  omit from the final rule the HON
requirement to demonstrate that 95 percent of the mass  of
organic HAP are removed.  The EPA did not consider it
necessary to perform a cost analysis specific to biological
treatment because the  additional cost to hard pipe affected
streams directly to the biological treatment unit or to
retrofit affected drains was  considered  to be less costly than
steam stripping.
      As  for the polystyrene  subcategory, the EPA has exempted
this  subcategory from the wastewater provisions in the  final
rule.   Because of  the  water solubility  characteristics  of
 styrene (300  ppm maximum) ,  wastewater streams containing only
 styrene cannot meet  the applicability criteria of a Group 1
wastewater stream,  which requires  an  organic HAP  concentration
 of 1000 ppm.   Based on the available  data,  styrene  is  the only
 organic HAP in wastewater at polystyrene affected sources.
 The EPA judged that there would never be any Group  1
                              7-6

-------
wastewater streams at polystyrene affected sources and
exempted these subcategories from the final wastewater
provisions.
     Comment:  One commenter  (IV-D-22) requested that
provisions be provided for existing wastewater treatment
plants versus requirements for future treatment plants.  No
specific concerns were indicated by the commenter.
     Response:  The difference in the requirements for
existing and new affected sources lies in the group
determination procedure.  For existing affected sources, the
wastewater provisions regulate a specific list of organic HAP
designated as volatile organic HAP  (VOHAP).  This list is
presented in Table 9 of the HON.  For new affected sources,
the wastewater provisions regulate the same list of VOHAP;
however, a subset of the list of organic HAP appearing on
Table 9 are listed in Table 8 of the HON.  [Note:  All the
organic HAP regulated by the thermoplastics rule are
considered VOHAP per Table 9 of the HON, except for ethylene
glycol, and butadiene also appears on Table 8 of the HON.]
For the VOHAP listed on Table 8, the applicability criteria
are more stringent (i.e., VOHAP concentration and flow rate
are lower) .   Therefore, a wastewater stream may not be
designated as Group 1  (i.e., be required to be controlled) at
an existing affected source but would be required to be
controlled at a new affected source because the applicability
criteria are more stringent.  Once a wastewater stream is
designated as Group 1, the control requirements are largely
identical regardless of whether it is located at an existing
or new affected source.
     Comment:  One commenter (IV-D-03) supported the
maintenance wastewater requirements of the proposed rule which
require compliance with §63.105 of the HON.   Two commenters
(IV-D-11, IV-D-18) contended that the proposed rule was
ambiguous about maintenance wastewater and should reference
the HON requirements.  Additionally, two commenters (IV-D-03,
IV-D-11) requested that the rule be revised to specifically
state that reactor and equipment washdown water also be
                             7-7

-------
handled as maintenance wastewater.  One cotnmenter (IV-D-11)
stated that because the proposed rule does not clearly
distinguish between maintenance and process wastewaters,
burdensome monitoring, recordkeeping, and control measures
could be necessary for intermittent maintenance wastewaters
that have negligible potential to release significant amounts
of pollutants.
     Response: The final rule has been clarified in the
following ways.  First, the definition of maintenance
wastewater has been added to the final rule.  Second, the
final rule has been clarified to designate reactor and
equipment washdown water as maintenance wastewater.  One
commenter requested that the final rule reference the HON
requirements  for maintenance wastewater.  In fact, the
proposed rule referenced §63.105 of  subpart F in §63.508(b),
and the final rule  [§63.1330(b)] retains this reference.
     Comment;  One commenter  (IV-D-22) stated that the
requirements  for Group  1 wastewater  streams in the HON  are
dictated by the  most  volatile chemicals listed in Table 9  of
the HON, regardless of  the  chemical  characteristics of  other
Table  9 chemicals.  For Group 1 wastewater streams that
require control  under the HON, treatment unit enclosures and
other  vapor entrapments are required throughout the wastewater
system starting  at the  process effluent, regardless of  the
volatility of the  individual  chemicals in the wastewater.   For
example,  ethylene  glycol and  1,4-dioxane are soluble in water,
have vaporization  temperature points above 100° C, and  other
chemical  characteristics,  such that  at small concentrations
 (i.e.,  10  ppm of any one component  from Table 9) in the
wastewater stream,  there will be  no significant air emissions.
The  commenter (IV-D-22)  stated that  great expenditure will be
taken to contain these insignificant emissions and felt that
resources can be better put to.use  elsewhere.  The commenter
 (IV-D-22)  suggested that  the  chemical characteristics of the
Table 9 listed chemicals  be evaluated in  regard to the
requirement to enclose wastewater treatment units and
specifically exempt ethylene glycol and 1,4-dioxane  from this
                              7-8

-------
requirement.  A one-size-fits-all approach to regulating
wastewater without regard to volatility of specific chemicals
is not useful or efficient.
     Response;  In developing the HON wastewater provisions,
the EPA considered the need to enclose waste management units
(e.g., sewers, tanks) for all organic HAP covered by the
wastewater provisions.  Based on this consideration, a number
of organic HAP preliminarily considered for regulation were
dropped from the final HON rule.  Organic HAP remaining on
Table 9 have, in the EPA's judgement, adequate propensity to
volatilize such that they be treated as volatile organic HAP
(VOHAP).  Specifically, 1,4-dioxane and methanol, two organic
HAP likely to be present in wastewater at PET facilities, have
Fe (fraction emitted) values of 0.18 and 0.17, respectively;
meaning that up to 18 percent of the 1,4-dioxane present in a
wastewater stream may be emitted to the atmosphere if the
wastewater stream is left exposed to the atmosphere long
enough.  Acetaldehyde, also present in PET wastewater, has an
Fe value of 0.48.  Acrylonitrile, the organic HAP likely to be
present in wastewater at nitrile resin facilities, has an Fe
value of 0.43.  The commenter did not provide data disputing
the Fe values for pertinent chemicals.  Because this issue was
considered in developing the HON, the EPA did not consider it
necessary to deviate from the HON provisions in the proposed
or final rule.
     As specified in §63.132(g) of the HON, a process
wastewater stream is a Group 1 wastewater stream for organic
HAPs listed in Table 9 if: (1) the total VOHAP average
concentration of the wastewater' stream is greater than or
equal to 10,000 parts per million by weight at any flow rate,
or (2) the total VOHAP average concentration is greater than
or equal to 1,000 ppm by weight and has a flow rate greater
than or equal to 10 liters/minute.  In the example provided by
the commenter for ethylene glycol and 1,4-dioxane, a
concentration of 10 ppm stated by the commenter would not
result in a wastewater stream being designated as Group 1.
Further, ethylene glycol is not included on the list of
                              7-9

-------
Table 9 organic HAPs.   [Note:  The commenter may have been
referring to ethylene glycol ethers, which are on the Table 9
list.]
7.2  MISCELLANEOUS
     Comment;  One commenter (IV-D-07) requested that the EPA
modify appendices A and C of the HON to allow anaerobic
wastewater treatment technology as an approved treatment
technology.  The commenter  (IV-D-07) stated that as proposed,
the thermoplastics rule would not allow the use of anaerobic
wastewater treatment.
     The commenter  (IV-D-07) indicated that the advantages of
anaerobic treatment over conventional aerobic treatment
include:  (1) the anaerobic process occurs in a closed vessel,
minimizing air emissions,  (2) the anaerobic process combines
stripping and combustion with biodegradation to destroy
volatile constituents,  and  (3) methane and other stripped
volatile organics are combusted for their energy value as a
substitute for purchased natural gas.
     The commenter  (IV-D-07) has also commented on this issue
in response to the proposed Part 60, Subpart YYY  (Wastewater
NSPS) .  The commenter  (IV-D-07) provided background
information describing  the  development of their technology
which has not been  repeated here.
     Response;  The  EPA has met previously with the commenter
to discuss their requests  for modifying appendices A and C of
the HON to allow anaerobic  wastewater treatment as an approved
treatment technology.   These discussions occurred as part of a
broader wastewater  issue that goes  beyond the thermoplastics
rule.  The EPA and  the  commenter agreed to continue
discussions on this topic.
      For  the  final  thermoplastics rule, anaerobic wastewater
treatment may be used as a compliance option if it follows the
wastewater provisions of §63.508  [§63.1330] of the final rule,
if the required  emission reduction  is achieved, and if
compliance  is demonstrated using the procedures in Appendix  C
of the HON.
                              7-10

-------
     Comment;  Two commenters (IV-D-11, IV-D-25) asserted that
the EPA substantially overestimated the emissions from
wastewater streams.  The commenters (IV-D-11, IV-D-25) noted
that at a meeting in August of 1993, the EPA estimated that
75 to 85 percent of the VOHAPs (e.g.,  acetaldehyde,
1,4-dioxane, and methanol) discharged to wastewater treatment
volatilized.  However, the commenters  (IV-D-11, IV-D-25) said
that in a  modeling study conducted by a CMA member company
using EPA approved models (i.e.,  WATERS and TOXCHEM) to
simulate maximum emissions (i.e., high temperature,  high
ventilation conditions) only 8 to 18 percent of the VOHAPs
volatilized.  Therefore, the commenters  (IV-D-11, IV-D-25)
concluded, the EPA has overstated emissions and emission
reductions from thermoplastic wastewater streams.
     Response:  As stated in a memorandum contained in the
SID, the estimation of baseline emissions.for wastewater
relied heavily on industry provided estimates.  Industry
provided data were not used when EPA estimates, made as part
of evaluating the application of controls, did not agree with
industry provided data; in these cases, the EPA estimates were
used.  The EPA estimates were made using the methodology found
in the Background Information Document  (BID) to the HON
 [EPA-453/D-92-.016b; November 1992] .  Wastewater stream
emissions were estimated based on flow rate, organic HAP
concentration, and Fe values presented in the final HON rule.
 [Note:  Fe values reflect the percent of a compound that will
be emitted to the atmosphere if the wastewater stream is left
exposed to the atmosphere long enough.]
     Commenters stated their concern that at a 1993 meeting
the EPA believed 75 to 85 percent of the VOHAPs discharged to
wastewater treatment volatilized.  The EPA estimates of
baseline emissions used Fe values provided in the final HON
rule.  For the chemicals of concern, the Fe values are as
follows:  0.18 for 1,4-dioxane; 0.17 for methanol; 0.48 for
acetaldehyde; and  0.43 for acrylonitrile.  These values are
more in agreement with the commenters cited estimates of 8 to
18 percent overall than were the 75 to 85 percent estimates
                             7-11

-------
from the 1993 meeting.   [Note:  In the majority of cases,
emissions provided by individual facilities were used in
developing the total baseline emissions.]
     Comment:  One commenter  (IV-D-09) suggested that proposed
§63.508(a)(9) of the wastewater provisions reference the
equipment' leak inspection provisions of the HON
(i.e.,  §63.148) directly, instead of referencing §63.509 which
in turn references these provisions.
     Response;  The EPA reviewed §63.508(a) in its entirety
and determined that proposed paragraph §63.508(a)(9) was not
required because it did not identify any differences between
the thermoplastics rule and the HON wastewater provisions.
Therefore, proposed paragraph §63.508(a)(9) has been deleted
from the final rule.  Under the final rule, an affected  source
is still required to  comply with the requirements of §63.148,
as specified in §63.508(a)  [§63.1330(a)].
                              7-12

-------
                       8.0  EQUIPMENT LEAKS

8.1  SELECTION OF THE STANDARDS/APPLICABILITY
     Commenters disagreed with the EPA's decision to require
compliance with subpart H for equipment leaks from all
thermoplastic manufacturing facilities  (IV-D-11, IV-D-25),. PET
facilities  (IV-D-16, IV-D-17, IV-D-19, IV-D-20, IV-D-21,
IV-D-22), MBS facilities  (IV-D-23), and polystyrene facilities
(IV-D-07, IV-D-08, IV-D-15, IV-D-23).  Reasonp the commenters
provided are summarized below.  Commenters  (IV-D-08, IV-D-11,
IV-D-15, IV-D-17, IV-D-20, IV-D-23, IV-D-25) stated that
subpart H was improperly selected as the required leak detection
and repair requirement.  Some commenters recommended that the
proposed equipment leak standards in §63.509 be withdrawn for:
all PET processes  (IV-D-25); PET, TPA continuous processes
(IV-D-17, IV-D-20, IV-D-22); MBS processes  (IV-D-23); and
polystyrene processes  (IV-D-07, IV-D-23).
8.1.1  Polystyrene and Methyl Methacrylate Butadiene Stvrene
     Comment:  Commenters  (IV-D-08, IV-D-11, IV-D-15, IV-D-23)
stated that the EPA has grossly overestimated the emissions and
potential emission reductions for equipment in styrene service by
using the average SOCMI emission factors.  Commenters (IV-D-23,
IV-D-15, IV-D-08, IV-D-11) stated that the use of the average
SOCMI emission factors resulted in an unrealistically low
estimate of the cost per ton of emission reductions and led to
the erroneous conclusion by the EPA that imposition of the
Subpart H equipment leak standards is cost effective for
equipment in styrene service.
     One commenter (IV-D-23) referred to the fact that the EPA
had determined that the subpart H equipment leak standards are
more stringent than the MACT floor.  This commenter (IV-D-23)
                               8-1

-------
pointed out that in such instances where the standards are more
stringent than the MACT floor, they must be justified on a cost
impact basis  (i.e., cost per ton of pollutant removed).  One
cpmmenter  (IV-D-11) stated that the overstatement of uncontrolled
emissions immediately led the EPA to conclude that a stringent
equipment leak program was justified in the proposed standards.
     Commenters  (IV-D-08, IV-D-11, IV-D-23) stated that by using
the average SOCMI  emission factors, which are based on data in
which a certain  fraction of components have leak rates greater
than 10,000 ppmv,  the EPA failed to take into account that the
vapor pressure of  styrene is  such that a leak at 10,000 ppmv is
impossible under normal operating temperatures.  One commenter
 (IV-D-23)  continued that the  concentration of styrene can not
reach 10,000  ppmv  until the temperature is approximately 84° F.
The commenter (IV-D-23) stated  that because styrene polymerizes
more readily as  temperature increases,  styrene  storage is
 frequently refrigerated and styrene is  kept as  cool as possible
prior to  polymerization.   The commenter (IV-D-23)  stated that
 normal  styrene storage and pumping temperature  would  rarely
 exceed  the maximum monthly average ambient  temperature, which  is
 less  than 80° F in virtually  the  entire United  States.  If it  is
 necessary to use equipment leak emission  factors to estimate
 emissions,  it would be more appropriate to use  the SOCMI
 Screening Value Range Emission Factors for components with a
 screening range <10,000 ppmv, stated the  commenter (IV-D-23).
 The commenter (IV-D-23)  stated that  use of these factors  from
 Table 2-4 of the June, 1993 EPA protocol  document would result in
 emission estimates from polystyrene  facilities  that  are
 96 percent lower  for valves and connectors and 91 percent lower
 for pumps than  estimates based on the average SOCMI  emission
 factors.   The commenter (IV-D-23) stated that any projected
 emission reduction would decrease by similar percentages.
      Given that these factors would significantly reduce  the
 estimated emissions and corresponding emission reductions,  the
 commenter  (IV-D-23) stated that the cost effectiveness of any
                                8-2

-------
emission reductions would be 10 to 25 times higher than
determined for the proposed standards.  The commenter (IV-D-23)
stated that the EPA must recalculate the cost effectiveness of
emission reductions using more appropriate emissions estimates
and reanalyze the cost impact.  Evaluation of more accurate cost
effectiveness values, according to the commenter (IV-D-23), will
likely result in a conclusion that the subpart H equipment leak
standards are not justified.  The commenter  (IV-D-23) recommended
that equipment in styrene service be excluded from any equipment
leak standards that are promulgated.  Alternately,  the commenter
(IV-D-23) requested that a leak detection and repair (LDAR)
program based on visual evidence of leaks be implemented for
equipment in styrene service.
     Two commenters  (IV-D-15, IV-D-08) stated that the EPA
inappropriately used the average SOCMI emission factors to
estimate fugitive emissions from polystyrene facilities and
ignored actual data provided by industry.  Another commenter
(IV-D-23) maintained that the most appropriate basis of emission
estimates is the data submitted by industry either in response to
the Section 114 questionnaires or as comments to the proposed
standards.  A previous commenter  (IV-D-15) maintained that the
EPA could not have projected the quantity of emission reductions
presented in the proposal material if the EPA had used the actual
data provided by industry.
     Two commenters  (IV-D-08, IV-D-15) requested that the EPA
adopt a visual leak detection program or, as an acceptable
alternative, an equivalent State LDAR monitoring program.
According to the commenters, only one of these two programs can
be justified for equipment in styrene service.
     Another commenter  (IV-D-07) strongly supported deleting the
requirements for LDAR programs for polystyrene producers.  This
commenter explained that two of their polystyrene production
facilities have implemented LDAR programs for pumps, valves, and
sample connection  systems under State RACT rules.  The commenter
(IV-D-07) stated that their experience shows that these programs
                               8-3

-------
have little or no environmental benefit because there were
essentially no emissions to begin with.
     Response:  The Clean Air Act requires the EPA to set
standards that are at least as stringent as the MACT floor.  In
evaluating equipment leak programs at existing facilities and
their impact in the analysis that was the basis for the proposed
standards, the EPA did not formally determine a MACT floor.
Instead, the EPA determined that the subpart H equipment leak
requirements were more stringent or equivalent to the equipment
leak programs at existing facilities and evaluated the impact of
the subpart ,H requirements on each existing facility given
information on their components and current equipment leak
programs.
     In re-evaluating the analysis, the EPA determined a MACT
floor for each styrene-based subcategory.  These MACT floor
programs are summarized in Table 8-1.  As seen in Table 8-1, the
MACT floor for each subcategory, except ASA/AMSAN, is some form
of equipment leak program.  Therefore, the EPA cannot provide a
blanket exemption for styrene-based subcategories because, at
minimum, the EPA must set standards that are at least as
stringent as the MACT floor.
     In addition to defining MACT floors for each subcategory,
the EPA re-evaluated the emission estimates, costs,  and cost
effectiveness of the proposed equipment leak standards for the
styrene-based subcategories.  (The results of the re-evaluation
are contained in Item IV-B-02, Docket A-92-45.)   With regards to
re-evaluating emission estimates, the EPA agrees with the
commenters that the average SOCMI emission factors are likely to
overestimate emissions from components containing or contacting
styrene.  Therefore, the EPA lowered the emission factors used to
estimate emissions from these components.  The resulting emission
factors for these components used by the EPA in this
re-evaluation are much closer to, but are not identical to,  the
SOCMI Screening Value Range Emission Factors for components with
                               8-4

-------
O

i
Cu

I
I
00
S
a

E
       o
       en
       s
      g
      •°«i
        'It
         as
       ^
       II
       i
               •n
               53 

a

         1
                                            SO
                                            —<
                                            3
                                                  « 0
                                                  es «-«
                 I
                 O)
                                         U
                 -
               « 53

               u
               w
               Q
               2

                                          I

                                                  rt
                                                  a
                                                  I
                                                  I
I

                                                        i
        «o

        3
      I
                                                               
-------


8
S
E
gpi

cog

c/a
i*
O
CO
2
M
O
U
o
§
To

o< ^*S
P J5 2,
J J a

w « Ha
fJKlr
~
i £"!
*""* S &*
t" f"5 »^^
> "-
pi
i°a
"c3 M
>


J3^
, c*
*p eo
co 5J
S

0
i
ts
0>
•a >
33 u
g go
u g.



.
o
Ul ,_
'G *S
M «*^
PQ S
o
to"
U
1.1
tS S3
0 .2
5 "S §
Z •« S
§
a

w §
S 2

§
a
CO
3
0
j_-.j3
< o
co U
     co u
1
g
I


8   ll
C   o e
             S
             •8    "8
             •a    2
                  I    8
                  *S    a
                  1
             pW    Vrf
             CT   Q
             
-------
 a screening range  <10,000  ppmv.   Table  8-2  summarizes  the  three
 sets  of emissions  factors.   Except  for  open-ended lines, the  EPA
 did not use the  "<10,000 ppmv" emission factors directly because
 information provided by the  commenters  and  information previously
 submitted by the industry  shows  that  some components containing
 or contacting styrene have leak  readings of greater than
 10,000  ppmv.
      With regards  to the costing of the equipment  leak programs,
 the EPA discovered several errors in  the costing program.  Most
 of the  errors were associated with the  cost for pumps  in light
 liquid  service.  The net effect  of correcting these costing
 errors  was to lower the estimated cost  of all equipment leak
 programs,  both those currently used by  the  facilities  and subpart
 H's equipment leak program.  In  general, the incremental
 differences decreased as well, but not  as much as  the  overall
 cost  estimates.
      As noted above,  the EPA determined the MACT floor for each
 styrene-based subcategory  (see Table  8-1) .  At a minimum,  each
 subcategory would  be required to  implement  these MACT  floors  (or
 their equivalent).   The HON  equipment leak  standards are more
 stringent  than each of these MACT floors.  To determine whether
 or not  subpart H's  equipment leak program was cost effective,  the
EPA examined  the costs and emission reductions of subpart  H's
equipment  leak program above each of these MACT floors. Table
8-3 summarizes the  results of this analysis.
     The primary ways subpart H is more stringent than the  MACT
floor programs are:
          a leak definition of 500 ppm instead of 10,000 ppmv for
          valves in gas/vapor and'in light liquid service
          a leak definition for pumps  of 5,000 ppmv in phase  II
          and phase  III of the program instead of  10,000 ppmv
          a leak definition of 500 ppmv instead of 10,000 ppmv
          for connectors
          requiring control of sampling connections
                               8-7

-------
                            TABLE 8-2

        COMPARISON OF EMISSION FACTORS (lbs/hr/component)
                 FOR STYRENE SERVICE COMPONENTS
COMPONENT
Valve, light liquid
Pumps, light liquid
Open-ended lines,
styrene
Sampling connections,
styrene
Flanges, light liquid
EMISSION FACTORS (lbs/hr/ component)
Average
Uncont rol led
SOCMI
0.00887
0.0438
0.00374
0.033
0.00403
Revised
Uncontrolled
0.000558
0.005587
0.0033
0.0033a
0.00043
<10,000
ppmv
SOCMI
0.00036
0.00412
0.0033
NA
0.00018
a Assumed to be same as for open-ended lines when taking a
  sample.
                                8-8

-------
                            TABLE 8-3
     SUMMARY OF THE AVERAGE  INCREMENTAL COST EFFECTIVENESS  OF
                   THE HON FROM  THE MACT FLOORS
                   ($/Mg of  Emission  Reduction)
Subcategory
ABS, Batch Emulsion
ABS, Continuous Emulsion
ABS, Continuous Mass
ABS, Latex
ABS, Suspension
ASA/AMSAN
EPS
MABS
MBS
PS, Batch
PS, Continuous
SAN, Batch
SAN, Continuous
No. of
Facilities
3
1
5
1
2
1
6
1
3
11
20
2
3
Average Incremental
Cost Effectiveness
(Subpart H - MACT
Floor)
-4,157
-1,972
-1505
-421
-2,963
511
-78
-936
-1,495
-291
-1,573
-3, 981
-743
     Based on the incremental differences between the MACT floor
programs and the proposed equipment leak standards, the EPA
determined that the subpart H equipment leak standards are cost
effective for each styrene-based subcategory.  Therefore, the EPA
has retained the equipment leak standards as proposed in the
final standards.
     Comment;  One commenter (IV-D-08) provided leak data from
three polystyrene plants.  The commenter (IV-D-08) stated that
these facility monitoring data do not support the EPA's protocol
for estimating the control effectiveness of a LDAR program.  The
                               8-9

-------
commenter (IV-D-08) stated that the EPA's protocol identifies
three important factors in determining control effectiveness:
(1)  how a "leak is defined, (2) the initial leak frequency before
the LDAR program is implemented, and (3) the final leak frequency
after the LDAR program is implemented.   The commenter (IV-D-08)
stated that the provided data clearly demonstrates that changing
the leak definition from 10,000 ppmv to 500 to 1,000 ppmv does
not result in a significant increase in the number of leaking
components.  The commenter  (IV-D-08) continued that this is
contrary to the basic assumptions used by the EPA to estimate
control effectiveness of LDAR programs.  Because of this faulty
premise, the commenter  (IV-D-08) stated that the proposed rule
will not result in a significant reduction in estimated
emissions.
     Response;  As noted in the previous response, the EPA
re_evaluated the emission  factors used to estimate emission
reductions.  The revised emission factors reduce estimated
emissions from most polystyrene plants  (i.e., batch, continuous,
and expandable polystyrene) by an order of magnitude.  In
developing the revised  emission factors, the EPA took into
account much lower leak frequencies in estimating the emission
factors for the various equipment leak programs used by the
industry, for the  MACT  floors, and  for the HON program than were
used previously.   For example, the  initial and final leak
frequencies used in the re-evaluation for valves in light liquid
service  (i.e., styrene  service) are, respectively, 0.3 and
0.05 percent  for a monthly leak detection and repair program with
a leak definition  of 10,000 ppmv.   Previously, these leak
frequencies were 4.34 and  0.54, respectively.  The effect of
these changes in leak frequencies is to lower the emission
factors used  to estimate each  equipment leak program, which in
turn lowers the magnitude  of the estimated emission reductions
between these programs.
     The  estimated leak frequencies used in the re-evaluation
were based on and  are  consistent with  the data presented by the
                               8-10

-------
commenter and on information provided previously by the industry.
For example, the data presented by the commenter shows that there
is an increase in the number of leaking components when the leak
definition is changed from 10,000 ppmv to 500 to 1,000 ppmv.  For
Plant No. 1, the number of components identified as leaking in
1993 increased from 2 (at a leak definition of 10,000 ppmv) to 4
 (at a leak definition of 500 to 1,000 ppmv), which corresponds to
a leak rate increase from a 0.35 to 0.5 percent, and from 1 to 40
 (an increase in the leak rate from 0.12 to 0.5 percent) in 1994.
For Plant No. 2, the number of components identified as leaking
increased 2 to 16  (an increase in the leak rate from 0.3 to
2.7 percent) in 1993 and from zero to three  (an increase in the
leak rate from 0 to 0.5 percent) in 1994.
     In summary, the EPA took into account data presented
specific to three polystyrene subcategories on indicated leak
rates and used that information in estimating the emission
factors and the resulting effectiveness of the various equipment
leak programs.  The results were then incorporated into the cost
and cost-effectiveness analysis, which still showed the proposed
program to be cost effective.
     Comment:  One commenter  (IV-D-09) supported proposed
§63.509(a)(1) which clearly defined when a loss of liquid from a
process was not a "leak."  One commenter (IV-D-08)  contended that
the exemption [proposed §63.509 (a)-(l) ] available to continuous
polystyrene manufacturers who use bleed ports on agitator and
pump seals should also be made available to batch and expandable
polystyrene plants that use this type of equipment.  The
commenter (IV-D-08) stated that there are no seals that can
handle this service without a self-lubricating bleed port.   The
commenter (IV-D-08) stated that the lack of alternative pumps and
seals is exactly why continuous process bleed ports were exempt
from the requirement of counting dripping liquids from bleed
ports as a leak.
     Response;  The EPA agrees with the commenter and has
extended this exemption to all polystyrene subcategories.
                               8-11

-------
8.1.2  Poly(ethvlene tereohthalate)
     Overall Objection to LDAR for -PET
     Comment;  Commenters had several objections concerning the
proposed provisions as applied to PET affected sources.
Commenters felt that emissions and emission reductions were
overestimated; that little environmental benefit could be
expected as  a result of implementing a LDAR program; that the
proposed provisions were not cost effective (largely due to the
overestimation of emissions and emission reductions) ; and that
the recordkeeping and reporting requirements were excessive.
     Commenters  (IV-D-16, IV-D-17) stated that emissions and
emission reductions from PET affected sources had been
overestimated.  Commenters  (IV-D-16, IV-D-17,  IV-D-25) also
stated that  this overestimation biased the EPA's cost analysis
and made the implementation of an equipment leak program appear
to be cost effective.  One commenter  (IV-D-17) stated that the
EPA's estimate of annual costs of  $744,000 for PET terephthalic
acid  (TPA) facilities using a continuous process are excessive
considering  that the equipment leak standards will result in
essentially  no emission reductions.
     One commenter  (IV-D-16) stated that the EPA had grossly
exaggerated  the emissions from PET affected sources using a TPA
process to justify the cost effectiveness of an equipment leak
program for  this subcategory.  The commenter  (IV-D-16) felt that
there may be some value in equipment leak programs for PET DMT
affected sources due to the presence of methanol and its
relatively high vapor pressure.  However, the commenter  (IV-D-16)
felt it did  not make any sense to  impose similar requirements on
processes where the only HAP present at a concentration greater
than five percent is ethylene glycol, which has a relatively low
vapor pressure.  The commenter  (IV-D-16) felt that an equipment
leak program designed to control HAP would not control a
significant  amount of emissions  for processes with ethylene
glycol.   [Note:  In reference to  the previous sentence, the EPA
assumes the  commenter is  implying  HAP that are more volatile than
                               8-12

-------
ethylene glycol.]  The commenter  (IV-D-16) challenged the EPA to
re-examine the HAP emissions and emission reductions from PET TPA
affected sources and base its assumptions on credible data.
     Another commenter (IV-D-17) specifically claimed that the
technical analysis is flawed because it does not accurately
reflect the emissions and emission reductions associated with
processes that operate in heavy liquid service and, as a result,
the EPA's emission and emission reductions estimates are
overstated, and the resulting cost per ton of control estimates
are substantially underestimated.
     In reaching the conclusion that emissions and emission
reductions were overstated, the commenter (IV-D-17) noted that
essentially all of the components potentially subject to the
equipment leak standards are in ethylene glycol service at
ambient temperature  (i.e., heavy liquid service).  Further,
stated the commenter  (IV-D-17), those components not in heavy
liquid service (i.e., in light liquid or vapor service)  are in
vacuum service, and the proposed equipment leak standards
specifically exempt components in vacuum service.  The commenter
(IV-D-17) also referred to the methodology used to estimate
emissions from connectors.  The commenter (IV-D-17) pointed out
that the emission factor used to estimate emissions for
connectors is for a component profile that includes some
connectors in each type of service (i.e., gas/vapor, light
liquid, and heavy liquid) and claimed that a more realistic
factor should be used for connectors in heavy liquid service.
The commenter  (IV-D-17) suggested that a factor of
0.00023 lbs/hr/component be used rather than the
0.00403 Ibs/hr/component for all services.  (This lower factor
was derived by the commenter from the factors for uncontrolled
valves and connectors in light liquid service.  For connectors in
light liquid service, the connector leak rate is 45 percent of
that for valves in light liquid service.  The commenter (IV-D-17)
applied the same ratio to the leak rate for valves in heavy
liquid service to derive a factor for connectors in heavy liquid
                               8-13

-------
service.)  The commenter  (IV-D-17) then noted that the resulting
emission factor of 0.00023 Ibs/hr/component is less than the
emission factor of 0.000286 Ibs/hr/component for connectors under
a subpart H-level equipment leak program.
     The commenter (IV-D-17) also referred to the 2,030 Mg/yr of
emission reductions  (23 percent of the baseline emissions)
estimated for the PET TPA subcategory.  The commenter  (IV-D-17)
stated that this estimate of emission reduction is inconsistent
with the approximately 90 percent reduction that would be derived
from comparing the emission factor for connectors under a subpart
H-level equipment leak program with the uncontrolled factor for
connectors they had  derived.  This inconsistency suggests,
according to the commenter  (IV-D-17) , that the analysis is flawed
and/or poorly conceived.  The commenter  (IV-D-17) continued that,
regardless, the argument  presented herein clearly limits
potential emission reductions associated with the application of
an equipment leak program for components in heavy liquid service
to inconsequential amounts.
     Finally, the commenter  (IV-D-17) noted that, in the
Supplemental Information  Document, the EPA recognizes  that
equipment leak programs are ineffective in reducing emissions
from valves and pumps in  heavy liquid service.  The commenter
 (IV-D-17) based their observation on the fact that the emission
factors  for these components reflecting an equipment leak program
are not  lower than the uncontrolled emission factors.  Therefore,
according to the commenter  (IV-D-17) , no substantive emission
reductions will result from the proposed equipment leak
standards.
     Another commenter  (IV-D-25)  felt that implementation of an
equipment leak program on PET affected sources would not result
in any significant environmental  benefit.  In support  of this
assertion, the commenter  (IV-D-25)  stated that, for two of their
PET DMT  continuous affected sources, approximately 220 components
are in light liquid  (i.e.,  methanol) service at one affected
source and approximately  570  are  in light liquid  (i.e., methanol)
                               8-14

-------
 service at the other affected source.   The commenter (IV-D-23)
 stated that the total expected emissions at these affected
 sources from these light liquid service components are 0.3 Ibs/hr
 (1.2 Mg/yr) and 1 Ib/hr (4 Mg/yr),  respectively.   The commenter
 (IV-D-25)  felt that little environmental benefit  could be
 realized by controlling these emissions.
      One commenter (IV-D-17)  stated that subpart  H,  §63.169,
 requires nothing more than what is  currently being practiced by
 the PET TPA continuous industry and,  therefore, the MACT floor
 determination is flawed for this source category.
      One commenter (IV-D-17)  stated that the "repair upon
 detection" requirement in §63.169 of subpart H is currently being
 practiced to a great extent by all  PET TPA affected sources.  The
 commenter (IV-D-17)  stated that subpart H recordkeeping and
 reporting requirements (§§63.181 and 63.182)  are  excessive
 because §63.169 requires nothing more  than what is currently
 being practiced by the PET TPA continuous industry.   Another
 commenter (IV-D-25)  stated that the proposed equipment  leak
 standards  impose a recordkeeping and reporting burden on PET
 sources that is unjustified.
      One commenter (IV-D-17)  concluded by stating that  the cost
 effectiveness of the proposed equipment leak standards  is
 unacceptable because (1)  emissions  from PET TPA affected sources
 have  been  greatly overestimated,  (2) nothing more  than  current
 industry practice is required by the heavy liquid  standards that
 cover essentially all  of this subcategory,  (3) no  substantive
 reduction  in HAP emissions can be expected from implementation of
 the proposed equipment  leak standards,  and (4) there  are
 substantial costs associated  with the  proposed recordkeeping and
 reporting  requirements.
      Two commenters  (IV-D-17,  IV-D-25)  stated that the  final
 standards  should exempt  all PET processes  from the equipment leak
provisions  because the regulatory analysis  is technically flawed.
One commenter (IV-D-17)  further stated that such an exemption
would be consistent with the  decision  made  in the development of
                               8-15

-------
the Polymer Manufacturing NSPS.  Another comraenter (IV-D-25)
stated that by exempting PET affected sources from the equipment
leak standards, the EPA would be in keeping with current
"reinventing government" initiatives to reduce unnecessary
recordkeeping and reporting burdens.
     Another commenter  (IV-D-22) suggested that the equipment
leak standards be amended to exempt PET TPA affected sources.
According to the commenter  (IV-D-22), such an exemption would
avert the use of resources on  recordkeeping for sources that have
no measurable emissions and allow those resources to be used on
process improvement to manage  the known emissions from the
process.  The commenter  (IV-D-22) specifically referred to
equipment in ethylene glycol service at PET TPA affected sources,
noting that such equipment  is  in heavy liquid service and will
have extremely  low  emissions.   The  commenter  (IV-D-22) stated
that past tests for ethylene glycol emissions have found no
detectable  emissions.  Thus, according to the commenter
 (IV-D-22),  complying  with the  proposed equipment  leak standards
would be an audit  and paper management effort.
     Response;  As  described below, the EPA re-evaluated the
emission estimates,  costs,  and cost effectiveness of  the proposed
equipment  leak standards for each  PET  subcategory.  The results
of the  re-evaluation are found in  Item IV-B-03, Docket A-92-45.
      Emission
Estimates.   Commenters felt that the EPA had
 overstated the emissions and emission reductions for equipment
 leaks from PET affected sources by using the average SOCMI
 emission factors.  Commenters felt that this was especially true
 for PET TPA facilities, where many of the components are in heavy
 liquid service.  One commenter highlighted the inap'propriateness
 of the emission factor for connectors in heavy liquid service.
      The EPA acknowledges that it is possible that emissions and
 emission reductions can be overstated whenever generic emission
 factors  (in this case, average SOCMI. emission factors) are used.
 The EPA used the average SOCMI emission factors because emissions
 estimates provided by  the industry were not universally available
                                8-16

-------
for the population of affected sources, some estimates that were
provided were not supported with a description of the estimation
methodology, and the EPA believes the average SOCMI emission
factors, while not ideal, are applicable to this industry.
Further, while industry has criticized the accuracy of the
average SOCMI emission factors as part of recent rulemakings,
data have not been provided to enable the EPA to develop a new
set of industry-specific emission factors.
     The EPA, nevertheless, revised the estimation of emissions
and emission reductions by using different emission factors for
several components.  Table 8-4 compares the emission factors used
to estimate uncontrolled emissions in the original analysis  (see
the column titled "Average Uncontrolled SOCMI") and in the
re_evaluation  (see the column titled "Revised Uncontrolled").  As
can be seen in Table 8-4, significantly lower emission factors
were used for most of the components in heavy liquid service.
For example, the emission factor used for flanges in heavy liquid
service decreased from 0.00403 Ibs/hr/component to
0.00018 Ibs/hr/component.  This latter emission factor is lower
than the one suggested by one commenter  (IV-D-17).  The use of
lower emission factors for components in heavy liquid service
should address the commenters' concerns about estimation of
emissions from these components at PET facilities.  As seen in
Table 8-5, using the revised emission factors, estimated
emissions from PET facilities decreased from 5,460 Mg/yr to
1,930 Mg/yr.   In addition, the use of revised emission factors
also decreased the estimated emission reduction for PET
facilities, from 1,790 Mg/yr to 1,136 Mg/yr  (see Table 8-5).
                               8-17

-------
                    TABLE 8-4
COMPARISON OF EMISSION FACTORS  (Ibs/hr/component)
         FOR COMPONENTS AT PET FACILITIES
Component
t
Valve , gas /vapor
Valve, light liquid
Valve, heavy liquid
Pump, light liquid
Pump, heavy liquid
Pressure relief device, gas
vapor
Pressure relief device,
light liquid
Pressure relief device,
heavy liquid
Open-ended line, gas/vapor
and light liquid
Open-ended line, heavy
liquid
Compressors
i
Sampling connection,
gas/vapor and light liquid
Sampling connection, heavy
liquid
Flange, gas/vapor and light
liquid
Flanges, heavy liquid
Emission Factors
( Ibs/hr /component )
Average
Uncontrolled
SOCMI
0.0131
0.00887
0.000506
0.0438
0.0190
0.2288
. 0.2288
0.2288
0.00374
0.00374
0.5016
0.033
0.033
0.00403
0.00403
Revised
Uncontrolled
0.0131
0.00887
0.000506
0.0438
0.00463
0.2288
0.09852
0.09852
0.00374
0.0033
0.5016
0.033
0.015
0.00403
0.00018
                       8-18

-------
                            TABLE 8-5

         COMPARISON OF EMISSIONS AND EMISSION REDUCTIONS
Subcategory
TPA, Continuous
TPA, Batch
DMT , Cont inuous
DMT, Batch
TOTALS
Emissions
(Mg/yr)
Previous
2,030a
90
2,150
1,190
5,460
Revised
345b
22
1,034
530
1,931
Emission Reductions
(Mg/yr)
Previous
473a
5.6
903
408
1,790
Revised
113b
2.7
704
316
1,136
a  Includes the Allied Signal facility.
b  Does not include the Allied Signal facility.

     It should be pointed out that, except for sampling
connections, no emission reduction credit is taken for components
in heavy liquid service.  Therefore, in contrast to the
commenters' concern, the predominance of components in heavy
liquid service does not contribute to an overstatement of
emission reductions.
     One commenter  (IV-D-17) claimed that the previous analysis
was flawed, in part, because the estimated emission reduction fox
the PET TPA subcategory from baseline was 23 percent, while
comparing uncontrolled and subpart H-level emission factors for
connectors shows a 90 percent reduction.  The commenter is
incorrect in comparing these two percent,reductions.   The PET TPA
subcategory emission reduction is based on percent reductions fox-
all components, not just connectors.  Many of the other
components have lower expected emission reductions.  In
particular, many of the components in the PET TPA subcategory are
in heavy liquid service.  Except for sampling connections, no
emission .reduction credits were taken for these components.
Thus, a large percentage of the components in the PET TPA
subcategory have a 0 percent reduction associated with them.  The
resulting overall emission reduction percentage is, therefore,
                               8-19

-------
significantly lower than the emission reduction percentage for
connectors alone.  In summary, the example used by the commenter
does not indicate a flaw in the analysis.
     Commenter IV-D-17 also noted that, because the emission
factors for components in heavy liquid service did not change
between the uncontrolled scenario and the controlled scenario, no
substantive emission reductions will result from the proposed
equipment leak standards.  The EPA disagrees with the commenter's
conclusion for two reasons.  First, based on the information
provided by the  industry, there are components in gas/vapor
service and in light liquid service at PET TPA facilities as well
as at other types of PET facilities.  There are sufficient
numbers of components in these other types of service to provide
the potential for substantial emission reductions.  Second, the
use of the same  emission factor does not reflect the belief by
the EPA that the equipment leak requirements for components in
heavy liquid service achieve no emission reduction, but rather
reflects the inability to quantify the emission reduction
achieved by the  work practice standard for these components.  The
EPA maintains that leaks from components in heavy liquid service
emit pollutants  to the air and that their repair achieves
emission reductions.  Therefore,  for these reasons, the EPA
disagrees with the commenter  that equipment leak standards
applied to PET TPA facilities will not achieve substantive
emission reductions.
     Another commenter  (IV-D-25)  felt  that an equipment leak
program would not achieve  in  any  significant environmental
benefit and provided emission estimates  from equipment leaks  at
two of their facilities.  .The EPA agrees that the impact of an
equipment  leak program may be relatively small for certain
individual facilities, because the  size  of the impact will vary
depending  on the number  of components  at the facility and the
type of service  they are in.  Nevertheless, based on all of the
information provided by  industry, the  EPA continues to believe
that the equipment  leak  standards will have a substantive
                               8-20

-------
 cumulative  impact on reducing emissions from equipment leaks at
 PET  facilities.
      Costs  and Cost Effectiveness.  In response to comments, the
 EPA  re-evaluated the costs and  cost effectiveness of the proposed
 equipment leak standards.  In the original analysis, the EPA
 evaluated the costs and  cost effectiveness of the proposed
 equipment leak standards for all PET facilities, both those that
 provided component profiles  (counts and types of service) and
 those that  did not.  For the latter, estimates of components
 counts and  service type  were extrapolated from the information on
 those facilities that provided  specific component profiles.  In
 the  re-evaluation of costs and  cost-effectiveness, the EPA
 limited the evaluation of costs and cost-effectiveness of the
 proposed equipment leak  standards to only those facilities that
 provided components profiles for determining whether or not to
 implement the proposed equipment leak standards.  The results of
 this re-evaluation are presented in Table 8-6.

                            TABLE 8-6
            COST EFFECTIVENESS RESULTS OF RE-EVALUATION
                        FOR PET FACILITIES
Sub-
category
TPA,
Continuous*3
DMT,
Continuous
DMT, Batch
Number of
Facilities3
5
3
4
Emission
Reductions,
Mg/yr
49.8
300
198
Annual
Costs,
$
75,913
98,050
121,000
Cost
Effectiveness,
$/Mg
1,524
327
611
a These numbers reflect the number of facilities that provided
  facility-specific information on the number of components and
  their type of service.   It does not reflect the total number of
  facilities in the subcategory.
b Does not include the Allied Signal facility.

     Based on the re-evaluation, the EPA found that the cost
effectiveness of the proposed equipment leak standards varied
                               8-21

-------
between affected sources.  However, the EPA looks at the cost
effectiveness values across all of the facilities for which cost
effectiveness values were estimated to determine overall
reasonableness and not just the impact on any one facility.
Using industry provided data, the re-evaluation continues to show
that, taken on an overall basis, the cost effectiveness of the
proposed equipment leak standards are reasonable for the PET
subcategories  (see response under "Allied Signal" for the one
exception).                                •
     Table 8-7 compares the previous and revised cost estimates
for all of the facilities in each of the four PET subcategories.
For those facilities that did not provide component profiles, the
EPA estimated component counts and types of service based on
information from those facilities that did provide component
profiles.  These estimated counts were then used to estimate
emissions, emission reductions, and costs of control.
     MACT Floor Determination.  Commenter IV-D-17 stated that the
MACT floor.determination, which was considered to be
"uncontrolled," was flawed because the proposed equipment leak
standards only require what PET TPA facilities are currently
doing for components in heavy liquid service.  Based on the
information provided by the industry prior to the receipt of the
comment letters on the proposed rule, there was insufficient
information for the EPA to determine that PET TPA facilities were
conducting these types of inspections.  Therefore, based on the
information available at the time, the EPA considered the MACT
floor to be "uncontrolled."  In light of this commenter's comment
and several other individual commenters, it appears that the MACT
floor for these types of facilities should have included a
visual-based inspection program.  This fact, however, does not
change the conclusion that components in heavy liquid service
should still be subject to this type of program.  As discussed in
the following paragraphs, some form of recordkeeping and
                               8-22

-------
                            TABLE 8-7
           COMPARISON OF COSTS AND COST EFFECTIVENESS
Subcategory
TPA,
Continuous
TPA, Batch
DMT,
Continuous
DMT, Batch
Previous
Analysis
Annual
Costs,
$
744,000a
17,700
988,000
569,000
Re -Analysis
Annual
Costs,
$
184,600b
2,140
251,300
216,900
Emission
Reduction,
Mg/yr
113b
2.7
704
316
Cost
Effectiveness ,
$/Mg
l,620b
792
357
686
a Includes the Allied Signal facility.
b Does not include the Allied .Signal facility.

reporting is still required for equipment leak programs that are
part of a MACT floor.
     Recordkeeriina and Reporting.  Several commenters stated that
the recordkeeping and reporting burden for components in heavy
liquid service were unjustified and costly.  Commenter IV-D-17
was concerned that the proposed equipment leak standards would
not have any emission reduction impact for PET TPA facilities and
yet would cost $744,000  per year  for what amounts to, according
to the commenter, only a recordkeeping and reporting exercise to
document what the facilities  are  already doing.
     The EPA believes that, for some of the commenters, their
concerns over the recordkeeping and reporting requirements of the
proposed rule are the result  of misunderstandings of how the rule
works.  The EPA, based on further contact with commenters, has
been led to believe  that industry believes that  "tagging" and
Method 21 monitoring is  required  for  components  in heavy liquid
service.   ("Tagging"  involves placing a physical tag on the
component to  identify it and its  type of  service (e.g., heavy
liquid) .)   In fact,  tagging and Method 21 monitoring were not
required by the  proposed equipment leak  standards for  components
                                8-23

-------
in heavy liquid service and are not required under the final
equipment leak standards for these components.  In lieu of
tagging, an owner or operator can identify components in heavy
liquid service on a process schematic or similar drawing.
Furthermore, Method 21 monitoring is not required under §63.169,
but is an option.  Alternatively under §63.169,  an owner or
operator can rely, solely on visual, auditory, or olfactory
evidence to determine the presence of a leak and then repair the
leak without ever using Method 21.  The latter alternative
greatly reduces the recordkeeping and reporting requirements
required when Method 21 is used.
     Commenter IV-D-17 was specifically concerned about the costs
of the proposed equipment leak standards as it affects PET TPA
facilities.  The EPA disagrees with the commenter's conclusion.
First, the estimated costs for the PET TPA facilities of
complying with equipment leak program is now estimated to be
$185,000 rather than $744,000.  This reduction in estimated cost
is due in part to the correction of errors in the original
costing program.  This averages to an annual cost of
approximately $17,000 per TPA PET facility.  Second, the EPA
disagrees that there is no emission reduction associated with
this cost.  As noted earlier, the degree of emission reduction
will vary from facility to facility.  Many PET TPA facilities
have reported the presence of components in gas/vapor service and
in light liquid service.  Thus, emission reductions can be
substantive for these facilities.  Third, the majority of these
costs are associated with those portions of the equipment leak
standards that affect components in gas/vapor service and those
in light liquid service.  Fourth, the average annual cost is
$17,000, but for facilities with few components in gas/vapor
service or  in light liquid service, the actual costs of complying
with the equipment leak program will be much less.
     In general, the EPA believes the recordkeeping and reporting
requirements found in subpart H of 40 CFR part 63 for components
in heavy liquid service are reasonable.  However, due to the
preponderance of components in heavy liquid service at PET
                               8-24

-------
facilities, the requirement to provide a list of identification
numbers for each individual heavy liquid component may be unduly
burdensome.  Therefore, the final standards do not require an
owner or operator to provide a list of identification numbers of
those components subject to §63.169  (i.e., pumps, valves,
connectors, and agitators in heavy liquid service; pressure
relief devices in liquid service, and instrumentation systems).
Given this change, these components would be subject to the
following recordkeeping requirements, which are currently
contained in subpart H:
     1.   Subject equipment may be identified on a plant site
          plan, in log entries, or by other appropriate methods;
          physical tagging of equipment to indicate that it is in
          organic HAP service is not required.
     2.   When a leak is detected, the following information is
          required to be kept:
               the equipment identification number and the
               operator name, initials, or identification number;
               the date the leak was detected and the date of
               first attempt to repair the leak;
               the date of successful repair of the leak;
               "repair delayed" and the reason for the delay if
               the leak is not repaired within 15 calendar days
               of its discovery.
     3.   Information, data, and analysis used to determine that
          a piece of equipment or process unit is in heavy liquid
          service.  Such a determination includes an analysis or
          demonstration that the process fluids do not meet the
          criteria of  "in light liquid or gas service."  Examples
          of information that could document this include, but
          are not limited to, records of chemicals purchased for
          the process, analyses of process stream composition,
          engineering calculations, or process knowledge.
     4.   If applicable, identification, either by list or
          location  (area or group), of equipment in organic HAP
                               8-25

-------
          service less than 300 hours per year within a subject
          process unit.
     Especially with the elimination of the requirement to
provide a list of identification numbers for each individual
component in heavy liquid service, the EPA does not find these
requirements burdensome.  A PET facility that is subject to the
equipment leak standards will need to identify the type of
service for each component in order to determine which portion of
the standards are applicable; the EPA does not consider this
requirement to be onerous.  Affected sources using the PET TPA
process should be able to identify easily components in heavy
liquid service through process knowledge and the use of chemical
purchase records.  Further, the provisions do not require owners
or operators to tag each component, but allow identification by
process area or similar, more general means.
     The primary recordkeeping requirements for components in
heavy liquid service  occur when evidence of a leak occurs.  The
information required  is what the  EPA considers to be the minimum
necessary to document leaks at an affected source.
     With regards  to  reporting requirements for components in
heavy liquid service,  there are none other than those associated
with the Notification of  Compliance Status, which requires the
reporting of the number of each equipment type  (including those
in heavy liquid service)  and the  method of compliance.  No
information  is  required to be  reported  on equipment  in heavy
liquid  service  in  subsequent periodic  (semiannual) reports.
The EPA again  does not consider such reporting to be burdensome.
The EPA considers  the reporting of  this information  necessary to .
the enforcement personnel to understand the entire compliance
program intended to be implemented at a facility.
     Recommendations.  Commenters requested that the EPA exempt
 affected  sources within some or all of  the PET  subcategories from
 the proposed equipment leak standards.   One commenter  stated that
 this decision would be consistent with decisions made  in
 developing the Polymer Manufacturing NSPS.  Another  commenter
 stated that exempting all PET affected sources  would keep with
                                8-26

-------
 the reinventing government initiatives to reduce unnecessary
 recordkeeping  and reporting burdens.
     The EPA cannot  exempt the PET subcategories because, as
 alluded to by  the commenters, there is a baseline equipment leak
 program implemented  by the majority of the industry, which would
 serve as the MACT floor.  Further, exempting all PET
 subcategories  would  not be consistent with the Polymer
 Manufacturing  NSPS,  which only exempted facilities using a
 continuous TPA process.  The continuous TPA process was exempted
 because data,  at that  time, showed that all components were in
 heavy liquid service.  Facilities using a continuous DMT process
 were already covered by the SOCMI Equipment Leaks NSPS.  Batch
 processes were not evaluated as part of the Polymer Manufacturing
 NSPS.  Finally,  as discussed earlier, the cost effectiveness of
 implementing the equipment leak standards was found to be
 reasonable.  For these reasons and the others presented earlier,
 the final rule does  not exempt PET affected sources from the
 equipment leak standards, except as discussed in the following
.paragraph and  in the last comment and response in this chapter.
     On the  basis of the comments received, the EPA has
 determined that.,  if  an owner or operator can demonstrate that all
 of the components in a PET TPPU are either in vacuum service or
 in heavy liquid service, it is not reasonable to require
 compliance with the  equipment leak standards for that TPPU.  The
 rationale for  this exclusion is equivalent to the rationale used
 in developing  the Polymer Manufacturing NSPS.  However,  since
 some PET TPA affected  sources have components in gas/vapor or
 light liquid service,  the EPA has determined that it is necessary
 for owners and operators of PET TPPUs exempted under these
 criteria  (i.e.,  vacuum service or heavy liquid service components
 only) to keep  a record that these criteria are met.  As contained
 in the final standards, the determination follows the guidance
 provided in  §63.181 (i) of subpart H.  The EPA recognizes that
 this imposes some burden on some affected sources without being
 able to show any emission reduction.  However, the EPA does not
 believe that the burden is significant because:  (1)  it is a
                               8-27

-------
one-time determination and  (2) most of the information to make
the determination will be readily available.  The EPA further
believes such a requirement is justified to ensure that those
affected sources without components in gas/vapor or light liquid
service are substantiated and identified to the EPA.
     Finally, changes may occur at an exempted PET TPPU such that
all of the components are no longer in either vacuum or heavy
liquid service, but now contains components in gas/vapor service
or in light liquid service.  Should such changes occur, the final
standards requires such PET TPPUs to comply with the equipment
leaks standards for all of  the components in the PET TPPU.
     Comment:  One cotnmenter'  (IV-D-17) referred to that portion
of the preamble for the Polymer Manufacturing NSPS which states
that the control of equipment leak emissions at PET TPA plants
had not been found to be cost effective and that the cost
effectiveness of controlling  fugitive emissions from pumps and
valves in heavy liquid service is greater than $10,000 per ton.
According to the commenter  (IV-D-17), the EPA had not presented
any new information in the  record for the proposed rule about the
methodologies for estimating  equipment leak emissions, control
effectiveness, or control costs that would refute the prior
conclusions  and allow the EPA to conclude that control of these
plants is now reasonable.
     Response:  Based on the  information provided by owners and
operators of PET TPA affected sources, the EPA has learned that
many PET TPA affected sources have components in gas/vapor
service and  in light liquid service.  This  is in marked contrast
to the information available  to the  EPA 10 years ago when it was
developing the Polymer Manufacturing NSPS.  At that time, the
information  available to  the  EPA indicated  that all components at
PET TPA affected  sources were in heavy liquid service, and the
EPA determined that equipment leak standards that affect only
components in heavy liquid service are not  cost effective.
Therefore, for the Polymer Manufacturing NSPS, the EPA decided
not to implement  equipment leak standards for PET TPA  continuous
facilities.  Data provided by the  industry  under this  rulemaking,
                               8-28

-------
however, shows that PET TPA affected sources have components in
gas/vapor service and/or in light liquid service.  The EPA,
therefore, believes that it is reasonable to evaluate the costs
and cost effectiveness of equipment leak standards as they apply
to components in gas/vapor service and in light liquid service at
facilities in this subcategory.
     Comment:  Commenters (IV-D-16, IV-D-21) stated that the
proposed equipment leak standards require the use of Method 21
for emissions monitoring and that Method 21 is inappropriate for
monitoring components in ethylene glycol service.  One commenter
(IV-D-21) noted that Method 21 does not provide a response to
ethylene glycol, which is not a vapor at ambient conditions.  The
commenter  (IV-D-21) suggested that the EPA allow a LDAR approach
that would utilize visual detection.  The commenter (IV-D-21)
then provided a draft description of such a program they have
prepared for use at one of their facilities.  The commenter
(IV-D-21) noted that their visual detection program would not
apply to components with high concentrations of methanol or
acetaldehyde; the intent was to implement the program for
components and lines containing primarily ethylene glycol liquid
or vapor.
     Response:  The EPA agrees with the commenters that Method 21
is inappropriate for detecting leaks from components in ethylene
glycol service if it is also in heavy liquid service.   (The EPA
maintains that Method 21 is appropriate where a component that
contains or contacts ethylene glycol is in gas/vapor service.)
Therefore, the final equipment leak standards eliminate the use
of Method 21 for determining if a leak exists for components in
heavy liquid service at PET affected sources.  Under the final
standards, an owner or operator is required to determine if a
leak exists by using visual, olfactory, audible, or any other
detection method  (other than Method 21) to determine the presence
of indications of potential leaks.  Repair of leaks determined in
this manner is required using procedures equivalent to those
contained in §63.169 (c) and  (d) .
                               8-29

-------
     With regards'to the proposed equipment leak program that
Commenter IV-D-21 submitted, the EPA contacted the commenter (see
Docket A-92-45, Item IV-E-7) and determined that, by clarifying
in the final standards that Method 21 was not required, the
commenter felt that complying with the §63.169 would be
acceptable.  Therefore, the EPA determined that further analysis
of the commenter's program was unnecessary.
     Comment;  One commenter  (IV-D-16) challenged the EPA to cost
justify extending the subpart H equipment leak standards to
unrelated sources that have a significantly different fugitive
HAP profile.
     Response;  Thirteen PET facilities provided specific
component profiles (counts and service type).  The EPA compared
these component profiles with those for the three model SOCMI
facilities used to analyze the impacts of the HON standards.  As
contended by the commenters, the EPA found that PET facilities
have a significantly higher proportion of components in heavy
liquid service than the three SOCMI model plants.  The EPA also
found that'PET facilities have a number of components in
gas/vapor and in light liquid service.  As noted earlier, the EPA
used the specific component profiles from these 13 facilities to
re-evaluate whether or not the implementation of the proposed
equipment leak standards was cost effective.  The re-evaluation
found that the equipment leak program was cost effective.
Therefore, regardless of the different profiles between PET and
SOCMI facilities, the EPA still finds that equipment leak
standards based on subpart H of 40 CFR part 63 is cost effective
for PET facilities even though their component profiles are
different from the component profiles associated with the three
SOCMI model plants.
     Comment:  One commenter  (IV-D-19) stated that affected
sources in the PET continuous TPA high viscosity multiple end
finisher process subcategory  should be exempted from the proposed
equipment leak standards.  The commenter  (IV-D-19) cited the
predominance of components  in heavy liquid service, the low vapor
pressure of ethylene glycol,  the low potential for equipment leak
                               8-30

-------
emissions from equipment in the raw material process section,  and
the vacuum conditions that prevent emissions from equipment in
the polymerization reaction process section.  Furthermore,  the
commenter (IV-D-19) stated that the cost of recordkeeping and
implementing a formal equipment leak program would far outweigh
the benefits of decreasing an already low amount of fugitive
emissions.
     The commenter (IV-D-19) pointed out that the nature of the
process and material are such that essentially all of the
components potentially covered by the proposed equipment leak
standards are in either vacuum service or heavy liquid service.
The commenter (IV-D-19) noted that, under the proposed equipment
leak standards,  components in vacuum service are exempt and
components in heavy liquid service are required to be monitored
within 5 calendar days of detecting a potential leak by the
method specified in §63.180(b).
     Addressing the requirement to monitor for emissions using
Method 21, the commenter  (IV-D-19) noted that equipment leaks
from components in ethylene glycol service would be detected as
dripping liquid, not as vapor leaks to the atmosphere, and that
overhead pipes are routinely inspected to detect leaks.  The
commenter (IV-D-19) also noted that at their facility a leak
check program is already in place for components in gas/vapor
service, which are located in the esterification section, for
safety and process quality control reasons.  The commenter
(IV-D-19) described their leak check program, which they have
used for 24 years as part of equipment startup:  Following an
outage where changes have been made to the esterification
equipment, a routine 2 to 3 hour leak check is conducted by
introducing hot ethylene glycol vapors into the system.  Leaks
are repaired by tightening flanges bolts before introducing new
materials into the process.
     The commenter (IV-D-19) provided equipment counts and an
emissions estimate (using average SOCMI emission factors, which
the commenter (IV-D-19) felt overestimate actual emissions) of
2.1 Mg/yr for their components in gas/vapor and light liquid
                               8-31

-------
service.  The commenter  (IV-D-19) stated that these emissions are
very small compared to point source emissions.
     Response:  The EPA re-evaluated the emission reductions,
costs, and cost-effectiveness of this subcategory, which contains
only one facility, based on the specific component counts
provided by that facility  (see Item IV-B-03, Docket A-92-45).
(Previously, the EPA had to estimate component counts for this
facility.)  Based on the component counts provided, the EPA has
determined that the proposed equipment leak standards are not
cost-effective, and the final standards exempt this subcategory
from the equipment leak standards.
                               8-32

-------
                    9.0  EMISSIONS AVERAGING

     Comment:  One commenter (IV-D-22) expressed confusion
concerning which emission points could be included together in
an emissions average.  The commenter  (IV-D-22)  stated that the
definition of affected source contained in proposed §63.500(b)
groups all similar emission point types and not the individual
production lines.  The commenter (IV-D-22) felt that other
text in the proposed rule grouped individual production lines.
The commenter  (IV-D-22)  stated that these two different
approaches represent different accounting of the same emission
points.  The commenter  (IV-D-22) stated that the emissions
averaging provisions should allow all similar emission points
on a contiguous site, except those points excluded as
presently defined, to be included in an emissions average.
The commenter  (IV-D-22)  explained that this would allow
industry to meet site emission limits on a whole through
flexibility and proficient economics as aspired in the
preamble of the Act.
     One commenter  (IV-D-09) stated that proposed
§63.510 (a) (2) unnecessarily restricts emissions averaging
within individual subcategories.  Another commenter (IV-D-10)
stated that one plant may contain numerous different NESHAP
requirements and recommended that the emissions averaging
provisions be modified to provide greater flexibility by
allowing an affected source a one-time (initial) option to
either average emissions, from the same subcategory or average
the cumulative emissions within its thermoplastic processes.
     Response:  The EPA agrees that the definition of affected
source contained in §63.500 (b)  of the proposed rule was
unclear.  The EPA has revised the provisions defining the
affected source in the final rule.  A discussion of these
revisions is included in Chapter 2.0 of this document.  The
                              9-1

-------
final rule defines an affected source as each group (i.e., one
or more)  of TPPUs at the same plant site manufacturing the
same thermoplastic product (i.e., one of the 19 subcategories)
as their primary product.  Also included in the affected
source are each wastewater stream and operation and each heat
exchange system that is associated with each group of TPPU.
     Emissions averaging is permitted among emission points
within the same source category or subcategory which are
located at the same plant site.  The final rule has 19
subcategories based on 7 source categories; a nineteenth
subcategory was created between proposal and promulgation for
a specific PET process.
     As described above, the TPPU within each subcategory make
up a separate affected source at a given plant site.
Therefore, while there may be more than one subcategory
 (i.e., affected source) present at a plant site, emissions
averaging is permitted only among emission points within the
same subcategory or affected source.  This differs from the
HON.  The HON regulates a single source category, and there is
only one affected source at a plant site that is regulated by
the HON.  Therefore,  emissions averaging is permitted among
all emission points eligible for emissions averaging at the
same plant site that  are covered^ by the HON.
     Finally, emissions  can not  be averaged across
subcategories regardless of the  potential for increased
flexibility.  Therefore, the final rule retains this
restriction and averaging across all subcategories at a single
plant site is not  allowed in this rule.
     Comment;  One commenter  (IV-D-05) contended that
compliance techniques such as  emissions averaging can improve
the overall effectiveness of this rule by reducing the cost of
 compliance and allowing affected sources flexibility as
needed.
     One commenter (IV-D-09) maintained that the emissions
 averaging provisions in the  proposed rule,  although modeled
 after the HON,  are more restrictive  than the HON.  The
 commenter (IV-D-09)  stated that  in order to be  useful to
                              9-2

-------
industry, the provisions need to be more flexible than the
HON.   [NOTE:  The commenter  (IV-D-09) did not indicate
specific areas where the proposed provisions are more
restrictive.]
     One commenter  (IV-D-18) supported the emissions averaging
concept contained in the proposed rule but was concerned about
some aspects of the proposal.  The commenter (IV-D-18) stated
that the emissions averaging provisions are long, complicated,
and difficult to understand, and the commenter (IV-D-18) was
unsure whether the approach could actually be used due to the
stringency of the limitations, such as the number of emission
points that may be averaged.
     One commenter  (IV-D-10) was concerned that wording in the
proposal Basis and Purpose document  [EPA-453/R-95-004a;
March 1995] stating that thermoplastic affected sources will
not find a large number of opportunities to generate
cost-effective credits may actually discourage industry from
investigating and identifying new methods of reducing
emissions.  The commenter  (IV-D-10) requested that EPA provide
as much operational flexibility as possible in the emissions
averaging provisions.
     Response;  The EPA thanks the commenters for their
support of the emissions averaging provisions.   The EPA allows
affected sources to use emissions averaging if they can
achieve a comparable level of emissions reduction and hazard
and risk benefit to point-by-point compliance.   The final rule
retains emissions averaging provisions.
     As stated in the Basis and Purpose Document
[EPA-453/R-95-004a; March 1995] for this proposed rule, the
emissions averaging provisions contained in the thermoplastics
rule are almost identical to the provisions contained in the
HON.  The proposed rule incorporated the emissions averaging
constraints included in the HON that were based on concerns
expressed during the HON public comment period.  Those
concerns were discussed in a supplementary Federal Register
notice published on October 15, 1993; 58 FR 53479.  While the
EPA understands how commenters may have perceived the proposed
                              9-3

-------
provisions concerning the number of emission points allowed in
an emissions average to be more restrictive than the HON, that
was not the EPA's intent, and the final provisions have been
modified to clarify this point.  In the final rule, the number
of emission points allowed in an emissions average has been
increased to 20 points for the entire plant site, with an
additional 5 points allowed if pollution prevention measures
are used.  This issue is discussed in detail later in this
chapter.
     As the EPA stated at proposal, although the EPA sees
limited potential for emissions averaging in this industry,
the EPA did not want to  exclude emissions averaging from the
rule.  The EPA's goal in providing emissions averaging
provisions in  this  and other rules has been to make emissions
averaging available to sources faced with some emissions
points that are particularly difficult or costly to control.
This  decision  is  in keeping with  the EPA's general policy  of
encouraging the use of  flexible compliance approaches where
they  can be properly  monitored and enforced.
      Comment:   One  commenter  (IV-D-25) requested that the
averaging period  for  a source  be  extended  from three months to
one year to provide adequate  flexibility.  The commenter
 (IV-D-25)  stated  that a source may require periodic shutdowns
of some  equipment,  and that because  of  these  shutdowns,
additional  time may be needed to  average out  debits.
      Response: • The compliance period for averaging that was
proposed,  an annual period with quarterly checks,  has been
 retained in the final rule.  Allowing averaging over a  year's
 time instead of just one quarter provides flexibility  for
 affected sources whose production rates vary over time.  The
 additional requirement that debits cannot exceed credits by
 more than 30 percent in any one quarter should assure  that
 wide-ranging  fluctuations in HAP emissions will not occur.
 The EPA concurs with commenters that an averaging period
 shorter than  one year would preclude the use of some emission
 points in averages.  An annual period allows inclusion of
 points that:   (1) do not have the same emission rates during
                               9-4

-------
 some periods of the year; and (2)  must undergo temporary
 maintenance shutdowns at different times during the year.
 Hence, an annual period provides affected sources the
 flexibility to construct the most cost-effective emissions
 averages.
      Comment;   One commenter (IV-D-10)  believed that the
 requirement in proposed §63.510(k)  to show relative risk is a
 strong disincentive to affected  sources to take part in
 emissions averaging.   The commenter (IV-D-10)  also considered
 this provision to be outside the scope of the  MACT Standards
 program.   The  commenter (IV-D-10)  believed the requirement for
 risk evaluation could hinder the achievement of or dissuade
 implementation of reductions of  HAP emissions  by emissions
 averaging that could be greater  than reductions achieved by
 strict application of the standards.
      Response:  The EPA disagrees  with the commenter on both
 issues.   First,  the EPA disagrees  that  the consideration of
 relative  risk  in emissions averaging is  outside the  scope  of
 the  MACT  standards program.   As  noted in the response to
 similar comments on the emissions  averaging provisions  in  the
 HON,  the  EPA believes that it has  the authority under the  Act
 to establish provisions as part  of  the alternative averaging
 system that will ensure that there  is no  increase in risk  or
 hazard as the  result  of a  source's  election to  comply through
 emissions averaging.   The  fact that section 112 (f) of the Act
 contemplates that  residual risk will be evaluated at a later
 time  and  that  other provisions specifically call for the
 consideration  of risk does not mean that the EPA is precluded
 from  considering risk or hazard in other contexts.
 Consequently,  the  EPA maintains that it has the authority to
 address risk and hazard in the emissions averaging provisions
 through a procedure such as  the one adopted in  the final rule
 -- the requirement  that affected sources that elect to use
 emissions averaging must demonstrate, to the satisfaction of
 the implementing agency, that compliance through emissions
averaging would  not result in greater risk or hazard than
compliance without emissions averaging.
                              9-5

-------
     Second, the EPA does not believe that the consideration
of risk will dissuade affected sources from using emissions
averaging.  Further, the EPA does not believe that the result
hypothesized by the commenter  (i.e., more emissions will occur
because fewer affected sources will emissions average) can be
reasonably expected to occur.  In contrast to this commenter's
concerns, the EPA received comments on the emissions averaging
provisions of the HON that indicated an owner or operator's
decision to emissions average will be based largely on
technical and economic criteria.  That there is no incentive
to propose emissions averages  that could increase risk or
hazard because owners or operators have strong incentives to
control the most hazardous emissions, such as protecting
community and worker safety, increasing product safety,
controlling pollution in other media, addressing State air
toxic laws, reducing residual  risks, and upcoming rulemakings
under section 112(g).  Therefore, the EPA does not believe
that the  requirement to  show relative risk will be a  strong
disincentive to  affected sources to take part  in emissions
averaging.
     Comment:  One commenter (IV-D-10) urged the EPA  not to
allow States to  have  the option of  disallowing emissions
averaging.  The  commenter (IV-D-10)  stated that  if the States
have open-ended authority in choosing whether  or not  to allow
the use of  emissions  averaging,  as  stated in proposed
 §63.503(b)(2),  affected sources would lose an  important
mechanism for  operational flexibility.   In addition,  the
 commenter (IV-D-10) explained, given the inherent delay before
 the States actually receive delegation of this rule,  the owner
 or operator of an affected source could potentially  invest  a
 year developing a compliance plan based on  emissions
 averaging,  after which time the State could decide  that
 emissions averaging is not an option.   Such a situation would
 be unduly burdensome and unfair to an affected source.  The
 commenter  (IV-D-10) felt that emissions averaging should  not
 be left to State discretion.
                               9-6

-------
     Response;  The final rule states that State or local
agencies have discretion to disallow emissions averaging.
Under Section 116 of the Act States have the authority to
impose more stringent requirements than the EPA if they
choose.  Thus, if a State believes that disallowing emissions
averaging is more stringent they have the authority to
disallow it.  The EPA maintains that States should have
discretion on whether to allow emissions averaging for a
number of reasons.  First, the EPA acknowledges that emissions
averaging can be more complex to administer than the
provisions allowing point-by-point compliance.  Allowing
emissions averaging could increase the administrative burden
to the State or local agency; this may be an important concern
for implementing agencies with limited personnel and
resources.  However, the determination of what constitutes too
much administrative burden will differ from State to State.
Some States may consider emissions averaging an acceptable
strategy for compliance and will retain the program.
     Second, the EPA recognizes that emissions averaging in
this rule could be inconsistent with ongoing air pollution
control programs for some States.  The EPA supports the use of
emissions averaging where it may be appropriate, and maintains
that the program has been designed to be enforceable and
protective of health and welfare.  However, the EPA also
acknowledges that its use must be balanced by the individual
needs of State and local agencies that bear the responsibility
                                f
for administering and enforcing the rule.
     Comment;  Five commenters (IV-D-06, IV-D-09, IV-D-11,
IV-D-19, IV-D-25) requested that the number of emission points
allowed in an emissions average be increased to 20 points from
the 5 points allowed at proposal.  If pollution prevention
measures are used, commenters (IV-D-06, IV-D-19) requested
that an additional 5 emission points be allowed from the
3 emission points allowed at proposal.  The commenters
(IV-D-06, IV-D-11, IV-D-19, IV-D-25) stated that this would be
consistent with the HON and with their understanding of the
draft Group I Polymers and Resins NESHAP.
                              9-7

-------
     Two commenters (IV-D-11, IV-D-22)  requested that there be
no limit on the number of emission points allowed in an
emissions average provided they are all within the contiguous
site.  As a result of this, the commenter (IV-D-22)
recommended that proposed §63.510(e)(3)(i) and §63.510(f)(1)
be deleted from the rule.
     One commenter  (IV-D-22) contended that by limiting the
number of emission points allowed in an emissions average to
5, the EPA is in fact segregating the individual production
lines.  The commenter  (IV-D-22) stated that the EPA provides
no rationale for quantifying the number of 5 or the additional
3 as the number of emission points  that can be averaged.
     Another commenter  (IV-D-10) stated that only allowing
emissions averaging between 5 emission points has no
scientific basis or technical merit.
     Response;  At proposal, the EPA concluded that the
maximum number of collocated affected  sources
 (i.e., subcategories)  at any one plant site would be 6.  The
EPA then determined that,  if the HON emission point limit of
20 to 25 emission points per affected  source
 (i.e., subcategory) were directly applied to this rule,  a
total of 120 to 150 emission points across 6 emissions
averaging plans could  potentially be involved in emissions
averaging at this worst-case plant  site.  Therefore, the EPA
determined that the maximum number  of  emissions points
included in an average for each affected  source
 (i.e., subcategory) without  the use of pollution prevention
would need to be 20 divided by 6, which equals 3 1/3.  Because
this number seemed  too small to allow  practical consideration
of emissions  averaging,  the number  was increased to 5 emission
points.
     However,  in response  to commenter requests and to be more
consistent with the HON and the Group  I Polymers and Resins
NESHAP,  the number  of  emission points  allowed in emissions
averages  for  the entire plant  site  has been  increased to
20 emission points, with an additional 5  emission points
allowed if pollution prevention measures  are used.  The  EPA
                              9-8

-------
determined  that  limiting the number of emissions points as
done at proposal was too restrictive and did not provide the
degree of flexibility necessary for emissions averaging.  This
change provides  more flexibility for owners or operators,
maintains consistency with the HON and Group I Polymers and
Resins NESHAP, and restricts the administrative burden of
implementing  and enforcing emissions averaging.   [Note:  The
restriction of emissions averaging to emission points within
the same subcategory is discussed in a previous response.]
     Two commenters requested that there be no limit on the
number of emission points allowed in an emissions average.
The EPA disagrees  with this request because, in summary, there
is a concern  about the burden and cost to implementing
agencies of overseeing and enforcing large numbers of emission
points in averages.  For a more detailed discussion of the
reasons for constraining the number of emission points to be
included in emissions averaging, see pages 19428-19429 of the
preamble to the  final HON rule  (59 FR 19402) promulgated on
April 22, 1994.
     Comment:  Five commenters  (IV-D-06, IV-D-09,  IV-D-1,0,
IV-D-13, IV-D-18}  requested that batch process vents be
allowed to  emissions average.  The commenters (IV-D-06,
IV-D-09, IV-D-10,  IV-D-13, IV-D-18)  explained that the EPA had
stated that there  was no acceptable and satisfactory
methodology available to quantify the emission reductions for
batch process vents because the accuracy and consistency
needed for  emissions averaging are greater than that needed
for applicability  determinations.   One of the commenters
(IV-D-18) argued that, if the EPA truly believes that batch
process vent emissions cannot be quantified with sufficient
accuracy to demonstrate the 90 percent reduction in emissions,
then there  should  be no emission standards for batch process
vents, only work practice or other operational .standards.   Two
commenters  (IV-D-10, IV-D-13)  maintained that emission
calculations should be acceptable for emissions  averaging.
One commenter  (IV-D-06)  stated that  they had conducted
extensive testing  of batch process vents at two  of their
                             9-9

-------
facilities and that methodology was available to accurately
quantify emissions from batch process vents.
     One commenter (IV-D-09) contended that for the EPA to
exclude all batch process vents from emissions averaging
because this has been done in the past and because the EPA
does not think adequate data are available is arbitrary and
overly restrictive.  The commenter  (IV-D-09) stated that batch
process vents represent a significant number of the process
vents in this industry and, thus, need to be included for an
emissions averaging scheme to be useful.
     One commenter (IV-D-10) stated that the equations 28 and
53 in proposed §63.510(g)(1) and  (h)(1) should be changed,
respectively, to include actual and allowable terms for batch
vents.
     Another commenter  (IV-D-05) believed the omission of
batch process vents from emissions averaging will facilitate
an easier implementation of the emissions averaging concept.
     Response:  With the changes made to the provisions for
estimating  emissions from batch process vents, the EPA finds
that, indeed, the emissions estimation techniques provided in
the  final rule are sufficient  for purposes of emissions
averaging.  Therefore,  the  final rule allows emissions
averaging of existing batch process vents, and the equations
in §§63.510 (g) (1) and  63.510(h) (1)  [§§63.1332(g) (1) and
63.1332 (h) (1)] have been  revised to  include both batch process
vents and aggregate batch vent streams.   [Note:  Changes to
the  emissions estimation  procedures  for batch process vents
are  discussed in Chapter  5.0 of this document. 3
      Comment:   One commenter  (IV-D-06) asserted that process
contact  cooling towers  should  be able to be used to generate
emissions  averaging  credits.   The  commenter  (IV-D-06)
explained  that,  if the rule is revised to allow cooling tower
water in contact condensers associated with vacuum systems,
the  opportunity would exist for  a  facility to control cooling
tower emissions to a level  more  stringent than the applicable
standard.
                              9-10

-------
      One commenter (IV-D-05) was uncertain of the ability to
 reliably quantify the generation of debits from process
 contact  cooling towers  through the proposed use of industry
 generated estimates.  The  commenter  (IV-D-05) recommended that
 if process contact cooling towers are to be included in
 emissions averaging, the EPA provide a reliable and accurate
 method for determining  emissions, perhaps based on the method
 provided by the commenter.
      Response:   Except  as  noted in the following paragraph,
 the  final rule  has been revised such that emissions from
 process  contact cooling towers at existing facilities are not
 regulated and process contact cooling towers are no longer
 part of  the affected source definition for existing
 facilities.  Therefore, process contact cooling towers can not
 be included in  emissions averaging.
      For PET affected sources using a continuous TPA high
 viscosity multiple end  finisher process, the final rule does
 regulate emissions from process contact cooling towers.
 However,  the EPA has determined that these emissions should
 not  be included in emissions averaging.  This decision was
 based upon consideration of the inconsistencies inherent in
 estimating emissions from process contact cooling towers.  The
 EPA  does not believe a  methodology exists that would
 consistently and accurately estimate emissions from this
 emission point.   Having an accurate method is especially
 important for this particular emission point within this
 subcategory because the emissions are estimated to be so large
 that  any error,  in either direction,  may overwhelm an
 emissions average that  included this emission point.   Also,  if
process  contact cooling towers were allowed in emissions
 averaging,  the  EPA cannot say with certainty that emissions
would not be greater than those that would occur under
point-by-point  compliance.  For these reasons,  the final rule
does  not  allow  the emissions averaging of process contact
cooling  towers  located  at existing PET affected sources using
a continuous TPA high viscosity multiple end finisher process.
                             9-11

-------
     Comment;  One commenter  (IV-D-09) requested that emission
points with controls installed prior to November 15, 1990, be
allowed in emissions averaging.   [Note:  Proposed §63.510
provides the Act date of November 15, 1990 as the date after
which controls must have been applied to utilize the achieved
emission reductions in an emissions average.  This is the same
date used in the thermoplastic and HON emissions averaging
provisions. ]
     Response;  Emissions averaging credits are not allowed in
the final rule for previous actions  (i.e., actions taken prior
to November 15, 1990, the date of passage of the 1990
Amendments to the Act).  Emission reductions from previous
actions occurred for  reasons  unrelated to the Amendments or
this rule  (such as State requirements) and are included in the
baseline.  As stated  in the proposal  preamble for the HON
 (57 FR 62608; December 31,  1992), if  the EPA allowed emission
reductions  from previous actions to  qualify as credits, then
the source  would be able to use  these previously acquired
credits to offset  debits generated after November 15, 1990.
This  situation would  result in more  total emissions occurring
than would occur under point-by-point compliance.
      For this reason, not  allowing credit for previous actions
should not be considered unfair or a penalty.  Rather, the
provision is necessary to  maintain emissions averaging as an
alternative means  of  compliance, achieving  equal or greater
emissions reductions  than achieved through  point-by-point
 compliance.  Likewise,  it  cannot be  considered a  "reward" to
 allow credit for previous  actions, which then enables  a  source
 to emit more pollution than would otherwise be allowed.   Also,
 if a previous action was required by another State  or  Federal
 rule,  the control can be used to meet the final  rule's
 requirements for Group 1 points as long as  the  control meets
 the level that the rule specifies.   However,  the control
 cannot be used to generate emissions averaging credit.
      Comment;  Two commenters (IV-D-09,  IV-D-10)  requested
 that process units that have been shutdown be eligible for
 emissions averaging.  The commenter  (IV-D-09)  stated that
                              9-12

-------
 proposed §63.510(d)(3)  prevents the  use  of  shutdown process
 units and that some  of  the most effective pollution prevention
 accomplishments occur when a new process unit, built with
 improved technology,  replaces an older process unit.  The
 commenter (IV-D-09)  stated that one  of their facilities
 reduced HAP emissions by 1,700 tons  per  year by precisely  this
 type  of process change  and that by disallowing such changes
 the EPA is discouraging pollution prevention and encouraging
 add-on controls.
      One commenter (IV-D-10)  felt that disallowing credits
 resulting from process  unit shutdowns  has no technical or
 regulatory validity.  The commenter  (IV-D-10) stated that,
 when  an owner  or operator makes a choice to shutdown a process
 unit  in order  to use the emission credits  (assuming that a
 federally enforceable commitment is  made to not restart the
 unit),  an actual emissions reduction has occurred for which
 credit should  be made available and  that the environmental
 impact of reduced emissions is achieved.
      Response;   It is not appropriate  to allow emissions
 averaging credits for process units  that have been shutdown.
 Regardless of  the motivation for a shutdown, the emission
 reduction from the production curtailment are not made
 permanent if emissions  averaging credit  is allowed.  If credit
 were  granted for the  emission reduction, the source could  then
 emit  an equal  amount  of emissions  from its debit generators.
 This  is in contrast  to  point-by-point  compliance,  where if a
 point  is shut  down,  the emissions  reduction is permanent.  To
 allow credit in emissions averaging  for  permanent shutdowns
 results in less stringent compliance and more total emissions
 than point-by-point compliance,  in which case emissions
 averaging does  not represent  an equivalent compliance
 alternative.
     Furthermore,  allowing emissions averaging credits for the
 addition of new process units  which are  replacing older
process units  that have been  shutdown  is not appropriate.
Although the EPA encourages affected sources to replace older
process units  to account  for  improved  technologies,  where cost
                              9-13

-------
 effective, emissions averaging credits will not be allowed in
 these circumstances because new sources are not allowed to
 emissions average under the provisions of the rule.  The
 rationale for not allowing emissions averaging for new sources
 is included in Volume 2C of the HON BID (EPA-453/R-94-003c;
 March 1994).
      Comment:  One commenter  (IV-D-09) stated that proposed
 §63.510 (d) (4) restricts the use of biologically treated
 wastewater streams from emissions averaging.  The commenter
 stated that biologically treated Group 2 streams should be
 allowed in emissions averaging.
      Response:  The final rule continues to prohibit the use
 of biologically treated wastewater streams in emissions
 averaging.  The EPA has based this decision on two reasons.
 The first reason addresses the comment that biologically
 treated streams should be included in general.  The second
 reason addresses the comment  that an owner or operator should
 receive credit for controlling Group 2 wastewater streams
 through'biological treatment.  First, under the emissions
 averaging provisions, an owner or operator must apply controls
 above and beyond what is required by other regulations or what
 is already being done in practice.  The EPA judged that the
 vast majority of affected sources would be required to send
 all wastewater to a biological treatment unit by existing
 water regulations or would already be sending all wastewater
 to a biological treatment unit as a matter of standard
 practice.  Based on this judgment, these wastewater streams
 would not be eligible for emissions averaging.  Second, the
' proposed and final rules provide several compliance options
 based on treating the combined wastewater stream from the
 affected source.  Under these compliance options, Group 2
 wastewater streams are controlled and the emission reduction
 achieved for these  streams  is applied toward the combined
 stream  control  requirement.   Therefore, the proposed and final
 rules already provide a benefit to  the owner or operator for
 the  control of  Group 2 wastewater streams through biological
 treatment.
                               9-14

-------
      Comment:   One  commenter  (IV-D-09) requested that
pollution prevention  credits be  allowed for  "out-of-process
recycling."  The  commenter  (IV-D-09) explained that
out-of-process  recycling may be  the most cost effective way to
achieve pollution prevention, and  it should  be allowed in the
calculations  [proposed  §63.510(j)(1)(i)] provided that all
emissions from  the  recycling process are included in the
calculations.   For  example, the  commenter  (IV-D-09) stated
that  solvent recovery on an out-of-process batch basis might
be more economical  than installation of in-process columns,
tanks, etc.
      Response:  Out-of-process recycling falls outside of the
definition of affected  source for  this rule.  Therefore,
emission  reductions from out-of-process recycling cannot be
included  in emissions averaging.
      Comment:   The  commenter  (IV-D-19) stated that the methods
for generating  credits  and debits  in §63.510(g) and (h) of the
proposed  rule include only methods for addressing those
sources that are  determined to be  Group 1 and Group 2 process
vents and have  their  control standards established by
following HON procedures.  The commenter stated that the PET
continuous TPA  process  vent standards in proposed
§63.505 (b) (3) are not developed by following the HON and are
not determined  to be  Group 1 or Group 2 sources using the HON.
The commenter  (IV-D-19) explained  that the PET continuous TPA
process vents have  a  separate process based emission standard
based on  production rate.  Therefore, proposed §63.510(g)  and
(h) appear to have  inadvertently omitted the procedure for
calculating credits and debits for PET continuous TPA process
vents.
      Response;  As  found in proposed §63.510(c)(5), continuous
TPA process vents (and  certain other PET continuous process
vents) subject  to proposed §63.505(b) were allowed in
emissions averaging.  In addition, proposed §63.510(c) (5)
stated that such  streams are to be considered Group 1 emission
points for the  purposes of emissions averaging.   These
                             9-15

-------
provisions are retained in the final rule in §63.510(c)(5)
[§63.1332(c)(5)].
     Comment;  One commenter  (IV-D-10) requested that the EPA
provide better guidance on acceptable procedures that an
affected  source must follow to identify and take credit for
pollution prevention efforts.  Specifically, the commenter
(IV-D-10) stated that the language in proposed §63.510(a)(3)
"An additional emission point may be added for each emission
point controlled through pollution prevention" is vague.
     Response:  The EPA has revised the language referred to
by the commenter in the final rule to read as follows:  "This
number may be increased by up to five additional emission
points if pollution prevention measures are used to control
five or more of the emission points included in the emissions
average." The EPA believes that this language clarifies the
commenter's  concern.
     Comment;  One commenter  (IV-D-10) felt that the  language
in proposed  §63.510(c) that reads  "... if sufficient
information  is available  to determine the appropriate value of
credits  ..." is very vague.   The commenter recommended  the
following language be used:  "... if reasonable information  is
available to calculate the actual  value of credits..." .
     The  commenter  (IV-D-10)  also  suggested that the  word
"control" in 63.510(c) be defined  or  clarified so  that  the
term is not  limited only to  added  or modified equipment, but
also includes change  in  operation, pollution prevention or
maintenance  and/or repair procedures.
     Response;  The EPA  disagrees  that the words "sufficient"
and "determine" need  to  be  revised per the commenter's
suggestion.  The  EPA also does not believe that  it  is
necessary to further  define  or clarify the word  "control" as
requested by the  commenter.   Within  §63.510  [§63.1332], all
types  of control  that can be used  to  achieve emission
averaging credits are identified.  These  types of  control
include both add-on controls, such as combustion devices, and
pollution prevention measures.   Pollution prevention measures
are,  as stated in §63.510 [§63.1332],  "...  any practice that
                              9-16

-------
results in a lesser quantity of organic HAP per unit of
product released to the atmosphere prior to out-of-process
recycling, treatment, or control of emissions ... " and may
include such measures as alterations to the production
process, equipment modifications, housekeeping measures, and
in-process recycling.  Therefore, the EPA does not find it
necessary to revise the wording in §63.510(c) [§63.1332(c)]
per the commenter's suggestions.
     Comment:  One commenter  (IV-D-05) recommended that the
EPA develop an emissions averaging guidance package, similar
to that being developed for the HON.  The commenter (IV-D-05)
recommended that such guidance thoroughly address the
interpollutant trading issue and contain a toxicity
measurement method similar, if not identical, to that of
section 112(g) in order to promote consistency.
     Response;  At this time, the EPA does not intend to
develop guidance material for emissions averaging specific to
this final rule.  However, two emissions averaging guidance
resources developed for the HON are available and may be
accessed via the Technology Transfer Network (TTN),  the EPA's
electronic bulletin board.  The EPA has developed an annotated
bibliography of resources to assist affected sources in
performing a risk hazard assessment titled "Technical Support
Document on Risk Hazard Assessment for Emissions Averaging
under the Hazardous Organic NESHAP" (EPA-452/R-95-007).   Also,
a computer program was developed to assist affected sources in
complying with the emissions averaging provisions of the HON.
The software, called "HON Averaging Assistant,"  leads the user
through input of needed site specific data and performs
calculations to assist in determining credits and debits for
emissions averaging.  However, the reader should remember that
the emissions averaging provisions of the thermoplastics rule
do vary slightly from the HON in several areas.   Areas  where
the final thermoplastics rule differs from the HON include:
(1)  the number of emission points that an affected source may
include in the emissions average; (2)  designation of a
separate Emissions Averaging Plan, where the HON includes
                             9-17

-------
emissions averaging information in the Implementation Plan;
(3) the requirement that under this rule some continuous
process vents are subject to an emission limit expressed as kg
of organic HAP per Mg of product; and  (4) batch process vents
are included in the emissions averaging provisions of the
thermoplastics rule.
     Comment;  One commenter  (IV-D-19) stated that §63.510(a)
of the proposed rule contains typographical errors, which
affect the scope of the emissions averaging provisions.  The
commenter  (IV-D-19) stated that §63.510(a) states  "this
section applies to owners or operators of existing sources who
seek to comply with the emission standards in §63.503(a) of
this subpart by using emissions averaging according to
§63.503 (f) of this subpart...11.  The commenter  (IV-D-19)
explained that the referenced §63.503(a) does not  refer back
to the §63.505 continuous process vent provisions.  The
commenter maintained this is  an error because process vents
from continuous operations are a group of emission points  that
the EPA intends to include in emissions  averaging.  Second,
the commenter stated that proposed  §63.510(a) refers to
§63.503(f), and this paragraph does not  exist.
     Response;  The commenter is correct that continuous
process vents are eligible for emissions averaging and this
was stated in the proposed rule.  The  commenter may have been
referring to a draft of  the proposed rule when they made their
comments.  The final rule states that:
     "This section applies to owners and operators of existing
affected sources who seek to  comply with §63.503(b)
 [§63.1313 (b) ] by using emissions averaging rather  than
following  the provisions of §§63.504,  63.505, 63.506, and
63.508  [§§63.1314,.  63.1315, 63.1316 through  63.1321, and
63.1330]."
                              9-18

-------
                    10.0  PERFORMANCE TESTING

      Comment;   One commenter (IV-D-18)  recommended that  owners
 or operators be allowed  to use  alternative test methods  which
 have been approved by the  Administrator.  The commenter
 (IV-D-18)  stated that in at least  six specific locations,  the
 proposed rule  requires owners or operators to use specified
 test methods.   In the past,  the commenter  (IV-D-18) stated, if
 there were circumstances where  it  was desirable and
 appropriate to use a  different  test method, or a modification
 of the method  specified  in the  rule, they have brought these
 circumstances  to the  attention  of  the Administrator and
 reached agreement on  an  acceptable protocol.  The commenter
 (IV-D-18)  noted that  this  has been a simple, quick, and
 effective  process,  and it  should be allowed to continue.
 However,  the commenter (IV-D-18) explained that for the  six
 locations  in the proposed  rule  that they identified,  the
 proposed rule  either  does  not allow for alternative test
 methods at all or validation of the alternative method
 following  EPA's Method 301 is required.  The commenter
 (IV-D-18)  stated that Method 301 is too complex and
 burdensome.  The commenter (IV-D-18) stated that both
 approaches,  specifying a particular test method or requiring
 the  use of Method 301, eliminate opportunities to make the
 rule work  better with less burden.   The commenter (IV-D-18)
 requested  that the phrase  "or other test methods approved by
 the  Administrator"  be added  to  the  provisions identified in
 their comments.
      Response;   The EPA  disagrees with commenter and has not
made the requested change  in the final rule.   The EPA
 considers  that adequate  flexibility has been added to the
                                            *
 final  rule with the addition of provisions allowing the use of
Method 25A and allowing  demonstrations of compliance  based on
                             10-1

-------
TOG.  The EPA cannot justify foregoing the use of Method 301
to validate alternative test methods when no specific comments
concerning its use were provided.  If commenters wish to
pursue changes to the procedures for obtaining permission to
use alternative test methods, and specifically the use of
Method 301, they should address concerns to the General
Provisions.   [Note:  Use of different EPA approved test
methods is discussed in the next response.]
     In response to the commenter's concern that some
provisions of the proposed rule specified a specific test
method and did not provide flexibility to pursue an
alternative test method, the final rule has been changed to
specifically cite that other methods validated according to
Method 301 are allowed.  However, even when not specifically
stated, owners or operators are always allowed to apply for
use of an alternative test method under §63.7 (f) of subpart A,
and the Administrator may either waive the use of or specify a
less rigorous validation procedure under paragraph 1.1.1 of
Method 301.
     Comment:  Two commenters  (IV-D-07, IV-D-18) requested
that a performance test not be required if a control device
has been tested in a similar manner using test methods
approved by the EPA.  The commenters  (IV-D-07, IV-D-18)
explained that the provisions in §63.506-4 (b) (3) of the
proposed rule allow the results of another performance test to
apply if the test was conducted using the same procedures
specified in the rule.  One commenter  (IV-D-18) stated that it
is unlikely that the performance test would have been
conducted using exactly the same procedures.  One commenter
 (IV-D-18) contended that any performance test performed for an
NSPS, NESHAP, or RCRA rule, a New Source Review permit, or
other programs administered or supervised by the EPA should be
considered sufficient.  Another commenter  (IV-D-07) requested
that RACT testing  satisfy the requirements for testing in MACT
standards.   The commenter  (IV-D-18) explained that performance
tests represent a  large investment of time, money, and
resources, and that it  is necessary for affected sources to
                             10-2

-------
 minimize duplication of these tests.   For example,  the
 commenter (IV-D-18)  explained,  one  of  the commenter's
 facilities conducts  at least  40 to  50  performance  tests  a
 year,  and this  number is expected to increase  dramatically.
 Currently, the  commenter {IV-D-18)  explained,  seven full-time
 stack testing personnel are employed to support this level of
 testing.  The commenter (IV-D-18) stated  that  the  total  cost
 of  each test  can range from $15,000 to $50,000, with an
 average of six  items monitored per  test.
      Response;   In response to this comment, the EPA
 considered the  open-endedness of the commenters' request and
 considered it unreasonable to allow a  test done using any EPA
 method to suffice for the performance  test required by this
 rule.   However,  the  EPA found it reasonable to allow tests
 done  for compliance  with an NSPS or NESHAP, where  either total
 organic HAP or  TOC or VOC was measured, to serve as  a
 substitute for  the performance test required under  this rule.
      Comment;   One commenter  (IV-D-18)  requested that a
 performance test not be required for RCRA incinerators that
 comply with 40  CFR part 264 when they  are used as control
 devices for complying with the  thermoplastics  rule.  The
 commenter (IV-D-18)  explained that  RCRA sets performance
 standards for incinerators.   These  standards require trial
 burns  to demonstrate compliance  with the standards and to
 determine destruction and removal efficiency.   The commenter
 (IV-D-18)  felt  it  would be redundant to require another
 performance test.
     Response;   The EPA agrees with the  commenter that a
 performance test done for RCRA should be  acceptable for
 compliance with  testing requirements under this rule; the
 final  rule allows  this.   The proposed rule already exempted
 boilers or process heaters burning hazardous waste that are in
 compliance with 40 CFR part 266.  In the  final rule,
 combustion devices that  are in compliance with either 40  CFR
part 264  or 40 CFR part  266 are exempt  from performance
 testing requirements.
                             10-3

-------

-------
                       11.0  MONITORING

11.1  OPERATING  PARAMETER MONITORING
     Comment:  Two commenters  (IV-D-09 and IV-D-10) requested
that more  flexibility be permitted for establishing compliance
levels for parameter monitoring.  The commenters  (IV-D-09 and
IV-D-10) asked the EPA to provide for the use of the HON range
concept  instead  of the level concept used in the proposed
rule.  The commenters  (IV-D-09 and IV-D-10) explained that the
HON uses ranges  in place of specific levels in recognition of
the fact that a  process or control device operates properly
over a range of  conditions.  One commenter (IV-D-09) cited
proposed §63.512(b)(1) which states that levels are to be
established as an average of minimum or maximum values from
test runs.  The  commenter  (IV-D-09) contended that using an
average  value leads to conditions where a portion of the time
the process or control device may be operating outside of the
acceptable performance range since an average may involve
values above and below the good performance range.
     One commenter  (IV-D-09) noted that the language of
proposed §63.513(b)(5)(ii)(B) is missing language from the HON
§63.152(b) (2) (ii) (A) and  (B) which discusses performance
testing  for determining ranges and the use of engineering
assessments.  The commenter  (IV-D-09) requested that this
important  language be added to the final rule.
     Response;   The approach for establishing parameter
monitoring criteria in the final rule incorporates the
methodology included in the HON rule, as well as keeping the
proposed methodology.  The methodology from the HON requires a
performance test, and then allows the owner or operator to use
the results of the performance test in concert with
engineering judgement to determine parameter monitoring
criteria that reflects proper operation of the control or
                             11-1

-------
recovery device.  Language from the HON [§63 .152 (b) (2) ]  was
added to the thermoplastics rule as part of incorporating this
methodology.  The methodology from the proposed rule specifies
that the parameter monitoring level is to be based on the
maximum or minimum recorded values, as appropriate,  from the
performance test.  For example, when determining the maximum
operating temperature for a condenser used to control a
continuous process vent, the owner or operator would conduct
three 1-hour test runs.  For each test run, four 15-minute
average values of the condenser temperature would be recorded.
The parameter monitoring level would be the average of the
maximum recorded temperature  (15-minute average value) for
each of the three'test runs.   [Note:  The final rule refers to
the criteria determined under either methodology as the
parameter monitoring level.]
     By allowing the affected source to choose between
methodologies,  the  final rule ensures that  site-specific
variation in emission point characteristics and control device
designs are considered  in  establishing appropriate parameter
monitoring  levels.
     Comment;   One  commenter  (IV-D-18) contended that the
calculated  average  is not  the  correct measuring point for
determining when  a  monitoring violation has occurred.  The
commenter  (IV-D-18)  explained  that whenever parameter ranges
are determined by a performance  test, they  constitute an
average of  multiple different  values.  Because  the approved
ranges represent  averages,  the commenter  (IV-D-18) explained
that it necessarily follows that some individual data points
were higher than the highest point of the approved range,  and
some were  lower than the lowest  point of  the  approved range.
Yet, on  the whole,  the values indicated compliance with
applicable regulations.  This means that  the  correct point  to
begin  looking for possible violations  is  not  the endpoint  of
an approved range;  instead, it is the  highest data point
 (above the tip of the range)  or the lowest data point  (below
 the bottom of the range) that contributed to  the approved
 average.   Unless parameters stray beyond those points,  there
                              11-2

-------
 is  no reason to believe  a violation may occur.  [Note:   This
 comment  uses the term "range,"  but  the  comment  has been
 interpreted by the EPA to be  applicable to  the  parameter
 monitoring "level" concept in the proposed  rule.]
      Response:   The EPA  believes the procedures contained in
 the proposed and filial rule for determining when a parameter
 monitoring violation has occurred are valid.  The EPA believes
 the basis  used to develop the parameter monitoring level  and
 the basis  used to calculate average values  for  the monitored
 parameter  are equitable  in that the averaging times are
 equivalent.   The data used to develop the parameter monitoring
 level are  typically based on  average values from a performance
 test.  The typical performance  test for a continuous process
 vent  entails 3  1-hour test runs, and the parameter monitoring
 level is calculated using a single  15-minute value from each
 test  run.   Likewise,  average  values for monitored parameters
 are based  on 15-minute values;  the  24-hour  average value  is
 calculated using all valid 15-minute values, and the 24-hour
 average  is compared to the parameter monitoring level.
 Further, as  stated in the proposal  Basis &  Purpose Document
 [EPA-453/R-95-004a,  March 1995], the EPA has concluded that
 operating  parameter monitoring  is an acceptable way to
 determine  whether control devices are being maintained and
 operated properly.   The  commenter is referred to
 Chapter  8.0,  Section 8.6  of the Basis and Purpose document for
 the rationale on this point that was presented at proposal.
      Comment;   One commenter  (IV-D-13)   agreed with the concept
 of parameter  monitoring  in conjunction with an initial
performance  test  of Group 1 emission points as a means to
 demonstrate  compliance.   [Note:  This commenter specifically
 addressed  batch process vents.]
     Response:  The parameter monitoring provisions are
 retained in the final rule.
     Comment:   One commenter  (IV-D-11)  contended that the
proposed batch  process vent monitoring requirements are not
 cost effective  and that the EPA did not consider more
practical monitoring.  The commenter (IV-D-11)  requested that
                             11-3

-------
alternate monitoring be allowed, such as the monitoring the
commenter has worked out with the State of New Jersey.
     The commenter  (IV-D-11) explained that a typical batch
building has approximately 20 condensers, scrubbers, and jets.
In most cases, control device instrumentation is located in
the field, and its readout is not transmitted back to a
central control room.  An additional.full-time employee would
be needed in each batch building to do hourly recordkeeping.
     The commenter  (IV-D-11) stated that companies affiliated
with CMA have numerous permits  for batch reactors with
condensers and/or scrubbers.  The commenter  (IV-D-ll)i
explained that under these permits owners or operators are
required to install and monitor contaminant side outlet
temperature on condensers and water flow and pressure drop for
scrubbers.  For recirculating scrubbers, pH is also required
to be monitored.  The commenter (IV-D-11) noted that only a
couple of their batch processes are required to keep records
on the control device instrumentation and that for the
processes where these records are required, continuous
recording has not been required; therefore, records are taken
manually.
     The commenter  (IV-D-11) explained that the State of New
Jersey initially required readings and records to be taken and
made every  shift.   For refrigerated  condensers, the records
meant nothing.  The condensers  were  on, but the temperature
readings varied substantially.  During a batch reactor charge
step, VOC emissions are  displaced to the condenser, and
uncondensed inerts  and vapors are vented.  The temperature
thermowell,  inserted at  the juncture of  the condenser outlet
and  the  stack, iteads the proper dew-point temperature of the
condensed and exiting gases. As the reaction mass  is agitated
and  reacted,  vapor  emissions cease.  Over time the  thermowell
in the  condenser outlet  conducts ambient heat from  the stack
and  shows a higher  temperature  that  falsely  indicates
noncompliance.  The commenter  (IV-D-11)  felt that the
recordkeeping requirements need to account for when the
                              11-4

-------
 control device is being used to control  emissions, not  just
 when it is operating.
      Subsequent to the  observation that  condenser temperature
 readings varied substantially,  the permit  authority  agreed to
 only require the recording  of the  control  device operating
 parameters during the worst-case emission  step of each  batch,
 stated the commenter (IV-D-11) .  This  ensures the control
 device will be in use and verifies its operation during the
 process'  worst-case emission step.
      In summary,  the commenter (IV-D-11) requested that a
 reading and record only be  required when the emission source
 is venting emissions to the control device.  The commenter
 (IV-D-11)  maintained that once per hour  is excessive, and  that
 once  per batch,  once per transfer  for  storage tanks  and
 transfer racks,  or once per shift  for  other continuous
 operations during the process'  worst-case emission step is
 more  appropriate.
      Response:   The EPA believes the proposed rule was
 misinterpreted by the commenters but acknowledges that  is  was
 potentially unclear.  Based on the  comments made, the EPA
 believes  the commenters were under  the impression that
 operating parameters for batch process vents had to be
 monitored 24-hours per  day;  this was not the intent.   In the
 final  rule,  the  EPA has clarified that operating parameters
 for batch process  vents must be  continuously monitored
 (i.e., measurements made at  least every 15 minutes)  during
 those  batch emission episodes, or portions thereof,  that the
 owner  or  operator  has selected to control as part of  meeting
 the requirement  to reduce emissions by 90 percent for the
batch  cycle.
     Several  changes or clarification were made to the final
 rule that  affect this area.  First, in §63.506-3(a)(1)
 [§63.1324 (a) (1)] of the final rule, changes were  made to
clarify that  "monitoring equipment shall be in operation at
all times  when batch emission episodes, or portions thereof,
that the owner or  operator has selected to control are vented
to the control device."  Second, it was necessary to  clarify
                             11-5

-------
§63.506-l(a)(2)  [§63.1322(a)(2)] to ensure consistency with
and understanding of the change described above.  In the
proposed rule, the requirement to achieve 90 percent for the
batch cycle was not expressed clearly.  In the final rule,
§63.506-l(a)(2)  [§63.1322(a)(2)] states that emissions for the
batch cycle must be reduced by 90 percent and that "Owners or
operators may  achieve compliance with this paragraph through
the control of selected batch emission episodes or the control
of portions of selected batch emission episodes."  The EPA
believes that, taken together, the two changes described above
make it clear  that parameter monitoring is only required when
emissions selected for control by the owner or operator are
vented to the  control device.
     A third point to be  made does not involve parameter
monitoring  for purposes of demonstrating continuous
compliance,  but  instead involves parameter monitoring during
performance tests.   In the final rule, an owner or operator
must either (1)  test  and  monitor the  entire batch emission
episode, or portion  thereof,  selected for control or  (2)  test
and monitor a selected period of the  batch emission episode,
or portion  thereof,  selected for control.  A  selected period
may be tested when the  emission rate  for the  entire episode
can be determined (i.e.,  when emissions are representative of
the average emission rate) or when  the emissions  are greater
than the average emission rate (i.e., at the  peak of the
episode).   The pertinent  point related to monitoring is that
an owner or operator is required to gathering parameter
monitoring data over the entire performance test.
 11.2   SELECTION OF MONITORED PARAMETERS
      Comment;  Two commenters (IV-D-08,  IV-D-23)  commended the
 EPA for allowing alternative monitoring  parameters  for batch
 process vents in proposed §63.506-3(c).   The  commenters
 (IV-D-08,  IV-D-23) stated that for many recovery and control
 devices,  there may be more appropriate  monitoring parameters
 than those specified in proposed section §63.506-3(a)  and
 §63.506-3(b).
                              11-6

-------
      Response:   The  final  rule  [§63.506-3(d)  [§63.1324(d)]3
 retains  the mentioned provisions.   In addition owners or
 operators of continuous  process  vents can  request  alternative
 monitoring parameters through similar provisions in the  HON,
 as  referred to  in §63.505-1  [§63.1315]  of  the  final rule.
      Comment;  One commenter  (IV-D-09)  suggested allowing more
 flexibility for designating monitoring parameters  for "other
 recovery devices"  than was provided  in Table 4 of  the proposed
 rule.  The commenter (IV-D-09) explained that  in the proposed
 rule  the only parameter  listed for  "other  recovery devices"
 was the  HAP concentration  level  at the outlet  of the recovery
 device.   The commenter  (IV-D-09) stated that it may not  be
 technically or  economically feasible to install continuous
 monitoring at the  outlet of a control device.   The commenter
 (IV-D-09)  requested  that owners  or operators be allowed  to
 base  the alternate parameter  on  engineering judgement and
 manufacturer's  recommendations.  The commenter (IV-D-09)
 included suggested language for  revising Table 4.
      Response:   As discussed  in  the  previous response, owners
 or  operators may request to monitor  alternative parameters
 through  provisions in §63.506-3  [§63.1324] for batch process
 vents and through  provisions  in  the  HON for continuous process
 vents.
      Comment;   One commenter  (IV-D-09)  maintained  that
 monitoring a valve position on a bypass line using a computer
 system should be an  acceptable alternative to  using  a flow
 indicator,  car-seal,  or  lock-and-key device.   The commenter
 (IV-D-09)  explained  that their facilities have  a number of
 valve positions monitored  by  the process computer system,  the
 commenter (IV-D-09)  found  this to be a reliable means to
 determine when  a bypass  is occurring.
     Response;   The  EPA agrees  with the commenter, and the
 final rule allows  the use  of  computer monitoring of bypass
 lines.
     Comment:   Table  5 of  the proposed rule requires that
 scrubber effluent  pH, scrubber liquid flow rate, and scrubber
gas flow rate be monitored for scrubbers used for halogenated
                              11-7

-------
process vents.  One commenter  (IV-D-13) maintained that
monitoring scrubber gas flow rate should not be required for
batch process vents because this parameter will not provide
meaningful compliance data due to intermittent gas flow.  The
commenter  (IV-D-13) stated that monitoring of scrubber liquid
flow rate when combined with scrubber liquid pH monitoring
should be an acceptable monitoring method for batch process
vents.
     Response;  The EPA agrees with the commenter that
monitoring scrubber gas flow rate would not be a meaningful
indicator of compliance for batch process vents.  In the final
rule, owners and operators of  batch process vents are required
to monitor scrubber liquid flow rate and scrubber effluent pH
only.
     Comment;  One commenter  (IV-D-13) stated that EPA should
only require monitoring of absorber liquid specific gravity on
total recirculating absorber systems where the VOHAP is
soluble  in the scrubber liquid.  The commenter  (IV-D-13) felt
that absorber systems  that operate either as once-through
systems  or with a  continual blowdown stream to organic HAP
recovery, disposal or  treatment should not be required to
monitor  absorber liquid specific gravity.
     Response:  The EPA agrees that in some situations it may
not be necessary to monitor the specific gravity of the
absorber liquid.   However, the decision to forego monitoring
the  specific  gravity of the absorber liquid in lieu of another
parameter is  best  determined on a case-by-case in concert with
the performance test.  If the  commenter wishes to monitor
another  parameter, he  or  she could pursue such an option
through  the  provisions for obtaining alternative monitoring
parameters which are discussed in an earlier response.
      Comment:   One commenter  (IV-D-13) suggested that in lieu
of monitoring the  total regeneration stream mass flow with an
integrating  mass  flow  monitor, which is not cost effective,
steam supply pressure  should  be allowed as an alternate
monitoring parameter for  steam regenerated activated carbon
systems.  On the  other hand,  the  commenter  (IV-D-13) stated
                              11-8

-------
that on carbon systems that regenerate using hot air or
nitrogen, the regeneration stream mass flow should be
monitored.
     Response;  The EPA does not agree with the commenter that
steam pressure would be a suitable alternative monitoring
parameter to steam mass flow rate, and the final rule
continues to require that the total regeneration stream mass
flow be monitored.  The commenter did not provide data
supporting their claim.  If the commenter wished to pursue
this further, he or she could do so through the provisions for
obtaining alternative monitoring parameters.
     Comment:  One commenter (IV-D-06) stated that for batch
process vents the exit temperature of the chilled water or the
exit flow rate of the chilled water, both of which will ensure
the condenser is operating properly, are suitable substitutes
to the exit temperature of the condenser.  The commenter
(IV-D-06) requested that Table 5 of the proposed rule be
revised to allow monitoring of these parameters.  The
commenter (IV-D-06) stated that monitoring the exit
temperature of the condenser may lead to erroneous readings
and/or permit exceedances for batch process vents.  The
commenter (IV-D-06) explained that for a batch process vent,
the condensible gaseous vapors generated are intermittent.
During certain times, there is no venting and a temperature
monitoring device will be measuring temperatures near ambient.
Because there is no liquid condensate during these non-venting
periods, the gaseous vapors at the exit of the condenser are
influenced by the outside environment resulting in potential
erroneous readings and/or permit exceedances.
     Another commenter  (IV-D-13) maintained that it is
critical that the EPA permit periodic monitoring of condenser
exit temperature as opposed to continuous monitoring.  The
commenter (IV-D-13) explained that batch process vents
typically have intermittent flow.  During periods of no stack
gas flow, the temperature monitor will gradually warm up due
to heat transfer from the vent.  The resulting temperature may
improperly show an excursion.  Therefore, the commenter
                             11-9

-------
(IV-D-13) suggested continuous recorders not be required
because the data they generate is misleading.
     Response:  Because of the changes made to the final rule
related to parameter monitoring for batch process vents,
discussed in an earlier response, the EPA is not persuaded to
change the specified operating parameter for condensers.  The
changes made to the final rule have addressed the issue of
monitoring a control device when no emissions are being vented
to  the control device.  However, an owner or operator is free
to  apply for  alternative parameter monitoring under
§63.506-3(d)  [§63.1324(d)l.
11.3  EXCURSIONS
      Comment:  Seven  commenters  (IV-D-03, IV-D-08,  IV-D-09,
IV-D-10,  IV-D-11,  IV-D-15,  IV-D-21) requested that  the
proposed rule duplicate  the  HON provisions in  §63.152 (c) (2).
for excused excursions.   One  commenter (IV-D-25)  recommended
that the EPA allow one excursion per  month. The commenter
 (IV-D-25)  stated that the HON system  of phasing down
excursions to one per semiannual period is  laudable but
 inadequate.
      One commenter (IV-D-21)  stated that this  phased reduction
 in the HON is premised upon the learning curve associated with
new e
 re
     iquipment and recognizes that a period of time will be
  :quired to experience the upsets or deviations from normal
operation which can be expected with new equipment.  Another
commenter  (IV-D-09) stated that control device operating
parameters will fluctuate for a variety of reasons that may
not be included in the start-up, shutdown, and malfunction
plan.  The commenter  (IV-D-09) stated that allowing excused
excursions would give sources more opportunity to understand
what kinds of fluctuations in the parameters can occur during
different  situations.  Several commenters  (IV-D-03, IV-D-08,
  IV
   -D-ll,  IV-D-21)  stated  that  it would be statistically
  impossible for a facility to never observe monitoring results
  greater than the emission limitation.
       The commenters (IV-D-03,  IV-D-08,  IV-D-11,  IV-D-13,
  IV-D-21) went on to state that allowing for no  excursions
                              11-10

-------
would be both unrealistic and would impose significant
additional capital and operating costs with negligible
corresponding reductions in air emissions.  The commenters
(IV-D-08, IV-D-15, IV-D-21) contended that by not allowing any
excused excursions in the proposed rule, the EPA has made the
thermoplastics rule more stringent than the HON.  One
commenter (IV-D-21) stated that if the EPA does not follow the
HON excused excursion scenario, the EPA must justify the
increased cost of imposing the more stringent no excused
excursion provision.  One commenter (IV-D-15) requested that
the EPA justify their position using actual facility data.
Another commenter  (IV-D-13) stated that the EPA's Section 114
survey did not collect sufficient data to demonstrate that
sources comprising the floor meet a set control efficiency
100 percent of the time.  Two commenters  (IV-D-13, IV-D-15)
insisted that the data would have to show that a facility
operating at the MACT floor level of control would never
deviate from this control level.
     Response;  In the final rule, the EPA decided to excuse a
certain number of excursions for each reporting period.  This
decision was based on data and information presented during
public comment on the HON and reiterated in public comments
received on this rule and during industry meetings held
subsequent to proposal.  The EPA also concluded that not
allowing excused excursions would impose significant
additional capital and operating costs on the affected source
for only negligible corresponding reductions in air emissions.
As is always the case, a State has the discretion to impose
more stringent requirements than the 'requirements of NESHAP
and other federal requirements and could choose not to allow
the excused excursion provisions of this rule.
     The EPA considered the number of excused excursions that
would be most appropriate for this rule and determined that
the number of excursions allowed in the HON would be
reasonable.  Therefore, the final provisions allow a maximum
of 6 excused excursions for the first semiannual reporting
period, decreasing by 1 excursion each semiannual reporting
                             11-11

-------
period.  Starting with the sixth semiannual reporting period
(i.e., the end of the third year of compliance) and
thereafter, affected sources are allowed one excused excursion
per semiannual reporting period.  The EPA determined that
allowing one excused excursion per month on an indefinite
basis, as one commenter requested, was not warranted and could
possibly adversely affect the emission reductions achieved by
the rule; this request was not granted.
     Comment:  One commenter  (IV-D-18) supported the
provisions that state that when multiple parameters are being
monitored for the same control or recovery device, an
excursion involving more than one of these parameters counts
as a single excursion, for the control or recovery device.
     Response:  The EPA thanks the commenter for the support,
and this provision is retained in the final rule.
     Comment;  One commenter  (IV-D-18) supported the
provisions that an exceedance is based on the  batch cycle
daily  average for a parameter, as opposed to hourly or
15-minute" averages.  The  commenter  (IV-D-18) stated that this
was important because parameters  can  change for brief periods
of time  for  no discernable  reason and return to previous
levels without causing  a  real problem.  The commenter
 (IV-D-18)  stated  that by  focusing on  the daily value, the
owner or operator has the opportunity to fix minor problems
quickly  and  continue  operation.
      Response;  The EPA thanks  the  commenter for  the  support,
and this provision is  retained  in the final rule.
      Comment:   One commenter (IV-D-18)  supported  the
provisions that  "insufficient data"  does not mean a  single
missing  data point.   The  commenter  (IV-D-18) explained  that
 any monitoring system is  subject to possible brief  failures  or
 outages,  most of which can be corrected quickly.  However,  if
 every missing data point  were a violation,  it  would be
 literally impossible to comply with the rule.
      Response:   The EPA thanks  the commenter for the support,
 and this provision is retained in the final rule.
                              11-12

-------
     Comment:  One commenter  (IV-D-18) supported  (1) the use .
of a daily average value to determine excursions and (2) the
provisions that it is not considered an excursion if these
situations occur during startup, shutdown, or malfunction of a
process and the startup, shutdown, or malfunction plan has
been followed.
     Response:  The EPA thanks the commenter for the support,
and these provisions are retained in the final rule.
     Comment:  One commenter  (IV-D-18) requested that the EPA
not consider a violation to have occurred every time a
parameter strays momentarily  beyond a pre-approved  range.
     Response:  The proposed  and final rule require a
comparison of a 24-hour average value against the parameter
monitoring level to determine operating parameter violations.
Momentary exceedances of the  level do not constitute a
violation.
11.4  DEMONSTRATION OF COMPLIANCE
     Comment;  One commenter  (IV-D-11) stated that  the
proposed'rule  (i.e.,  §63.114  of subpart G, §63.505, and
§63.506-3) do not seem to  recognize that absorbers  may be used
as control devices  (i.e.,  no  HAP recovery) on nonhalogenated
process vents.  Further, the  commenter  (IV-D-11)  stated  that
the provisions do not seem to establish monitoring  parameters
for absorbers used as control devices.
     Response:   The  EPA believes the monitoring  parameters
specified  for absorber  systems  are sufficient, regardless of
the intended use of the system  (i.e., as a control  or recovery
device).   The commenter did not provide suggested alternate
monitoring parameters for  the EPA to  consider.  Nor did  the
commenter  indicate how  the proposed rule  failed to  recognize
the use  of absorbers  as control devices.  The monitoring
provisions for absorbers have not been changed in the final
rule  to  specifically  address  this comment.
                              11-13

-------

-------
              12.0  RECORDKEEPING AND REPORTING
12.1  GENERAL BURDEN
     Comment:  Three commenters (IV-D-07,  IV-D-08, IV-D-11)
requested that the recordkeeping and reporting burden of the
proposed rule be reduced.  One commenter (IV-D-25) stated that
the EPA should include all the provisions of the HON that
reduce burden.  One commenter  (IV-D-18) contended that the
cost of complying with "paperwork" requirements  (monitoring,
recordkeeping, and reporting) will exceed the cost of actually
installing and operating air pollution control equipment.
     Response;  The EPA believes that the recordkeeping and
reporting requirements in the  final thermoplastics rule are
the minimum necessary to determine compliance on a continuous
basis.  Because monitoring, recordkeeping, and reporting data
are used directly to determine compliance, the EPA considers
these requirements to be an integral part of this and all
regulations developed under the Act.  Consistent with the
Paperwork Reduction Act, the EPA has attempted to reduce the
burden of recordkeeping  and reporting requirements to the
maximum extent, while still maintaining the enforceability of
the rule.
     The monitoring, recordkeeping, and reporting provisions
of the proposed rule were based on the concepts  from the HON
 (i.e., batch process vents)  or directly referenced the
provisions  of the HON  (e.g., continuous process  vents).
Therefore,  the EPA believes  the benefits of the  HON mentioned
by the commenter were  included in the proposed rule.
     The EPA reviewed  the monitoring,  recordkeeping, and
reporting provisions of  the  proposed rule and decided further
reductions  in the amount of  information required to be
monitored,  recorded, or  reported  were  possible.   Two reports
required at proposal,  the Initial Notification and the
                              12-1

-------
Implementation Plan, are not required in the final rule.
As part of deleting these reports, the EPA added a new report,
the Precompliance Report, to allow the owner or operator of an
affected source to request an extension of compliance or to
request approval to use alternative monitoring parameters,
alternative continuous monitoring or recordkeeping, or
alternative controls.  At proposal, these items were submitted
in the Implementation Plan.  Overall, these changes, deleting
the Initial Notification and the Implementation Plan and
adding the Precompliance Report, result in a reduction of the
reporting burden for the affected source.
     The final rule has been changed to require, with one
exception, that only hourly average values of continuously
monitored values be recorded; the proposed rule required the
monitoring and initial recording of all 15-minute values.  The
exception to this is that batch process vents must record each
measured data value but are no longer required to calculate
and record batch cycle averages.  The reasons for this
difference are discussed in more detail in Chapter 11.0 of
this document.
     The final rule also reduces the recordkeeping and
reporting frequency of batch cycle records for those batch
process vents required to comply with the batch cycle
limitation.  The final rule requires a record of the number
and type of batch cycles every three months, as opposed to the
requirement for daily records in the proposed rule.  The final,
rule requires that every second Periodic Report include the
batch cycle records for the previous 12-month period, as
opposed to the requirement for quarterly reporting in the
proposed rule.  The EPA considers that the changes and
clarifications made to the batch process vent monitoring,
discussed in Chapter 11.0, also serve to reduce the overall
burden.  In addition to specific changes to the requirements,
the final rule includes provisions for implementing a reduced
recordkeeping program  [§63.513(h)  [§63.1335(h) ] ] .
                              12-2

-------
     In summary, the EPA considers the final provisions
related to monitoring, recordkeeping, and reporting to be the
minimum required to demonstrate compliance with the rule.
12.2  RECORDKEEPING
     Comment;  Two commenters  (IV-D-18, IV-D-25) supported the
provisions which state that if each is recorded value is in
compliance with the parameter monitoring level for an
operating day, the affected source may simply record that fact
and is not required to calculate a daily average value or
batch cycle daily average value.
     Response;  These provisions have been retained in the
final rule.  Compliance with the parameter monitoring required
by the rule is based on the daily average value of
continuously monitored parameters.  If all recorded values for
a monitored parameter during an operating day are in
compliance with the parameter monitoring level established in
the Notification of Compliance Status or operating permit, the
owner or operator may record this fact and forego calculating
a daily average  (or batch cycle daily average) value.
     Comment;  One commenter  (IV-D-25) requested that manual
recordkeeping be allowed in the final rule as it was in
§63.151(g)(3) of the HON; however, the commenter (IV-D-25)
wanted readings of the operating parameters to be allowed at
4-hour intervals for situations where the control device
normally operates in steady-state.
     Response:  The proposed  [§63.513(e) (3)] and final
 [§63.513(g)(2)  [§63.1335(g)(2)]] rule allow non-automated
monitoring and recording systems.  These provisions state that
manual reading and recording of operating parameters must be
performed no  less frequently than once per hour.  These
provisions are based on those found in §63.151 (g) (3) of the
HON.  The decision to retain the hourly timeframe and to not
switch to 4-hour intervals, as requested by the commenter, was
made to ensure consistency with the continuous monitoring
requirements  requiring retention of hourly values for use in
calculating the daily average values.
                             12-3

-------
     Comment:  One commenter  (IV-D-25) urged the EPA to reduce
the recordkeeping requirements of the thermoplastics rule
related to monitoring control devices in the same way as the
HON requirements were reduced.  Another commenter  (IV-D-18)
urged the EPA to eliminate "universal" recordkeeping
requirements from the proposed rule.  The commenter  (IV-D-18)
contended that these records demonstrate routine compliance
and cannot be of assistance to the EPA in its enforcement role
in cases where there have been no violations.  The commenter
(IV-D-18) stated that the records also have no value to the
owner or operator after initial review has verified the
absence of atypical conditions.  The  commenter  (IV-D-18)
stated that a missing data point demonstrating routine
compliance could result in a  fine of  up to $25,000 per day for
an affected source.  The commenter  (IV-D-18) provided 7
specific examples of ways the recordkeeping requirements of
the rule could be reduced:
      (1)  Proposed  §63.506-5(c) requires retention of records
of all results of parameter monitoring for batch process
vents.  The commenter  (IV-D-18) recommended that they only be
required to keep the results  that show an excursion.
      (2)  Proposed  §63.506-5(f)(l) of the batch process vent
provisions requires retention of all  hourly records  for
flares.  The commenter  (IV-D-18) recommended that they only be
required to keep the records  that relate to pilot flame
outages.
      (3)  Proposed  §63.506-5(f)(3) of the batch process vent
provisions requires retention of all  hourly records of
"whether" a  flow indicator was operating and "whether" flow
was detected.  The  commenter  (IV-D-18) recommended that they
only  be  required to keep  records that show a diversion of
flow.
      (4)  Proposed  §63.513(f)(1) through  (f)(3) requires
retention of records  of all monitored data values, or records
of block average values for  15-minute or shorter periods.
Then,  if the daily  average value shows compliance, the block
average  values  or  individual  batch  cycle average values must
                              12-4

-------
be retained.  The commenter  (IV-D-18) claimed no records
should have to be retained if there was not an excursion.
      (5)  Proposed §63.513(f)(4) states that if the daily
average value shows noncotnpliance  (i.e., an excursion
occurred), all the data recorded on that operating day must be
retained.  The commenter  (IV-D-18) regarded this as an
excessive burden and requested that they only be required to
retain the daily average value because this is all that would
be needed to demonstrate a violation had occurred.  The
commenter (IV-D-18) contended that the short-term data points
which make up the daily average value are unnecessary.
      (6)  Proposed §63.513 (f) (5) requires retention of daily
average values for 5 years,  even if no violation occurred.
The commenter  (IV-D-18) disagreed with these provisions
because if an excursion occurred, there are separate
provisions requiring retention of the daily average value.  If
there was no excursion, the  commenter contended that retention
of the daily average values  was unnecessary.  The commenter
 (IV-D-18) recommended deleting these provisions in their
entirety.
      (7)  Proposed §63.513(f)(6)  states that if all measured
values were within established  limits, then the owner or
operator could record that fact and  retain the statement for
5 years.  The commenter believed  this was an excessive burden
and that if all the measured values  were within the
established limits, records  should not have to be kept.
      Response:  Under section 114(a) of the Act, owners or
operators have the burden of demonstrating that each emission
unit  remains in compliance at all times.  There must be a
means of verifying compliance on  a continuous basis.
Regulations must,  therefore, establish monitoring requirements
that  are  capable  of determining continuous compliance with the
applicable  standards.   The EPA  believes that the recordkeeping
and requirements  of the final thermoplastics rule are
necessary to show compliance with the  rule.  The EPA will
continue  to require owners or operators to keep records,
regardless  or  whether there was an excursion or not.  These
                              12-5

-------
records are necessary to prove compliance when no excursion
has occurred.  These records are also used to determine the
severity of a violation and the extent of the penalty to be
assessed once an excursion has occurred.  If an excursion
occurred and no records had been kept regarding the severity
of the excursion, a maximum fine of up to $25,000 per day per
control device could be assessed.
     The EPA has made every effort to reduce the recordkeeping
requirements of the final thermoplastics rule.  The EPA
reviewed the recordkeeping required by the proposed rule and
has made even further reductions in the amount of information
that is required to be recorded.  For instance, as mentioned
earlier, the final rule has been changed to require recording
and retention of hourly average values of continuously
monitored parameters.  The proposed rule required recording
15-minute averages.  Under the proposed rule, if the daily
average value was above the minimum or below the maximum
established levels  (i.e., excess emissions occurred), the
15-minute values had to be retained;  if the daily average
value was within the established level, the 15-minute values
could be converted to hourly  averages and the hourly averages
could be retained instead of  the 15-minute averages.  Upon
reconsideration, the EPA  found the proposed two-step process,
of first computing and  recording 15-minute averages, and then
being allowed to convert  them to hourly averages for record
retention,  to be burdensome and unnecessary.  Hourly average
values provide  a sufficient record to support the calculation
of the daily average value  of a monitored parameter.
Therefore,  to reduce the  recordkeeping  burden, the  final rule
has been changed to  specify that the  value of the parameter
must be  measured at  least once every  15 minutes, but only  the
hourly  average  must  be  recorded  and retained.  Further, only
hourly  averages must  be retained, regardless  of whether or not
excess  emissions occurred.   The  final rule does not require
recording  or retention of 15-minute  average values.. The
15-minute  monitoring frequency is consistent  with the  General
Provisions and previous NSPS and NESHAP for  emission points
                              12-6

-------
from similar industries.  The monitoring and recordkeeping of
control devices for batch process vents does not follow the
above and is discussed in more detail in Chapter 11.0.
Briefly, for batch process vents all measured data values are
required to be retained in the same way that hourly block
averages are retained for the monitoring of continuous process
vents.
     The final thermoplastics rule and the General Provisions
to Part 63 require records to be kept for five years, which is
consistent with the recordkeeping requirements of the
operating permit program and other NESHAP.  Storage of records
for more than 5 years would be burdensome, and any compliance
issues  should be identified within 5 years.  Furthermore, the
statute of - limitations  for enforcement is 5 years so there is
no reason to keep records for a longer period of time.  For
these reasons, the final rule continues to require the
retention of records for 5 years.
     The following paragraphs respond to the commenter's
 (IV-D-18) 7 examples of where the recordkeeping requirements
of the  proposed rule could be reduced.
      In response to example  1, in the proposed  [§63.506-5(c)]
and  final rule  [§63.1326 (c)], requires that documentation of
information used to establish the parameter monitoring levels.
This  includes parameter monitoring data collected during
performance tests.  These provisions do not require that all
results of parameter monitoring be kept for batch process
vents  as the commenter stated.
      In response to example  2, proposed §63.506-5(f)(1) refers
to table 8 of  the proposed rule which  requires monitoring for
the  presence of a  flame at the pilot light.  The provisions,
which are  in  §63.506-5(e)(1)  [§63.1326(e)(1)] of the  final
rule,  require  hourly  records of whether the pilot  flame was
present.
      In response  to example  3, proposed §63.506-5(f)(3)
 [§63.1326(e)(3)  of the final rule]  requires hourly records of
whether the  flow indicator was  operating  and whether flow was
 detected during that  period.
                              12-7

-------
     In response to example 4, proposed §63.513(f) (1) through
(f)(3)  [§63.513(d)(1) through  (d)(3)  [§63.1335(d)(1) through
(d) (3)] of the final rule], the EPA has decided to remove the
requirement to record 15-minute average values as required at
proposal.  As discussed earlier in this response, the EPA no
longer finds it necessary  to have a two-step process for
calculating daily  averages and 15-minute  averages are no
longer required.   The EPA  made this change to reduce the
recordkeeping burden for continuously monitored parameters.
The EPA is unable  to require that records only be kept when an
excursion occurs,  as the commenter suggested.
     In response to example 5, proposed §63.513 (f) (4)
[§63.513 (d) (4)  [§63.1335(d) (4)] of the final rule], the owner
or operator is required to keep each measured data value or
block average values if the daily average value is below the
minimum level or above the maximum established level.  As
stated in the response to  example 4, the  EPA has already
reduced the recordkeeping  burden.  However, for days when
operating parameters are not in compliance, the EPA considers
retaining each measured data value or the block average values
to be essential and is unable to only require retention of the
daily average value.
     In response to example 6, proposed §63.513(f)(5)
[§63.513(a)  [§63.1335(a)]  of the final rule], all records and
reports required by the final rule are required to be kept for
5 years.
     In response to example 7, §63.513(f)(6) of the proposed
rule  [§63.513(c)(6)  [§63.1335(c)(6)] of the final rule], the
EPA is unable to allow only a statement that a source is in
compliance.  Hourly values are required to be kept to
demonstrate continuous compliance.
     Comment:  One commenter  (IV-D-18) questioned whether the
records of the individual  measured data values monitored
during a batch cycle would need to be retained if they are all
within the approved range.  The commenter  (IV-D-18) maintained
that the EPA should have no need for the measured data values
if they are all within the compliance range and keeping them
                             12-8

-------
would be a burden with no corresponding environmental value.
The commenter  (IV-D-18) requested that proposed
§63.506-5(f)(2)(iv) be revised to state that the individual
measured data values do not need to be kept.  The commenter
(IV-D-18) included suggested language.
     Response;  The provisions in §63.506-5 (e) (2) (iii)
[§63.1326(e)(2)(iii)] of the final rule state that an owner or
operator may record that all recorded values for a monitored
parameter during an operating day are above the minimum level
or below the maximum level, instead of having to calculate and
record a batch  cycle daily average.  The owner or operator
must still keep the recorded values to allow the EPA to verify
compliance.  As discussed a previous response, the EPA must
require records to verify compliance.
12.3  REPORTING
     Comment:   One commenter  (IV-D-09) recommended clarifying
changes to the  Initial Notification provisions contained  in
§63.513(b)(1) of the proposed rule.  The commenter  (IV-D-09)
stated that the references in the §63.513(b)(1)  lead-in
paragraph are  in error  [there is no  §63.9(b)(6)  of subpart A]
or are not consistent with the Table  6 exclusion of §63.9(b)
of subpart A.   The commenter  (IV-D-09) explained that a review
of the provisions of proposed §63.513(b)(1) compared to
§63.9(b) of subpart A quickly leads  one to  the conclusion that
the exclusion  in Table 6 is correct  because proposed
§63.513(b)(1)  contains all the needed information.  The
commenter  (IV-D-09) contended that the references in proposed
§63.513(b)(1)(ii)(B) and  (C) to  "§63.5(d) of subpart A" are
redundant to the provisions in §63.513 (b) (2) and should be
replaced by a  reference to "paragraph (b) (2) of  this section."
The commenter  (IV-D-09) included suggested  language.
     Response;  The  final rule has been changed  to delete the
requirement for an Initial Notification.  This decision was
made to eliminate  duplicative requirements  and reduce the
recordkeeping  and  reporting burden of the final  rule.
      It was determined that the  Initial Notification  is not
necessary  for  thermoplastic manufacturers because it would
                              12-9

-------
duplicate information that will be available through the
operating permit program.  Major sources are required to
obtain Title V operating permits, and the deadline for permit
applications will be prior to the 3-year compliance deadline
for the thermopolastics rule.  The permit applications will
identify that the thermoplastic plants are subject to this
rule.  Therefore, a separate Initial Notification would be
duplicative and has been eliminated from the final rule.
     Comment;  One commenter  (IV-D-09) supported allowing
150 operating days after the compliance date to submit the
Notification of Compliance Status.
     Response:  The final rule has retained the provisions
included at proposal allowing 150 days after the compliance
date for submittal of the Notification of Compliance Status.
     Comment:  One commenter  (IV-D-23) stated that the
preamble to the proposed rule erroneously states that
quarterly reporting of the number of batch cycles accomplished
for Group 2 batch process vents would be required.  The
commenter  (IV-D-23) explained that the proposed rule requires
this information to be reported in every other semi-annual
Periodic Report, which is an appropriate reporting frequency.
     Response:  In fact, the proposed rule did require
quarterly reporting of the number of batch cycles accomplished
for those Group 2 batch process vents with a batch cycle
limitation.  However, in an effort to reduce the reporting
burden, the final rule requires submittal of the type and
number of batch cycles accomplished and a statement that the
Group  2 batch process vent is either in or out of compliance
with the batch cycle limitation on an annual basis only.
     Comment;  One commenter  (IV-D-25) contended that all
routine (i.e., periodic) reports should be removed from the
rule because  they are redundant with the annual operating
permit certification requirement.  The commenter  (IV-D-25)
stated that these reports are merely a continuation of the
reporting  required prior to  operating permits and their
required annual  certifications.  The commenter  (IV-D-25)
recommended including  in the  rule  that the annual
                             12-10

-------
certification is required by the operating permit regulation.
The commenter (IV-D-25) reminded the EPA that
President Clinton has mandated a 25 percent reduction in
paperwork.
     Response;  The periodic reporting system of semiannual
reporting is in conformance with §70.5(c) of the operating
permits program, which states that facilities are required to
submit reports no less frequently than once every six months.
The operating permits program contains monitoring,
recordkeeping, and reporting requirements minimally sufficient
for a facility to verify compliance and requires submittal of
compliance certifications.  The requirements in the operating
permit program serve as a baseline minimum for all facilities
subject to Title V.  However, for facilities subject to other
federal standards  (e.g. NESHAP), more frequent or substantial
monitoring, recordkeeping, and reporting may be necessary to
certify compliance.  Consequently, regulations developed under
Title III, including the thermoplastics rule, will incorporate
additional requirements that go beyond the minimal
requirements contained in the operating permits program.
Facilities must comply with only the most stringent
requirements.  Therefore, duplicate records or reports are not
required.
     Comment;  One commenter (IV-D-09) recommended that the
EPA clarify proposed §63.513 (b) (6) (ii) (C) [lack of monitoring
data] and correct the wording of §63.513 (b) (5) (i) (A)
[performance tests].  The commenter  (IV-D-09) included
suggested language.
     Response;  The EPA agreed with the suggestions and has
made these changes to the final rule:
     §63.513(e) (6) (iii) (C)  [§63.1335(e) (6) (iii) (C)] ,  Periodic
     Reports, now  reads:
     (C) The periods when monitoring data were not collected
     shall be specified; and
     §63.513(e) (5} (i) (A)  [§63.1335(e) (5)(i)(A)],  Notification
     of Compliance Status, now reads:

                             12-11

-------
     (A) ... For additional tests performed for the same kind
     of emission point using the same method,  the results and
     any other required information shall be submitted, but a
     complete test report is"not required.

     Comment:  Two commenters  (IV-D-18, IV-D-25) suggested
that §63.513(b)(7)(i) of the proposed rule be revised to state
that the start-up, shutdown, and malfunction plan is not
required to be routinely submitted to the Administrator.  One
commenter (IV-D-18) explained that these plans are massive and
that it would be better to allow these plans to remain at the
facility and be available for EPA inspection.   The commenter
(IV-D-18) requested that this requirement be moved from the
reporting section to the recordkeeping section of the rule.
     The other commenter  (IV-D-25) stated that the requirement
to document and report any deviation from the start-up,
shutdown, and malfunction plan required in subpart A should be
modified to only be required when the deviation results in the
underlying control/recovery/emission limit not being met.
     Response:  The commenters are correct that the start-up,
shutdown, and malfunction plan required by §63.10(d) (5) of
subpart A is not required to be submitted to the
Administrator.  The provisions of the final rule have been
changed to clarify that the start-up, shutdown, and
malfunction plan should be kept onsite.  If, at any time, the
start-up, shutdown, and malfunction plan is not followed, this
should be described and reported in the Periodic Report as
specified in  §63.513(e) (6) (iii) (E)  [§63.1335(e) (6) (ill) (E) ] .
     With respect  to the commenter's request to only document
and report deviations  from the start-up, shutdown, and
malfunction plan when  they result in a violation, the
requirements  in the  final rule have not been changed.  The EPA
considers the documentation and reporting of all deviations to
be necessary  for the EPA to evaluate the adequacy of the
start-up, shutdown,  and malfunction plan.  Because the owner
or operator is allowed to add  to the start-up,  shutdown, and
malfunction plan as  needed, deviations that occur frequently
can be  added  to the  plan  to avoid a repetitive  recordkeeping
                             12-12

-------
and reporting burden.  Further, while a specific deviation may
not result in a violation, it could be indicative of a problem
that would eventually lead to a violation.
     Comment:  One commenter  (IV-D-09) suggested that the
Implementation Plan and the Emissions Averaging Plan submittal
timing should not be tied to the affected sources operating
permit application timing.  The commenter (IV-D-09) noted that
proposed  §63.513(b)(3) and §63.513(b)(4) both reference the
"operating permit application" in their lead-in paragraphs.
The commenter  (IV-D-09) explained that this concept was lifted
from the  HON rule.  However, the commenter  (IV-D-09) explained
that this HON paragraph will most likely be revised by the HON
technical correction notice to be published later in 1995 to
exclude any reference to the operating permit application.
The commenter  (IV-D-09) stated that the original intent was to
tie the submittal of the Implementation Plan or Emissions
Averaging Plan to a certain time period measured from the
compliance date, not an earlier date which might be imposed by
an operating permit application submittal.  The commenter
(IV-D-09) explained that the owner or operator needs time to.
determine, plan, and design the means to achieve compliance
with the  rule, hence the practical timing deadlines in the
amended HON make sense.
     The  commenter  (IV-D-09) went on to explain that there may
be times  when the due date for submittal of an Implementation
Plan is approximately the same as for submittal of an
operating permit application.  The commenter (IV-D-09)  stated
that under these circumstances, the owner or operator,  in
agreement with the permit authority, may submit the
Implementation Plan information. as part of the operating
permit application.  The commenter (IV-D-09) explained that in
this case, the provisions of proposed §63.513 can and should
be used to guide the applicant on the information to submit.
The commenter  (IV-D-09) clarified that removing the reference
to the operating permit application from proposed
§63.513 (b) (3), Implementation Plan, would not forgo this
combined  submittal opportunity.
                             12-13

-------
     Response:   In an attempt to reduce the reporting
contained in the proposed rule,  the EPA has removed the
Implementation Plan requirements from the final thermoplastics
rule.  The EPA anticipates that the affected sources at a
plant site subject to this rule would already have submitted
an operating permit containing the same information.
     The provisions for the Emissions Averaging Plan exclude
any reference to timing of submittal of the operating permit
application.  Instead, the final rule requires submittal of
the plan no later than 18 months before the compliance date.
As the commenter stated, this prevents the Emissions Averaging
Plan from having to be submitted by an earlier date imposed by
the operating permit  application.
                              12-14

-------
    13 . 0   POLYMER MANUFACTURING NSPS ADOPTED  PROVISIONS FOR
   CONTINUOUS PROCESS VENTS AT PET AND POLYSTYRENE AFFECTED
              SOURCES USING CONTINUOUS PROCESSES
13.1  SELECTION OF STANDARDS
     Comment:  Four commenters  (IV-D-08, IV-D-15, IV-D-18,
IV-D-24) supported the proposal to modify the requirements
specified in the Polymer Manufacturing NSPS  (Part 60,
Subpart ODD) to eliminate references to polystyrene and PET.
     Response:  The EPA agrees that affected sources should
not be subject to the thermoplastics rule and another rule,
and at proposal the EPA proposed to remove all reference to
polystyrene and PET from the Polymer Manufacturing NSPS.
However, upon review, the EPA has realized that the Polymer
Manufacturing NSPS applies to both area and major sources.
The EPA did not intend to forego the regulation of new or
modified/reconstructed area sources producing polystyrene and
PET.  As a result, the EPA has withdrawn the proposed
revisions to the Polymer Manufacturing NSPS.  In the final
rule, §63.501(i)  [§63.1311(1)] specifies that affected sources
(i.e., sources subject to the thermoplastics rule) producing
PET or polystyrene that are also subject to the Polymer
Manufacturing NSPS are required to comply only with the
thermoplastics rule and are no longer subject to the Polymer
Manufacturing NSPS.
     Comment:  Two commenters  (IV-D-08, IV-D-24) requested
that polystyrene affected sources using a continuous process
be given the maximum amount of flexibility by allowing them
the option to comply with either proposed §63.505(a), which is
based on the HON, or proposed §63.505(c), which is based on
the Polymer Manufacturing NSPS.  The commenters  (IV-D-08,
IV-D-24) stated that both are equivalent to the MACT floor, as
                             13-1

-------
determined by the EPA, and, therefore, both are appropriate
compliance options.
     Response;  The final rule continues to require continuous
process vents in the material recovery section at continuous
polystyrene affected sources to meet the provisions based on
the Polymer Manufacturing NSPS (i.e., §63.505-2 [§63.1316]).
The EPA disagrees with the commenter's statement that the
requirements for continuous process vents from the HON and the
Polymer Manufacturing NSPS are equivalent.   In fact, the EPA
found the Polymer Manufacturing NSPS requirements to have the
potential to be more stringent than the HON (see Chapter 6.0
of Basis and Purpose Document  [EPA-453/R-95-004a, March
1995]) .  Further, because the Polymer Manufacturing NSPS was
found to be equivalent to the MACT floor for these affected
sources  (also discussed in Chapter 6.0), the EPA must require
compliance with the Polymer Manufacturing NSPS requirements in
order to set a standard at least as stringent as the MACT
floor.
     Comment;  Three commenters  (IV-D-08, IV-D-18, IV-D-24)
objected to the proposed requirement to control continuous
process vents from material recovery sections at continuous
polystyrene affected sources regardless of their emission
rate.  One commenter  (IV-D-29) stated there were no
applicability criteria for these continuous process vents in
the proposed rule.  Another commenter  (IV-D-18) encouraged the
EPA to implement a group determination procedure for
continuous process vents in the material recovery section of
continuous polystyrene affected sources.
     One commenter  (IV-D-18) provided an example illustrating
their objection.  The commenter  (IV-D-18) stated that a
process condenser  is used  to collect liquids in the material
recovery section of a continuous polystyrene process. . The
commenter  (IV-D-18) explained  that the system normally
operates at a negative pressure, which is achieved by a small
vacuum pump.  In order to  prevent air from being pulled into
the system by the vacuum,  a continuous stream of nitrogen is
used to blanket the tank below the process condenser.  As a
                             13-2

-------
result of the nitrogen blanket, the exhaust of the vacuum pump
contains a high concentration of nitrogen along with a small
amount of HAP.  If the dilute exhaust vent on the vacuum pump
required emission controls, the equipment would be very large
and very costly.  The cost of controlling the dilute stream
would be significantly out of proportion with respect to the
minimal amount of emission reduction that would be achieved.
     Response;  The proposed and final rules, which are
modeled after the provisions in the Polymer Manufacturing
NSPS, do not require each continuous process vent in the
material recovery section to be controlled.  The requirements
of the proposed  [§63.506(c)] and final rule  [§63.1316(c)]
provide for a compliance option of maintaining emissions from
continuous process vents within the material recovery section
at or below an emission rate of 0.0036 kg organic HAP/Mg of
product.  This may allow some continuous process vents to go
uncontrolled as long as the overall emission rate of the
material recovery section is at or below 0.0036 kg organic
HAP/Mg of product.  However, other compliance options that
allow an owner or operator to potentially take advantage of
existing control devices do require the control of each
individual continuous process vent.  Paragraphs
§63.505-2(c) (1) (iii) (B) and (C) [§63.1316(c) (1) (iii) (B) and
(C) 3 of the final rule allow an owner or operator to comply by
using a boiler/process heater or a flare.  An owner or
operator choosing to comply with either of these paragraphs
would be required to control each and every continuous process
vent within the material recovery section.
     The proposed and final rules do not contain applicability
criteria for continuous process vents from continuous
polystyrene affected facilities because the MACT floor was
found to be equivalent to the emission limits in the Polymer
Manufacturing NSPS  (see Chapter 6.0 of the Basis and Purpose
Document  [EPA-453/R-95-004a, March 1995]).   Because the
emission limits from the Polymer Manufacturing NSPS are part
of the MACT floor, all affected sources are required to meet
the emission limits, regardless of the cost effectiveness of
                             13-3

-------
controlling any one continuous process vent.  On the other
hand, the proposed and final provisions for continuous process
vents within material recovery sections from existing
continuous PET affected sources using a dimethyl terephthalate
process do contain applicability criteria.  For these
continuous process vents, the MACT floor was found to be less
stringent than the requirements in the Polymer Manufacturing
NSPS.  Because the Polymer Manufacturing NSPS requirements are
more stringent than the MACT floor for this subcategory, they
have been implemented with the use of applicability criteria
to ensure that only those material recovery sections that are
cost effective to control are required to be controlled.
     Comment:  Two commenters  (IV-D-08, IV-D-11) suggested a
change to the provisions that allow compliance by limiting the
final condenser  outlet temperature.  The commenters  (IV-D-08,
IV-D-11) explained that one of the compliance options, which
is based on  the  Polymer Manufacturing NSPS, is to maintain a
final C9ndenser  outlet temperature of  -25  °C.  One commenter
 (IV-D-11) explained  that the  reporting provisions from the
Polymer Manufacturing NSPS  state that this  parameter boundary
 (i.e., -25  °C  limit)  is  only  exceeded when  the average
condenser operating  temperature for a  3-hour period is more
than 6 °C above  (i.e., warmer) the average  operating
temperature  established  during the most recent performance
test at which  compliance was  demonstrated.
      Three  commenters (IV-D-08, IV-D-15,  IV-D-24) requested
that the  final thermoplastics rule incorporate those
monitoring,  test method,  and recordkeeping and reporting
requirements from the Polymer Manufacturing NSPS that provide
this flexibility (i.e,  the six degree  window) .  Additionally,
one commenter (IV-D-24)  provided  their interpretation and
understanding of the monitoring,  test  method, and
recordkeeping and reporting provisions of the Polymer
Manufacturing NSPS.   This commenter  (IV-D-24) requested  that
 their understanding be included  in the promulgation preamble.
      One commenter (IV-D-11)  explained that affected sources
 using the final condenser outlet  temperature compliance  option
                              13-4

-------
have experienced severe plugging and icing problems with their
condensers during continuous operation below -25 °C.  However,
they are able to maintain effective operations when operating
within a 6 °C window of approximately -25 to -19 °C.  The
commenter  (IV-D-11)- explained that because this six degree
window provision is not included in the proposed rule, the
proposed rule is more stringent the Polymer Manufacturing
NSPS.  The commenter  (IV-D-11) requested that the proposed
thermoplastics rule  [§63.505(c)(2)(i)]  be changed to state
that the final condenser outlet temperature should never
exceed a limit of -19 °C.
     Response:   The EPA intended for the requirements of the
proposed rule to be equivalent to the Polymer Manufacturing
NSPS.  The EPA has revised the final rule to provide the
flexibility found in the Polymer Manufacturing NSPS (i.e., the
six degree window) for determining compliance with the final
condenser outlet temperature limit.
     In addition, the EPA has disassociated the six degree
window from the results of the performance test and has
instead associated it with the temperature limit of -25 °C.
The final rule allows all owners or operators complying with
the final condenser operating temperature limit to use -19 °C
as the limit for their 3-hour averages.  The EPA considered
that as proposed, the provisions did not achieve an
even-handed implementation because some affected sources would
be allowed to have 3-hour averages at -19 °C while others
would be limited to some colder value (e.g., -21 °C) because
their performance test results indicated a temperature colder
than -25 °C (e.g., -27 °C) .  The EPA also added other items
from the Polymer Manufacturing NSPS monitoring, test method,
and recordkeeping and reporting provisions as appropriate.
These items appear in §§63.505-3 through 63.505-6 [§§63.1317
through 63.1320] of the final rule.  The EPA chose to add
provisions to the thermoplastics rule as opposed to
incorporating the monitoring,  test method, and recordkeeping
and reporting provisions from the Polymer Manufacturing NSPS
or referring to them; the EPA did not want affected sources
                             13-5

-------
subject to both §63.505-1 and §63.505-2 [§63.1315 and
§63.1316] to be required to comply with two sets of similar,
but not identical, monitoring, test method, recordkeeping, and
reporting provisions.
     Comment;  Two commenters (IV-D-17, IV-D-20) supported
including the Polymer Manufacturing NSPS adopted provisions in
the proposed rule provided that only organic HAP are regulated
and not total organic compounds.  The commenters (IV-D-17,
IV-D-20) contended that requiring the regulation of only
organic HAP provides greater control flexibility at lower
costs, while achieving equivalent emission reductions.  The
commenters  (IV-D-17, IV-D-20) stated that industry is already
familiar with their suggested approach and initial and ongoing
compliance and monitoring costs will be less with this
approach.
     Another commenter  (IV-D-18) asked the EPA  to remove  the
requirements to express total organic compound  emissions  from
polystyrene TPPUs as the sum of the actual compounds and  allow
the use'of  carbon equivalents instead.  The commenter
 (IV-D-18) maintained it can be very difficult to analyze  and
speciate  all the components in a gas stream.  The commenter
 (IV-D-18) explained that Method 18, which  is required by  the
proposed rule,  is the method normally used when speciating
compounds.  However, the commenter  (IV-D-18) stated that  this
method is very  difficult to use if the stream has numerous
components.  Additionally, the commenter  (IV-D-18) contended,
the burden  of speciating each compound, just to add them  all
together again,  is  substantial and is not  supported by a
corresponding environmental benefit-.  The  commenter  (IV-D-18)
explained that  Method  25A  is  an equivalent method that could
be used to  determine non-speciated total organic compound
emissions.   The commenter  (IV-D-18) noted  that  Method 25A is
accepted by EPA in  numerous NSPS regulations and can provide  •
the same assurance  of  compliance, even though it does not
quantify each individual  compound.  Based  on further contact
with the commenter,  the EPA understands the primary concern of
                              13-6

-------
the commenter to be the potential requirement to conduct
speciated testing for an unending number of compounds.
     Response:  The EPA has restricted the provisions from the
Polymer Manufacturing NSPS, as adopted for the thermoplastics
rule  [§63.505-2  [§63.1316]], to regulate only organic HAP in
order to stay within the mandate of the MACT standards
program, which is to regulate organic HAP.
     Because, with one exception, the best controlled existing
sources were meeting the requirements of the Polymer
Manufacturing NSPS, these requirements represent the MACT
floor for these continuous process vents.  Because the MACT
floor analysis only considered the control of organic HAP and
the best controlled existing sources were meeting the emission
limits in the Polymer Manufacturing NSPS, these emission
limits have been retained without modification,  even though
they are being used, to regulate organic HAP and not TOG.  For
the one exception, which is the set of continuous process
vents within the material recovery section at existing
affected sources producing PET using a continuous DMT process,
the threshold emission rate, which serves as an applicability
cutoff, was required for implementing the Polymer
Manufacturing NSPS emission limits in the thermoplastics rule.
Because the available emissions data show that the regulated
organic HAP are the only emissions from these continuous
process vents, the threshold emission rate value was not
changed and is expressed as kg organic HAP/Mg product.
     Addressing the testing concerns raised by one commenter,
the definition of organic HAP in the final rule  specifies the
organic HAP that are regulated for each subcategory.   These
and other organic HAP identified by the owner or operator of
an affected source are the only organic HAP regulated by the
rule for the subcategory.  The previous definition of organic
HAP included all compounds from the CAA list of  189 that were
commonly considered to be organic.   It is this previous
definition that led the commenter to consider the potential
testing burden to be unreasonable.
                             13-7

-------
     One commenter suggested allowing the use of Method 25A as
a way to alleviate the testing burden.  As discussed
previously in Chapter 1.0, Section 1.2, the EPA has judged it
reasonable to allow compliance demonstrations and some
applicability determinations to be based on emissions of total
organic compounds  (TOO .  As part of adding this flexibility,
the EPA also has allowed  the use of Method 25A, which measures
TOC.  The final rule allows owners or operators to make
applicability determinations and demonstrate compliance with
the Polymer Manufacturing NSPS adopted provisions  [§63.505-2
 [§63.1316]] based  on TOC, and the use of Method 25A is
allowed.
     Comment:  One commenter  (IV-D-10) requested clarification
on the intent of the term "limit" when used in proposed
§63.505(c)(2)(i) and  (c)(2)(ii).  The commenter  (IV-D-10) was
uncertain whether  the  requirement is  zero emissions or
emission reduction, as specified in the HON.
     Response;  The EPA intended for  the term  "limit" to be
associated with the emission  standards presented in the same
paragraph or the emission standards referred to by the same
paragraph.   In both proposed  §63.505(c)(2)(i)  and  (c)(2)(ii),
the text goes on to  identify  an emission standard.  For
example, §63.505(c) (2) (i) (A)  of the proposed rule  specifies
that emissions  shall  not exceed 0.0036 kg TOC/Mg of product.
Alternatively,  §63.505(c) (2) (ii) of the proposed rule refers
 to  another paragraph within the section which  itself refers  to
 emission standards outside of the proposed  subpart (e.g.,  the
 HON provisions  for continuous process vents  in subpart G).   In
 either case,  the EPA means for the  term  "limit"  to mean  "limit
 emissions per the  emission standards  identified by the
 paragraph."
      Comment:   One commenter (IV-D-10) requested clarification
 on the intent of the term "limit" when used in proposed
 §63.505 (c) (2) (ii)  .  The commenter (IV-D-10)  was uncertain
 whether the requirement to limit  emissions  from batch  and
 other continuous process vents, and any batch process  vent
 from the material recovery section,  suggests that  all  of  these
                              13-8

-------
process vents would have to be physically separated and
controlled to demonstrate compliance.
     Response;  The paragraph cited by the commenter, now
§63.505-2(c)(2) and  (c)(3)  [§63.1316] in the final rule,
requires the owner or operator to control continuous process
vents not subject to proposed §63.505(c)(2)(i)
[§63.505-2(c)(1)  [§63.1316] of the final rule]  by complying
with the HON provisions.  Proposed §63.505(c)(2)(ii) also
requires the owner or operator to control batch process vents
by complying with proposed §63.506.  The EPA does not see a
basis for the commenter's concern that the cited paragraph
requires all of the process vents in the TPPU to be physically
separated and controlled.  Neither proposed §63.506, the batch
process vent provisions, nor the HON provisions specify that
individual process vents must be "physically separated and
controlled."  These provisions do require that owners or
operators demonstrate compliance for individual process vents,
but they do not require that unique control devices be used to
achieve control.
13.2  APPLICABILITY
     Comment:  Two commenters (IV-D-17, IV-D-20) stated that,
according to their interpretation of the definition of
"affected source" as applied to the PET process, the affected
source would be the collection of all process units at a site
that produce PET from TPA or DMT using either a batch or
continuous process.  Further, the commenters (IV-D-11,
IV-D-17, IV-D-20) stated that the continuous process vent
provisions of proposed §63.505 (b) specify that the emission
limits apply to owners or operators of an affected source.
The commenters  (IV-D-11, IV-D-17, IV-D-20)  stated that this
regulatory construction implies that the continuous process
vent emission limits apply to the collection of all material
recovery sections, raw material preparation sections, and
polymerization sections, respectively,  within an affected
source and not to each individual line as under the Polymer
Manufacturing NSPS.  The commenters  (IV-D-11, IV-D-17,
                             13-9

-------
IV-D-20)  supported this approach and believed  it should be
clarified in the  final rule.
     The  commenters (IV-D-11, IV-D-17,  IV-D-20) stated that
applying  the emission limits more broadly will simplify the
compliance demonstration, monitoring,  recordkeeping,  and
reporting requirements of the rule.  Further,  this will
simplify the administrative burden and lower compliance costs
while  achieving the same level  of emission  reductions.
Additionally,  the commenters  (IV-D-11,  IV-D-17, IV-D-20)
stated, this approach will avoid the burden of allocating
emissions from shared vessels to individual lines; often,
vessels in the material recovery and/or raw material
preparation sections of PET plants serve more  than one
continuous polymerization line.  Also,  explained the
commenters (IV-D-11, IV-D-17, IV-D-20),  because the control
requirements  in this rule are the same for  new and existing
affected  sources, this broader  definition will result in  the
same level of  emission reduction as  line-by-line compliance.
      Response;  The EPA disagrees with the  commenters
interpretation of the proposed  provisions,  but the EPA agrees
'that a broader implementation of the Polymer Manufacturing
NSPS adopted  provisions will simplify  compliance while
achieving the same emission reductions.   Therefore, the final
rule applies  the emission  limits adopted from  the Polymer
Manufacturing NSPS  [§63.505-2  [§63.1316]] to the collection  of
all material  recovery sections, raw  material preparation
sections, or  polymerization reaction sections, as appropriate,
within an affected source.  For example, a  plant site has five
polystyrene TPPUs using a  continuous process,  all with
independent material recovery sections.   Because all  five
TPPUs are part of the same  affected  source, the emission  limit
applies  to the collection  of  continuous process vents within
the five material recovery sections.   [Note:   Chapter 2.0
discusses the concept of  affected  source in detail.]  In
another example,  a plant  site has  three continuous .PET TPPUs.
One TPPU uses a DMT process and has  independent process
 sections  (i.e., material recovery  and  polymerization
                             13-10

-------
reaction).  The other two TPPUs use a TPA process and share a
raw material preparation section and have independent
polymerization reaction sections.  At this plant site, there
are four process sections subject to the provisions.  Because
the PET TPA TPPU and the PET DMT TPPU are separate affected
sources, the owner or operator cannot combine the
polymerization reaction sections from all three TPPUs.
Therefore, there are two process section within the PET DMT
TPPU (i.e., material recovery section and polymerization
reaction section), and there are two process sections within.
the two PET TPA TPPUs  (i.e., raw material preparation section
and polymerization reaction section).
     The misinterpretations of the definition of affected
source and the applicability of the Polymer Manufacturing NSPS
adopted provisions represented in the provided comments are
addressed below.
     The EPA acknowledges that the definition of affected
source in the proposed rule was unclear.  As discussed in
Chapter 2.0 of this document, the definition of affected
source and the applicability provisions of the rule have been
revised.  Paraphrasing the final rule, the affected source has-
been defined as the group (i.e., one or more) of TPPUs
manufacturing the same thermoplastic product as their primary
product; the definition of TPPU states that the primary
product of a TPPU is a thermoplastic product; and
thermoplastic product is defined as one of the 19
subcategories covered by this rule.  In other words, an
affected source contains one or more TPPUs producing the same
thermoplastic product  (i.e., one of the 19 subcategories)  as
their primary product.  With this understanding of the
affected source, the commenters' statement that the affected
source would be "the collection of all process units at a site
that produce PET from TPA or DMT using either a batch or
continuous process" would no longer be a valid interpretation;
the commenter is mixing subcategories within an affected
source.  Using the commenters' words, the affected source
would be, for example,  (1) the collection of all process units
                             13-11

-------
at a site that produce PET from DMT using a continuous process
or (2) the collection of all process units at a site that
produce PET from TPA using a continuous process.  Under the
final rule, the affected source would not be the combination
of (1) and  (2) .
     The commenters stated that the emission limits adopted
from the Polymer Manufacturing NSPS apply to the affected
source.  The EPA acknowledges that this portion of the rule
was unclear.   In the Polymer Manufacturing NSPS, the
provisions apply to each affected facility, which is defined
as each process section  (e.g., material recovery section).
The intent of  the proposed rule was to be similar to the
Polymer Manufacturing NSPS and require each process section to
meet the emission limits.  However, as stated above, the final
rule will apply the emission limits to the affected source
 (i.e., the collection of process sections within an affected
source).
     Finally,  by stating that  "because the control
requirements are the same for new and existing sources. . . will
result in the  same level of emission reduction as line-by-line
compliance," the commenters inferred that existing and new
TPPUs will be  part of the same affected source.  The EPA has
determined  that existing and new TPPUs will not be part of the
same affected  source; the provisions of proposed §63.500(g)
were not clear on this point.  The final  [§63.1310 (i)] rule
states that if a TPPU with the same primary product as TPPUs
already present at a plant site is added to the plant site and
the added TPPU emits greater than 25 tpy of organic HAP or
greater than  10 tpy of any individual organic HAP, the added
TPPU  is a  separate affected source.
      Comment:  Two commenters  (IV-D-17, IV-D-20) requested
clarification on which provisions apply to paste feed tanks
and additive  mixing tanks.  The commenters  (IV-D-17, IV-D-20)
explained  that these tanks are involved in preparing raw
materials  and catalysts  for polymerization.  One commenter
 (IV-D-11)  stated that paste feed tanks and additive mixing
tanks  at thermoplastic manufacturing facilities could be
                             13-12

-------
 regulated as process vents  under proposed §63.505(b)
 (i.e.,  the Polymer Manufacturing NSPS adopted provisions)  or
 as surge control vessels  under proposed §63.509  (i.e.,  the
 equipment leak provisions).
      Another commenter (IV-D-07)  requested modification of the
 definition of raw materials preparation section.   The
 commenter (IV-D-07)  stated  that the definition excludes
 certain operations that are traditionally considered part  of
 feed preparation.   For example,  the commenter (IV-D-07)
 explained that mixing and dissolving should be added to the
 list of activities specified in the definition.
 Alternatively, the commenter (IV-D-07)  suggested making it
 clear that the unit operations list includes, but  is not
 limited to,  the activities  cited.
      Response;  Two issues  related  to the definition of raw
 material preparation section were raised by the commenters.
 First,  commenters  expressed concern that paste feed tanks  and
 additive mixing tanks may be regulated as process vents under
 the  Polymer Manufacturing NSPS adopted provisions or as  surge
 control vessels under the equipment leak provisions of  this
 rule (i.e.,  §63.509  [§63.1331]).  For two reasons,  these tanks
 will not be regulated under  the  Polymer Manufacturing NSPS
 adopted provisions.   First,  these tanks do not meet the
 definition of continuous  process vent,  which are the only
 emission points regulated by said provisions.  Second,  careful
 reading of the emissions  standard for the raw material
 preparation section  [§63.505-2(b) (2) (i)  [§63.1316(b) (2) (i)]]
 reveals that only  emissions  from the  continuous process vents
 associated with the  esterification  vessels are subject to the
 emission limit.  On  the other hand, it  is possible  that these
 tanks could be regulated  as  surge control vessels,  but  the
 commenter did not  provide enough information to make  a
definitive assessment.
     Second,  one commenter requested  that the definition of
raw material preparation  section be expanded to include
activities such as mixing and dissolving.  Because  the  only
emissions  regulated  are those from  the continuous process
                             13-13

-------
vents associated with the esterification vessels, the EPA does
not find it necessary to specifically include the activities
of mixing and dissolving in the definition.  However, the EPA
does not consider the definition to be written to exclude any
activities and considers the phrase "otherwise treated" to be
adequate to include these activities.
     Comment;  One commenter  (IV-D-10) stated that the
definitions of raw materials preparation section,
polymerization reaction section, and material recovery  section
do not  adequately define the limits of the process to be
regulated.
     For example, one  commenter  (IV-D-10)  stated that the
definition of  "storage vessel" would suggest that raw material
storage tanks might be considered  part of  the affected  source.
However, the proposed  definition for  "raw  materials
preparation  section" specifically  excludes "...the  storage  of
raw materials."   The commenter (IV-D-10) believed the EPA does
intend that  raw material storage vessels be  part of  the
affected source and suggested that one way of  including these
storage vessels is  to  modify the definition of  "thermoplastic
product process unit"  as follows:  ..."connected by  pipes and
ducts  to store,  deliver and process raw  material...".
      Response':   The purpose of the "section" definitions is to
 define those equipment subject to  the Polymer Manufacturing
NSPS adopted provisions.  The purpose of these definitions is
 not to define the entire process or affected source.  In fact,
 the Polymer Manufacturing NSPS requirements do not  cover large
 portions of the process.  For example,  the raw materials
 preparation section is not regulated at PET affected sources
 using a DMT process.  The final rule has not been modified to
 address this specific comment.
      Comment:  One commenter  (IV-D-04)  stated that it appears
 the EPA has not considered which emission limitation applies
 to continuous process vents associated with recovery equipment
 that is clearly part  of the material recovery section but
 which  is als.o used to recover ethylene glycol from the
 polymerization reaction section.  The commenter (IV-D-04)
                              13-14

-------
stated that it appears the EPA has envisioned two distinct
recovery operations; one servicing the polymerization reaction
section and one servicing the material recovery section.
When, in fact, the recovery operations in the materials
recovery section may also be used to purify material
(e.g., ethylene glycol) from the polymerization reaction
section.  For example, the commenter  (IV-D-04) explained,
there may be instances where solvent recovered from the
polymerization reaction section is sent directly to the
material recovery section, mixed with other material
(e.g., methanol) in a single vessel, and then sent to a series
of stills designed to separate the methanol and ethylene
glycol from the recovered process streams.   The commenter
(IV-D-04) recommended that the EPA either establish a process
for determining which emission limitation applies when
material from the polymerization reaction section is recovered
in the material recovery section or make continuous process
vents from these "shared" operations subject to proposed
§63.505(a) (i.e., the HON provisions).  The commenter
(IV-D-04) stated that if the EPA decided to establish a
process for determining which emission limitation applies,
such a process be submitted for public comment prior to
promulgation of the final rule.  [Note:  The Polymer
Manufacturing NSPS did not identify this particular situation
when being developed.]
     Response;  The EPA agrees with the commenter that, as
defined in the proposed rule, it was unclear in which process
section these continuous process vents should be placed.
     The EPA considered two options for dealing with this
specific situation and others like it.  First, allow the
affected source and permitting authority to work out a
case-by-case answer.  Second, considering the definitions of
the'process sections involved and the relative stringency of
the emission limits for each process section, the EPA could
designate the described continuous process vents as being part
of a particular process section.  The EPA considered the first
option and decided not to select it because it would be
                             13-15

-------
difficult to ensure an even-handed implementation of the
provisions, and it would increase the burden on the affected
source and the permitting authority.  The EPA considered the
second option and judged that designating a particular process
section was reasonable.  The EPA considered the relative
stringency of the two emission limits as the primary factor in
making a selection because it would provide the most
protection to the environment.  The EPA considered the
emission limit for the polymerization reaction section
 (0.020 kg organic HAP/Mg of product) to be more stringent than
the emission limit for the material recovery section
 (0.018 kg organic HAP/Mg of product) because there is no
threshold emission rate, which serves as an applicability
cutoff, associated with the polymerization reaction section.
Therefore, the EPA decided that equipment that would be
included in both the material recovery section and the
polymerization reaction section based on the definitions would
be subject to the provisions for:the polymerization reaction
section.  The definition of material recovery section has been
modified to provide for this determination.  The commenters
requested that if a process for determining the applicability
of recovery equipment was developed, that the EPA submit it
for public comment.  Because the EPA has made a set decision
as opposed to developing a decision-making process, the EPA
does  not consider it necessary to submit it for public
comment.
      Comment:  Three commenters  (IV-D-17, IV-D-20, IV-D-22)
requested clarification of the phrase "...other parts of the
process..."  in proposed §63.505(b) (2) (iii) and  (b) (3) (iii) of
the continuous process vent provisions.  One commenter
 (IV-D-22)  stated that this phrase offers opportunity for
misinterpretation.  The commenter  (IV-D-22) suggested the
phrase  "auxiliary equipment" as an alternative and requested
 further clarification in §63.502, definitions.  Two commenters
 (IV-D-17,  IV-D-20)  stated that post-polymerization steps
 (e.g.,  spinning or  extrusion) should be excluded from the rule
                             13-16

-------
to specifically eliminate inconsistent interpretation of the
phrase "other parts of the process."
     Response:  . The EPA considers the proposed [§63.505(b)
and (c)]  and final rules [§63.505-2 [§63.1316]] to be clear on
what process vents are covered by the Polymer Manufacturing
NSPS adopted provisions.  The definitions of raw materials
preparation section, polymerization reaction section, and
material recovery section,  which were discussed previously,
describe the purpose of and/or process for the equipment
included in each process section; only continuous process
vents that are part of one of these three process sections are
covered by the Polymer Manufacturing NSPS adopted provisions.
By default, those process vents not covered by the Polymer
Manufacturing NSPS adopted provisions are subject to §63.505-1
or §63.506 [§63.1315 or §63.1321],  as appropriate.  For
example,  process vents in the raw materials preparation
section of a continuous PET affected source using a DMT
process are not covered by the Polymer Manufacturing NSPS
adopted provisions and would be covered by either §63.505-1 or
§63.506 [§63.1315 or §63.1321].
     To improve the clarity of these provisions,  the EPA has
restructured §63.505 of the proposed rule, which implements
the HON provisions and the Polymer Manufacturing NSPS adopted
provisions.  The provisions contained in §63.505 of the
proposed rule are found in §§63.505-1 through 63.505-6
[§§63.1315 through 63.1320] of the final rule.  While this
restructuring serves many purposes, one of them is that the
restructured provisions provide an opportunity to clarify what
is meant by "process vents from other parts of the process."
For example, in the final rule if paragraph x imposes a
Polymer Manufacturing NSPS adopted emission limit,
subparagraph y will say "continuous process vents not subject
to paragraph x are subject to §63.505-1 [§63.1315]" (i.e., the
HON requirements), and subparagraph z will say "batch process
vents are subject to §63.506 [§63.1321]."
     The final rule implements the HON requirements in
§63.505-1  [§63.1315], the Polymers Manufacturing NSPS adopted
                             13-17

-------
emission limits in §63.505-2 [§63.1316], and monitoring,
testing, recordkeeping, and reporting requirements specific to
the Polymer Manufacturing NSPS adopted provisions in §63.505-3
through 63.505-6  [§63.1317 through 63.1320].
     Commenters requested that post-polymerization steps, such
as spinning and extrusion, be specifically excluded from the
rule.  These steps were never intended to be covered by the
rule and to ensure clarity, these processes were exempted from
the affected source in §63.500(d)  [§63.1310(d)].
     Comment;  One commenter (IV-D-07) considered the
definition of material recovery section to be confusing.  The
commenter  (IV-D-07) noted that as part of their polystyrene
materials recovery they have several tanks that store
recovered styrene condensate.  They expressed concern that the
definition of "materials recovery section" could be
interpreted to classify these storage tanks as process vents.
The commenter (IV-D-07) suggested clarifying the definition
by:  (1) deleting the reference to process section and process
line unless these terms are defined,  (2) ensuring that storage
tanks and .surge control vessels are not captured by the
material recovery section definition, (3) clarifying the
boundaries between the raw material preparation,
polymerization reaction, and material recovery sections, and
(4) removing any disincentives to pollution prevention and
internal recycling in the definition.
     Response;  The EPA has revised the definitions related to
the Polymer Manufacturing NSPS adopted provisions to address
this commenter's concerns, as well as other issues found
during  internal EPA review. ' The EPA considers that these
modifications address three of the commenter's concerns.
First,  process section has been defined in the final rule, and
the term process  line is no longer used in the definitions of
raw material preparation section, material recovery section,
and polymerization reaction section.  Second, an exemption has
been added to the definition of material recovery section to
exclude "equipment used to store recovered materials."  Third,
in response to a previous comment, the EPA modified the
                             13-18

-------
definition of material recovery section to clarify the
boundary between the material recovery section and the
polymerization reaction section, specifically related to
equipment used to recover ethylene glycol and other
(i.e., methanol) materials.  In addressing the fourth comment,
the EPA does not find any disincentives to pollution
prevention or internal recycling to exist in the definitions.
     The EPA considers these modifications sufficient to
address the commenter's general concerns, as well as their
specific concern with tanks storing recovered condensate.
13.3  DEMONSTRATION OF COMPLIANCE
     Comment:  One commenter (IV-D-21) recommended that the
final rule allow the use of engineering assessments to
demonstrate compliance with the Polymer Manufacturing NSPS
adopted provisions, as allowed in §63 .115 (d) (1) of the HON.
[Note:  §63.115(d) (1) of the HON allows an owner or operator
to use engineering assessments to demonstrate that the TRE
value  (i.e., the applicability criteria) for a continuous
process vent is greater than 4.O.]  The commenter (IV-D-21)
noted that the proposed rule requires testing to determine
applicability of the standard.  The commenter  (V-D-21)
explained this would mean testing would be required even if
credible engineering assessments show that emissions are well
below the threshold emission rate.  The commenter (IV-D-21)
asserted that this extensive testing is unnecessary and
imposes unjustified expense and delay on polymer and resin
production facilities.   [Note:  Proposed §63.505(b) (2) (i)
contains a threshold emission rate.  If emissions from the
material recovery section at an existing PET affected source
using a DMT continuous process are at or below this value,
control of continuous process vent emissions is not required.
Process §63.511 (d), which refers to proposed §63.505(b) (2) (i> „
requires a performance test based on Method 18 to determine
whether emissions for an affected source are above or below
the threshold emission rate.]
     Response;  The EPA agrees with the commenter that,  in
certain circumstances, engineering assessments should be
                             13-19

-------
allowed for determining applicability.  Therefore,  the EPA has
modified the final rule to allow the use of engineering
assessments to demonstrate that emissions from the material
recovery section at existing PET affected sources using a DMT
continuous process are less than or equal to the threshold
emission rate of 0.12 kg organic HAP/Mg of product in certain
circumstances.  Like the HON, which only allows the use of
engineering assessment to demonstrate that a continuous
process vent is not subject to the provisions within a certain
range  (i.e., a TRE value greater than 4.0), the final rule
only allows the use of engineering assessments to demonstrate
that emissions are less than or equal to 0.07 kg organic
HAP/Mg of product.  If engineering assessment shows that
emissions are greater than 0.07 kg organic HAP/Mg of product,
and the owner or operator wishes to demonstrate that emissions
are below the threshold emission rate of 0.12 kg organic
HAP/Mg of product, testing is required as  specified in
§63.505-4(b)  [§63.1318(b)].  The value of  0.07 kg organic
HAP/Mg of product is the approximate midpoint between the
emission limit  (i.e.,  0.018 kg organic HAP/Mg of product) and
the threshold emission rate.  The EPA felt this value provided
an appropriate measure of flexibility to owners and operators
while ensuring compliance with the standard.
     Comment;  One commenter  (IV-D-09) objected to the
requirement in proposed §63 .511 (d) (2) that production rate be
determined by direct measurement of polymer pulled.  The
commenter  (IV-D-09) explained that this would require the
polymer produced during the testing to be  weighed.  The
commenter  (IV-D-09) went on to say this would result in an
unnecessary and  additional cost because production rates are
not normally  obtained  through direct product measurement.  The
commenter  (IV-D-09) explained that production rate is normally
determined by raw material charges into the process reactor or
is based on  extruder throughput.  The production rate can be
easily and  accurately  calculated based on  raw material charge
weights  and extruder yield histories.  The commenter  (IV-D-09)
recommended that the normal method for determining production
                             13-20

-------
rate should be allowed without prior approval of the
Administrator.
     Response;  The EPA agrees with the commenter, and the
final rule allows the rate of polymer produced to be
determined by alternate means without prior approval of the
Administrator.  To support this change, the final rule
requires that the owner or operator selecting an alternate
means of determining polymer produced submit the methodology
to be used, including all data, assumptions, and procedures
required by the methodology, as part of the Notification of
Compliance Status.
13.4  MISCELLANEOUS
     Comment:  Two commenters  (IV-D-17, IV-D-20) stated that
proposed §63.505(b)(1) is redundant to proposed §63.505(b)
and, therefore, a more descriptive introduction is needed in
proposed §63.505(b)(1) to reduce confusion.
     Response;  As discussed in a previous response, §63.505
of the proposed rule has been restructured to improve the
clarity of these provisions.  This restructuring has addressed
the commenter's concern.
                             13-21

-------

-------
                           14.0 OTHER
     Comments  that  did not  fit into the other chapters have
been summarized and responded to here.
     Comment;   One  commenter  (IV-D-18) supported the
differentiation between Group 1 and Group 2 emission points
for storage vessels, process vents, and wastewater streams.
The commenter  (IV-D-18) stated that it made the rule more cost
effective by not requiring  controls for Group 2 emission
points.
     Response:   The Group I/Group 2 concept has been retained
in the final rule.
     Comment;   One  commenter  (IV-D-12) maintained that
equipment suitable  for cost effective control of organic HAP
emissions from TPPUs is widely available.  The commenter
(IV-D-12) emphasized that the equipment is capable of meeting
the removal efficiencies specified in the proposed rule for
process vents,  wastewater, and storage vessels in both new and
retrofit applications.
     One commenter  (IV-D-02) explained that catalytic
oxidation is an available, cost effective control option which
can be designed to  achieve in excess of 98 percent VOC
reductions.  The commenter  (IV-D-02)  stated that this
technology can be used to comply with the VOC aspects of the
rule.
     Response;  The  thermoplastics rule is written as a
performance standard so that any control device  may be used if
it can meet the performance requirements.   Nothing in the
proposed or final rule precludes any technology  as long as it
can achieve the level of emission control required by the
rule.  The EPA even provides provisions for alternative
control technologies and promotes innovative control
strategies.
                             14-1

-------
     Comment:   One commenter (IV-D-25)  recommended that all
control device efficiency requirements have associated with
them the same 20 ppm concentration alternative that is
associated with the continuous process vent provisions in the
proposed rule..  The commenter  (IV-D-25) stated that the HON
and the proposed Petroleum Refinery NESHAP both contain this
alternate concentration limit for process vent control
devices.  The commenter (IV-D-25) believed that 20 ppm may
represent the lower bound of the theoretical kinetic limit;
as the inlet concentration drops, the control efficiency drops
and the outlet concentration levels off.
     Response:  The EPA agrees with the commenter and,
therefore, the final rule provides the 20 ppm concentration
alternative for all control device efficiency requirements
associated with continuous process vents, wastewater
operations, and storage vessels.  The EPA did not allow the
20 ppm concentration alternative for batch process vents
because the emission standard  for batch process vents  is not
an instant-in-time limit, but  rather a set emission reduction
for a given event  (i.e.,  90 percent reduction for the  batch
cycle).
                              14-2

-------