& EPA
         United States
         Environmental Protection
         Agency
            Office of Air Quality
            Planning and Standards
            Research Triangle Park. NC 27711
EPA-453/R-93-059
February 1994
         Air
HAZARDOUS AIR POLLUTANTS
FROM MAGNETIC TAPE
MANUFACTURING
         Background
         Information for
         Proposed
         Standards

-------
4
o
    Hazardous Air Pollutant Emissions from Magnetic
       Tape Manufacturing Operations-Background
            Information for Proposed Standards
                         Emission Standards Division
                    U. S. ENVIRONMENTAL PROTECTION AGENCY
                          Office of Air and Radiation
                     Office of Air Quality Planning and Standards
                    Research Triangle Park, North Carolina 27711

                            February 1994   ,, c Cn...

-------
This report has been reviewed by the Emission Standards Division of the Office of Air Quality  Planning and
Standards, EPA, and approved for publication. Mention of trade names or commercial products is not intended
to constitute endorsement or recommendation for use.  Copies of this report are available through the Library
Services Offices (MD-35), U. S. Environmental Protection Agency, Research Triangle Park,  N.C. 27711, or from
National Technical Information Services, 5285 Port Royal Road, Springfield, Virginia 22161.

-------
                 ENVIRONMENTAL  PROTECTION AGENCY

                      Background Information
                            and Draft
                  Environmental Impact Statement
              for Hazardous Air Pollutant Emissions
           From Magnetic Tape Manufacturing Operations
                           Prepared by:
-Bruce C. Jordan                                    (Date)
Director, Emission Standards Division
U. S. Environmental Protection Agency
Research Triangle Park, N.C.  27711

1.  The proposed standards of performance would reduce hazardous
    air pollutant emissions from existing and new magnetic  tape
    manufacturing facilities that are major sources of hazardous
    air pollutant emissions.  Under Section 112 of the Clean Air
    Act as amended in  1990, EPA is authorized to require  the
    maximum degree of  reduction in emissions of hazardous air
    pollutants that is achievable, taking into consideration the
    cost of achieving  such emission reductions, and any nonair
    quality health and environmental impacts and energy
    requirements.

2.  Copies of this document have been sent to the following
    Federal Departments:  Labor, Health and Human Services,
    Defense, Transportation, Agriculture, Commerce, Interior, and
    Energy; the National Science Foundation; the Council  on
    Environmental Quality; members of the State and Territorial
    Air Pollution Program Administrators; the Association of
    Local Air Pollution Control Offices; EPA Regional
    Administrators; and other interested parties.

3.  The comment period for review of this document is 45  days
    from the date of publication of the proposed standard in the
    Federal Register.  Ms. Gail Lacy may be contacted at
     (919) 541-5261 regarding the date of the comment period.

4.  For additional information contact:

    Ms. Gail Lacy
    Standards Development Branch (MD-13)
    U. S. Environmental Protection Agency
    Research Triangle  Park, N.C.  27711
    Telephone:  (919)  541-5261

5.  Copies of this document may be obtained from:

    U. S. EPA Library  (MD-35)
    Research Triangle  Park, N.C.  27711
                               111

-------
                        TABLE OF CONTENTS

                                                             Page

LIST OF FIGURES	      ix
LIST OF TABLES	       x

1.0  SUMMARY	     1-1

     1.1  STATUTORY AUTHORITY 	     1-1
     1.2  REGULATORY ALTERNATIVES	     1-1
     1.3  ENVIRONMENTAL IMPACT	     1-2
     1.4  COSTS AND ECONOMIC IMPACTS	     1-6

2.0  INTRODUCTION	     2-1

     2.1  BACKGROUND AND AUTHORITY FOR STANDARDS  ....     2-1
     2.2  SELECTION OF POLLUTANTS AND SOURCE CATEGORIES  .     2-5
     2.3  PROCEDURE FOR DEVELOPMENT OF NESHAP	     2-6
     2.4  CONSIDERATION OF COSTS	     2-9
     2.5  CONSIDERATION OF ENVIRONMENTAL IMPACTS  ....    2-10
     2.6  RESIDUAL RISK STANDARDS	    2-11

3.0  THE MAGNETIC TAPE MANUFACTURING INDUSTRY PROCESSES
     AND POLLUTANT EMISSIONS	     3-1

     3.1  GENERAL	     3-1
          3.1.1  Industry Description 	     3-1
          3.1.2  Industry Growth	     3-8
     3.2  PROCESSES AND THEIR EMISSIONS	    3-10
          3.2.1  Process Descriptions . . *	    3-10
          3.2.2  HAP Emission Points	    3-27
     3.3  BASELINE EMISSIONS	    3-30
          3.3.1  Existing Emission Limits	    3-30
          3.3.2  Determination of Baseline Level  ....    3-34
     3.4  REFERENCES FOR CHAPTER 3	    3-47

4.0  EMISSION CONTROL TECHNIQUES  	     4-1

     4.1  INTRODUCTION	     4-1
     4.2  SOLVENT HAP CONTROL SYSTEMS 	     4-1
          4.2.1  Adsorption	     4-3
          .4.2.2  Condensers	    4-15
          4.2.3  Incinerators	    4-21
          4.2.4  Flare Systems	    4-24
          4.2.5  Conservation Vents and Pressure Relief
                 Valves	    4-25
          4.2.6  New Solvent HAP Emission Control
                 Technologies 	    4-28
     4.3  PARTICULATE HAP CONTROL SYSTEMS 	    4-30
          4.3.1  Enclosed Particulate Transfer Methods   .    4-31
          4.3.2  Pulse-Jet Fabric Filters 	    4-31

-------
                  TABLE OF CONTENTS (continued)
                                                             Page

     4.4  VOC EMISSION CAPTURE SYSTEMS	    4-34
          4.4.1  Capture of Emissions From Solvent
                 Storage Tanks	  .    4-34
          4.4.2  Capture of Emissions From Mix Rooms   .  .    4-35
          4.4.3  Capture of Emissions From Coating
                 Operations	    4-36
     4.5  CONTROL OF PACKAGING AND LABELING EMISSIONS  .  .    4-41
     4.6  CONTROL OF CLEANING EMISSIONS 	    4-41
          4.6.1  Vapor Capture and Control	    4-41
          4.6.2  Solvent Substitution 	    4-44
          4.6.3  Work Practice	    4-44
          4.6.4  Specialized Cleaning Techniques  ....    4-45
          4.6.5  Mechanical Cleaning  	    4-46
     4.7  CONTROL OF EMISSIONS FROM WASTE HANDLING  ...    4-46
          4.7.1  Reduction of Emissions from Waste
                 Transfer and Loading	    4-46
          4.7.2  Emission Capture and Control	    4-47
     4.8  CONTROL OF LEAKS FROM SOLVENT TRANSFER PIPING  .    4-47
     4.9  LOWER-VOC OR NON-HAP SOLVENT TECHNOLOGY ....    4-48
     4.10  REFERENCES FOR CHAPTER 4	    4-49

5.0  MODIFICATION AND RECONSTRUCTION	     5-1

6.0  MODEL LINES AND REGULATORY ALTERNATIVES	     6-1

     6.1  GENERAL	     6-1
     6.2  MODEL LINES	     6-2
     6.3  MODEL LINE PARAMETERS :'	     6-3
          6.3.1  Solvent Storage Tanks  	     6-3
          6.3.2  Mix Room Equipment	     6-4
          6.3.3  Coating Application/Drying 	     6-5
          '6.3.4  Solvent Recovery	     6-6
          6.3.5  Waste Handling	     6-7
          6.3.6  Packaging/Labeling 	     6-8
          6.3.7  Particulate HAP Emissions	     6-8
          6.3.8  Cleaning Activities	     6-8
     6.4  REGULATORY ALTERNATIVES 	     6-9
          6.4.1  Existing Major Sources 	    6-11
          6.4.2  New Lines	    6-13
          6.4.3  Control Options Not Included in the
                 Regulatory Alternative 	    6-13
     6.5  REFERENCES FOR CHAPTER 6	    6-28

7.0  ENVIRONMENTAL IMPACT	     7-1

     7.1  AIR POLLUTION IMPACTS	     7-1
          7.1.1  Primary Air Pollution Impacts	     7-3
          7.1.2  Secondary Air Pollution Impacts  ....     7-6
     7.2  WATER POLLUTION IMPACTS	    7-11

-------
                  TABLE OF CONTENTS (continued)
                                                             Page

     7.3  SOLID WASTE IMPACTS	    7-14
     7.4  ENERGY IMPACTS	    7-15
     7.5  OTHER ENVIRONMENTAL IMPACTS	    7-18
     7.6  OTHER ENVIRONMENTAL CONCERNS	    7-18
          7.6.1  Irreversible and Irretrievable
                 Commitment of Resources	    7-18
          7.6.2  Environmental Impact of Delayed
                 Standard	    7-19
     7.7  REFERENCES FOR CHAPTER 7	    7-19

8.0  COSTS	     8-1

     8.1  INTRODUCTION	     8-1
     8.2  MODEL LINE PARAMETERS	     8-1
     8.3  CONTROL COST ESTIMATING METHODOLOGY	     8-2
          8.3.1  Total Capital Investment	     8-5
          8.3.2  Total Annual Control Costs 	    8-15
          8.3.3  Summary of Control Costs	    8-24
     8.4  COMPLIANCE, REPORTING, AND RECORDKEEPING
          COSTS	    8-26
          8.4.1  Compliance Requirements and Costs for
                 Existing Sources 	    8-26
          8.4.2  Reporting and Recordkeeping Costs
                 for Existing Sources 	    8-36
          8.4.3  Compliance, Reporting,  and Recordkeeping
                 Costs for Model Lines	    8-46
     8.5  EQUIPMENT LEAKS IN PIPING	    8-48
     8.6  RESULTS OF THE COST ANALYSIS	    8-57
          8.6.1  Cost Effectiveness	    8-57
          8.6.2  Cost per Unit Area of Tape Coated  ...    8-57
          8.6.3  Incremental Cost Effectiveness	    8-62
     8.7  REFERENCES.FOR CHAPTER 8  	    8-62

9.0  ECONOMIC IMPACTS	"  .     9-1
     9.1  INDUSTRY PROFILE  	     9-1
          9.1.1  Introduction	     9-1
          9.1.2  Market Structure 	     9-2
          9.1.3  Production	    9-12
          9.1.4  Foreign Trade	    9-15
          9.1.5  Consumption	    9-17
          9.1.6  Pricing	    9-21
          9.1.7  Outlook	    9-21
     9.2  ECONOMIC IMPACT ANALYSIS	    9-23
          9.2.1  Methodology	    9-23
          9.2.2  Industry-Level Impacts 	    9-35
          9.2.3  Facility-Level Impacts	    9-37
          9.2.4  Conclusions	    9-43
                               vii

-------
                  TABLE OF CONTENTS  (continued)
                                                             Page
     9.3  SMALL BUSINESS IMPACTS	    9-44
     9.4  REGULATORY FLEXIBILITY ANALYSIS 	    9-45
          9.4.1  Introduction	    9-45
          9.4.2  Demographic Analysis	    9-45
          9.4.3  Findings of the Economic Impact Analysis    9-45
          9.4.4  Consideration of Flexibility in
                 Regulatory Requirements	    9-46
     9.5  REFERENCES FOR SECTION 9	    9-48

APPENDIX A.  EVOLUTION OF THE BACKGROUND INFORMATION
             DOCUMENT	     A-l

APPENDIX B.  INDEX TO ENVIRONMENTAL IMPACT
             CONSIDERATIONS	     B-l

APPENDIX C.  EMISSION SOURCE TEST DATA	     C-l

             C.I  DATA FROM EPA-SPONSORED TESTS ON
                  CARBON ADSORBER RECOVERY EFFICIENCIES  .     C-l
             C.2  DATA FROM INDUSTRY ON CARBON ADSORBER
                  RECOVERY EFFICIENCIES 	     C-2
             C.3  DATA FROM STATE COMPLIANCE TESTS  ...     C-3
             C.4  DATA FROM EPA-SPONSORED TESTS FOR
                  RELATED INDUSTRIES  	     C-3

APPENDIX D.  EMISSION MEASUREMENT AND CONTINUOUS
             MONITORING	     D-l

             D.I  EMISSION MEASUREMENT TEST PROGRAM AND
                  METHODS	     D-l
             D.2  PERFORMANCE TEST METHODS	     D-5
            .D.3  MONITORING SYSTEMS AND DEVICES  ....    D-13

APPENDIX E.  ENVIRONMENTAL AND ENERGY IMPACTS OF THE
             CONTROL OPTIONS  	     E-l

APPENDIX F.  COST IMPACTS OF THE CONTROL OPTIONS  ....     F-l

             F.I  SAMPLE CALCULATIONS	     F-l
             F.2  CALCULATION OF COSTS TO CONTROL
                  EQUIPMENT LEAK EMISSIONS	    F-13
             F.3  REFERENCES FOR APPENDIX F	    F-25

APPENDIX G.         	     G-l
                               Vlll

-------
                         LIST OF FIGURES
Figure 3-1.

Figure 3-2.


Figure 3 - 3.

Figure 3-4.

Figure 3-5.

Figure 3-6.

Figure 3-7.


Figure 4-1.


Figure 4-2.

Figure 4-3.


Figure 4-4.


Figure 4-5.

Figure 4-6.


Figure 4-7.

Figure 9-1.


Figure 9-2.

Figure 9-3.
Annual shipments of blank magnetic tape

Schematic drawing of magnetic tape
coating plant  	 .
Coating head configurations

Metering-type coating heads

Air flotation drying oven

Festoon oven	,
Baseline solvent HAP emissions from
magnetic tape industry 	
Diagrammatic sketch of a two unit,
fixed-bed adsorber 	
Fluidized-bed carbon adsorption system .  .

Schematic of condensation system using
nitrogen 	

Flow diagram of condensation system using
an air atmosphere in the drying oven .  .  .
Diagram of conservation vent
Schematic of total enclosure ventilation
system	! .

Schematic of room ventilation systems

Vertical relationships in the magnetic
recording media industry 	
Economic impact analysis methodology .  .

Effects of installing pollution control
equipment  	
 Page

 3-9



3-11

3-18

3-19

3-20

3-22



3-37



 4-4

4-10



4-16



4-19

4-27



4-38

4-39



 9-8

9-24



9-26
                               ix

-------
                          LIST OF TABLES
                                                             Page
TABLE 1-1.   POTENTIAL NATIONWIDE EMISSIONS REDUCTION
             AND NATIONWIDE AIR, WASTEWATER, SOLID
             WASTE, AND ENERGY IMPACTS FOR EXISTING
             MAJOR SOURCES	     1-3

TABLE 1-2.   POTENTIAL NATIONWIDE EMISSIONS REDUCTION
             AND NATIONWIDE AIR, WASTEWATER, SOLID
             WASTE, AND ENERGY IMPACTS FOR NEW LINES   .  .     1-4

TABLE 1-3.   NATIONWIDE REGULATORY ALTERNATIVE COST
             IMPACTS FOR EXISTING MAJOR SOURCES  	     1-7

TABLE 1-4.   NATIONWIDE REGULATORY ALTERNATIVE COST
             IMPACTS FOR NEW LINES	     1-8

TABLE 3-1.   USES OF HAP'S AT MAGNETIC TAPE MANUFACTURING
             PLANTS	     3-2

TABLE 3-2.   MAGNETIC TAPE PRODUCT PARAMETERS	     3-4

TABLE 3-3.   PLANTS COATING MAGNETIC TAPE	     3-6

TABLE 3-4.   SELECTED COATING MIX PROPERTIES	    3-13

TABLE 3-5.   CURRENT STATE REGULATIONS ON VOC EMISSIONS
             FROM THE MAGNETIC TAPE COATING INDUSTRY   .  .    3-31

TABLE 3-6.   SUMMARY OF NSPS REGULATION	    3-35

TABLE 3-7,   POINT-SPECIFIC BASELINE EMISSIONS   	    3-38

TABLE 4 -1.   CONTROL DEVICES USED ON COATING OPERATIONS  .     4-2

TABLE 4-2.   PROCESS PARAMETERS FOR MAGNETIC TAPE PLANTS
             CONTROLLED BY FIXED-BED CARBON ADSORBERS  .  .     4-6

TABLE 6-1.   SUMMARY OF EMISSION POINTS FOR MODEL LINES  .    6-16

TABLE 6-2.   SUMMARY OF NSPS REGULATION	    6-17

TABLE 6-3.   MODEL SOLVENT STORAGE TANK PARAMETERS  ...    6-18

TABLE 6-4.   MODEL MIX ROOM PARAMETERS	    6-19

TABLE 6-5.   MODEL COATING OPERATION PARAMETERS  	    6-20

TABLE 6-6.   SOLVENT RECOVERY PARAMETERS FOR MODEL
             LINES	    6-21

-------
       LIST OF TABLES (continued)
WASTE HANDLING PARAMETERS FOR MODEL LINES

PARTICULATE HAP PARAMETERS FOR MODEL LINES
Page

6-22

6-23
TABLE 6-7.

TABLE 6-8.

TABLE 6-9.   CLEANING ACTIVITIES--PARAMETERS FOR MODEL
             LINES	    6-24

TABLE 6-10.  CONTROL OPTIONS FOR HAZARDOUS AIR POLLUTANT
             EMISSIONS FROM MAGNETIC TAPE MANUFACTURING
             FACILITIES	    6-25

TABLE 6-11.  IMPACT OF REGULATORY ALTERNATIVES FOR HAP
             EMISSION POINTS--EXISTING MAJOR SOURCES
             AND NEW LINES	    6-26

TABLE 6-12.  IMPACT OF REGULATORY ALTERNATIVES ON NEW
             LINES	    6-27

TABLE 7-1.   REGULATORY ALTERNATIVES FOR HAP EMISSION
             POINTS--EXISTING SOURCES AND NEW LINES  ...     7-2

TABLE 7-2.   ENVIRONMENTAL IMPACT OF THE REGULATORY
             ALTERNATIVE ON EXISTING MAJOR SOURCES   ...     7-3

TABLE 7-3.   IMPACT OF REGULATORY ALTERNATIVES ON NEW
             LINES	     7-4

TABLE 7-4.   IMPACT OF REGULATORY ALTERNATIVES ON NEW
             LINES--HAP EMISSIONS AND INCREMENTAL HAP
             EMISSION REDUCTION 	     7-5

TABLE 7-5.   SUMMARY OF ANNUAL SECONDARY POLLUTANT
            ' EMISSIONS FROM THE COMBUSTION OF NATURAL
             GAS--EXISTING MAJOR SOURCES  	     7-7

TABLE 7-6.   SUMMARY OF ANNUAL SECONDARY POLLUTANT
             EMISSIONS FROM THE COMBUSTION OF FUEL OIL--
             EXISTING MAJOR SOURCES 	     7-9

TABLE 7-7.   ANNUAL INCREMENTAL FUEL OIL CONSUMPTION
             FOR STEAM GENERATION--NEW LINES	     7-9

TABLE 7-8.   ANNUAL SECONDARY POLLUTANT EMISSIONS FROM
             FUEL OIL COMBUSTION--NEW LINES	    7-10

TABLE 7-9.   ANNUAL INCREMENTAL WASTEWATER DISCHARGES
             AND EMISSIONS FROM EXISTING MAJOR SOURCES   .    7-13

-------
                    LIST OF TABLES (continued)
TABLE 7-10.


TABLE 7-11.


TABLE 7-12.


TABLE 7-13.


TABLE 7-14.


TABLE 8-1.



TABLE 8-2.


TABLE 8-3.



TABLE 8-4.



TABLE 8-5.



TABLE 8-6.


TABLE 8-7.


TABLE 8-8.

TABLE 8-9.


TABLE 8-10.
ANNUAL INCREMENTAL WASTEWATER DISCHARGES
AND EMISSIONS FROM NEW LINES 	
ANNUAL INCREMENTAL ENERGY REQUIREMENTS OF
REGULATORY ALTERNATIVES--EXISTING SOURCES   .

ANNUAL INCREMENTAL STEAM REQUIREMENTS OF
REGULATORY ALTERNATIVES--NEW LINES  	

ANNUAL INCREMENTAL ELECTRICITY REQUIREMENTS
OF REGULATORY ALTERNATIVES--NEW LINES  .  .  .

TOTAL ANNUAL INCREMENTAL ENERGY REQUIREMENTS
OF REGULATORY ALTERNATIVES--NEW LINES  .  .  .

CONTROL OPTIONS FOR HAZARDOUS AIR POLLUTANT
EMISSIONS FROM MAGNETIC TAPE MANUFACTURING
FACILITIES 	

REGULATORY ALTERNATIVES FOR HAP EMISSION
POINTS--EXISTING SOURCES AND NEW LINES .  .  .

MAGNETIC TAPE NSPS AND NESHAP--ASSUMED
NUMBER OF STORAGE TANKS; NSPS TOTAL CAPITAL
INVESTMENT 	

NESHAP COST ANALYSIS--TOTAL CAPITAL
INVESTMENT FOR CONTROL OF STORAGE TANK
EMISSIONS WITH A COMMON CONTROL DEVICE .  .  .

NESHAP COST ANALYSIS--TOTAL CAPITAL
INVESTMENT TO CONTROL MIX PREPARATION
EQUIPMENT WITH A COMMON CONTROL DEVICE .  .  .

TOTAL CAPITAL INVESTMENT--TOTAL ENCLOSURE
FOR COATING APPLICATION/DRYING PROCESS .  .  .

MODEL PARAMETERS USED IN TOTAL ANNUAL COST
METHODOLOGY  	
UNIT COSTS USED IN NESHAP ANALYSIS
FIXED BED ADSORBER--DIRECT AND INDIRECT
COSTS--EXAMPLE CALCULATION 	
INCINERATOR TOTAL ANNUAL COSTS--EXAMPLE
CALCULATION  	
Page


7-13


7-16


7-17


7-17


7-18



 8-3


 8-4



 8-3



 8-9



8-11


8-12


8-17

8-17


8-18


8-19
                               xzz

-------
                    LIST OF TABLES  (continued)

                                                             Page

TABLE 8-11.  CALCULATION OF NATURAL GAS DEMAND FOR
             INCINERATORS  	    8-20

TABLE 8-12.  FACILITY-SPECIFIC COST-EFFECTIVENESS
             OF CONTROL	    8-25

TABLE 8-13.  ENHANCED MONITORING REQUIREMENTS 	    8-27

TABLE 8-14.  ESTIMATED INITIAL PERFORMANCE TEST COSTS  .  .     8-31

TABLE 8-15.  ASSUMPTIONS FOR MAGNETIC TAPE TESTING
             BUDGET	     8-32

TABLE 8-16.  MONITORING EQUIPMENT CONTROL COSTS
             (EXCLUDES MATERIAL BALANCE OPTION) 	     8-35

TABLE 8-17.  COST OF COMPLIANCE IF METHOD 25A IS
             REQUIRED	     8-37

TABLE 8-18a. ANNUAL BURDEN TO EXISTING SOURCES TO
             IMPLEMENT REPORTING AND RECORDKEEPING
             REQUIREMENTS--FIRST YEAR	     8-39

TABLE 8-18b. ANNUAL BURDEN TO EXISTING SOURCES TO
             IMPLEMENT REPORTING AND RECORDKEEPING
             REQUIREMENTS--SECOND YEAR  	     8-42

TABLE 8-19a. ANNUAL BURDEN TO MODEL LINES TO
             IMPLEMENT REPORTING AND RECORDKEEPING
             REQUIREMENTS--FIRST YEAR	     8-49
      •
TABLE 8-19b. ANNUAL BURDEN TO MODEL LINES TO
             IMPLEMENT REPORTING AND RECORDKEEPING
             REQUIREMENTS--SECOND YEAR  	     8-52

TABLE 8-20.  TOTAL INDUSTRY-WIDE COST FOR THE
             NESHAP--RA  I	     8-58

TABLE 8-21.  TOTAL INDUSTRY-WIDE COST FOR THE
             NESHAP--RA  II	     8-59

TABLE 8-22.  TOTAL ANNUAL COST, COST PER UNIT AREA
             OF TAPE COATED, AVERAGE AND INCREMENTAL
             COST EFFECTIVENESS FOR RA I	     8-60

TABLE 8-23.  TOTAL ANNUAL COST, COST PER UNIT AREA OF
             TAPE COATED, AVERAGE AND INCREMENTAL COST
             EFFECTIVENESS FOR RA II	     8-61
                               Xlll

-------
                    LIST OF TABLES  (continued)
TABLE 9-1.


TABLE 9-2.


TABLE 9-3.


TABLE 9-4.


TABLE 9-5.


TABLE 9-6.


TABLE 9-7.


TABLE 9-8.



TABLE 9-9.


TABLE 9-10.


TABLE 9-11.


TABLE 9-12.


TABLE 9-13.


TABLE 9-14.


TABLE 9-15.


TABLE 9-16.
VALUE OF SHIPMENTS, 1986-1990:  SIC 3695,
MAGNETIC AND OPTICAL RECORDING, MEDIA   .  .

VALUE OF SHIPMENTS, 1981-1990:  FLEXIBLE
COMPUTER DISKS 	
VALUE OF SHIPMENTS, 1986-1990:  COMPUTER
CASSETTES AND CARTRIDGES 	
VALUE OF SHIPMENTS; 1986-1990:  COMPUTER
REEL TAPE  	
VALUE OF SHIPMENTS, 1986-1990:  BLANK AUDIO
TAPE	,

VALUE OF SHIPMENTS, 1986-1990:  BLANK VIDEO
TAPE	

WORLD'S TOP 25 PRODUCERS OF MAGNETIC
RECORDING MEDIA, 1990	,
YEAR-TO-YEAR CHANGES IN THE OUTPUT OF
SELECTED MAGNETIC RECORDING MEDIA,
1982-1990  	 .
FACTORY SHIPMENTS, 1986-1990:  BLANK AUDIO
AND VIDEO CASSETTES  	
VALUE OF EXPORTS, 1989-1990:  SIC 3695,
MAGNETIC AND OPTICAL RECORDING MEDIA .  .  .

VALUE OF IMPORTS, 1989-1990:  SIC 3695,
MAGNETIC AND OPTICAL RECORDING MEDIA .  .  .

VALUE OF EXPORTS, 1989-1990:  SELECTED
MAGNETIC RECORDING MEDIA 	
VALUE OF IMPORTS, 1989-1990:  SELECTED
MAGNETIC RECORDING MEDIA 	
U.S. APPARENT CONSUMPTION OF SELECTED
MAGNETIC RECORDING MEDIA, 1989-1990  .  . .

FACTORY SHIPMENTS OF BLANK AUDIO CASSETTES
AND AUDIO TAPE EQUIPMENT, 1986-1990  .  . .

FACTORY SHIPMENTS OF BLANK VIDEO CASSETTES
AND VIDEO CASSETTE RECORDERS, 1986-1990
Page


 9-3


 9-4
    •


 9-4


 9-5


 9-5


 9-6


9-10



9-13


9-14


9-16


9-16


9-17


9-17


9-18


9-19


9-19
                               xiv

-------
                    LIST OP TABLES  (continued)

                                                             Page

TABLE 9-17.  PERCENT CHANGE IN PRICE REALIZATION:
             SELECTED MAGNETIC RECORDING MEDIA,
             1982-1990	    9-22

TABLE 9-18.  ECONOMIC IMPACTS ON THE MAGNETIC RECORDING
             MEDIA INDUSTRY	 . . .    9-36

-------
                           1.0   SUMMARY

1.1  STATUTORY AUTHORITY
     National emission standards for hazardous air pollutants
(NESHAP) are established under Section 112 of the Clean Air Act
(CAA) (42 U.S.C. 7412), as amended in 1990.  Emission standards
under Section 112 apply to new and existing sources of hazardous
air pollutants  (HAP's) as listed in Section 112(b).
Section 112(c) directs the Administrator to use the HAP list to
develop a list of source categories for which NESHAP will be
developed.  Magnetic tape manufacturing operations have been
selected for regulation.  This background information document
supports proposed standards regulating HAP emissions from
magnetic tape manufacturing facilities.
1.2  REGULATORY ALTERNATIVES
     Solvent HAP emission points identified for the magnetic tape
manufacturing industry include solvent storage, coating mix
preparation,  application and drying of the magnetic coating,
waste solvent handling, packaging and labeling of magnetic tape
products, cleaning activities (fixed surfaces, removable parts,
and fixed piping lines), and leaks in piping systems that
transport HAP's.  Coating mix preparation is also a source of
particulate HAP emissions.  Many regulatory alternatives are
possible from various combinations of the control options listed
above.  Each alternative was evaluated based on technical
feasibility and cost; alternatives that were not feasible were
eliminated.  This process yielded two Regulatory Alternatives for
which cost and environmental impacts were determined.
Environmental and cost impacts were determined for these
alternatives.

                               1-1

-------
   "  Five NESHAP model lines have been established.  These are:
(1) a small line; (2) a medium line built with concurrent
construction of a volatile organic compound (VOC) control device;
(3) a medium line built without concurrent construction of a VOC
control device;  (4)  a large line built with concurrent
construction of a VOC control device; and (5)  a large line built
without concurrent construction of a VOC control device.  The
three line sizes (small, medium, and large)  are analogous to the
research, small, and typical model lines developed for the NSPS,
which were categorized by the major design parameters of
production rate, hours of operation, coating solvent content, and
coating thickness.   It is necessary to distinguish between the
situation when the plant concurrently constructs a VOC control
device with the new coating line and when the plant uses an
existing control device with a new coating line because the
requirements of the NSPS, and hence baseline emissions, are
different.
     Regulatory Alternative I represents the maximum achievable
control technology (MACT) floor for this industry.  Regulatory
Alternative I includes control of solvent storage, coating mix
preparation, coating application/drying, waste solvent handling,
cleaning of removable parts, flushing of fixed piping lines, and
particulate emissions in the mix room.  Regulatory Alternative I
represents 51 percent control of solvent HAP's and 70 percent
control of particulate HAP's from baseline levels for existing
sources.
     Regulatory Alternative II requires control of equipment
leaks and the use of closed containers for all cleaning
activities in addition to the controls specified in Regulatory
Alternative I.  Regulatory Alternative II represents 61 percent
control of solvent HAP emissions and 70 percent control of
particulate HAP emissions from baseline levels for existing
sources.
1.3  ENVIRONMENTAL IMPACT
     Tables 1-1 and 1-2 summarize the environmental impacts of
Che regulatory alternatives for existing major sources and for
                               1-2

-------
H H

          8 I
          s ea
U3
I S3
          1s
       1
          2
     1c
     «3 e
     S I


     tl
      * i i j $
      1 3 111
      11lls;
      2 § .2 I ^>
       Q. E **" S
      1!I||
wide
         .« V rt
         III
         e
         o

      «5l-l^

      1*11^
      .1 g c i 2
      i I .3 c «
      2 3 J .2
          2» 2
90 ^

8" 2"
              0 0,
               3
              0
                 82
         o s,
          S
                 0 0
                 b o,
              01
              to
         •s
         3

         I
         2
         s

         3
         2

         I
                       00
                       c
                      I
              .Si
                      .2
              a
              ^3

              •3 I-

              S I

              1 = 1
              •00%
              5 a
                      15
                             at -^
                             ai •••
                             3 g
                 o -a
                 3 'M
                  9
                             C -O
                            **
              i^-aj
              I I'll
              •3 lb * «
              1 S II u H  u

              ISzsftfs
                              1-3

-------
Cd
EH

O
04
M
W
J


I
1 ^ § 1?

1 i
en

iia,!J.i*£
j !. i "i jo •§
2 o.
ilUjisf
||11*J*4
I M1 s.r=
JiM-Hls
2 * w •"
.a
CM
S
—•
«
3



^









*
1
S
1
i*













v*T
r*-
i
a
S
a


€
o
f
a
o

£
a
d
~
o
a
s
0
S
a
S
o
3
Ol

a





01

m
£
S

€
0

"!
S
P
a
vO
e

a
r-
d

f
0,
S
o
f
a
a
Q-

oo
m
J3
§
•S
j
8
3
1

^^
Ot

€
o
f
a
VI
d


0

o



o
f
a
8
0
g

£

.A

J
S
3
1
g
*
n

S
o

2
S
(S

i.
o

m1

0,

O

a
o
S
o

m

jO
3

j
1


1

g
e

/— .
ci
oo
s
8
a
i
o

s
8
d
,-,
a
§
0

J
o
§

—


.
-S
j
§













=3
|
I
i
**













V)
a
§
a
s
o


a
o
f
a
o

a
d
^
o
a
m
3
0
g
a
S
o
5
Ot

s





I


g

€
o



«
S
0

a
o
g
5

S?

•a.

^
i
o
I


§
a

g
e

^,
ci
I
a
e

S
a
S
d
^.
a
§
o

a
o
1

r-
m


^
jj
i
J
                                                         IzsT
                                1-4

-------
new coating lines.  Included in Tables 1-1 and 1-2 are solvent
HAP emission reductions and the impacts of HAP control
technologies on secondary air pollution, wastewater, solid waste
and energy requirements.  Secondary pollutants (particulate
matter [PM] , SOX, and NOX) result from the burning of fuel oil to
generate steam for carbon adsorbers.  Carbon monoxide (CO) is
produced in addition to PM, SOX, and NOX from incineration of
solvent HAP's.
     As shown in Table 1-1, Regulatory Alternative I  (the MACT
floor) would reduce solvent HAP emissions from existing major
sources by 2,080 megagrams per year  (Mg/yr)  (2,300 tons per year
[tons/yr]).  As shown in Table 1-2, emissions from new coating
lines that fall under the category of the small model line would
be reduced by 19 Mg/yr  (21 tons/yr).  Emissions from new lines
classified as medium-sized would be reduced by 3.3 Mg/yr
(3.6 tons/yr) for lines not built concurrently with a control
device and 1.1 Mg/yr (1.2 ton/yr) for lines built concurrently
with a control device.   Emissions from new lines classified as
large-sized would be reduced by 30 Mg/yr (33 tons/yr) at lines
not built concurrently with a control device and 12 Mg/yr
(13 tons/yr) at lines built concurrently with a control device.
     Regulatory Alternative II would reduce solvent HAP emissions
from existing major sources by 2,470 Mg/yr (2,720 tons/yr).
Emissions from new coating lines that fall under the category of
the small model line would be reduced by 19 Mg/yr (21 tons/yr).
Emissions from new lines classified as medium-sized would be
reduced by 4.5 Mg/yr (5.0 tons/yr)  at lines not built
concurrently with a control device and 2.3 Mg/yr (2.5 tons/yr) at
lines built concurrently with a control device.  Emissions from
new lines classified as large-sized would be reduced by 43 Mg/yr
(47 tons/yr) at lines not built concurrently with a control
device and 24 Mg/yr (27 tons/yr) at lines built concurrently with
a control device.
     Regulatory Alternative I reduces particulate HAP emissions
at existing major sources by 0.27 Mg/yr (0.3 ton/yr).  For model
lines classified as small, Regulatory Alternative I reduces
                               1-5

-------
particulate HAP emissions by 0.02 Mg/yr (0.02 ton/yr).  For model
lines classified as medium, the particulate HAP emissions would
decrease by O.OS Mg/yr (0.06 ton/yr).  For model lines classified
as large, particulate HAP emissions would decrease by 0.5 Mg/yr
(0.6 ton/yr).  Regulatory Alternative II does not provide any
additional particulate HAP emission reduction beyond the baseline
or Regulatory Alternative I.
1.4  COSTS AND ECONOMIC IMPACTS
     The nationwide cost impacts of the regulatory alternatives
on existing major sources and new lines are summarized in
Tables 1-3 and 1-4, respectively.  For existing major sources,
Regulatory Alternative I requires a total annual cost of $400,120
and Regulatory Alternative II requires a total annual cost of
$2,754,790.
     For new coating lines that fall under the category of the
small model line, the total annual cost of Regulatory
Alternative I is $98.660.  The total annual costs for new lines
classified as medium-sized are $56,500 for lines not built
concurrently with a control device and $15,370 for lines built
concurrently with a control device.  The total annual costs for
lines classified as large-sized are $47,260 for lines not built
concurrently with a control device and $36,140 for lines built
concurrently with a control device.  The total annual costs for
lines built concurrently with a control device are lower because
the NSPS requires lines under this category to control mix room
emissions to the extent required by Regulatory Alternative II.
     The total annual cost of Regulatory Alternative II for new
lines classified as small is $104,060.  The total annual costs
for lines classified as medium are $62,900 for lines not built
concurrently with a control device and $21,770 for lines built
concurrently with a control device.  The total annual costs for
lines classified as large are $77,740 for lines not built
concurrently with a control device and $66,620 for lines built
concurrently with a control device.  The total annual costs for
emissions to the extent required by Regulatory Alternative II
lines built concurrently with a control device are lower because
                               1-6

-------
TABLE 1-3.
NATIONWIDE REGULATORY ALTERNATIVE COST IMPACTS
    FOR EXISTING MAJOR SOURCES




Regulatory
Alternative
I

n





Total annual
cost, $
400,120

2,754,790


Annual
emission
reduction from
baseline, Mg
(tons)
2,080
(2,300)
2,470
(2,720)



Cost
effectiveness,
$/Mg ($/ton)
190
(170)
1,120
(1,010)
Annual
emission
reduction from
previous
alternative, Mg
(tons)
N/A

386
(425)


Incremental
cost
effectiveness,
$/Mg ($/ton)
N/A

6,100
(5,540)
                           1-7

-------
  TABLE  1-4,
NATIONWIDE REGULATORY ALTERNATIVE COST  IMPACTS
              FOR  NEW LINES
Regulatory
Alternative

Small
Medium
without*
Medium
with*
Large
without*
Large with*
Total
annual
cost, $
Annual
emission
reduction from
baseline, Mg
(tons)
Cost effectiveness,
$/Mg ($/ton)
Annual emission
reduction from
previous
alternative, Mg
(tons)
Incremental
cost
effectiveness,
$/Mg ($/ton)
Regulatory Alternative I
99,520
57,360
16,230
48U20
37,000
19.1
(21)
3.3
(3.6)
1.1
(1.2)
29.7
(32.7)
11.5
(12.7)
5,210
(4,740)
17,380
(15,930)
14,750
(13,530)
1,620
(1,470)
3,220
(2,910)
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Regulatory Alternative Q
Small
Medium
without*
Medium
with*
Large
without*
Large with*
104,920
63,760
23,630
78,600
67,480
19.1
(21)
4.5
(5.0)
2.3
(2.5)
42.5
(46.8)
24.3
(26.8)
5,480
(4,970)
14,170
(12,750)
9,840
(9,050)
1,850
(1,680)
2,780
(2,520)
0.1
(0.1)
1.3
(1.4)
1.2
(1.3)
12.8
(14.1)
12.8
(14.1)
54,000
(54,000)
4,920
(4,570)
5,330
(4,920)
2,380
(2,160)
2,380
(2,160)
a"Without" means that the coating line was built without concurrent construction of a control device.
 "With" means that the coating line was built with concurrent construction of a control device.
                                      1-8

-------
the NSPS requires lines under this category to control mix
room emission to the extent required by Regulatory
Alternative II.
     The economic analyses indicate that the worst-case maximum
industrywide price impacts for existing major sources are
0.0082 percent for Regulatory Alternative I and 0.0857 percent
for Regulatory Alternative II.  Detailed analyses of the costs
and the economic impacts are presented in Chapters 8 and 9.
                               1-9

-------
                        2.0  INTRODUCTION

2.1  BACKGROUND AND AUTHORITY FOR STANDARDS
     According to industry estimates, more than 2.4 billion
pounds of toxic pollutants were emitted to the atmosphere in 1988
("Implementation Strategy for the Clean Air Act Amendments of
1990," EPA Office of Air and Radiation, January 15, 1991).  These
emissions may result in a variety of adverse health effects,
including cancer, reproductive effects, birth defects, and
respiratory illnesses.   Title III of the Clean Air Act Amendments
of 1990 provides the tools for controlling emissions of these
pollutants.  Emissions from both large and small facilities that
contribute to air toxics problems in urban and other areas will
be regulated.  The primary consideration in establishing national
industry standards must be demonstrated technology.  Before
national emission standards for hazardous air pollutants  (NESHAP)
are proposed as Federal regulations, air pollution prevention and
control methods are examined in detail with respect to their
feasibility, environmental impacts, and costs.  Various control
options based on different technologies and degrees of efficiency
are examined, and a determination is made regarding whether the
various control options apply to each emissions source or if
dissimilarities exist between the sources.  In most cases,
regulatory alternatives are subsequently developed that are then
studied by EPA as a prospective basis for a standard.  The
alternatives are investigated in terms of their impacts on the
environment, the economics and well-being of the industry, the
national economy, and energy and other impacts.  This document
summarizes the information obtained through these studies so that
                               2-1

-------
interested persona will be able to evaluate the information
considered by EPA in developing the proposed standards.
     National emission standards for hazardous air pollutants for
new and existing sources are established under Section 112 of the
Clean Air Act as amended in 1990 [42 U.S.C. 7401 et seq., as
amended by PL 101-549, November IS, 1990], hereafter referred to
as the Act.  Section 112 directs the EPA Administrator to
promulgate standards that "require the maximum degree of
reduction in emissions of the hazardous air pollutants subject to
this section (including a prohibition of such emissions, where
achievable) that the Administrator, taking into consideration the
cost of achieving such emission reductions, and any non-air
quality health and environmental impacts and energy requirements,
determines is achievable ... ."  The Act allows the Administrator
to set standards that "distinguish among classes, types, and
sizes of sources within a category or subcategory."
     The Act differentiates between major sources and area
sources.  A major source is defined as "any stationary source or
group of stationary sources located within a contiguous area and
under common control that emits or has the potential to emit
considering controls, in the aggregate, 10 tons per year or more
of any hazardous air pollutant or 25 tons per year or more of any
combination of hazardous air pollutants."  The Administrator,
however, may establish a lesser quantity cutoff to distinguish
between major and area sources.  The level of the cutoff is based
on the potency, persistence, or other characteristics or factors
of the air pollutant.  An area source is defined as "any
stationary source of hazardous air pollutants that is not a major
source."  For new sources, the amendments state that the "maximum
degree of reduction in emissions that is deemed achievable for
new sources in a category or subcategory shall not be less
stringent than the emission control that is achieved in practice
by the best controlled similar source, as determined by the
Administrator."  Emission standards for existing sources "may be
less stringent than the standards for new sources in the same
                               2-2

-------
category or subcategory but shall not be less stringent, and may
be more stringent than--
     (A) the average emission limitation achieved by the best
performing 12 percent of the existing sources (for which the
Administrator has emissions information), excluding those sources
that have, within 18 months before the emission standard is
proposed or within 30 months before such standard is promulgated,
whichever is later, first achieved a level of emission rate or
emission reduction which complies, or would comply if the source
is not subject to such standard, with the lowest achievable
emission rate (as defined by Section 171) applicable to the
source category and prevailing at the time, in the category or
subcategory for categories and subcategories with 30 or more
sources, or
     (B) the average emission limitation achieved by the best
performing five sources (for which the Administrator has or could
reasonably obtain emissions information) in the category or
subcategory for categories or subcategories with fewer than
30 sources."
     The Federal standards are also known as "MACT" standards and
are based on the maximum achievable control technology previously
discussed.  The MACT standards may apply to both major and area
sources, and to both existing and new sources, although the
existing source standards may be less stringent than the new
source standards, within the constraints presented above.  The
MACT is considered to be the basis for the standard, but the
Administrator may promulgate more stringent standards which have
several advantages.  First, they may help achieve long-term cost
savings by avoiding the need for more expensive retrofitting to
meet possible future residual risk standards, which may be more
stringent  (discussed in Section 2.6).  Second, Congress was
clearly interested in providing incentives for improving
technology.  Finally, in the Clean Air Act Amendments of 1990,
Congress gave EPA a clear mandate to reduce the health and
environmental risk of air toxics emissions as quickly as
possible.
                               2-3

-------
     For area sources, the Administrator may "elect to promulgate
standards or requirements applicable to sources in such
categories or subcategories which provide for the use of
generally available control technologies or management practices
by such sources to reduce emissions of hazardous air pollutants."
These area source standards are also known as "GACT"  (generally
available control technology)  standards, although MACT may be
applied at the Administrator's discretion.
     The standards for hazardous air pollutants (HAP's), like the
new source performance standards (NSPS) for criteria pollutants
required by Section 111 of the Act (42 U.S.C. 7411), differ from
other regulatory programs required by the Act (such as the new
source review program and the prevention of significant
deterioration program) in that NESHAP and NSPS are national in
scope (versus site-specific).   Congress intended for the NESHAP
and NSPS programs to provide a degree of uniformity to. State
regulations to avoid situations where some States may attract
industries by relaxing standards relative to other States.
States are free under Section 116 of the Act to establish
standards more stringent than Section 111 or 112 standards.
     Although NESHAP are normally structured in terms of
numerical emissions limits, alternative approaches are sometimes
necessary.  In some cases, physically measuring emissions from a
source may be impossible or at least impracticable due to
technological and economic limitations.  Section 112(h) of the
Act allows the Administrator to promulgate a design, equipment,
work practice, or operational standard, or combination thereof,
in those cases where it is not feasible to prescribe or enforce
an emissions standard.  For example, emissions of volatile
organic compounds  (many of which may be HAP's, such as benzene)
from storage vessels for volatile organic liquids are greatest
during tank filling.  The nature of the emissions  (i.e., high
concentrations for short periods during filling and low
concentrations for longer periods during storage)  and the
configuration of storage tanks make direct emission measurement
                               2-4

-------
impractical.   Therefore, the MACT or GACT standards may be based
on equipment specifications.
     Under Section 112(h)(3), the Act also allows the use of
alternative equivalent technological systems: "If, after notice
and opportunity for comment, the owner or operator of any source
establishes to the satisfaction of the Administrator that an
alternative means of emission limitation" will reduce emissions
of any air pollutant at least as much as would be achieved under
the design, equipment, work practice, or operational standard,
the Administrator shall permit the use of the alternative means.
     Efforts to achieve early environmental benefits are
encouraged in Title I.  For example, source owners and operators
are encouraged to use the Section 112(i)(5) provisions, which
allow a 6-year compliance extension of the MACT standard in
exchange for the implementation of an early emission reduction
program.  The owner or operator of an existing source must
demonstrate a 90 percent emission reduction of HAP's (or
95 percent if the HAP's are particulates) and meet an alternative
emission limitation, established by permit, in lieu of the
otherwise applicable MACT standard.  This alternative limitation
must reflect the 90 (95) percent reduction and is in effect for a
period of 6 years from the compliance date for the otherwise
applicable standard.  The 90 (95) percent early emission
reduction must be achieved before the otherwise applicable
standard is first proposed, although the reduction may be
achieved after the standard's proposal  (but before
January 1, 1994) if the source owner or operator makes an
enforceable commitment before the proposal of the standard to
achieve the reduction.  The source must meet several criteria to
qualify for the early reduction standard, and
Section 112(i) (5) (A) provides that the State may require
additional reductions.
2.2  SELECTION OF POLLUTANTS AND SOURCE CATEGORIES
     As amended in 1990, the Act includes a list of 189 HAP's.
Petitions to add or delete pollutants from this list may be
submitted to SPA.  Using this list of pollutants, SPA will
                               2-5

-------
publish a list of source categories (major and area sources) for
which emission standards will be developed.  The EPA will publish
a schedule establishing dates for promulgating these standards.
Petitions may also be submitted to EPA to remove source
categories from the list.  The schedule for standards for source
categories will be determined according to the following
criteria:
     "(A) the known or anticipated adverse effects of such
pollutants on public health and the environment;
     (B) the quantity and location of emissions or reasonably
anticipated emissions of hazardous air pollutants that each
category or subcategory will emit; and
     (C) the efficiency of grouping categories or subcategories
according to the pollutants emitted, or the processes or
technologies used."
     After the source category has been chosen, the types of
facilities within the source category to which the standard will
apply must be determined.  A source category may have several
facilities that cause air pollution, and emissions from these
facilities may vary in magnitude and control cost.  Economic
studies of the source category and applicable control technology
may show that air pollution control is better served by applying
standards to the more severe pollution sources.  For this reason,
and because there is no adequately demonstrated system for
controlling emissions from certain facilities, standards often do
not apply to all facilities at a source.  For the same reasons,
the standards may not apply to all air pollutants emitted.  Thus,
although a source category may be selected to be covered by
standards, the standards may not cover all pollutants or
facilities within that source category.
2.3  PROCEDURE FOR DEVELOPMENT OF NESHAP
     Standards for major and area sources must  (1) realistically
reflect MACT or GACT;  (2) adequately consider the cost, the non-
air quality health and environmental impacts, and the energy
requirements of such control; (3) apply to new and existing
                               2-6

-------
sources; and (4) meet these conditions for all variations of
industry operating conditions anywhere in the country.
     The objective of the NESHAP program is to develop standards
to protect the public health by requiring facilities to control
emissions to the level achievable according to the MACT or GACT
guidelines.  The standard-setting process involves three
principal phases of activity:  (1) gathering information,
(2) analyzing the information, and (3) developing the standards.
     During the information-gathering phase, industries are
questioned through telephone surveys, letters of inquiry, and
plant visits by EPA representatives.   Information is also
gathered from other sources, such as a literature search.  Based
on the information acquired about the industry, EPA selects
certain plants at which emissions tests are conducted to provide
reliable data that characterize the HAP emissions from well-
controlled existing facilities.
     In the second phase of a project, the information about the
industry, the pollutants emitted, and the control options are
used in analytical studies.  Hypothetical "model plants" are
defined to provide a common basis for analysis.  The model plant
definitions, national pollutant emissions data, and existing
State regulations governing emissions from the source category
are then used to establish "regulatory alternatives."  These
regulatory alternatives may be different levels of emissions
control or different degrees of applicability or both.
     The EPA conducts studies to determine the cost, economic,
environmental,  and energy impacts of each regulatory alternative.
From several alternatives, EPA selects the single most plausible
regulatory alternative as the basis for the NESHAP for the source
category under study.
     In the third phase of a project, the selected regulatory
alternative is translated into standards, which, in turn, are
written in the form of a Federal regulation.  The Federal
regulation limits emissions to the levels indicated in the
selected regulatory alternative.
                               2-7

-------
     As early as is practical in each standard-setting project,
EPA representatives discuss the possibilities of a standard and
the form it might take with members of the National Air Pollution
Control Techniques Advisory Committee, which is composed of
representatives from industry, environmental groups, and State
and local air pollution control agencies.  Other interested
parties also participate in these meetings.
     The information acquired in the project is summarized in the
background information document (BID).  The BID, the proposed
standards, and a preamble explaining the standards are widely
circulated to the industry being considered for control,
environmental groups, other government agencies, and offices
within EPA.  Through this extensive review process, the points of
view of expert reviewers are taken into consideration as changes
are made to the documentation.
     A "proposal package" is assembled and sent through the
offices of EPA Assistant Administrators for concurrence before
the proposed standards are officially endorsed by the EPA
Administrator.  After being approved by the EPA Administrator,
the preamble and the proposed regulation are published in the
Federal Register.
     The public is invited to participate in the standard-setting
process as part of the Federal Register announcement of the
proposed regulation.  The EPA invites written comments on the
proposal and also holds a public hearing to discuss the proposed
standards with interested parties.  All public comments are
summarized and incorporated into a second volume of the BID.  All
information reviewed and generated in studies in support of the
standards is available to the public in a "docket" on file in
Washington, D.C.  Comments from the public are evaluated, and the
standards may be altered in response to the comments.
     The significant comments and EPA's position on the issues
raised are included in the preamble of a promulgation package,
which also contains the draft of the final regulation.  The
regulation is then subjected to another round of internal EPA
review and refinement until it is approved by the SPA
                               2-8

-------
Administrator.  After the Administrator signs the regulation, it
is published as a "final rule" in the Federal Register.
2.4  CONSIDERATION OF COSTS
     The requirements and guidelines for the economic analysis of
proposed NESHAP are prescribed by Presidential Executive
Order 12866 (EO 12866) and the Regulatory Flexibility Act  (RFA).
The EO 12866 requires review by Office of Management and Budget
(OMB) if the regulatory action is considered significant.  A
significant regulatory action is one that is likely to result in
a rule that may:
     1.  Have an annual effect on the economy of $100 million or
more, or adversely affect in a material way the economy, a sector
of the economy, productivity, competition, jobs, the environment,
public health or safety, or State, local, or tribal governments
or communities;
     2.  Create a serious inconsistency or otherwise interfere
with an action taken or planned by another agency;
     3.  Materially alter the budgetary impact of entitlements,
grants, user fees, or loan programs or the rights and obligations
of recipients thereof; or
     4.  Raise novel legal or policy issues arising out of legal
mandates, the President's priorities, or the principles set forth
in the Executive Order.
     The RFA requires Federal agencies to give special
consideration to the impact of regulations on small businesses,
small organizations, and small governmental units.  If the
proposed regulation is expected to have a significant impact on a
substantial number of small entities, a regulatory flexibility
analysis must be prepared.  In preparing this analysis, SPA takes
into consideration such factors as the availability of capital
for small entities, possible closures among small entities, the
increase in production costs due to compliance, and a comparison
of the relative compliance costs as a percent of sales for small
versus large entities.
     The prime objective of the cost analysis is to identify the
incremental economic impacts associated with compliance with the
                               2-9

-------
standards based on each regulatory alternative compared to
baseline.  Other environmental regulatory costs may be factored
into the analysis wherever appropriate.  Air pollutant emissions
may cause water pollution problems, and captured potential air
pollutants may pose a solid waste disposal problem.  The total
environmental impact of an emission source must, therefore, be
analyzed and the costs determined whenever possible.
     A thorough study of the profitability and price-setting
mechanisms of the industry is essential to the analysis so that
an accurate estimate of potential adverse economic impacts can be
made for proposed standards.  It is also essential to know the
capital requirements for pollution control systems already placed
on plants so that the additional capital requirements
necessitated by these Federal standards can be placed in proper
perspective.  Finally, it is necessary to assess the availability
of capital to provide the additional control equipment needed to
meet the standards.
2.5  CONSIDERATION OF ENVIRONMENTAL IMPACTS
     Section 102(2)(C) of the National Environmental Policy Act
(NEPA) of 1969 requires Federal agencies to prepare detailed
environmental impact statements on proposals for legislation and
other major Federal actions significantly affecting the quality
of the human environment.  The objective of NEPA is to build into
                                                        •
the decision-making process of Federal agencies a careful
consideration of all environmental aspects of proposed actions.
     In a number of legal challenges to standards for various
industries, the United States Court of Appeals for the District
of Columbia Circuit has held that environmental impact statements
need not be prepared by EPA for proposed actions under the Clean
Air Act.  Essentially, the Court of Appeals has determined that
the best system of emissions reduction requires the Administrator
to take into account counterproductive environmental effects of
proposed standards as well as economic costs to the industry.  On
this basis, therefore, the Courts established a narrow exemption
from NEPA for EPA determinations.
                               2-10

-------
      In addition to these judicial determinations,  the Energy
 Supply and Environmental Coordination Act (ESECA)  of 1974
 (PL-93-319)  specifically exempted proposed actions under the
 Clean Air Act from NEPA requirements.  According to
 Section 7(c)(1) ,  "No action taken under the Clean Air Act shall
 be deemed a major Federal action significantly affecting the
 quality of the human environment within the meaning of the
 National Environmental Policy Act of 1969" (15'U.S.C. 793(c)(1)).
      Nevertheless,  EPA has concluded that preparing environmental
 impact statements could have beneficial effects on certain
 regulatory actions.  Consequently, although not legally required
 to do so by Section 102(2)(C)  of NEPA, EPA has adopted a policy
 requiring that environmental impact statements be prepared for
 various regulatory actions,  including NESHAP developed under
 Section 112 of the Act.  This voluntary preparation of
 environmental impact statements, however, in no way legally
 subjects the EPA to NEPA requirements.
      To implement this policy,  a separate section is included in
 this document that is devoted solely to an analysis of the
 potential environmental impacts associated with the proposed
 standards.  Both adverse and beneficial impacts in such areas as
 air and water pollution, increased solid waste disposal, and
 increased energy consumption are discussed.
 2.6  RESIDUAL RISK STANDARDS
      Section 112 of the Act provides that 8 years after MACT
 standards are established (except for those standards established
 2 years after enactment, which have 9 years),  standards to
 protect against the residual health and environmental risks
 remaining must be promulgated,  if necessary.   The standards would
 be triggered if more than one source in a category or subcategory
 exceeds a maximum individual risk of cancer of 1 -in 1 million.
• These residual risk regulations would be based on the concept of
 providing an "ample margin of safety to protect public health."
 The Administrator may also consider whether a more stringent
 standard is necessary to prevent--considering costs, energy,
 safety, and other relevant factors--an adverse environmental
                                2-11

-------
effect.  In the case of area sources controlled under GACT
standards,  the Administrator is not required to conduct a
residual risk review.
                               2-12

-------
          3.0   THE  MAGNETIC TAPE MANUFACTURING INDUSTRY
                PROCESSES AND POLLUTANT EMISSIONS

3.1  GENERAL
     This chapter describes the magnetic tape manufacturing
industry, the processes used in this industry, the sources of
pollutant emissions, the factors affecting emissions, baseline
emissions for the industry, and the methodology used to calculate
baseline emissions.
3.1.1  Industry Description
     The magnetic tape manufacturing industry is included in the
Standard Industrial Classification  (SIC) Codes 3695, "Magnetic
and Optical Recording Media," and 2675, "Die-Cut Paper and
Paperboard and Cardboard."  In the manufacturing process, a
mixture of magnetic particles, resins, and solvents  (the coating
mix) is applied on either a plastic film or paper.  The hazardous
air pollutants (HAP's) used in this industry are listed in
Table 3-1 (solvents and particulates).  The HAP solvents used to
                              •
the greatest extent are methyl ethyl ketone (MEK), toluene, and
methyl isobutyl ketone (MIBK).  Other, non-HAP solvents used in
this industry include cyclohexanone, tetrahydrofuran (THF),
acetone, and to a lesser extent, isopropyl alcohol and
1,1,1-trichloroethane.1
     The solvent HAP emissions result from the evaporation of
solvents from the  (1) storage tanks, (2) coating mix preparation
area, (3) coating application/flashoff area,  (4) drying oven,
(5) packaging and labeling operations,  (6) waste handling area,
                               3-1

-------
TABLE 3-1.  USES OF HAP'S AT MAGNETIC TAPE MANUFACTURING  PLANTS1
HAP
Methyl ethyl ketone
Methyl isobutyl ketone
Toluene
Toluene diisocyanate
Ethylene glycol
Methanol
Xylenes
Chromiuma
Cobalt3
Ethylbenzene
Acetaldehyde
Use
Coating, cleaning,
packaging/labeling (1 plant)
Coating, cleaning
Coating, cleaning
Coating
Packaging and labeling
(1 plant)
Coating (1 plant) , packaging
and labeling (1 plant)
Coating, cleaning (1 plant)
Coating
Coating
Contaminant in another solvent
(1 plant)
Reaction product (1 plant)
aParticulate HAP's,
                               3-2

-------
(7) cleaning activities, and  (8) piping fittings.  The drying
oven is the largest solvent HAP emission source.
     The magnetic particles are combinations of iron, chrome, or
cobalt.  Particulate HAP emissions can result from handling and
from dry media being added to mix tanks during coating mix
preparation.
     Magnetic media on plastic film is used primarily for audio
and video recording and computer information storage.  Table 3-2
presents a summary of the physical parameters of the three major
types of magnetic tape products.2  Other uses of magnetic media
on plastic film include magnetic cards, credit cards, bank
transfer ribbons, instrumentation tape, and dictation tape.
Magnetic coatings on paper are used for toll tickets, airline
boarding passes, and security badges.
     Over'the last decade, manufacture of products such as Beta
video tape, 8-track audio tape, dictation tape, and magnetic
cards for word processing has decreased.  Products developed in
recent years include 8-millimeter (mm) tape for video and data
storage and rotary head digital audio tape for data storage.3
     Table 3-3 presents the names, locations, and products of the
domestic plants that make up the source category.  There are
25 plants  (a 26th plant is in the planning stage), representing
21 companies located in 17 States.  Five plants apply coatings to
paper; two purchase rather than prepare their coatings.
     Unlike many manufacturing processes, magnetic tape
manufacturing is not restricted by raw material or market
requirements to select regions of the country.  California, with
five plants, has the largest population.  Twenty (76 percent) of
the plants coat only one type of product, three  (12 percent)  coat
two types,  and three (12 percent) coat three types.  Each plant
contains from one to nine coating lines.1
     Since 1988, 17 magnetic tape manufacturing facilities have
shut down.   The primary reasons for these closures appear to be
stiff foreign competition that provides lower-cost products and,
relatedly,  the high labor cost of production in the
United States.4"6  Both small, independent companies and large,

-------




















W
w
E-i
s

*3
2
s
04
Ei
U
O
O
OS
04

u
O4
EH


H
B
2
O
f^

*

.
CM
I


W
0)
EH































M
a
0)
c
u
2:
H

































































m
w1
4J
«
o


^*
14


0>
c
•£
id
u

4J
0)








£
•o
jj










JJ
JJ
0»
g
a>
>j



















•












>t

"a




a
I


a
^
"e



id
1






C













-p








































































































_5
*rt



u

3
Q.
o

f4
•
o




m
•
CN


in
in

d


en
•
i-i





in
d





m
i^






O
o
I-l
1-1





vO

m











•a

u
^
14
(ft
in
u

(flj
•U
id
Q





rn
•
O




vO
*
*"•


t- o

O'CN

i
ao
•
r- 0
1-1 in





in
0





(T)
I-l






O
0
CN





CN
O.
f^









*-*
— 1
0)
0)
u

m .
O.C

o











at
-U
jj
01
a
0)
 0
• •
CN CM
1-1 n



A
J3 in
in cvU3
ro in ao





C\ f\ f\
cn ro fj
oo n o
»H CN
























0
.X
(0
.^4
•a

0
J3

X
01
r-4
U.





































































(It

•a
e

Q
.-4
•0
3
«
j^
C
1-4
03


































































03
0
4J
01
a
a








o o
0 0
CN O
t^d
^^




in ao
on ^
^ 0

CN n











u
01
0)
(ll
HI
U
c
id
cu







CN

d




i-i
t
m


i
* ,
1-4 «H

1
on en
• •
r- in
CN m




in
i-i
d



CO

ro







1 in
in ao
CN 00
CN






1 O
en t^
VO CN









0)
Oi
id
4J

^4
0)

a
c
o
u










































































u
>d
u

1
ao





























m
CN
d



^,
»
vO







O
O
00*





O
VO
in
«
CN












(0
01

u
c
id
cu







o

d




0

-1

i
m in
i-i in

do

i
ao en
* •
fO
n 3
3 O1
jj 01
U]









in
r-l





m CN
CN •-!













in
CN
o



^
v
vO







0
0
CO
1-1





ai
^
tn




r-l
0
0
Vj

^
r-t
It) JJ
C C
0 0)
—4 ^H
m id
ID >
VM 3
0 O1
U a)
a.






3-4

-------
•O
 0)
 3

•H
-P

 O
 O
(N
00






a
0
01
u
£
H








0»
JJ
id
0
U
14
o
o.
c
7)
id
u

^
0)
s



jfi
-p
•o
3!




•U
0)



id
S



id
g
id
••*
8



id
g
a.


c





1

£


g


















































0 H
0




•














in
r*
O



l-l
a\
<-t
0 0
o o
\O CN

n so
00 ^O
0



o
•- t
[J
id
g
3




































S,
id

u
0

o
C
0
u



CN
d




*
i
. .
O fl

i
CN *O
• •
o m
^

in
 CN
0



X



4J
01
03





































2,
id
4J



oo




                                                                                                   01
      u
      d)
      JJ
      V4
      o
      a
      JJ

    • O
   a
   -* 0
   a jj
   -^
   TJ u
      0
   14 --I
   id u
   -4 ft

   U 01
   U 04
 « -^ «
C O 4J
                                                                                            rH  0-1
                                                                                            O     3
                                                                                            o  o a

                                                                                            O  <0  Qt
                                                                                                •U -4
                                                                                             ii   -H
                                                                                             u -i  a»
                                                                                             o  9 01

                                                                                             g  a
                                                                                                01  0)
                                                                                             II  3  O»
                                                                                               iH  VI
                                                                                             e  id  id

                                                                                           id A  u
                                                                           3-5

-------
TABLE 3-3.  PLANTS COATING MAGNETIC TAPEa
Facility/location
Alabama
Sony, Dothan
Ampex, Opelika
JVC, Tuscaloosa
Arizona
3M, Tucson
California
Bav Area
Verbatim, Sunnyvale
Ampex, Redwood City (research)
Memorex, Santa Clara
Tandy, Santa Clara
Ventura County.
3M, Camarillo
Georgia
Sony Music, Carrollton
Illinois
Rand McNally, Skokie
XDP, Gary
Indiana
Sealtran, Richmond (planned for construction)
Maryland
Malco Plastics, Garrison
Minnesota
3M, Hutchinson
3M, St. Paul
Nebraska
Anacomp, Omaha
New Jersey
Fidelipac
North Carolina
EDM, High Point
Oklahoma
3M, Weatherford
South Carolina
Fuji, Greenwood
South Dakota
Syncom, Mitchell
Tennessee
NCR, Morristown
Type of product
Audio
X
X

X
X
X
X


X

X





Video
X
X
X

X




X




X


Computer
X
X
X
X
X





X


X

X

Paper5




X
X

X


X




Other0
X

X



X


X




X
                   3-6

-------
                               TABLE 3-3.    (continued)
Facility/location
Texas
Graham Magnetics, Inc. Graham
Magnetic Ticket & Label, Dallas
Virginia
Audiopak, Winchester
Type of product
Audio

X
Video


Computer
X

Paper5
X

Other0

X
aAs of August 1992.
Includes toll tickets, airline boarding passes, and security cards.
°Includes bank transfer ribbons, magnetic cards, credit card tape, and instrumentation tape.
                                             3-7

-------
multinational corporations have exited the magnetic coating
business.  Other large companies have consolidated production
into fewer facilities, both domestic and foreign.  Data are not
available to indicate whether or not magnetic tape production
overall has decreased.
     Other activities that take place at some of the plants
include audio tape prerecording, record production, compact disc
production, and injection molding and assembly of plastic parts
such as cassettes.  Other activities that take place at some
plants on the same equipment that is used for magnetic media
production include production of "leader" tape (tape at the
beginning of an audio cassette without magnetic particles),
graphite liners for cassette boxes, and plastic film with
adhesive coating.
3.1.2  Industry Growth
     The magnetic tape recording industry began in the late
1940's with reel-to-reel audio tape recorders.  Commercial
industrial development of video and computer tape began in the
1950's, and individual home use of these products has increased
greatly since the mid-1970's.7  The value of magnetic tape
shipments in the United States from 1984 to 1990 is shown in.
Figure 3-I.8"13  In 1990, the total value of magnetic tape
shipped was approximately $3 billion, of which 12 percent was
audio tape, 42 percent was computer tape, 36 percent was video
tape, and 10 percent was "other" tape.13
     Some sources estimate that between 1989 and 1994, the
production of 5.25-inch  (in.) disks  (floppies) will decrease by
17 percent while the production of 3.5-in. floppies will increase
by 23 percent.14  Another source estimated that up to 11 new
coating lines would be built in the United States by the end of
1992, at both new and existing plants.1^  Information supplied
directly by domestic companies in this industry indicates that,
while modifications of existing lines may occur, construction of
new lines is unlikely.
     Current data indicate that five new lines were built between
1988 and April 1992; two more lines are planned for construction
                               3-8

-------
  3,500
  3,000
e
O
in
2,500
  2,000
CO
HI
CO

3
CO
  1,500
  1,000






    500





         I        I     '	T—
      0 U	'	L

        1984    1985   1986    1987   1988    1989   1990

                              YEAR


         TOTAL COMPUTER VIDEO AUDIO OTHER
    Figure 3-1.  Annual shipments of blank magnetic tape.


                           3-9

-------
in 1992.  Of the seven new or planned lines, four are in existing
facilities and three are in a new facility.1  For further details
of the historical and future growth in this industry, see
Section 9.1.
3.2  PROCESSES AND THEIR EMISSIONS
3.2.1  Process Descriptions
     The process for manufacturing magnetic tape consists of
mixing the coating ingredients (magnetic particles, resins, and
solvents), conditioning the base film, applying the coating to
the base film, orienting the magnetic particles, removing the
solvents by evaporation in a drying oven, and finishing the tape
by calendering, rewinding, slitting, testing,  and packaging.
Figure 3-2 presents a schematic of a magnetic tape coating line.
The solvent HAP emissions primarily result from coating and
drying of the tape and, to a lesser degree, solvent storage, mix
preparation, transferring solvent through piping, cleaning of the
equipment, treating solvent-laden waste material, and packaging
and labeling the final product.  The following sections describe
the raw materials used and the magnetic tape manufacturing
process in more detail.
     3.2.1.1  Raw Materials.  The two fundamental components of
magnetic tape are the magnetic coating and the base substrate to
which the coating is applied.  Several types of plastic base
films have been used, but polyester is the most common substrate
currently used because it has the best combination of chemical
and mechanical properties, availability,  and cost.16  Polyester
film can be used with any magnetic tape coating formulation.  The
thickness of the base film varies with the product, ranging from
5.1 to 192 microns (pm) (0.20 to 7.50 mils).16'17  The width of
the film ranges from 15 to 107 centimeters (cm)   (6 to 42 in.).1
     Paper is used as the base support for some products.  One
manufacturer reported that the width of the paper substrate at
the facility is 53.3 cm (21 in.), on which a stripe or stripes of
magnetic coating are'applied at a thickness of 50.8 to 76.2 /zm
(2 to 3 mil),18
                               3-10

-------
                                                        C
                                                        A
                                                        cn

                                                        •H
                                                        -P
                                                        fl
                                                        o
                                                        u

                                                        0)
                                                        o
                                                        •H
                                                        -P
                                                        (1)
                                                        C
                                                         c
                                                        •H
                                                         O
                                                        •H
                                                        -P
                                                         (fl

                                                         a)
                                                        £
                                                         u
                                                         I
                                                        n
                                                         3

                                                        •H
                                                        tu
3-11

-------
     The exact composition of the coating may vary slightly with
the desired end use of the magnetic tape and the targeted
"quality" of reproduction.  All types of coating mix, however,
combine three types of components:  magnetic particles, binder,
and solvents (solvent serves as the vehicle for the particles and
resins).  Table 3-4 presents coating composition ranges by both
weight and volume.1'2
     Six types of magnetic particles are used in magnetic tape
manufacturing:  7-ferric oxides, cobalt-doped 7-iron oxides,
chromium dioxide, barium ferrite, magnetite and metallic
particles that usually consist of elemental iron, cobalt, or
nickel.1^  Of these particulates, the chromium and cobalt
compounds are HAP's.  They are used primarily in audio tape.  The
magnetic particles normally make up about 10 to 39 percent  (by
weight) of the coating mix.20'21  The magnetic particles are
bonded to the film by a permanent coating.  The binder (resin
and/or cross-linker) is an organic polymer that holds the
magnetic materials together in a flexible matrix that adheres to
the base film or matrix.  Most coating mixes contain thermoset
binders, particularly polyurethanes and polyvinyls.22  The
coating mix normally contains about 2 to 10 percent  (by weight)
binder.20'23
     Solvents are used to dissolve the binder polymers and to
provide a fluid medium for the dispersion of particles in the
coating mix.24  The major solvents used in the coating mix when
applied to a plastic film are THF, MEK, MIBK, toluene, and
cyclohexanone.1  Of these, MEK, MIBK, and toluene are HAP's.
Other HAP compounds used in lesser quantities in coatings are
xylene, methanol, and toluene diisocyanate.  Various combinations
of the solvents may be used.  Coating mixes applied to paper
substrates contain MEK, THF, and acetone in various combinations.
     Factors affecting solvent selection are toxicity, avail-
ability, cost, ease of solvent recovery after use, desired rate
of evaporation, and effect on solvent recovery equipment.24'25
The solvent in the coating mix for plastic substrates ranges from
50 to 85 percent  (by weight),1'25  The solvent content of the
                               3-12

-------
TABLE 3-4.  SELECTED COATING MIX PROPERTIES1'2
Parameter
Solids
VOC
Density of coating
Density of coating
solids
Res ins /binder
Magnetic particles
Density of magnetic
material
Viscosity
Unit
% by weight
% by volume
% by weight
% by volume
kg/L
Ib/gal
kg/L
Ib/gal
% by weight of solids
% by weight of solids
kg/L
Ib/gal
Pa.s
Ib/ft.s
Range
15-50
10-26
50-85
74-90
1.0-1.2
8-10
2.2-4.0
18-33
15-21
66-78
1.2-4.8
10-40
2.7-5.0
1.81-3.36
                     3-13

-------
coatings for paper substrates is approximately 60 percent  (by
weight}.1'26  In some cases, evaporation of the solvent causes
the coating to harden; other coatings are chemically reactive and
cure through polymerization of the resin oligomers.  Solvent
evaporation in the drying oven is the primary source of volatile
organic compound (VOC) emissions from the coating facilities.
     Other coating components, which are generally not HAP's, may
include (1) dispersants (1 to 5 percent by weight) to prevent the
magnetic particles from agglomerating; (2) conductive pigments
(1 to 4 percent by weight) to prevent the buildup of static
charge; (3) lubricants (less than 2 percent by weight) to
minimize head-tape friction and wear on the tape; and
(4) miscellaneous additives (1 to 3 percent by weight), such as
mild abrasives for cleaning the head or fungicides to control
mildew.16'23
     After the substrate is coated and dried, the product must be
packaged and labelled for shipping.  Small amounts of three HAP's
(MEK, methanol, and ethylene glycol)  have been identified as a
component of inks used for printing onto packaged products.  One
plant uses MEK, and the second uses all three HAP compounds in
printing inks.1  One plant uses a small amount of methanol in the
tape cartridge manufacture.
     Cleaning of coating mix preparation, coating application,
and packaging equipment also involves using solvents.  Solvents
that are used for cleaning in this industry, either individually
or in blends, are THF, toluene, acetone,  MEK, cyclohexanone, and
to a lesser extent isopropyl alcohol and 1,1,1-trichloroethane.1
Of these,  toluene and MEK are HAP's.  Usually, one of the solvents
used in the coating formulation is also used for cleaning,
because it is the one that can best dissolve the binder and resin
in the coating.
     3.2.1.2  Storage of HAP's.  Generally, small tanks are used
to store solvent HAP's used in magnetic tape production.  These
tanks may be horizontal or vertical and are sometimes below
ground.  The tanks operate at atmospheric pressure or slightly
above atmospheric pressure.  Typically, there are from l to
                               3-14

-------
12 tanks at a facility, ranging in total capacity from 757 to
75,700 liters (L) (200 to 20,000 gallons [gal]).1  Solvents used
in small amounts, either for specialty purposes or at plants with
small production rates, are sometimes received and stored in 30-
or 55-gal drums.
     Particulate HAP's are received and stored at the plants in
paper bags or in tote bins.  The sealed containers arrive at the
plant and are stacked in room(s) near the mix preparation area.
     3.2.1.3  Coating Mix Preparation.  The coating mix
preparation room is separate from the coating line.  One room or
several rooms may contain the mix preparation equipment for all
of the coating lines at the plant.  One set of mix equipment can
be used for more than one line or product.   The number of pieces
of mix equipment serving a line or product varies widely.  One
plant may have as few as 1 piece of mix equipment, while another
may have over 100 pieces of mix equipment.1
     The coating mix is prepared in batches.  The duration and
frequency of batch preparation depends on company practice and
the hours of operation.  The following describes the production
sequence for most, but not all, magnetic coatings.  The process
begins with the blending of the components in low-shear mixers.
The mix is then transferred to a series of mills  (e.g.,  sand,
ball, high-speed, colloid, small-media, or roll), where the
dispersing action of beads, combined with the high shearing
forces of the centrifugal mixing action, thoroughly disperse the
aggregates of magnetic particles without reducing particle size.
The final step in the process is polishing, during which the
conductive carbon black is added.  Polishing improves the final
appearance of the tape by making it smoother or glossier.  The
completed mix is then continuously circulated and filtered in
holding tanks to prevent binders from curing and metal particles
from settling out and to remove any oversize contaminants.27-29
Table 3-4 presents the range of values for selected properties of
coating mixes used in the industry.1'2
     Some plants may skip one of these steps (e.g., polishing)  or
include additional mixing steps.  The facilities that purchase
                               3-15

-------
coating mixes in drums have only one or two mixers to ensure that
the purchased coatings are thoroughly blended before
application.30
     In the majority of the plants, the solvents are pumped to
and from the different pieces of mix equipment through closed
lines.  At a small operation, employees may pour the solvents by
hand into the mixers and mills.  Particulates can either be
transferred through closed systems or be manually poured through
hatches in the covers of the mix equipment.
     3.2.1.4  Conditioning.  Before the coating mix is applied,
the plastic substrate may undergo conditioning steps.  Some
precision products, such as videotape, have a nonmagnetic coating
on the back of the tape (backcoating), which provides a
conductive surface that minimizes static buildup, enhances
handling, and increases abrasion resistance.1**'31  Backcoatings
are sometimes used to increase the "slipperiness" of the tape
(reduce friction) so it will unwind within the cassette more
easily.  Backcoating is done before the magnetic coating is
applied, using the same solvents and the same equipment that are
used in coating the tape with the magnetic material.  The
thickness of the backcoat generally ranges from 1.0 to 1.5 pm
(0.04 to 0.06 mils).7
     As the plastic film is unwound, it can be cleaned by some
method such as "tacky11 rolls or jets of air.  It is then passed
over rollers-, which may be heated, to remove wrinkles from the
film.28
     3.2.1.5  Coating Application.  In the coater, the substrate
passes over a backup or support roll while the coating mix is
applied either by another roll or by extrusion under pressure
through a narrow slot in a die.  The layer of wet coating mix
that is applied ranges in thickness from 2.4 to 63.5 pm (0.09 to
2.5 mils).1/2  The amount of coating mix applied by a coater is
precisely measured and controlled.  When coating a plastic film,
the entire width of the substrate is coated, with the possible
exception of a narrow band on either edge.  When coating a paper
substrate, the magnetic coating is applied in a narrow band or
                               3-16

-------
bands (up to 0.5 in. in width) that extend the length of the
roll.18
     Four types of coaters are used for production of all types
of magnetic tape:  extrusion  (slot die), gravure, knife, and
reverse roll (3-roll and 4-roll).28  Figures 3-3 and 3-4 present
schematic drawings of these coaters.  Coaters range from 15 to
107 cm (6 to 42 in.) in width and operate at speeds ranging from
15 to 366 meters per minute (m/min) (50 to 1,200 feet per minute
[ft/min]).1  Extrusion and gravure coaters apply coatings
uniformly at speeds in the higher end of the range.  Knife
coaters are not typically used in manufacturing precision
products such as computer tape.  Reverse roll coaters are used in
tape manufacturing when thicker coatings are required.28'32  The
range of coating mix viscosity that can be applied varies with
the type of coater.
     Immediately following the coater, the plastic film is guided
through an orientation field consisting of an electromagnet or
permanent magnet, which aligns the individual magnetic particles
in the direction of the intended recording.  Films from which
flexible disks are produced do not go through the orientation
process because magnetic particle alignment is not required.28'33
     High-performance tapes require clean working conditions,
especially in the coating application and oven areas, where dirt
and unclean work areas may lead to poor tape quality.  Therefore,
coating areas are sometimes maintained as clean rooms.  A clean
room uses filtration to limit the quantity of certain size
particles that can enter a given volume of space.  The quantity
and size of particles that can enter depend on the "class" of the
clean room.  A Class 10 clean room allows only a small quantity
of the smallest size particles to enter a given volume.  A Class
100,000 clean room allows larger quantities of particles to enter
a given volume of space.
     3.2.1.6  Drying.  The coated plastic film then passes
through a drying oven, where the solvents in the coating mix
evaporate.  Figure 3-5 presents a schematic of an air flotation
oven, which is the type of oven typically used in this industry.

-------

\
i.
                                                      DC
                                                      HI
 o
 w
 Q


 §
 CO,


 O
 CO
 Z>
 cc

X
LU
                                                                                        c
                                                                                        o
                                                                                        •H
                                                                                        -P
                                                                                        (0
                                                                           C
                                                                           O
                                                                           u

                                                                           •a
                                                                           fl
                                                                           0)
                                                     DC |
                                                     Hi
                                                     S
                                                     O
                                                     LU
                                                    O
                                                    g
                                                    GL
                                                                                       •H
                                                                                       -U
                                                                                        03
                                                                                        O
                                                                                       o
                                   
-------
                                                  CC
                                                  UJ|
                                                  O
                                                  u
                                                  O

                                                  g
                                                  (0
                                                  cc
                                                  Ul
                                                  \u
                                                  cc
                                                 O
                                                 cc
                                                 UJ
                                                 Ul
                                                 QC
                                                cc
                                                I
                                               2
                                               Ul
                                               8
                                               Ul
                                               Ul
                                               flC
                                               O
                                               tr
                                               cc

                                               O
  en
 TJ
  (fl
  (U
 x:
                                                           •H
                                                           JJ
                                                            fi
                                                            O
                                                            0

                                                            0)
                                                            a
                                                            I
                                                            o>
                                                            c
                                                            (U
                                                           -p
                                                            0)
                                                           s
 I
 r>

 
-------
/\
                                           c
                                           0)

                                           o
                                           c
                                           •H
                                           O
                                           ItJ

                                           -P

                                           O
 M -
                                           in
                                           I
                                           0)
                                           Li
                                           3
3-20

-------
In this oven, the film is supported by jets of drying air and
never touches any metal, thus reducing abrasion and deformation
and eliminating the need for reorientation of the magnetic
particles after drying.  Some ovens are blanketed with a nitrogen
atmosphere in the oven  (see Section 4.2.2.I).1  For paper
magnetic products, a festoon oven may be used, in which the
uncoated side of the paper is in contact with a series of
rollers.30  A schematic of a festoon oven is provided in
Figure 3-6.  Ovens range from 0.6 to 1.2 m (2 to 4 ft) in width
and 12.2 to 30.5 m (40 to 100 ft) in length.34'35  Oven
temperature settings range from 16° to 177°C  (60° to 350°F).1
     The airflow within the oven is countercurrent to the
direction the film is traveling.  The air is conditioned before
entering the oven to "remove" dust particles and to adjust the
temperature and humidity.  Negative pressure ovens reduce leakage
of solvent emissions from the oven into the room but special
conditions may be required to minimize contamination problems.
At magnetic tape manufacturers, most ovens are operated at
negative pressure.1  The airflow rate in the oven is adjusted so
that the solvent concentration is maintained at less than
25 percent of the lower explosive limit (LEL) of the solvent/air
mixture for the particular solvent or solvent mix used.1  Airflow
rates vary with the line size, solvent evaporation rates, and
company practice.  Total individual oven exhaust flow rates range
from 0.5 to 7.1 standard cubic meters per second (m3/sec)
(1,000 to 15,000 standard cubic feet per minute [scfm]).1  Air
from other parts of the coating line may be used as oven makeup
air, and the air may be recirculated within the oven.  Practices
include pulling oven makeup air from the atmosphere, oven room,
coater room, or combinations of these sources.
     3.2.1.7  Finishing Processes.  For products coated on
plastic film, the dried film may be squeezed between sets of
calender rolls to compact the dry coating and smooth the surface
finish.29'33  The amount of calendering varies with the product;
not all products require compressing.  The final dry coating
                               3-21

-------
                                   f
                                   <5
                                   •a
                                   co
                                   o
•3

•Si
                                                                                 c
                                                                                 0)

                                                                                 o

                                                                                 c
                                                                                 o
                                                                                 o
                                                                                 4J
                                                                                 U)
                                                                                 
-------
thickness on the film ranges from 1.0 to 10.8 jim (0.04 to
0.4 mil), depending on the product specification.1
     Nondestructive testing is performed on up to 100 percent of
the final product.  The percentage of tape tested increases as
the level of precision required in the final product increases.
     3.2.1.8  Packaging and Labeling.  After being coated and
rewound, the plastic-based tape may then be slit into the desired
tape widths by means of a rotary shearing operation.  The tape
used to make flexible disks is not slit; dies are used to punch
                                      o fl
the disks from the finished substrate.^°
     Some facilities mold the various plastic parts, such as
cassettes, and assemble the complete final product.  Only one
plant was identified as using a HAP  (methanol) in the molding
process.  Some plants purchase the plastic casings and assemble
the final product.  Some plants ship the coated film in bulk to
other facilities for slitting and final packaging.
     Whatever the final form of the product, printed materials
such as labels, boxes, and inserts are usually part of the final
package.  Only two facilities were identified as using very small
amounts of HAP's for "printing" in-house.1  In fact, these
operations are minuscule in that only product identification
codes are applied to boxes.  Most facilities purchase these items
preprinted.
     3.2.1.9  Cleaning.36"70  The frequency and type of cleaning
depend on the type of equipment being cleaned, production rate,
number of different products, and company practice.36"70  Data
collected from the industry do not indicate that frequency of
cleaning is a function of the type of product manufactured.71
Some plants clean equipment between each batch of coating, even
if the same coating is being made in the successive batches.
Other companies clean only between product changes.36'37'39  A
few facilities clean the mix tanks as seldom as once or twice per
year.  Sometimes fresh solvent is used once and immediately
treated as a waste.  At other plants, cleaning solvent may be
used several times before it is considered "spent."  All plants
were found to either regenerate cleaning solvent in-house or send
                               3-23

-------
it to a permitted offsite facility for reclamation or
disposal.36'70
     The cleaning activities identified in the magnetic tape
manufacturing industry were considered as one of four categories.
These were:   (1) flushing fixed lines; (2) cleaning open-top and
closed-top tanks; (3) cleaning fixed exterior surfaces; and
(4) cleaning miscellaneous removable parts.  The following
                                                  *       *
paragraphs describe each in more detail.
     One type of cleaning performed at magnetic coating plants is
the flushing of fixed lines, such as piping that transfers the
coating mix to the coater head.  Depending on the size and length
of the item being cleaned, from 38 to 190 L (10 to 50 gal) of
clean solvent was repeatedly used for each flush.  Flushing
frequency was found to vary from several times each day to less
than once per month.47-49'54'57'63'65'67-71
     For open- and closed-top tank cleaning, spraying is often
used and can be either manual or automated.  Manual spraying
employs a hose or wand through which the cleaning solvent is
pumped.  With automated spraying, a multi-orificed spray head,
either permanent or removable, is fitted onto the top of the
tank, through which cleaning solvent is pumped.  Manual spraying
occurs in an open tank, so the operator can reach in with the
hose or wand.  Each cleaning event lasts from 15 minutes to
4 hours and uses from about 4 to 250 L (1 to 65 gal) of
solvent.41"71  Data provided by the industry were counter-
intuitive in that they indicated similar solvent losses for both
open- and closed-top cleaning methods.
     Internal tank surfaces may also be scrubbed with brushes or
abrasive pads.  Manual spraying frequently, but not always, is
accompanied by scrubbing.  Scrubbing may last from a. few minutes
to several hours, depending on the type of dirt being removed
(slurry or dried residual dirt).  When cleaning is complete, the
dirty solvent is either drained from the bottom of the mix tank
and stored in a 55-gal drum or piped directly to a waste solvent
holding tank.
                               3-24

-------
     Tank cleaning may involve a combination of open- and
closed-top methods; i.e., a "hybrid" method.  Solvent may be
pumped in while the tank lid is on.  After solvent is initially
flushed through the system, an operator may open the lid to do
some scrubbing.  The final rinse may then be done with the lid
opened or closed.
     Cleaning fixed external surfaces with items such as brushes,
rags, paper towels, and lint-free wipes is another cleaning
category.  The items cleaned in this manner are the calender roll
bars, external parts of the coater head, and the slitter blades
for slicing the wide roll of tape into the final product widths.
Cassette assembly machinery and substrate smoothing bars are
other examples of fixed external surfaces.  A small can
containing a few gallons of cleaning solvent is usually kept near
the item to be. cleaned.  A rag is dipped into the can or pressed
to a mesh covering over the can to acquire some solvent, which is
then wiped over the item to be cleaned.  Mesh-covered safety cans
are discussed in Section 4.6.3.  Rags are usually used more than
once.  Sometimes the can of solvent is used until it is empty,
while at other facilities it is used until visible contamination
is seen in the can.  Some parts require wiping several times per
day, others as infrequently as once per week.  Spills of solvent
or coating are cleaned with rags or with absorbent material that
is then swept up.  Both the rags and absorbent material are then
disposed of properly, depending on the waste classi-
fication.41'49'51-53'58-^
     Another category of cleaning is the cleaning of small
removable parts such as pump parts, coater parts, filter
housings, and mix blades.  Most magnetic tape manufacturing
plants have one to four wash sinks for this purpose.  The surface
area of wash sinks ranges from about 0.4 to 1.1 m2  (4 to 12 ft2).
Cleaning solvent is pumped in through hoses.  The sink drains to
either a waste solvent tank or 55-gal drums.  Some facilities use
the solvent for only one cleaning event.  Others retain it in the
sink until it is considered too dirty for further use.  Parts are
scrubbed in the sink using brushes.  Some sinks are used
                               3-25

-------
frequently (several times per day),  others infrequently  (every
few days).41'49'51-66
     3.2.1.10  Waste Handling.72"79  The wastes produced at
magnetic coating plants fall into two categories:  liquid and
solid.  The liquid wastes consist of dirty cleaning solvents and
waste coatings that were left over at the end of a production
run, did not meet specifications, or were "planned wastes" for
quality control purposes.  Solid wastes are primarily soiled mix
rags, towels, brushes, absorbent used to clean spills, and cups
used to dip coating samples for analysis.  Some facilities treat
some or all of their wastes onsite to recover the solvent, while
others ship the waste offsite for treatment or disposal.
     If handled onsite, liquid wastes are treated in a
distillation process.  Following generation of the waste, the
liquid, is pumped or manually transferred into 55-gal drums or
pumped directly to a waste holding tank.  The liquids are either
poured or pumped from the drums or piped from an intermediate
waste holding tank into the treatment device.  The device can be
either a pot still or a thin film evaporator.1  In either case,
the accumulated liquid waste is heated for up to 16 hours to
evaporate the solvent, which then passes through a condenser, and
the condensed solvent is collected in a receiving vessel.
Depending on the cleanliness of the recovered solvent, its
intended use, and company practice,  the recovered solvent is
either sent to a storage tank for reuse or sent to the
distillation system handling the solvent recovered from the main
VOC emission control device.  The dry residue (still bottoms)
that remains in the device is sent offsite as discussed in
subsequent paragraphs.
     If handled onsite, solid wastes are treated with heat via a
filter dryer to recover the solvents.1  Plant-generated waste is
typically stored in 55-gal drums.  Sometimes mechanical
compaction is used to fit more waste into the drum.  The drum is
then transported to the filter dryer and emptied.  The filter
dryer contents are heated, the evaporated solvents pass through a
condenser, and the condensed solvent collects in a receiving
                               3-26

-------
vessel.  One filter dryer used in the industry does not have a
condenser but is vented directly to the facility's control
device.1  As with the liquid waste treatment, the fate of the
recovered solvents varies.  Options include onsite distillation
and offsite reclamation.
     Most facilities ship some, if not all, of their liquid and
solid wastes offsite.  Dirty solvent, discarded mix, sludges, and
still bottoms can be (1) sent *offsite for reclamation and
repurchase; (2) sold to a fuel blender; or (3) sold for use in
other industries, such as paint manufacturing.  Solid waste is
usually considered a hazardous waste and must be disposed of as
such.  Often such waste is treated offsite in a hazardous waste
incinerator or cement kiln.
3.2.2  HAP Emission Points
     3.2.2.1  Sources of Emissions and Factors Affecting
Emissions.  Emissions from a magnetic tape facility occur as a
result of solvent storage, mix preparation, coating application
and drying, packaging and labeling, cleaning, onsite waste
handling, and equipment leaks from piping.  Each of these
emission points is discussed in more detail below.
     Emissions from outdoor solvent storage tanks result from
both working losses and breathing losses due to diurnal
temperature changes.  The solvent HAP emissions depend on the
tank size, solvent vapor pressure, number of volume turnovers,
and temperature.
     In the coating mix preparation room,  solvent HAP's are
emitted from the individual mixers and holding tanks, from the
transfer of the coating mix between equipment (piping equipment
leaks), and from intermittent activities such as changing the
filters in the holding tanks.  The emissions from mixing the
coating are intermittent because the coating is.made in batches.
     Emissions vary with the solvent vapor pressure, room
temperature, volume of mix equipment, fill rate, surface area
exposed to the atmosphere, and method of solvent transfer (i.e.,
closed lines versus manual pouring).  Particulate HAP emissions
result from transferring the material from the bag to the
                               3-27

-------
intermediate carrier, such as a tote bin or hopper, and then to
the mixer.  Once the particulates are thoroughly dispersed in the
coating during the mixing process, there are little, if any,
particulate emissions from subsequent steps such as milling or
polishing.  No HAP's are emitted from mills that are permanently
sealed and operate under pressure.
     Emissions from the coating area come from the evaporative
loss of solvent HAP's at the coating head and from the coated
film as it travels, exposed, from the coater to the oven
entrance.  The magnitude of these losses is a function of line
width and speed, coating thickness, volatility of the solvent(s),
temperature, distance between coater and oven, and air turbulence
in the coating area.  Particulate HAP's are not emitted at the
coater.
     In the drying oven, the rate of evaporation of the solvent
HAP's is affected by the temperature, airflow rate and direction,
and the line speed.  The airflow rate is adjusted to keep the VOC
level below the LEL.  Particulate HAP's are not emitted in the
oven.
     Of the total solvent HAP emissions from the mix preparation
room and the coating operation (coater head and drying oven),
approximately 10 percent are emitted from the mix room.80  Of the
total solvent HAP emissions from the coating operation,
approximately 10 percent are emitted from the application/
flashoff area and the remainder from the oven.81  Because the
tape is somewhat delicate and must be dry, ovens are operated to
remove all solvent.  Hence, downstream of the oven, solvent HAP
emissions from the coated tape are negligible.
     The packaging and labeling steps may have solvent HAP
emissions from inks used for printing.  Emissions vary with the
solvent content of the ink and the quantity of ink used.
     Factors affecting solvent HAP emissions from waste handling
include the type and amount of waste, solvent concentration of
the waste, and the temperature and humidity.  The method of
storing  (e.g., sealed drums, wastebaskets) and transferring
 (e.g., closed piping, poured into drums) dirty solvents and
                               3-28

-------
solvent-laden objects such as rags and brushes also influences
evaporation.  For onsite treatment devices--stills, thin film
evaporators, and filter dryers--the frequency with which they are
opened, the types of seals, and the length of treatment time
affect solvent HAP emissions.
     Solvent HAP emissions from cleaning activities are affected
by numerous factors, such as the cleaning methods employed by a
facility, the quantity of solvent used for the particular
cleaning activity, and the frequency and/or time duration of the
cleaning.  For example, in the absence of capture and control
methods, 100 percent of the HAP solvent used for fixed exterior
surface cleaning evaporates.71  Other factors include the method
of transfer (e.g., bucket, closed piping, faucet), degree of
solvent agitation during cleaning, and whether cleaning takes
place in an open or covered container.  The physical
characteristics of the container such as the freeboard ratio
(defined as the distance from the top of the container to the
solvent level divided by the width or length of the container)
affect solvent HAP emissions in removable parts cleaning.  The
temperature and humidity of the room and the vapor pressure of
the solvent influence the degree of evaporation and thus
emissions.  The ventilation rate in the vicinity of the cleaning
activity will also affect solvent HAP emissions.  Because
cleaning is generally not automated, employee work practices can
also affect HAP solvent emissions from cleaning.
     Emissions result when liquids leak from pipe fittings.
Factors affecting the typical leakage rate include the number and
types of fittings, the HAP content  (percent by weight) of the
solvent in contact with the fittings, and the length of time the
fitting is in contact with the solvent.
     3.2.2.2  Total Industry Uncontrolled Emissions.82  Of the
25 plants operating in the United States, 6 have no VOC emission
control device. Four of the uncontrolled plants do not use
solvent HAP's.  The HAP emissions from the two uncontrolled
facilities using solvent HAP's are approximately 16 and
                               3-29

-------
178 megagrams per year (Mg/yr) (18 and 196 tons per year
[tons/yr]),  respectively.
     Of the 11 facilities in this industry using particulate
HAP's, 9 do not use an add-on control device.  The HAP
particulate emission estimates for these nine range from <
0.01 to 0.29 Mg/yr (< 0.01 to 0.32 tons/yr).
3.3  BASELINE EMISSIONS
     The baseline emission level represents the level of control
that is expected without a national emission standard for
hazardous air pollutants (i.e., the existing level of emissions
control).  This level of control was developed from plant-
specific data for each facility in this industry combined with
emission factors.  The baseline is used to evaluate the impacts
of the regulatory alternatives to be selected for analysis.
3.3.1  Existing Emission Limits
     Table 3-5 summarizes the State and local regulations that
apply to VOC emissions from magnetic tape manufacturing plants.
Solvent HAP's are VOC's and, therefore, are regulated
accordingly.  Twenty-eight States (which include 12 operating
plants and 1 planned plant)  limit VOC emissions to 347 grams per
liter (g/L)   (2.9 pounds per gallon [Ib/gal]) of coating applied,
excluding water, and 2 States  (1 plant) limit VOC emissions to
359 g/L (3.0 Ib/gal).  The former was recommended by a Federal
control techniques guideline  (CTG) published in 1977 that
addressed control of volatile organic emissions from existing
stationary sources, specifically surface coating of cans, coils,
paper, fabrics, automobiles, and light-duty trucks (document No.
EPA-450/2-77-008).  For typical coatings used by this industry,
this is equivalent to approximately 83 percent control.  Sources
in California  (five plants)  are limited to coatings with VOC
contents of either 120 g/L  (1.0 Ib/gal) or 264 g/L (2.2 Ib/gal),
depending on the local district regulations.
     For 15 States (in which four plants are located), the
National Ambient Air Quality Standard for ozone is the only
applicable regulation.  As of 1988, new or modified lines in any
State are subject to the new source performance standard  (NSPS)
                               3-30

-------
TABLE 3-5.
       FROM
CURRENT STATE REGULATIONS ON VOC EMISSIONS
THE MAGNETIC TAPE COATING INDUSTRY31
State
Alabama1"
Alaska
Arizona*
Arkansas
California1"
Bay Area
San Diego County
SCAQMDd
Ventura County
Colorado
Connecticut
Delaware
Florida
Georgia*
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland1*
Massachusetts
Michigan
Minnesota1"
Mississippi
Regulation15
kg/<
0.35


0.36
0.12
0.265
0.12
0.265
0.120
0.35
0.35
0.35
0.35
0.35


0.35
0.35



0.35
0.35
0.35
0.35
0.35


Ib/gal
2.9


3.0
1.0
2.2
1.0
2.2
1.0
2.9
2.9
2.9
2.9
2.9


2.9
2.9



2.9
2.9
2.9
2.9
2.9


NAAQS

X








X
X


X
X






X
X
Comments0


Maricopa County requires BACT for all add-
on controls, and facilities emitting
<2.0 tons/yr are exempted.

0.265 for low solvent technology; 0. 12 for
add-on controls; for exemptions for sources
emitting <6.5 kilograms per day (kg/d)
(14.3 pounds per day [lb/d]).


Sources emitting < 10 lb/d are exempted
along with sources located in Kent and
Sussex Counties, and any conforming
landfills.
Sources emitting < 15 lb/d and <3 Ib/hr are
exempted.







Affected facility must not discharge
> 15 percent by weight of VOC compounds
net input into facility; facility is exempted if
coating is <2.9 Ib/gal VOC (less water).
Sources emitting < 100 lb/d are exempted.



Sources within one geographical site emitting
< 100 lb/d or < 2,000 Ib/mo are exempted.


             3-31

-------
TABLE 3-5.   (continued)
State
Missouri
Montana
Nebraska*
Nevada
New Hampshire
New Jersey*
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota*
Tennessee*
Texas*
Utah
Vermont
Virginia*
•
Washington
West Virginia
Wisconsin
Regulation13
kg//
0.35



0.35
0.35
*
0.35
0.35

0.35

0.35
0.35
0.35
0.35

Ib/gal
2.9



' 2.9
2.9

2.9
2.9

2.9

2.9
2.9
2.9
2.9

NAAQS

X
X
X


X


X






X
0.20 kg/kg per month
0.35
0.35
0.35
0.36
0.35

0.35
2.9
2.9
2.9
3.0
2.9

2.9





X

Comments0
Applies only in Clay, Jackson, and Platte
Counties and sources emitting <6.8 kg/d or
2.7 tons/yr are exempted.

New sources or modifications of existing
sources will fall under state air toxics rule if
emitting >2.5 tons/yr or 75 Ibs/day (state air
toxics rule incorporates 40 CFR 61 by
reference).

Sources emitting < 100 tons/yr are
exempted.

Nonmethane hydrocarbons must be
5.0 /tg/nr over 3 hour period.



Coating lines using <3 gallons of coating
per day are exempted.
No discharge of more than 3,000 Ib of
organics in 1 day or more than 450 Ib in
1 hour, 90 percent reduction by incineration,
85 percent reduction by adsorption or any
process of equivalent reliability and
effectiveness.





Facilities emitting < 45 mg/yr are exempt.
Emissions limits are determined regionally.


Sources emitting up to 7 tons/yr, 40 tb/d,
and 8 Ib/h are exempted.
Uncontrolled emissions allowed if VOC from
coater, dryer, and flashoff area <_l& Ib in
any given 24-h period.


          3-32

-------
                                TABLE 3-5.    (continued)
State
Wyoming
Regulation"
kg/I

Ib/gal

NAAQS
X
Comments0

aState regulations are based on emissions of VOC's from the surface coating of paper and fabric for existing
 sources.
"NAAQS = National Ambient Air Quality Standard for hydrocarbons only. These States had no specific
 regulation for VOC's for surface coating of paper and fabric and emissions from magnetic tape coating, or
 surface coating emissions were not listed under the State's air toxics rule. It is assumed that NAAQS will
 apply.
 Puerto Rico will comply with NAAQS unless the source has no ventilation control system.  Exemptions are
 industrial surface coating operations with adequate ventilation built prior to the regulation (in which case,
 RACT will be in force) and where  the coating solvent make-up is water-based and does not exceed 20 percent
 organic solvents by volume.
"South Coast Air Quality Management District.
•"States with existing or planned magnetic tape coating facilities.
                                              3-33

-------
for this industry.  The level of control required by the NSPS is
shown in Table 3-6.83  Five operating coating lines are currently
subject to this rule.  Two lines planned for construction will
also be subject to the NSPS.
     One existing regulation and one pending regulation limiting
VOC emissions from cleaning operations have been identified.
Illinois currently requires covering of vessels during cleaning.
This regulation applies to magnetic tape manufacturing facilities
in nonattainment counties with emissions of greater than 91 Mg/yr
(100 tons/yr).  California has proposed a regulation that would
require that mix tank cleaning be a closed process.  This
regulation would apply to the coating, ink, and adhesive
manufacturing industry.   Although it is not intended to apply to
magnetic tape facilities, this regulation, or a similar
regulation, may apply in the future.  The Bay Area Air Quality
Management District requires that wash sinks maintain a minimum
freeboard ratio or that wash sink emissions be vented to a
control device.
     No State or local regulations governing particulate HAP
emissions from magnetic media were identified.
3.3.2  Determination of Baseline Level
     3.3.2.1  Summary of Plantwide/Industrywide Baseline
Emissions.  Information on HAP use  (solvents and particulates),
emissions, and control technologies was acquired from all of the
25 operating facilities.  The data were obtained from site
visits, written information requests, telephone contacts,
emission test reports, and State preconstruction and operating
permits.  These data were used to estimate the HAP emissions from
each emission point at each facility under their current
operating conditions. The emission points considered in this
analysis include solvent storage tanks, mix preparation
equipment, coating application, drying, packaging and labeling,
onsite waste handling, cleaning, and piping leaks.  Baseline
emissions were estimated based on 1990, 1991, or 1992 data.
Plants that ceased operations as of 1992 were not included in the
calculations.
                               3-34

-------
TABLE 3-6.  SUMMARY OF NSPS REGULATION83
Emission point
Storage tanks
MIX
New mix equipment with
concurrent construction of
control device (other than a
condenser)
All other mix equipment
Regulation
No controls required
ROOM
Use covers and vent emissions
to a 95 percent efficient
control device
Use covers alone or use covers
and vent emissions to a
control device
COATING OPERATION
New coating operation
Modified or reconstructed
operations with existing level
of control a 90 percent
Modified or reconstructed
operations with existing level
of control 2 90 percent that
install a new control device
Control 93 percent of VOC
content of coating applied at
the applicator
Maintain existing control or
93 percent, whichever is lower
Install a 95 percent efficient
control device and maintain
previous control level up to
93 percent
                  3-35

-------
     Total annual solvent HAP emissions per plant from the
25 operating plants in this industry ranged from zero (some
plants in this industry do not use any HAP solvents) to over
1,480 Mg/yr (1,630 tons/yr).   Figure 3-7 summarizes the
distribution of the annual solvent HAP emissions per plant.  The
annual HAP solvent emissions per plant were found to vary widely
and not be simply a function of the number of coating lines but
rather the percentage of the solvent use that is HAP's,  total
annual production, and level of control.  Total industrywide
baseline solvent HAP emissions were estimated as 4,100 Mg/yr
(4,510 tons/yr).
     Particulate HAP's are also used in the coatings at eleven
plants in this industry.1  Of these plants, only two use an add-
on control device specifically for the control of particulate
emissions1.  Of those plants using particulate HAP's, estimates
of baseline particulate emissions ranged from < 0.01 to
0.29 Mg/yr (<0.01 to 0.32 tons/yr) per plant.  Total industrywide
baseline particulate HAP emissions were estimated as 0.55 Mg/yr
(0.60 tons/yr).
     3.3.2.2  Methodology for Estimating Point-Specific Baseline
Emissions.  Baseline emissions were estimated for each of the HAP
emission points within a magnetic tape coating facility.
Emissions were estimated using plant-specific data on HAP use and
controls in combination with emission factors.  Some of the
emission factors were developed specifically for this industry,
while other factors had been developed by EPA from experience
with other industries.  The following paragraphs present the
estimating strategy for each emission point.  Table 3-7
summarizes plantwide and nationwide emissions from each of the
emission sources identified below, and the emission.factors that
were used to determine uncontrolled emissions from each source.
     3.3.2.2.1  Storage of HAP's.  The EPA emission factors
published in AP-42 for volatile organic liquid storage tanks were
used to estimate uncontrolled solvent HAP emissions from bulk
storage tanks at the magnetic tape manufacturing facilities.84'85
Plant-specific data on the number and size of the tanks and on
                               3-36

-------
jo
    3-37

-------
cn

§
H
CO
CO
H
1

u
H
b
H
U
CO
 I
o
cu
r*
 i
m

W




3
•i
i
§
1
09







-C
X
i
i
i
1



I
1
&
4)
IS,
£




Emission factor applied

a
2
g
=
1

ci



^
N

c?
—
a


-:
i
0


| AP-42 equations





p Solvent storage

1
d


o
s
N

§
i
a,


I
i
o


tt «
MK *?ft
si
s
V
§
1
*
|| Mix equipment, coaler,

a



XI
b

??
b
4


a
b
i
o
V
.S a*
& i
0.25% of total annual particulate usa
emitted to air (uncontrolled
0.3 Ib emitted per ton on particulate i
(enclosed transfer)





Particulatea

P
e



•r
1*1

^^
7
a


Os
i
O


Site-specific



•I

Q.
1
.1
•5
£

53.



§


1

s,


o
a
i
o


AP-42 factors
0.2 Ib VOC emitted/ton of waste
0.72 Ib VOC emitted/ton of waste
3.3 Ibs VOC emitted/ton of waste





Ion-site waste handling
- Spills
- Loading
- Condenser vent

£ f f s
«i -, -, a



P 5 - •=
-r — ON o

rt w to
e o o S
a a a s


? ? ?
0 0 O 9
0999

. S1 I
In the absence of control:
-48.5% of solvent used for removabi
parts cleaning ia emitted
-7.1 % of solvent used for tank clean!
ii emitted
-100% of solvent used for fixed extel
surface cleaning is emitted
-Flushing of fixed lines results in
negligible emissions




3
a i 1
• 06 ^ T?
Q* CS O 4*
1 J | |
a» > u " .5
1 | | 1 1
-"C fr * ,

s- f ~
S, 0, !C,



_ 0 v.
oo m •»

-^ *-*
i i i
S9 O
*£* *&


? S 2
i i *
O O O

i
-0.36% of the total annual HAP solv.
used for coating is emitted
-1 .7% of total annual HAP used is
emitted
-Size-dependent





Piping equipment leaks
- Mix preparauon
- Solvent recovery
- Waste Handling
                                        3-38

-------
the solvent type and annual throughput were used.  Solvents used
for both coating and cleaning were included in the analysis.  The
annual solvent HAP volume throughput and the tank volume were
used to determine the annual number of tank turnovers.  Tank
parameters such as height were estimated using standard
manufacturer's dimensions for the known plant-specific volumes.85
Standard default values were used for the remaining values in the
equations, such as temperature change and paint factor.85
     The equations in AP-42 estimate storage tank emissions in
the absence of control.  After the uncontrolled emissions were
calculated using the emission factors, control values were
applied based on the type of control device operated by a
facility.  Five magnetic tape manufacturing facilities control
storage tank emissions using an onsite carbon adsorber.1'85  All
of the remaining solvent HAP storage tanks were assumed to have a
pressure relief valve that controls emissions by 35 percent, the
value used in the NSPS analysis.86
     Plants using exclusively non-HAP solvents were not included
in the analysis described above.  Plants storing solvent HAP's in
underground storage tanks or in 55-gal drums were assumed to have
no air emissions resulting from solvent storage.
     Baseline emissions for plants with bulk'storage tanks range
from <0.01 to 1.1 Mg/yr (<0.01 to 1.2 tons/yr).  The total
industrywide baseline emissions were approximately 2.4 Mg/yr
(2.6 tons/yr).
     All of the particulate HAP's used in this industry (chromium
and cobalt) are delivered in sealed bags or drums.1  Therefore,
it was assumed that there are no air emissions from storing
particulate HAP's.
     3.3.2.2.2  Mix equipment, coater. and oven.  Information on
total annual solvent HAP used (purchased and recycled) to prepare
the coating actually applied at the coater was obtained from each
plant in operation in this industry.  This does not include
solvent used to make coatings that, for whatever reason, were not
applied to a substrate.  The following factors were applied to
the annual HAP solvent usage to estimate the relative
                              3-39

-------
uncontrolled solvent HAP emissions from coating mix preparation,
                        QHf
application, and drying:0'
     Coating mix preparation       10 percent
     Coating application            9 percent
     Coating drying                81 percent
     In addition to the 10 percent uncontrolled emissions from
the mix equipment, allowance was made for piping leaks that occur
in the mix preparation area.  A discussion of equipment leak
emissions is provided in Subsection F.
     Once the uncontrolled emissions from mix preparation,
coating and drying were calculated, plant-specific information on
capture and control technologies was used to determine control
efficiencies.  Based on information from the magnetic coating
industry, the following control efficiencies were used:2'92'88
     Covers on mix equipment                 40 percent
     Mix equipment vent to control device    95 percent
     No enclosure + adsorber/condenser       83 percent
     No enclosure + incinerator              86 percent
     Partial enclosure + adsorber/condenser  87 percent
     Partial enclosure + incinerator         90 percent
     Total enclosure + adsorber/condenser    95 percent
     Total enclosure + incinerator   "        98 percent
     These efficiencies were applied to the uncontrolled
emissions values to determine the baseline solvent HAP emissions
from the three emission points  (mix equipment, coater, and oven)
at each plant.
     Plants not using solvent HAP's were assumed to have no
solvent HAP emissions from mixing, coating, or drying.  Plants
that do not mix their own coatings (i.e., purchase prepared
coatings) were assumed to have no solvent HAP emissions from mix
preparation.
     Baseline solvent HAP emissions resulting from the mix
equipment, coater, and oven ranged from 0.3 to 1,020 Mg/yr
(0.3 to 1,120 tons/yr) for those plants using solvent HAP's in
these operations.  The total industrywide baseline emissions from
                               3-40

-------
these operations (not including leaks from piping) were
2,630 Mg/yr (2,890 tons/yr).
     Particulate HAP emissions occur from the mix preparation
step.  Data on annual use were obtained from each of the
11 facilities using particulate HAP'a.  This value was then
multiplied by 0.25 percent to estimate the uncontrolled
particulate emissions that would result from manually
transferring particulates from the storage container into the mix
vessel.  The emission factor was supplied by one magnetic coating
facility that performed a material balance of the amount of
particulate HAP poured by hand from the storage container
relative to the amount of particulate HAP in the final
coating.89'90  To estimate particulate HAP emissions for those
plants using enclosed transfer mechanisms, annual particulate HAP
usage was multiplied by an AP-42 emission factor of 0.3 pounds
particulate HAP emitted per ton used. See Section 4.3.1 for a
description of the enclosed transfer mechanisms used by the
magnetic tape manufacturing industry.  For the two plants using
add-on control devices to control particulate emissions, the
appropriate emission value was further multiplied by the control
efficiency of.the device.
     Baseline particulate HAP emissions ranged from < 0.01 to
0.29 Mg/yr (< 0.01 to 0.32 tons/yr) for those plants using
particulate HAP's.  Total industrywide baseline particulate HAP
emissions were 0.55 Mg/yr  (0.60 tons/yr).
     3.3.2.2.3  Packaging and labeling.  Only three facilities in
this industry use HAP's in the packaging and labeling steps.  Two
use solvent HAP's in inks that are printed on package labels.1
One plant uses solvent HAP's in tape cartridge manufacture.
Information on solvent HAP use is available from one of the
plants using HAP's in inks and the one using HAP's in tape
cartridge manufacture.  Neither plant controls these emissions.
Therefore, baseline emissions were assumed to be equal to solvent
HAP use.  The total emissions for these plants were 3.4 Mg/yr
(3.7 tons/yr).
                               3-41

-------
     3.3.2.2.4  Onsite waste handling.91"97  Five facilities in
this industry conduct operations to recover solvent HAP's from
dirty cleaning solvents, discarded coatings, and dirty rags,
filters, etc.  Another facility also performs such operations,
but HAP solvents are not involved.  Recovery activities include
using batch stills, thin film evaporators, and filter
dryers.91'97
     To estimate baseline emissions from the batch stills, thin
film evaporators, and filter dryers, data were obtained from each
applicable facility regarding the volume of liquid wastes
delivered to the still and the mode of transfer  (manual versus
piped).  If the solvent HAP's are transferred to the still in
drums and hand poured, previously established emission factors
were used to estimate emissions from these activities (0.2 pound
HAP emitted per ton of HAP waste for spills and 0.72 pound HAP
emitted per ton of HAP waste for manual loading).9^  If closed
piping systems are used to transport the liquid wastes,  it was
assumed there were no fugitive emissions from spills or loading.
     For all facilities, it was assumed that all solvent HAP's
introduced into the still, evaporator, or dryer are evaporated.
The purpose of these operations at-all of the facilities is to
recover the solvent for reuse.  Therefore, none of the stills,
evaporators, or dryers are vented directly to the atmosphere
 (i.e., "uncontrolled").  However, the primary condenser, which is
used to recover solvent and- is part of the waste handling device,
has an atmospheric vent through which emissions could enter the
air.  Some facilities direct emissions from this vent to the
onsite air pollution control device.  If the device has a primary
condenser associated with it, emissions from the condenser vent
were based on plant-specific data.  In the absence of plant-
specific data, emissions for condenser vents were estimated using
an EPA-established emission factor  (3.3 pounds of HAP emissions
per ton of HAP waste treated)."  If the condenser vent is
connected to the onsite carbon adsorber, the value obtained using
the emission factor was adjusted to account for the 95 percent
efficiency of the carbon adsorber.  For facilities venting the
                               3-42

-------
still, evaporator, or dryer directly to the onsite carbon
adsorber (e.g., no primary condenser), a control efficiency of 95
percent was assumed because capture efficiency is 100 percent and
carbon adsorber destruction efficiency is 95 percent.
     Baseline solvent HAP emissions from onsite waste handling
ranged from <0i01 to 220 Mg/yr (<0.01 to 240 tons/yr).
Industrywide baseline solvent HAP emissions from these operations
were approximately 320 Mg/yr  (350 tons/yr).
     3.3.2.2.5  Cleaning.  The following cleaning activities were
identified in Section 3.2.1.9 as emission points in the magnetic
tape industry: (1) flushing of the fixed lines that carry the
coating mix from the mix room to the coater, (2) cleaning of
tanks used in the coating mixing process, (3) cleaning of fixed
exterior surfaces, and  (4) removable parts cleaning.  Data on
solvent used for and emitted from these activities were not
available,  because most plants do not record usage or emissions
for each cleaning activity.  Therefore, protocols were sent to a
sample of the industry in order to quantify solvent usage and
emissions associated with each of the different cleaning
operations.41"70  These protocols instructed plants to estimate
emissions by performing material balance calculations around each
type of cleaning activity.  The results were used to (1) obtain
HAP usage and emissions data for specific plants, and (2) develop
usage and emission factors that could be applied to facilities
for which plant-specific data were not available.  Information on
tank cleaning emissions was also received from one plant that did
not receive a protocol.
     Some facilities use a combination of HAP and non-HAP
solvents for cleaning.  When calculating usage and emission
factors using the protocol data,  all solvents were included so
that data from different facilities could be compared and
averaged.  Because the NESHAP deals only with HAP emissions,
however, only HAP solvents were included in calculating usage and
emission estimates (using the factors) for plants that did not
provide such information for their cleaning activities.   Emission
factors estimated uncontrolled emissions  (emissions in the
                               3-43-

-------
absence of pollution control devices).   Plant-specific data were
then used to determine the level of control for each cleaning
activity at each plant.  Some facilities use non-HAP's
exclusively for cleaning; therefore, they contribute no
emissions.
     A usage factor of 0.18 was calculated for total cleaning
solvent usage.71  Therefore, to calculate total HAP solvent used
for cleaning at a plant, the total quantity of HAP solvent used
for coating should be multiplied by 0.18.
     A usage factor of 0.22 was calculated for removable parts
cleaning.  That is, 22 percent of total cleaning solvent usage is
used for removable parts cleaning.  The emission factor
calculated for removable parts cleaning estimates that
0.485 pounds of solvent are emitted per pound of solvent used for
parts cleaning.71  Annual solvent HAP emissions from removable
parts cleaning ranged from 0 to 300 Mg/yr (0 to 330 tons/yr)
plantwide.  Industrywide parts cleaning emissions were
approximately 470 Mg/yr  (515 tons/yr).82
     A usage factor of 0.72 was calculated for tank cleaning;
i.e., 72 percent of cleaning solvents are used to clean tanks.
The emission factor estimates that 0.071 pounds of solvent are
emitted per pound of solvent used.71  Annual solvent HAP
emissions from tank cleaning ranged from 0 to 55 Mg/yr (0 to
60 tons/yr).  Industrywide tank cleaning emissions were
approximately 140 Mg/yr  (150 tons/yr).82
     A usage factor of 0.015 was calculated for fixed exterior
surface cleaning; 1.5 percent of total cleaning solvent usage is
for this activity.  Most magnetic tape facilities do not control
evaporative losses from cleaning of exterior surfaces.
Therefore, for the purpose of determining baseline emissions, it
was assumed that solvent emissions for exterior surface cleaning
are equal to solvent usage.71   (For the plants that do control
these emissions, plant-specific data on capture and control
efficiency were used to calculate actual emissions.)  Yearly
emissions from exterior surface cleaning ranged from 0 to
30 Mg/yr  (0 to 35 tons/yr) plantwide.  Industrywide emissions
                               3-44

-------
from surface cleaning were approximately 91 Mg/yr
(100 tons/yr),82
     A usage factor of 0.045 was calculated for flushing fixed
lines indicating that 4.5 percent of total cleaning solvent use
is for flushing lines.  Most facilities did not provide enough
data regarding actual emissions from flushing of fixed lines to
calculate an emission factor.  Since line flushing is typically a
closed-system operation, it was assumed that emissions from this
activity are negligible.71
     Baseline total solvent HAP emissions from cleaning
operations ranged from 0 to 355 Mg/yr (0 to 390 tons/yr)
plantwide.  Industrywide emissions were approximately 700 Mg/yr
(770 tons/yr),82
     3.3.2.2.6  Emissions from piping leaks.  For purposes of
this analysis,  the piping used to transfer solvent HAP's and
coatings were divided into three areas:   the mix preparation
area, the solvent recovery area, and the waste handling area.
"Mix preparation" includes piping from bulk storage to the mix
room, within the mix room, and from the mix room to the coater.
The "solvent recovery" area includes piping associated with the
solvent•recovery device and transfer of dirty solvent to and from
the waste solvent storage tanks and to and from recovered solvent
storage.   The "waste handling" area includes piping associated
with the waste handling device  (filter dryer, batch still, thin
film evaporator) and piping associated with spent and recovered
material storage.
     Plants were contacted for detailed information on (1) the
number of piping fittings in the mix preparation and solvent
recovery areas;  (2) the number of hours these areas are in
operation; and  (3)  the HAP concentrations in the piping in these
areas.99"104  Three plants provided information for the mix
preparation area.  Two plants also provided information for the
solvent recovery area.
     The number of fittings in the waste handling area were based
on representative waste handling devices (units) developed by EPA
for the transfer, storage, disposal facilities  (TSDF)
                               3-45

-------
 standards.105   These model  units  are sized as small,  medium,  and
 large based on the total  solvent  throughput.   Using plant-
 specific data  on solvent  throughout  to its waste handling
 device(s),  each device was  classified as  small,  medium,  or large.
 (All  of  the devices in the  magnetic  tape  industry were small  or
 medium).  The  types and number of fittings and operating hours
 identified for the model  unit were used in conjunction with
 plant-specific data on HAP  solvent throughput to calculate piping
 leak  emissions in the waste handling area.
      The EPA-published emission factors for piping leak emissions
 from  pumps,  flanges,  valves,  sampling lines,  and open lines were
 applied  to these plants to  determine total piping leak emissions
 from  the three areas at each plant.106'107 Two emission factors
 specific to the magnetic  coating  industry were then developed for
 the mix  preparation area  and the  solvent  recovery area using  the
 estimate of pipe leakage  relative to their plant-specific total
 annual solvent HAP used to  make applied coatings.   (These factors
 are identified in the following paragraphs.)   To estimate piping
 equipment leak emissions  from the remaining facilities in this
 industry,  these emission  factors  were then applied to the plant-
.specific total annual solvent HAP usage values,  as applicable.
 The solvent recovery emission factor was  not  applied to plants
 that  do  not recover solvents onsite.  Plants  not using solvent
 HAP's were also excluded  from the analysis.   Because the number
 and type of fittings in the waste handling area were based on the
 model units described above-,  no factor was necessary.
      Piping leak emissions  in mix room.  A piping equipment
 leakage  factor of 0.36 percent was calculated for the mix
                  Q*)
 preparation area.0"*  That is,  0.36 percent of the total annual
 solvent  HAP used to make  applied  coatings at  a facility are
 emitted  from leaks in piping in the  mix preparation area.  This
 0.36  percent is assumed to  be in  addition to  the 10 percent of
 total uncontrolled solvent  HAP emissions  attributed to this area
 (see  Section 3.3.2.2B).  It is also  assumed that piping leak
 emissions in the mix preparation  area remain  uncontrolled unless
 mix room air is vented to a control  device.
                                3-46

-------
     For those facilities using solvent HAP's, baseline piping
leak emissions in the mix preparation area ranged plantwide from
0.01 to 44 Mg/yr (0.01 to 49 tons/yr).  Industrywide baseline
equipment leak emissions from piping in this area were
approximately 81 Mg/yr (90 tons/yr).
     Piping emissions from solvent recovery area.  A piping
equipment leak emission factor of 1.7 percent was determined for
the solvent recovery area.82  If a facility recovers solvents
onsite, emissions from leaky piping in the area are assumed to
equal 1.7 percent of the total annual HAP solvent used to make
applied coatings.  These emissions were assumed to be
uncontrolled.
     For the nine plants that use solvent HAP's and recover
solvent onsite, baseline emissions from leaky piping in the
solvent recovery area ranged plantwide from 0.6 to 140 Mg/yr
(0.7 to 155 tons/yr).  Total industry piping leak emissions in
this area were approximately 350 Mg/yr (390 tons/yr.)
     Piping leak emissions from waste handling.  For the five
plants that handle solvent HAP's in onsite waste handling
devices, baseline emissions from leaky piping in the waste
handling area ranged plantwide from 0.1 to 1.5 Mg/yr (0.1 to
1.6 tons/yr).  Total industrywide piping leak emissions in this
area were 4.5 Mg/yr  (5.0 tons/yr).
3.4  REFERENCES FOR CHAPTER 3
                              •
  1.  Memorandum and attachment from Angyal,  S., to Magnetic Tape
      Project File.  November 9, 1992.   Summary of confidential
      and nonconfidential information from U. S. magnetic tape
      manufacturing facilities.
  2.  U. S. Environmental Protection Agency.   New Source
      Performance Standard.  Magnetic Tape Manufacturing
      Industry--Background Information Document for Proposed
      Standards.  Chapter 6.  EPA-450/3-85-029a.  December 1985.
  3.  Telecon.  Beall,  C., MRI, with Brief, H., ITA.  June
      26,  1991.  Information on products.
  4.  Telecon.  Williams,  D., MRI,  with Kosley, A.,  Brown Disk
      Manufacturing.  July 23 and 25, 1991.  Information on plant
      closings.
                              3-47

-------
 5.  Telecon.  Williams,  D.,  MRI, with Lewis, J.,  Certron.
     July 23 and 25, 1991.  Information on plant closings.

 6.  Telecon.  Williams,  D.,  MRI, with Hyde, B., Audiopak.
     July 25, 1991.  Information on plant closings.

 7.  Perry,  R. H.,  and A. A.  Nishimura, Magnetic Tape.  In:
     Kirk-Othmer Encyclopedia of Chemical Technology.  Third
     Edition, Volume 14.

 8.  U.S. Department of Commerce.  Bureau of the Census.
     Computers and Office and Accounting Machines.  1985.

 9.  U.S. Department of Commerce.  Bureau of the Census.
     Computers and Office and Accounting Machines.  1986.

10.  U.S. Department of Commerce.  Bureau of the Census.
     Computers and Office and Accounting Machines.  1987.

11.  U.S. Department of Commerce.  Bureau of the Census.
     Computers and Office and Accounting Machines.  1988.

12.  U.S. Department of Commerce.  Bureau of the Census.
     Computers and Office and Accounting Machines.  1989.

13.  U.S. Department of Commerce.  Bureau of the Census.
     Computers and Office and Accounting Machines.  1990.

14.  Electronic Market Data Book.  1990.

15.  Magnetic Media Information International Newsletter.
     XI(4):9.  May 1990.

16.  Reference 7, p. 734.

17.  Du Pont Mylar* price list.  June 1982.

18.  Telecon.  Beall, C., MRI, with Childress, S., EDM.
     August 15, 1991.  Information on coating operations.

19.  Reference 7, p. 737.

20.  Telecon.  Thorneloe, S., MRI, with Perry, R., Ampex Corp.
     May 19, 1983.  Information on typical coating mix
     formulations.

21.  Telecon.  Thorneloe, S., MRI, with Hudson, J., E. I.
     du Pont de Nemours and Company, Inc.  May 19, 1983.
     Information on typical coating mix formulation.

22.  Reference 7, pp. 740-741.
                              3-48

-------
23.  Letter and attachments from Petersen, J., Pfizer, Inc., to
     Johnson, W., EPA/CPB.  October 31, 1983.  Comments on the
     Draft BID Chapters 3-6.

24.  Reference 7, p. 743.

25.  Telecon.  Thorneloe, S., MRI, with Cannon, T., VIC
     Manufacturing Company.  February 15, 1983.  Information on
     the design of carbon adsorption systems.

26.  Telecon.  Beall, C., MRI, with Chagnon, M., Omniquest.
     July 31, 1991.  General information on magnetic coatings.

27.  Telecon.  Thorneloe, S., MRI, with Waxmonsky, J.,
     Moorehouse Industries, Inc.  August 24, 1983.  Information
     on mix room equipment.

28.  Reference 7, pp. 743-747.

29.  Telecon.  Meyer, J., MRI, with Missbach, P., Netszch-
     Feinmahltechnix GmbH.  August 25, 1983.  Information on mix
     room equipment.

30.  Telecon.  Beall, C., MRI, with Childress, S., EDM.
     July 30, 1991.  Information on coating operations.

31.  Letter from Petersen, A., Ampex Corp., to Meyer, J., MRI.
     March 9, 1983,  Information on Ampex Corp., Opelika, AL,
     facility.

32.  Zink, S.  Coating Processes.  In:  Kirk-Othmer Encyclopedia
     of Chemical Technology.  Third Edition, Volume 6.

33.  Letter and attachments from Harris, T., Tandy Magnetic
     Media, to Wyatt, S., EPA/CPB.  October 28, 1983.  Comments
     on Draft BID Chapters 3-6.

34.  Telecon.  Glanville, J., MRI, with Heinfeld, S., Passavant
     Corp.  July 11, 1983.  Information on drying ovens.

35.  Telecon.  Glanville, J., MRI, with Whitmore, G., Egan
     Leesona Corp.  July 11, 1983.  Information on drying ovens.

36.  Telecon.  Strum, M., and M. Serageldin, EPA/CPB, with
     Fritzemeier, J., Syncom.  June 3, 1991.  Information on
     cleaning.

37.  Telecon.  Strum, M. and M. Serageldin, EPA/CPB, with
     Rainey, C., Carlisle Memory Products.  June 13, 1991.
     Information on cleaning.

38.  Telecon.  Strum, M. and M. Serageldin, EPA/CPB with
     Falco, M., 3M.  June 20, 1991.  Information on cleaning.


                              3-49

-------
39.  Telecon.  Strum, M. and M. Serageldin, EPA/CPB, with
     Gilbert, R., Ampex.  June 11, 1991.  Information on
     cleaning.

40.  Telecon.  Strum, M. and M. Serageldin, EPA/CPB, with
     Danandeh, D., Disksystems.  July 3, 1991.  Information on
     cleaning.

41.  Telecon.  Angyal, S., MRI, with Fritzemeier, J., Syncom.
     November 19, 1991.  Discussion of cleaning data information
     request.

42.  Response to Section 114 information request for cleaning
     operations data.  Received from Fritzemeier, J., Syncom.
     February 1992.

43.  Telecon.  Angyal, S., MRI, with Fritzemeier, J., Syncom.
     February 10, 1992.  Clarification of cleaning data
     information request response.

44.  Telecon.  Angyal, S., MRI, with Anderson, N.,  and B.
     Balagot, Verbatim.  November 20, 1991.  Discussion of
     cleaning data information request.

45.  Letter and attachments from Anderson,  N., Verbatim, to
     Angyal, S., MRI.  February 7, 1992.  Response to
     Section 114 information request for cleaning operations
     data.

46.  Telecon.  Angyal, S., MRI, with Anderson, N.,  Verbatim.
     February 12, 1992.  Clarification of cleaning data
     information request response.

47.  Telecon.  Angyal, S., MRI, with Hughes, R.,  et al., Sony
     Music Operations.  December 11, 1991.   Discussion of
     cleaning data information request.

48.  Letter and attachments from Danas, M., Sony Music
     Operations, to Strum, M.,  EPA/ESD.  February 13, 1992.
     Response to Section 114 information request response for
     cleaning operations data.

49.  Telecon.  Angyal, S., MRI, with Danas, M., Sony Music
     Operations.  February 2, 1992.  Clarification of cleaning
     data information request response.

50.  Letter and attachments from Rainey, C., Carlisle Memory
     Products, to Strum, M., EPA/ESD.  September 10, 1991.
     Information on tank cleaning operations.

51.  Telecon.  Angyal, S., MRI, with Farmer, M.,  and B. Yoemans,
     Sony Magnetic Products.  November 20,  1991.   Discussion of
     cleaning data information request.


                             3-50

-------
52.  Letter and attachments from Farmer, M., Sony Magnetic
     Products, to Strum, VL ,  EPA/ESD.  February 1992.  Response
     to Section 114 information request for cleaning operations
     data.

53.  Telecons.  Angyal, S., MRI, with Farmer, M., Sony Magnetic
     Products.  February 19 and May 4, 1992.  Clarification of
     cleaning data information request response.

54.  Telecon.  Schmidtke, K., MRI, with White, R., and P.
     Roberts, Tandy Magnetic Media.  December 17, 1991.
     Discussion of cleaning data information request.

55.  Letter and attachments from White, R., Tandy Magnetic
     Media, to Strum, M., EPA/ESD.  March 1992.  Response to
     Section 114 information request for cleaning operations
     data.

56.  Telecon.  McManus, S., MRI, with White, R., Tandy Magnetic
     Media.  March 18, 1992.   Clarification of cleaning data
     information request response.

57.  Telecon.  Strum, M., EPA/ESD, with White, R., Tandy
     Magnetic Media.  April 21, 1992.  Clarification of cleaning
     data information request response.

58.  Telecon.  Schmidtke, K., MRI, with Emerich, D., JVC
     Magnetics.  December 5,  1991.  Discussion of cleaning data
     information request.

59.  Letter and attachments from Emerich, D., JVC Magnetics, to
     Strum, M., EPA/ESD.  March 1992.  Response to Section 114
     information request for cleaning operations data.

60.  Telecon.  Strum, M., EPA/ESD, with Emerich, D., JVC
     Magnetics.  March 17, 1992.  Clarification of cleaning data
     information request response.

61.  Telefax and attachments from Emerich, D., JVC Magnetics, to
     Strum, M., EPA/ESD.  April 1, 1992.  Information of
     cleaning operations.

62.  Telecon.  Schmidtke, K., MRI, with Falco, M. , and
     W. Neumann, 3M.  December 4, 1991.  Discussion of cleaning
     data information request.

63.  Letter and attachments from Falco, M., 3M, to Strum, M.,
     EPA/ESD.  February 5, 1992.  Response to Section 114
     information request for cleaning operations data from the
     Camarillo, CA, facility.
                              3-51

-------
64.  Letter and attachments from Falco, M.,  3M, to Strum, M.,
     EPA/ESD.  April 14, 1992.  Response to Section 114
     information request response for cleaning operations data
     from the Hutchinson, MN, facility.

65.  Telecon.  Angyal, S., MRI,  with Falco,  M., 3M.
     February 24,  1992.  Clarification of cleaning data
     information request response for the Camarillo, CA,
     facility.

66.  Telecon.  Strum, M., EPA/ESD,  with Udo, J.,  Fuji Photo
     Film.  April 16, 1992.  Information on cleaning operations.

67.  Telecon.  Angyal, S., MRI,  with Badger, B.,  et al., Ampex
     Recording Media Corporation.  November 22, 1991.
     Discussion of cleaning data information request.

68.  Letter and attachments from Vaughan, B., Ampex Recording
     Media Corporation, to Angyal,  S., MRI.   February 19, 1992.
     Response to Section 114 information request for cleaning
     operations data.

69.  Telecon.  Angyal, S., MRI,  with Vaughan, B.,  Ampex
     Recording Media Corporation.  March 2,  1992.   Clarification
     of cleaning data information request response.

70.  Telefax and attachments from Vaughan, B., Ampex Recording
     Media Corporation, to Angyal,  S., MRI.   March 2, 1992.
     Information on cleaning operations.

71.  Memorandum from Angyal, S., MRI, to Magnetic Tape NESHAP
     project file.  January 6, 1993.  Statistical evaluation of
     cleaning protocol data.

72.  Letter and attachments from Coulter, B., Tandy Magnetics,
     to  Weigold,  J., EPA.  May 8,  1991.  Response to Section
     114 information request.

73.  Letter and attachments from White, R.,  Tandy Electronics,
     to  Weigold,  J., EPA.  May 10, 1991.  Response to Section
     114 information request.

74.  Memorandum and attachments from Adase,  J., MRI, to
     Strum, M., EPA/CPB.  February 26, 1991.  Report on site
     visit to Audiopak, Winchester, VA.

75.  Memorandum and attachments from Beall,  C., MRI, to
     Strum, M., EPA/CPB.  August 20, 1991.  Report on site visit
     to JVC, Tuscaloosa, AL.

76.  Letter and attachments from Falco, M.,  3M, to  Weigold, J.,
     EPA.  March 6, 1991.  Response to Section 114 information
     request.


                              3-52

-------
77.  Memorandum and attachments from Beall, C., MRI, to
     Strum,  M.,  EPA/CPB.  September 19, 1991.  Report on site
     visit to Sony Music, Carrollton, GA.

78.  Letter and attachments from Farmer, M.,  Sony  Magnetic
     Products Inc. of America, to  Weigold, J., EPA.
     June 28, 1991.  Response to Section 114 information
     request.

79.  Memorandum and attachments from Beall, C., MRI, to
     Strum,  M.,  EPA.  November 13, 1991.  Report on site visit
     to Ampex, Opelika, AL.

80.  Memorandum from Beall, C., MRI, to Magnetic Tape Project
     File.  June 22, 1984.  Distribution of Emissions Between
     Mix Preparation Area and the Coating Line.

81.  Memorandum from Beall, C., MRI, to Magnetic Tape Project
     File.  June 22, 1984.  Distribution of Emissions Between
     Coating Application/Flashoff Area and Drying Oven.

82.  Memorandum from Angyal, S., MRI to Magnetic Tape Project
     File.  December 1, 1992.  Calculation of baseline
     emissions.

83.  Code of Federal Regulations.  Standards of performance for
     magnetic tape coating facilities.  Office of Federal
     Register.  40 CFR Part 60, Subpart SSS,  July 1, 1989.

84.  U. S. Environmental Protection Agency.  Compilation of Air
     Pollutant Emission Factors  (AP-42).  Fourth Edition.
     September 1985.  pp. 4.3-1--4.3-35.

85.  Memorandum and attachments from McManus, S., MRI, to the
     project file.  January 11, 1993.  Solvent storage emissions
     from the magnetic tape manufacturing industry.

86.  Reference 2, Chapter 4.

87.  Reference 2, Chapter 3.

88.  The Measurement Solution:  Using a Temporary Total
     Enclosure for Capture Efficiency Testing.  U. S.
     Environmental Protection Agency.  Research Triangle Park,
     NC.  Publication No. EPA-450/4-91-020.  August 1991.
                                                             •
89.  Facsimile from Bryant, M., XDP Magnetics, Incorporated, to
     Beall,  C.,  MRI.  November 13, 1991.  Estimation of facility
     particulate emissions.

90.  Facsimile from Bryant, M., XDP Magnetics, Incorporated, to
     Angyal, S., MRI.  December 2, 1991.  Particulate emission
     estimate.


                              3-53

-------
 91.  Telecon.  Beall,  C.,  MRI, with Falco, M.,  3M.
      November 19,  1991.  Waste handling practices.

 92.  Telecon.  Beall,  C.,  MRI, with Badger, W.,  Ampex.
      November 12,  1991.  Waste handling practices.

 93.  Facsimile from Rainey, C., Carlisle Memory Products, to
      Angyal,  S.,  MRI.   November 8, 1991.  Waste handling
      practices.

 94.  Telecon.  Williams, D.,,MRI,  with Rainey,  C., Carlisle
      Memory Products.   December 2, 1991.  Waste handling
      practices.

 95.  Telecons.  Beall, C., MRI, with Emerich, D., JVC Magnetics.
      November 14 and 20, 1991.  Waste handling practices.

 96.  Telecons.  Beall, C., MRI; with Farmer, M.,  Sony Magnetics.
      November 14 and December 3, 1991.  Waste handling
      practices.

 97.  Telecon.  Beall,  C.,  MRI, with Danas, M.,  Sony Music.
      November 21,  1991.  Waste handling practices.

 98.  Reference 84,  pp. 4.7-1--4.7-8.

 99.  Letter and attachments from Rainey, C., Carlisle Memory
      Products, to Angyal,  S., MRI.  November 11,  1991.
      Information on process piping.

100.  Telecon*.  Angyal, S., MRI, with Rainey, C.,  Carlisle Memory
      Products.  November 19, 1991.  Information on process
      piping.

101.  Facsimile from Neumann, W., 3M, to Williams, D., MRI.
      October 1,  1991.   Information on process piping.

102'.  Telecons.  Angyal, S., MRI, with Neumann,  W., 3M.
      November 21 and December 18,  1991.  Information on process
      piping.

103.  Telecon.  Beall,  C.,  MRI, with Farmer, M.,  Sony Magnetics.
      October 15,  1991.  Information on process piping.

104.  Telecon.  Angyal, S., MRI, with Farmer, M.,  Sony Magnetics.
      November 20,  1991.  Information on process piping.

105.  Memorandum and attachments from Zerbonia,  R., and York, S.,
      RTI, to Colyer, R., EPA/CPB.   September 25,  1987.  Draft
      Model Unit Parameters-Post Proposed Analysis.   •

106.  Memorandum from Hausle, K. J., and D. J. Whatt, Radian
      Corporation,  to Markwordt, D., EPA/CPB.  February 28, 1992.
      Final cost impacts analysis for HON equipment leaks.
                               3-54

-------
107.  Memorandum from Whitt, D., Radian Corporation, to
      Markwordt,  D.,  EPA/CPB.  June 5, 1991.  Impacts from the
      control of VHAP emissions from equipment leaks in non-SOCMI
      process units for HON.
                             3-55

-------
                 4.0   EMISSION CONTROL TECHNIQUES
   *       *

4.1  •INTRODUCTION
     Emissions from magnetic tape manufacturing plants result
from solvent storage, coating mix preparation, coating
application, drying of the coating in the oven, packaging and
labeling, cleaning, piping leaks, and waste handling.  A complete
air pollution control system for solvent-hazardous air pollutant
(HAP) emissions from a magnetic tape manufacturing plant consists
of a capture or containment system and an emission control
device.  Several magnetic coating facilities also use and emit
particulate HAP's such as chromium compounds, which are emitted
from coating mix preparation.  This chapter describes the
technology available for capture and control of emissions from
all of the sources mentioned above and the expected levels of
control achievable.
4.2  SOLVENT HAP CONTROL SYSTEMS
     Table 4-1 presents control devices currently used to control
the coating operation solvent emissions.  The add-on technologies
used to control solvent HAP emissions are absorption, adsorption,
condensation, and incineration.  The theory and principles of
these control systems are discussed briefly.  The design and
operation of the systems are presented with emphasis on factors
that- affect their use by the magnetic tape manufacturing
industry.  Absorption systems are not discussed even though one
is operated by a magnetic tape plant.  This older system
reportedly achieves a lower removal efficiency (85 to 95 percent)
than do other types of control devices.1"^
                               4-1

-------
     TABLE 4-1.  CONTROL DEVICES USED ON COATING OPERATIONSa
Control device
Carbon adsorber
-- Fixed-bed
-- Fluidized-bed
Rotor concentrator/
"incinerator
Condenser system
Incinerator
Absorber
None
No. of control
devices"
11
2
1
4
9
1
4
Percentage0
34
6
3
13
28
3
13
^Includes plants that do not use solvent HAP's.
bSome facilities have more than one device.
Percentage is  equal  to the number  of a certain  type  of device
  divided by the total number of devices.
                               4-2

-------
4.2.1  Adsorption
     Carbon adsorption has been used for many decades by various
industries to recover a variety of organics from solvent-laden
air (SLA) streams.4  This technology reduces volatile organic
compound (VOC) emissions by adsorbing the organic compounds from
the SLA onto the activated carbon bed.  The organics are
subsequently desorbed and recovered.  The exhausts from more than
one coating operation are commonly vented to the same carbon
adsorber.  Two types of carbon adsorption systems are commonly
used in the magnetic tape manufacturing industry, fixed bed and
fluidized bed.  Adsorption-based technologies used in other
industries include disposable canisters and continuous rotor
concentrators; a continuous rotor concentrator also iri use in the
magnetic tape industry.
     4.2.1.1  Fixed-Bed Carbon Adsorbers.  For most of the time
since carbon has been used as a commercial adsorbent, it has been
available only in a fixed-bed process.  The typical thickness of
a carbon bed within a vertical or horizontal metal vessel is 15
to 76 centimeters (6 to 30 inches).5  The SLA is fed beneath the
bed, if vertically aligned, and the organics are adsorbed as the
SLA passes up through the bed.  The SLA can also be fed with a
downward flow to minimize bed lifting.6'7  Most fixed-bed
adsorbers have multiple beds to allow simultaneous adsorption and
desorption and, thus, continuous operation.  In this industry,
there are usually two to six beds.8  Figure 4-1 is a schematic of
a two-unit fixed-bed adsorber.
     When the VOC concentration in the air discharge from a bed
starts to increase,  or at a preset time interval, the inlet SLA
is routed to a different carbon bed, and the nearly saturated bed
is regenerated.  Regeneration is usually accomplished using low-
pressure steam, although heated nitrogen is sometimes used.3'9
The steam heats the bed to desorb the solvents and acts as a
nonflammable carrier gas.  Typical steam requirements range from
4 to 9 kilograms (kg) of steam per kg of recovered solvent
(4 to 9 pounds (Ib)  of steam per Ib of recovered solvent)  or
0.13 to 0.44 Ib of steam per Ib of carbon.8'10'11  After
                               4-3

-------
                                                         0)
                                                        ,Q
                                                         ^
                                                         o
                                                         CD
                                                        •O
                                                         us

                                                        T3
                                                         (U
                                                        T)
                                                         0)
                                                         X
                                                         C
                                                         3
                                                         i

                                                         i
                                                        4J
                                                        0
                                                        XJ
                                                        0)
                                                        ^
                                                        CQ

                                                        U
                                                        •H
                                                        4J
                                                        01
                                                        ^
                                                        cn
                                                        nJ
                                                        •H
                                                        Q
                                                        ^

                                                        0)
                                                        Cn
                                                       -H
4-4

-------
regeneration, the carbon bed is dried and cooled to improve the
ability of the carbon to adsorb organic compounds.  The mixture
of steam and organic vapors exhausts from the adsorber and is
condensed in a heat exchanger, and the condensate is routed to a
decanter (see Figure 4-1).   In the decanter, the solvent floats
on the organic-soluble water layer.  Both water and organics are
drawn off and sent to storage or further treatment.
     The parameters considered in the design of a fixed-bed
adsorption system are:
      1.  Type of solvent(s);
      2.  SLA inlet concentration;
      3.  SLA flow rate;
      4.  Temperature of the inlet SLA;
     -5.  Relative humidity of the inlet SLA;
      6.  Type and amount of carbon;
      7.  Superficial bed velocity;
      8.  Bed pressure drop;
      9.  Cycle time;
     10.  Degree of regeneration of the carbon bed;
     11.  Pressure and temperature of steam; and
     12.  Condenser water outlet temperature.
The first five parameters are characteristics of the production
process.  The next three are design parameters for the adsorber.
The remaining parameters are operating variables that affect the
performance of the adsorber.  Table 4-2 presents process
parameters representative of several magnetic tape plants
presently controlled by carbon adsorbers using steam or nitrogen
desorption; there appear to be no significant operating
differences between the two systems.
     The SLA discharge from the oven must be cooled to below
approximately 38°C (100°F)  to optimize adsorption.  A minimum of
20 to 40 percent relative humidity should be achieved, especially
if ketones are to be adsorbed, because heat dissipated by
evaporation of water helps prevent heat buildup and subsequent
bed fires.12  Filtration equipment may also be required if there
is particulate matter in the dryer exhaust.  Particulate matter
                               4-5

-------
    TABLE 4-2.  PROCESS PARAMETERS FOR MAGNETIC TAPI
           CONTROLLED BY FIXED-BED CARBON ADSORBERS*
PLANTS
Parameters
SLA flow rate
SLA inlet concen-
tration
SLA temperature
SLA relative
humidity
Range3-
3.78 to 48.14 cubic
meters per second
(m3/sec)
(8,000 to 102,000
actual cubic feet
per minute [acfm] )
100 to 9,000 ppm
24° to 43°C
(75° to 110°F)
25 to 50 percent
Typical range13
4.7 to 9.4 m3/sec
(10,000 to 20,000
acfm)
2,000 to 9,000 ppm
27° to 38°C (80° to
100°F)
40 to 50 percent
These ranges include all data points.
    majority of the data lie in this range.
                              4-6

-------
will coat the carbon and plug the voids between carbon particles,
decreasing adsorptivity and increasing pressure drop.
     The effectiveness of carbon adsorption systems can be
affected by fouling, corrosion, and bed fires.  Carbon beds foul
when the carbon cannot be regenerated with normal steam
temperature and pressure.  Fouling reduces adsorption and
requires early replacement of the carbon.  Carbon beds can be
fouled by high boiling compounds that may polymerize or oxidize
on the carbon particles.1-^  In the magnetic tape industry, carbon
life has been reported to vary from 6 months to over 5 years,
with 6 to 24 months being the most common.8'10
     Corrosion can be a problem in fixed-bed carbon adsorption
systems used to recover ketones because of the formation of
acidic compounds in the wet steam.  Ketones are commonly used at
magnetic tape coating plants; corrosion can be avoided by the use
of corrosion-resistant materials.  Additional problems with
ketones include plugging of the carbon bed by oxidation of the
ketones.  The oxidation reaction products of cyclohexanone are
solids, which can rapidly plug the carbon adsorption system and
cause corresponding loss in activity.6
     Adsorption is an exothermic phenomenon; typically, the
amount of heat generated is 465 to 700 kilojoules (kJ) per kg
(200 to 300 British thermal units [Btu] per pound)  of solvent
adsorbed.14  If sufficient air is not present to carry this heat
off, the bed may overheat, resulting in poor adsorption and, in
extreme cases, bed fires.15  Fires are predominantly associated
with adsorption of ketones and are more likely to occur after
addition of fresh carbon.16
     Solvents recovered by an adsorber may be purified by
distillation and reused in the mix formulation.  Typical
purification systems consist of a decanter and several
distillation columns.  Caustic drying systems may also be used to
remove water from the solvent.  The materials of construction of
the distillation system are a function of the types of solvents
to be recovered.  If ketones are present, expensive corrosion-
resistant materials are required.  The complexity and the
                               4-7

-------
recovery efficiency of the distillation system will vary with the
amount of water in the recovered solvent, the number of solvent
components, the desired purity of the recovered solvents, and
plant preference.  A plant may ship the solvents recovered by the
adsorber offsite for purification.  Alternatively, a plant may
ship the recovered solvents offsite for use as a fuel.
     Long-term average VOC control efficiencies of at least
95 percent are readily achievable with well-designed and well-
operated fixed-bed adsorbers.1'  Several currently operating
carbon adsorbers in this industry are routinely achieving control
efficiencies of 97 to 99 percent.8'17"19  In contrast, one plant
claims to achieve only 94 percent control.2^  The keys to
achieving high levels of control efficiency appear to be:
(1) designing for the specific component in the solvent that will
"break through" the carbon bed first; (2) taking the bed
offstream as soon as solvent begins to break through the bed;
(3) thoroughly desorbing the bed; and (4) replacing the bed well
before fouling reduces adsorptive capacity to the point where the
regeneration cycle must be compromised.21  Most of these factors
are operating parameters and, as such, are influenced by economic
decisions that the plant operator must make regarding operation.
The cost of more frequent and longer desorption of the bed and
more frequent carbon replacement is at least partially offset by
the value of the extra solvent that is recovered.
     There are nine fixed-bed carbon adsorption units operating
in the magnetic tape manufacturing industry.  Most were built
during the last 10 to 12 years.  Two, tested by EPA, are
described below.
     Plant A installed its system in 1980 to recover toluene and
tetrahydrofuran  (THF) from a tape coating operation.22  The
system consists of three beds and a purification section.  The
design flow for the adsorption unit is 6.3 cubic meters per
second (m3/s)  (13,400 standard cubic feet per minute  [scfm] ).
Solvent concentrations in the inlet stream range from 50 to over
2,400 parts per million by volume (ppmv), depending on the number
of coating lines operating, line speed,  and coating thickness.
                               4-8

-------
Outlet solvent concentrations vary from near 0 to 100 ppmv,
depending upon adsorption-desorption cycle timing.  The EPA test
measured an average VOC removal efficiency of 99.9 percent, using
a gas chromatograph with flame-ionization detector.
     Plant B installed its adsorber in 1975 to recover solvent
from two coating lines.  The adsorber was redesigned in 1978.22
The solvent used at the plant is a mixture of toluene, THF,
                       •       »
methyl ethyl ketone (MEK), methyl isobutyl ketone (MIBK), and
cyclohexanone.  The recovery section consists of three pairs of
fixed carbon beds and a purification system.  Only two pairs of
beds are operated at any given time while the third pair is
either down for maintenance or on standby.  The design flow rate
for the adsorber units is 9.7 m3/s at 27°C  (20,600 acfm at 80°F)
with inlet solvent concentrations ranging from 2,000 to
5,000 ppmv depending on the line speed, number of lines in
operation, and the type of magnetic tape being produced.  Outlet
concentrations from the beds varied from 35 to 350 ppmv.  The
efficiencies of the beds ranged from 91 to 98 percent.  The
efficiency varied with the age of the carbon.  Solvent vapor
concentrations were measured using the same procedure used for
Plant A.
     4.2.1.2  Fluidized-Bed Carbon Adsorbers.  In fluidized-bed
systems, adsorption and desorption are carried out continuously
in either the same vessel or side-by-side vessels.  Figure 4-2
presents a flow diagram of a fluidized-bed carbon adsorber in
which adsorption and desorption sections are in a single vessel.
The system consists of a multistage, countercurrent,
fluidized-bed adsorption section; a pressure-sealing section; and
a moving bed desorption section.  Heated nitrogen gas is used as
a carrier to remove the solvent vapors from the desorption
section.  The pressure-sealing section prevents-air from entering
the mixture of solvent and nitrogen vapors.  The regenerated
carbon is carried by air from the bottom to the top of the
column.
     The SLA is introduced into the bottom of the adsorption
section of the column and passes upward countercurrent to the
                               4-9

-------
                                               CLEAN AIR
ADSORPTION
 SECTION
                                                             PRESSURE-SEALING
                                                                 SECTION
ADSORPTION
 SECTION
(SHELL-ANO-
TU8EHEAT
EXCHANGER)
                                                              MIXTURE OF SOLVENT    JL
NITROGEN
RECYCLE
BLOWER
         AIRLIFT
         BLOWER
        AIR LIFT NOZZLE
      FOR CARBON RECYCLE
•» CARBON FLOW
         Figure  4-2.    Fluidized-bed carbon  adsorption  system.
                                         4-10

-------
flow of carbon particles.  Adsorption occurs on each tray as the
carbon is fluidized by the SLA.  The carbon falls down the column
through a system of overflow weirs.  Below the last tray, the
carbon falls to the desorption section where indirect heating
desorbs the organic compounds from the carbon.  Hot nitrogen gas
passes through the bed countercurrent to the carbon flow and
carries the desorbed organic compounds to the condenser.  The
desorption temperature is normally around 121°C (250°F) but can
be raised to 260°C (500°F) to remove buildup of high-boiling
materials.  The desorption section is maintained continuously at
the temperature required to volatilize the adsorbed compounds.23
The solvent and nitrogen mixture is directed to a condenser where
the solvent can be recovered for reuse.  The nitrogen is sent
through the "secondary adsorber" (top layer of carbon in the
desorption section),  which removes residual solvent from the
nitrogen, and is then recycled.
     The microspherical particles of carbon (0.7 mm in diameter)
used in a fluidized-bed are formed by spray-drying molten
petroleum pitch.  The carbon particles are easily fluidized and
have strong attrition resistance. ^  The adsorptive properties of
carbon made this way are similar to those of other activated
carbons.24
     The parameters considered in design of a fluidized-bed
carbon adsorber system are:
      1.  Type of solvent(s);
      2.  SLA inlet concentration;
      3.  SLA flow rate;
      4.  Temperature of the inlet SLA;
      5.  Relative humidity of the inlet SLA;
      6.  Superficial bed velocity;
      7.  Bed pressure drop;
      8.  Rate of carbon flow;
      9.  Degree of regeneration of the carbon (bed);  and
     10.  Condenser water outlet temperature.
The first five parameters are characteristics of the production
process.  The next two are design parameters for the adsorber.
                               4-11

-------
The next three are operating parameters.  The rate of carbon flow
is set by the operator to achieve desired control efficiency.
Just as with the fixed-bed, the dryer exhaust gas (the SLA) must
be cooled before it reaches the adsorber in order to optimize the
carbon's performance.  Pressure drop per stage normally ranges
from 1 to 2 kilopascals (kPa) (4 to 8 in. water), with six to
eight stages required, depending on the application.  The
pressure drop across the entire bed is 6 to 16 kPa  (24 to 64 in.
water).  The gas velocity through the adsorption section is as
high as 1 m/sec (200 ft/min), which is two to four times that
used in fixed-bed adsorbers.  For a given flow rate, a high gas
velocity reduces the cross-sectional area of the bed. ^
     Two primary concerns with operation of fluidized-bed
adsorbers are fouling of the carbon and attrition of the
activated carbon particles.  The same factors that affect fouling
of carbon in fixed-bed adsorbers also affect the carbon used in
fluidized-bed adsorbers.  Suppliers claim that a recent advance
in fluidized bed technology, the addition of a side stream
regenerator, has successfully addressed the problem of fouling.^
Attrition is the wear and loss of particles in the course of
repeated use as they are transported through the system.
     Corrosion is generally not a problem in fluidized-bed
adsorbers.  Because stripping is by nitrogen rather than steam,
the water content of the recovered solvent is low, typically
5 percent or less.2^  The only water present in the recovered
solvent is that humidity which was adsorbed from the SLA.  Thus,
generally, the carbon adsorber need not be made from expensive
corrosion-resistant materials.
     Bed fires usually do not occur in fluidized-bed adsorbers
because the relatively high superficial velocities and the
intimate contact between the SLA and activated carbon minimize
the possibility of hot spot formation.  They are also not a
problem when the desorption section is inerted, e.g., when N2 is
used as the carrier gas.  However, hot spots can form, depending
on the solvents adsorbed, if the bed is shut down before being
                               4-12

-------
completely stripped.  Shutdowns resulting from mechanical
problems could create conditions leading to a bed fire.6
     A distillation system may not be required for a
fluidized-bed adsorption system because of the low water content
of the recovered solvent (less than 5 percent water by weight).26
Cleanup can be as simple as drying by the addition of caustic
soda.27  Of the two facilities in this industry using a
fluidized-bed system, one facility dries with caustic soda, and
one distills the recovered solvent.
     There are two fluidized carbon bed systems in this industry,
and only inlet and outlet monitoring data are available.
Consequently, the data regarding long-term control efficiency are
limited.  One facility averages a control efficiency of about
98 percent.28  The other facility achieves control efficiencies
ranging from 91 to 99, varying primarily with carbon age.8
     Suppliers of the fluidized bed system note that the product
has recently undergone considerable evolution, and many
variations have been developed to meet specific needs.2^
     4.2.1.3 "Disposable" Carbon Adsorbers.  Although no magnetic
tape line is known to use disposable adsorbers, they have been
used to control emissions from solvent storage tanks and reactors
in other industries.2^  In this system, a prefabricated canister
containing activated carbon is connected to the emission source
vent.  The principle of operation is the same as that of a. fixed-
bed carbon adsorber except that there is no in-situ regeneration
of spent carbon.  Rather, a new canister is installed and the old
canister and its contents are shipped from the site for disposal.
The actual useful life depends on the size of the canister and
the type and amount of vapors to which the carbon is exposed.29
     As with other fixed-bed systems, bed overheating can be a
problem if these systems are not properly operated.  Overheating
can be circumvented by keeping the carbon damp.30  Keeping the
carbon damp, however, reduces the effectiveness of carbon
adsorption; consequently, a larger unit must be provided than
would be indicated from the design calculations involving
adsorption rates on dry carbon.^
                               4-13

-------
     Disposable canisters are used for flows generally less than
0.05 m/sec (100 acfm)  with low organic loading.  They can be used
to control emissions from solvent storage tanks and mix room
equipment, which have inherently low flow rates and solvent
concentrations.
     4.2.1.4  Continuous rotor concentrator.  At least one
facility uses a continuous rotor concentrator technology to
control emissions from magnetic tape manufacturing operations.
The emission points are the exhaust from the main control device
and fugitive emissions from the mix room, oven room, and coating
room.  These emission points contain relatively dilute solvent
concentrations.31
     Rotor concentrators have been installed to control VOC's by
a variety of other industrial coating operations including auto
parts painting, film coating, rotogravure printing, and adhesive
coating.32'33  These systems concentrate a high-volume, dilute
(less than 300 ppm) SLA stream into a low-volume stream which may
be subsequently catalytically or thermally oxidized or recovered
by condensation.  Operated independently, a rotor concentrator by
itself is not a control system; it must be followed by a device
which destroys or recovers the solvent in the concentrated SLA
stream.  The magnetic tape facility operating a rotor
concentrator destroys the concentrated emission stream with a
catalytic incinerator.31
     A continuous rotor concentration system consists of either a
single or multiple rotary cylindrical wheels containing
adsorbent.  The adsorbent is coated on a paper or fiberglass
substrate in a honeycomb configuration contained within the
cylinder.  The honeycomb shape provides a large gas-solvent
contacting area with minimal pressure drop as the gas flows
though.32"38  The systems presently available in the United
States use activated carbon or hydrophobic zeolite as an
adsorbent.   (The rotor concentrator in use in the magnetic tape
industry uses zeolite.)31  The carbon systems were developed
first and are in greater use.  Hydrophobic zeolite technology is
still considered an emerging adsorption technology.37
                               4-14

-------
     Regardless of the adsorbent, the wheel rotates so that a
portion of it is exposed sequentially to adsorption, desorption,
and the cooling zone.  Incoming SLA passes through the adsorption
zone and is adsorbed onto the adsorbent in the honeycomb.
Purified air is exhausted.  As the wheel slowly rotates
(typically 1 to 3 revolutions per hour), the adsorbed species
passes through the desorption zone, where a relatively small
quantity of hot air is passed through to desorb the solvent.  The
net result is a small-volume, concentrated stream that exits the
rotor.  The solvent in the concentrated stream could be
incinerated or recovered.  The wheel rotates through the cooling
zone where the sorbent is cooled to ensure high-efficiency
adsorption when it again traverses through the adsorption
zone.32'38
     Bed fires are not a problem with a well designed carbon or
hydrophobic zeolite system.32'3S"37  in addition, hydrophobic
zeolite is inflammable and may be subjected to desorption
temperatures of 180°C (360°F) and higher.37  High humidity is not
a problem for hydrophobic zeolite rotors because the hydrophobic
zeolite, by its synthetic design, does not adsorb water up to
relative humidities of 95 percent.35'37
4.2.2  Condensers
     Condensers recover VOC emissions by cooling the SLA below
the dew point of the solvent or solvent mixture and collecting
the droplets of solvents.  The temperature reduction necessary to
condense the solvent vapor depends on the vapor pressure of the
solvents in the gas stream.  Two types of commercially available
condensation systems have been used to recover VOC from drying
ovens at magnetic tape plants.  These systems differ (1) in the
design and operation of the drying oven (i.e., use of nitrogen or
air in the oven) and (2) in the method of cooling the SLA (i.e.,
liquid nitrogen or refrigeration).
     4.2.2.1  Condensation System Using Nitrogen Atmosphere.
Figure 4-3 presents a flow diagram of a condensation system that
uses a nitrogen-blanketed drying oven and a nitrogen-cooled
condenser.39  The inerting curtains shown in Figure 4-3 minimize
                               4-15

-------
§
                                       M
                                  •'M
                                  OT
                                  i
                                               CD
                                              4J
                                              rt
                                              00
                                              a
                                              0)
                                              T3
                                              c
                                              o
                                              o
                                              u
                                              •H
                                              JJ
                                              nJ

                                              a;
                                              J3
                                              o
                                                     0)
                                                     ^
                                                     3
                                                     01
                                                     •H
4-16

-------
both airflow into the oven and VOC flow from the oven.  Frame
collection hoods may also be located near the ovens and curtains
to capture any gases escaping from these areas.
     Nitrogen is used in the drying oven to permit operation with
high solvent vapor concentrations without danger of explosion.
The nitrogen recycled through the oven is monitored and operated
to maintain solvent vapor concentrations of 10 to 30 percent, by
volume.4^  The higher the solvent concentration, the smaller the
auxiliary equipment required and the less makeup nitrogen
required.  This allows economical solvent recovery.
     Solvents are recovered by sending a bleed stream of
approximately 1 percent of the recycle flow through a shell-and-
tube condenser.41  The liquid nitrogen is on the tube side, and
the solvent-laden nitrogen passes over the outside of the tube
surfaces.  Solvent vapors condense on the tubes and drain to a
collection tank.42  The nitrogen gas is directed to the oven and
inerting curtains.  To avoid solvent condensation in the oven and
to maintain the product cure rate, the temperature in the oven
must be substantially above the dew point of the solvent vapor.
     This nitrogen-blanketed system results in considerably less
water being present in the recovered solvent relative to steam
desorption of carbon beds.  Therefore, the solvent purification
steps following recovery may be simpler.  Two magnetic tape
plants use this type of condenser.  Both use a molecular sieve
for water removal.^
     The parameters considered in the design and operation of an
inert gas condensation system are:
     1.  Type of solvent(s);
     2.  Temperature of the solvent -laden nitrogen bleed stream;
     3.  Solvent-laden nitrogen flow rate; and
     4.  Concentration of VOC in nitrogen.
The first two parameters are characteristics of the production
process.  The remaining parameters are design characteristics of
the condenser.
     The primary operating problem anticipated with this
condenser design is the possibility of air leaking into the oven
                               4-17

-------
and loss of an inert atmosphere.  If air leakage occurred, an
explosion hazard might exist; therefore, it is necessary to
maintain the system at a slight positive pressure.  Because of
the inert atmosphere and low water content, corrosion is not a
problem; therefore, no special materials of construction are
required.
     According to manufacturers of the nitrogen condensation
system, advantages of this system over a carbon adsorption system
include minimal waste generation, elimination of the need for
wastewater treatment, and avoidance of solid waste disposal.
This consideration along with compliance with the New Source
Performance Standard (NSPS),  reliability, capability for
production expansion, and minimization of waste solvent handling
led one magnetic tape manufacturer to select this control
technique for recovering solvent emissions from the drying
oven.39
     There are no exhausts from the nitrogen condensation system.
Therefore, the overall efficiency of controlling coating
application and drying oven emissions with this type of system
can be determined either by a total enclosure or by measuring the
solvent used at the coater and the quantity of solvent
recovered.43  Based on a material balance performed by a magnetic
tape facility, its nitrogen condensation system operates
consistently at an overall efficiency above the 93 percent
required by the NSPS and up to 97 percent.  That is, 93
to 97 percent of the solvent applied at the coater is
recovered.8'39
     4.2.2.2  Condensation System Using an Air Atmosphere.  Two
magnetic tape plants use a condenser with an air atmosphere; one
uses the system to recover cyclohexanone from the SLA.8
Figure 4-4 presents the flow diagram in which SLA is drawn from
the drying oven through a counterflow heat exchanger.  There, the
SLA is cooled to reduce the heat load on the control device, a
refrigerated condenser.  The condensate mixture of solvent and
water from the refrigerated condenser is held in intermediate
storage tanks to await further processing.  The cool solvent-free
                               4-ia

-------
                          A
                           s
 0)
                                         03
                                         JJ
                                         ti
                                         (fi

                                         c
                                         td


                                         B1
                                        -H
                                         CD
                                         3

                                         S
                                         (U
                                        -U
                                         00
                                           
-------
discharge air from the condenser is then circulated back through
the heat exchanger where it is preheated before being returned to
the oven.  Drying ovens used with this type of system must be
relatively tight, i.e., have minimum air leakage and, for safety,
be equipped with vapor concentration monitors.  Typically, ovens
of this type are designed to operate up to 40 or 50 percent of
the lower explosive limit (LED, although the one magnetic tape
plant using this system maintains the oven solvent concentration
below 28 percent of the LEL and has allowed it to fall as low as
8 percent.44'45
      Small amounts of water in the condensate (from the water
vapor in the oven exhaust)  can be removed by caustic drying or by
distillation, depending on the desired purity.44  One of the
plants in this industry that uses this type of condensation
system sends the solvent offsite for reclamation.45  The other
reclaims the solvent onsite.8
     The factors important in the design and operation of a
condenser using a counterflow heat exchanger are:
     1.  Type of solvent(s);
     2.  SLA flow rate;
     3.  Temperature of the SLA at the heat exchanger inlet;
     4.  SLA concentration in the oven exhaust;
     5.  Temperature of the refrigerated air returned to the heat
exchanger; and
     6.  Operating temperature of the refrigeration coil.
The first four parameters are characteristics of the production
process.  The remaining parameters are operating variables that
may affect the performance of the condenser.
     This condensation system requires careful control of the
evaporators when high water vapor concentrations are present in
the SLA to prevent freezing.  Corrosion problems are not expected
for this system if the water content of the recovered solvent is
less than 5 percent and mild steel is an acceptable construction
material.
     As with the nitrogen condensation system, the efficiency of
this type of condensation system can be measured either by a
                               4-20

-------
total enclosure or by comparing the quantity of solvent applied
at the coater with the quantity of solvent recovered.  This is
because there is no exhaust stack  (the air is recirculated);
hence, all losses are fugitive emissions.  Based on a material
balance of solvent applied at the coater relative to the amount
of solvent recovered (less water), one of the air condensation
system used in this industry achieves control efficiencies
ranging between 91 and 96 percent. ^
4.2.3  Incinerators
     Incineration is the combustion of organic compounds by
exposure to high temperatures in the presence of air within a
combustion chamber.  Carbon dioxide and water are the products of
complete combustion.  Incinerators are used to control VOC
emissions at nine magnetic tape plants (see Table 4-1).^
Incinerators are typically the device of choice if (1) recovery
of solvents is not economically feasible or practical, such as at
small lines and research lines that may use a variety of solvent
mixtures; or (2) a higher level of control is desired.46
Incinerators used in this industry may be of thermal or catalytic
design and may use primary or secondary heat recovery to reduce
energy consumption.
     4.2.3.1  Thermal.   Thermal incinerators are usually
refractory-lined combustion chambers with a burner located at one
end.  In these units, part of the SLA. is passed through the
burner along with an auxiliary fuel; the rest of the SLA is
bypassed to the combustion chamber.  The combustion gases exiting
the burner blend with the bypassed SLA, and the VOC's in the SLA
are combusted.   With most solvents, complete destruction is
obtained in 0.75 seconds at temperatures of 870°C (1600°F).47/48
     Factors important in incinerator design and operation
include:
     1.  Type and concentration of VOC;
     2.  SLA flow rate;
     3.  SLA temperature at incinerator inlet;
     4.  Burner type;
     5.  Efficiency of flame contact (mixing);
                              4-21

-------
     6.  Residence time;
     7.  Auxiliary fuel firing rate;
     8.  Amount of excess air;
     9.  Firebox temperature; and
    10.  Preheat temperature.
The first three parameters are characteristics of the production
process.  The next three parameters are characteristics of the
design of the incinerator.  The auxiliary fuel firing rate is
determined by the type and concentration of VOC in the SLA, the
SLA flow rate, the preheat temperature, and desired firebox
temperature. The last four parameters are operating variables
that can be set by the operator and will affect the performance
of the incinerator.  Based on studies, well-designed and
-operated incinerators can achieve VOC destruction efficiencies
of 98 percent or better.47'48
     At magnetic tape plants using thermal incinerators for
control of VOC emissions, a typical VOC concentration is about
3,500 ppmv of solvents such as MEK, MIBK, toluene, and
cyclohexanone.8  The SLA flows range from about 0.5 to 3.1 m3/sec
(1,000 to 6,500 acfm), discharge temperatures range from 732° to
788°C  (1350° to 1450°F), and the residence time between 0.75 and
            Q                                   •
1.5 seconds.   At the facility with the highest residence time,
an emission test of the incinerator indicated a destruction
efficiency of 99.99 percent.8
     The cost of incineration can be reduced by recovering heat
from the incinerator exhaust gases.  "Primary" recovery refers to
the transfer of heat directly from the hot incinerator effluent
to the relatively cool incinerator inlet VOC stream.  "Secondary"
heat recovery refers to exchange of heat from the incinerator to
any other process.  Heat recovery can be accomplished through
recuperative or regenerative methods.  With recuperative heat
recovery, the hot incinerator effluent is cooled using a heat
exchanger.  With regenerative heat recovery, the incinerator
effluent is cooled using a ceramic bed.  In either case, the
cooler incoming SLA is preheated by the hot exhaust gases.
                               4-22

-------
     Plants using primary and secondary heat recovery with
typical thermal efficiencies can achieve overall heat recoveries
of 70 to 80 percent.49  Actual overall energy savings obtained
vary with the VOC concentration in the oven exhaust, the
incinerator operating temperature, and the ability of the plant
to incorporate primary and secondary heat recovery.
     4.2.3.2  Catalytic Incinerators.  Catalytic incinerators use
a catalyst to promote the oxidation rate of VOC.  The SLA is
preheated by a burner or heat exchanger and then brought into
contact with the catalyst bed, where oxidation occurs.  Common
catalysts used are platinum or other noble metals on alumina
pellets or honeycomb support.  Catalytic incinerators can achieve
destruction efficiencies similar to those of thermal incinerators
while operating at lower temperatures, i.e., 400° to 540°C (750°
to 1000°F).  Thus, catalytic incinerators can operate with lower
energy costs than thermal incinerators.5^  Construction material
may also be less expensive because of the lower operating
temperatures.  Primary and secondary heat recovery, as described
in Section 4.2.3.1. can also be done with catalytic incineration.
     Factors important in the design and operation of a catalytic
incinerator include those that affect thermal incinerators but
also the acceptable operating temperature of the catalyst.  The
maximum temperature for the catalyst indirectly sets the upper
VOC concentration that can be incinerated.  For most catalysts,
temperatures greater than about 700°C (1300°F)  severely reduce
catalyst activity.51  Consequently, the heating value of the
inlet stream must be adjusted to avoid the higher temperature.
Typically, inlet VOC concentrations must be less than 25 percent
of the LEL.  The possibility of catalyst poisoning is also a
limiting factor in the use of catalytic incineration.6
     One catalytic incinerator in use in this industry controls
acetone (a non-HAP) emissions.  The other catalytic incinerator
in this industry is used in conjunction with the rotor
concentrators,  as discussed in Section 4.2.1.4.  Emission testing
conducted by the plant controlling acetone emissions indicates a
destruction efficiency of 98 to 99 percent.
                               4-23

-------
4.2.4  Flare Systems
     Flares are another incineration method for controlling VOC
emissions.  Although flares are a proven technology used for
controlling a wide range of gaseous emissions in other
industries, flares are not presently used in the magnetic tape
coating industry.  A brief description of flare technology,
factors affecting performance, and the potential of flares as a
VOC control method for ovens, mix preparation equipment, and
solvent storage tanks are discussed in this section.
     4.2.4.1  Operating Principles.  Flaring is an open
combustion process in which the oxygen required for combustion is
provided by the air around the flame.  Good combustion in a flare
is governed by flame temperature, residence time of components in
the combustion zone, turbulent mixing of the components to
complete the oxidation reaction, and the amount of oxygen
available for free radical formation.52
     There are two types of flares:  elevated flares and ground
level flares.  In an elevated flare system, process off-gases are
sent to the flare through the collection heater.  The off-gases
entering the heater can vary widely in volumetric flow rate,
moisture content, VOC concentration, and heat value.  They can be
used for almost any VOC stream and can handle fluctuations in VOC
concentration, flow rate, and inert content.1
     The VOC stream enters at the base of the flame, where it is
heated by already burning fuel and pilot burners at the flare
tip.  If the gas has sufficient oxygen and residence time in the
flame zone, it can be completely oxidized.  The high volume of
fuel flow in a flare requires more combustion air at a faster
rate than simple gas diffusion can supply, so flare designers add
steam or forced air injection nozzles to increase gas turbulence
in the flame boundary zones and, thus, draw in more combustion
air and improve combustion efficiency.  Steam injection promotes
smokeless flare operation by minimizing the cracking reactions
that form carbon, but it also causes increased noise and cost.
Typically 0.15 to 0.5 kg of steam per kg of flare gas  (0.15 to
0.5 Ib of steam per Ib of flare gas) is required.  With
                               4-24

-------
air-assisted flares, steam is not required.  Air assist is rarely
used on large flares because airflow is difficult to control when
the gas flow is intermittent.  Gases with heating values of below
approximately 18 MJ/scm (500 Btu/scf) may be flared smoke-free
with steam or air assist.52
     Ground flares are usually enclosed and have multiple burner
heads that are staged to operate based on the quantity of gas
released to the flare.  The energy of the flared gas itself
(because of the high nozzle pressure drop) is usually adequate to
provide the mixing necessary for smokeless operation, and air or
steam assist is not required.  A fence or other enclosure reduces
noise and light from the flare and provides some wind protection.
Ground flares are less numerous and have less capacity than
elevated flares.  Typically they are used to burn gas
"continuously," while steam-assisted elevated flares are used to
dispose of large amounts of gas released intermittently.52
     4.2.4.2  Design Factors.  The factors important in the
design and operation of a flare are:
     1.  Flammability limits of the gases flared;
     2.  Auto-ignition temperatures of the gases;
     3.  Heating values of the gases;
     4.  Density of the gases; and
     5.  Efficiency of flame contact (mixing).
     4.2.4.3  Control Efficiency.  Based on the results of
several studies, EPA has concluded that 98 percent combustion
efficiency can be achieved by steam-assisted flares burning gases
with exit flow velocities less than 19 m/sec (63 ft/sec) and with
heat contents over 11 MJ/scm  (300 Btu/scf).  Flares operated
without assist can also achieve 98 percent combustion given exit
flow velocities less than 18 m/s  (60 ft/sec) and burning gases
with heat contents over 8 MJ/scm  (200 Btu/scf),52
4.2.5  Conservation Vents and Pressure Relief Valves
     Conservation vents have long been used to minimize solvent
emissions from tanks used in a variety of industries including
magnetic tape manufacturing.3'53  These vents are valves that are
permanently attached to the outside of sealed,  vapor-tight
                               4-25

-------
vessels.  These valves open when either positive or negative
pressure within a vessel exceeds the design values of the valve.
Conservation vents minimize the VOC emissions that occur because
of cyclic changes in the temperature of the liquid inside a
vessel; these are known as "breathing losses."   With the vents,
emission losses to the atmosphere are reduced because they occur
only when the vapor pressure within the tank exceeds the pressure
setting of the vent.  As long as the pressure within the tank
remains below the vent setting, the tank does not breathe and
there are no losses.
     Figure 4-5 presents a diagram of a conservation vent.  The
vessel pressure is applied to the underside of the pressure
pallet and the top side of the vacuum pallet.  As long as the
vessel pressure remains within the valve pressure and vacuum
settings, the pallet remains in contact with the seat rings, and
no venting takes place.  The pressure pallet lifts from its seat
ring when the vessel pressure reaches the valve pressure setting
and allows the excess pressure to vent to the atmosphere.  As the
vessel pressure drops below the valve setting, the pressure
pallet returns to the closed position.  For a negative pressure
(vacuum), the vacuum pallet lifts from its seat ring when the
vessel vacuum reaches the valve vacuum setting, allowing air to
flow into the vessel to relieve the excess vacuum condition.  The
vacuum pallet returns to its normal position as the vessel vacuum
drops below the valve vacuum setting.54  Conservation vents do
not prevent the tank from venting when it is filled  (working
losses) because the internal pressure exceeds the set pressure on
the valve.
     The amount of VOC emission reduction achieved by
conservation vents depends on the solvent vapor pressure, the
diurnal  temperature change, the tank size, and the vent pressure
and vacuum settings.  Breathing and working losses from solvent
storage  tanks can be estimated using emission equations developed
by the American Petroleum Institute.55  Assuming a yearly average
diurnal  temperature change of 11°C  (20°F); 10,000 gallon tanks
holding  THF, toluene, and cyclohexanone; 20 turnovers per year;
                               4-26

-------
I
o

i

C
                                                                                                                        C
                                                                                                                        o
                                                                                                                       C
                                                                                                                       o
                                                                                                                       o
                                                                                                                       fO
                                                                                                                       ^
                                                                                                                       en
tn
0)
^
3
cn

-------
and conservation vents set at 0.215 kPa (0.5 ounce  [oz]) vacuum
and 17.2 kPa (2.5 psig) pressure, an average overall efficiency
of 50 percent is achieved.56
     Pressure relief valves operate in a manner similar to that
of conservation vents but require higher pressures to activate.
They usually do not protect against vacuum.  The pressure relief
valves prevent any breathing losses and much of the working
losses.  Based on the average vapor pressures of the solvents in
this industry and a pressure setting of 103 kPa (15 psig), a
control efficiency of 90 percent was calculated for pressure
relief valves.56
     Conservation vents or pressure relief valves are usually
installed as standard equipment on bulk storage tanks.  Virtually
all of the storage tanks in the magnetic coating industry have
vents of some sort, ranging in settings from 0.5 oz to 10 psig.3
4.2.6  New Solvent HAP Emission Control Technologies
     There are two relatively new control technologies that could
feasibly be applied to control solvent HAP emissions from the
magnetic tape manufacturing industry.  These technologies have
been reasonably well developed over the last decade and have been
used to control solvent emissions from a variety of operations on
both a full-scale and pilot-scale basis.
     4.2.6.1  Brayton-Cycle Condensation.   Development work has
continued since the early 1980's on the use of the Brayton Cycle
Heat Pump (BCHP) for solvent recovery.  Two distinct systems in
which the BCHP condenses solvents from a gas stream using the
reverse "Brayton cycle" have been developed.  One condenses
solvent directly from the SLA oven discharge.  The other uses the
BCHP in conjunction with nitrogen desorption of a fixed-bed
carbon adsorber.  The BCHP condenses the solvent from the
nitrogen.  This discussion elaborates on the first system.  The
second is a minor modification of a standard carbon adsorption
system.57
     The "reverse Brayton cycle" can be described in three steps.
In Step l, solvent-laden gas from the process is compressed.
During compression, the gas pressure is raised to two
                               4-28

-------
atmospheres.  Compression increases the gas temperature.  During
Step 2, a heat exchanger or "regenerator" drops the temperature
of the gas stream low enough to condense most of the solvents.
In Step 3, the gas pressure is relieved through a turbine.  The
Joule-Thompson effect lowers the gas temperature further and the
remaining solvents condense.  The turbine assists in driving the
compressor referenced in Step 1.  The clean gas stream is passed
through the regenerator to recoup heat and then is returned to
the process.57'58
     A demonstration project for the BCHP was conducted on a
magnetic tape coating line.  The system treated an 3.8 m3/sec
(8,000 cfm) exhaust stream from a drying oven containing MEK,
toluene, and cyclohexanone.  This system collected 227 kg
(500 Ib) of solvent per hour and was reported to operate with an
efficiency of greater than 99 percent.  Data have not been
collected on long-term reliability of the system.57
     Another installation involving the BCHP consists of the BCHP
in conjunction with a fixed-bed carbon adsorber to control
solvent emissions from one of the magnetic tape coating lines.
Control efficiencies of 96 to 99 percent have been reported, with
an average of 97.9 percent control.18
     4.2.6.2  Ultraviolet (UV)-Oxidation.  Over the past 6 years,
several facilities (aerospace and furniture manufacturing) in
southern California have installed UV-oxidation systems to
control VOC emissions from a variety of operations such as spray
booths, enclosed ovens, mix rooms, and large preparation rooms.
The system uses a combination of UV light and activated
oxygen/ozone to destroy solvent emissions in an onsite, self-
regenerative process.  Systems have been designed for a variety
of air flows and pollutant mixtures.  (The manufacturer reported
the process to be relatively insensitive to variations in flow
rates and pollutant concentrations.}  Limited testing at two
aerospace facilities indicated destruction efficiencies in the
range of 99 percent for solvent mixtures containing MIBK and
toluene.5^
                               4-29

-------
     The UV oxidation control system treats SLA in a four-step
sequence.  In Step 1, the air passes through a chemically inert
filter designed to remove large particles from gas.  Particles on
the order of 1 micron in diameter are reported to be removed at
an efficiency rating of 99 percent, but no detailed test data are
available.  In Step 2, the air is forced into a counter-current
water scrubber.  Oxidants are continually injected into the water
to enhance oxidation of VOC's to carbon dioxide and water and of
metals to metal oxides.  The exit water is again oxidized to
complete oxidation of the solvents.  The water is filtered to
remove any metal oxides present, then it is recycled to the
process.  Step 3 is a "coalescer," which removes water droplets
entrained in the air stream leaving the scrubber.  The gas stream
is then passed through activated carbon to remove the remaining
light-end, nonsoluble solvents.^
     This system has not been used by a magnetic tape coating
facility, but it has been used on systems with similar operations
and solvents, such as MEK, toluene, and cyclohexanone.  Although
the system has relatively high capital costs, it is reported to
have low long-term operating costs.  Other advantages of the
system are its low maintenance demand and zero hazardous waste
generation.
4.3  PARTICULATE HAP CONTROL SYSTEMS
     Eleven plants use solid raw materials that are HAP's (e.g.,
chromium, cobalt), all in fine particulate form.3  Particulates
are transferred into the mix either manually or using an enclosed
automatic device during mix preparation.  Of the facilities using
particulate HAP's, only two use add-on control devices (filters)
to control emissions.8  One plant, based on an estimate of annual
inlet and outlet particulate loadings, reports that greater than
99.9 percent emission control is achieved.50  Certain emissions
result depending on the method of transfer and whether or not an
add-on control device is used.  Emission factors have been
developed to calculate emissions from manual and enclosed
transfer, as discussed in Chapter 3, Section 3.3.2.2,
                               4-30

-------
Subsection B.  Enclosed transfer equipment and fabric filters are
discussed below.
4.3.1  Enclosed Particulate Transfer Methods
     Two types of enclosed particulate transfer methods are used
by the industry to transfer particulate to the mix tanks:  vacuum
injection and bag slitter devices.8  The vacuum injection system
draws particulate from a storage container, such as a drum, into
an enclosed hopper.  The hopper gravity-feeds a conveyer, which
is also enclosed, that carries the material to the mix
preparation vessel.  Users of such systems have reported that
there are negligible emissions from the system because all
components are in self-contained housing.60'61
     Bag slitting devices used by two magnetic tape manufacturers
also control most particulate HAP emissions.  The particulate raw
materials are received in bags.  A bag is placed into a hopper,
the hopper is closed, and an internal mechanism slits the bag,
releasing the particulate into an enclosed conveyor that feeds
the mix preparation vessel.  Alternatively, the hopper may be
located directly above the mix preparation vessel and feed it
directly.
     The control efficiency of the above methods has not been
measured in a magnetic tape manufacturing plant.  Using an
emission factor published in EPA's "Compilation of Air Pollutant
Emission Factors"  (AP-42) for a similar operation, 0.3 Ib of
particulate are emitted per ton of particulate transferred,
corresponding to a control efficiency of 99.9 percent.62
4.3.2  Pulse-Jet Fabric Filters63'64
     The primary components of a pulse-jet fabric filter system
are the bags and auxiliary equipment, the housing that contains
the bags, the inlet or dirty-side plenum that receives combustion
gases and distributes them to the bags, the clean-air plenum that
receives the cleaned combustion gas from the bags before it is
discharged to the atmosphere, and the hopper and discharge
system.
     The bag compartment is separated from the clean-air plenum
by a flat metal plate called a tube sheet.  Within this
                               4-31

-------
compartment, bags are suspended from the tube sheet and supported
internally by rings or cages.  Dust-laden air is filtered through
the bag, depositing dust on the outside surface of the bag.  All
pulse-jet systems filter the gas from the outside to the inside
of the bag.63'64
     Periodically, the dust cake is removed from each bag by a
blast of compressed air injected into the top of the bag tube.
The high-pressure back-flow of air shakes the bag and blows the
deposited particles from the outside surface of the bag, allowing
most to fall to the bottom of the enclosure.  The pressures
involved are commonly between 414 and 689 kPa (60 to 100 psig).
     Most pulse-jet filters use bag tubes that are 10 to 15 cm
(4 to 6 in.) in diameter and 1.8 to 5.8 m (6 to 19 ft) long.63
The bags are usually arranged in rows and are cleaned one row at
a time, in sequence.
     The performance of a fabric filter for metals emissions is
affected by the fabric filter design and operating parameters and
by process operating parameters.  Key pulse-jet filter design and
operating parameters are:
     1.   Air-to-cloth ratio (A/C) (or filtration velocity);
     2.   Bag material;
     3.   Operating temperature;
     4.   Pressure drop across the filter;
     5.   Cleaning frequency and back-flow air pressure;
     6.   Variations in temperature;
     7.   Variations in flow rate;
     8.   Variations in pollutant concentration; and
   " 9.   Humidity of air stream.
Process startup/shutdown procedures can also affect long-term
performance.
     The A/C ratio, a measure of the superficial gas velocity
through the filter medium, is the ratio of the actual flow rate
of gas through the fabric filter to the bag area.  It is usually
measured in units of m3/sec/m2  (acfm/ft2).  Fabric filter systems
that operate satisfactorily at a high A/C ratio are desirable
because it minimizes the size and cost of the filter.  In
                               4-32

-------
general, however, the lower the A/C ratio, the lower the
maintenance (bag replacement rate) and higher the removal
efficiency.  Although short-term exceedance of the design ratio
is not likely to have a substantial effect on emissions, long-
term exceedances increase bleed-through of particulate and
increase wear and potential bag failure.
     Beginning with new bags, the pressure drop across the fabric
increases as the particulate is filtered from and attaches to the
outside of the bag.  Some "cake" on the outside of the bag is
essential for good filtration; particulate often otherwise
penetrates the fabric.  This pressure differential would continue
to increase until there is essentially no flow through the
filter.  The frequency of cleaning is then adjusted so that the
pressure drop of the process air across the bag wall is
maintained between predetermined minimum and maximum levels.
This may be done manually or with a feed-back control loop
initiated by a differential pressure sensor across the dirty-
clean air plenums.  Cleaning too frequently can disrupt the cake
of particulate, resulting in an increase in emissions of
particulates.   On the other hand, cleaning that is not frequent
enough will allow excessive cake buildup, increased pressure drop
through the system, and reduced flow.  Frequency and length of
cleaning vary with the inlet loading, the A/C ratio, and plant
preference.  The cleaning frequency can range from several times
per hour to only once or twice per shift.
     The particle size distribution and loading of particulate
must be considered during the design (and operation) of a fabric
filter.  Within certain limitations  (+/- 10 to 20 percent of
design values), however, changes in these parameters reportedly
do not seriously affect fabric filter efficiency.65  A major
advantage of a properly designed fabric filter system is its
ability to perform well over the normal variation in exhaust gas
characteristics.  Fabric filters can collect particle sizes
ranging from submicrbn to several hundred microns in diameter at
efficiencies generally in excess of 99 or 99.9 percent.66  One of
the magnetic tape facilities that uses a fabric filter to control
                               4-33

-------
particulate HAP's can only estimate the efficiency because inlet
and outlet concentrations have not been measured.
4.4  VOC EMISSION CAPTURE SYSTEMS
     Although some web coaters merely sweep solvent HAP's from
the coating room to minimize worker exposure, many others use a
capture system that gathers emissions from storage tanks, mix
preparation equipment, coating operations, cleaning operations,
and waste handling to deliver them to an emission control
device.8  If a source is enclosed, total capture may be achieved.
Otherwise, the ratio of solvent HAP emitted to solvent HAP
delivered to the control device, or "capture efficiency," is a
function of the design and the specifics of the operation.  This
section contains a description of capture systems used on solvent
storage tanks, mix rooms, coaters, and drying ovens and
identifies factors that affect their performance.  Capture of
emissions from cleaning and waste handling operations is
discussed in Sections 4.6 and 4.7.
4.4.1  Capture of Emissions From Solvent Storage Tanks
     The HAP emissions from solvent storage tanks in this
industry can be 100 percent captured if the tanks are sealed and
the conservation vents are vented to the control device.  Five
magnetic tape coating plants are known to control VOC emissions
from solvent storage tanks by venting into the carbon adsorber
that serves the coating operation.9
     As discussed earlier, losses from storage tanks are either
from diurnal temperature changes or "fill losses."  The latter
occur when liquid is pumped into a partially empty tank.  As the
liquid level rises, the pressure in the saturated vapor space
increases until the overpressure safety device, which protects
the tank from physical damage, releases the vapors, usually to
the atmosphere.  No plants in this industry are known to use
"vapor balancing" to control emissions from the storage tanks
during product delivery.  With a vapor balance system, while
product is being delivered, a hose connecting the top of the tank
to the delivery truck allows the displaced vapors in the tank
                               4-34

-------
headspace to transfer to the headspace of the truck.  In turn,
the truck ultimately returns the vapors to a bulk storage system.
Vapor balancing is practiced by many gasoline distribution
systems that supply local service stations.
4.4.2  Capture of Emissions From Mix Rooms
     The VOC emissions from equipment in the mix room may be
captured to varying degrees by placing covers on the mix tanks,
by covering mix tanks using pressure-tight covers and venting to
the control device, or by venting part or all of the mix room air
to the control device.
     By covering and venting the mix equipment  (e.g., premixers,
holding tanks) to a control device, effective capture and control
of emissions can be achieved with minimum airflow.  Eight plants
are known to cover and vent mix equipment directly to a control
device.^  At some, however, not all of the mix equipment is
vented to the control device reportedly because of physical
constraints of the mix equipment or layout of the mix room.  One
plant has installed flexible hosing that is connected to the main
SLA duct to capture emissions from mobile mix tanks.19
     Room ventilation systems evacuate air from the room or rooms
in which the coating mix is prepared.  Ideally, the air exhausted
by this type of ventilation system may be used as oven make-up
air and then sent to a control device.  Currently, it is more
often discharged directly to the atmosphere.  Three plants are
known to vent part of the air from the mix preparation area to
the oven (and indirectly to the control device) or the control
device.^
     The large majority of plants in this industry cover the mix
equipment for most of the coating preparation cycle.  Some covers
or hatches may occasionally be opened to add raw material to the
coating batch.  The covers on mixers may have a circular opening
in the center through which the mix shaft passes.  Types of
"covers" include aluminum foil or plastic wrap, sheet metal, and
covers that are gasketed and bolted in place.  Four plants that
have completely sealed some or all of the mix tanks limit
                               4-35

-------
emissions using conservation vents that exhaust to the atmosphere
(see Section 4.2.5 for an explanation of conservation vents).8
4.4.3  Capture of Emissions From Coating Operations
     4.4.3.1  Coatingr Application/Flashoff Capture System.  Total
enclosures, room ventilation, partial enclosures, local point
ventilation, and overhead hoods are used to capture fugitive
emissions from the coater.  Typically, ventilation rates are
adjusted to avoid concentrations that exceed either the threshold
limit value for occupational exposures or 25 percent of the LEL.^
Total enclosures or room ventilation are generally used at new
plants because of regulatory requirements.67  Total enclosures
may also be desirable because of the improved product quality
control that can be achieved by controlling the quality of the
air which enters the system.
     The most effective emission capture system is a "total
enclosure."  The EPA has defined a total enclosure according to
the following criteria:68
     1.  Any natural draft opening shall be at least four
equivalent opening diameters from each solvent emitting point  (a
natural draft opening is any opening that remains open during
routine operation of the process and is not connected to a duct
with a fan or blower attached);
     2.  The total surface area of all natural draft openings
shall not exceed 5 percent of the surface areas of the
enclosure's four walls, floor,  and ceiling;
     3.  The average face velocity of the air through all natural
draft openings shall be at least 3,600 m/hr (200 ft/min).  The
direction of airflow through all natural draft openings shall be
into the enclosure;
     4.  All access doors and windows whose areas are not
included in the area determined in Item 2 and not included in the
calculation in Item 3 shall be closed during routine operation of
the process; and
     5.  All solvent emissions must be captured and contained for
discharge through a control device.
                               4-36

-------
     Three coating lines in the magnetic coating industry have
demonstrated to air pollution officials that their enclosures
meet this definition.9  Two others reportedly fulfill the above
criteria but formal evaluation was not required because the
plants were not subject to the NSPS.
     Twelve plants have installed enclosures that they claim to
be total because the enclosures completely surround the coating
application area.9  These enclosures, however, have not been
demonstrated to meet the EPA definition.  Figure 4-6 illustrates
one design of a total enclosure, in which the coater head is
surrounded by a small roomlike structure.  At some plants the SLA
from the coater head and web enclosure is directed to the drying
oven.  This ventilation approach reduces the total airflow rate
to the control system.  At other plants, the exhaust from the
coater head is sent directly to the control device.  Some plants
rely on the pressure difference between the coater enclosure and
the oven to draw the SLA into the oven.  The total enclosures may
also include point, or local, ventilation in areas of high
solvent concentration.
     In many web coating industries such as magnetic tape,
publication rotogravure, and pressure sensitive tapes and labels,
the coating room itself serves as a total enclosure.  It is also
possible for the coating room ventilation air to be directed to a
control device (see Figure 4-7(a)).  The cost of such a control
scheme can be significantly reduced by recycling the bulk of the
coating room ventilation air and withdrawing only a portion for
delivery to the control device.  For example, as shown in
Figure 4-7(b), perhaps only 10 percent of the coating room
ventilation air need be directed to the control device; the rest
is recirculated.   This dramatically reduces the size and cost of
the control equipment required.  The coating room ventilation air
may also be used as oven makeup air, thereby enriching it and
further minimizing the total air volume delivered to the control
device.  Obviously, it is advantageous to withdraw the air
                               4-37

-------
                         •
         \
COATER
(REVERSE
 ROLL)
0
 UNWIND
                                                                   TO CONTROL
                                                                     DEVICE
                                              OVEN
                                                                     REWIND
  Figure 4-6.   Schematic of total enclosure ventilation system.
                                  4-38

-------
             TO CONTROL
               DEVICE
                                                           AIR

H


t-LS
»•
1

OVEN


WALL— *
•^
COATEH
                                                             •AN
       TO CONTROL
         DEVICE
                   OVEN
                                                     U|
                                  AIR
                                 INLET
                                         RfORCULATEOAIR
                                         u
                             w
       TO CONTROL
         DEVICE
  CT-
          V
                   OVEN
f
                                 AIR
                                 INLET
                             (0)
Figure 4-7.   Schematics  of  room ventilation systems,
                             4-39

-------
richest in organics to the oven to minimize organic build-up in
the workroom areas.  This might be accomplished by situating a
hood as shown in Figure 4-7(c).
     A less effective method of confining VOC emissions is
partial enclosure of the coating application/flashoff area.  A
wide range of capture efficiencies is achievable depending on the
design of the partial enclosure and the airflow velocities.  For
a given design, the larger the air velocity, the higher the
capture efficiency.  However, the cost of the system and control
device (if applicable) depend on the airflow rate, and it becomes
uneconomical to operate with large flow rates and high capture
efficiencies.
     Two magnetic tape facilities are known to use partial
enclosures on some of their production lines.8  Types of partial
enclosure systems observed at magnetic tape plants include small
boxes around the coater with one side left open or with the sides
not reaching the ground.19  With these systems, some confinement
of fugitive emissions is achieved while ease of access is
maintained.  The VOC emissions are removed from inside the
enclosure by point ventilation systems or natural draft from the
oven and vented through the drying oven or directly to the main
SLA duct.  Canopy hoods may also be used to capture part of the
coater fugitive emissions.  To achieve good capture, large
airflow rates must be used with either of these hood systems.
Capture efficiencies achieved by these partial enclosures are
generally unknown and can be low because the openings allow
turbulence that can aggravate fugitive emissions.
     At least five magnetic coating facilities have lines with no
capture or control of the emissions from the application/flashoff
area.3/0°  Some of these are research and development lines.
Others, however, are production lines.
     4.4.3.2  Drying Oven Ventilation.  With the exception of
ovens with a nitrogen atmosphere, ovens are generally designed to
operate at a slightly negative pressure.  Despite this, a few
facilities in this industry are known to operate positive
pressure ovens, presumably 'to prevent dust from entering the oven
                               4-40

-------
and settling on the wet web.8  Two facilities operate ovens with
negative pressure at the inlet zone of the oven and positive
pressure at the outlet.8
     The negative pressure ovens are designed to prevent any loss
of VOC vapor to the room air.  The area of web entry and exit
openings is minimized with no substantial pressure differentials
across these openings.  It is also possible to install curtain
fans at the entrance and exit of the oven, in which SLA is pulled
from the area directly near the opening and vented to the control
device.19  This prevents both SLA from escaping the oven and
particles from entering the oven.  There is still, however, the
possibility of fugitives from the oven.
4.5  CONTROL OF PACKAGING AND LABELING EMISSIONS
     Only three plants in the magnetic coating industry use HAP's
in packaging and labeling for small-scale printing operations
(e.g., printing serial numbers on labels) associated with the
final product.8  No control technology is used at any of these
plants.  The use of inks without HAP's, such as water-based inks,
would result in 100 percent reduction of HAP emissions from this
source.  The remaining facilities purchase all preprinted labels
and inserts.
4.6  CONTROL OF CLEANING EMISSIONS
     Emissions from industrial cleaning operations may be reduced
in any of several ways:  HAP vapor capture and control, solvent
substitution, work practices, specialized cleaning equipment, and
mechanical cleaning.  The control technologies are discussed
below for each of the four categories of cleaning defined in
Chapter 3, Section 3.2.1.9.  These are referred to as cleaning
tanks, removable parts, fixed exterior surfaces, and flushing
fixed lines.  Not all of these control methods are presently used
in the magnetic tape manufacturing industry.
4.6.1  Vapor Capture and Control
     In some instances, vapor capture and control may be feasible
for controlling HAP solvent emissions from all four cleaning
categories.  Examples of vapor capture and control systems are
provided below.
                              4-41

-------
     As discussed in detail in Chapter 3, Section 3.2.1.9, tank
cleaning can involve manual or automated cleaning or a
combination of both methods ("hybrid" method).  At least two
facilities that clean their tanks using the hybrid method direct
emissions to an add-on control device via hoods and vent
emissions to the SLA.8'69  The capture efficiency of different
hoods will vary, but will not be perfect (100 percent).  One
facility measured the capture efficiency as 80 percent by
comparing the solvent sent to the SLA to the solvent input for
cleaning.8'69
     A commercial tank washing system for portable tanks less
than 600 gallons in size has recently become available and is in
use by one magnetic tape facility.  The commercial system
consists of a tank cleaner and control device in one unit.70'71
The system is totally closed with a cover assembly designed to
fit tanks of varying diameter.  Solvent is pumped into the tank
to be cleaned.  The system operates under vacuum to evaporate the
solvent vapors in the tank and to eliminate fugitive emissions.
All of the systems are equipped with a carbon adsorber or
condenser.  After the cleaning cycle, a fresh air purge occurs:
fresh air is pumped through the system bringing with it any
remaining solvent vapor.  Air is continuously pumped until the
pressure in the tank is exceeded and the air/solvent mixture is
released through the pressure relief valve to the carbon adsorber
or condenser unit.70'71  Theoretically, the only emissions are
fugitive emissions from pipe fittings and pumps, and from any
solvent that remains in the tank after cleaning.  No data are
available, however, to support this claim.  The system is limited
in that it is only appropriate for tanks that are less than
600 gallons in capacity.  Also, the system manufacturer does not
recommend that the system be used on stationary tanks with
existing cover assemblies.
     Closed-top cleaning is a requirement for paint and ink
manufacturers in Illinois.72  One coatings manufacturer in the
State has noted a substantial savings in the amount of solvent
used by switching to a closed-top cleaning system for both
                               4-42

-------
stationary and portable mix tanks.73  The in-house system
developed for their stationary tanks uses the same concept as the
commercial system described above.73
     At least two existing magnetic tape facilities have in-house
enclosed tank cleaning systems that are used on some of their
tanks.8'69  Interestingly, data collected by these facilities do
not indicate that closed-top cleaning results in fewer emissions
than cleaning the same size vessel using an open-top method.
However, the small time frame over which the emission test was
conducted could have masked the difference in emissions between
open- and closed-top cleaning.69  Also, the closed-top cleaning
experiments involved small quantities of solvent.  Therefore,
small errors in measurement accuracy could skew the overall
results.
     Vapor capture and control methods are also used to control
emissions from cleaning removable parts in wash tanks or sinks.
At least three magnetic tape manufacturers use hoods to vent
emissions from the wash sink to the SLA.69  At one, the air from
the room in which the wash sink is located is also vented to the
SLA, enhancing capture and control of fugitive emissions.8'69  At
facilities that use only hoods above the wash sink, the capture
efficiency is unknown but is less than 100 percent.  Most plants
have ventilation fans above the wash sink to draw solvent vapors
away from the worker, but not all vent to a control device.8'69
     At least three magnetic tape facilities with total
enclosures control emissions from the cleaning of fixed exterior
surfaces located in the enclosure.8'69  The capture efficiency,
however, is unknown and is most likely low if the enclosure doors
remain open for worker access.  At two facilities, the room in
which the enclosure is located is also vented to the SLA thus the
capture efficiency would be improved.8'69
     Finally, flushing of fixed lines may be controlled either by
flushing in a totally closed system in which the cleaning solvent
is never exposed to ambient air, or, by conducting the activity
in an enclosure, as discussed in Section 4.4.3.1, that vents to
the SLA or its control device.  The capture efficiency will be
                               4-43

-------
less than 100 percent if the enclosure does not meet the EPA
criteria for a total enclosure.
4.6.2  Solver^
     The HAP emissions from cleaning can be reduced or eliminated
if HAP materials can be replaced with cleaners which contain
nonhazardous materials.  The criteria used to select a cleaning
material include solubility of the resin in the coating mix and
cost.  Also, the cleaning agent must not harm the substrate and
must be easily rinsed from equipment so there is no risk of
cross -contamination.  The cleaning material normally used is one
of the solvents in the coating mix.3  If a facility does not use
solvent that contain SAP's in the coating operation, it may not
use them in the cleaning operation.  If a solvent other than one
used in the coating is selected for cleaning, it may or may not
be a HAP solvent.
     Nonsolvent cleaning agents include caustic solutions and
aqueous detergent/surfactant solutions.  At least one plant in
this industry uses soap and water to clean the calender rolls;
all other cleaning  (e.g., mix equipment) uses solvent.74   The
facility that uses this technology, however, indicates that soap
and water cannot be used universally in the industry, but that it
can use soap and water because of the uniqueness of its calender
roll substrate, which is a paper substrate.  Its calender is an
older technology and not widely used in the industry.  The
facility cautioned that most facilities use a plastic calender
that could not be cleaned with soap and water.74  The coating
mixes contain resins that harden fairly quickly.  As with any
cleaning agent, to be acceptable, an aqueous cleaner must be
capable of  (1) dissolving the resins;  (2) not harming the
surfaces being cleaned; and (3) being rinsed completely from mix
preparation equipment, so as not to contaminate the coating.
4.6.3  Work Practice
     Work practices that reduce solvent evaporation include
activities such as conservation of cleaning solvent; recovery and
reuse of solvent; and reduction of fugitive emissions during
cleaning.
                               4-44

-------
     Conservation--e.g., a reduction in solvent used for cleaning
operations--may also be effective in reducing solvent emissions.
The maintenance of solvent inventory records may help achieve a
reduction in usage.  Solvent used for cleaning has been reduced
in one plant by approximately 60 percent in response to
California regulation SB-14.69
       Most facilities in this industry contain dirty cleaning
solvent and/or dirty cleaning rags for solvent recovery either
onsite or offsite.8'69  At the large majority of the plants,
dirty cleaning solvents and rags are kept in closed containers,
such as 55-gallon drums, waste tanks, or covered buckets, in an
attempt to reduce fugitive emissions during storage prior to
treatment or disposal.  It is crucial that the container chosen
for dirty rags be impermeable to the solvent contaminant;
solvent vapors can permeate certain storage materials.  Thus, the
container the solvent is stored in can also affect solvent HAP
emissions.
4.6.4  Specialized Cleaning Techniques
     Facilities reduce the quantity of solvent emitted from
cleaning of removable parts by maintaining a minimum freeboard
ratio in the wash sink.  Freeboard ratio is the ratio of the
vertical distance from the evaporative area to the top of the
sink (freeboard height) divided by the smaller of the length or
width of the sink evaporative area.7^  Based on data collected
from the industry, uncontrolled wash sink emissions from
facilities with 75 percent freeboard requirements are 88 percent
lower than those facilities without a freeboard requirement or
any other control method.69  At many facilities, wash sinks are
covered when not in use, thereby reducing fugitive emissions when
the sink is inactive.
     At least two facilities dispense the fresh solvent used for
rag wiping using a commercial spring-loaded can that delivers
solvent to a mesh surface at the top only when the rag is pressed
to the mesh.69'76  This reduces the surface area of solvent
manually exposed to the atmosphere and the amount of solvent
                               4-45

-------
delivered to the rag.  The control efficiency of such cans is
unknown.
4.6.5  Mechanical Cleaning
     This category includes activities such as grit blasting,
water jet cleaning, and manual scraping and brushing.  No
mechanical cleaning by itself is used in the magnetic tape
manufacturing industry.  During mix tank cleaning, manual
scraping and brushing is used at many plants in conjunction with
a solvent.  Given the hardness of the resin, it is unlikely that
scraping alone would be able to clean the mix tanks.  Two plants
clean small parts with solvents that contain no HAP's using sonic
cleaning.20'77  This method has not had wider applications for
other types of cleaning in the industry, such as internal
surfaces of piping and mix tanks.
4.7  CONTROL OF EMISSIONS FROM WASTE HANDLING
     As described in Chapter 3, all facilities generate waste
solvent, coatings, rags, filters, etc.  For those that perform
onsite waste handling, there are two categories of emissions
control:  (1) reduction of fugitive emissions from transferring
and loading waste and  (2) emissions capture and control.
4.7.1  Reduction of Emissions from Waste Transfer and Loading
     A means of controlling emissions from the transfer of wastes
from their point of generation to their destination is to
transfer in a closed system.  Dirty solvent from cleaning tanks
and lines is pumped directly to the waste storage tank or
treatment device  (e.g., batch still) through a closed system,
never exposing the solvent to the atmosphere.60  This method of
control is used at several facilities.  Others drain dirty
solvent from the cleaned mix tank into a bucket or drum, carry it
to the waste handling device, and pour it in.60
     Dirty rags, filters, mops, etc. are often placed in drums or
buckets after use.  By their nature, however, they must be
physically "dumped" into the treatment device (e.g., filter
dryer) rather than transferred in a closed loop system.
     Proper design and operation of the treatment device
maximizes the solvent recovered from the waste,  thereby reducing
                               4-46

-------
emissions.  One plant continuously monitors the condensate from
the filter dryer with a gas chromatograph to ensure that the
treatment is not terminated before a minimum solvent
concentration is reached.19
     A leak detection and repair  (LDAR) program may also be
implemented in the waste handling area of a facility to minimize
leaks from pipe fittings.  Piping leak emissions from the waste
handling area were estimated using factors presented by the EPA.
An LDAR program, as it pertains to control of leaks from pipe
fittings, is discussed in Section 4.8.
4.7.2  Emission Capture and Control
     As with other operations at a magnetic coating plant, it is
possible to capture emissions from the waste handling device
process vent and direct them to a control device.  The most
effective method of control is to direct the waste handling
device process vent to the facility add-on control device.  Based
on EPA emission factors published in AP-42, a condenser vent
emits 3.3 pounds of VOC per ton of VOC waste loaded (99.8 percent
control).78  Additional control by an add-on control device
achieves 99.9 percent overall control.  Eight waste handling
devices in the magnetic tape industry use solvent HAP's and have
this level of control for the onsite waste handling devices.8
     One waste handling device in the industry is not vented to
the facility add-on control device.  There is some control of
emissions from the waste handling device with the primary
condenser (integral to the unit).  Plant-specific data indicate
that the control efficiency of this device is approximately
80 percent.8'60
4.8  CONTROL OF LEAKS FROM SOLVENT TRANSFER PIPING
      As discussed in Chapter 3,  leaks of HAP containing solvent
occur from the fittings, pumps, etc. associated with the piping
that transfers the solvent HAP's and the coating mix.   These
materials usually evaporate into the atmosphere.  The EPA
developed a leak detection and repair  (LDAR)  program to minimize
emissions from such leaks.  None of the magnetic coating plants
                               4-47

-------
currently have a program to routinely detect and limit these
emissions although other industries have adopted such programs.
     It is possible to inspect all pipe fittings and pumps on a
routine basis using a hand-held device that detects the presence
of VOC's.  When a rise in VOC level at a fitting is detected,
some type of repair or replacement activity is then made within a
short time frame.  The nature of the repair would vary with the
type of equipment that is leaking.
     Leaks from piping can be estimated using emission factors
developed by EPA.79  The EPA has also developed factors to
estimate piping leak emissions from various fittings that are
subject to an LDAR program. 9  The efficiency of an LDAR program
is dependent on the number and types of fittings, the VOC content
of the solvent in contact with them, and the duration of the
contact.  Such information is facility-specific.  The efficiency
of LDAR is calculated by comparing piping leak emissions before
and after the adoption of the LDAR program.
4.9  LOWER-VOC OR NON-HAP SOLVENT TECHNOLOGY
     The use of lower-VOC coatings such as high-solids coatings
or waterborne coatings is a potential method of reducing both
HAP and VOC emissions.  Although research in the areas of lower-
VOC coatings for magnetic tape is reported, such technologies are
not used in any existing or planned commercial facility.^
Coatings based on organic solvents are expected to continue to be
needed for many years, in large part because of the high cost of
the research and development of the lower-VOC or waterborne
coatings.80'81
     Eight plants in the magnetic coating industry use solvents
that do not contain HAP's  (cyclohexanone, THF, and acetone) in
the coating operation.^0  The use of acetone is limited to
operations that apply coatings to a paper substrate.  The list of
HAP compounds in the 1990 Clean Air Act is not final; compounds
may be added or deleted.  The possibility exists that
tetrahydrofuran and cyclohexanone may eventually be classified as
HAP's.  As a consequence, EPA has not recommended that the
industry switch to these solvents.  Even if a viable non-HAP
                               4-48

-------
option were available, the industry indicates that the research

and development costs would probably be the major deterrents to

plants switching as long as abatement remains a less expensive

option.80'81  Further, companies are concerned that research into

solvents not currently on the list of HAP's may be futile as new

compounds are added in the future.

4.10  REFERENCES FOR CHAPTER 4

 1.  Telecon.  Thorneloe, S., MRI, with Fischer, G., Memorex
     Corp.  May 17, 1983.  Information on control devices at
     Memorex, Santa Clara, CA, facility.

 2.  Telecon.  Meyer, J., MRI, with Fischer, G., Memorex Corp.
     May 23, 1983.  Information on absorber at Memorex, Santa
     Clara, CA, facility.

 3.  Memorandum and attachments from Beall, C., MRI, to
     Strum, M., EPA/ESD.  March 4, 1991.  Meeting with industry
     representatives.

 4.  Meyer, W.  Solvent Broke.  Vulcan-Cincinnati, Inc.,
     Cincinnati, Ohio.  Presented at TAPPI Test/PAP Synth. Conf.,
     Boston, October 7-9, 1974.  pp. 109-115.

 5.  Neveril, R.B., (GARD, Inc.).  Capital and Operating Costs of
   ' Selected Air Pollution Control Systems.  U. S. Environmental
     Protection Agency.  Research Triangle Park, NC.  Publication
     No. 450/5-80-002.  December 1978.  pp. 5-41.

 6.  Letter and attachment from Forber, R., Eastman-Kodak, to
     Johnson, W., EPA.  September 26, 1984.  Comments on draft
     BID of Magnetic Tape Manufacturing Industry NSPS.

 7.  Stern, A.  Air Pollution.  Third Edition.  Volume IV.
     Engineering Control of Air Pollution.  New York, Academic
     Press,  p. 343.

 8.  Memorandum from Angyal, S., MRI, to Strum, M., EPA/ESD.
     November 9, 1992.  Summary of confidential and
     nonconfidential information from magnetic cape manufacturing
     facilities.

 9.  Cubix Corporation, Austin, Texas.  Test Report on Exhaust
     Emissions from Weatherford 51 Coater at 3M's Magnetic Media
     Division, Weatherford, OK.  Prepared for 3M Company.
     February 1989.  p. 29.
                               4-49

-------
10.  Radian Corporation.  Full-Scale Carbon Adsorption
     Applications Study:  Draft Plant Test Report--Plant 3.
     Prepared for U. S. Environmental Protection Agency.
     Cincinnati,  OH.  EPA Contract No. 68-03-3038.  August 19,
     1982.  p. 29.

11.  Stunkard, C.  Solvent Recovery From Low Concentration
     Emissions.  Calgon Carbon Corp.  Undated,  p. 6-9.

12.  U. S. Environmental Protection Agency.  Control of Volatile
     Organic Emissions From Existing Stationary Sources--
     Volume I:  Control Methods for Surface-Coating Operations.
     Publication No. EPA-450/2-76-028.  Research Triangle
     Park, NC.  November 1976.

13.  Reference 4, p. ill.

14.  Reference 12, p. 33.

15.  Reference 12, p. 34.

16.  Reference 4, p. 115.

17.  Crane, G.  Carbon Adsorption for VOC Control. U. S.
     Environmental Protection Agency.  Research Triangle Park,
     North Carolina.  January 1982.  p. 23.

18.  Cubix Corporation, Austin, TX.  Test Report on Exhaust
     Emissions from a Solvent Recovery Unit at 3M's Consumer
     Video/Audio Division.  Hutchinson, M.N.  Prepared for 3M
     Company.  July 1991.  p. 6.

19.  Memorandum from Beall, C., MRI, to Strum, M., EPA/ESD.
     September 13, 1991.  Report of visit to Ampex Recording
     Media Corporation in Opelika, AL.

20.  Letter and attachments from Falco, M., 3M, to EPA.  June 13,
     1991.  Section 114 information request response for the 3M,
     Camarillo, CA, facility.

21.  Barnett, K.  W. et al. (Radian Corporation).  Carbon
     Adsorption for Control of VOC Emissions:  Theory and Full-
     Scale System Performance.  Prepared for the U. S.
     Environmental Protection Agency.  Research Triangle
     Park, NC.  June 6, 1988.

22.  Radian Corporation.  Full-Scale Carbon Adsorption
     Applications Study:  Draft Plant Test Report--Plant 2.
     Prepared for U. S. Environmental Protection Agency.
     Cincinnati,  OH.  EPA Contract No. 68-03-3038.  July 30,
     1982.  p. 27.
                               4-50

-------
23.  Basdekis, H. (IT Enviroscience).  Emission Control Options
     for the Synthetic Organic Chemicals Manufacturing Industry.
     Control Device Evaluation, Carbon Adsorption.  Prepared for
     the U. S. Environmental Protection Agency, Research Triangle
     Park, NC.  EPA Contract No. 68-02-2577.  New Orleans.
     February 1980.  pp. 11-25 to 11-26.

24.  Golba, N.,  and J. Mason.  Solvent Recovery Using Fluidized
     Bed Carbon Adsorption.  Union Carbide Corporation,
     Tonawanda,  NY.  Presented at the Waterborne and Higher
     Solids Coating Symposium.  New Orleans.
     February 17-19, 1982.  18 pp.

25.  Letter and attachments from Cowles, H., E.C.& C.
     Environmental, to Strum, M., EPA/ESD.  June 2, 1992.
     Information on air pollution control devices.

26.  Telecon.  Thorneloe, S., MRI, with Pfeiffer, R., Union
     Carbide Corp.  August 22, 1983.  Information on the cost of
     fluid-bed adsorbers.

27.  Telecon.  Thorneloe, S., MRI, with Pfeiffer, R., Union
     Carbide Corp.  September 8, 1083.  Information on use of
     caustic drying with fluid-bed carbon adsorbers.

28.  Letter and attachments from Danas, M., Sony Music
     Operations, to Beall, C., MRI.  August 6, 1991.
     Nonconfidential HAP questionnaire.

29.  Letter and attachments from Wetzel, J., Calgon Carbon Corp.,
     to Beall, C., MRI.  February 13, 1984.  Information on
     Calgon Ventsorb unit.

30.  Telecon.  Beall, C., MRI, with Byron, B., Tigg Corp.
     February 8, 1984.  Information on disposable-canister carbon
     absorption system.

31.  Telecon.  Strum, M., EPA/CPB, with Ramachandran, R., Memorex
     Technologies, Inc.  July 20, 1992.  Information on rotor
     concentrator.

32.  Kenson, R.E.  Paint Booth Solvent Emission Control
     Technology.  Plating and Surface Finishing. July 1988.
     pp. 18-19.

33.  Daikin Industries, Ltd.  Honeycomb Type Deodorization
     Equipment and Deodorization Systems.  EDX lOb.  August 1989.

34.  Daikin Industries, Ltd.   Honeydacs Technical Information.
     EDH-1.  October 1989.
                               4-51

-------
35.  Gronvaldt, J.  Advanced Rotor Concentrators Using
     Hydrophobic Zeolites.  Moenters Zeol.  Amesbury, MA.
     Presented at U. S. Department of Energy Industrial Recycling
     Conference.  San Diego, CA.  May 23-24, 1991.  4 pp.

36.  ABB Flakt.  Environmental Control Systems Product Data.
     APF144491.  Madison Heights, MI.

37.  Memorandum from Strum, M.,  EPA/CPB, to Berry, J., EPA/CPB.
     April 7, 1992.  Meeting minutes from ABB Paint Finishing
     presentation to EPA.

38.  Izumo, M.  Development of a Deodorization and Solvent
     Recovery System.  Conservation and Recycling (Great
     Britain).  i£:195-208.  1987

39.  Jeffcoat, I.A., and M.D. Heil.  Case Study:  Assembling
     Systems to Optimize Volatile Organic Compound Recovery.
     Pollution Prevention Review.  Spring 1992.  pp. 209-217.

40.  Rothchild, R.  Curing Coatings with an Inert Gas Solvent
     System.  Journal of Coatings Technology.  5_3_(675) :53-56.
     April 1981.

41.  Nikityn, J.  Inert Atmosphere Solvent Recovery.  Reprinted
     from the Journal of Industrial Fabrics,  i,(4).  Spring
     1983.

42.  Erikson, D.  (IT Enviroscience).  Emission Control Options
     for the Synthetic Organic Chemicals Manufacturing Industry--
     Control Device Evaluation,  Condensation.  Prepared for the
     U. S. Environmental Protection Agency, Research Triangle
     Park, NC.  EPA Contract No. 68-02-2577.  July 1980.
     p. II-l.

43.  Telecon.  Beall, C., MRI, with Rieman, D., Airco Industrial
     Gases.  February 15, 1984.   Information on the Airco
     condensation system.

44.  Telecon.  Thorneloe, S., MRI, with Memering,  L., United Air
     Specialists.  May 4, 1983.   Information on the Kon-den-
     Solver™ system.

45.  Section 114 response with no cover letter from
     Fritzemeir, J., Syncom Computer Group.  Received May 8,
     1991.

46.  Telecon.  Meyer, J., MRI, with Harper, S. Verbatim Corp.
     March 3, 1983.  Information on the control system at
     Verbatim, Sunnyvale, CA, plant.

47.  Memorandum from Mascone, D., EPA, to Farmer,  J., EPA.
     June 11, 1980.  Thermal incinerator performance for NSPS.


                              4-52

-------
48.  Memorandum from Mascone, D., EPA, to Farmer, J., EPA.
     June 11, 1980.  Thermal incinerator performance for NSPS,
     Addendum.

49.  U. S. Environmental Protection Agency.  Pressure-Sensitive
     Tape and Label Surface Coating Industry--Background
     Information for Proposed Standards.  Publication
     No. EPA-450/3-80-003a.  Research Triangle Park, NC.
     September 1980. p. 4-18.

50.  Reference 12, p. 51.

51.  Reference 12, p. 54.

52.  U. S. Environmental Protection Agency.  VOC Emissions from
     Petroleum Refinery Wastewater Systems-Background Information
     for Proposed Standards.  Preliminary Draft.  Research
     Triangle Park, NC.  July 1984.   pp. 4-25 to 4-32.

53.  Letter and attachment from Crist, W., W.P. Crist Company,
     Inc., to Beall, C., MRI.  February 6, 1984.  Pollution and
     Gas Control Equipment Bulletin CP-6003-B.  P. 55.

54.  Reference 6, p. 52.

55.  U. S. Environmental Protection Agency.  Compilation of Air
     Pollution Emission Factors  (AP-42).  Fourth Edition.
     September 1985.  pp. 4.3-1 through 4.3-35.

56.  Memorandum from Glanville, J., MRI, to Magnetic Tape Project
     File.  October 19, 1984.  Conservation vent control
     efficiency.

57.  Jain, N., and P. Scheihing.  Solvent Recovery Using the
     Brayton Cycle Heat Pump.  18 pp.

58.  Scheihing, P.  The U. S. Department of Energy Brayton Cycle
     Solvent Recovery Heat Pump Project Status Report.  U. S.
     Department of Energy, Office of Industrial Programs.
     February 1990.

59.  Telecon.  Lambert, K.', MRI, with Jackson, T., Terr-Aqua
     Enviro Systems, Inc.  August 27, 1991.  Information on the
     TABS UV Oxidation System.

60.  Memorandum and attachments from Angyal, S., MRI, to the
     Magnetic Tape NESHAP Project File.  December 1, 1992.
     Calculation of baseline emissions for the magnetic tape
     manufacturing industry.

61.  Letter and attachments from Ramachandran, R., Memorex
     Technologies, Inc., to Williams, D., MRI.  January 2, 1992.
     Information regarding particulate transfer.


                               4-53

-------
62.  Reference 55,  pp. 8.15-1 through 8.15-11.

63.  Beachler, D., and M. Peterson.  APTI Course SI:412A,
     Baghouse Plan Review Student Guidebook.  Publication
     No. EPA 450/2-82-005.  April 1982.

64.  U. S. Environmental Protection Agency.  Control Techniques
     for Particulate Emissions from Stationary Sources, Volume I.
     Publication No. EPA 450/3-81-OOSa.  September 1982.

65.  U. S. Environmental Protection Agency.  Operation and
     Maintenance Manual for Fabric Filters.  Publication
     No. EPA/625/1-86-020.  June 1986.

66.  U. S. Environmental Protection Agency.  OAQPS Cost Manual.
     Publication No. EPA 450/3-90-006.  January 1990.

67.  Code of Federal Regulations.  Standards of performance for
     new stationary sources.  Standards of performance for
     magnetic tape coating facilities.  40 CFR Part 60,
     Subpart SSS.  July 1, 1989.

68.  U. S. Environmental Protection Agency.  The Measurement
     Solution:  Using a temporary total enclosure for capture
     efficiency testing.  Publication No. EPA-450/4-91-020.
     Research Triangle Park, NC.  August 1991.

69.  Memorandum and attachments from Angyal, S., MRI, to the
     Magnetic Tape NESHAP project file.  Janaury 6, 1993.
     Statistical evaluation of cleaning protocol data.

70.  Letter and attachments from Poulos, F., Washtech Systems,
     Inc., to Strum, M., EPA/ESD.  October 2, 1991.  Information
     on the Port-A-Wash system.

71.  Telecon.  Angyal, S., MRI, with Poulos, F., Washtech
     Systems, Inc.  June 11, 1992.  Information on the
     Port-a-Wash system.

72.  Memorandum from Dart, D., EPA Region V, to the file.
     February 7, 1992.  Report of visit to Universal Chemicals
     and Coatingsi

73.  Telecons.  Strum, M., SPA/ESD, with Chin, F., and C. Corp,
     Universal Chemicals and Coatings.  June 8 and 15, 1992.
     Information on tank cleaning.

74.  Telecon.  Angyal, S., with Hughes, R., et al, Sony Music
     Corporation.  December 11, 1991.  Information on cleaning
     operations.

75.  Bay Area Air Quality Management District.  Regulation 8,
     Rule 16.  8-16-203, 204.


                               4-54

-------
76.  Sales brochure.  C&H Distributors, Inc.  Milwaukee, WI.
     Information on Eagle containers.  1991.

77.  Letter and attachments from Falco, M.,  3M, to Weigold, J.,
     EPA/ESD.  June 13, 1991.  Section 114 information request
     response.

78.  Reference 55.  Pp. 4.7-1 through 4.7-8.

79.  Memorandum from Whitt, D., Radian Corporation, to
     Markwordt, D., EPA/ESD.  June 5, 1991.   Impacts from the
     control of VHAP emissions from equipment leaks in non-SOCMI
     process units for HON.

80.  Telecon.  Beall, C., MRI, with Chagnon, M.,  Omniquest.
     July 31, 1991.  Information on non-HAP coatings.

81.  Telecon.  Strum, M., EPA, with Chagnon, M.,  Omniquest.
     September 1991.  Information on non-HAP coatings.
                              4-55

-------

-------
               5.0  MODIFICATION AND  RECONSTRUCTION

     National emission standards for hazardous air pollutants
(NESHAP) apply to both new and existing facilities that are major
sources of hazardous air pollutants  (HAP's).  As defined in
Section 112(a)(1) of the Clean Air Act (CAA),  a major source is
any source that emits or has the potential to emit, considering
controls, 10 tons per year (tons/yr)  of any one HAP or 25 tons/yr
of any combination of HAP's.
     For both new and existing sources, the standard must be at
least as stringent as the maximum achievable control technology
(MACT)  floor.  According to the CAA,  the MACT floor for new
sources shall not be less stringent than the emission control
achieved by the best controlled source in the source category.
For this industry (less than 30 sources),  the MACT floor for
existing sources shall not be less stringent than the average
emission limitation achieved by the best performing five sources.
     The situation may occur in which an existing major source,
currently subject to the standard, undergoes a modification or
reconstruction as defined in Sections 112(a)(5)  and 40 CFR 63.2,
respectively.  Depending on the extent of the modification or
reconstruction, and the associated change in HAP emissions, the
existing facility may become subjecc to the standards for new
sources.  For this industry, however, the proposed standard will
be the same for new and existing sources and no changes will be
required in the level of control.   Therefore,  a further
discussion of what constitutes a modification or reconstruction
under this NESHAP is not necessary.
                               5-1

-------

-------
           6.0  MODEL  LINES AND  REGULATORY ALTERNATIVES
                         •       •
6.1  GENERAL
     Model lines are descriptions of lines that will be
encountered in new or modified magnetic tape manufacturing
plants.  The purpose of this chapter is to define these model
lines, to identify the emission points associated with these
lines, and to present the baseline emissions from the model line
emission points.  The term model line is being used to describe a
new or modified plant.  This is because new plants are expected
to be constructed with one coating line plus sufficient space to
add additional coating lines in the future.   Existing plants are
expected to expand capacity by adding only one coating line at a
time.  The emission points identified in this chapter that are
associated with the coating operations are included in the "model
line description;" these emissions points would be encountered in
new or modified facilities.
     Because the number of existing major sources in the magnetic
tape industry is relatively small, it was not necessary to define
models for existing sources.   Instead, plant-specific data were
used to estimate baseline emissions for existing facilities,  as
described in Chapter 3.
     In addition to defining model lines, this chapter also
presents the regulatory alternatives for new lines and existing
facilities and the impact of these alternatives on, reducing
hazardous air pollutant (HAP)  emissions.  The regulatory
alternatives represent various courses of action that the
U.S. Environmental Protection Agency  (EPA) could take in
controlling HAP emissions from magnetic tape manufacturing
plants.  Regulatory alternatives are limited to those control
                               6-1

-------
methods that meet or exceed the maximum achievable control
technology  (MACT) floor, as described in Section 6.4.  The
environmental, energy, and economic impacts associated with
applying the alternatives to each of the model lines and to
existing facilities are presented in subsequent chapters.
Finally, control techniques that were not considered to be
feasible for this industry are discussed.
6.2  MODEL LINES
     Five model lines representing three line sizes--small,
medium, and large, have been selected to characterize the
manufacturing lines expected to be constructed, modified or
reconstructed in the future.  The emission points associated with
these lines are described in detail in Chapter 3, and are
summarized in Table 6-1.  This table shows that most, but not
all, emission points are common to all three model line sizes.
Because newly constructed, modified or reconstructed lines are
subject to the new source performance standard (NSPS) for the
industry, the model lines defined in this chapter incorporate
control devices or methods required by the NSPS.1  A summary of
the NSPS requirements is provided in Table 6-2.
     The three line sizes are analogous to the research, small,
and typical model lines developed for the NSPS, which were
categorized by the major design parameters of production rate,
hours of operation, coating solvent content, and coating
thickness.  From these sizes, five NESHAP model lines have been
established.  These are:  (1) a small line;  (2) a medium line
built with concurrent construction of a volatile organic compound
(VOC) control device; (3) a medium line built without concurrent
construction of a VOC control device; (4) a large line built with
concurrent construction of a VOC control device; and  (5) a large
line built without concurrent construction, of a VOC control
device.  It is necessary to distinguish between the situation
when the plant concurrently constructs a VOC control device with
the new coating line and when the plant uses an existing control
                               6-2

-------
device with a new coating line because the requirements of the
NSPS, and hence baseline emissions, are different.
6.3  MODEL LINE PARAMETERS
     Model line parameters and HAP emission rates are given
separately in Tables 6-3 through 6-9 for each emission point.
The model line parameters were based on data gathered from
existing facilities and from NSPS requirements.1"'  The basis for
defining these parameters is discussed below.  Some of the
information used in defining model lines was taken from plants
that use solvent mixtures that are a combination of HAP and non-
HAP VOC's, and some was taken from plants where none of the
solvents used are HAP's.  For the purposes of determining the
maximum potential HAP emissions for different types of model
lines, it was assumed that all solvents used by the model lines
are HAP's.
6.3.1  Solvent Storage Tanks
     The model storage tanks described in Table 6-3 are of the
size and number required to supply appropriate solvent quantities
to the model mix room.  The number of storage tanks for large
lines is six; this is based on the average number of storage
tanks used by existing facilities that have lines that could be
classified as large.1   The number of storage tanks for small and
medium lines is equal to the number of different solvents used in
the industry.1  Analysis of existing plant data indicates that
toluene, methyl ethyl ketone (MEK), and methyl isobutyl ketone
(MIBK) are the three HAP solvents that are used in significant
quantities in the industry.5'9  The sizes selected for the three
storage tanks are the most common used in the industry.2
     For the purpose of calculating storage tank emissions, total
solvent usage was apportioned as follows:  40 percent MEK,
40 percent toluene, and 20 percent MIBK  (the approximate relative
usage of these solvents in existing lines).5'9  The uncontrolled
VOC emission levels shown in Table 6-3 were calculated using
equations published in EPA's compilation of Air Pollutant
Emission Factors (AP-42).11  Model storage tanks were assumed to
be fixed roof (cylindrical horizontal) vented to the atmosphere.
                               •3 - J

-------
Emissions may be reduced by installing conservation vents or by
ducting the emissions to a control device such as a carbon
adsorber.  The conservation vents have an average control
efficiency of 35 percent and the carbon adsorption control method
reduces emissions by 95 percent.
6.3.2  Mix Room Equipment
     The model mix rooms described in Table 6-4 are defined as
the total mix equipment (mixers, polishing tanks, and holding
tanks) that supplies coating to each model coating operation.
The assumed number of mills, mixers, polishing tanks, and holding
tanks  (identified in Table 6-4) is based on a reasonable minimum
number that could supply a coating line.  Because there is
considerable variation in types and numbers of equipment used for
coating mix preparation among magnetic tape lines, the model
chosen for mix room equipment represents only one of many
possible combinations.^
     The uncontrolled HAP emission levels presented in Table 6-5
for mix preparation, application, and drying processes were
calculated using emission factors developed for the NSPS for the
magnetic tape industry.  These factors, presented below,
apportion uncontrolled solvent HAP emissions as a percentage of
total solvent HAP usage.  (Solvent usage is, as defined in
Chapter 3, the total annual quantity of solvent used to make
coatings that are applied to a substrate.)
          Coating mix preparation            10 percent
          Coating application                 9 percent
          Coating drying                     81 percent
In the mix room, HAP emissions occur as a result of mix
preparation and leaks from piping as discussed below.
     6.3.2.1  Mix Preparation.  The NSPS defines two cases for
new medium and new large model lines:  those that are built with
concurrent construction of a VOC control device and those that
are built without concurrent construction of a VOC control
device.  Different levels of control are required for each case
because, in the case when a new VOC control device is built
concurrently with the new line, the design loading of the

-------
adsorber can incorporate the mix room  equipment.  Mix equipment
for those lines built concurrent with  a  control device must  be
covered and vented to the control device (with the  exception of
pressurized vessels such as mills); for  lines "without control
devices," mix equipment is only required to be covered (no
venting to the control device required).  Covering  mix equipment
reduces uncontrolled emissions by approximately 40  percent
whereas covering and venting to a control device reduces
emissions by 95 percent  (see Chapter 3,  Section 3.3.2.2,
Subsection B).  The controlled emission  rates presented in
Table 6-4 are based on these control efficiencies.
     6.3.2.2  Emissions Resulting from Piping Leaks.   Chapter 3,
Section 3.3.2.2, Subsection F, presents  the rationale for
deriving a factor  (0.36 percent of total HAP solvent  used  to make
applied coatings) to estimate emissions  from piping leaks  in the
mix preparation area of existing plants.  The same  factor  was
used for the model lines.  Because most  existing plants do not
control piping leaks and the NSPS does not require  control,
"baseline for the model lines is no control of piping  leak
emissions in the mix room.
6.3.3  Coating Application/Drying
     As discussed in Section 6.3.2, uncontrolled HAP  solvent
emissions from coating application and drying constitute
90 percent of the HAP solvent used to make applied  coating.   Of
that, 9 percent emanates from the applicator and the  remainder
(81 percent) from the dryer.  These factors were used to
calculate the uncontrolled emissions from coating operations for
the model lines as presented in Table  6-5.
     For new facilities, the NSPS requires control  of 93 percent
of the VOC applied at the coater.  This  control level may  be
achieved by totally enclosing the line and venting  it to a carbon
adsorber.  Since development of the NSPS, EPA has established a
policy (as stated in the EPA document, "The Measurement Solution:
Using a Temporary Total Enclosure for Capture Efficiency
Testing"), that credits a source with  100 percent capture  if the
source is contained within an enclosure  that meets  the criteria
                               5-5

-------
presented in Chapter 4.  For enclosures built with those
characteristics, no testing is required to substantiate
effectiveness .
     Carbon adsorbers can be consistently operated at
efficiencies of at least 95 percent.  For the NESHAP analysis, it
follows that the combined efficiency of a total enclosure/carbon
adsorber is 95 percent, not 93 percent as identified in the NSPS.
For that reason, the controlled model lines are presumed to
control 95 percent of the VOC applied at the coater.  Small
lines, as defined herein, do not require control under the NSPS
because they are below the applicability cutoff.  Table 6-5 lists
uncontrolled and controlled emissions from coating operations, as
well as data necessary in sizing a carbon adsorber or
incinerator.
6.3.4  Solvent Recovery
     Both medium and large model lines are assumed to vent to a
carbon adsorber and perform on- site solvent recovery.  On- site
solvent recovery is a current practice at existing facilities
with coating lines that could be classified as medium and
large. ^  The model small lines do not have solvent recovery
capability because no existing facility with small lines performs
on-site solvent recovery operations.
     Piping associated with solvent delivery and recovery is a
potential source of leaks, as is the piping associated with
transfer of solvent from the solvent recovery device and to and
from the waste and recovered solvent storage tanks.  An emission
factor  (1.7 percent of annual HAP solvent used to make applied
coatings) was developed using data from existing plants to
estimate piping leak emissions in the solvent recovery area of
existing facilities (see Chapter 3, Section 3.3.2.2,
Subsection F) .   This factor was also used for model lines.
     Because no existing plant has instituted leak control
activities, piping leaks in the solvent recovery area of model
lines are uncontrolled.  Quantitative estimates of leak emissions
from piping in the solvent recovery area are presented in
Table 6-6.
                                -

-------
6.3.5  Waste Handling
     Existing plant data reveal that only those facilities with
existing lines that could be classified as large perform on-site
waste handling operations.5'9  Onsite waste handling involves the
use of equipment such as filter dryers, thin film evaporators,
and/or batch distillation units.  Most of the waste handling
equipment at these facilities have integral condensers to recover
the solvent mixture.  The waste handling device process vent is
then usually directed to the facility control device.  Only one
facility operates waste handling equipment (a filter dryer) that
has no integral condenser, but the facility still directs the
waste handling device process vent to the plant carbon adsorber.^
Large model lines were assumed to operate waste handling devices
equipped with integral condensers.  Because the NSPS does not
require control of the waste handling device process vent, it was
assumed that model lines discharge process vent emissions to the
atmosphere.  Emissions from waste handling were calculated using
EPA-established emission factors to be applied to the amount of
solvent processed by the waste handling device annually (see
Chapter 3, Section 3.3.2.2, Subsection D).  These factors also
may be used to calculate fugitive emissions from spills and
loading based on the amount of waste solvent processed.  An
average value of 771 tons per year of waste loaded per large line
was estimated from data on existing lines that use waste handling
devices.   Estimated HAP emissions from waste handling operations
are presented in Table 6-7.
     Leaks in piping associated with waste handling operations
are also a source of solvent HAP emissions.  Piping leak
emissions from waste handling operations were estimated for
existing facilities and model lines using parameters from
representative waste handling units defined by EPA (see
Chapter 3, Section 3.3.2.2, Subsection F)'-.  Equipment leak
emissions in the waste handling area are estimated as 1.0 Mg/yr
(1.1 tons/yr)  for large lines.
                               <$-'

-------
6.3.6
     Data from existing plants reveals that only 3 of the
25 operating facilities use HAP solvents for packaging and
labeling operations.9  Therefore, model lines are not assumed to
use HAP's in packaging and labeling operations.
6.3.7  Particulate HAP Emissions
     Particulate emissions can occur when dry chromium dioxide or
cobalt, (both of which are HAP's), is added to the coating mix
during mix preparation.  Because some existing plants of every
size use such particulates to make coatings, all model lines were
assumed to emit particulate HAP's.9 Approximately one-half of the
existing facilities control particulate emissions from transfer
into the coating mix.9  Because the NSPS does not require control
of particulate emissions, model lines were assumed to have no
controls for particulate HAP's.
     An average particulate HAP usage of 221.4 tons/yr was
calculated using data obtained from existing "large" lines and
that value was used to estimate parameters for small and medium
lines using relative solvent use ratios.^'9  A factor of
0.25 percent of particulate HAP usage was used to estimate model
line particulate emissions, as described in Chapter 3,
Section 3.3.2.2, Subsection B.  Table 6-8 presents particulate
HAP emissions for model lines.
6.3.8  Cleaning Activities
     Information obtained from the industry identifies four
cleaning categories describing routine cleaning activities that
are significant sources of solvent HAP emissions: (l) cleaning of
tanks used in the coating mixing process, (2)  removable parts
cleaning,  (3) cleaning of fixed exterior surfaces, and
(4) flushing of fixed lines that carry the coating mix from the
mix room to the coater. ' 10  These activities occur to some
degree at all of the existing facilities regardless of size;
therefore, each model line is assumed to perform them.5'10
     Emissions from cleaning activities at model lines were
estimated using factors developed from data provided by existing
facilities (for a detailed explanation, see Chapter 3,
                               6-8

-------
Section 3.3.2.2, Subsection E).  These factors estimate emissions
from each cleaning activity as a fraction of total cleaning
solvent usage.   Total cleaning solvent usage was also estimated
(as a fraction of total solvent usage) using data provided by
existing facilities.  Solvent HAP usage and uncontrolled
emissions from each cleaning activity are summarized in
Table 6-9.
6.4  REGULATORY ALTERNATIVES
     Control options have been developed for each of the HAP
emission points identified in this industry.  They are:  solvent
storage tanks, mix preparation (solvent and particulate),  coating
application/drying, waste handling, piping leaks, and
packaging/labeling.  Emissions also result from cleaning
activities such as tank cleaning, removable parts cleaning,
cleaning of fixed exterior surfaces, and flushing of fixed lines.
Control options for solvent and particulate HAP emissions are
presented in Table 6-10.
     The control levels of the control options are based on
statements from industry representatives, test data from related
industries, and engineering judgement.2,5,10  information
regarding control levels for each emission point is provided in
Chapter 3, Section 3.3.2.2 and in Chapter 4.
     These control options were combined in various ways to form
regulatory alternatives for this industry.  Tables 6-11 and 6-12
present the alternatives, the associated control techniques, and
the potential industrywide HAP emission reductions for both
existing major sources and new sources.  The HAP emission
reduction associated with each alternative was calculated using
the estimated emission rates, capture device efficiencies, and
control device efficiencies.  The MACT floor for new sources is a
combination of the best control method for each emission point
that is currently used by existing major facilities.  In this
case the MACT floor for new sources is the same as the MACT floor
for existing sources.
     The control options for each emission point identified in
Table 6-11 have equivalent or greater emission reductions than
                               6-9

-------
the average of the best-performing five sources in the source
category or the "MACT floor."  (Less stringent control options
were not analyzed.)  In order to identify the impacts of the MACT
floor, the EPA first needed to determine which existing
facilities were "major sources."  A major source is defined as a
facility that emits greater than 10 tons per year of any HAP
compound, or greater than 25 tons per year of multiple HAP
compounds.
     To determine which sources are major, baseline emissions
were estimated for existing magnetic tape facilities using the
assumptions and methodology outlined in Chapter 3.  As explained
in Chapter 3, emission estimates were based on the plant-specific
total annual HAP solvent usage.  The definition of major source,
however, requires that the determination be based on potential
emissions.  Therefore, for the purposes of this analysis, a
"potential" operating schedule of 24 hour per day, 7 days per
week, and 52 weeks per year was identified.  The initial baseline
emission estimates for each plant were subsequently scaled-up
using the plant reported operating schedule and the assumed
potential operating schedule.
     Any existing facility that had the potential to emit 10
tons/yr of any HAP or 25 tons/yr of multiple HAP's was considered
major in our analysis.  (Fourteen of the twenty-five operating
magnetic tape manufacturing plants were classified as major.)
Cost impacts, emission reductions, and other environmental
impacts presented in this document are based on the assumption
that these sources will be classified as major.  In determining
which facilities will actually be subject to the NESHAP
requirements, the appropriate regulatory agency will designate
sources.  This designation will likely be based on more specific
information on each plant than is possible for the EPA to use in
its analysis.  Thus, our designation of a source as major should
not be used to determine applicability of the NESHAP requirements
if more specific plant information is available.
                               6-10

-------
6.4.1  gxisting Manor Sources
     In Table 6-11, the regulatory alternatives are presented in
increasing order of stringency.  Alternative I is the "MACT
floor."  It requires:  (1) 95 percent control of emissions from
storage tanks, (2) installation and use of covers and venting of
all mix preparation equipment emissions to a control device
operated at 95 percent or greater efficiency, and (3) 95 percent
control of emissions from coating application and drying
operations.  In addition, waste handling devices would be
controlled at a rate of at least 95 percent, which can be
achieved by directing the waste handling device process vent to a
95 percent efficient control device.  A 75 percent freeboard
ratio is required for wash sinks used for removable parts
cleaning (freeboard ratio is defined in Chapter 4,
Section 4.6.4).   Equivalence can be achieved by venting the wash
sink to a control device provided the overall control efficiency
is demonstrated to be 88 percent.  Flushing of fixed lines would
be a closed-system operation, or solvent supply and collection
vessels would be vented to a 95 percent efficient control device.
Alternative I also requires that particulate HAP's be controlled
by 99.9 percent.   This control efficiency can be achieved by
using an enclosed transfer method.  Alternative I does not
require control of piping leaks in the mix room, solvent
recovery, and waste handling areas, or control of HAP emissions
from packaging and labeling activities, tank cleaning, and
cleaning of fixed exterior surfaces.
     Implementation of Alternative I would result in industrywide
emission reductions of 50 percent of HAP solvents and 70 percent
of HAP particulates from existing major sources.  The impacts to
sources that are not major sources based on our analysis have noc
been calculated.
     Note that Table 6-11 presents no levels of control other
than the MACT floor for the storage tanks, mix preparation,
coating application/drying, waste handling (excluding piping
leaks), tank cleaning, flushing fixed lines, and particulate
transfer emission points.  For these emission points, the Agency
                               S-ll

-------
believes the MACT floor represents the highest level of control
feasible.
     Regulatory Alternative II includes a leak detection and
repair (LDAR) program, as defined in Section 4.8, to control
piping leak emissions in the mix preparation, solvent recovery,
and waste handling areas.  (In the mix room, equivalence can be
achieved by venting the mix room air to the control device.)  An
LDAR program is required by the Resource Conservation and
Recovery Act (RCRA),  Subtitle C, for treatment, storage, and
disposal facilities that require a RCRA permit.  An LDAR program
is also being proposed for the synthetic organic chemical
manufacturing industry in the hazardous organic NESHAP  (HON).
The LDAR program examined for the magnetic tape manufacturing
industry is based on the program proposed in the HON
(57 PR 62608).   None of the magnetic tape facilities are now
RCRA-permitted.  Implementation of an LDAR program was included
in each of the regulatory alternatives above the MACT floor
because  (1) it results in the most dramatic emission reduction of
the available control options and (2) it has been found to be
reasonable in other manufacturing industries in which solvents
flow through equipment components.  .Alternative II also requires
the use of closed containers for collecting and dispensing
solvent.  The latter two control options were also included in
each alternative above the MACT floor because the emission
reductions are either so small, or could not be quantified, that
the percent reduction from baseline does not change.  By
including these control options in each regulatory alternative, a
number of regulatory alternatives having negligible additional
impact on emissions were eliminated.  As indicated on Table 5-11,
industrywide emission reductions of 60 percent of HAP solvent and
70 percent of particulate HAP's from existing major sources would
result from implementation of Alternative II. The particulate HAP
emission reduction for every alternative is 70 percent for
existing facilities,  because the only feasible control option for
particulates constitutes the MACT floor.
                               6-12

-------
6.4.2  New Lines
     Regulatory alternatives for new lines are the same as those
for existing sources, as presented in Table 6-11.  The emission
reductions associated with these alternatives are given in
Table 6-12.  As indicated in Table 6-12, emissions from the small
line are reduced by approximately 94 percent for both
Alternatives.  This is because the only additional control
required by Alternative II is the LDAR program and uncontrolled
equipment leak emissions from the small model line are less than
0.1 Mg/yr  (0.1 ton/yr).   The emission reductions from medium and
large lines designated as "A" are greater than those for the
medium and large "B" lines because the mix equipment of the "A"
lines is currently uncontrolled.
6.4.3  Control Options Not Included in the Regulatory Alternative
     Some of the control techniques that were discussed in
Chapter 4 were not included in either regulatory alternative
because they were not considered feasible for this industry.
Specifically eliminated were control technologies for controlling
emissions from tank cleaning and fixed-exterior surface cleaning.
     As discussed in Chapter 4, Section 4.6.1, emissions from
tank cleaning may be controlled by venting emissions through a
hood that is exhausted to a control device or by using either a
commercial or in-house closed-top tank cleaning system.  Also,
although not currently practiced by anyone in the industry, tanks
could conceivably be cleaned in a total enclosure that is vented
to an add-on air pollution control device.  In such a scenario,
the mix room in which the tanks are located would operate as a
total enclosure.
     None of the above control techniques is included in the
regulatory alternatives.  Control of emissions from tank cleaning
by venting emissions through a hood to a control device is not
recommended because, based on cost estimates, it is costlier than
cleaning the tanks in a total enclosure and provides less
emission reduction.  The emission control achieved by venting
emissions from tank cleaning through a hood is less than that
                               a-13

-------
which would be achieved by cleaning within a total enclosure
because the capture efficiency of a hood is less.
     The control strategy that requires that tanks be cleaned in
a total enclosure that vents to a control device was further
evaluated.  After analyzing this alternative, it was determined
that such a control system may not be technically feasible at
many facilities.  First, a facility may not be able to operate
its mix room as a total enclosure because the number and location
of existing windows, doors, and other openings may prevent it
from meeting the total enclosure criteria (see Section 4.4.3).
Second, even if the mix room could operate as a total enclosure,
the emission stream from this activity is a high volume, low
concentration stream.  A facility may not be able to control such
a stream with an existing control device, or the stream may
reduce the efficiency of the control device.  The installation of
a separate control device to control this stream would be costly,
and was therefore not considered.
     The final control strategy that was considered for tank
cleaning is the use of either a commercial or in-house closed-top
tank cleaning system.  Given the limitations of these systems
that were discussed in Chapter 4, Section 4.6.1, and the fact
that emission tests conducted by the industry did riot conclude
that closed systems were necessarily better than open-top
cleaning systems, the use of a closed-top tank cleaning system is
not required by either regulatory alternative.
     Emissions from the cleaning of fixed exterior surfaces are
controlled at some facilities by venting the enclosure in which
they are located to the control device during cleaning
activities.  The efficiency of such a system is unknown because
the capture efficiency is unknown.  The capture efficiency will
almost always be less than 100 percent because a worker will be
present in the enclosure during cleaning.  While the worker is
present, doors to the enclosure will likely be open and the
enclosure will not be "total."  The technical feasibility of
requiring facilities to run their enclosure while cleaning
surfaces is questionable.  As with tank cleaning, the emission
                               6-14

-------
stream will be a high volume, low concentration stream that a
facility may not be able to control with an existing control
device, or that may adversely affect the performance of an
existing control device.  Given this possibility, the regulatory
alternatives do not require this control technique.
                              5-15

-------
     TABLE 6-1.  SUMMARY OF EMISSION POINTS FOR MODEL  LINES
Emission point

Storage tanks
Mix room
Equipment leak emissions
from piping from mix room
to coating operation
Coating application/
drying
Equipment leak emissions
from solvent recovery
operations
Waste handling
Packaging/labeling
Cleaning activities
Particulates
Model line
Small
*
*
*
*



*
*
Mediuma
*
*
*
*
*


1C
*
Largea
*
*
*
*
*
*

if
*
alncludes lines that are built with and without concurrent
 construction of a control device.

 Note; '*'  indicates emission point is included in model line
                               6-16

-------
             TABLE 6-2.  SUMMARY OF NSPS REGULATION1
EMISSION POINT
STORAGE
All storage tanks
REGULATION
TANKS
No controls required
MIX ROOM
New mix equipment with
concurrent construction of
control device (except when
control device is a condenser)
All other mix equipment
Use covers and vent emissions
to a 95 percent efficient
control device
Use covers alone or use covers
and vent emissions to a
control device
COATING OPERATION
New coating operation
Modified or reconstructed
operations with existing level
of control .>90 percent
Modified or reconstructed
operations with existing level
of control .>90 percent that
install a new control device
Control 93 percent of VOC
content of coating applied at
the applicator
Maintain existing control or
93 percent, whichever is lower
Install a 95 percent efficient
control device and maintain
previous control level up to
93 percent
NOTES:
      Emission points not addressed by the NSPS:
         -emissions from piping leaks
         -emissions from cleaning activities
         -emissions from waste handling
         -emissions from packaging/labeling
         -particulate HAP emissions

      The NSPS requirements apply to all new sources that use
      greater than 38 m3 (10,039 gal) of solvent in coating
      operations.

-------
TABLE 6-3.  MODEL SOLVENT STORAGE TANK PARAMETERS2'10
Line designation:
Sol vent .usage per
line, nr/yr
(gal/yr)
No. of different
solvents used
No. of storage tanks
Capacity of each
tank, mj (gal)
HAP emissions per
line, Mg/yr
(tons/yr)
Small
26
(6,810)
3
3
4
(1,000)
0.02
(0.02)
Medium
77
(20,445)
3
3
4
(1,000)
0.04
(0.04)
Large
770
(204,445)
3
6
40
(10,000)
0.96
(1.06)
                         6-18

-------
                     TABLE  6-4.    MODEL  MIX  ROOM PARAMETERS
Line designation:
Line subcategory:
Substrate width, m (in.)
Line speed, m/s (ft/min)
Operating, hr/yr
2. Mix room information
Equipment, No.:
Mixers
Millsb
Holding tanks
Polishing tanks
Equipment ventilation
rate per item, nP/hr
(acfh)c
Uncontrolled HAP
emissions (10 percent of
total usage), Mg/yr
(ton/yr)
Mix room controls
Control efficiency, %
Controlled HAP
emissions, Mg/yr (ton/yr)
Piping leak emissions,
Mg/yr (ton/yr)
Small
0.15(6)
1.3 (250)
2,000
2
1
1
1
5.7 (200)
2.1 (2.3)
Cover
40
1.3 (1.4)
0.07 (0.08)
Medium9
A
0.15(6)
1.3 (250)
6,000
2
1
2
2
5.7 (200)
6.4 (7.0)
Cover
40
3.8 (4.2)
0.23 (0.25)
B
0.15(6)
1.3 (250)
6,000
2
1
2
2
5.7 (200)
6.4 (70.0)
Cover/CA
95
0.3 (0.4)
0.23 (0.25)
Large*
A
0.66 (26)
2.5 (500)
6,000
2
1
2
2
5.7 (200)
64.0 (70.4)
Cover
40
38.4 (42.2)
2.3 (2.5)
B
0.66 (26)
2.5 (500)
6,000
2
1
2
2
5.7 (200)
64.0 (70.4)
Cover/CA
95
3.2 (3.5)
2.3 (2.5)
Ref.
3
3
4
2
2
2,5
1
2
5
5
aLine "A" is a new facility that was not built concurrently with a control device.  Line "B" is a new facility
 that was built concurrently with a control device.
     HAP emissions from working losses in sealed mills will be pushed into the next tank and subsequently
 controlled by that tank's control device.
cFor systems purging tanks and ducting emissions to control device.
                                              1-19

-------






03
&
W
EH
i
J3
CU

3
O
OPERATI

i
*••*
H
C-,
g
O
U
J
w
Q
i


in
I
V^
W
j
§
EH








^


«S
IJ
<











s«
.3 -9
il








1
V3










e designation:
e aubcategory:
c e
33
*> -o <•>
S^»c
e
si|
—







3S§
*•*
§








3S§
§














Coaling formulation:
% VOC, weight
% VOC, volume
Density, kg/m3 Ob/gal)



W* S"
SS8
111
II
s> •-
1 *
s f
2



*^ S-* .*•*
« « «
ss§,
^*-* '*»«' *^

.M o ^
±(2 z
Methyl ethy
Methyl isobuly




*~* ^* .*•*
« « *
§§§
j* ~3 n
c^
">, T.
1 1
•£
">* .2
1 f
1








Solvent mixture


r~
<"*
E
s
b







f
<2,
2
b







f
2
b












^?
c
•a
iting head area ventilation
Total enclosure in3/* (ft3/
3
0
r-
«» « g-s>
^•a








oo « sr jr-
N^SS?
o o 55, «









o« >O «• IS-
««8S?
oo^a














Ovea ventilation rate
actual m/sec
atandard nr/sec
actual (ft3/min)
standard (tt /min)


•o
1
N_^
v»
VI
m







i

VI
V|








i
s^
s














Oven temperature, K (°F)


m
|
N-*-
••«
S







|
N^*









|
•^
«n








c
o
^-/
X
w"
1
4J
a.
Carbon adsorber inlet tern


M
44(160)
m







1

5
to







1
m








I
w
U
o.
jU
IS
_~a
Incinerator heat exchangei
K("F)


oo
S|
rc 20°C (0»"
mil control efficiency u Uic pc
2 u
111 5
« -3 J -3
5-20

-------
TABLE 6-6.  SOLVENT RECOVERY PARAMETERS FOR MODEL LINES5'10
Line designation:
Line subcategory:
Onsite solvent recovery
capability:
HAP solvent used to make
applied coatings, Mg/yr
(ton/yr)
Solvent recovery piping leak
emissions (0.017 x HAP
coating solvent usage) ,
Mg/yr (ton/yr)
Small
No
23.5

N/A

Medium
(A and B)
Yes
64.1
(70.5)
1.1
(1.2)

Large
(A and B)
Yes
641.2
(705.3)
11.0
(12.1)

                            6-21

-------
TABLE 6-7.  WASTE HANDLING PARAMETERS FOR MODEL LINES5'10
Line designation:
Line subcategory:
Onsite waste handling
capability
HAP waste loading, Mg/yr
(ton/yr)
Condenser vent emissions
(3.3 Ib/ton HAP waste
loaded) , Mg/yr (ton/yr)
Fugitive emissions from
spills (0.2 Ib/ton HAP waste
loaded) , Mg/yr (ton/yr)
Fugitive emissions from
loading (0.72 Ib/ton HAP
waste loaded) Mg/yr (ton/yr)
Total HAP emissions from
waste handling, Mg/yr
(ton/yr)
Small
No
N/A
N/A
N/A
N/A
N/A
Medium
(A and B)
No
N/A
N/A
N/A
N/A
N/A
Large
(A and B)
Yes
699
(771)
1.2
(1.3)
0.1
(0.1)
0.2
(0.3)
1.5
(1.6)

-------
TABLE 6-8.  PARTICULATE HAP PARAMETERS FOR MODEL LINES5'10
Line designation:
Line subcategory:
HAP used
Annual HAP particulate usage
per line, Mg/yr (ton/yr)
HAP emissions (0.0025 x
particulate HAP's used for
coating) , Mg/yr (ton/yr)
Small
Cr02
or Cobalt
6.7
(7.4)

0.017
(0.019)
Medium
(A and B)
Cr02
or Cobalt
20.0
(22.1)
-
0.05
(0.055)
Large
(A and B)
Cr02
or Cobalt
200.8
(221.4)

0.5
(0.55)
                           S-23

-------
   TABLE 6-9.  CLEANING ACTIVITIES--PARAMETERS FOR MODEL LINES
Line designation:
Line subcategory :
Quantity of HAP solvent used
for cleaning (0.17), Mg/yr
(ton/yr)a
Quantity of HAP solvent used
for tank cleaning (0.72),
Mg/yr (ton/yr)°
Uncontrolled emissions from
tank cleaning (0.071), Mg/yr
(ton/yr)c
Quantity of HAP solvent used
for removable parts cleaning
(0.22), Mg/yr (ton/yr)b
Uncontrolled emissions from
removable parts cleaning
(0.485), Mg/yr (ton/yr)c
Quantity of HAP solvent used
for exterior surface cleaning
(0.015), Mg/yr (ton/yr)b
Uncontrolled emissions from
exterior surface cleaning
(1.0) , Mg/yr (ton/yr)c
Quantity of HAP solvent used
for flushing fixed lines
(0.045), Mg/yr (ton/yr)*3
Uncontrolled emissions from
flushing fixed lines (0.0),
Mg/yr (ton/yr)c
Small
3.6
(4.0)
2.6
(2/9)
0.2
(0.2)
0.8
(0.9)
0.4
(0.4)
0.1
(0.1)
0.1
(0.1)
0.2
(0.2)
0.0
(0.0)
Medium
(A and B)
10.8
(12.0)
7.8
(8.6)
0.5
(0.6)
2.4
(2.6)
1.2
(1.3)
0.2
(0.2)
0.2
(0.2)
0.5
(0.5)
0.0
(0.0)
Large
(A and B)
108
(120)
78.2
(86.3)
5.5
(6.1)
24.0
(26.4)
11.6
(12.8)
1.6
(1.8)
1.6
(1.8)
4.9
(5.4)
0.0
(0.0)
aThe factor in parentheses is applied to the total quantity of
 HAP solvent used for coating.
°The factor in parentheses is applied to the total quantity of
 HAP solvent used for cleaning.
cThe factor in parentheses is applied to the quantity of HAP
solvent used for a particular cleaning activity.
                               1-24

-------
     TABLE 6-10.
       EMISSIONS
  CONTROL OPTIONS FOR HAZARDOUS  AIR POLLUTANT
FROM MAGNETIC TAPE MANUFACTURING  FACILITIES
Emission point
Control
Overall HAP
control,
percent*
COATING OPERATIONS
Storage tanks
Mix preparation
Coating application/drying
Waste handling
Piping leaks (plantwide)
Packaging and labeling
Participates
Vent to 95% efficient control device
Cover/vent mix vessels to 95% efficient control
device
Total enclosure vented to 95% efficient control
device
Direct the waste handling device process vent to
a 95% efficient control device.
1. No control
2. Implement LDAR program
No control
Enclosed transfer
95
95
95
95
0
89
0
99.9
CLEANING ACi'IVfmiS
Tank cleaning
Removable parts
Fixed exterior surfaces
Flushing of fixed lines
All cleaning activities
No control
75% Freeboard ratio or equivalent through
venting to control
No control
Closed system or vent supply and waste
collection vessels to 95% efficient control device
1. No control
2. Closed containers for dispensing solvent; lids
on solvent collection containers when not in
use.
0
88
0
95
n/a
aThe control efficiency indicated is the control for that specific emission source, not the entire facility.
                                   6-25

-------
 I
CO
& at
             e
             f>
                 I
                 *
                   ll

                 if! i
                 l^a-s
                 Svi 3 o
                 •353
                 i * |

                 *j*.
                 w J O O
                 £ « '3 •—
                 * 2 ,-• ^

                    ,
                        i
"S JS S o
> 2 '3 .a
i H S 5
U > u -3
3 a -
«• o O
II!
                        •fe
        s

        I
        S
        Si
        •
        ?
                                1   1
                                1   .s
                                \   I
                                •s
                                 «
                                 41
                                 2
                                    S-26

-------
   TABLE 6-12.  IMPACT OF REGULATORY ALTERNATIVES ON NEW LINES
Reg. Alt.a
I
II
Percent reduction in solvent HAP emissions*3
Small
line
93.1
93.4
Medium linec
A
38
52
B
16
35
Large linec
A
35
50
B
17
37
aRefer to Table 6-11 for descriptions of the regulatory
 alternatives.
"The reduction in particulate HAP emissions is the same for each
 model line and each regulatory alternative.  This is because the
 MACT floor contains the most stringent control option being
 considered for particulates (enclosed transfer).  The percent
 reduction in particulate HAP emissions is 94 percent.
GLine A is a new line that was not built concurrently with a
 control device.  Line B is a new line that was built
 concurrently with a control device.
                               5-27

-------

-------
6.5  REFERENCES FOR CHAPTER 6

 1.  U. S. Environmental Protection Agency.  Code of Federal
     Regulations.  Part 60.  Standards of performance for new
     stationary sources.  Subpart SSS.  Standards of performance
     for magnetic tape coating facilities.  Office of the Federal
     Register.  40 CFR Part 60, Subpart SSS, July 1, 1989.

 2.  New Source Performance Standard for the Magnetic Tape
     Manufacturing Industry.  U.S. Environmental Protection
     Agency.  Research Triangle Park, NC.  Publication
     No. EPA-450/3-85-029a.  December 1985.

 3.  Memorandum from Beall, C., MRI, to Magnetic Tape Project
     File.  June 22, 1984.  Summary of nonconfidential
     information on U.S. magnetic tape coating facilities.

 4.  Memorandum from Glanville, J., MRI, to Magnetic Tape Project
     File.  June 22, 1984.  Typical process parameters for
     magnetic tape coating facilities using fixed-bed carbon
     adsorbers.

 5.  Memorandum from Angyal, S., MRI, to Magnetic Tape Project
     File.  December 1, 1992.  Calculation of baseline emissions
     for the magnetic tape manufacturing industry.

 6.  Memorandum from Glanville, J., MRI, to Magnetic Tape Project
     File.  October 15, 1984.  Revised calculation of utility
     requirement for control devices.

 7.  Memorandum from Meyer, J., MRI, to Magnetic Tape Project
     File.  June 10, 1983.  Uncontrolled VOC emission rates for
    - model plants.

 8.  Memorandum from Meyer, J., MRI, to Magnetic Tape Project
     File.  June 13, 1983.  Calculation of oven and coating head
     area ventilation rates for preliminary model plants.

 9.  Memorandum from Angyal, S., MRI, to Magnetic Tape Project
     File.  November 9, 1992.  Summary of confidential and
     nonconfidential information from U.S. magnetic tape
     manufacturing facilities.

10.  Memorandum from Angyal, S., MRI, to Strum, M.,  SPA/CPB.
     December l, 1992.  Model lines development.

11.  U. S. Environmental Protection Agency.  Compilation of Air
     Pollutant Emission Factors (AP-42).  Fourth edition.
     September 1985.  Pp.  4.3-1 to 4.3-35.
                              6-28

-------
12.  Memorandum from Rasor,  S.,  MRI,  to Magnetic Tape Project
     File.  September 10,  1992.   Emission Standards for
     Treatment,  Storage,  and Disposal Facilities under RCRA and
     their applicability to  Magnetic  Tape Manufacturing
     Facilities.
                              5-29

-------
                    7.0  ENVIRONMENTAL IMPACT

    . The environmental and energy impacts of the regulatory
alternatives as applied to existing and new major sources are
presented in this chapter.  The regulatory alternatives are
identified in Chapter 6, Model Lines and Regulatory Alternatives
and are also summarized in Table 7-1.  (Appendix E presents the
environmental and energy impacts for each of the emission points
identified in the tables.)  In this chapter, the incremental
increase or decrease in air pollution, water pollution, solid
waste generation, and energy consumption that would result from
implementing the regulatory alternatives are discussed for
existing major sources and model lines.  Existing major sources
are those sources that have the potential to emit greater than
10 tons per year (tons/yr) of any hazardous air pollutant (HAP)
or 25 tons/yr of multiple HAP's.  Baseline HAP emissions from the
magnetic tape industry are provided in Chapter 3.  Baseline HAP
emissions from model lines are presented in Chapter 6.
7.1  AIR POLLUTION IMPACTS
     Hazardous air pollutants are components of the solvents and
particulates used by the magnetic tape industry.  The HAP
compounds are emitted from several points during magnetic tape
production and ancillary activities such as cleaning and waste
handling.  The HAP emission points can be controlled by use of
add-on control equipment such as carbon adsorbers, incinerators,
and the other control methods that are described in detail in
Chapter 4.  The two regulatory alternatives for existing and new
sources involve control of various combinations of the HAP
emission points.  The primary and secondary air pollution impacts

-------
CO
w
o
o
CO


i
H
E->
CO
H
X
a
CO
H
O
cu


§
H
CO
CO
CO
w
04


I
ED
o
u
OS
W
J
              0

              oa.
              u
              as
                I
              H
              l
            -ill
               '5
              iO
              3 2
                      s a •?

                      til
                    e
                   s
                       -
                       !
                    o
                           o
                         o o
                   •g g^ 2

                    5-3 5 I
                   y s S 3
  •a _

  .3 3 -3

*sfi § S
^ 1> O ^
                                        o
                  -
                  r-
                      1
                                        « §
                                        Z u
                                        o

                                        0
                                       SI
                  I   jala  f





                  Q S ^ -3 Q. > 
                                                       au
                                                       3-3

                                             * 1 ~
                    > "1
                    x *S s

                    'IS 3
                                        a  "
                                        -  u 3 y

                                        J5II2
                                                             oo
                                                             e
                                                             •a

                                                             S
                                                             2

                                                             c

                                                             '1


                                                             I
                                                             I
                                                             u
                                                             >

-------
that would  result from implementing each alternative  are
described below.
7.1.1  primary Air pollution Impacts
     Primary air pollutant impacts  are total HAP and  volatile
organic compound (VOC) emission  reductions resulting  from control
of emission points at major sources in the magnetic tape
manufacturing industry.  In this industry, all HAP solvent
compounds are VOC's; thus, all control options in the regulatory
alternatives would reduce HAP and VOC emissions to the same
extent.  The annual HAP solvent  emission level that would result
from implementation of each regulatory alternative at existing
major sources is presented in Table 7-2.  This table  also
provides the percent reduction in HAP solvent emissions from the
current baseline level as well as the megagrams per year (Mg/yr)
reduction from baseline.  Note that the reduction in  particulate
HAP emissions from existing major sources is the same for both
alternatives,  70 percent, because Alternative I presents the
highest level of particulate control considered feasible.
Particulate HAP emissions would  decrease to 0.12 Mg/yr [0.13 tons
per year  (tons/yr)], compared to the baseline particulate HAP
emission level of 0.39 Mg/yr  (0.43  ton/yr) for existing major
sources.

        TABLE  7-2.   ENVIRONMENTAL IMPACT OF  THE  REGULATORY
              ALTERNATIVES ON EXISTING MAJOR SOURCES3
Regulatory
Alternative15
Baseline
I
n
Percent reduction in
solvent HAP emissions
N/A
51
61
Emissions after alternative
implementation,
Mg/yr (tons/yr)
4,060 (4,470)
1,980(2,170)
1,590(1,750)
HAP emission reduction
from baseline,
Mg/yr (tons/yr)
N/A
2,080 (2,300)
2,470 (2,720)
  aThe percent reduction in particulate HAP emissions is the same for each alternative, 70 percent.
   Particulate emissions from existing major sources will be reduced from 0.39 to 0.12 Mg/yr (0.43 to
   0.13 ton/yr).
  "Refer to Table 7-1 for a description of the regulatory alternatives for existing sources.
                                 7-3

-------
For new sources,  the percent reduction in  solvent HAP  emissions
from the baseline level is provided in Table  7-3.  As  with
existing sources,  the percent  reduction in particulate HAP
emissions is the  same for each alternative, 94 percent,  because
Alternative I includes the highest level of particulate control.
The HAP solvent emission reduction in Mg/yr associated with the
regulatory alternatives for each new line  is  presented in
Table  7-4.  This  table also provides the reduction in  HAP solvent
emissions in Mg/yr from the baseline level.   As noted  in
Table  7-4, for the small model line, the particulate HAP
emissions would decrease to 0.001 Mg/yr  (0.001 ton/yr)  compared
to the baseline level of 0.02  Mg/yr (0.02  ton/yr).  For the
medium model lines,  the particulate HAP emissions would decrease
to 0.003 Mg/yr  (0.003 ton/yr)  compared to  the baseline level of
0.05 Mg/yr (0.06  ton/yr).  For the large model lines,  the
particulate HAP emissions that would decrease to 0.03  Mg/yr
(0.03  ton/yr) compared to the  baseline level  of 0.5 Mg/yr
(0.6 ton/yr).

    TABLE  7-3.   IMPACT  OF REGULATORY ALTERNATIVES  ON NEW LINES
Regulatory
Alternative3
I
n
Percent reduction in solvent HAP emissions"
Small line
93.1
93.4
Medium line0
A
38
52
B
16
35
Large linec
A
35
50
B
17
37
  aRefer to Table 7-1 for descriptions of the regulatory alternatives.
  ''The reduction in particulate HAP emissions is the same for each model line and each regulatory
   alternative. This is because the MACT floor contains the most stringent control option being considered
   for particulates (enclosed transfer).  The percent reduction in particulate HAP emissions is estimated as
   94 percent.
  cLine A is a new line that was not built concurrently with a control device.  Line B is a new line that was
   built concurrently with a control device.
      The HAP solvents used in  the magnetic  tape industry are
VOC's.   Because  all control options in the  regulatory
alternatives reduce VOC emissions to the  same extent  as HAP
emissions, the primary impact  of a HAP emission reduction is a
                                  7-4

-------
1
H
CO
CO
H
O.
CO
w
Kiti
** f-i
J O
  H

gs


• §
O 04
0]
             1^
            * us
                 1
I

                     a.
                     
-------
potential decline in ambient VOC levels and thus a reduction in
ozone and photochemical smog formation.
7.1.2  Secondary Air Pollution Impacts
     Secondary emissions of air pollutants result from generation
of the energy needed to operate the control devices required by
the regulatory alternatives for existing and new sources.
For those facilities that currently operate a control device, the
energy requirements of the regulatory alternatives are
incremental; i.e., in addition to the current energy expended at
a facility.  The combustion of natural gas in incinerators will
result in particulate matter (PM) ,  nitrogen oxides (NO..) , and
                                                      J^
carbon monoxide (CO) emissions.  The combustion of fuel  oil in
the boiler used to produce steam for the fixed-bed carbon
adsorption system will result in PM, NOX, and sulfur oxide  (SOX)
emissions.
     The combustion of natural gas as supplemental fuel  in
incinerators results in secondary air pollutants.  Assuming the
incinerator generates pollutants at a rate comparable to that of
an industrial process boiler, the secondary emissions were
calculated using the following emission rates:1
     - particulates      0.001 kilograms/gigajoule (kg/GJ)
                         [0.0024 pound/million British Thermal
                         Units (lb/106 Btu)]
     - NOX               0.057 kg/GJ
                         (0.133 lb/106 Btu)
     - CO                0.014 kg/GJ
                         (0.033 lb/106 Btu)
The natural gas consumption and resultant secondary pollutant
emissions presented in Table 7-5 reflect additional consumption
and pollutants for existing sources currently using incineration
as the primary means of VOC emission control.  Because the
regulatory alternatives require control of HAP emission  points
that may not currently be controlled, the fuel consumption of the
existing incinerator, and subsequent secondary pollutants, will
increase over the present amounts.  For existing sources, the
impacts are the same for Alternatives I and II because
                               7-6

-------
inclementing  the  additional control options under Alternative  II
does not result in increased loading of  the control device.  It
is assumed that model lines are currently controlled, or will  be
controlled, with  a fixed-bed carbon adsorption system.  Thus,  no
secondary pollutants from an incinerator's natural gas fuel
consumption are expected.

   TABLE 7-5.  SUMMARY OF ANNUAL SECONDARY POLLUTANT EMISSIONS
    FROM THE COMBUSTION OF  NATURAL GAS'-- EXISTING MAJOR SOURCES
Regulatory
Alternative"
I
a
Natural gas
consumption,
GJ/yr (10°Btu/vr)
70(65)
70(65)
Emission levels, Kg/yr (lb/yr)a
PM
0.1 (0.2)
0.1 (0.2)
NOX
4(9)
4(9)
CO
1(2)
1(2)
  aPM =• particulate nutter
  NOX =- nitrogen oxides
  CO =» carbon monoxide
  ''For a description of the regulatory alternatives for existing sources, refer to Table 7-1.
     The major  secondary air pollution  impacts for fixed-bed
carbon adsorption systems are the emissions  from the boiler used
to produce steam.   The steam is used to strip the carbon bed of
adsorbed VOC at a ratio of 4 kg of steam per kg of recovered
solvent  (4 Ib steam/Ib recovered solvent).2   Assuming that
existing and new sources use fuel oil containing 1.5 percent
sulfur by weight,  the following estimates can be made of the
levels of secondary pollutant emissions:
     - particulate
     - SO
         x
     - NO,
(1.25(S) +  0.38)  kg/1,000 liters  (103  L)
of fuel oil used
[(10(S) + 3)lb/l,000 gallons (103 gal)
of fuel oil used]
19(3) kg/103 L
[157(S) lb/103 gal]
6.6 kg/103  L
[55 lb/103  gal]
                                7-7

-------
where S is the percent sulfur by weight.3  The fuel oil consump-
tion was calculated assuming a heating value of 150,000 Btu per
gal of fuel, oil and a boiler thermal efficiency of 80 percent.
     The fuel oil consumption and resultant secondary pollutant
emissions are presented in Table 7-6 for existing sources.  The
fuel oil consumption for some existing facilities is the
incremental fuel oil consumption over the current fuel oil
consumption.  Additional fuel oil is required to control
additional HAP emission points, such as the mix preparation
equipment, waste handling device, or storage tanks, that may not
currently be controlled.  For existing facilities that would need
a new fixed-bed carbon adsorber to comply with a given regulatory
alternative, the fuel oil consumption is based on the operation
of a new fixed-bed carbon adsorption device.  Fuel oil
consumption and associated secondary pollutants for each new line
are presented in Tables 7-7 and 7-8.  For the small model line,
the fuel oil consumption is a result of a new control device
required to meet Regulatory Alternatives I and II.  For the other
model lines, the fuel oil consumption is the additional fuel
required as a result of controlling emission points that are not
currently controlled with the existing control device.  Medium
Line B has negligible fuel consumption for either alternative.
This is because the only control options not already being
implemented are 95 percent control of storage tank emissions, the
leak detection and repair (LDAR) program, and freeboard ratio.
The HAP solvent loading from the storage tanks is small, only
80 Ib/yr.  Therefore, the fuel oil requirements are also very
low, less than 5 gal/yr.  Neither the LDAR nor the freeboard
ratio options require energy as explained below.
     For existing and new sources, the implementation of the LDAR
program identified in Table 7-1 has negligible energy impacts.
This is because an LDAR program requires monitoring of pipe
fittings by a worker.  The hand-held device that is used for
monitoring is electronic.  A freeboard ratio has no energy
requirements because it simply requires that solvent in a wash
sink be maintained at a certain level.
                               7-a

-------
  TABLE 7-6.   SUMMARY OF ANNUAL  SECONDARY  POLLUTANT EMISSIONS
    FROM  THE  COMBUSTION  OF FUEL  OIL--EXISTING MAJOR SOURCES
Regulatory
Alternative11
I
n
Fuel oil
consumption,
KPL/yr (Kpgal/yr)
605 (160)
605 (160)
Emission levels,* Kg/yr (Ib/yr)
PM
1,300 (2,870)
1,300 (2,870)
S°x
17,240 (38,000)
17,240 (38,000)
NOX
4,000 (8,800)
4,000 (8,800)
aPM =» particulate matter
 SOX =» sulfur monoxide
 NOX = nitrogen oxides
bFor a description of the regulatory alternatives for existing sources, refer to Table 7-1.
 TABLE 7-7.
ANNUAL INCREMENTAL FUEL OIL  CONSUMPTION  FOR  STEAM
          GENERATION--NEW LINES
Regulatory
Alternative
I
n
Fuel oil consumption, l(rL/yr (10^ gal/yr)
Small line
6.0 (1.6)
6.0 (1.6)
Medium lines
A
23.6 (6.2)
23.6 (6.2)
B
0(0)
0(0)
Large lines
A
40.5 (10.7)
40.5 (10.7)
B
0.9 (0.2)
0.9 (0.2)
aFor a description of the regulatory alternatives for new sources, refer to Table 7-2.
"Line A is a new line that was not built concurrently with a control device. Line B is a new line that was
built
 concurrently with a control device.

-------
H

O
a
oo
 i
u









'C*
1
2
at
2
05
5
fc
§
£
>
S
ECOND
91











^
j
41
2?



-a
J
§>


1«J
S
3
•R



•<
j
H


e
a
S


ox
z
S"

i
X
0
Z
ox

g
ox
z
X
o
01



1
ox

ox
at

I
X
O
X
o

7
a.
a- J
2 -|

ii
r» yf
0
10 S"
« S,

-S

S 2
«s
§1
- ci
«!

^"^
0 g



«€
<0 <0
*'
o§



0 g
>o •»

if
"^
- S1
"* —
s|
o o"
r* r*-
"^




a

                               — ea
                               •s -i
                                         7-10

-------
     The magnitude of the secondary pollutants generated by the
operation of the control device systems is much smaller.than the
magnitude of solvent emissions being reduced.  In the worst case
scenario, 23 Mg/yr (25 tons/yr) of secondary pollutants are
generated from existing major sources as a result of fuel oil
combustion when either Alternative I or Alternative II is
implemented  (Table 7-6}.  The calculated HAP solvent emission
reduction from existing major sources, however, is 2,080 Mg/yr
(2,300 tons/yr) and 2,470 Mg/yr (2,720 tons/yr) when
Alternatives I and II, respectively, are implemented (Table 7-2).
7.2  WATER POLLUTION IMPACTS
     Wastewater impacts arise from the use of fixed-bed carbon
adsorbers that use steam for desorption.  In a fixed-bed carbon
adsorption system, water is used to produce steam, which is then
used to strip adsorbed solvent from the carbon beds.   (As
previously stated, it is assumed that 4 pounds of steam are used
to strip 1 pound of HAP solvent.)   Upon completion of the
stripping operation,  the solvent-steam vapors are condensed and
fed to a decanter where the water-insoluble organic layer
separates from the water and water-soluble organic layer.
Although the HAP solvents separate from the water phase in the
decanter, the HAP solvents used in the industry (methyl ethyl
ketone, methyl isobutyl ketone, and toluene) are not entirely
immiscible.  Therefore, some organic solvents are present in the
wastewater discharge from the decanter.  The water phase is,
therefore, usually then sent to a stripping column for further
separation.
     Data from three magnetic tape facilities report solvent
concentrations ranging from 5 to 500 ppm of total VOC,  some of
which are HAP, in the wastewater discharge from their stripping
columns.4  Based on typical stripper column design and the type
of compounds present in wastewater from magnetic tape
manufacturing facilities, the aqueous bottoms from the stripper
column contains up to 50 ppm HAP.5"7  This concentration is based
on the maximum outlet HAP concentration reported by magnetic tape
manufacturing facilities.  The wastewater discharges and
                               7-11

-------
associated waterborne HAP emission levels that would result from
each regulatory alternative are presented in Table 7-9 for
existing major sources.  The wastewater quantities are
incremental increases over current wastewater discharges for
facilities that currently use solvent recovery but would have to
control additional HAP emission sources to comply with the
regulatory alternatives.

     TABLE 7-9. ANNUAL INCREMENTAL WASTEWATER DISCHARGES AND
              EMISSIONS FROM EXISTING MAJOR SOURCES
Regulatory
Alternativea
I
II
Wastewater
discharges,
103 L/yr
(10J gal/yr)
5,600 (1,480)
5,600 (1,480)
HAP solvent emissions,
Kg/yr (Ib/yr)
280 (615)
230 (615)
  aFor a description of the regulatory alternatives,  refer to
   Table 7-1.
     Existing magnetic tape facilities using a fixed-bed carbon
adsorption system for VOC control may dispose of the solvent and
water phases that are generated from carbon desorption as
hazardous waste.  There are no wastewater discharges or
waterborne VOC emissions from these activities.  Magnetic tape
facilities that do not use fixed-bed carbon adsorption systems
for control are also assumed to have no additional wastewater
discharges resulting from implementation of the regulatory
alternatives.
     There are other operations conducted at magnetic tape
facilities that result in wastewater discharges.  These include
wastewatar from sparge units and molecular sieves.  Wastewater
discharges from these operations are not expected to increase as
a result of the regulatory alternatives.  Also, some facilities
operate condensers that have a water phase from moisture present
in the air.  However, in the case of facilities operating a
condenser, it was assumed that the condenser would not be used to
control any additional HAP emission points that require control
                               7-2.2

-------
under Alternatives I and II but are  not already controlled by the
facility.
     For model  lines, it was assumed that the medium and  large
lines currently have solvent recovery associated with an  existing
carbon adsorption system.  Thus, the above assumptions also apply
to the medium and large model lines.   Small lines are assumed to
dispose of  wastewater as hazardous waste.  The annual wastewater
discharge quantities that would result from each regulatory
alternative for medium and large model lines are presented in
Table 7-10.   The waterborne HAP emission levels for medium and
large model lines are also presented in Table 7-10.  As with the
existing facilities, the wastewater  discharges and waterborne HAP
emissions result from controlling HAP emission points that are
not currently controlled with an existing control device.
    TABLE  7-10.
ANNUAL INCREMENTAL WASTEWATER DISCHARGES AND
    EMISSIONS  FROM NEW LINES
Model
Line*
Small
Medium: A
B
Large: A
B
Regulatory Alternative Ib
Wastewater
discharge,
103 L/yr (103 gal/yr)
0(0)
24 (6.5)
0.15 (0.04)
250(67)
8 (2.1)
HAP solvent
emissions,
Kg/yr (to/yr)
0(0)
1.2 (2.7)
0.01 (0.02)
13 (28)
0.4 (0.9)
Regulatory Alternative II11
Wastewater
discharge,
103 L/yr (103 gal/yr)
0(0)
24 (6.5)
0.15 (0.04)
250(67)
8 (2.1)
HAP solvent
emissions,
Kg/yr (lb/yr)
0(0)
1.2 (2.7)
0.01 (0.02)
13 (28)
0.4 (0.9)
  aLine A is a new line that was not built concurrently with a control device. Line B is a new line that was built
  concurrently
  .with a control device.
  kpor a description of the regulatory alternatives, refer to Table 7-1.

7.3  SOLID  WASTE  IMPACTS
     The only solid waste impacts  from the  add-on control  systems
come from carbon  adsorption units.   Solid waste impacts resulting
from either of the regulatory alternatives  are only  those  impacts
that are in addition to the solid  waste currently generated at  a
plant.

-------
     The activated carbon in carbon-adsorption units gradually
degrades during normal operation.  The efficiency of the carbon
eventually drops to a level such that replacement is necessary,
thereby creating a solid waste load.  The average carbon life is
estimated to be from 6 months to 5 years for fixed-bed carbon
adsorbers and from 3 to 5 years for fluidized-bed carbon
adsorbers.8"10  For existing facilities that are currently
operating a carbon adsorption system, it is assumed that the
control of the additional sources not currently controlled at the
plant that require control under the regulatory alternatives
would not decrease the carbon life.  Thus, there are no
incremental solid waste impacts from these sources.  Solid waste
will result from existing sources that require new carbon
adsorption systems as a result of the regulatory alternatives.
This is the case for one existing major source.11  The annual
solid waste impact from the source is 70 kg (160 Ib).11
     All model lines except for the small model line are assumed
to currently use fixed-bed carbon adsorption systems for VOC
control.  As with existing facilities, it is assumed that the
control of additional emission points would not decrease the
carbon life.  Thus, there are no solid waste impacts from medium
and large model lines.-  The small model line would require a
fixed-bed carbon adsorption system to meet Alternatives I and II.
The annual solid waste impact from the small model line is
70 kg  (160 Ib)-11
     Three alternatives are available for handling the waste
carbon material:   (1) landfilling the carbon,  (2) reactivating
the carbon and reusing it in the adsorber, and  (3) using the
carbon as fuel.  Landfilling is a common practice and no
environmental problems would occur if the landfill site has been
properly constructed. If the site is not secured by a lining of
some type (either natural or artificial), possible soil leaching
could occur.  The leachate may contain traces of organics which
have been left on the carbon as residue.  Transmission of this
leachate into ground and surface waters would represent a
potential environmental impact.  Even if no problems exist at a
                               7-14

-------
landfill, however, it is desirable to reduce the quantity of
wastes sent to landfills whenever possible.-
     The second and most common alternative for handling the
waste carbon material does not create any significant amount of
solid waste.  Most of the carbon is reactivated and reused in the
carbon adsorber.  Disposal of waste carbon represents only 5 to
10 percent of the carbon used.  Disposal of this waste by
landfilling poses minimal environmental problems provided the
landfill site is properly constructed.
     The third method involves selling the waste carbon as a
fuel.  The physical and chemical structure of the carbon in
combination with the hydrocarbon residues make the waste a fuel
product similar to other solid fuels such as coal.  Potential
users of this fuel include industrial and small utility boilers.
Because activated carbon generally contains very little sulfur,
furnace 862 emissions resulting from combustion would be
negligible.  Particulate and NOX emissions from the burning of
activated carbon would be comparable to those of coal-fired
operations.  The use of this disposal method, however, would be.
limited because of the small quantities of carbon generated by
facilities in this industry.
7.4  ENERGY IMPACTS
     The air emission control equipment for the magnetic tape
manufacturing industry utilizes two forms of energy:  electrical
            *
energy and fossil fuel energy.  Electrical energy is used in
carbon adsorber and incinerator control systems.  The electrical
energy is required to operate fans, cooling tower pumps and fans,
boiler support systems, and all control system instrumentation.
Fuel oil is used for atsam generation for fixed-bed carbon
adsorption units, and natural gas is used for supplemental fuel
in incineration units.  Electrical energy and steam are also
required for the distillation systems used to separate and purify
recovered solvents at those existing facilities that currently
have onsite solvent recovery.  It is assumed that medium and
large model lines perform onsite solvent recovery.  As with the
other environmental impacts, the energy impacts presented here

-------
are incremental.  That is, the  impacts  are in addition to the
current energy used at existing sources and model lines.
     The total energy demands for  existing major sources that
would result from implementing  the regulatory alternatives are
presented in Table 7-11.  The energy demands include the
additional natural gas required by those facilities currently
using incineration, the additional steam required by existing
carbon adsorption system, and the  additional electricity
requirements associated with control device operation as well as
operation of the ventilation fan required for controlling
particulate HAP emissions.

      TABLE  7-11.   ANNUAL INCREMENTAL ENERGY REQUIREMENTS  OF
            REGULATORY ALTERNATIVES--EXISTING SOURCES
Regulatory
Alternative*
I
n
Natural gas
requirements,
GJ/(106Btu/yr)
70(65)
70(65)
Steam
requirements,
GJ/yr (106 Btu/yr)
20,160(19,125)
20,160(19,125)
Electricity
requirements,
GJ/yr (106 Btu/yr)
600 (570)
600 (570)
Total energy
requirements,
GJ/yr (106 Btu/yr)
20,830 (19,760)
20,830 (19,760)
  aFor a description of the regulatory alternatives for existing sources, refer to Table 7-1.

     For new lines, there  is no natural  gas demand because
fixed-bed carbon adsorption is  the  assumed control method.  The
additional steam requirements of  new lines that would result from
implementing the regulatory alternatives are presented in
Table 7-12.  The additional electricity  requirements for new
lines are provided in Table 7-13, with the total energy
requirements of new lines  presented in Table 7-14.  The emission
points that are not currently controlled but would require
control under Regulatory Alternatives I  and II  include storage
tanks, mix preparation, waste handling,  and equipment leaks.  As
explained in Section 7.1,  Alternatives I and II have the same
impacts because the only, higher level control  options of
Alternative II are the  LDAR program and  closed containers for
cleaning solvents.  These  have  no energy requirements.
                               7-1S

-------
      TABLE 7-12.   ANNUAL INCREMENTAL  STEAM REQUIREMENTS OF
                  REGULATORY ALTERNATIVES--NEW LINES
Regulatory
Alternative2
I
H
Steam requirements, GJ/yr (106 Btu/yr)
Small line
200 (189)
200 (189)
Medium lines"
A
790 (750)
790 (750)
B
0.3 (0.3)
0.3 (0.3)
Large lines''
A
1,360 (1,290)
1,360 (1,290)
B
32 (30)
32 (30)
aFor a description of the regulatory alternatives for new sources, refer to Table 7-1.
"Line A is a new line that was not built concurrently with a control device.  Line B is a new line that was
 built concurrently with a control device.
   TABLE  7-13.   ANNUAL  INCREMENTAL ELECTRICITY REQUIREMENTS
                OF  REGULATORY  ALTERNATIVES--NEW LINES
Regulatory
Alternative3
I
n
Electricity requirements, GJ/yr (106 Btu/yr)
Small line
17 (16)
17 (16)
Medium lines"
A
3.2 (3)
3.2 (3)
B
1.6 (1.5)
1.6(1.5)
Large lines'1
A
3.2 (3)
3.2 (3)
B
1.6 (1.5)
1.6 (1.5)
aFor a description of the regulatory alternatives for new sources, refer to Table 7-1.
"Line A is a new line that was not built concurrently with a control device.  Line B is a new line that
 built concurrently with a control device.
was
                                     7-17

-------
  TABLE 7-14.
 TOTAL ANNUAL INCREMENTAL- ENERGY REQUIREMENTS OF
REGULATORY ALTERNATIVES--NEW LINES
Regulatory
Alternative*
I
n
Total energy requirements, GJ/yr (106 Btu/yr)
Small line
217 (205)
217 (205)
Medium lines
A
793 (753)
793 (753)
B
1.9 (1.8)
1.9 (1.8)
Large lines'3
A
1,365 (1,295)
1,365 (1,295)
B
32 (32)
32 (32)
  aFor a description of the regulatory alternatives for new sources, refer to Table 7-1.
  "Line A is a new line that was not built concurrently with a control device.  Line B is a new line that was
   built concurrently with a control device.
7.5  OTHER ENVIRONMENTAL IMPACTS
     The impact, of  increased noise levels  is  not  a significant
problem with  the emission control systems  used in the magnetic
tape coating  industry.   No noticeable increases in noise levels
occur as a result of increasingly stricter regulatory
alternatives.   Fans and motors, present in the majority of the
systems, are  responsible for the majority  of  the  noise in the
control operations.
7.6  OTHER ENVIRONMENTAL CONCERNS
7.6.1  Irreversible and Irretrievable Commitment  of Resources
     As discussed in Section 7.4, implementing the regulatory
alternatives  will result in an increase in the irreversible and
irretrievable commitment of natural resources.   This increased
energy demand for pollution control equipment,  however, is
insignificant compared to the total energy demand of a magnetic
tape facility.
7.6.2  Environmental Impact of Delayed Standard
     Because  the water pollution, solid waste,  and energy impacts
are small, there is no significant benefit to be  obtained from
delaying the  proposed standard.  Furthermore,  there does not
appear to be  any emerging emission control technology that
achieves greater HAP emission reduction or achieves a HAP
emission reduction  equal to that of the regulatory alternatives
at a lower cost than those represented by  the control devices
                                7-13

-------
considered here.  Consequently, there are no benefits or

advantages to delaying the proposed standard.

7.7  REFERENCES FOR CHAPTER 7

 1. U. S. Environmental Protection Agency.  Compilation of Air
    Pollutant Emission Factors (AP-42).  Fourth Edition.
    September 1991.  p. 1.4-2

 2. Background Information Document for the New Source
    Performance Standard for the Magnetic Tape Manufacturing
    Industry.  U. S. Environmental Protection Agency.  Research
    Triangle Park, NC.  Publication No. EPA-450/385-029a.
    December 1985.

 3. Reference 1.  October 1986.  p. 1.3-2.

 4. Memorandum from Angyal, S., MRI, to magnetic tape NESHAP
    project file.  November 9, 1992.  Summary of confidential and
    nonconfidential information from magnetic tape manufacturing
    facilities.

 5. Telecon.  Thompson, L., MRI,  with Vaughan, B., Ampex
    Corporation.  July 26, 1993.   Information on steam strippers.

 6. Telecon.  Thompson, L., MRI,  with Farmer, M., Sony Magnetic
    Products.  July 29, 1993.  Information on steam strippers.

 7. Telecon.  Angyal, S., MRI, with Garner, G., 3M Environmental
    Engineering and Pollution Control.  September 16, 1993.
    Information on steam strippers.

 8. Lee, J., 3M, to Farmer, J., EPAiESD.  October 24, 1983.
    Response to Section 114 information request for the 3M
    facility in Camarillo, California.

 9. Telecon.  Glanville,  J., MRI,  with Mason, J., Union Carbide.
    September 10, 1984.  Information on fluidized-bed carbon
    adsorbers.

10. Telecon.  Angyal, S., MRI, with Cowles, H., Environmental
    C & C.  October 20, 1992.  Information on fluidized-bed
    carbon adsorbers.

11. Memorandum and attachments from McManus, S., MRI, to magnetic
    tape NESHAP project file.  February 1, 1993.  Documentation
    of cost impacts estimating methodology.

-------
                            8.0   COSTS

8.1  INTRODUCTION
     The estimated cost impacts of implementing the regulatory
alternatives presented in Chapter 6 for existing facilities and
model lines are provided in this chapter.  This chapter also
contains a. summary of the methods and assumptions used to
calculate costs.1  Appendix F to this chapter contains example
calculations for some of the more complicated costing procedures.
A discussion of the economic impact of the regulatory
alternatives on magnetic tape manufacturers is presented in
Chapter 9.
     In this chapter, the model lines for which cost impacts have
been calculated are briefly summarized.  A detailed discussion of
model lines is provided in Chapter 6.  An explanation of the
methodology used in the cost analysis follows.  Requirements and
costs for installing and operating control equipment, compliance
monitoring, and reporting and recordkeeping for new and existing
sources are discussed.  Except where noted, all costs are
reported in first quarter 1992 dollars.
8.2  MODEL LINE PARAMETERS
     Five model lines representing three line sizes--small,
medium, and large--have been selected to characterise the
manufacturing lines expected to be constructed, modified, or
reconstructed in the future.  A detailed discussion of model
lines is provided in Chapter 6.  The three line sizes are
analogous to the research, small, and typical model lines
developed for the new source performance standard (NSPS) for the
magnetic tape industry, which were categorized by the major
design parameters of production rate, hours of operation, coating
                               a-i

-------
solvent content, and coating thickness.2  (Designations have been
changed from research, small, and typical to small, medium, and
large, respectively, to avoid confusion with a research facility
as defined in Section 112(c)(7)  of the Clean Air Act.)  From
these sizes, five NESHAP model lines have been established.
These are (1) a small line;  (2)  a medium line built without
concurrent construction of a VOC control device;  (3) a medium
line built with concurrent construction of a VOC control device;
(4) a large line built without concurrent construction of a VOC
control device; and (5)  a large line built with concurrent
construction of a VOC control device.  It is necessary to
distinguish between the situation when the plant concurrently
constructs a VOC control device with the new coating line and
when the plant uses an existing control device with a new coating
line because the requirements of the NSPS, and hence baseline
emissions, are different for each case.
8.3  CONTROL COST ESTIMATING METHODOLOGY
     The total annual cost of a control technique is the sum of
the direct costs and indirect costs less any recovery credits.
Indirect costs are based on the total capital investment (TCI)
required to implement a control option; therefore, the
calculation of the TCI for each control option is discussed
first.  The total annual cost calculations follow.  Costs to
control equipment leaks in piping are discussed in a separate
section due to the uniqueness of the costing methodology used for
this emission point.  Control options for each emission point are
summarized in Table 8-1.  The regulatory alternatives considered
for the NESHAP for existing facilities and model lines are
summarized in Table 8-2.
     To determine total annual costs, five scenarios were defined
to describe the current level of control and to identify what
controls are needed under the NESHAP.  These scenarios encompass
every situation found in industry and the model lines and are
described as follows:
                               3-2

-------
    TABLE 8-1.   CONTROL  OPTIONS FOR HAZARDOUS AIR  POLLUTANT
     EMISSIONS  FROM MAGNETIC  TAPE  MANUFACTURING FACILITIES
Emission point
Control
Overall HAP
control,
percent*
COATING OPERATIONS
Storage tanks
Mix preparation
Coating application/drying
Waste handling
Piping leaks (plantwide)
Packing/labeling
Particulates
Vent to 95% efficient control device
Cover/vent mix vessels to 95% efficient control
device
Total enclosure vented to 95 % efficient control
device
Direct the waste handling device process vent to
a 95% efficient control device.
1. No control
2. Implement LDAR program
No control
Enclosed transfer
95
95
95
95
0
89°
0
99. 9b
CLEANING ACnvmES
Tank cleaning
Removable parts
Fixed exterior surfaces
Flushing of fixed lines
All cleaning activities
No control
75% Freeboard ratio or equivalent through
venting to control
No control
Closed system or vent supply and waste
collection vessels to 95% efficient control device
1. No control
2. Closed containers for dispensing solvent; lids
on solvent collection containers when not in
use.
0
88b
0
95b
n/a
aThe control efficiency indicated is the control for that specific emission source, not the entire facility.
 Engineering estimate.
                                   3-3

-------
CO
o
cm

S
o
H
CO
CO
H
  CO
CO
  u
a s

Is
*8
S H
IB
§3
ga
o
a
CN

CO

a
j
    a

II
         fi
II
            oo
 I
         u
           u
           an
                  8
              O
              u

              o
     a

     O
     Z
              11
                1
              I
                   u

              o g ^ \n
              H 5 > os
                     '>
              i
              5J
              o
              CJ
          s
        O •**
        sg S
        0 ^3
     o\

     2g^
                 s-s
                   o
                   Z
                            o
                             1
                   I
                   SI
                           .
                   3 ~ B o* to a >
                   s2 S 2 ai 8-8
                   §5«

                   11 -8.8
                   Sill
                   -.228
                   Ilil
                          3-4

-------
     1.  The facility has an existing carbon adsorber that can be
used for controlling additional emission points required by the
NESHAP;
     2.  The facility has an existing incinerator that can be
used for controlling additional emission points required by the
NESHAP;
     3.  The facility does not have a control device and will
need a new control device to meet the requirements of the
regulatory alternatives;
     4.  The facility only uses hazardous air pollutant  (HAP)
solvent for cleaning and thus only has to meet requirements
pertaining to cleaning; and
     5.  The facility currently meets all MACT floor requirements
with the exception of controlling equipment leak emissions and
the use of closed containers for cleaning.
If a facility or model line has an existing control device, it is
assumed that it has adequate capacity to control additional
emission points.
     The basic cost estimating methodology explained in this
section is the same for each scenario.  The combination of
control options required, however, is different for each.  Sample
calculations showing how the costing methodology was used for
each scenario are presented in Appendix F.
     Because newly constructed, modified, or reconstructed lines
are subject to the NSPS, the NESHAP model lines representing new
facilities use the control devices or methods required by the
NSPS.  Costs of additional controls required by the NESHAP for
new lines were estimated using the same methodology used for
existing facilities.  The small model line is described by
Scenario 3, as discussed above.  All other model lines are
described by Scenario 1.
8.3.1  Total Capital Investment
     The total capital investment (TCI)  associated with each
control option has been estimated to determine the impacts of
controlling hazardous air pollutant (HAP) emissions from: solvent
storage, mix preparation, coating application and drying, waste
                               3-5

-------
handling, particulate transfer, cleaning activities, and
equipment leaks from piping.   (The total annual cost to control
equipment leaks is discussed in Section 8.5.)  The methodology
used to calculate the TCI for controlling individual emission
points is the same for the five scenarios identified.  The TCI
will differ for the facilities described by each scenario
depending on the current level of HAP emission control at each
particular facility.
     Many of the control techniques being considered for the
NESHAP were also considered in the NSPS analysis.^'^
Consequently, control costs developed for the NSPS have been used
in this analysis to the greatest extent possible.  Control costs
developed for the NSPS were developed for each model line
size--small, medium, and large.  Therefore, for this cost
analysis, existing facilities were classified as small, medium,
or large based on the total solvent usage  (HAP plus non-HAP
solvent) per line.  Control costs identified for line sizes
defined in the NSPS could then be applied to existing facilities
to determine control cost impacts if no plant-specific data were
available.  Unless a facility was known to have a unique line or
lines (i.e., research and development), the facility
classification was based on total aggregate solvent usage divided
by the number of lines.
     As described in Chapter 6, the total solvent usage
associated with model lines (as determined by the NSPS) is
6800 gallons per year  (gal/yr) for a small line, 20,445 gal/yr
for a medium line, and 204,445 gal/yr for a large line.
Therefore, if a plant uses 800,000 gal/yr of solvent and has four
coating lines, this line has an average coating usage of
200,000 gal per line.  Thus, the facility falls under the "large
line" classification.  Control costs developed for the NSPS for
large lines would be used for this facility.
     8.3.1.1  Solvent Storage Tanks.  The MACT floor level of
control for storage tanks is to capture and control 95 percent of
HAP emissions.  For the purposes of the cost analysis, it is
                               3-6

-------
assumed that this level of control will be achieved by venting
storage tank emissions to a 95 percent efficient control device.4
     This control technique was also considered as part of the
NSPS analysis.  Therefore, the TCI required to control emissions
from storage tanks was estimated for existing facilities and
model lines using cost data from the NSPS.2'3  The TCI estimated
for the NSPS analysis included labor and material costs to
install ductwork from the tanks to the control device.  It is
assumed that the existing control device  (or the control device
that will be built to meet the MACT floor for coating'
application/drying emissions) will be used.  Thus, the TCI to
vent storage tanks does not include the cost of a new control
device.
     The costs used in the NSPS are expressed in March 1983
dollars.  For the NESHAP analysis, costs were updated from
March 1983 to March 1992 using cost indices for fabricated
equipment from Chemical Engineering magazine. '^  The March 1983
index is 326.8; the March 1992 index is 362.2.
     The number of storage tanks associated with the small and
medium model lines developed for the NESHAP analysis is the same
as the number of tanks assumed for a medium line in the NSPS
analysis.  Therefore, the TCI to control storage tank emissions
for small- and medium-sized NESHAP model lines is simply the
costs used in the NSPS adjusted for inflation as described above.
The TCI for existing facilities was further adjusted to account
for the retrofit costs associated with installing equipment in
existing facilities  (the TCI is multiplied by a retrofit factor
of 1.4).7  The calculation of the TCI for small and medium model
lines is presented below:
Small and medium size lines:
TCI for new small and medium size NESHAP lines:
     $24,948 * 362.2 / 326.8 - $27,650
TCI for existing small and medium size NESHAP lines:
     $27,650 * 1.4 =. $38,710
     The TCI calculation for the large NESHAP model line is
somewhat more complicated because the assumed number of tanks per
                               a-7

-------
line, six, is different than the three tanks assumed for the
corresponding large NSPS line.  The assumed numbers of tanks for
NSPS and NESHAP model lines are presented in Table 8-3, along
with the cost estimates developed for the NSPS.  A "cost per
tank" was developed by dividing the difference in the number of
tanks between the small and medium NSPS model lines by the
corresponding difference in the TCI.  This "cost per tank" was
multiplied by the number of additional tanks assumed for the
large NESHAP model line (6 - 3 = 3) to estimate the additional
capital investment to control the large NESHAP model line.  The
calculation is presented below:

           TABLE 8-3.  MAGNETIC TAPE NSPS AND NESHAP--
       ASSUMED NUMBER OF STORAGE TANKS; NSPS TOTAL CAPITAL
                            INVESTMENT
NSPS
Model
Small
line sizea
(Research)
Medium (Small)
Large
(Typical)
No. of
tanks
5
3
3
TCI required
with existing
control
device
$31,389
$24,943
$24,948
NESHAP
Model No. of
line size tanks
Small
Medium
Large
3
3
6
  aFor the NESHAP,  model line designations (research,  small,
   and typical)  were changed to small,  medium,  and large,
   respectively.
Large size lines:
Difference in no. of tanks between small and medium NSPS lines:
     5-3-2
Difference in TCI between small and medium NSPS lines:
     $31,390 - $24,950 = $6,440
Incremental cost per tank:
     $6,440 / 2 - $3,220
Difference in no. of tanks between medium and large NSPS line:
     6-3-3
                               3-3

-------
Additional capital investment for new large NESHAP line:
     $3,220 * 3 * 362.2 / 326.8 » $10,710
TCI for new large NESHAP line (equals TCI for new small or medium
line, plus additional capital investment):
     $10,710 + $27,650 - $38,360
TCI for existing large NESHAP line:
     $38,360 * 1.4 - $53,700
     Table 8-4 summarizes costs developed for the NESHAP to
estimate the TCI associated with controlling storage tank
emissions for existing facilities and model lines.

                TABLE 8-4.   NESHAP  COST ANALYSIS--
       TOTAL  CAPITAL  INVESTMENT  FOR CONTROL  OF  STORAGE  TANK
              EMISSIONS WITH A COMMON CONTROL DEVICE
Model line
Small
Mediuma
Largea
No. of storage
tanks
3
3
6
Total capital
investment for
existing
facilities13
$38,710
$38,710
$53,700
Total capital
investment for
model lines*3
$27,650
$27,650
$38,360
  alncludes lines built without concurrent construction of a
   control device and lines built with concurrent construction
  of a
   control device.
  bCosts reported in first quarter 1992 dollars.
     8.3.1.2  Mix Preparation.  Mix preparation refers to the
process of mixing the coating ingredients  (solvent, magnetic
particles, resins, etc.) to formulate the magnetic coating.  Of
these ingredients, the solvent and the magnetic particles may be
HAP's.  The cost associated with particulate control is discussed
in Section 8.2.1.5.
     The MACT floor level of control for this emission point is
capture and control of 95 percent of mix preparation emissions.
This, level of control is based on covering all mix vessels and
venting the vessels to a 95 percent efficient control device.4
                               3-9

-------
This control requirement excludes pressurized vessels such as
mills because no air emissions are expected from those vessels.
     This control technique was also considered for the NSPS, and
the TCI estimates were derived from those developed for the
NSPS. ' ^  The number and types of mix equipment needed to support
one coating line are identical for NESHAP and NSPS model lines.
Thus, the TCI estimates used for the NESHAP are essentially the
same as those developed for the NSPS with adjustments for
inflation (for both new and existing lines) and retrofitting  (for
existing lines only).   The inflation and retrofit factors cited
above for storage tanks were also used to update the TCI estimate
for mix preparation controls.
     The TCI represents the total expense involved in installing
ductwork from the mix equipment to the SLA duct, including labor,
materials, overhead, taxes, and administration.  As with the
storage tanks, it is assumed that the existing control device  (or
the control device that will be built to meet the MACT floor for
coating application/drying emissions), will be used by existing
facilities and model lines.  Thus, the TCI to vent mix equipment
does not include the cost of a new control device.
     As previously stated, the TCI associated with controlling
mix preparation equipment that was used in the NESHAP analysis
are essentially the NSPS costs adjusted for inflation and
retrofitting.  As indicated in Table 8-5, the TCI per coating
line associated with new and existing small lines is $3,900 and
$5,400, respectively.   The TCI associated with controlling mix
preparation emissions for existing medium and large size lines is
$7,400; for new medium and large size lines, the cost is $5,300.
Because the model line parameters and associated costs are Che
costs to control mix equipment supporting one coating line, the
TCI for existing facilities was multiplied by the number of
coating lines at a facility to determine the TCI for the entire
facility.
     8.3.1.3  Coating Application/Drying.  Coating application/
drying refers to the process in which the substrate is coated and
the solvent is evaporated from the coating mix.  The MACT floor
                               8-10

-------
level of control for this emission point is capture and control
of 95 percent of emissions.  The assumed method of meeting the
MACT floor for coating application/drying is to install a total
enclosure meeting EPA criteria around the coating and drying
processes, and vent the enclosure to a carbon adsorber.4

                TABLE  8-5.   NESHAP  COST ANALYSIS--
       TOTAL  CAPITAL INVESTMENT TO  CONTROL MIX  PREPARATION
              EQUIPMENT WITH A COMMON CONTROL DEVICE
          Size designation
 TCI'
  Small
       New
       Existing
  Medium
       New
       Existing
  Large
       New
       Existing
$3,900
$5,400
$5,300
$7,400
$5,300
$7,400
  aAll costs are reported in 1992 dollars.

     The TCI associated with controlling an uncontrolled line at
a magnetic tape facility is the cost of equipment, materials and
labor to install a carbon adsorber, a total enclosure, and the
necessary ductwork.  The TCI for individual facilities depend on
the control equipment that currently exists at a facility.  For
example, some facilities currently vent coating/drying emissions
to an existing control device with a partial enclosure; the TCI
for these facilities is just the cost to install a total
enclosure.  Some other facilities have several coating lines but
only control emissions from a portion of their lines.  The TCI
for these facilities is the cost of installing total enclosures
and ductwork to vent emissions from uncontrolled lines to the
existing control device.
     This control option was also considered for the NSPS;
therefore, TCI estimates from the NSPS were adjusted for
3                                 -7 i
                                ~" ~Lo.

-------
inflation and. retrofitting using the same factors used to adjust
mix preparation and storage tank control costs for the NESHAP
analysis.2'3'5'6'8  The TCI estimates from the NSPS analysis for
total enclosures are presented in Table 8-6 along with costs
calculated for the NESHAP.  The figures in Table 8-6 represent
the TCI per coating line for facilities with existing control
devices that currently vent coating application and drying
emissions to a control device but do not have total enclosures.
Cost calculations used for the NSPS were used in the NESHAP
analysis to calculate the TCI for uncontrolled facilities and
facilities that have a control device but do not control all of
their coating lines.

      TABLE  8-6.   TOTAL  CAPITAL  INVESTMENT--TOTAL ENCLOSURE
              FOR COATING APPLICATION/DRYING  PROCESS
Model line
size
Small
Medium
Large
NSPS TCI for a
total
enclosure, $a
13,500
13,500
15,000
TCI for an
existing
NESHAP^ line,
$b
20,900
20,900
23,300
TCI for a
NESHAP model
line, $B
14,900
1,4900
16,600
  5J1983 dollars,
  °1992 dollars.
     8.3.1.4  Waste Handling.  Waste handling refers to the
processing of solvent-laden waste, such as cleaning rags and
filters from various operations, to remove and recover the
solvent.  The solvent is evaporated from the waste and condensed;
noncondensible gases  (along with small amounts of solvent vapor)
are vented either to the control device or to the atmosphere.
The MACT floor level of control for reducing waste handling
emissions is capture and control of 95 percent of the emissions
from the waste handling unit process vent.  This control level is
based on directing the process vent to a 95 percent efficient
control device.4
                               3-12

-------
     This control option affects only one facility; all other
facilities performing onsite waste handling achieve the control
efficiency required by the MACT floor.  That facility provided an
estimate of the TCI that would be required if it vents its waste
handling process vents to the solvent-laden airstream (SLA)
header leading to the facility's existing carbon adsorber.  The
cost provided by the facility is based on a previous project that
involved venting tanks to the SLA header.  This cost was compared
to costs calculated for controlling storage tank emissions from a
large facility, a project similar in scope, and was found to be
within approximately 30 percent of that cost.
     8.3.1.5  Particulate Emissions in Mix Room.  Particulate
emissions result from the addition of magnetic particles
containing chromium or cobalt to the coating mix.  The MACT floor
control level is 99.9 percent control of particulate emissions,
which can be accomplished by making particulate transfer a closed
process.  This control method results in a 94 percent reduction
in particulate emissions from the baseline level.4
     Two facilities with enclosed transfer systems provided
estimates of the TCI required for enclosed systems; the estimates
are both of the same order of magnitude.  Although the two
systems are not identical, both include features such as
ductwork, a dust collector, and a fan.  The average of the two
cost estimates is $80,000, which was used to represent the TCI
required to implement this control option.1  It is assumed that a
particulate transfer device will be needed for each coating line
at a facility, because facilities with more than one line may
perform mixing simultaneously.  Also, particulate transfer
equipment in existing facilities is fixed, and is integral to mix
equipment.  Therefore, the TCI for a plant is $80,000 multiplied
by the number of coating lines.
     8.3.1.6  Removable Parts Cleaning.  Removable parts cleaning
refers to the cleaning of small removable parts such as pump
parts, coater parts, filter housings, and mix blades.  Cleaning
takes place in wash sinks or tanks filled with solvent.
                               8-13

-------
     The MACT floor for controlling emissions from removable
parts cleaning is 88 percent control, which can be accomplished
by maintaining a 75 percent freeboard ratio.4  Freeboard ratio is
the ratio of the vertical distance from the evaporative area to
the top of the sink (freeboard area)  divided by the smaller of
the length or width of the sink evaporative area.
     The TCI associated with installation of a sink designed to
provide a 75 percent freeboard ratio has previously been
calculated as $200.**  This estimate is expressed in 1976 dollars;
to obtain a current estimate of the TCI, this figure was adjusted
to first quarter 1992 dollars using cost indices for fabricated
equipment in 1976 (192.1--average for the year)  and March 1992
(362.2),5'6'8  The 1.4 retrofit factor was also applied to the
cost for an existing facility.7  The updated TCI for a sink that
can accommodate a 75 percent freeboard ratio is $470 per sink for
an existing facility and $340 per sink for a new facility.  The
TCI is less for a new facility because it does not include the
retrofit factor.
     8.3.1.7  Use of Closed Containers for All Cleaning
Activities, and Lids on Collection Containers When Not in Use.
This control option is considered as a work practice to further
reduce emissions from cleaning activities.  The cost incurred by
a facility would be the cost to purchase closed containers.  An
example of a closed container is a spring-loaded can that
delivers solvent to a mesh surface at the top only when the
cleaning rag is pressed to the mesh.   This reduces the surface
area of solvent exposed to the atmosphere and the amount of
solvent delivered to the rag.  A cost of $40 for a typical
1-gallon plunger can was obtained from an industrial equipment
vendor catalog.^  It was assumed that 10 of these containers
would be needed for each facility.  Therefore, the TCI associated
with this control is $400.
     8.3.1.8  Flushing of Fixed Lines.  Fixed piping carries
coating from the coating mix tank(s)  to the coater.  These lines
are flushed with solvent periodically to remove buildup of
coating mix.
                               3-14

-------
     The MACT floor for controlling emissions from fixed line
flushing is capture and control of 95 percent of the emissions
from this operation.4  This control level is based on a closed
operation, or venting solvent supply and collection vessels to a
95 percent efficient control device.  All major sources already
meet this level of control and it is assumed that new lines will
also have this level of control.  Thus, there are no capital
costs associated with this control option.
     8.3.1.9  Summary.  The TCI estimates described above were
applied to existing facilities depending on the existing level of
control.  Some existing facilities have more than one coating
line.  Some controls, such as total enclosures for coating
operations, are required for each line in the facility.
Similarly, many facilities have more than 1 wash sink, and
control of emissions from removable parts cleaning is required
for each sink.  Consequently, the TCI associated with these
controls is different for each facility depending on the number
of coating lines, wash sinks, etc.
8.3.2  Total Annual Control Costs
     This section describes the methodology used to calculate the
total annual costs associated with each control option for
existing facilities and model lines.  As discussed previously,
the total annual cost of a control option includes direct costs
and indirect costs.  Credit for reduction in solvent usage due to
solvent recovery was also considered, where applicable, and is
discussed as part of direct costs.  Each of these cost elements
is discussed in this section.  The calculation of direct costs is
discussed first, followed by discussions of the methods used to
calculate indirect costs and recovery credits.  Appendix F
provides sample calculations of total annual costs related to
control device operation for Scenarios 1 through 5.
     8.3.2.1  Direct Costs.  Direct costs include labor,
maintenance, fuel, and raw materials.  These costs were
calculated using the guidelines specified in the OAQPS Cost
Manual and methods used in the NSPS.2'3'7  Direct costs were
calculated for the following emission points:  mix preparation,
                               3-15

-------
coating application/drying, waste handling, equipment leaks, and
transfer of magnetic particles.  As discussed later in this
section, direct costs were not calculated for control of storage
tank emissions, use of a freeboard ratio, or use of closed
containers.  The model parameters and unit costs used to estimate
direct costs are summarized in Tables 8-7 and 8-8.  Model
parameters are from the NSPS; unit costs are 1992 costs.2'10"17
The model parameters were used for both existing sources and
model lines.  An existing source's size was based on the coating
usage per line, and parameters for that line size were then
applied.
     Direct costs to control emissions in the mix preparation,
coating application/drying, and waste handling areas are related
to the operation of the control device.  Direct costs for
emission points that are controlled by venting emissions to a
control device are a function of the additional HAP solvent
loading to the control device, additional airflow, frequency of
the operation, and HAP solvent concentration.3'7  These costs
include maintenance and operating labor, utilities (water, steam,
electricity, natural gas),  and raw materials (for example,
replacement carbon for carbon adsorbers).  For facilities that
currently operate a control device, the direct costs are
incremental costs; i.e., in addition to the current annual
operating expenses associated with the device.  Calculation of
direct cost components for control devices is summarized for
carbon adsorbers in Table 8-9 and for thermal incinerators in
Table 8-10.3'7'ia/19  The calculation of natural gas demand for
incinerators is presented in Table' 8-II.3
     The equations, factors, and percentages used to calculate
these costs are from the OAQPS Cost Manual and the NSPS cost
analysis.3'7  As Tables 8-9 and 8-10 show, operating labor and
maintenance are a function of the additional time needed to
operate and maintain the control device, and to supervise the
workers.  If a facility already has a control device, it is
assumed that the facility also has the operating, maintenance,
and supervisory labor necessary to operate it.  Therefore,
                               3-16

-------
    TABLE 8-7.   MODEL PARAMETERS USED IN TOTAL ANNUAL COST
                         METHODOLOGY
Emission point
Model flow rates ft^/min) :a
Mix preparation vessels
Total enclosures
Drying oven
Particulate control (flow
from mix vessels)
Carbon requirements. lb:a
Model operatina parameters:
Operating, hr/yr
Operating, d/yr
No. of shifts/d
Model line size
Small
13.2
300
600
13.2
784
2,000
250
1
Medium
19.8
300
600
19.8
784
6,000
250
3
Large
19.8
500
6,000
19.8
7,940
6,000
250
3
aModel flow rates and carbon requirements are from the NSPS
 cost analysis.
"A default value was not used for one existing facility with
 a small line.  This facility is identified in Section IV of
 the cost documentation memorandum.1
    TABLE 8-8.  UNIT COSTS USED  IN NESHAP ANALYSIS7'10'16
 Parameter
Unit cost, 1992  (quarter)
 Labor cost
 Electricity
 Steam
 Water
 Carbon
 Natural gas
$11.38/hour (2nd)
$0.049/kilowatt-hour  (1st)
$6/1,000 pounds  (2nd)
$0.936/1,000 gallons  (2nd)
$2.18/pound (2nd)
$2.77/1,000 cubic feet.  (1st)
                             3-17

-------
 TABLE 8-9.
FIXED  BED  ADSORBER--DIRECT AND  INDIRECT COSTS-
          EXAMPLE  CALCULATIONa

Direct costs
1 . Operating lab»r
Shifti per day
Hours per shift
Operating days/yr
Annual labor cost @ $11.38/hr
Supervisor (15% of labor cost)
2. Utilities
• Electricity: (_acfm)(5 hp/1,000 acfin)
(0.746 kW/hp)(_Jir/yr)($0.049/kWh)
• Steam: (4 Ib steam/lb HAP)
(__lb HAP/yr)($671,000 Ib steam)
• Water: (3.43 gal H2O/lb steam) (_lb
steam/yr) ($0.936/1,000 gal H2O)
• Initial carbon charge, Ib
•' Initial carbon cost, Cc ($2.18/lb)
• Carbon replacement (CRF (1.08 x Cc +
CcL»
4. Maintenance
• Labor (1 10% of operating labor)
• Materials (100% of maintenance labor)

5. Solvent credit/disposal cost
• Quantity disposed (4.75 x HAP loadings,
ton/yr)^
• Quantity recovered (0.95 x 0.9 x 	 HAP
loading, ton/vr)1
• Disposal cost ($69/ton)3'23'24
• Solvent credit ($372/ton)3'22
Indirect costs
1. Overhead (60% of operating and maintenance
labor and maintenance materials)
2. Administrative charges (2%)b
3. Property tax (l%)b
4. Insurance, (l%)b
5. Capital recovery cost. i=7%, n="10 years
Total Annual Cost:

Small

1
0.5
250
$1,423
$213
$219
$1,013
$542
784
$1,709
$460
$1,565
$1,565

100
N/A
$6,900
N/A

$2,860
$3,218
$1,609
$1,609
$22,913
$47,820
Line size
Medium

3
0.5
250
$4,268
$640
$658
$3,038
$1,626
784
$1,709
$460
$4,695
$4,695

300
N/A
$20,700
N/A

$8,579
$3,273
$1,636
$1,636
$23,302
$80,915

Large

3
0.5
250
$4,268
$640
$6,580
$30,394
$16,263
7,940
$17,309
$4,650
$4,695
$4,695

N/A
541
N/A
$201,396°

$8,579
$35,626
$17,813
$17,813
578,722
$46,650

Comments/Source of
Data

• Table 8-7
• OAQPS Cost
Manual7
• Table 8-7
• Table 8-8
• Table 8-8
The example calculation
assumes control of the
coating application/drying
emission point.
Capital recovery factor
(CRF)) is 0.2439 based
on a 7% interest rate and
5-year life. The cost of
labor for carbon
replacement, Cc^, is
$0.05/lb carbon.


The example is based on
HAP loading from the
coating application/drying
emission point.

The example is based on
the capital expenditure
for a total enclosure and
carbon adsorber at a new
plant.

*1992 dollars.
 Administrative charges, property tax, and insurance are calculated as a percentage of the total capital investment.
°Cost savings.
                                     3-18

-------
          TABLE  8-10.
INCINERATOR TOTAL ANNUAL COSTS-
EXAMPLE CALCULATION


Direct costs
1. Operatine labor
Shifts per day
Hours per shift
Labor (L): ($11.38/hr)
Supervisor (0.15)(L)
Total
2. Utilities - Assumes control of coating
application/drying emissions
• Electricity:
(0.049/kWh)1.7xl
-------
TABLE 8-11.
CALCULATION OF NATURAL GAS DEMAND FOR
      INCINERATORS
Calculation method: Assumed control of coating
application/drying. Refer to Table 8-7 for model parameters.
• Assume 35% heat recovery
• Temperature of gas entering incinerator: Tmc =
[0.35(1400 - 160)] + 160 = 594"F
• Mass flow rate of gas, lb/hr
[(_scfm)(60 min/hr)]/[13. 1 f^/lb air]
• Heat required, 106 Btu/hr
Enthalpy590 = 253.46
Enthalpy ^00 = 493.64
Q = (_lb/hr)(493.64 - 253.46)
Assuming 10% heat loss: Q1 = Q x 1.1
• Heat available from VOC, 106 Btu/hr
-25% LEL = 13 Btu/scf (13 Btu/scf)(_scfm)(60 min/hr)
• Heat to be supplied, 106 Btu/hr:
Q!f = Q1 - Q VOC
• Natural gas, 106 f^/yr:
[(_hr/yr)(_Btu/hr)]/[l,000 Btu/ft3]
• Natural gas cost, $:
(_ft3/yrx IQfix&TTVl.OOOft3)
Small


2,748
0.66
0.73
0.47
0.26
0.52
$1,440
Medium


2,748
0.66
0.73
0.47
0.26
1.56
$4,321
Large


27,481
6.6
7.3
4.7
2.6
15.6
$43,212
                       3-20

-------
facilities would not have to hire additional labor or
supervisors, nor would existing personnel need to work additional
hours, simply because additional emission points are being
controlled by the control device.  Thus, there are no operating
labor or maintenance costs to the facilities and model lines with
existing control devices.
     Utility costs are a function of HAP solvent loading and flow
rate to the control device.3'7  Utility requirements for
operation of a carbon adsorber include electricity, steam, and
water.  The calculation of carbon adsorber utility requirements
and costs are presented in Table 8-9.3/7/18/19  If a facility
currently operates a carbon adsorber, the utility requirements
are a result of the incremental increase in air flow and HAP
solvent loading to the device that would result from controlling
additional emission points.  Utility requirements for operation
of a thermal incinerator include electricity and natural gas.
The calculation of incinerator utility costs are shown in
Table 8-10; the calculation of natural gas requirements is shown
in Table a-11.3'7'18'19  As with carbon adsorbers, if a facility
currently operates a thermal incinerator and is simply
controlling additional emission points to comply with the NESHAP,
the annual utility costs are a function of the incremental
increase in air flow and HAP solvent loading.
     If a facility will require a new carbon adsorber to comply
with the NESHAP, one component of the direct cost is the
replacement carbon cost.  The calculation of carbon replacement
cost is shown in Table 8-9.7  If the facility currently operates
a carbon adsorber, it is assumed that the carbon will not degrade
any sooner due to controlling additional sources.  Therefore,
carbon replacement costs are the same as they would be in the
absence of a NESHAP; the incremental cost to the facility is
zero.  This assumption is supported by the fact that all existing
sources and model lines with carbon adsorbers use them to control
the coating application and drying operations, the largest
emission source.  The additional sources requiring control
                               3-21

-------
contribute much smaller quantities of emissions.  These points
include solvent storage, mix preparation, and waste handling.
     If a facility has an onsite distillation system, the costs
resulting from additional loading to this system were also
considered.  Operation of a distillation system requires cooling
water, electricity, and steam.3  When these costs were calculated
for the NSPS, electricity and cooling water costs combined
amounted to only 1 percent of the total cost.3  Therefore, for
the NESHAP only the annual cost of steam was calculated.  In the
NESHAP analysis, only those existing facilities and model lines
that currently operate a distillation system were assumed to
incur these costs.  The cost of steam is based on the additional
heat transfer requirements of the distillation device; the
additional heat transfer requirements are based on the solvent
loading.17   (A sample calculation of the cost of steam for
distillation is provided in Appendix F).  For the small existing
facility and the small model line, which would require new
control devices, it was assumed that the solvent/water mixture
from the carbon adsorber would be disposed as a hazardous waste.
     Direct costs for controlling emissions.from transfer of
magnetic particles are based on the option of transfer using an
enclosed bag slitting device.  This device typically has a
ventilation system that creates negative pressure inside the
slitter housing and pulls the particles into a filter.  The only
direct cost considered for this control option is the electricity
required to operate the fan.  Electricity requirements are based
on the same airflow rate used for the NSPS and NESHAP to
calculate carbon adsorber annual costs for mix equipment.
Airflow rates are presented in Table 8-7.  Electricity costs for
controlling particulate HAP's were calculated as shown in
Table 8-9 for carbon adsorbers.
     For this analysis, it was assumed that there are no direct
costs associated with controlling emissions from storage tanks.
Industrywide HAP solvent emissions from storage tanks were
calculated as 1.4 Mg/yr (1.6 tons/yr), with individual facility
storage tank emissions ranging between 0.004 and 0.8 Mg/yr  (0.004
                               8-22

-------
and 0.91 ton/yr).20  The airflow rate is estimated by one
facility as 0.2 ft3/min per storage tank.21  Utility costs are a
function of HAP solvent loading and airflow rate.  Therefore,
utility costs to control storage tank emissions with an existing
control device would be negligible compared to current utility
costs.  Maintenance and operating labor are assumed to be
negligible because the control device operating hours would not
be affected by controlling emissions from storage tanks.  Also,
because of the low emission rate from tanks, solvent disposal
charges are assumed to be negligible.  Therefore, there are no
direct costs associated with this control option.  The total
annual cost is then equal to the sum of the indirect costs.
     Direct costs were also not calculated for the control
options considered for cleaning operations  (freeboard ratio,
closed containers for cleaning) because there are no direct costs
associated with them.  In fact, an annual savings would result
from implementing these control options, although the amount of
solvent saved is difficult to quantify and is probably not
significant.
     Solvent credits were calculated for facilities with onsite
distillation systems if they were compatible with the additional
streams requiring control.  These credits were calculated as
shown in Table 8-9 assuming a 95 percent control efficiency for
the carbon adsorber, a 90 percent distillation efficiency, and a
solvent credit of $372/ton.3'22   The market value of solvent was
obtained from a chemical supplier.  The actual solvent credit was
calculated as 60 percent of the total solvent value to account
for the fact that solvent recovered onsite may not be pure enough
to use in the coating process.2  Recovered solvent may be used
for secondary purposes such as cleaning.
     Disposal charges were calculated for facilities that
currently dispose or would have to dispose of the solvent/water
mixture from carbon adsorbers.  These costs were calculated
assuming a 95 percent efficiency of the carbon adsorber, a
solvent/water mixture comprised of 4 pounds of water per pound of
solvent, and a disposal cost of $69/ton.3'23'24
                               3-23

-------
     8.3.2.2  Indirect Costa.  Indirect costs include overhead
charges, administrative charges, property taxes, insurance, and
capital recovery charges.  Overhead charges are calculated as
60 percent of the operating and maintenance labor and maintenance
materials.7  If no operating labor or maintenance costs are
incurred by a facility or model line, there are no overhead
charges.  As shown in Tables 8-9 and 8-10, administrative
charges, property tax, and insurance are a percentage of the
TCI.7  The capital recovery cost is a function of the equipment
cost and interest rate.  For this analysis, the capital recovery
cost was calculated by multiplying the TCI by a capital recovery
factor of 0.1424, corresponding to an annual interest rate of
7 percent and an equipment life of 10 years.7  Indirect costs
were calculated for control options related to solvent storage,
mix preparation, coating application/drying, waste handling, and
particulate transfer.  Only the capital recovery component of
indirect costs was calculated for control of removable parts
cleaning and for closed containers for all cleaning activities.
There are no overhead costs because there is no operating or
maintenance labor required for these control options.  The
remaining indirect costs (taxes, administrative charges, and
insurance) were assumed to be negligible because these costs are
calculated as 1 to 2 percent of the total capital investment,
which is small for these control options.
8.3.3  Summary of Control Costs
     Table 8-12 compares the cost-effectiveness of Regulatory
Alternatives I and II for existing facilities.  For Regulatory
Alternative I, the cost-effectiveness values range from a cost
savings of $260/Mg  (due to reuse of recovered solvent) to a cost-
effectiveness of $ll,200/Mg.  For Regulatory Alternative II, the
cost-effectiveness values range from $0/Mg to $16,900/Mg.
Industry-wide, the cost-effectiveness of Regulatory Alternative I
is $80/Mg, and the cost-effectiveness of Regulatory
Alternative II is $l,020/Mg.
     There are several things to consider when reading
Table 8-12.  First, these costs reflect only the cost of control,
                               3-24

-------
 TABLE 8-12.   FACILITY-SPECIFIC COST-EFFECTIVENESS OF CONTROL
Facility
1
2
3
4
5
6
7
B
9
10
11
12
13
14
Industry wide:
Cost - effectiveness
of RA I, $/Mg
$1,120
$1,160
$11,200
$0
$0
$1,180
$0
($180)
$0
$10
($230)
($260)
$1,550
$6,240
$80'
Cost - effectiveness
of RA II, $/Mg
$1,260
$1,190
$16,900
$0
$330
$6,470
$0
$1,910
$2,820
$10
$370
$690
$1,620
$3,260
$1,020
1.  These costs reflect only the cost of control and do not
    reflect the cost of compliance, recordkeeping, and
    reporting.

2.  Incremental cost-effectiveness to industry for
    Alternative II is $6,100/Mg.

3.  Differences in cost are due to existing level of control,
    operating conditions,  and onsite solvent recovery versus
    incineration.

4.  A solvent usage exemption has been developed that exempts
    Facility 4 from control requirements in this NESHAP.
                             3-25

-------
and do not include monitoring or reporting and recordkeeping
costs (discussed in Section 8.4).  Second, differences in cost
are due to the existing level of control and the operating
conditions at each facility, as well as whether that facility
performs onsite solvent recovery or incineration.  Finally, a
solvent usage cutoff has been developed that exempts Facility 4
from control requirements in this NESHAP.
8.4  COMPLIANCE, REPORTING, AND RECORDKEEPING COSTS
     In addition to the control costs described in Section 8.3, a
source subject to the NESHAP would be required to conduct an
initial performance test, perform ongoing compliance monitoring,
keep records, and provide emission reports to the appropriate
regulatory agency.  A description of these requirements and their
costs are provided in this section.25
     The compliance, reporting, and recordkeeping costs incurred
by a facility are the same for Regulatory Alternatives I and II.
This is because the only difference between the alternatives is
the LDAR program and the use of closed containers for cleaning
solvents.  Any additional compliance, reporting, and
recordkeeping requirements of the LDAR program are inherent in
its annual cost, as explained in Section 8.5.  There are no
compliance or reporting and recordkeeping requirements associated
exclusively with the use of closed containers.
8.4.1  Compliance Requirements and Costs for Existing Sources
     The compliance requirements of the NESHAP include conducting
an initial performance test and performing ongoing compliance
monitoring.  Table 8-13 discusses the monitoring requirements of
the NESHAP on which the cost estimate was based.  The following
sections describe in more detail the test methods required to
conduct the initial performance tests, their costs, the
compliance monitoring requirements and costs, and the total cost
of compliance for existing sources.
     8.4.1.1  Initial Performance Test.  There are two test
methods that could be used in the magnetic tape industry for
establishing initial compliance with the 95-percent HAP removal
efficiency required by the two regulatory alternatives being
                               3-26

-------
col
!
§
                        111
                                                 a 73

tt
H
S
w
a!

i
H
«
  f
                                                  1
                                                  u
                                                  o
                                                  •a
                                                  3|
                                                  s &
                                                  I!
                                                  if
                                                  U 3
ml
col
    I
S a   j
IS   a
                        4?
                         o
                         "3
                         3
   u
  _!)
  '>
   U
  •o
     O


-------
8-28

-------
considered for the NESHAP.  One method is Method 18, which
provides a breakdown of the species in the inlet and outlet
streams of the control device.  This method reveals the HAP
species being sorbed, the rates at which they are sorbed, and
whether a certain species is being differentially sorbed over
another species.  A percent HAP removal can be calculated by
using this method.  The other test method is Method 25A, which
indicates the total VOC's in the inlet and outlet of the control
device.  This method allows a percent VOC reduction to be
calculated, but not a percent HAP reduction.
     For this industry, all of the HAP's controlled by control
devices are also VOC's.  It is assumed that VOC removal is equal
to HAP removal.25  Performance tests conducted on carbon
adsorbers in this industry indicate that even when certain
species are preferentially sorbed over other species, the HAP and
non-HAP removal efficiencies are still above 95 percent.
     Method 25A is approximately 20 percent less costly than
Method 18.2S  This is largely due to the fact that seven
facilities that have continuous VOC monitors on their control
device would not have to conduct an initial performance test if
Method 25A is required, because their monitors would be able to
demonstrate that a 95-percent VOC removal is being achieved
continuously.  All of these facilities, however, would be
required to perform testing if Method 18 is required because none
of them have data on the removal efficiency of specific HAP
compounds.
     Given this information, the cost analyses for performance
tests for the magnetic tape industry were performed under the
assumption that Method 25A will be the test method chosen for
conducting initial performance tests.
     Magnetic tape manufacturing facilities that currently have a
continuous emissions monitor (GEM)  in place to monitor emissions
of VOC's will be exempt from conducting an initial performance
test.  The use of CEM's is discussed further in Section 8.4.1.2,
Monitoring Requirements and Costs.
                               3-29

-------
     The cost of conducting an initial performance test on the
control device is estimated to be $16,450 per facility. The
breakdown of this cost estimate is presented in Table 8-14.  For
the purpose of this analysis, it was assumed that all facilities
that are required to conduct an initial performance test would do
so using Method 25A.  Other assumptions used to derive this
estimate are listed in Table 8-15.  The cost of the initial
performance test for a total enclosure has been estimated to be
$7,945.  This cost estimate is based on the Agency's previous
experience with total enclosure demonstrations.25
     8.4.1.2  Monitoring Requirements and Costs.   Enhanced
monitoring is used to confirm ongoing compliance with the NESHAP.
Monitoring requirements generally use a site-specific operating
parameter to determine compliance.  This parameter can be chosen
either by the owner or operator of a facility or by the
regulating authority.  During the initial performance test, a
value is established for the operating parameter that corresponds
with the level of emission control required by the regulation.
During future monitoring, if the control device is operating
within the operating parameter value range specified during the
performance test, the facility is considered to be in compliance
with the regulation.  This type of enhanced monitoring is
generally required at emission points where the NESHAP requires a
percent reduction in emissions.  If the regulation requires
equipment standards to achieve compliance, enhanced monitoring
would entail a one-time visual inspection of the equipment to
make sure that the equipment is in place and operating according
to manufacturer's instructions.  If the regulation requires work
practice standards, ongoing compliance would be determined by
recordkeeping that describes how the work practice has been
executed.
     In the magnetic tape manufacturing industry, emissions from
the coating operation are generally controlled by either a carbon
adsorber, an incinerator,"a condenser, or a total enclosure.  The
following paragraphs discuss possible monitoring requirements for
                               3-30

-------
      TABLE 8-14.    ESTIMATED INITIAL PERFORMANCE  TEST  COSTS
Cost elements

Method 25 Aa
1. Site Survey
2. Site specific test plan
3. Equipment preparation
4. Testing
5. Post test equipment
6. Report/data reduction
7. Management
Subtotal
1. Travel (1 car, 1 day; 1 truck, 4 days -f gas)
2. Room/board (2x3 days/2 nights)
3. Calibrating gases (propane)
4. Other gases
5. Supplies
6. Services (phone/reproduction/ word processing)
Subtotal
Total
Total enclosure1*
Initial performance test
Technical labor
Management labor
Clerical labor
Administrator determination
Technical labor
Management labor
Clerical labor
Total
8 hr x $75
4 hr x $75 + 28 hr x $70
8 hr x $70 + 24 hr x $60
30 hr x $70 + 30 hr x $60
12 hr x $60
40 hr x $70
8 hr x $75
$600
$2,260
$2,000
$3,900
$720
$2,800
$600
$12,880
$525
$430
$500
$750
$1,200
$165
$3.570
$16,450

155 x $33
7.8 x $49
15.5 x $15
60x$33
3x$49
6x$15
$7,947
$5,115
$382
$233
$1,980
$147
$90
* $7,945
aSee Table 8-15 for the assumptions used in the cost analysis.
"The person-hour requirements are derived from the SF-83 and supporting statement prepared for the New
 Source Performance Standard for the magnetic tape industry. This standard was promulgated in 1988.
                                       8-31

-------
 TABLE 8-15.  ASSUMPTIONS FOR MAGNETIC TAPE TESTING BUDGET
                    (Method 25A Analysis)

Facility is located 4 hours driving time from testing firm
sampling crew office.	

Site survey by one senior person for 1 day.	

Conduct continuous Method 25A analysis for VOC's from the
inlet and outlet of APCD for 2 hours per test run.  Perform
three test runs for program.	

Conduct temperature, airflow, and moisture determinations
at the inlet and outlet in conjunction with Method 25A.	

Program schedule is:
  Day 1 - travel and set up equipment at facility
  10 hr
  Day 2 - conduct runs 1, 2, and 3
  12 hr
  Day 3 - pack equipment and travel
   8 hr	

Moderate preparatory costs for equipment calibration, check
out,  and capital and expendable supplies.	

Dedicated analyzers will be used at each location to allow
simultaneous analyses.	

One common laboratory supplied by the facility will be used
to house the two GC's.  The analyzers will use common gas
supplies as much as possible.	

Field crew has one mid-level person to run the analyzers
and to be in charge.  A technician-level individual will
conduct temperature and moisture tests and flow
calculations.	

Room and board costs $100 per person per day.	

A site specific test plan containing a QA plan is required.
Only a minimalist report will be prepared.	

No facility upsets,  weather related delays or equipment
problems will alter the schedule.
                            3-32

-------
each type of control device used to control emissions from
coating operations.-
     For emissions controlled by a carbon adsorber, a facility
would likely monitor the inlet and outlet total VOC concentration
such that a percent VOC removal could be calculated.  (Because in
this industry all HAP's are VOC's, VOC measurement is allowed as
a surrogate for HAP measurement.)  If a facility has VOC CEM's in
place, an initial performance test would not be necessary,
because the CEM's would be able to indicate the percent VOC
removal.  Ongoing compliance would then be determined by
continuously monitoring the inlet and outlet concentrations.
Alternatively, a facility could select the outlet VOC
concentration as the operating parameter, setting the value as
that which corresponds to a 95 percent control efficiency.  The
outlet VOC concentration would then be continuously monitored to
demonstrate ongoing compliance.
     Facilities that plan to install VOC CEM's in order to comply
with the NESHAP would still be required to conduct an initial
performance evaluation of the CEM to ensure that it functions
according to manufacturer's specifications.  The cost of this
initial performance evaluation is approximately $6,500.25
Quarterly audits would also be required to ensure that proper
operation of the CEM's is ongoing.  The cost of these audits is
$1,480 per year.
     For emissions controlled by a thermal incinerator,  the
combustion temperature would be the operating parameter chosen to
demonstrate ongoing compliance with the NESHAP.  For thermal
incinerators, the minimum allowable combustion temperature that
corresponds to a control efficiency of 95 percent would be set by
the owner or operator during the initial performance test.  For
catalytic incinerators, the gas temperature both upstream and
downstream of the catalyst bed would be the site-specific
operating parameter chosen.
     For emissions controlled by a condenser, the site-specific
operating parameter would be a continuous measure of the
temperature of the exhaust stream.
                               3-33

-------
     All of these control devices would require a flow switch on
each bypass line.  This switch would need to be installed at the
entrance to any bypass line that could vent emissions to the
atmosphere, in order to record times during which flow is
diverted from the control device.  Based upon a conversation with
one magnetic tape manufacturer, the sole bypass line is
associated with the solvent laden air (SLA)  duct, so one flow
switch would be needed for each control device.25
     For the total enclosure capture system, any of the following
parameters would suffice for the site-specific operating
parameter:  air flow rate, absolute pressure within the
enclosure, or differential pressure across the enclosure.  For
the purpose of this cost analysis, it is assumed that
differential pressure will be measured; this is the parameter
monitored by facilities currently required to do such monitoring
under the NSPS.  The equipment required to monitor total
enclosures are four pressure transducers, a controller, and a
data logger.
     Table 8-16 shows the capital cost, annual cost, maintenance
cost, and total annual cost for each of the monitoring
instruments discussed above.  For carbon adsorbers, the capital
cost of a VOC GEM is $25,000.  The total annual cost of the GEM,
incorporating maintenance and performance demonstration costs,
is $20,092.  For condensers and incinerators, the capital cost of
a thermocouple to monitor temperature is $1,500, and the total
annual cost is $3,896.  The capital cost of a flow switch for a
carbon adsorber, a condenser, or an incinerator is $225.  For a
total enclosure, the capital cost of equipment required to
monitor differential pressure across the opening are four
pressure transducers at $300 each, a controller at $450, and a
data logger at $3,000.  The total capital cost for monitoring a
total enclosure is $4,650.  The total annual cost for this
equipment is $662  ($171 for the pressure transducers, $64 for the
controller, and $427'for the data logger).  For the total annual
costs, maintenance was assumed to be $17.50 per hour, and a
capital recovery factor of 0.1424 was applied to the capital
                               3-34

-------
H

tt
a

H
a
H
a
o
u


§
H
t*
O
2
U}

H
 i

00
         h
             I  H
             s
              X «.

             a
VOC Contin

Monitor* (C

Flow iwitch
                   6  H
                      wi
                      s
                   S1
                   3
                        i
                        s
H
                         o

                         |


                         u

                         3-



                        £
                              il>

                              i.sl
                              31
                                                       •o 3-2     B

                                                       •I5  2=3

                                                               '"

                                   3-35

-------
coat.  This corresponds to a 10-year equipment life and a
7-percent interest rate.
     8.4.1.3  Total Cost of Compliance.  Table 8-17 shows the
total cost of compliance per facility assuming Method 25A is used
for the initial performance test.  The total annual cost of
compliance, including the performance test, ranges from $32 to
$26,021.  The facilities with the lowest annual compliance costs
are those that have already installed a carbon adsorber and a
total enclosure as required by the NSPS, and either have
installed a VOC GEM or use a. material balance to demonstrate
ongoing compliance with the NSPS.  The facility with the highest
annual compliance cost has no control device on the coating
operation and two total enclosures in existence.  This facility
will be required to install a control device, conduct an initial
performance test on it, and conduct performance tests on both
total enclosures.
     Industry-wide, 33 performance tests are anticipated to be
required for the NESHAP, resulting in a total performance test
cost of $365,650.  The total industry-wide annual cost pf
compliance with the NESHAP is estimated to be $115,640.
8.4.2  Reporting and Recordkeepincr Costs for Existing Sources
     The facilities subject to the magnetic tape NESHAP will be
subject to the reporting and recordkeeping requirements outlined
in the proposed General Provisions to 40 CFR Part 63.  In
addition to those requirements, both Regulatory Alternatives I
and II require sources to measure and record the freeboard ratio
for the wash sinks used to clean removable parts.  The reporting
and recordkeeping costs for existing sources are described in
Tables 8-lSa and 8-lSb.  These two tables represent first-year
costs and second-year costs, respectively.  During the first year
of compliance with the NESHAP, the total national burden and cost
to the magnetic tape manufacturing industry is estimated to be
$143,514, with a corresponding average first-year cost per
facility of $11,040.  For the second and third years, the total
national burden and cost to the magnetic tape manufacturing
industry is estimated to be $93,594 each year, with a
                               3-36

-------
                                              a
Q
Ed
O
W
01
H
                         55°?}
             §  s
             oo  n
                  §  2
                                        O O °°.
                                           m
                                           (S
                                                 00
                                                    00"
                                      •a
                                      i
                                   00.
                                                              0 0

                      i
         o o*^
         ZZz
                              I  «
                  -oo
                           zz
                              zz
               "s
rber
            §2°:
            ISIS
            U-oa
ler

Ad
1  12
Z  pi
IS
                          o r* «
                          o-a
   •3
!  i
j  i
                       is
                       §2
                5
                     w
                    Sir
§2
3!
                                                         •S PUJ
                                                         5^H
                                                         •ii

•e
o
C4
^H
eH
12
5%
                                      B-37

-------
•o
0)


g
•H
4J
£
O
O
 i

CO



W
     ©II
I
         •§•
          -fj


          1
   15
                 II
                 13
       •3 S

      > II


       "4
                     1
                     •§
                     5  ill
              *S3
            •S  * ».S

            i  511
            s  u Si
           t
                          .1  I1
                          o  . ^3
  S Oei S ^JBO

llJlilll
J4 •= . S - "3 3 -a
si3*a«g^


liilfj'il
                                  a
                                  o
                                  I
                        =


                        I


                        '3
                        a

                        s
                        •g


                        I
                        o
                        g
             1
             8
                                 •8
                                 •s
                                 u
                                 3
               .
           3 — 5 "» " 3 2 ir
           !|ijslf|
           I § z= ;§ •*.; .s §
                                 3-38

-------
       I
       3


       I
       •8
                          i
.3

                     I1
                      a
                      3^
                      r
.s
s



1 s


s i
2- *


I"


|
3-39

-------
I
1
§
•8
             1
                        8
                   «^a  s
i
3
>i
e
I
                                             <•
                                             o
     2
     2
                          3-40

-------
          I

          I

          2

         » *
        a. s *
        r ii

        !f
        2
        J?
        •3
        3
II
 '
        I
        1 '
        I
        I
        3 •
        I ^
3-41

-------
        .1
II
tl
           •8
           li
                           I
—*
2 S


i
>> M



II
"a s

I S
9-42

-------
1
     3

     |
         .1
                I
                   1

                         .1
                           11
                             •a "8

                                                  II
                                                           • I
                                                           i J
                                                           11

                                                           n
                                                           11
                                                           M
                                                           *1

                                                           II
11
M.I
                                                           If!
                                                           *JJ

                                                           III
                                                           axe
                                                        -  V
                                                        •a  i.=
II

8-3

3 i
= a




 .3
                                                             Is
   H
   «"i
   3^
   | |


   I -i
   — «

   ii

3 H-?
u — J _jj

His
* *z *
  toon
  V c i

e a ^S
s g I|
2 s = 1
«5 Ij2 S
                                                           1 i"
                                                           8 |
                                                           I
                            3-43

-------
I

            JS
            3
            3
       5
       •3
             I
             !

             3
             3
             .8
      *   I  1
             «!
1  ii
1  la
          *  1
          i  !*
          1  ti
       28
             5-"
             1:
       11 !  ij
       I 3  «  5 =
       J- 1  ij
          I  i!
          f  i-J
          t2  P 8
         1-44

-------
corresponding average cost per facility of  $7,200.   The  ongoing
reporting and recorkdeeping costs are calculated as  an average  of
the first, second, and third year costs.  The average
industrywide cost over 3 years is therefore
 ($143,514 + $93,594 + $93,594)/3, or $110,234.  With 13  existing
facilities subject to the rule,  the average cost per facility is
$8,480.
     8.4.2.1  Assumptions Made in Estimating Reporting and
Recordkeeping Costs.  The following paragraphs discuss the
assumptions made in estimating reporting and recordkeeping  costs
for magnetic tape manufacturing  industries  subject to this
NESHAP.
     The average facility burden was calculated by dividing the
total facility burden by the number of facilities.   An actual
cost per facility cannot be calculated because it is not known
which facilities will have excess emissions and which will  not.
Technical person-hours were estimated at $33/hour, management
person-hours at $49/hour, and clerical person-hours  at $15/hour.
     The costs that result from  the following activities will
only be incurred during the first year of compliance with the
NESHAP:  (1) reading the instructions for the reporting
requirements, (2) preparing notification of applicability of  the
standard for new and existing sources, (3)  preparing notification
of applicability of the standard for new and reconstructed
sources, (4) preparing notification of the  initial performance
test, (5)  preparing a compliance status information  report,
(6) submitting a startup, shutdown and malfunction plan,
(7) submitting a quality control plan for a continuous monitoring
system  (CMS), (8) preparing a waiver application, and
(9) developing a record system.
     For this NESHAP,  both Regulatory Alternatives I and II
recommend identical reporting and recordkeeping requirements for
both new and existing sources,  so notification of construction,
reconstruction,  and physical or operational changes are not
necessary.
                               3-45

-------
     Within this source category, 14 sources are considered to be
major sources.  One of these sources is expected to fall below a
solvent usage cutoff that will be in the regulation, and would
therefore not be subject to the NESHAP.  A one-time applicability
report will be required, however, for this source, as well as an
annual report certifying the solvent usage level.  All but three
of the remaining 13 facilities expected to be subject to the
NESHAP will be required to conduct a performance test, either for
the add-on control devices or for the total enclosure(s).  All
but one facility plans to operate a CMS; that facility has only
one control device and will perform a material balance to
demonstrate compliance.  None of the existing area sources are
expected to exceed the solvent usage cutoff or to become major
sources.
     The reports of monitoring exceedances, periods of
noncompliance, and no excess emissions will include data based on
CMS performance and/or material balance results.  It is assumed
that 90 percent of the facilities in this source category will
have no excess emissions; therefore, these facilities will only
have to report quarterly in the second and subsequent years.  It
is estimated that ten percent of facilities will have excess
emissions; these facilities will need to report quarterly during
the second and subsequent years.  All facilities will be required
to maintain a 75-percent freeboard ratio in their wash sinks, and
use a closed system to flush fixed lines and for particulate
transfer.  A record system will need to be developed to indicate
that a 75-percent freeboard ratio is being maintained in the wash
sinks.  Other records that will need to be maintained at the
facility include those associated with the CMS.  The burden of
reporting and maintaining these records is included in the
maintenance costs for CMS in Table 8-16.
8.4.3  Compliance. Reporting, and Recordkeeping Costs for Model
       Lines
     The following paragraphs will discuss NESHAP requirements
and costs that differ for new sources.  Initial performance
tests, compliance monitoring, and reporting and recordkeeping
                               3-46

-------
costs are slightly different for new sources from the
requirements and costs for existing sources.
     8.4.3.1  initial Performance Test Recruirementg and Costs for
New Sources.  For this NESHAP, the requirements and costs for
conducting initial performance tests are the same as those for
existing sources.  As they are defined in Chapter 3, however,
small model plants were not subject to the NSPS because they fell
below the solvent usage cutoff.  New small facilities will have
to conduct initial performance tests for carbon adsorbers and
total enclosures, incurring costs from which they were exempted
under the NSPS.  The costs for these tests are presented in
Table 8-14, and are the same as the costs for initial performance
tests for existing facilities.  New facilities that fall within
the medium and large model plant parameters will be subjects to
the NSPS.  They will therefore already be required to install
total enclosures, carbon adsorbers, and VOC CEM's on the carbon
adsorbers, and to conduct initial performance tests on these
devices.  These facilities would be analogous to existing
facilities with CEM's, and would not be required to repeat
initial performance tests for compliance with the NESHAP.
     8.4.3.2  Compliance Monitoring Requirements and Costs.  The
compliance monitoring requirements and costs would be the same
for both new and existing sources.  Small model plants would be
required to install and operate VOC CEM's, total enclosure
monitoring equipment, and a flow switch on the SLA bypass line.
Other model plants would only need to install the flow switch, as
the other equipment would already be in place in order to comply
with the NSPS.  Like existing facilities, they would be required
to conduct an intial performance test for the CEM'3 and conduct
quarterly audits.  Table 8-16 presents the compliance monitoring
equipment and costs for existing magnetic tap« manufacturing
facilities; the costs would be the same for new facilities.
     8.4.3.3  Reporting and Recordkeepincr Requirements and Costs
for Model Lines.  New facilities would be subject to the same
reporting and recordkeeping requirements and costs as existing
facilities.  The only additional requirement for new lines would
                               3-47

-------
be incurred by preparation of notification of anticipated start-
up and actual start-up reports, which are only required for new
lines.  This would result in the costs for reporting and
recordkeeping being higher for new facilities than for existing
facilities.  The reporting and recordkeeping costs for new
sources are given in Tables 8-19a and 8-19b.  These two tables
represent first-year costs and second-year costs, respectively.
             •       •
     For model lines in their first year of operation, the total
cost for the five projected model lines would be $56,534.  The
average cost per line would be $11,307.  For model lines in their
second and third years of operation,  the total cost for the five
projected model lines would be $37,061.  The average costs per
line per year is therefore $7,412.  As with existing facilities,
the ongoing reporting and recordkeeping is calculated as an
average of the costs over the first 3 years.  Therefore, for new
lines, the total average cost is approximately $43,552
($56,534 + $37,061 + 37,061/3) with an average cost per line of
$8,710.
8.5  EQUIPMENT LEAKS IN PIPING
     As with other control options, there are direct and indirect
costs, as well as recovery credits, associated with controlling
equipment leak emissions.  The calculation of these costs is
discussed below.
     The only control option above the MACT floor that was
considered for controlling equipment leak emissions is the
program specified in the Negotiated Regulation for controlling
equipment leaks in the synthetic organic chemical manufacturing
industry (SOCMI).2S  The Negotiated Regulation requires control
of equipment leak emissions through equipment modifications that
reduce emissions, as well as a leak detection and repair (LDAR)
program.                             •
     Components of the total capital investment  (TCI) associated
with implementing the equipment leaks control program described
in the HON Negotiated Regulation include the cost of equipment
modifications and initial costs associated with the LDAR program.
Initial LDAR costs include the cost of a monitoring instrument to
                               3-48

-------
<
?
\f.
Z

I
a
z
i
z
I

i
p
Q
2£

3
z
z
£

r 3 S
— ' w ^^
J) ^^
•=, i S 2
3 ll>a
£ -It"
^ 5 *• 0

•z £ *-
5 5 sS
E - 3 o
r™ *• u ™ • **« *
i- 5» i ^ a
•^ a S u
= » .u II
s|~-=.
ili?
51 J 1> ^
- 1 1 s "
si £ a. a
u —
i.
^
a L.
u n
-v •= "
^ S >
^Zf A U
1£

eg ^
= 1 S?
J ^f^f
~ ^Ji:
MU. ^i
*•£
"S § S
B J{ §>•
•— ' saw
II-
n
2 2
< 1!
J*
a- a.






Burden llein































<
Z









pplicalions
<
—





























I









jrveys and Sludies
to
ri
g

1
:?8
^
1
£"?
S:-:;
;•:•:•
Si:
•-.-•
:-K
•:•:•'•
is;
::4:
&
K*
1
?V.
,:Av
&'
r:
fe

i-
^
w.
1
1
•ft "•
£S
P
P

1
1
•:::::
|;
|iv:
:|:S

.




eporling Requirements
as.
m'
VI
00



—




o




VI



VI







-~








u



A. Read instructions


i

1?
1
;,::':;:
::.v
ift
i;i
II
11
^
•:#
i;S;:
•?:•':•
is
»s
;••.•:•
«
•:•:•:••
1
1
•:•::
X- -.
*?;•
&*
;-..
1
;i;s
l;|
i
i:':>'
:••':"'•
i?l
:•?•;
^

$1
:&:
;:V.;
:::S:
te
•;:;x
1

•W1




B. Required aclivilies

































«r
00
1
Z
v>


Q
£
<
8
W
U
c
™
I
"^
[5

































•T
00
U
I
8
CO


8
a.
u
u
u
C
?3
•r
i
!2
's
^3
H
V
a:

































•^
QO
£
S
en


1 enclosure
2
Initial performance lesl-

































T
00
1
u
u
CO
u
s
u
^
3
1
u
CJ
e
«
Repeal of initial perform































-S a
•o •»
11
^ a
^ m






e
«
!s
Iu
3
5
CJ































-StU
-o •»•
U 13
ll
Is





a
O. Gather existing intormali


y

8$
i:'!::::
J5::
$$'
iz
ii
ss?
sfe
Hi
&
l&

1
1?
1;
••••':
?S*
:;-:$:
^1
>£••!
S
%
:::-:;::
[•••[•••
&?:
^
1
1
:1:
1
&1":
:;:: :
1
::::':;.
•:•:•.
:|>x
:".":••
-:*:¥:
:;*v
i
^
i
w-




E. Write report































<
Z





•o

a
s
a
•A
hi
;j
Notification of conslr./re































^
Z





•a
»i
E
lional chanj
2
i
Nolificalion of physical/u


3
r-
m


—




flm




0



VI




rM


—


M





U

Q.
S
a
>i
~3
U
13
c.
:§
s
'H
^
u
3
z


s
r-
n


•—




_.T




O



VI




n


—


. 3
Notification of applicabil
slandard--new/recoiiM


rt



0




9




ulii>l |'l
ttionilot int> syMcins

i
                                      3-49

-------
>•
5
I

u
36
M
e
'a.
|
|
9

9
9
9




9

IM



"

(M




"" U U
?!
ii
u 2
Ot E
u c
> .-^
Report when exceed so
report area source t>



3


9
•O




-

•0



*"*






'^f





Waiver application



S



3




-

3



f

"™



„
•* "3
2L
1
A

i
u
X
Report of monitoring e
of noncompliance



!*•



a




-r

?!



^

00




-"


^


Report of no excess em






























£
ordkeeping Requiremcn
3
at
•»


















.2

_a







Read Instructions
^



















<
i









Plan Activities
a



















O
deduH

JC







Implement activities
U


I
r*

o
g
N



•"

o



"

S




«"





u
•2
}
a

i

|
j
i
1
1


F:
I
:-:-::
K:

1
•:?:
i
S
1
£s
?:'






^
Time to enter informali
U

I
a
«


s
00



"

r*



*

10
9




8 c
(A
3

S 1-
~Z 3
Facilities above cutoff
- including records ol
shutdown, malluncli
of freeboard lalio



9

9
9
9




9





^






•~





Facilities below cutoff





















Z









Time to train personnel
tL




















Z









Time for audits
0


i/l
8
~
>>
S






-.
,


'
'
,.
'-
V

Z
O
z
y^

J
S2.
u-t
L BURDEN AND COS
<
r"*
t-
r-
^




•















a ii
I

s
-
.is
Average
















g
.2
|
'3
>
3
TJ
11
73
3
a
2
n
a
JI
s
z
1
1
1
2
2
4
e
i
3
TS
C
§
U
2
a
C
o
j;
c
1

e
u
e
ao
CQ
C
I
i«r
s
n
73
U
2?
a
"3
u
•3
3
o
s

i
a
's
1
I
3
i
C
a
••A
t
0
s
i
I
3
§
3
3
1
e
assumi

u
&
H
u
at
c
3
T3
T3
••
3
U
^*
|
1
J)
0
•f
ng sources. Thus, notification of construction, reconstruction, am
^
'x
1
T
«j
^
3
2
|
U
a
33
u
3
i
3
£
|
ift
C
U
M
y
'5
sr
u
^"
^
1
'
^
1












£>
3
u
1
I
-J)
A
^
C
™
3
^
1.
<
3
"3.
u
r5
-o
c
p"
3
-5
i
__p
"3
S

s

"3
|
i
^
15
^
J
o
•3
U
1
S
3
I

                                                                                                     c  >
                                                                                                    '>  i
                                                                                                •=•   H -=    =
                                                                                                      .
                                                                                                     7
                                                                                                 i
                                                                                                <   c •=
                                                   a-50

-------
3-51

-------
s

z
UJ
z
&

z
z
z
z
I Js
„§,,=
i, jj s w
w S.
a. ^
— v» — ..
C b i/7

b f S^"
S£ fe
2 ^
U 3?
s i??v
r- 2 a. a
u ^-
a.
C u
" ^1
Of
2 g
J| s?
2 | |t If
U b ~
a. u
a.
° 1 1 "
2 8
3 e
- 1
< "Z u
^ y
a °
a. Q.




c
•j
a

























Z









pplicalions
_
























Z









jrveys and Studies
VI
"'"


•;

$

"4
s
"''





••







.
?v



"'
',':
%-"'



eporting Requirements
K
9
e


9



9





9


-






"







u

A. Read instructions

-i
:

''•

j2
?
%•£
% :
-":
i


;

K :




-
f
f
iff
•,
[',

'



,



B. Required activities



























z
00
J
F-
u
V)


Initial performance test-APCD



























2
00
JJ
a
8
V]


Repeat initial performance lest— APCD



























2
00
JJ
f-
u
VI


Inilial performance lest --Total enclosure



























2
00
u
J3

U
U
V]

8
Repeal of inilial performance test-Total

























.5 tu

1i
3 n
"a a
.= en





C. Create information


























^ •*
u -a
3 A
Is





D. Gather existing information

''
,„;

•
,
^
'i
'•
i
•~
"• .


•
•

:




;
'
••
yv

*

>:

s



E. Write report

























Z







•o

Nolificalion of conslr./rcconslr./slartup

























Z







•a
8
BJ
Notification of pliysical/operalional chanj

a
o


9



0





9


l.m lor conlinuou
niomtoiin^ iy^U'ins

                                   9-52

-------
3
•J
I
C
u


a
a
9


9
9

9
9
r-l

^







"" u *»
2, |
^ i
= 0
» «
il
5 3
-J -C
Report when exceed 1
report area source

9


9
9

9
9
•a

•a








wi

^5:

s




B
O

Time to enter informa
U

3
<•?

-r

00

8
9








U
2
i
£•0
3 a
2 a"
VI 2

Facilities above culofi
- including records c
shutdown, malfunc
of freeboard ratio





o

9
9
(N

(S

.










Facilities below cutofl













<
Z









a
s
i
'3
o
u
g
P
U.












<
z'









Time for audits
O

r*~
m
§
0

i
-
s>
^
-
f.



•* :

z
O
z
p
^
Crt
C/l
L BURDEN AND CO
i

ff





















Average cost per facil

                                                               I
                                                               I


                                                               I
                                                               S.
s

I

•
    4
                                                                   d.
                                                                   S

                                                                   S
                                                                   •a  9
                                                                      J
  11
i. = a
§ I J
                                                                                  -3-3
                                                                                  2 - -3
                                                                                  -23
         3

         Ji
         o

                                                                                    S 3

s
avoid f radio

o
1
3
k.
U
a


g
3
w
3
re charged al

n
3
«
e
i
^3
,3
C
*
£
1
2
i red during 1
3
'£
i
-%
3
5
w!
g

j)
g
3
*
menls for nt

i
u
^
1
Vl
.2

••
rp
a

ges is not nei

2
s
o
I
•Q
"&

>I
"a.
TJ
U
vf
J.
a
|

^
•3
;H
.3
5.
3
a.
u
-C

,s
2 -
=
u •
j
1
= :
~
| |
"5 U*
3* —
T" "J -

3 ^ ^
Z i- -f
                                        3-53

-------
s
.S
I
oe
90
e
•s.
u
J<
•2

dj
a
•o
e
«
04

i
u
C£
a
u
g
i.
s
0
S
2
13
"8
2
o
g
u
3
u
1
M
^C
'c
5
c
'3
S
"5t
c
'•5
L«
4J
"o
1
3
^
U
fS
VI
S
'^>
^
s
I
.2
n
^


^
1
e
3
u
c
•™
•v
S.
5
o

§
I
n
s
s
c
•S
~
•3
E
"3 .
| 1
i 1
n
II
1 3
   •a
   ^3
   ~
   s
  12
       ^ 1
     -2 .S
   s '
  C .2
       •a
  11
     ~ •*

     "3 2
!i
       &.S

       SI
3-54

-------
detect leaks and initial monitoring and repair costs which are
more extensive and costly than subsequent monitoring and repair
costs.27"31  To calculate the TCI, capital costs pertaining to
control of equipment leaks were annualized using the following
capital recovery factors:27"31
     0.55   For rupture disks and pump seals  (7 percent
            interest, 2 year life)
     0.21   For monitoring instruments  (7 percent  interest,
            6 year life)
     0.163  For all other types of equipment  and initial LDAR
            costs
     The total annual cost includes direct costs (periodic
monitoring, maintenance, and repair of piping equipment) and
indirect costs (maintenance of modified piping equipment and
monitoring equipment,  taxes,  insurance, administrative expenses
and capital recovery).  Solvent cost savings resulting from
reduced emissions are also taken  into account.27'31  Appendix F
explains in detail the approach used to calculate equipment leak
control costs.
     Certain facility-specific information is needed to calculate
control costs for equipment leaks.  This information includes
equipment counts for each operation that is controlled  (mix
preparation, solvent recovery, and waste handling), the weight
percent HAP's in the fluid contained by the piping and the time
that the HAP solvent is in contact with the fitting.  Only two
magnetic tape manufacturing facilities provided representative
equipment counts for piping in the solvent recovery and mix
preparation areas, and no facilities provided equipment counts
for waste handling areas.  Consequently, different approaches had
to be developed to estimate costs based on the information
available.
     For the solvent recovery area, the cost effectiveness of
controlling equipment leak emissions was calculated for the two
facilities that provided equipment counts.  The average cost
effectiveness for the two facilities was used as a "cost factor"
to estimate the total annual cost for the remaining existing
                               3-55

-------
facilities.  For the mix preparation area, the cost effectiveness
for two of the three facilities supplying information was used
because the other facility's mix room operating hours are not
representative of mix room operations for the industry in
general.  The cost effectiveness values calculated for the
solvent recovery and mix preparation areas are $l,026/Mg
($931/ton) and $5,889/Mg ($5,343/ton), respectively.  To estimate
the amount of HAP's controlled, uncontrolled equipment leak HAP
emissions from each operation were multiplied by the average
emission reduction resulting from control of equipment leaks
(89 percent).  The amount of HAP's controlled was then multiplied
by the cost effectiveness to estimate the total annual cost of
controlling equipment leak emissions for facilities that did not
provide specific equipment counts:

   hSSS^"*"'] ' °'83 • (Effectiveness tin) ' <*»trol Cost
     A different approach was necessary to calculate equipment
leak control costs for waste handling because none of the
facilities provided equipment counts for waste handling piping.
Instead, equipment counts were obtained using model parameters
developed in regulating waste solvent treatment, storage, and
disposal facilities (TSDF).31  The TSDF analysis defines three
model unit sizes: small, medium, and large.  Waste handling
operations at magnetic tape facilities were classified as small,
medium, or large based on the amount of solvent processed
annually  (solvent reclamation rate).  The equipment count of the
corresponding model TSDF was then used to estimate the annual
cost of controlling equipment leak emissions from waste handling
piping.  Existing facilities with waste handling operations
provided the weight percentage of HAP's in the waste fluid stream
so that HAP emissions could be estimated; an average of .
70 percent was calculated for model lines.
                               3-56

-------
8.6  RESULTS OP THE COST ANALYSIS
     Tables 8-20 and 8-21 present total industry-wide costs for
RA's I and II, respectively, for the magnetic tape industry and
for model lines.  Tables 8-22 and 8-23 present total annual cost,
cost per unit area of tape coated, and average and incremental
cost-effectiveness of RA's I and II for existing and new sources.
These costs are discussed below.
8.6.1  Cost Effectiveness
     The cost-effectiveness value is the total annual cost to
control 1 Mg  (or ton) of HAP's.  The average cost effectiveness
of an alternative is determined by dividing the total annual
control cost of an alternative by the annual HAP reduction in
megagrams.
     Although the industrywide cost effectiveness is low, there
is significant variation in the cost effectiveness to individual
facilities under RA I.  Some facilities would actually save money
by implementing additional controls (up to $240/Mg saved due to
solvent recovery credits), whereas the highest cost effectiveness
value for a facility is $27,480/Mg of HAP controlled.  This
variation in cost effectiveness is due to the diversity of  .
operational characteristics among magnetic tape manufacturing
facilities.  Cost effectiveness values for new facilities under
RA I range from $l,620/Mg to $17,380/Mg HAP controlled.
     Under RA II, cost-effectiveness values for existing sources
                                             •
range from $360/Mg to $32,070/Mg HAP controlled.  For new
sources, the cost effectiveness values for RA II range from
$l,850/Mg to $14,170/Mg HAP controlled.
8.6.2  Cost per Unit Area of Tape Coated
     The cost per unit area of tape coated was calculated for
model lines with RA I and RA II.  Tables 8-22 and 8-23 show the
range of cost per unit area of tape coated for RA I and RA II,
respectively.  Under RA I, the cost per square meter of tape
coated is as low as *0.10/m2, and as high as *6.86/m2.  Under
RA II, the costs ranged from <{:0.2/m2 to <:7.2/m2 of tape coated.
                               3-57

-------
H


g
ft
CO
g
CO
o
u

w
o
H
H
O
O)


CO


CO
     ft
     I'
      !•£
       u.

           a

            
-------
H
Oi



I
w



Ed
t*
O
fc

EH
CO

8

w
Q
H
EH


§
H
g
CN
 i

CO
    11
     I1
    r

         s
           S

             3
s
                  fS

s

s
                         a
g
                         S

                                §

                                            1
                                        g


                                              s

                                                   3
                                                            i
                                        (S
                                        rs
                                                     
-------
CN
01
 i
00
a


I
                     g,
                     'a S-2
                     ill
        111
                        9 *-S on
                        |lf
                        < «*»
                      .— .-•
                      11.i
                        o
                        U
          §•!
         II
]**
lf
                             <
                             Z
                             8
             00 <7\
             o n
             rirf
                             s
                             I
                             'S
                              5
    ll

    11
    = J
                  ^
                  Z
         i
                  2?
                  vf V
S
                                   r
                                   a
                                 Jl"
             •<
             Z
I
                                       oo en
                                       en ON
                                       — O
O
"5
S

'•§
    <
    Z
<
Z
                 ON 5
                 en p
                 O O
a
I
    <
    Z
<
Z
                     s
                              o\
                              25
                              00
                              •f
                                               u
    •<
    Z
<
Z
            29
            6 o
                                                   o
             S
             2
             3
             •o
             S
             00
             u
             3

             .3
             1
                                                       •5
                                                        8
                                                       ,2
                              I
                              =3
                              >t
                                                       £
                                       Z
                                                3-oO

-------
w



03





Q"



O


w
  H
fe
  u
88
CO
g

             oo
             2
* " V
Q u> o
8 § 2
         * •-

         II!
       .ill-
       BJ
           J
           "o
           ? 3
           t|
           3s
           ,0 >
i


1
               en
                 ts
               82'
               0 O
               00
                so
               .3
               "V 's
               ^ J
                      o
                      m
                      r~
                      (H
                      3
                ss
                          — 0
                          8
                          3
                               CO
                               ts en
                               °°. °.
                               ^ ^
                               m
                    
                                        CO —
                                        00 —
                                        **! *"
                                        r»* pf
                                        co oo
                                          oo
                                        O 0

                                               oo

                                               i
                                               2
                                    00
                                    o

                                    I
                                    .s

                                    1
                                    q
                                              i
                                               00
                                              '•3
                                               >,
                                 3-
                                    <3 J.

-------
8.6.3  Incremental Cost Effectiveness

     Incremental cost effectiveness is a measure of the

additional annual cost required to achieve the next higher level

of emission reduction.  The incremental cost effectiveness is

calculated by dividing the incremental increase in the total

annual cost by the incremental emission reduction.

     Industry-wide, the incremental cost effectiveness of RA II

over RA I is $6,100/Mg ($5,540/ton) HAP controlled.  For small

model lines, the incremental cost effectiveness is $54,000/Mg

($54,000/ton).  For medium lines built without concurrent

consturction of a control device, the incremental cost

effectiveness is $4,920/Mg ($4,570/ton); for those built with a

control device, the cost-effectiveness is $5,330  (4,920) .  For

both large lines, the incremental cost effectiveness is $2,380/Mg
($2,160/ton).

8.7  REFERENCES FOR CHAPTER 8

 1.  Memorandum from McManus, S., MRI, to Project File.
     February 1, 1993.  Documentation of methods used to
     determine costs of control options and regulatory
     alternatives for the magnetic tape manufacturing industry.

 2.  Magnetic Tape Manufacturing Industry--Background Information
     for Proposed Standards.  U. S. Environmental Protection
     Agency, Research Triangle Park, NC.  Publication
     No. EPA-450/3-85-029a.  December 1985.  Ch. 6, pp. 6-1
     through 6-9.

 3.  Memorandum and attachments from Beall, C., and J. Glanville,
     MRI, to Johnson, W., EPA/CPB.  March 15, 1985.  New source
     performance standards for the magnetic tape manufacturing
     industry -revised final tabular costs.

 4.  Memorandum from Angyal, S., MRI, to Magnetic Tape NESHAP
     Project File.  December 9, 1992.  Major Source
     designation/MACT floor determination.

 5.  Economic Indicators.  Chemical Engineering.  .9J3(13) :7.
     June 27, 1983.

 6.  Economic Indicators.  Chemical Engineering.  9_9_(5):202.
     June 1992.

 7.  U. S. Environmental Protection Agency.  Office of Air
     Quality Planning and Standards  (OAQPS) Cost Manual.
     Publication No. EPA-450/3-90-006.  January 1990.


                               3-62

-------
 8.   U.  S. Environmental Protection Agency.  Control of Volatile
     Organic Emissions from Solvent Metal Cleaning.  Publication
     No. EPA-450/2-77-022.  November 1977.

 9.   C&H Distributors (industrial supply company), 1991 catalog.

10.   U.  S. Department of Labor.  Monthly Labor Review.
     April 1992.  p. 69.

11.   U.  S. Department of Energy.  Monthly Energy Review.
     February 1992.  p.  116.

12.   Telecon.  Angyal, S., MRI, with Water Department, City of
     Raleigh, NC.  June 28, 1992.  Cost of water for industrial
     users.

13.   Telecon.  Williams, D., MRI, with Ellam, E., Calgon Carbon
     Corporation.  June 30, 1992.  Cost of carbon for fixed-bed
     carbon adsorbers.

14.   Telecon.  Williams, D., MRI, with Stoehr, D., TIGG
     Corporation.  June 30, 1992.  Cost of carbon for fixed-bed
     carbon adsorbers.

15.   Telecon.  Williams, D., MRI, with Focal, T., Enviatrol.
     June 30, 1992.  Cost of carbon for fixed-bed carbon
     adsorbers.

16.   Reference 11.  p. 119.

17.   Peters, M., and K.  Timmerhaus.  Plant Design and Economics
     for Chemical Engineers.  New York, McGraw-Hill Book Company.
     1980.  pp. 390-392.

18.   Richardson Engineering Services,  Inc.  Process Plant
     Construction Estimating Standards.  1983-1984 Edition.
     Volumes 1 and 3.

19.   Neveril, R.B., GARD, Inc.  Capital and Operating Costs of
     Selected Air Pollution Control Systems.  U. S. Environmental
     Protection Agency.   Research Triangle Park, NC.  Publication
     No. EPA-450/5-80-002.  December 1978.

20.   Memorandum and attachments from Angyal, S., MRI, to the
     Magnetic Tape NESHAP Project File.  December l, 1992.
     Calculation of baseline emissions from the magnetic tape
     manufacturing industry.

21.   Letter and attachments from Falco, M., 3M, to Weigold, J.,
     EPA:BSD.  June 13,  1991.  Section 114 information request
     response for the Camarillo, CA; Hutchinson, MN; Tucson, AZ;
     and Weatherford, OK facilities.
                               3-63

-------
22.  Telecon.  McManus, S., MRI, with Grissett, G., Ashland
     Chemical Company.  June 18, 1992.  Solvent prices.

23.  Telecon.  Angyal, S., MRI, with Carothers, D., Southland
     Environmental Systems.  July 20, 1992.  Liquid waste
     disposal costs.

24.  Telecon.  Angyal, S., MRI, with Heitzman, T., Chemical Waste
     Management.  July 21, 1992.  Liquid waste disposal costs.

25.  Memorandum from Angyal, S., MRI, to Lacy, G., EPA/SDB.
     December 30, 1993.  Format of the Standard, Enhanced
     Monitoring Requirements, and Costs.

26.  U.S. Environmental Protection Agency.  Equipment Leaks
     Negotiated Regulation.  Washington, DC.  U.S. Government
     Printing Office  (Federal Register Notice).  March 6, 1991.
     (29.) 44:9315.

27.  Fugitive Emission Sources of Organic Compounds--Additional
     Information on Emissions, Emission Reductions, and Costs
     (Section 2).  EPA-450/3-82-010.  U.S. Environmental
     Protection Agency, Research Triangle Park, North Carolina.
     April 1982.

28.  Richardson Engineering Services, Inc.  Richardson Process
     Plant Construction Estimation Standards.  Mechanical and
     Electrical.  Volume 3.  Mesa, Arizona.  1988.
     Sections 15-0, 15-42, 15-43, and 15-55.

29.  Memorandum from Hausle, K.J., and D.J. Whitt, Radian
     Corporation, to Markwordt, D., EPA/CPB.  February 28, 1992.
     Final cost impacts analysis for HON equipment leaks.

30.  Memorandum from Whitt, D., Radian Corporation, to
     Markwordt, D., EPA:CPB.  June 5, 1991.  Impacts from the
     control of VHAP emissions from equipment leaks in non-SOCMI
     process units for HON.

31.  Memorandum from Zerbonia, R., RTI, and S. York, RTI, to
     Colyer, R., EPA/SDB.  September 25, 1987.  Draft model unit
     parameters  (accelerated rule) Post-Proposal Analysis.
                               3-64

-------
                      9.0  ECONOMIC IMPACTS


9.1  INDUSTRY PROFILE

9.1.1  Introduction
     Products that will be regulated by the national emissions

standard for hazardous air pollutants (NESHAP) are manufactured

in SIC 3577, Computer Peripheral Equipment, Not Elsewhere

Classified, and SIC 3695, Magnetic and Optical Recording Media.

Firms in these SIC's manufacture the following products:

     SIC 3577:
          Magnetic Stripe Cards
          Other*

     SIC 3695:
          MAGNETIC RECORDING MEDIA:
               Blank Audio Media
                 audio cassettes
                 audio reel tape
                 other audio tape products

               Blank Video Media
                 video cassettes
                 other video tape products

               Blank Computer Recording Media
                 rigid disks*
                 flexible disks
                 computer reel tape
                 computer cassettes
                 computer cartridges
                 other computer recording
                   media

          OPTICAL RECORDING MEDIA*
           (Optical Disks and Tapes)

     * not regulated'
                               9-1

-------
     Certain products will not be profiled because, as noted,
they are not regulated by the NESHAP.  All other product lines
are covered under the NESHAP and therefore will be profiled.  It
should be noted that regulated products classified in SIC 3695
often undergo further fabrication or finishing downstream at
facilities included in other SIC's.  For example, the manufacture
of prerecorded audio tapes is part of SIC 3652.  Establishments
producing prepackaged computer software are classified in
SIC 7372.  Prerecorded video tape and disk production is
classified in SIC 78.  These downstream activities do not involve
coating operations in which solvents containing hazardous air
pollutants (HAP's) are used and therefore are not covered under
the NESHAP.
     The rest of the sections in this profile detail market
information that is relevant to determining the nature and
magnitude of economic impacts resulting from the NESHAP.  To the
extent that information was available, the industry profile
addresses the following market characteristics:  market
structure, production, foreign trade, consumption, pricing, and
industry outlook.  The discussion of the manufacture of magnetic
stripe cards is limited due to a paucity of publicly available
information.
9.1.2  Market Structure
     In 1990, there were 58 companies in SIC 3695; total revenue
was $3,696.8 million.1  The Bureau of the Census publication
County Business Patterns reported that there were 214 facilities
employing 25,539 personnel in SIC 3695 in 1989.2  Of these
establishments, 118  (55 percent) employed 1 to 19 personnel, 39
(18 percent) employed 20 to 99, and 57 (27 percent) employed 100
or more.  However, it is not known how reliable these statistics
are; discussions with an analyst from the Bureau of the Census
revealed that there was duplication in these numbers.^
     It should be noted that not all of the facilities in
SIC 3695 are regulated by the NESHAP.  Many of these facilities
manufacture unregulated products, such as optical recording
media.  Chapter 3 states that there are 25 existing magnetic
                               9-2

-------
recording media and magnetic stripe and facilities that will be
covered by.the standard.  These 25 facilities are owned by
21 companies.
     Conversations with industry contacts indicated that
17 facilities have closed since 1988.4~6  These facilities were
owned by both small, independent companies and large
multinational corporations.  Competition from foreign producers,
who have lower production costs (primarily labor costs) than
domestic producers, was blamed.4"6  The industry contacts also
noted that both domestic and foreign competitors have been
consolidating their operations.
     Table 9-1 presents value of shipments in SIC 3695 from 1986
to 1990.  Shipments increased from $3,672.8 million in 1989 to
$3,696.8 million in 1990.  In real dollars, though, the value
declined by 3.3 percent in 1990 from a 5-year high of
$3,823.8 million in 1989.

      TABLE 9-1.  VALUE OF SHIPMENTS, 1986-1990:  SIC 3695,
              MAGNETIC AND OPTICAL RECORDING MEDIA1

1990
1989
1988
1987
1986
Current dollars
3,696,800,000
3,672,800,000
3,149,700,000
3,226,300,000
2,964,900,000
1990 dollars3
3,696,800,000
3,823,841,749
3,414,679,098
3,613,687,276
3,426,045,759
  aThe data are normalized using the implicit price deflator
   for GNP.
     Tables 9-2 through 9-6 detail historical trends for the
product groups under consideration.  (For Tables 9-3, 9-5, and
9-6, value of shipments data are not available for years prior to
1986) .   As can be seen in Table 9-2, the real value of flexible
computer disks shipped grew for two consecutive  years to reach
$667.6 million in 1990.  However, this was lower than the decade-
high in 1983  ($873.5 million).  Real-dollar shipments of computer

                               9-3

-------
    TABLE 9-2.  VALUE OF SHIPMENTS, 1981-1990:  FLEXIBLE
                       COMPUTER DISKS1

1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
Current dollars
667,648,000
593,377,000
444,783,000
N.A.
N.A.
453,963,000
561,537,000
690,163,000
458,297,000
277,336,000
1990 dollarsa
667,648,000
617,779,282
482,201,865
N.A.
N.A.
538,317,325
685,637,363
873,513,479
602,625,904
387,991,046
aThe data are normalized using the implicit price deflator
 for GNP.
N.A. - Not Available.
         TABLE 9-3.   VALUE OF SHIPMENTS,  1986-1990:
              COMPUTER CASSETTES AND CARTRIDGES1

1990
1989
1988
1987
1986
Current dollars
463,343,000
438,621,000
514,779,000
322,564,000
403,704,000
1990 dollarsa
463,343,000
456,659,032
558,086,513
361,294,803
466,494,107
aThe data are normalized using the implicit price deflator
 for GNP.
                              9-4

-------
     TABLE 9-4.
VALUE OF SHIPMENTS. 1981-1990:
         REEL TAPE1
COMPUTER

1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
Current dollars
167,895,000b
234,969,000b
245,051,000b
229,592,000b
229,812,000
221,947,000
179,867,000
116,784,000
40,631,000
34,174,000
1990 dollarsa
I67,895,000b
244,631,963b
265,666,739b
257,159,498b
265,555,812
263,188,664
219,617,827
147,809,138
53,426,693
47,809,177
aThe data are normalized using the implicit price deflator
 for GNP.
bExcludes other computer reel tape including instrumentation
 tape and high density recording tape (HHDR).
         TABLE 9-5.
    VALUE OF SHIPMENTS, 1986-1990:
     BLANK AUDIO TAPE1

1990
1989
1988
1987
1986
Current dollars
361,494,000
337,748,000
382,430,000
285,821,000
245,800,000
1990 dollars3
361,494,000
351,637,689
414,603,209
320,140,009
284,030,506
aThe data are normalized using the implicit price deflator
 for GNP.
                             9-5

-------
cassettes and cartridges (Table 9-3) fluctuated from 1986 to 1989
before increasing by 1.5 percent in 1990 to $463.3 million.
Real-dollar shipments of computer reel tape declined from 1988 to
1990, as shown in Table 9-4.  The decline from 1989 to 1990 was
especially precipitous: from $244.6 million to $167.9 million
(-31.3 percent).  Current value and real value peaked in 1988 at
$245.0 million and $265.7 million, respectively.  In 1990, the
real value of blank audio tape shipments (Table 9-5) grew
2.8 percent to $361.5 million.  Blank audio tape shipments also
peaked in 1988  ($382.4 million in current dollars and
$414.6 million in real dollars).

           TABLE 9-6.   VALUE OF  SHIPMENTS,  1986-1990:
                        BLANK VIDEO TAPE1

1990
1989
1988
1987
1986
Current dollars
1,095,054,000
1,027,444,000
674,419,000
795,584,000
741,228,000
1990 dollars3
1,095,054,000
1,069,697,033
731,156,765
891,111,111
856,514,906
  aThe data are normalized using the implicit price deflator
   for GNP.
     The $1,095.0 million real dollar value of blank video tape
shipments in 1990 represented a 2.4 percent increase over 1989
(Table 9-6).  Blank video tape shipments accounted for the
largest share of industry value of shipments in 1990 among the
products under consideration.  In current dollars, the five
product groups accounted for the following percentages of the
industry's total value of shipments in 1990: flexible computer
disks, 18.1 percent; computer cassettes and cartridges,
12.5 percent; computer reel tape, 4.5 percent; blank audio tape,
9.8 percent; blank video tape, 29.6 percent.
     In 1990, the value of all magnetic recording media shipped
was $3,041.0 million, 82 percent of SIC 3695's total value of

-------
shipments.1  Thus,  82 percent  of  this  SIC'a  revenue  will  be
impacted by the NESHAP.
     Financial data specific to the  companies  that produce
magnetic recording  media are difficult to  obtain.  However,
survey data reported in Dun and Bradstreet's Industry Norms and
Key Business Ratios for SIC 3695  can be used to  represent the
average financial position of  producers in the industry.7  These
data may be representative of  the source category because the
products that will  be impacted by the  NESHAP account for
82 percent of the total value  of  shipments in  SIC 3695.   In 1991,
the average-sized firm in SIC  3695 had net sales of  $940,519 and
total assets of $489,517.  Net profit  after  tax was  $92,171
 (9.8 percent of net sales).  Long-term debt  ($48,952) accounted
for 10 percent of total liabilities  plus net worth,  while net
worth  ($255,038) accounted for 52.1  percent.
     Activities other than the manufacture of  the recording
medium  (e.g., tape) may occur  at  many  facilities in  the magnetic
recording media industry.  These  other activities include audio
tape prerecording,  record production,  compact  disc production,
injection molding,  assembly of plastic parts such as cassettes,
leader tape production, production of  graphite liners for
cassettes, and production of plastic film with adhesive coating.
These activities suggest that  there  is  a fair  amount of vertical
integration.
     Figure 9-1 outlines the vertical,  relationships  in the
industry.  Magnetic recording  media  can be either intermediate  or
final goods.  In Stage 1, for  example,  coated  blank  video tape  is
manufactured as an  intermediate good.   (Stage  1 is the level at
which some facilities employ regulated  coating operations).  With
further fabrication in Stage 2, blank video  tape can be sold
either as a final good (e.g.,   a blank video  cassette sold
directly to consumers)  or as an intermediate good to be used in a
downstream manufacturing process  (e.g., prerecording) in  Stage  3.
While a good deal is known about vertical relationships,  the
extent of horizontal integration  is unknown.
                               9-7

-------
 STAGE 1
                  Manufacture of
             magnetic recording medium
             (includes coating operations),
                e.g., blank video tape
                                  intermediate
                                     good
 STAGE 2
 STAGES
 STAGE 4
                  Manufacture of
                  blank magnetic
                 recording medium
                (fabricated product),
              e.g., blank video cassette
                   intermediate
                      good
   Manufacture of
prerecorded magnetic
  recording medium,
  e.g., prerecorded
   video cassette
                      final
                      good

                     _L
    Consumption*
                                        final
                                        good
Consumption'
* Both consumer and professional (e.g., recording studios) applications.
Figure 9-1.
 Vertical  relationships in the  magnetic
 recording media industry.
                             9-3

-------
     The magnetic recording media industry is spread throughout
the world.  In 1989, more than 300 companies worldwide were
involved in the production of magnetic recording media; only
about a hundred, though, actually coated various magnetic
recording media.8  Two-thirds of these firms were located in
Asia, primarily the Far East.  While the United States was the
world's largest market for magnetic recording media in 1989,
Europe and Japan were also significant markets.
     Table 9-7 lists the world's top 25 producers of magnetic
recording media in 1990.  It is evident that a great deal of
revenue is concentrated among the top six firms: 3M Company
 (United States), TDK Corporation (Japan), Sony Corporation
 (Japan), BASF AG (Germany), Hitachi Maxell (Japan), and Fuji
Photo Film (Japan).  These six firms accounted for almost
51 percent of an estimated $13,586 million in industry revenue in
1990.1:L
     3M is the largest worldwide producer (as well as  domestic
producer) of magnetic recording media, with sales of $1.7 billion
in 1990  (almost half a billion dollars more than its nearest
competitor, TDK Corporation).^"  Four of the six leading
producers are Japanese companies.  The top four Japanese
producers were projected in 1990 to have combined revenues of
$4,177 million as opposed to 3M and BASF's combined
$2,747 million.  Asian producers account for almost 60 percent of
revenue among the top 25 producers.
     The strong presence of Asian firms in the industry is in
large part due to their dominance in the production of downstream
audio and video products.12  For example, TDK was the world's
largest producer of consumer audio cassettes in 1990.l3  However,
not all of the markets served by the industry are dominated by
Asian firms.   Asian firms, particularly Japanese firms, tend to
focus on high-volume consumer products rather than professional
products (with the exception of floppy disks and professional
video media)  because of their ties to the consumer electronics
business, on which the magnetic recording media industry is
highly dependent.
                               9-9

-------
           TABLE 9-7.   WORLD'S TOP 25 PRODUCERS* OF
               MAGNETIC RECORDING MEDIA,  199010
                                     Estimated 1990  revenue,
 Company	106 $b	
 3M Company                                   1,712
 TDK Corporation                              1,258
 Sony Corporation                             1,148
 BASF AG                                      1,035
 Hitachi Maxell                                 986
 Fuji. Photo Film                                785
 SKC/Sunkyong Mag                               410
 Saehan Media                                   356
 Nippon Victor (JVC)                             326
 Agfa-Gevaert AG                                322
 Matsushita Denki                               314
 Ampex Corporation                              285
 IBM Corporation                             •   266
 Seagate Magnetics                              246
 Verbatim/Kasei V                               206
 Lucky Goldstar                                 165
 Nashua Corporation                             161
 Konica Ltd.                                    146
 Komag-Asahi Komag                              131
 PD Magnetics                                   123
 Tandy/Tandy-Memorex                            115
 Graham Magnetics                               111
 KAO Corporation                                105
 MTI/MCC (Memorex)                              103
 Kolon Industries                               102

aOnly those companies which coat magnetic recording media
 are included.
"Revenue figures are preliminary estimates made in
 September 1990.
                             9-10

-------
     Many Western firms are well positioned with respect to
magnetic recording media used in professional applications  (e.g.,
recording studios).  For example, the market for data recording
products (computer tape, cartridges, disks, etc.) was estimated
to account for a third of industry revenues in 1990 and Asian
firms were relatively minor players in this segment.12  In 1990,
3M was the leading producer of flexible disks and all forms of
computer tape and cartridges.13  The 3M controlled over
75 percent of the world market for these products, according to
Datamation.14  BASF AG and Graham Magnetics were major players in
the computer tape market.1^  In 1990, Ampex was the world's
largest producer of professional audio tape.  Ampex has been
described as having monopoly power in the instrumentation tape
market (a type of computer tape used primarily by Government
defense agencies).1°  In 1988, Magnetic Media Information
Services reported that Ampex controlled at least 95 percent of
the market.16
     As previously mentioned, published information concerning
the magnetic stripe card industry is scarce.  However,
discussions with an industry trade association revealed that the
industry -is comprised of two market segments.17  The first
segment is the market for magnetic stripe cards in ticketing
(e.g., transit)  applications.  According to the trade
association, Rand McNally is the dominant competitor in this
segment.1   The second segment is the market for magnetic stripe
cards as financial transaction cards'(e.g., credit cards).  Major
competitors include Malco Plastics as well as the Datacard
Corporation.17
     Information about the horizontal relationships in the
magnetic stripe card industry is not available.  It is known that
companies manufacturing magnetic stripe cards are not vertically
integrated beyond the manufacture of the cards as intermediate
goods.17  End users (e.g., banks) purchase magnetic stripe cards
as inputs in the "production" of the various services they offer.
For example, a bank uses a credit card as a part of the financial
service package it offers to a customer.  A transit ticket (e.g.,
                               a _

-------
a monthly subway pass) is part of the transportation service
offered by a transit agency.
9.1.3  Production
     Table 9-8 reports the year-to-year change in unit output
from 1982 to 1990 for selected magnetic recording media.   (While
year-to-year changes in unit output can be presented, it is not
possible to present actual levels of unit output because the
Bureau of the Census does not define its standard unit).  The
largest percentage change in output occurred in 1982 for flexible
computer disks (224.5 percent).  Growth probably followed the
trend in the computer industry, which experienced tremendous
growth in the early part of the decade.  Another example of
explosive growth is the 78.9 percent increase in computer reel
tape output in 1984.  The largest percentage change from 1989 to
1990 was for computer cassettes and cartridges; quantity shipped
grew 121 percent.  Computer reel tape was the only product whose
shipments declined in 1990  (-11.9 percent).  The quantity shipped
of blank video tape, flexible computer disks, and blank audio
tape increased in 1990 by 14.5 percent, 13.1 percent, and
7.3 percent, respectively.
     It is important to note that a number of foreign producers
have manufacturing facilities located in the United States.
Output figures therefore include production from foreign-owned
manufacturing facilities in the United States.
     No information concerning capacity, either industry-specific
or company-specific, is available.  However, it is known that the
capacity utilization rate in SIC 3695 declined from 84 percent in
1989 to 82 percent in 1990  (a 2.4 percent decline).18  The
decline in capacity utilization seems to contradict the growth in
output reported in Table 9-8.  However, the decline could be
attributable to capacity expansion.
     In Table 9-9, factory shipments of blank audio and video
cassettes are presented.  Blank audio cassette shipments grew by
4.8 percent in 1990; shipments of blank video cassettes grew
13.2 percent in the same year.  Both of these growth rates are
                               9-12

-------
I
u
w
w
CO
  o
  a\
  a\
O fs
  oo
W en
CO
pa










~
g.
3
0
'S
1
U
!
Y
£
















8,
1
>
1
5
1
_o

a
1
CO



8.
1
1

1 S
en so
a T3
o 'C
I i
3
n
o
u
computer
isks
.U -o
^
u
itl




1
!<<«M«M<
Jrzzzzzzz



i
O**'!'J*

^•zzzzzzz
en
t-'





§ i i 1 1 § 1
S,S,a.S,8.8,8.<
gv-^O^oOONONZ
= g 8 S J5 H JS

~ *~> ~ -z?
4J ^ u P
P ^ O o

0-H.S.S.'"1:'^
q.Jvo^ZZZZ

"1 X CS X
cj S- «n g


g S f 1
!!<<<< 1 i
— enZZZZvoO
S3 & » S
SI* c5


^ 00 OO OO OO OO 00 00
ON ON ON ON ON ON ON ON


<
Z





z






1
1
o
ON





Z




1
"5
fS

cs
00
ON

                                                    i
                                        9-13

-------
comparable to the growth rates exhibited by the larger product
groupings, blank audio tape and blank video tape, in Table 9-8

   TABLE  9-9.   FACTORY SHIPMENTS,  1986-1990:   BLANK AUDIO AND
                        VIDEO CASSETTES19

1990
1989
1988
1987
1986
Blank audio
103 Units
440,799
420,512
394,506
419,242
323,680
cassettes
Percent
change
4.8
6.6
(5.9)
29.5
N.A.
Blank video cassettes
103 Units
381,269
336,670
340,568
320,415
367,391
Percent change
13.2
(l.l
6.3
(12.8)
N.A.
  N.A.  -  Not available.
  NOTE:   These products  are part of the product categories
         detailed in Table 9-8.
     The following raw materials are used in the production of
magnetic recording media: plastic films, plastic resins', aluminum
and other metals, solvents, and magnetic pigments. °  Many of
these raw materials are petroleum-based products; thus, the cost
of magnetic recording media production is ultimately tied to the
price.of oil.21  The industry uses a great deal of polyester
film.  In 1989, the tonnage used to produce magnetic recording
media represented 25 to 30 percent of worldwide polyester film
production.20  Many producers are vertically integrated into the
raw materials market.  For example, many produce their own
magnetic pigment, a raw material specific to the industry.22
 (The production of magnetic recording media accounts for the
majority of magnetic pigment consumption).  Worldwide raw
material inputs into the manufacture of magnetic recording media
were approximately $2.5 billion in 1989.20  This represents
18 percent of projected worldwide revenue in 1990 (approximately
$13.6 billion).10
                               9-14

-------
     There is no published information available concerning the
production characteristics of magnetic stripe card facilities.
9.1.4  Foreign Trade
     Tables 9-10 and 9-11 show exports and imports in SIC 3695
for 1989 and 1990.  From 1989 to 1990, exports increased
35.5 percent to $1,380.6 million.  This was approximately
37.3 percent of the total value of shipments in the U.S.  Imports
were larger in value than exports.
     Their value grew 6.6 percent in 1990 to $1,566.3 million
(40.3 percent of domestic apparent consumption).  Thus, imports
control a sizable share of the U.S. market.  The trade balance in
1990 was a deficit of $185.7 million.
     Product-specific export data are reported in Table 9-12.
While all of the selected products exhibited considerable growth
from 1989 to 1990, the largest change (53.8 percent)  occurred in
blank video tape exports.  However, blank video tape exports
accounted for only 12 percent of the total value of blank video
tape shipped in 1990.  In contrast, the $698 million in exports
of flexible and rigid magnetic recording media for computers
accounted for 53 percent of total shipments of these media.   (As
noted in the table, rigid magnetic recording media for computers
are not regulated by the NESHAP).
     Computer cassette and cartridge imports increased
19.4 percent in 1990 (Table 9-13).  Imports of flexible and rigid
magnetic recording media for computers increased 11.8 percent.
Imports of blank audio and video tape decreased by 15.1 percent
and 0.1 percent, respectively.  Imports as a percent of 1990
domestic apparent consumption ranged from 34 percent for blank
audio tape to 19 percent for computer cassettes and cartridges.
     Discussions with an analyst from an industry trade
association revealed that there is virtually no foreign
competition in the magnetic stripe card industry.17  The market
for magnetic stripe cards in ticketing applications is dominated
by domestic firms because transit agencies, for example, give
preference to American firms; the agencies are government
                               9-15

-------
     TABLE 9-10.   VALUE OF EXPORTS,  1989-1990:  SIC 3695,
            MAGNETIC AND OPTICAL RECORDING MEDIA1
Current dollars
1990
1989
1988
1987
1986
1,380,
1,018,
N
N
N
600,000
817,000
.A.
.A.
.A.
Percent of total
value of shipments
37.3
27.7
N.A.
N.A.
N.A.
N.A. - Not available.
     TABLE 9-11.   VALUE OF IMPORTS,  1989-1990:   SIC 3695,
            MAGNETIC AND OPTICAL RECORDING MEDIA1
               Current dollars
   Percent of U.S.
apparent consumptionc
1990
1989
1988
1987
1986
1,566,300,000
1,468,376,000
N.A.
N.A.
N.A.
40.3
35.6
N.A.
N.A.
N.A.
aSee Section 9.1.5.

N.A. - Not available.
                             9-16

-------
        TABLE  9-12.   VALUE OF EXPORTS,  1989-1990:  SELECTED
                       MAGNETIC RECORDING  MEDIA1

Flexible and rigid
magnetic recording
media for computers3
Computer cassettes
and cartridges
Blank audio tape
Blank video tape
Current
1989
542,256,000
76,888,000
53,144,000
83,524,000
dollars
1990
698,800,000
88,100,000
60,100,000
128,500,000
Percent change
28.9
14.6
13.1
53.8
Exports as
a percent of value
of shipments, 1990
53.0
19.0
17.0
12.0
  alncludes rigid disks used in Winchester (hard-disk) drives and optical media; these products are not
   regulated by the NESHAP.
   TABLE 9-13.  VALUE  OF IMPORTS,  1989-1990:   SELECTED  MAGNETIC
                             RECORDING MEDIA1
                           Current dollars
                        1989
               1990
                          Imports as a percent
                             of apparent
             Percent change   consumption'7, 1990
   Flexible and rigid
   magnetic recording
   media for computers4
   Computer cassettes
   and cartridges
   Blank audio tape
   Blank video tape	
261,480,000   292,300,000
 72,440,000

179,168,000
345,540,000
 86,500,000

152,100,000
346,000,000
 11.8

 19.4

(15.1)
 (0.1
32.0

19.0

34.0
26.0
  alncludes rigid disks used in Winchester (bard drives) drives and optical media; these products are not
   regulated by the NESHAP.
  "See Section 9.1.5.
entities.   The  domestic market  for magnetic stripe cards as
financial  transaction cards has no foreign competition.
9.1.5  Consumption
      Domestic apparent consumption is defined as domestic
production plus imports minus exports (assuming  no change  in
inventory).   In current dollars,  apparent  consumption declined
from $4,122.4 million in  1989 to $3,882.5  million in  1990,  a
                                    9-17

-------
change of 5.8 percent.   Real  dollar consumption totalled
$3,263.9 million in  1989 and  $2,952.5 million in 1990.x
     Domestic apparent  consumption for selected magnetic
recording media is shown in Table 9-14.   Consumption of blank
video tape jumped 65 percent  from 1989 to 1990.  Consumption of
flexible and rigid magnetic recording media for computers grew
28.8 percent while consumption of computer cassettes and
cartridges grew 12.3 percent.   Only the consumption of blank
audio tape decreased from 1989 to 1990 (-15.4 percent).  No
published consumption data for magnetic stripe cards are
available.

        TABLE 9-14.   U.S. APPARENT CONSUMPTION  OF SELECTED
               MAGNETIC RECORDING MEDIA,  1989-19901

Flexible and rigid magnetic
recording media for computers2
Computer cassettes and cartridges
Blank audio tape
Blank video tape
Current
1989
712,053,000
410,974,000
535,983,000
794,014,000
dollars
1990
916,900,000
461,700,000
453,500,000
1,312,600,000
Percent change
28.8
12.3
(15.4)
65.0
  alncludes rigid disks used in Winchester (hard drives) drives and optical media; these products are not
  regulated by the NESHAP.
     The demand  for magnetic recording media is directly related
to the consumption of  the  electronic hardware in which they are
used  (e.g., computers,  VCRs,  audio tape players/recorders).   In
other words, magnetic  recording media products and the hardware
in which they are used are complementary goods.  Table 9-15
reports the shipment data  previously shown in Table 9-9 for blank
audio cassettes  and compares them to shipment data for audio  tape
equipment  (home, portable,  and automotive recorders and players).
Changes in hardware shipments were mirrored by blank audio
cassette shipments except  for 1990.  In Table 9-16, a similar
comparison is made between blank video cassette shipments and

                                9-13

-------
   TABLE 9-15.  FACTORY SHIPMENTS  OF BLANK AUDIO CASSETTES AND
                AUDIO TAPE  EQUIPMENT,  1986-199019


1990
1989
1988
. 1987
1986
Blank
103 Units
440,799
420,512
394,506
419,242
323,680
audio cassettes
Percent change
4.8
6.6
(5.9)
29.5
N.A.
Audio
103 Units
68,162
74,960
'60,800
61,200
48,457
tape equipment
Percentage change
(9.1)
23.3
(0.7
26.3
N.A.
  N.A. - Not available.
    TABLE 9-16.  FACTORY SHIPMENTS  OF BLANK VIDEO CASSETTES
            AND VIDEO CASSETTE  RECORDERS,  1986-199O19


1990
1989
1988
1987
1986
Blank
103 Units
381,267
336,670
340,568
320,415
367,391
video cassettes
Percent change
13.2
(1.1)
6.3
(12.8)
N.A.
Video cassette
103 Units
13,079
12,076
12,733
13,231
13,533
recorders
Percentage change
8.3
(5-2)
(3.8)
(2.2)
N.A.
  N.A. - Not available.
videocassette recorders.  Blank video  cassette shipments mirror
videocassette recorder shipments  except  for 1988.       The price
elasticity of demand for magnetic recording media  is most likely
relatively inelastic.  The choice of which  magnetic recording
product to use is primarily influenced by the  choice of hardware.
Hardware decisions may sometimes  involve a  preference for certain
magnetic recording media.  However, once a  hardware decision has
been made, no substitution between the various magnetic recording
media is possible.  For example,  if one  buys a notebook computer
with a 3.5-inch drive, then optical disks or 5.25-inch floppy
                               9-19

-------
disks cannot be used.  The only feasible choice is a 3.5-inch
disk.  Because one cannot readily substitute among the various
magnetic recording media without entailing costs for substitute
hardware, demand will tend to be price- inelastic.  Another
reason why demand will tend to be price- inelastic is that
magnetic recording media probably account for a small percentage
of users' budgets.  For example, the cost of.floppy disks for a
                              •       »
computer most likely accounts.for a small percentage of the
computer user's budget, which includes the cost of the hardware
(e.g., the computer).
     According to Marshall's rules, the demand elasticity of an
input varies directly with (l)  the demand elasticity of the
product or service provided with the input; and (2) the share of
the input in the cost of production.^  The products manufactured
with magnetic recording media (e.g., prerecorded music cassettes,
prepacked software) are probably more unitary elastic than price
inelastic.  However, the percentage of manufacturing costs that
magnetic recording media account for is probably small.  Thus, we
can assume that the demand for magnetic recording media as
intermediate goods is relatively price inelastic.
     Much like the demand for magnetic recording media, the
demand for magnetic stripe cards if probably price inelastic.
The services provided with magnetic stripe cards (e.g., banking
services) have relatively price inelastic demand.   For example,
there is no convenient substitute for the services provided by a
bank.  Also, the cost of a credit card, for example, is a small
part of a bank's production costs.  These same characteristics
apply to other uses of magnetic stripe cards, such as transit
tickets and airline tickets.  Thus, it can be assumed that the
demand for magnetic stripe cards is price inelastic.
     Two new technologies that may eventually supplant.magnetic
recording media are optical recording media and solid state
recording (SSR) media.    Optical recording devices use lasers to
record electrically encoded information onto a reflective disk.
Currently, optical disks are used to store extremely large
amounts of data, much like computer tape and cartridges.  Optical
                              9-20

-------
recording devices are generally much more expensive than current
technologies for recording purposes, and have therefore not been
used extensively.  Solid state recording devices eliminate the
need for mechanical devices  (to move tape or rotate a disk) by
using semiconductor memory, for example.24  This technology is in
the beginning stages of development.
9.1.6  Pricing
     The most accurate gauge of product pricing is average
realized price per unit.  It is defined as the value of shipments
divided by total unit output.  Average realized price is more
accurate than list price because it accounts for product
discounts.  Because standard sizes for the products under
consideration are not specified by the reference source (Bureau
of the Census), average realized price cannot be reported for a
defined unit.  However, percent changes in average realized price
per unit are meaningful and therefore can be reported.  Year-to-
year changes in average realized price per unit (price
realization) for selected magnetic recording media are reported
in Table 9-17.  No price data are available for magnetic stripe
cards.
     The ratio declined from 1989 to 1990 for all five product
groups reported in the table, with the greatest decline
experienced by computer cassettes and cartridges.  This is most
likely attributable to the 121 percent growth in output in 1990.
Audio tape and flexible computer disks experienced slight
declines in price realization from 1989 to 1990.  These declines
are most likely due to increases in output (refer to Table 9-8).
9.1.7  Outlook
     Magnetic Media Information Services forecast that the
worldwide magnetic recording media market will be worth
$18.1 billion (1989 dollars)  by 1996.2S  This represents a
49.5 percent increase (8.2 percent growth annually)  from their
estimate of 1990 revenue.  Magnetic recording media will prosper
because they provide a low cost, reliable means of making a
recording.26  However,  by the late 1990s,  production of these
products will have reached their peaks, and they will
                               9-21

-------
 TABLE 9-17.  PERCENT CHANGE IN PRICE REALIZATION3:
             MAGNETIC RECORDING MEDIA, 1982-199O1
SELECTED

1990
1989
1988
1987
1986
1985
1984
1983
1982

Flexible
computer disks
(1-2)
N.A.
N.A.
N.A.
N.A.
N.A.
9.2
(50.1)
(49.1)

Computer
cassettes and
cartridges
(50.8)
(14.6)
(5.0)
101.1
(61.7)
15.1
(29.9)
109.6
15.3
aPrice realization is computed by dividing value
changes are presented in lieu of actual realized
the Bureau of the Census.
Year-to-year change
Computer reel
tape
(18.9)
. 55.5
(47.2)
0.1
(19.0)
19.8
(13.9)
N.A.
9.1

Blank
audio tape
(0.2)
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
of shipments by number of units shipped.
prices because standard unit sizes were not

Blank
video tape
(7.0)
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
Percent
reported by
N.A. - Not available.
                              9-22

-------
increasingly be replaced by optical recording media, particularly
in prerecorded applications (e.g., prerecorded music and
movies).27  Market value is expected to fall to $17.9 billion by
the year 2000 as optical media and SSR media penetrate the
market.25
     The annual growth rate projected by MMIS of over 8 percent
may be optimistic.  The yj.Sj. Industrial Outlook 1992 forecasts
that shipments of consumer electronics will grow 2.5 percent per
year between 1992 and 1997.28  Disk/Trend Inc., a California-
based market research firm, estimates that shipments of floppy
disk drives will grow 3 percent annually from projected 1991
levels to 1993.29  Because it is probable that there is a strong
relationship between magnetic recording media and the electronic
hardware in which they are consumed, it is more likely that
annual growth of magnetic recording media products will be
closely related to the moderate annual growth projections for
consumer electronics and floppy drives.
     During the 1990s, MMIS expects the major American, Japanese,
and European producers (3M, TDK Corporation, Sony Corporation,
BASF AG, Hitachi Maxell,  and Fuji Photo Film) to shift their
emphasis to higher-margin, higher value added magnetic recording
media (e.g., 8mm video cassettes).30'31  These firms will no
longer manufacture the standard grades of audio, video, and
computer recording media.  As a result, the Asian firms other
than the Japanese will become more prominent.32  Korean firms
will supply a considerable percentage of standard grade media of
all types to the six firms.32  A large percentage of molding and
fabrication of plastic parts as well as loading of tape
(inserting the tape into its shell) will shift to low-cost labor
areas- such as southern China,  Malaysia, and Thailand.30'32'33
Component and raw material production will be controlled by Asian
producers to a greater extent than in the 1980s.33
9.2  ECONOMIC IMPACT ANALYSIS
9.2.1  Methodology
     The structure and sequence of the economic impact analysis
is depicted in Figure 9-2.  The analysis is designed to assess
                               9-23

-------
     2
     I


                         "•gas" £
                                                            Cn
                                                            O
                                                           rH
                                                            O
                                                           T3
                                                            O
                                                           X!
                                                           4J

                                                            I

                                                            QQ
                                                           •H
                                                            CQ
                                                           rt
                                                           a
                                                           m
                                                           j_)
                                                           u
                                                           u
                                                           -H

                                                           I
                                                           a
                                                           o
                                                           O
                                                           CT\

                                                           0)
                                                           ^

                                                           01
                                                           -H
il
C. ffl
                    n
                          9-24

-------
economic impacts at the industry level and the facility level.
Note that impacts are not assessed at the industry level for.
magnetic stripe card manufacturing due to a lack of aggregate
industry data.
     9.2.1.1  Industry-Level Analysis.  Emission control costs
are available for the entire magnetic recording media industry
only, and not for the impacted market segments delineated in the
industry profile.  Therefore, industry-level impacts may mask
differences between the market segments.
     Figure 9-3 illustrates the effects on an industry of having
to install emission control equipment.  The implementation of
emission control techniques leads to increased costs of
production for the regulated firms in the industry.  These
increased costs are reflected in a leftward and upward shift of
the industry supply curve, from SQ to S1.  (The industry supply
curve is by definition the horizontal summation of the supply
curves of all firms in the industry).  The leftward shift
represents a decrease in supply, which in turn leads to a higher
average price for the industry  (P1 instead of PQ).  The vertical
distance cd is equal to the industrywide annualized cost of
installing and operating emission control equipment.  It
represents the price increase necessary for the regulated portion
of the industry to fully recover annualized control costs.
     The first step in the analysis is to calculate the value of
the price increase.  To do so, total annualized control costs for
a regulatory alternative attributable to the manufacture of
magnetic recording media are divided by the revenue of SIC 3695
attributable to these media.
     The ability of the regulated portion of the industry to
achieve the price increase necessary to fully recover annualized
control costs is constrained by the unregulated portion.
Unregulated producers will manufacture the same magnetic
recording media at costs unaffected by the NESHAP.  The actual
price increase that will occur is a function of a variety of
factors, including market concentration,  the extent of vertical
and horizontal integration, and the number and size of firms.
                               9-25

-------
                                        t
X	
is
li
                                         o
G

I
•H
O1
                                                     JJ
                                                     G
                                                     O
                                                     U

                                                     G
                                                     O
                                                     •H
                                                     4J
                                                     Cn
                                                     G
                                                     •H
                                                     4J
                                                     tn
                                                     G
                                                     CO
                                                     iJ
                                                     CJ
                                                     0)
                                                     4-4
                                                     0)
                                                     M
                                                     3
                                                     tn
§
                      9-26

-------
Unfortunately, many aspects of pricing dynamics in the industry
are unknown.  However, it can be assumed that regulated
facilities that are not price leaders may have to absorb
that percentage of control cost per unit that exceeds the
industry-level price increase.
     The strength of foreign competition is another important
factor in determining the degree to which unregulated magnetic
recording media producers will constrain the industry-level price
increase.  Foreign producers are not subject to the NESHAP.  In
the industry profile, it was reported that imports accounted for
significant percentages of 1990 domestic apparent consumption of
magnetic recording media (refer to Table 9-13).  The implication
is that there is a good deal of import price elasticity.  Thus,
we can-expect that as long as transportation costs are not
prohibitive, foreign producers will be able to supply magnetic
recording media at lower prices.  In order to avoid increased
import competition, regulated domestic producers may not raise
prices to fully recover emission control costs.
     The increase in price will result in a decrease in quantity
demanded unless demand is perfectly inelastic.  Because the
demand for magnetic recording media is price inelastic, a
decrease in quantity demanded can be expected.  In"the short run,
the reduction in quantity demanded could be large enough to lead
to the closure of one or more marginal firms.  A marginal firm is
manufacturing products at the highest average cost in the
industry; the implication is that it is not using the most
technologically efficient process.  Even if the reduction in
quantity demanded is not large enough to lead to closure in the
short run, cost increases could diminish profitability so that
capital will be redeployed in the long run.  In this case,
marginal firms could exit,  and more technologically efficient
ones could replace them.
     The industry's percent reduction in quantity demanded is
calculated as the product of the percent price increase and the
price elasticity of demand.  This procedure is derived from the
definition of the price elasticity of demand.
                               9-27

-------
                            E
                            E
where:
                E »  price elasticity of demand
     percent A QD »  percentage change in quantity demanded
      percent A P »  percentage change in price
Rearranging the formula:
                           %AQD = %P x E

     The price elasticity of demand measures the percent change
in the quantity demanded resulting from a 1 percent change in
price.   The more inelastic demand is, the greater the ability of
producers to increase price without losing output.  Conversely,
relatively elastic demand restricts the ability to increase
prices without losing output.
     Magnetic recording media can be either consumer goods or
factors of production.  Therefore, in order to estimate the price
elasticity of demand, we must evaluate the determinants of price
elasticity for consumer goods and the determinants for factors of
production.  The determinants for consumer goods relevant to the
analysis include the following:
     1.  Nature of the goods — necessities, such as basic
transportation, have inelastic demand curves, while luxury goods,
like weekend flights to Bermuda, have elastic demand curves.
     2.  Availability of close substitutes — goods that can be
easily substituted for have elastic demand curves.
     3.  Fraction of income absorbed — goods that absorb only a
small fraction of income, such as salt,  have inelastic demand
curves.  Conversely, goods that absorb a large fraction of
income, such as an automobile, have elastic demand curves.
     Based on these determinants, it can be assumed that the
demand for magnetic recording media as consumer goods is
relatively price inelastic.  While magnetic recording media are
not necessities, there are no readily available substitutes for a
particular magnetic recording medium.  Substitution would entail

                               3-28

-------
buying new hardware compatible with the substitute magnetic
recording medium, a costly undertaking.  Another indication that
demand is inelastic is the fact that the magnetic recording
medium (e.g., a blank consumer audio tape) most likely accounts
for a small percentage of the consumer's income.
     Demand for factors of production is said to be derived,
i.e., the need for factors derives from the demand for the
products which are produced by the factors.  Determinants of the
price elasticity of demand for factors of production are similar
to those for consumer goods, and are concisely defined by Alfred
Marshall.  They are known as Marshall's rules.  Two of them are
particularly relevant to this analysis.  Demand elasticity varies
directly with the following:
     1.  The demand elasticity for the product the factor
produces.
     2.  The share of the factor in the cost of production.
     The demand for products produced with blank magnetic
recording media is most likely unitary elastic or slightly
elastic.   For example, the demand for prepackaged computer
software is probably close to unitary elasticity.  Prepackaged
software represents a substantial percent of the computer user's
budget.  Nevertheless, an accountant, for example, views a
"spreadsheet" program as a necessity.  In many cases, certain
prepackaged software is specifically needed to accomplish certain
tasks  (e.g., complicated numerical calculations performed by
statistical software).  With respect to the second determinant,
the magnetic recording medium most likely accounts for a
small percentage of the producer's budget.  For example, blank
audio mastering tape probably accounts for a small percentage of
a recording studio's production costs.
     Another consideration is that the demand for factors of
production generally tends to be inelastic in the short-run.
This is due to the fact that firms are "locked into" a technology
process,  and speedy substitution is difficult.  Thus, considering
this point and the aforementioned determinants,  it is likely that
                               9-29

-------
the demand for magnetic recording media as factors of production
is slightly price inelastic.
     In general, it can be assumed that the price elasticity of
demand for magnetic recording media is in the inelastic range
between -1 and -0.5.  An estimate of -0.75 is used to calculate
the reduction in quantity demanded.
     Any change in the industry's output will affect employment.
A linear relationship between output and employment is assumed
due to a lack of published material indicating otherwise.  Thus,
the industry's percent reduction in quantity demanded also
represents the percent reduction in employment.
     Both the percent increase in price and percent reduction in
quantity demanded result in a change in industry revenue.  The
change is calculated as the sum of the percent price increase,
the percent reduction in quantity demanded, and their cross
product.  The general formula:
                %ATR = [(%AP + %AQD) * (%AP x %AQD) ]

where:
     percentTR = percent change in total industry revenue
     percentAP = percent change in price
    percentAQD = percent change in quantity demanded
     9.2.1.2  Facility-Level Analysis.   The facility-level
analysis is designed to assess impacts on those facilities
regulated by the NESHAP.  Three types of magnetic recording media
manufacturing facilities are impacted.   The first type includes
facilities that manufacture coated magnetic recording media
(e.g., blank video cassettes) as intermediate goods and use HAP
solvents in their production.  These Stage l facilities are the
ones actually regulated by the NESHAP.
     The next two types of facilities are impacted by the NESHAP
if they use, as raw-material inputs,  magnetic recording media
that have been coated using HAP solvents.  They include, in
Stage 2, facilities that manufacture fabricated magnetic
recording media (e.g., blank video cassettes), which can be
                               9-30

-------
either intermediate or final goods.  Also included are facilities
in Stage 3 that produce prerecorded media as final goods.
     Impacts are measured under the assumption that regulated
facilities are fully integrated through Stage 2 and produce final
goods using regulated raw material inputs.  The magnitude of a
regulated facility's impacts resulting from the regulation of its
Stage l operations will depend on its ability to increase prices
at Stage 2.  While a facility's Stage 1 impacts may be
significant, the facility may be able to fully recover emission
control costs if the intermediate good accounts for a
small percentage of the cumulative value added to the final good.
This means that at Stage 2 the facility would not have to
increase product price significantly to recover increased raw-
material costs.  The regulated facility will also be able to
fully recover emission control costs via a price increase at
Stage 2 if there are no lower-cost alternatives for obtaining the
inputs currently produced in Stage 1.  Examples of these
alternatives are substitute manufacturing processes and job shops
with lower production costs.
     Impacts on Stage 3 facilities using regulated raw materials
will not be examined because they are conservatively represented
by impacts on final-good producers at Stage 2,  which has less
cumulative value added.
     For regulated magnetic stripe card manufacturing facilities,
impacts are assessed under the assumption that these facilities
only produce magnetic stripe cards.  The impacts are therefore
the "worst-case" scenario because the facility's revenue basis is
less than if the facility is also vertically integrated into the
manufacture of such products as financial transaction cards.
However, it should be noted that the regulated magnetic stripe
card facility may be able to fully recover emission control costs
through a price increase if magnetic stripe accounts for a
small percentage of the cumulative value added to the final good
(e.g., a credit card).
     The facility-level price increase necessary to fully recover
emission control costs is calculated as the ratio of total
                               9-31

-------
annualized facility control costs for a regulatory alternative to
facility revenue associated with the regulated production.  The
cost of emission control equipment represents increased costs of
production for the regulated facility.
     Revenue information for the regulated facilities is not
available.  Therefore, a regulated facility's revenue basis must
be estimated.  The first step in determining the revenue basis is
to calculate facility output.  Three model lines (small, medium,
large) are used to represent coating operations at a regulated
facility.  The maximum amount of substrate (coated magnetic
recording medium)  that each of the three model lines is capable
of coating is calculated based on the model coating operation
parameters detailed in Table 6-5.  (See Appendix G for the
specific details of the calculations).  The results are the
following:
     Total amount of sgybfltrate that can be coated in year
     Small model line:  1,404,000 m2 (15,112,163 ft2)
     Medium model line:  4,212,000 m2 (45,336,488 ft2)
     Large model line:  35,640,000 m2 (383,616,438 ft2)
     The size and number of coating lines at each of the
regulated facilities have been estimated based on survey
information gathered from the industry.   Using this information
to choose the appropriate size and number of model lines, the
total amount of substrate that a regulated facility is capable of
coating can be determined.  For example, if emission control
costs are calculated assuming that a facility has one small line
and one medium line, then the total amount of substrate that the
facility can coat in a year is 5,616,000 m2 (1,404,000 m2 +
4,212,000 m2).
     However, the total amount of substrate coated is not a
measure of a regulated facility's output of magnetic recording
media and/or magnetic stripe cards.  The total amount of
substrate coated must be divided by the standard unit sizes
(i.e., the amount of coated substrate used per unit) of the
magnetic recording media and/or magnetic stripe cards produced to
determine a facility's unit output.
                               9-32

-------
     Unfortunately, the specific products produced by a regulated
facility are unknown.  However, the market segments  (i.e., blank
audio, video, computer recording media or magnetic stripe cards)
it operates in are known  (Table 3-3) .  Facility output can be
calculated under the assumption that capacity is dedicated to the
production of only one magnetic recording medium or only magnetic
stripe cards.  For example, if we know that a facility is
producing blank computer media, we can use a 3.5-inch flexible
disk as a "model unit" of output.  Dividing the model unit's
dimensions into the total amount of substrate coated by the
facility gives us the total unit output that the facility is
capable of producing.*  This procedure is repeated for each
market segment in which a regulated facility is known to operate;
output is calculated for each of the segment's model units.  The
regulated facility's revenue basis can be estimated by
multiplying a model unit's total output by a retail price per
unit.  Note that by using a retail price the revenue basis is
effectively calculated for Stage 2.
     Since the actual product mix manufactured by a regulated
facility is unknown, the percent price increase is calculated
assuming that the facility is .manufacturing the model unit which
would result" in the least amount of revenue.  This means that
the percent price increase is a worst-case estimate.
     A regulated facility's price increases may be significant if
they are greater than l percent and deviate considerably from the
industry price increase.  The values of l percent and the
industry price increase are used as screening values to isolate
impacts that may be significant.  Because the industry-level
price increase is not calculated for magnetic stripe card
manufacturing, only the value of 1 percent is used as a screening
* Note  that  the  total  unit  output  figure  calculated is  the
  maximum unit output  obtained  assuming dedicated facility
  capacity.   In  reality,  no facility has  dedicated all  of its
  capacity to the manufacture of one product.  However,  due  to  a
  lack  of production data,  the  assumption of dedicated  capacity
  is made in order  to  determine facility  output.
                              9-33

-------
criterion for regulated facilities in this industry.  As
previously mentioned, regulated facilities that are not price
leaders will have to absorb that percentage of control cost per
unit that exceeds the screening values.  Otherwise, these
facilities could lose their market share to unregulated
facilities (including domestic and foreign production plants).
     For regulated facilities that may not be able to achieve the
full price increase, two questions are asked:  (1) will absorbing
control costs result in an "facility-closing" decline in
earnings, and (2) will capital be available to finance the
investment in capital controls?  If neither is a problem,
production using the regulated solvents will continue, with the
addition of emission control equipment.
     The impact on a regulated facility's earnings of unabsorbed
control costs is calculated by dividing the percentage annualized
control costs not recovered through a price increase by the
facility's before-tax net income.  Before-tax is the appropriate
measure of earnings because emission control costs are before
taxes and are tax-deductible.
     In the short run, closure of a regulated facility would, in
theory, occur if annualized control costs result in an excess of
variable costs over revenue.  An indication of this would be a
ratio of annualized control costs to before-tax net income of at
least 100 percent.
     In the long run, a firm is free to deploy its capital in
investments that yield the highest rates of return.  In theory,
closure should result if the return on investment falls below the
rate of return on the best alternative investment  (which is equal
to the cost of capital).  A threshold decrease in net income,
20 percent, is established for indicating the possibility of
closure in the long run.
     The impact on capital availability is assessed by dividing
capital control costs for all the regulated facilities owned by a
company by the company's before-tax net income.  Before-tax net
income serves as a proxy for cash flow.  The ratio indicates the
extent to which capital costs can be financed from one year's
                               3-34

-------
cash flow.  Capital costs do not have to be paid from cash flow,
but the ability to do so suggests that either external financing
is not needed or would not be difficult to obtain.  If the ratio
exceeds 100 percent, it is possible that debt will have to be
issued.   (For investment in emission controls, it is normally
assumed that equity will not be issued because the investment
does not add to a firm's productive capacity).
     If debt needs to be issued, the impact on a firm's capital
structure can be assessed by dividing capital costs for all of a
firm's regulated facilities by the firm's total liabilities.
Firms that own regulated facilities may not be able to obtain
external financing if capital costs represent a
significant percentage of total liabilities.  A threshold value
of 20 percent is used to assess the impact on capital structure.
     An increase in total liabilities of 20 percent is not, to be
sure, a definitive threshold beyond which no facilities will be
able to obtain external financing.  There will always be some
facilities that are able to take on debt and, as a result, expand
total liabilities by 20 percent, or even more.  Conversely, some
facilities will be limited to expanding total liabilities by far
less than 20 percent.  However, an average increase in total
liabilities of 20 percent is likely to' make external capital
difficult to obtain for at least some facilities.  Creditors are
reluctant to lend to a firm with a high degree of financial
leverage  (i.e., high ratio of debt to net worth)  because there is
a high risk that the debt cannot be repaid.  If capital
availability is significantly impacted,  then the facility will
terminate the use of HAP solvents and may close the facility.
     If there are no adverse effects on the regulated facility's
earnings or the ability of its company to obtain capital,  then
the regulated facility will continue production with the addition
of pollution control equipment.
9.2.2  Industry-Level_Impacts
     The economic impacts of the NESHAP on the domestic magnetic
recording media industry are presented in Table 9-18.   As
reported in the table, annualized control costs for either
                               9-35

-------
regulatory alternative account for a negligible share of the
domestic industry's total value of shipments.  To fully recover
the annualized control costs associated with RA I, price would
have to increase by 0.01 percent.  For RA II, a 0.09 percent
increase is necessary to fully recover annualized costs.

         TABLE  9-18.  ECONOMIC  IMPACTS ON THE MAGNETIC
                    RECORDING MEDIA  INDUSTRY
Regulatory
alternative
l
2
Percent -
price
increase
0.01
0.09
Percent
change in
quantity
demanded
-0.01
-0.07
Percent
change in
employment
-0.01
-0.07
Percent
change in
revenue
0.00
0.02
  NOTE:   The figures  are rounded.
     The fact that annualized control costs account for a
negligible share of the industry's value of shipments indicates
that a large portion of industry production is unregulated.  In
other words, a large number of domestic magnetic recording media
facilities do not use HAP solvents in their production processes.
Unfortunately, information concerning the relationship between
regulated and unregulated domestic production is unavailable.
     The full pass-through of annualized control costs to the
end-user via a price increase will result in a reduction of
quantity demanded.  The price increase necessary to fully recover
RA I's annualized control costs would reduce quantity demanded by
0.01 percent.  For RA II, the reduction in quantity demanded
would be 0.07 percent.  The magnitude of the reduction in
quantity demanded is less than that of the price increase because
the demand for magnetic recording media is inelastic  (E = -0.75).
The more inelastic demand is, the less output producers will lose
as a result of increasing price.
     As stated in the methodology, it is assumed that there is a
linear relationship between output and employment.  Therefore,
the percent reduction in quantity demanded also  represents the

                               9-36

-------
decrease in industry employment.  In the industry profile it was
reported that, in 1989, 25,539 personnel were employed in
SIC 3695.  For RA I, the percent decrease in employment
translates to the loss of two workers.  If RA II is implemented,
17 workers will lose their jobs.  Obviously, the employment
impact on the entire magnetic recording media industry is small.
However, in terms of a marginal facility, the employment impacts
could be very significant.
     The price and quantity adjustments result in a change in
industry revenue (except when demand in unitary elastic with
respect to price).   Industry revenue would not change
significantly if RA I is implemented.  However, industry revenue
would be increased 0.02 percent by the implementation of RA II.
The increase in revenue is explained by the fact that demand is
price inelastic.  Because demand is inelastic, the effect on
industry revenue of the price increase more than offsets the
effect of the reduction in quantity demanded.
9.2.3  Facility-Level Impacts
     9.2.3.1  Introduction.  The magnetic recording media and
magnetic stripe card manufacturing facilities regulated by the
NESHAP are presented in Table 9-19.  It should be noted that
facility-level impacts will be discussed without identifying the
regulated facilities to which they pertain.  This is done for the
purpose of confidentiality.
     9.2.3.2  Facility Price Increases.  Tables 9-20 and 9-21
report the percent price increases that regulated facilities will
need to fully recoup control costs stemming from RA's I and II.
The price increase for a regulated facility is calculated
assuming the smallest possible revenue basis.  In other words, it
is assumed that facility capacity is dedicated to the manufacture
of the product which generates the least revenue.  However,
because regulated facilities manufacture a variety of products
(Table 3-3), the aforementioned procedure has the effect of
underestimating facility revenue and, as a result, overestimating
the price increase.  Thus, the reader should keep in mind that
the price increases calculated are worst-case estimates and that
                               9-37

-------
                TABLE 9-19.  REGULATED FACILITIES
                            Ampex, AL
                           Anacomp,  NE
                           Audiopak,  VA
                             Fuji, SC
                             JVC, AL
                            Malco, MD
                           Memorex,  CA
                             Sony, AL
                             Sony, GA
                            Syncom, SD
                              3M,  CA
                              3M,  MN
                              3M,  OK
the actual price increases will be less than those reported.  How
much less is unknown due to the lack of information concerning
regulated facility revenue.
     Two regulated facilities will need percent price increases
exceeding the screening values to fully recover the annualized
control costs of RA I.  As shown in Table 9-20, the Facility C
will need a 4.79 percent price increase to fully recover
annualized costs if facility capacity is dedicated to the
production of blank audio media.  A price increase of
2.57 percent will be needed to fully recover annualized costs if
Facility B dedicates its capacity to the production of blank
computer media.
     For RA II, four regulated facilities will need percent price
increases exceeding the screening values (Table 9-21).
Facility C will need a price increase of 6.01 percent for the
production of blank audio media.  Facility B will need a
2.63 percent price increase to fully recover annualized costs
incurred by blank computer media production.  For the manufacture
of blank video media, Facility F will need a 1.45 percent price
                               9-38

-------
      TABLE  9-20.
FACILITY PRICE INCREASES, REGULATORY
    ALTERNATIVE ONE
Regulated
facility
C
B
K
A
H
E
D
L
G
I
F
J
Worst -case percent price
increase3
4.79
2.57
0.81
0.76
0.19
0.11
0.06
0.04
0.01
0.00
0.00
-0.04
Market segment
Blank audio media
Blank computer media
Blank computer media
Blank computer media
Blank computer media
Blank computer media
Blank video media
Blank video media
Blank audio media
Blank computer media
Blank computer media
Blank audio media
aThe figures are rounded.

NOTE:  The code letters do not directly correspond to the
       facility names in Table 9-19.

       One regulated facility is not included in the table so
       as to maintain its anonymity.  Its percent price
       increase will be 0.91 percent in the worst-case
       scenario.
                             9-39

-------
      TABLE 9-21.
FACILITY PRICE INCREASES, REGULATORY
    ALTERNATIVE TWO
Regulate
d
Facility
C
B
F
K
A
D
J
H
E
I
L
G
Worst -case percent price
increase3
6.01
2.63
1.45
0.98
0.91
0.59
0.53
0.19
0.11
0.05
0.04
0.03
Market segment
Blank audio media
Blank computer
media
Blank video media
Blank computer
media
Blank computer
media
Blank video media
Blank video media
Blank computer
media
Blank computer
media
Blank computer
media
Blank video media
Blank audio media
aThe figures are rounded.

NOTE:  The code letters do not directly correspond to the
       facility names in Table 9-19.

       One regulated facility is not^included in the table so
       as to maintain its anonymity.'  Its percent price
       increase will be 1.19 percent in the worst-case
       scenario.
                             3-40

-------
increase.  The fourth facility, discussed in the table notes,
will need a 1.19 percent price increase.
     The ability of the regulated facilities to fully recover
annualized control costs via a price increase is a function of
the industry's market structure.  In a perfectly competitive
market, cost increases for only a few firms cannot be directly
translated into price increases because firms face a perfectly
elastic demand curve.  As the market structure deviates from
perfect competition, however, it is easier to recover control
costs via the pricing mechanism.  For this analysis, a l percent
price increase is thought to be achievable.  Only those firms
with price increases above 1 percent will have to be analyzed for
cost absorption  (as well as capital availability impacts).
     9.2.3.3  Earnings Impacts.  The regulated facilities with
price increases exceeding 1 percent will experience earnings
impacts.  These facilities will have to use their earnings
(represented by before-tax income) to pay for the percentage of
control costs which are absorbed, i.e., not covered by a
1 percent price increase.
     Table 9-22 details the impacts on earnings for the regulated
facilities unable to fully recover annualized control costs.  For
RA I, the costs which Facility B would not be able to recover via
a price increase would account for 14.79 percent of the
facility's earnings.  Costs that would have to be absorbed under
RA II would account for 15.36 percent of facility earnings.  For
Facility C, unabsorbed costs would account for 35.80 percent of
earnings under RA I and 47.24 percent of earnings under RA II.
Facility F will only have an earnings impact if RA II is
implemented; the facility's unrecovered control costs will
account for 4.28 percent of facility earnings.  For the
unidentified facility discussed in the table notes, absorbed
annualized costs associated with RA II represent 1.75 percent of
the facility's before-tax net income.
     In the short run, none of the listed facilities would have
to shut down.  Annualized control costs would not result in an
excess of variable costs over revenue, as indicated by the fact
                               9-41

-------
    TABLE 9-22.  IMPACTS OF ABSORBED ANNUALIZED CONTROL COSTS
               ON FACILITY BEFORE-TAX NET  INCOME
Percent of facility before- tax net
income
Regulated facility
B
C
F
RA 1
14.79
35.80
-
RA 2
15.36
47.24
4.28
  NOTE:   The  regulated  facility discussed in Table 9-21's notes
         will have  to absorb  a  percentage of its  annualized
         control  costs  associated  if  Regulatory Alternative II
         is implemented.  Absorbed annualized control  costs
         account  for 1.75 percent  of  the  facility's before-tax
         net  income.
that control costs do not exceed earnings for any of the four
facilities.
     In the methodology,  a threshold decrease in net income,
20 percent, was established to indicate the possibility of
closure in the long-run.   The value is exceeded for Facility C
for both regulatory alternatives.  Absorbed annualized control
costs may reduce the facility's earnings below the opportunity
cost of capital.  It is not known if this is the case for
Facility C.  However, it is possible that capital will be
redeployed from magnetic recording media production to other
investments as a result of the NESHAP.
     9.2.3.4  Capital Availability Impacts.  The capital costs
incurred by a regulated facility will impact the capital
structure of the parent company.  These impacts are reported in
Table 9-23.  The highest impact on before-tax net income under
either regulatory alternative would be approximately 10 percent
for Facility C's parent company.  Facility B's parent company
would have to use approximately 1.5 percent of its before-tax net
income to pay for its regulated facility's capital costs stemming
from either regulatory alternative.  The parent company of
Facility F would only have to use 0.05 percent of before-tax net

                               9-42

-------
income to finance capital costs for RA II.  Capital control  costs
account for a negligible share of before-tax net income  for  the
parent company of the regulated facility discussed in the  table
note.  All four companies will be able to finance capital  costs
from cash flow, and will not require external financing.

       TABLE  9-23.   PER-FACILITY CAPITAL CONTROL COSTS AS A
             PERCENT OF  COMPANY BEFORE-TAX NET INCOME
                         Percent of company's before-taxnNet
                                        income
   Regulated facility	RA 1	RA 2	
           B                  1.45                 1.45
           C                  9.86                 9.95
           F                    -                   0.05
  NOTE:   Capital control costs as a percent of before-tax net
         income are less than 0.01 percent for the regulated
         facility discussed in Table 9-21's notes.

9.2.4  Conclusions
     A regulated facility will be adversely affected by the
NESHAP if it has a significant price increase as well as a
significant impact with respect to either earnings or capital
availability.  Based on the impact calculations, one regulated
facility  (Facility C) will be adversely affected by the NESHAP.
     There i's a possibility that, in the long run, Facility C's
company will choose to redeploy its assets from the magnetic
recording media industry to a more profitable venture.  However,
it is not known whether the threshold decrease in income applies
to the company.  Thus, it cannot be said with any certainty that
Facility C's operations will be discontinued in the long run.
     The exit from the industry of any marginal firm may lead to
entry by an unregulated producer.  It is not known whether the
entrant would be a domestic or foreign competitor.
     However, there are indications that entry by a foreign
producer is a distinct possibility.  A large amount of magnetic
recording media is produced overseas.  In 1990, MMIS reported

                               9-43

-------
that only 8.2 percent of all audio cassettes were produced in the
United States.34  The trade association also reported that in
1990 only 15.1 percent of coated videotape was manufactured in
the U.S.35  A final point is that magnetic recording media
imports account for a significant percentage of domestic apparent
consumption (see Section 9.1.4).
9.3  SMALL BUSINESS IMPACTS
     According to the Regulatory Flexibility Act, the proposing
agency must determine whether the regulation will have a
significant impact on a substantial number of small entities.  If
the regulation does result in adverse effects on a substantial
number, then a Regulatory Flexibility Analysis (RFA) must be
undertaken.
     In order to make this determination, the first step is to
define exactly what a small entity is for magnetic recording
media producers and magnetic stripe card producers.  The
definitions of small entities provided in the Act, which are
based on the Small Business Administration's size standards, are
used.  Magnetic stripe card producers are part of SIC 3577.  An
entity in this SIC is considered small if it employs less than
1,000 personnel.  The same standard is used for SIC 3695, of
which magnetic recording media producers are a part of.
     Ten companies will be regulated by the NESHAP.  According to
thetaforementioned size standards, three of the ten companies are
considered to be small entities.
     A significant impact occurs if any of four criteria are
satisfied:
     1.  Annual compliance costs increase total costs of
production by more than 5 percent.
     2.  Compliance costs as a percent of sales are at least
10 percentage points higher for small entities.
     3.  Capital costs of compliance represent a significant
portion of the capital available to small entities.
     4.  The requirements of the regulation are likely to result
in the closure of small entities.
                               9-44

-------
The first and fourth criteria are satisfied for one of the three
small companies.  The affected company owns Facility C, which is
a closure candidate.
     Although the regulation is expected to cause significant
impacts for only one small company  (of an industry which consists
of 21 companies),  a Regulatory Flexibility Analysis will be
conducted.  The analysis will be done in order to assess the
possibility of providing additional regulatory flexibility to the
small companies expected to comply with the proposed regulation.
9.4  REGULATORY FLEXIBILITY ANALYSIS
9.4.1  Introduction
     Under Section 112 (d) of the 199.0 Clean Air Act, the U.S.
Environmental Protection Agency is developing a National Emission
Standard for Hazardous Air Pollutants (NESHAP) for the magnetic
recording media and magnetic stripe card industries.  The primary
hazardous air pollutants (HAP's) used in these industries are
methyl ethyl ketone, methyl isobutyl ketone, and toluene.  Two
regulatory alternatives are being considered.
     The results of the economic impact analysis (Section 9.3 of
the EIA) indicate that one small entity will be significantly
impacted.  This RFA focuses on this entity.  Particular attention
is given to why its economic impacts are considered significant.
However, the reader should note that neither the company nor its
facility are identified for the sake of confidentiality.
9.4.2  Demographic Analysis
     To define what qualifies as a small company in the two
industries, the Small Business Administration's (SBA) size
standards were used.  As discussed in Section 9.3,  three
companies have been identified as being small.
9.4.3  Findings of the Economic Impact Analysis
     Based on the results of the EIA, it has been determined that
two of the criteria for ascertaining whether there is a
significant economic impact were satisfied.  The first criterion
is that annual control costs increase total costs of production
by more than 5 percent.  Production costs for the three small
companies are unknown; however, they can be approximated by
                               9-45

-------
revenue data.  The percentage of revenue accounted for by
annualized control costs represents the product price increase
necessary for a regulated facility to fully recoup these costs.
One facility will require a price increase of approximately
5 percent for RA I and 6 percent for RA II.  This facility is
identified as facility C in the EIA.  Facility C is owned by one
of the small companies.
      Furthermore, absorbed annualized costs would re'sult in a
significant decrease in facility earnings under-either regulatory
alternative.  For RA I the decline in earnings will be
approximately 36 percent; the decline in earnings will be
47 percent for RA II.  In the long run, the owners of the company
may find the decrease in earnings unacceptable and redeploy their
assets in another, more profitable venture.  Thus, closure of the
company, the other criterion, is a possibility.
     The possibility of closure is made stronger by the fact that
the majority of the large entities in the industry will be not
significantly impacted.  Those large firms which are regulated
will not be significantly impacted with respect to production
costs.  Furthermore, an unregulated portion of the industry
exists; the facilities that comprise this portion do not use the
regulated HAP's in their production processes.  In the EIA, it
has been estimated that the industry-level price increase that
will result from RA I will be approximately 0.01 percent; for
RA II, the most stringent alternative, the price increase will be
approximately 0.09 percent.  The small industry-level price
increases mean that it is unlikely that the small company in
question will be able to recoup annualized control costs by
increasing price.  The company will possibly be able to recoup
annualized costs if it sells a differentiated product.
9.4.4  Consideration of Flexibility in Regulatory Requirements
     One of the purposes of conducting this analysis is to assess
the feasibility of providing additional regulatory flexibility
for small business to comply with the proposed regulation.  The
possibility of providing additional regulatory flexibility is
examined due to the significant economic impacts this regulation
                               9-46

-------
is expected to impose on one of the small companies in this
industry."  Two possibilities for providing flexibility were
considered, both of which are explained below.
     For small businesses in general, one mechanism that was
identified for providing regulatory flexibility was a solvent
usage cutoff.  The solvent usage cutoff was evaluated as a
mechanism for mitigating the impacts on small facilities in the
industry that may be initially considered major sources as a
result of the "potential to emit" definition.  The solvent usage
cutoff allows a facility to be exempt from the proposed
regulation if it uses less than 9.1 Mg/yr of any one HAP or
22.7 Mg/yr of multiple HAP's.  The rationale for establishing
this cutoff is that although some facilities may initially be
considered major sources based on their potential to emit HAP's,
their actual emissions may fall well below the cutoff level.
This may be the situation for some facilities that perform
magnetic coating operations as only one part of their total
manufacturing process.
     Based on information currently available to the Agency, one
existing facility owned by a small company will be exempt from
the proposed regulation because its estimated future solvent
usage is below the proposed solvent usage cutoff.  Because this
facility will not be required to control its emissions, it was
not included in the economic impact analysis.  However, the
facilities owned by the three small companies identified in
Section 9.3 have solvent usage levels above the cutoff.  The
small company that was identified as being significantly impacted
is included in this group.
     Any facility whose solvent usage exceeds the cutoff level
will have operations similar to those located at large
facilities.  The facilities owned by the three small companies
have a similar potential to emit HAP's as the large facilities.
Therefore,  there are no technical reasons for examining different
requirements for these three small companies as opposed to the
remainder of the regulated industry.
                               9-47

-------
     The Agency also examined the monitoring, recordkeeping, and
reporting activities required in the proposed regulation as a
possibility of providing regulatory flexibility.  For the small
company expected to experience significant economic impacts, the
monitoring, recordkeeping, and reporting activities are the least
costly options that could be implemented to achieve the
requirements of the Clean Air Act.  The recommended recordkeeping
and reporting requirements are also the minimum requirements
proposed for the General Provisions of the NESHAP program.
     Recordkeeping and reporting costs developed for the industry-
assume that some facilities may fail to comply with the proposed
regulation.  These requirements are more detailed (and therefore,
more costly) for facilities that fail to comply.  The small
company in question could minimize its recordkeeping and
reporting activities (and therefore, its costs) by continuing to
stay in compliance with the proposed regulation.
     To provide regulatory flexibility to small companies,
several requirements of the proposed regulation have been
reconsidered.  First, the Agency is providing a solvent usage
cutoff that allows facilities classified as major sources to be
exempt from the proposed regulation if their solvent usage level
is below the cutoff level.  The Agency has also attempted to
provide regulatory flexibility by requiring the minimum
monitoring, recordkeeping, and reporting activities that would
ensure that the objectives of the proposed regulation are being
achieved.
9.5  REFERENCES FOR SECTION 9
 1.  U.S. Department of Commerce, Bureau of the Census.  Current
     Industrial ReportS-MA35R, Computers and Office and
     Accounting Machines, 1981-1990.
 2.  U.S. Department of Commerce, Bureau of the Census.  County
     Business Patterns 1989, United States.
 3.  Telecon.  Jenkins, A.K., with Beckett, L., Bureau of the
     Census, Industry Division.  August 31, 1992.  Information on
     Bureau of the Census data.
                               9-48

-------
 4.  Telecon.  Williams, D., MRI, with Kosley A., Brown Disk
     Manufacturing.  July 23 and 25, 1991.  Information on plant
     closings.

 5.  Telecon.  Williams, D., MRI, with Lewis, J., Certron.  July
     23 and 25, 1991.  Information on plant closings.

 6.  Telecon.  Williams, D., MRI, with Hyde B. Audiopak.  July
     25, 1991.  Information on plant closings.

 7.  Industry Norms and Key Business Ratios.  Dun and Bradstreet,
     p. 121.

 8.  Magnetic Media International Newsletter, Volume XI, Number
     Three.  Magnetic Media Information Services, Honolulu, HI.
     November 30, 1989_.  p. 60.

 9.  Reference 8, p. 65.

10.  Magnetic Media International Newsletter, Volume XI, Number
     Six.  Magnetic Media Information Services, Honolulu, HI.
     October 19, 1990.  p. 66.

11.  Reference 10, p. 52.

12.  Reference 10, p. 54.

13.  Reference 10, p. 53.

14.  Datamation, June 15, 1990.  pp. 92-93.

15.  Magnetic Media International Newsletter, Volume XI, Number
     One.  Magnetic Media Information Services, Honolulu, HI.
     June 20, 1989.  p. 69.

16.  Reference 8, p. 87.

17.  Telecon.  Jenkins, A.K., with Halliday, S., Automatic
     Identification Manufacturers, U.S.A.  July 29, 1992.
     Information on the magnetic stripe card industry.

18.  U.S. Department of Commerce, Bureau of the Census.  Current
     Industrial Report-MQ-C1, Survey of Plant Capacity.

19.  1992 Electronic Market Data Book, Electronic Industries
     Association, Washington, DC.  pp. 13-14^ 23.

20.  Reference 3, p. 26.

21.  Dealerscope Merchandising.  January 1991.  p. 46.

22.  Reference 8, p. 38.
                               9-49

-------
23.  Layard, P.R.G., and Walters, A.A.,  Microeconomic Theory.
     McGraw-Hill, 1978.  pp. 259-267.

24.  Magnetic Media International Newsletter.  Volume XI, Number
     Four.  Magnetic Media Information Services, Honolulu, HI.
     May 10, 1990.  p. 55.

25.  Reference 24, p. 66.

26.  Reference 10, p. 19.
           *       •
27.  Reference 10, p. 18.

28.  U.S. Industrial Outlook 1992, U.S.  Department of Commerce,
     International Trade Administration, Washington, DC.
     p. 37-17.

29.  Standard and Poor's Corp.  Industry Surveys, "Computers-
     Basic Analysis."  New York, NY, October 17, 1991.  p. C107.

30.  Reference 10, p. 14.

31.  Reference 10, p. 15.

32.  Reference 10, p. 5.

33.  Reference 10, p. 6.

34.  Reference 10, p. 61.

35.  Reference 10, p. 57.
                               9-50

-------
                 APPENDIX A.




EVOLUTION OF THE BACKGROUND INFORMATION DOCUMENT

-------
                           APPENDIX A.

         EVOLUTION  OF THE BACKGROUND  INFORMATION DOCUMENT


     The purpose of this study was to provide data to support the
development of the proposed national emission standard for
hazardous air pollutants  (NESHAP) for the magnetic tape
manufacturing industry.  To accomplish the objectives of this
program, technical data were gathered on the following aspects of
the industry:  (1)  the. operation of solvent storage tanks, mix
preparation equipment, coating application and drying equipment,
waste handling devices, transfer equipment (piping), and cleaning
devices, (2)  the release and controllability of hazardous air
pollutants  (HAP's)  emitted into the atmosphere from the above
emission points,  and  (3) the types and costs of demonstrated
emission control technologies.  The bulk of the information was
gathered from the following sources:

     1.  Technical literature;

     2.  Plant visits;

     3.  Questionnaires sent to industry;

     4.  Industry representatives;

     5.  State and regional air pollution control agencies; and

     6.  Equipment vendors.

     Significant events relating to the evolution of the
background information document are itemized in Table A-l.
                               A-l

-------

-------
TABLE A-l.  EVOLUTION OF THE BACKGROUND INFORMATION DOCUMENT
Date
12/4/90
12/14/90








1/10/91
2/20/91
2/21/91
2/22/91
Company, consultant, or
agency/ location
International Tape
Association (ITA) ,
New York, NY
Radio Shack
(Tandy Corporation)
Fort Worth, TX
Anacomp Corporation
(formerly Xidex)
Santa Clara, CA
Ampex Corporation
Redwood City, CA
Memorex Corporation
Santa Clara, CA
3M Magnetic Media
Hutchinson , MN
BASF Corporation
Bedford, MA
JVC Magnetics, Inc.
Tuscaloosa, AL
Syncom Technologies
Mitchell, SD
NCR Corporation
Morristown, TN
Audiopak, Inc.
Winchester, VA
Sony Music Operations
Carroll ton, GA
Ampex Corporation
Opelika, AL
JVC Magnetics America
Tuscaloosa, AL
Nature of action
Letter from U. S. EPA
describing the magnetic tape
NESHAP project
Section 114 information
request sent by the U. S.
EPA








Plant visit
Plant visit
Plant visit
Plant visit
                            A-2

-------
TABLE A-l.  (continued)

Date
4/4/91























4/12/91



7/25/91


8/6/91

10/2/91

10/2/91

Company, consultant, or
agency/ location
Sony Magnetic Products
Dothan, AL

Graham Magnetics
(formerly, Carlisle
Memory Products)
Graham , TX
Fidelipac Corporation
Moorestown, NJ
Malco Plastics Company
Garrison, MD
Radio Shack
(Tandy Corporation)
Fort Worth, TX
Memorex Corporation
Santa Clara, CA
BASF Corporation
Bedford, MA
Syncom Technologies
Mitchell, SD
3M Environmental
Engineering and
Pollution Control
St. Paul, MN
Environmental C&C
Tempe , AZ


Sony Magnetic Products
Dothan , AL

Rand McNally
Skokie, IL
XDP Magnetics
Gary, IL
Washtech Systems
Springfield, MO

Nature of action
Section 114 information
request sent by the U. S.
EPA





















Letter from MRI requesting
test data for Kureha
fluidized-bed carbon
adsorber
Letter from U. S. EPA
requesting information on
HAP emissions
Letter from MRI requesting
information on HAP emissions
Letter from MRI requesting
information on HAP emissions
Product information from
vendor
          A-3

-------
TABLE A-l.  (continued)
Date
10/23/91

11/8/91









11/8/91
11/26/91
11/22/91
12/6/91
Company, consultant, or
agency/ location
Magnetic Ticket and
Label
Dallas, TX
Ampex Corporation
Opelika, AL
Disksystems , Inc.
Sunnyvale, CA
JVC Magnetics America
Tuscaloosa, AL
Memorex Technologies
Santa Clara, CA
Sony Music Operations
Carrollton, GA
Sony Magnetic Products
Dothan, AL
Syncom Technologies
Mitchell, SD
Verbatim Corporation
Sunnyvale, CA
3M Environmental
Engineering and
Pollution Control
St. -Paul, MN
Graham Magnetics
Graham , TX
Sony Music Operations
Carrollton, GA
Tandy Magnetic Media
Santa Clara, CA
Ampex Corporation
Opelika, AL
Nature of action
Section 114 questionnaire
from U. S. EPA

Section 114 information
request from U. S. EPA,
focused on cleaning
activities









Information sent from plant
regarding waste handling
operations
Information sent from plant
regarding waste handling
operations
Section 114 information
request from U. S. EPA,
focused on cleaning
activities
Information sent from plant
regarding waste handling
operations
          A-4

-------
                    TABLE A-l.   (continued)
Date
 Company,  consultant,  or
     agency/location
      Nature  of  action
1/29/92
3M Environmental
Engineering and
Pollution Control
St. Paul, MN
Information sent from plant
regarding coater enclosures
1/23/92
Electronic Data
Magnetics
High Point, NC
Plant visit
2/17/92
Syncom Technologies
Mitchell, SD
Information sent from plant
regarding coater enclosures
2/21/92
Anacomp, Inc./Omaha, NB
Information sent from plant
regarding coater enclosures
5/20/92
Mailed to members of
Work Group	
Mailout to EPA Work Group
members (briefing package)
6/26/92
Mailed to all known
industry members and
selected vendors
Request from U. S. EPA for
comment on draft BID
Chapters 3, 4, and 6 and
request for cost information
from industry
11/18/92
U. S. Environmental
Protection Agency,
National Air Pollution
Control Techniques
Advisory Committee
(NAPCTAC), and industry
representatives	
NAPCTAC meeting
                               A-5

-------
                APPENDIX B.




INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS

-------
                           APPENDIX B.

           INDEX TO  ENVIRONMENTAL  IMPACT  CONSIDERATIONS
    • This appendix consists of a reference system which is
cross-linked with the October 21, 1974, Federal Register
(39 FR 37419)  containing the Agency guidelines concerning the
preparation of environmental impact statements.  This index can
be used to identify sections of the document which contain data
and information germane to any portion of the Federal Register
guidelines.
                               a-i

-------

-------
TABLE B-l.  CROSS-INDEXED REFERENCE SYSTEM TO HIGHLIGHT
     ENVIRONMENTAL IMPACT PORTIONS OF THE DOCUMENT
Agency guidelines for preparing
regulatory action environmental
impact statements (39 FR 37419)
1. Background and
summary of regulatory
alternatives
Summary of the
regulatory alternatives
Statutory basis for
proposing standards
Relationship to other
regulatory agency actions
Industries affected by the
regulatory alternatives
Specific processes affected
by the regulatory
alternatives
2. Regulatory alternatives
Control techniques
Regulatory alternatives
3. Environmental impact of
the regulatory
alternatives
Primary impacts directly
attributable to the
regulatory alternatives
Secondary or induced
impacts
4. Other considerations
Location within Background Information Document

The regulatory alternatives from which standards will be chosen
for proposal are summarized in Chapter 1, Section 1.1.
The statutory basis for proposing standards is summarized in
Chapter 2, Section 2. 1
The relationships between EPA and other regulatory agency
actions are discussed in Chapter 3.
A discussion of the industries affected by the regulatory
alternatives is presented hi Chapter 3, Section 3.1. Further
details covering the business and economic nature of the
industry are presented in Chapter 9, Section 9.1.
The specific processes and facilities affected by the regulatory
alternatives are summarized in Chapter 1, Section 1.1. A
detailed technical discussion of the processes affected by the
regulatory alternatives is presented in Chapter 3, Section 3.2.

The alternative control techniques are discussed in Chapter 4.
The various regulatory alternatives are defined in Chapter 6,
Section 6,4. A summary of the alternatives is also included in
Chapter 1, Section 1.1.

The primary impacts on mass emissions and ambient air quality
due to the alternative control systems are discussed hi Chapter 6,
Section 6.4 and Chapter 7, Section 7.1. Tables summarizing the
environmental impacts are included in Chapter 1 .
Secondary impacts for the various regulatory alternatives are !
discussed in Chapter 7, Sections 7.1, 7.2, 7.3, 7.4, and 7.5. j
A summary of the potential adverse environmental impacts
associated with the regulatory alternatives is included in
Chapter 1, Section 1.2, and Chapter 7. Potential socioeconomic
and inflationary impacts are discussed in Chapter 9, Section 9.2.
Irreversible and irretrievable commitments of resources are
discussed in Chapter 7, Section 7.6.
                          3-2

-------

-------
       APPENDIX C.




EMISSION SOURCE TEST DATA

-------
                           APPENDIX C.
       *       »
                    EMISSION SOURCE TEST DATA


     The emission source test data presented here were obtained
during the NSPS analysis.  Sources of the data include:
(1) EPA-sporisored testing, (2) magnetic tape industry data on
carbon adsorbers, (3) State compliance tests, and (4) EPA-
sponsored testing for a related industry.  The following sections
discuss these data.

C.I  DATA FROM EPA-SPONSORED TESTS ON CARBON ADSORBER RECOVERY
     EFFICIENCIES

     Tests were conducted at two magnetic tape coating plants to
determine the solvent recovery efficiencies of the fixed-bed
carbon adsorbers.  At Plant 2, the carbon adsorption system
recovers a mixture of toluene and tetrahydrofuran (THF) solvent
from the tape coating process.  The system features three annular
carbon beds and processes 4.6 normal cubic meters per second
(Nm3/s) (9,800 standard cubic feet per minute [scfmj) of
solvent-laden air (SLA).  The three beds repetitively undergo
adsorption and desorption in a staggered sequence, which is
controlled by a timer.  The adsorption period is set at
64 minutes per bed, and desorption is set at 32 minutes per bed.
The cycle does not include a bed cooldown period after the
desorption period.  A continuous distillation train separates
solvent that is removed from the beds during steam desorption
into toluene- and THF-rich fractions.

     During the 3-week test period, a hydrocarbon analyzer
semicontinuously monitored the inlet and outlet solvent
concentrations.  The analyzer data were digitized and input ro an
onsite computerized data acquisition system.  Table C-1 presents
the operating conditions encountered during the tests.  Table c-2
presents a summary of the results of the tests.  The inlet
toluene/THF concentration averaged 1,230 parts per million by
volume (ppmv), which corresponds to an inlet solvent mass rate of
70.5 kg/hr (156 Ib/hr).  Inlet concentrations varied from about
50 to over 2,400 ppmv, depending on the number of operating
coating lines, the coating speed, and the coating thickness.  The
time-averaged outlet concentrations from Beds 1, 2, and 3 were
5.2, 2.2,  and 2.4 ppmv, respectively, which correspond to outlet


                               C-l

-------
solvent mass rates of 0.15, 0.071, and 0.078 kg/hr  (0.33,  0.16,
and 0.17 Ib/hr), respectively.  Therefore, average  system
volatile organic compounds  (VOC) removal efficiencies were 99.8,
99.9, and 99.9  percent for  Beds 1, 2, and 3, respectively,  based
on solvent mass rates.  Outlet solvent concentrations varied  from
near O to over  100 ppmv, depending primarily upon cycle timing.
Laboratory analysis indicated that bed carbon adsorption capacity
was significantly below virgin carbon capacity levels.  However,
the reduced adsorption capacity apparently did not  severely
affect system performance,  as indicated by the bed  VOC removal
levels approaching 100 percent.

     Plant 3 manufactures magnetic tape by coating  a polyester
film or web with ground magnetic iron oxide slurried with  a
solvent formulation of primarily THF and toluene with small
amounts of methyl ethyl ketone (MEK), methyl isobutyl ketone
(MIBK), and cyclohexanone.  The carbon adsorption system recovers
the solvent driven off during the coating process.  The system,
which has been  operational  since 1978, processes
9.4 Nm3/s  (19,800 scfm) of  SLA.  The system features three pairs
of annular carbon beds, with two pairs on-line and  one pair
maintained as a spare.  The two on-line pairs operate on a
90-minute timed adsorption/ regeneration cycle.  The system
processes SLA with an approximate inlet solvent concentration of
2,000 or 5,000  ppmv, depending on the type of magnetic tape being
produced.  The  system is regenerated by low-pressure steam
desorption and  ambient air  cooldown.  A batch distillation train
separates solvent that is removed from the beds during steam
desorption into component fractions.

     During the 2-week test period, hydrocarbon analyzers
semicontinuously monitored  the inlet and outlet solvent
concentrations.  The analyzer data were digitized and input to  an
onsite computerized data acquisition system.  Table C-3 presents
a summary of the operating  parameters encountered during the
tests.  Table C-4 presents  a summary of the results of the VOC
sampling.  The  data show VOC removal efficiencies ranging  from
91 to 98 percent.  The measured VOC removal efficiency for each
bed was generally based on  an average of 10 cycles  of adsorption
and regeneration.  The carbon in Beds 3A and 3B had been in
service for only 1 to 2 weeks at the time of the testing.   These
beds had substantially better VOC reduction performance than  did
Beds 1A, IB, 2Af which all  contained carbon with up to 5 months
.of service.

C.2  DATA FROM  INDUSTRY ON  CARBON ADSORBER RECOVERY EFFICIENCIES

     A two-bed, fixed-bed carbon adsorption system  installed  in
1980 controlled the VOC emissions from the magnetic tape coating
operation at the IBM Corporation facility in Tucson, Arizona.
Table c-5 presents a summary of the normal operating parameters
of this system. Table C-6  presents a summary of the monthly
                                C-2

-------
average VOC control efficiencies for 1982.  Control efficiencies
ranged from 94 to 99 percent.

     The 3M Company facility in Camarillo, California, coats
computer tape and disks.  Two separate fixed-bed carbon
adsorbers, one with four beds and one with two beds, control VOC
emissions.  The systems were installed in 1975 and 1979.
Table C-7 presents the actual operating parameters of these
systems.  Table C-8 presents a summary of the monthly average
control device efficiencies for a 12-month period.  Monthly
average control efficiencies ranged from 89 to 97 percent.

C.3  DATA FROM STATE COMPLIANCE TESTS

     The Allied Media Technology facility in Sunnyvale,
California, operated two coating lines that produced video and
computer tape products.  A nitrogen condensation system
controlled oven VOC emissions.  The control device recovery
efficiency was determined by conducting a 3-hour liquid solvent
material balance on February 25, 1983.  The amount of solvent
applied was calculated from the operating parameters of both
lines.  Table C-9 presents a summary of these data.  The total
amount of solvent applied during the test period was 110.2 kg
(242.9 Ib).  The amount of solvent recovered was calculated from
the weight of the solvent drum after collection, minus the weight
of the drum and the weight of the solvent in the drum before
collection.  These data are presented in Table C-10.  The total
amount of solvent recovered during the test was 102.5 kg
(226.0 Ib).  The solvent recovery efficiency of this coating
operation VOC control system was 93.0 percent (102.5 kg -s-
110.2 kg) (226.0 Ib + 243.0 Ib),

C.4  DATA FROM EPA-SPONSORED TESTS FOR RELATED INDUSTRIES

     The EPA conducted tests at plants in the pressure-sensitive
tape and label (PSTL) industry.  This is an industry with coating
and control processes very similar to those used in the magnetic
tape manufacturing industry.  In both industries, a solvent-based
coating is applied to a continuous supporting web.  Fixed-bed
carbon adsorbers are the most commonly used control device in
both industries, and similar total enclosures around the coating
application/flashoff area are used to capture fugitive VOC
emissions.  The following paragraphs describe relevant test data
from the PSTL industry.

     One PSTL facility was examined over a 4-week period
(January 15, 1979, to February 9, 1979).  The facility consists
of four adhesive coating lines controlled by a single carbon
adsorption system.  There are three lines that are each 28 inches
wide, and one line that is 56 inches wide.  The plant operation
is characterized by many short runs at slow line speeds.
Table C-ll summarizes the operations of each line and the total
system.  This facility is an example of a facility that is hard


                               C-3

-------
to control because slow coating lines are the most difficult to
control (e.g., they have the greatest potential for fugitive
solvent emissions).

     During the 4-week test period, the controlled facility used
28.7 m3 (7,589 gallons) of solvents in its adhesive formulations
and recovered 26.7 m3  (7,065 gallons) from the carbon adsorption
facility.  This represents an overall VOC control of
93.1 percent.  The system performed 140 separate runs and used
the following solvents: toluene, acetone, hexane, ethyl acetate,
MEK, rubber solvent, heptane, mixed solvents, recovered pro lam
solvents, xylene, ethyl alcohol, and isopropanol.

     The makeup air for the ovens is pulled directly from the
work area.  The building that houses the coaters is tight enough
to allow a slight negative pressure in the work area as compared
to the outside of the building.  Also, there is a slight negative
pressure in the coater ovens with respect to the room air.  With
a fully enclosed, tight system, the overall result is that all
makeup air flows into the building, through the oven, and out to
the carbon adsorption system.  Therefore, essentially 100 percent
of all solvent emissions are captured.  The facility also uses
hoods over the coater areas to capture fugitive solvent emissions
near the coating applicator.  Ductwork directs hood gases into
the drying oven.
                               G-4

-------
I

U


W
J
03
   3,
tes
64
         7 J j.
oo
r»
           tt.


        S .-^ T3 O


         *
        >o d
 

                     1
                     IS

                                                   0
erature

ve humidity
Outlet te

Outlet re
                                                                 •a

                                                                 .2
                                            u.

                                            £

                                            1
                                                           o

                                                           U

                                                           3
                                              .s
                                              Ul

                                              •3
                                                  73



                                                  1
                                                                 o
                                                                 8

                                                                 •3
                                              .2
                                              T3
                                                  I

                                                  f-
                                                  O
                                              o


                                              ^5



                                              I

                                              u
           !!i
           'i.8 >

           -I.I
           o 3 3
           -2 < a
           a .a  a -
                                                 I
                               C-5

-------



CM
04
1
<

J
$
O
B
OS
u
o
06
a
CO
I ADSOR
«E>
O
CO
OS
CJ
Q
W
N
H
OS

^3
a
5
CO

*
CM
1
O
a

0
?j




1
8^
J |
o o
r
i
§ §
s ~
II

^* §.
§ s
nge of sol v
entrations,
CO O

p




>
c^ S
c g-

^ •
s.I
o tS
2* S
> u
»






















ib
X
H
•«
c
«


2
"S
£

2
's
3
a
i
j{
i
e
m
3
Vt
M
1*
S-*
til Z
C-6

-------










JH»
n
E^
04
1
rt!
g
««
O
U
O
i
z
o
H
H
04
tt
O
en
Q
<
•
n
o
u
.J
oa
g
™













S»
uV
1








j
«
&
3?
^












5
3
a






^
Z








„
(^














^
I
|
1
!
u
§
«•*
*
s
i
*
2 «
•^x
09
^
B
Z
00
Ot
<*
>n
'^
00



^1
s 1
r^ N
,= 1
^ >
^ 1
^ 5
s |.
S i
1 S
C en
u is
3 i
? !
a o
S 9.
oo" 2
^ g
* "
S oC «
-^ «
«n jj»
1 a
2 Z
s s
* -t
s i
« •*



~ ?
1 s
a o
5? 5
« o •«,
ss c <

S
•*»
0.
ss
Cj tn
o
r»

SSa
•*•*?»

o











pressure drop - 1A/1B
pressure drop - 2A/2B
pressure drop - 3A/3B
u 1J ij


^^
F
s
*^
*o
•J o
u
a
CJ
00






c!
a O
O f\
ec
,y^
="s
o
f^






?
•3
W
3
o
>
^
sl-(.
1-1
ll
3* 5
^ 0
5 J
V. W
33






O








.*
^J*
H
ap
1
01
                     .2
                      Jtf
                      1
                     "S
                     .0
                      o
                     a

                      i
                      Wl
                      3
                      u
                      2
                     I
2
u
a.
VI
O
S
&
u
e
Q
3
1

oj
1 s
& a 1
» » G
sl-a
2 S 3
a -a -j
y
o
~*
u
S i-
o
t/» w
•a "°
i§
» 2 S
III
at ^ |
Jl*
111
> « g
* s c
— .S 0
.5 « is
-? £.3
C-7

-------


*
0*
61
Q
j
£
0

S
0
f^
^^

OS
w
ADSORB
§
CO
u
o

OS

CO

.
1
o
TABLE


•

1
00
.M
1
c
1
2
J
S
«
r
1





],
J
.2
1
o
1
j£
i
o
1
^



ja


1
4> _
r
so
if
°
•§ .0
•3 p.
1|
3 j
°



1





!l
i B

60
|l
•O gi
*8 -1
i


1
»-•


a
II
o
1 Outlet sampling
locution
q
*
NO


VO





00





5


s




g
(S



j

tS
J<
o m
•* rf
ON ON
T **?
O f>

oo q
«•> 2





«o w>
«s \o





^ «r>


s5




|g
»~ •*



> -S
Is

?
ca

ON ON
O 
s

i i
S o
S 2
2 &
^ *«
__l
1 J
0 oo
1 1
M S
•2 8
1 1 i :•

1^1 8
oil I
l|! 3
5 |-S £
1- § 1
* S "3
II i
*•* *«4 O *^
0 g - "2

i I § •§
'! s | 1
111^!
8-a > 1 i s
ipJ!2S
Ill^s
3 "" "° J. JJ <
^ "S "S a a ^
4 a a u u «
* g « 3 3 -a
llllil
Illiil
7 a <3 « a a
u oa CQ u u u
<• ja a -o u <-r
c-a

-------
TABLE OS.  CARBON ADSORBER OPERATING PARAMETERS  FOR IBM FACILITY
Parameter
SLA flow rate, m3/s (acfm)
Nominal
Maximum
SLA inlet temperature °C (°F)
SLA inlet, % relative humidity
SLA inlet concentration, ppmv
Typical
Maximum
Carbon lifetime, months
Total carbon charge, kg (Ib)
Design value
NA*
7.5 (1500)
29-32 (85-90)
55
30
65
N/Ab
N/A
Actual value
6.85 (13,700)
7.5 (15,000)
29 (85)
55
6
21
60
17,237 (38,000)
*NA « Not available.
bN/A » Not applicable.
                                  C-9

-------
    TABLE C-6.   SUMMARY OF CARBON ADSORBER
     EFFICIENCIES FOR IBM FACILITY IN 1982
Month
January
February
March
April
May
June
July
August
September
October
November
December
Control device, %
94.2
94.2
95. Oa
95. Oa
98.3
99.0
98.6
98.7
99.1
99.4
98.5
97.7
aEstimated

-------
TABLE C-7.  CARBON ADSORBER OPERATING PARAMETERS FOR 3M FACILITY
Parameter
SLA flow rate, m3/s (acftn)
Maximum
SLA inlet temperature, 8C (°F)
SLA inlet, % relative humidity
SLA inlet concentration, ppmv
Typical
Maximum
Carbon lifetime, months
Total carbon charge, kg (Ib)
Carbon Adsorber No. 1 Values
8.0 (16,000)
43 (110)
50-65
400
1,200
18
29,030 (64,000)
Carbon Adsorber No. 2 Values
4.0 (8,000)
38 (100)
50-65
125
375
60
9,072 (20,000)
                              Oil

-------
 TABLE C-8.  SUMMARY OF CARBON ADSORBER RECOVERY
           EFFICIENCIES FOR 3M FACILITY
Month
1
2
3
4
5
6
7
8
9
10
11
12
Control device
No. 1, %
94
96
97
96
95
94
92
90
91
90
90
89
Control device
No. 2 %
96
96
96
95
95
94
94
93
93
91
91
90
aControl device efficiency data are at end of
 adsorption cycle, which has the lowest
 efficiency at this point.
                       C-12

-------
TABLE C-9.  CALCULATION OF THE AMOUNT OF SOLVENT APPLIED DURING
           COMPLIANCE TEST AT ALLIED MEDIA TECHNOLOGY
Parameter
Coating thickness, fan (mil)
Coating speed, m/s (ft/min)
Coating time, min
Coating width, cm (in.)
Coating formulation, % solids by wt.
Coating usage, kg/m2 (Ib/ft2)
Solvent density, kg/f (Ib/gal)
Solvent use, kg (Ib)
Line No. 1
3.1 (0.125)
1.0 (200)
226
31.1 (12.25)
32
0.037 (0.0075)
0.88 (7.35)
66.7 (147)
Line No. 2
4.1 (0.165)
0.8 (150)
149
31.1 (12.25)
32
0.037 (0.0075)
0.88 (7.35)
43.5 (96)
                             C-13

-------
TABLE 010.  CALCULATION OF AMOUNT OF SOLVENT RECOVERED DURING
          COMPLIANCE TEST AT ALLIED MEDIA TECHNOLOGY
Parameter
Total weight of drum and collected solvent
Drum tare weight
Weight of solvent collected before time of
test
Total solvent recovered
Value, kg (Ib)
118 (260)
12.5 (27.6)
3 (6.6)
105.5 (232.6)
                             C-14

-------
TABLE C-ll.   SUMMARY  OF COATING  LINE OPERATIONS AT PSTL FACILITY

Line width, m (in.)
No. of runs
Average line speed, m/sec
(ft/min)
Avenge weight percent solvent
Total solvent used*
kg
(lb)
L
(gal)
Line No.
1
1.42 (56)
25
0.21 (41)
57.5
12,750
(28,110)
15,630
(4,129)
2
0.71 (28)
68
0.24 (46.5)
62.2
4,915
(10,837)
5,761
(1,522)
3
0.71 (28)
23
0.24 (46.5)
66.0
3,747
(8,262)
4,323
(1,142)
4
0.71 (28)
24
0.22 (42.5)
62.4
2,309
(5,091)
3,017
(797)
Total

140
0.23 (44.8)
60.3
23,723
(52,300)
28,731
(7,589)
aMeasured during 4-week test period.
                                 C-1S

-------

-------
                 APPENDIX  D.




EMISSION MEASUREMENT AND CONTINUOUS MONITORING

-------
                           APPENDIX D.

          EMISSION MEASUREMENT AND CONTINUOUS MONITORING


D.I  EMISSION MEASUREMENT TEST PROGRAM AND METHODS

     Sufficient  information  on  the emissions  associated  with
magnetic  tape  manufacturing  operations  was  available  such that
additional data gathering through source  testing was not performed
by the Emission Measurement Branch  (EMB)  of the U. S. Environmental
Protection Agency (EPA)  as part of the background support study for
the  National  Emission  Standards  for Hazardous  Air Pollutants
(NESHAP) for this industry.  EMB has assisted in the writing of the
data gathering protocols which were needed to obtain  estimates of
emissions  from;  (1)  the cleaning of removable  parts,  (2)  the
cleaning  of tanks used in the  coating  mixing process,  (3)  the
cleaning of fixed exterior surfaces,  and  (4)  the flushing of fixed
lines that carry the coating mix from the mix room to the coater.
Protocols were developed to aid industry in gathering the necessary
data  without having to perform an  extensive and costly source
testing program involving collection and  stack testing of fugitive
cleaning  emissions.    Protocols   were   deemed   unnecessary  in
determining emissions from other sources,  such as storage tanks,
coating mix preparation,  application, and  drying, packaging and
labeling, waste handling, and piping leaks.   These latter source
emissions   were  estimated  by  using   plant-specific   data  in
conjunction with  previously  determined  emission  factors  for the
magnetic  tape  industry or for  another  industry having  related
processes.  The procedures used  in the  data collection study are
described below by emission source.

D.I.I  Cleaning of Removable Parts
     In determining emissions  from the cleaning of removable parts,
the following  summarizes  what was done  in  the collection of the
data:
     Facility A subtracted the quantity used for tanks  and surfaces
from the total usage.
     Facility  C  used  weight  measurements.    Solvent  in  was
determined by weight and solvent out by  a volume measurement.  The
facility showed that weight measurements  compared well with volume
measurments, so the reported quantity of  solvent used  annually for
parts cleaning was extrapolated from the test results.
     Facility D used the reported quantity of solvent used annually
for parts cleaning is based on records maintained at the facility.
     Facility E used the  annual solvent used and recovered from
parts cleaning  based on records specifically maintained  for the
wash sink.
     Facility F solvent  input  and output was determined by a volume

                               D-l

-------
measurement;.   The  annual  solvent usage  for parts  cleaning was
extrapolated from the test data.
     Facility  H solvent usage  for parts  cleaning was  based  on
records.  The  quantity  of  solvent recovered was not known by the
plant, so it was determined by applying an average percent recovery
based on the tests to the total quantity of solvent used.
     Facility  I solvent usage  for parts  cleaning was  based  on
records.  The quantity of solvent recovered was determined using  an
average percent recovery from the tests.
     Facility J annual solvent usage was calculated by multiplying
the total quantity  of solvent added to the sink by the  frequency
that the sink is drained and reloaded.

D.I.2  Cleaning of Tanks
     Emissions from the cleaning of the tanks used in the coating
mixing process were estimated based on either data provided by the
plant or by completing the  data  gathering protocol.  There are two
types of cleaning systems for tanks:   closed top tank cleaning and
open  top  tank cleaning.    The  following  summarizes   the  data
gathering methods:
     Facility A (open-top) solvent recovery was calculated using  an
average percent recovery from the tests.  Annual solvent used for
tank cleaning is based on records.
     Facility B (open-top)  used weight measurements.
     Facility C (open-top)  annual solvent used and recovered were
derived from the protocols.
     Facility D (hybrid; a combination open-top:closed-top system)
annual solvent used and recovered were derived from the protocols.
     Facility E (closed-top) annual usage is based on an estimate
of quantity used per tank cleaning, number of tanks, and  frequency
of cleaning.  Emission estimate is based on the AP-42 equation.
     Facility  F (hybrid)  annual usage  and  recovery  are based  on
cleaning protocol.
     Facility G (open-top)  annual usage and recovery is based on  6-
month records multiplied by two.
     Facility G  (closed-top)  solvent  in was a level measurement,
solvent out was a weight measurement.
     Facility  H  (open-top)  annual solvent usage obtained  from
records.   The  annual  recovery data  was  extrapolated  from the
protocol results.
     Facility  I (hybrid) solvent  in  determined by  a flowmeter,
while the solvent out was weighed.
     Facility  J annual usage was calculated by multiplying the
estimated  quantity  of  solvent  used   per  tank  cleaning by  the
frequency of tank cleaning.

D.I.3  Cleaning of Fixed Exterior Surfaces
     The approach used fdr estimating  emissions  from the cleaning
of fixed exterior surfaces  was  the same as the cleaning of tanks
and  the cleaning  of removable  parts.  The plant  could either
provide  data   on  cleaning  solvent usage,  or  perform  the  test
described  in  the protocol.   The  following  summarizes  the  data

                               D-2

-------
gathering methods:
     Facility A annual usage was extrapolated based on frequency of
cleaning and assumed to equal emissions.
     Facility C  solvent in and number and  type of parts cleaned
were monitored for four hours and extrapolated to usage per year.
It was assumed that usage equalled emissions.
     Facility D annual surface cleaning usage is based on records.
     Facility E usage records were maintained for eight months to
come up  with an  average use per  surface cleaning  event.   This
number was extrapolated to estimate annual surface cleaning usage.
     Facilities F and I annual surface cleaning  usage are based on
records.
     Facility J annual usage was calculated from an estimate of the
quantity of solvent used for each cleaning multiplied by the number
of cleaning events.

D.I.4  Flushing of Fixed Lines
     Data concerning flushing of fixed lines were provided by five
facilities,  none  of  which conducted experiments  to  determine
cleaning solvent used for and emissions from flushing activities.
All facilities provided  such  information based  on plant records.
Three of the five facilities described their operations as closed
systems  and reported  no emissions.   One  facility  provided  an
emission  estimate based  on plant  records.   According to  that
facility, its emissions are  captured and controlled with an overall
control efficiency of 94 percent.   Total  annual  emissions from the
facility  were estimated  as  0.4  ton  per year.    The  remaining
facility also controls emissions through capture and control, but
an  emission  estimate  was  not  provided  because  the  capture
efficiency was unknown.

D.I.5  Other Sources of Emission
     Emissions from other sources  such as storage tanks,  coating
mix preparation,  application  and drying,  packaging and labeling,
waste handling,  and piping leaks were estimated by using plant data
along with emission factors for either the magnetic tape industry
or other industries having  similar processes.
     The storage  tanks emissions,  along with the  waste handling
emissions, were determined by using EPA emission  factors published
in the  "Compilation of  Air Pollutant Emission  Factors (AP-42).
Fourth Edition.   September  1985."
      The mix preparation,  application and  drying emissions were
estimated using the factors determined from the work done on the
Magnetic Tape New Source Performance Standard (NSPS)  (Publication
Number EPA-450/3-85-029a (December 1985)).
      The piping  leaks emission factors were obtained  from the
"Protocols  for  Generating  Unit-Specific Emission  Estimates  for
Equipment Leaks  of VOC and VHAP" (Publication Number EPA-450/3-88-
010, October 1988).

D.2  PERFORMANCE TEST METHODS
                               D-3

-------
     The  reference  test  methods  and  procedures  available for
determination of compliance with an emission limitation, along with
the costs of each procedure,  are discussed  in this section.  Stack
emission  testing  methods  and  methods  for determining solvent
volatile organic compound (VOC) concentrations are similar to those
discussed  in the Background  Information Document  (BID)  for the
Magnetic Tape Manufacturing Industry NSPS.

D.2.1  Stack Emission Testing
     D.2.1.1  Testing Locations,  stack emission testing techniques
would be needed to measure the VOC concentration and gas  flow rate
in stacks and ducts such as:  inlets and outlets of vapor control
devices;  exhaust streams  from mixing equipment and/or storage
tanks;  uncontrolled  exhaust  streams  venting directly to  the
atmosphere; intermediate process  streams such as hood exhausts and
drying  oven  exhausts  venting  to  other  process  units.    The
particular  streams to  be  measured depends  upon the  applicable
regulation.

     D.2.1.2  Overall Control Efficiency. Current performance test
methods and procedures can be used to determine  the overall control
efficiency of the add-on pollution control  system.   A complete air
pollution control system for controlling solvent VOC  emissions from
a magnetic tape  manufacturing plant  consists of  a capture  or
containment  system and  an emission control  device.  The add-on
technologies  used to control solvent  VOC emissions are  carbon
adsorption, condensation, and incineration.  The control efficiency
of each component is determined separately  and  the overall control
efficiency is the product of the capture system and  control device
efficiencies.  This measured overall control efficiency will not,
and is not intended to,  reflect control or  emission  reductions due
to process and operational changes.

     D.2.1.3  Control Device Efficiency.  Three  types of processing
devices are used in the magnetic tape industry; carbon adsorbers,
condensers, and incinerators. The test procedure that can be used
to determine efficiency is the same for each control technology.
     To determine the efficiency of the emission control device,
the solvent VOC mass flow rate in the  inlet and outlet gas streams
must be determined.  To  determine the  mass  of VOC in a gas stream,
both  the concentration  and  flow  rate  must   be  measured.    The
recommended methods are discussed later.

     D.2.1.4   Capture System Efficiency.  The efficiency of the
vapor capture system is defined as the ratio of  the mass of gaseous
VOC emissions  directed  to  the vapor control device to  the total
mass of gaseous VOC emissions available to the capture system for
the magnetic tape coating line, storage tanks, the mixing  room, and
other sources of emission.  The mass of VOC  in each applicable vent
is determined by measuring  the  concentration  and  the  flow rate
using  standard EPA  test  methods.   The recommended methods are
discussed later.

                               D-4

-------
     In  order to determine capture  efficiency,  all fugitive  VOC
emissions  from the coating area,  storage  tanks,  mixing room,  or
other sources must be captured and vented  through  stacks suitable
for  testing.   Furthermore, the  operation  being tested should be
isolated  from  any  fugitive  emissions originating  from  other
sources.  All doors and other openings  through which fugitive  VOC
emissions might escape, or enter,  must  be  closed.

     D.2.1.5  Stack Emission Testing -  Time and Cost.  The length
of a performance test is specified  in the applicable regulation  and
is selected to be representative  for  the process being tested.  The
length of a performance test should  be  selected to be  long enough
to account for variability in emissions due to changing operation
times, routine process problems, and the manufacture of different
products.  Also,  the performance test time period should correspond
to the cycles of  the emission control device.
     In general,  a performance test  would  consist  of three to  six
runs, each lasting from 1/2 to 3  hours.   For most magnetic tape
operations, the field testing could  probably be completed  in 2 to
3 days  (i.e.,  two or three 8-hour work shifts)  with an extra  day
for  setup, instrument preparation, and  cleanup.
     The cost of the testing varies with the length of the test  and
the  number of vents to  be tested:   inlet,  outlet, intermediate
process, and fugitive vents.  The cost to measure VOC concentration
and  flow rate is  estimated at $12,880,  excluding travel expenses.

     0.2.1.6  Details on Gas VQlu.iH?£ric Flow Measurement Method.
          Recommended methods. Reference Methods 1, LA, 2,  2A,  2C,
2D, 3 and 4 are recommended as appropriate for determination of  the
volumetric flow rate of gas streams.
          Large stacks with steady flow. Methods  1 and 2 are used
in  stacks with steady  flow and  with  diameters greater  than  12
inches.  Reference Method  1  is used to select the  sampling site,
and Reference Method 2 measures the volumetric flow rate using a S-
type pitot tube  velocity  traverse technique.   Methods 3 and 4
provide fixed gases analysis and moisture  content,  which are used
to determine the gas stream molecular weight and density in Method
2.  The results are in units of standard cubic meters per  hour.
          Small ducts.  If the duct  is  small  (less  than 12  inches
diameters) then  alternative flow  measurement techniques  will  be
needed using Method 2A, Method 2D,  or Methods 2C and 1A.  Method 2A
uses an in-line turbine meter to  continuously and directly  measure
the volumetric flow.  Method 2D  uses rotameters,  orifice  plates,
anemometers,   or  other  volume rata  or pressure  droo measuring
devices to continuously measure the flow rate.  Methods 1A and 2C
(in  combination)  modify Methods  l  and  2 and  use a  small standard
pitot tube traverse technique to measure the flow  in small ducts,
and apply when the flow is constant and continuous.
          Unsteady flow.  If the flow in a large duct (greater than
12 inches diameter)  is  not  steady or  continuous, then Method 2  may
be modified to continuously monitor the changing flow rate in  the
stack.  A continuous 1-point pitot tube measurement is made at a

                               D-5

-------
representative  location  in  the  stack.    For  small  ducts with
unsteady  flow,  continuous measurement  with Method  2A or  2D is
recommended.
          Adjustment for moisture.  The results do not need to be
adjusted to dry conditions (using Method 4 for moisture)  if the voc
concentrations  are  measured  in  the gas  stream  under actual
conditions; that  is, if  the  VOC concentrations are  reported as
parts of VOC per million parts of actual (wet)  volume  (ppmv).  If
the concentrations are measured on a dry basis (gas chromatographic
techniques  or  Method 25)  then the  volumetric  flow  rate must
correspondingly be adjusted to a dry basis.
          Method  18.   With  Method 18,  "Measurement  of  Gaseous
Organic Compound Emissions by Gas Chromatography (GC)," the major
organic  components  of  a  gas mixture  are separated  by GC  and
individually  quantified  by   flame ionization,  photoionization,
electron  capture,  or  other   appropriate  detection  principles.
However, the sample technique is not a continuous measurement and
would be impractical because of the length of the testing.  Also,
it  would be  costly  and time  consuming to  calibrate for each
compound.   It is a  method which  can  be used if  the  facility is
willing  to  incur  the  additional  costs  as  compared  to  the
recommended method.
          Method 25A.   The recommended  VOC measurement method is
Reference  Method 25A,  "Determination of  Total Gaseous  Organic
Concentration  Using a  Flame  Ionization Analyzer  (FIA)".   This
method  was selected because  it  measures  the expected  solvent
emissions  accurately, is  practical for long-term,  intermittent
testing,  and  provides a. continuous record  of  VOC concentration.
Measurements that are not continuous may  not give a representative
indication of  emissions.   The coating lines in this industry may
operate  intermittently,  and  the  vent  concentrations may vary
significantly.  Continuous measurements and records are easier to
use for  intermittent processes,  and the short-term variations in
concentration can be noted.  The continuous records are averaged or
integrated  as  necessary to  obtain  an  average  result  for  the
measurement period.
     Method 25A applies  to the measurement of total gaseous organic
concentration  of  vapors consisting  of alJcanes,  and/or arenes
(aromatic hydrocarbons).  The instrument is  calibrated  in  terms of
propane  or another  appropriate  organic compound.   A  sample is
extracted from the  source through  a heated sample line and glass
fiber  filter  and routed to a FIA.  Provisions are  included for
eliminating the heated sampling line and glass fiber filter under
some sampling  conditions.  Results are reported as concentration
equivalents of the calibration gas  organic constitutent  or organic
carbon.
     Instrument  calibration  is  based  on  a  single  reference
compound.    For  the magnetic tape  industry,  the  recommended
calibration compound is  propane or butane.   However,  if  only one
compound is used as the sole  solvent at a plant, then that solvent
should  be used as  the  calibration compound.   As a  result,  the
sample  concentration  measurements are on the  basis  of that

                               D-6

-------
reference  compound  and  are  not necessarily  true  hydrocarbon
concentrations.  The response of an FIA is proportional to carbon
content for  similar compounds.  Thus, on  a  carbon number basis,
measured concentrations based on the reference compound are close
to the true hydrocarbon concentrations.  Also, any minor biases in
the FIA concentration results are less significant if the results
will be used in  an efficiency calculation (both inlet and outlet
measurements are made and compared) and biases in each measurement
will tend to cancel out.   For calculation of emissions on a mass
basis,  results  would  be  nearly equivalent  using  either  the
concentration and molecular weight based on a reference gas or the
true  concentration  and  true  average  molecular  weight  of  the
hydrocarbons.
     The advantage of using a single component calibration is that
costly  and  time  consuming  chromatographic techniques  are  not
required to isolate and quantify the individual compounds present.
Also, propane and butane calibration gases are readily available in
the concentration  ranges needed for this industry.
     The predominant VOC compounds used by the industry are methyl
ethyl ketone, methyl isobutyl ketone, and toluene.  Most plants use
a mixture of different  compounds  for  solvent.   Since the solvent
mixtures may vary from day-to-day and from  plant-to-plant, there is
no standard  solvent mixture to use  for calibration.  Also,  the
individual  compounds  in  the  mixture  will  evaporate  and  be
controlled  at different  rates,  so the  gaseous VOC  mix in  the
exhaust stream is not the same mix as the original multi-component
liquid solvent.  Furthermore,  if  incineration  is  used,  any semi-
destructed gaseous compounds at  the  incinerator outlet will  be
different  from the  compounds in  the original solvent  mixture.
Thus, there is no advantage in calibrating the inlet and outlet FIA
with the mixture of solvents being used.

D.2.2  Liquid Solvent Material Balance
     If a plant's  add-on control device recovers solvent (such as
carbon adsorption  or condenser systems),  then a  liquid  solvent
material balance approach could be used to determine the overall
efficiency of the vapor control system.  This is done by comparing
the solvent used versus the solvent recovered.   The difference is
assumed to be emitted.  These values may be obtained from a plant's
inventory records  or through actual measurements.   The EPA has no
test  procedure to independently  verify  the plant's  accounting
records.   However/  the  plant must  set   up and  submit  to  the
enforcement  agency through  the permitting process its  proposed
inventory  accounting  and  record-keeping system  prior  to  any
performance testing if it is determined that  is the means by which
the material balance will be performed.
     For this compliance demonstration approach,  the averaging time
(compliance demonstration time period) usually needs to be 3 days.
This  averaging period  allows  for  a  representative  variety  of
coatings and tape products,  as well as  reducing the impact  of
short-term variations due to process  upsets, solvent  spills,  and
variable amounts of solvent in use in the process.

                               D-7

-------
     The volume of solvent: recovered may be determined by measuring
the level of solvent  in  the  recovered solvent storage tank.  The
storage  tank  should  have  an  accurate,  easily readable  level
indicator to  within  one-tenth  of one  percent,  i.e.  1  mm.   To
improve the precision of  the  volume measurement,  it is recommended
that the recovered solvent tank have a relatively small diameter,
so that small changes in volume result in greater changes in tank
level.   Alternatively,  the  solvent  recovered  may be  measured
directly by using a  liquid volume meter  in the solvent return line
that is capable of one cubic centimeter determinations.     The
volume of solvent used may be determined from plant inventory and
purchasing records or  by measuring the level  in the solvent storage
tank.  Alternatively,  a liquid volume  meter  can be used to measure
the amount of solvent drawn off from the solvent storage tank.

D.3  MONITORING SYSTEMS AND DEVICES

     The purpose of  monitoring  is to  ensure that  the  emission
control system is being properly operated  and maintained to assure
continual compliance  after the performance  test.  One can either
directly monitor the regulated pollutant, or instead,  monitor a set
of operational  parameters  for the  process  and  emission control
system.  The aim is to select a relatively inexpensive and simple
method  that will determine  that  the  facility  is  in  continual
compliance with the standard.
     The three types of vapor control  devices that are expected to
be used in the  magnetic  tape  industry are carbon  adsorbers,
condensers, and incinerators.  Possible monitoring approaches and
philosophy for each part of  the VOC control system are discussed
below.
D.3.1  Monitoring of Vapor Control Devices
     D.3.1.1  Monitoring  in Units of Efficiency.    This monitoring
requires measuring inlet and exhaust VOC concentrations, along with
the inlet and exhaust volumetric flow rates.  An overall cost for
a complete monitoring system is difficult to  estimate due to the
number  of   component  combinations possible.    The  purchase  and
installation cost of  an  entire monitoring  system  (including VOC
concentration  monitors,  flow  measurement  devices,   recording
devices, and automatic data reduction) is  estimated to be $25,000.
The total annual costs are estimated at $17,691 per year.
     D.3.1.2  Monitoring in Units of Mass Emitted.   Monitoring in
units of mass of VOC emitted would require concentration and flow
measurements only at the exhaust location.  The cost is estimated
at $12,500 for purchase and installation plus $12,500 annually for
operation,  maintenance, calibration,  and data reduction.
     D.3.1.3  Monitoring  of Exhaust VOC  Concentration.  Monitoring
equipment  is  commercially  available,  however,  to  monitor  the
operational or process variables  associated  with vapor control
system  operation.     The  variable  which would yield  the  best
indication  of  system operation   is  VOC  concentration  at  the
processor outlet.  Accurate measurements are required because the
purpose  of  the  monitoring  is to  determine the outlet  emission

                               D-8

-------
concentration to  be used  for  continuous compliance purposes.
Monitors for this type of continuous VOC measurements, including a
continuous recorder,  typically cost about $6000  to purchase and
install, and $6000  annually to calibrate,  operate, maintain, and
reduce  the data.   To  achieve representative  VOC concentration
measurements at the control outlet,  the concentration monitoring
device  should  be  installed in the  exhaust vent at  least two
equivalent stack diameters from the exit point,  and protected from
any interferences due to wind, weather, or other processes.
     0.3.1.4  Monitoring of Process Parameters.   For  some vapor
control systems,  there may be a set  of  process parameters other
than the exhaust VOC concentration which is an accurate indicator
of system operation.  Because control system design is constantly
changing  and being upgraded  in this  industry,   all  acceptable
process   parameters  for   all   systems  cannot   be  specified.
Substituting the  monitoring of vapor processing  system process
parameters for the monitoring of exhaust VOC concentration is valid
and acceptable if the process parameter is deemed to be an equally
accurate measure of  system performance.  If parameter monitoring is
selected  as  the method  of  establishing  compliance  with  the
standard,  sources will,  in most cases, be required to establish
site specific limits for the appropriate control device parameters
during  the  initial  compliance determination.    Site  specific
determination  of  these  compliance  limits  would  address  the
variabilities in  operating conditions and designs of  individual
control   devices.   Parameter   monitoring   equipment,    such  as
temperature  monitors,  would  typically cost  $1500  to  purchase.
Operating  costs,  including maintenance,  calibration,   and  data
reduction, would be about $3896 annually.
     O.3.1.5    Monitoring  of  Carbon  Adsorbers.    For  carbon
adsorption  vapor processing devices,  the preferred  monitoring
approach  is  the use of  a  continuous VOC exhaust concentration
monitor.    Alternatively,  the  carbon  bed  temperature  (after
regeneration and completion of any  cooling cycles), and the amount
of  steam  used  to  regenerate  the bed have  been  identified  as
parameters which  could demonstrate compliance  with  the emission
standard.    Temperature  monitors  and  steam  flow meters  which
indicate  the  quantity of  steam used over a  period of  time are
available.
     D.3.1.6 Monitoring of Condensers.  For condenser devices, the
temperature  of  the  exhaust stream    has  been  identified  as  a
parameter  which  could demonstrate  compliance with  the emission
standard, and condenser temperature monitors are available.
     D.3.1.7    Monitoring  of   Incinerators.    For  incineration
devices, the exhaust concentration is quite low and is difficult to
measure accurately with the inexpensive VOC monitors.  Instead, the
firebox temperature has  been identified  and  demonstrated to  be a
process parameter which determines the efficiency of the level of
emissions  from the  device.   Thus, temperature  monitoring  is the
recommended monitoring approach for incineration control devices.
Since a temperature monitor  is usually included as  a standard
feature  for  incinerators,  it  is  expected  that  this  monitoring

                               D-9

-------
requirement will not incur additional costs to the plant.
     O.3.1.8  Use of Monitoring Data.  The use of monitoring data
is the same regardless of whether the VOC outlet concentration or
an  operational  parameter is  selected  to  be monitored.    The
monitoring system should be installed and operating properly before
the first performance test.  Continuous compliance is determined by
comparing the monitored value of the  concentration or parameter to
the  value which  occurred during  the last  successful  test,  or
alternatively, to a preselected value which demonstrate compliance.
0.3.2  Monitoring of Vapor Capture Systems
     O.3.2.1  Monitoring  in  Units  of Efficiency.   Monitoring the
vapor  capture  system  in  the  units of  efficiency would be  a
difficult and  costly procedure.   This monitoring  approach would
require measuring the VOC concentration and volumetric flow rate in
the inlet to the vapor processing device and in each fugitive VOC
vent and then combining the results to calculate an efficiency for
each  time  period.     Such  a  monitoring system  has  not  been
commercially demonstrated.   The  purchase and  installation of an
entire monitoring system  is  estimated at  $12,500  per  operation,
with an  additional  $12,500 per year for operation,  maintenance,
calibration, and data reduction.  Thus, monitoring in the units of
efficiency is not recommended.
     D.3.2.2  Monitoring of Flow Rates.   The key to a good capture
system is maintaining proper flow rates in each vent.   Monitoring
equipment  that   can   monitor  these  flow  rate   parameters  is
commercially available.  Flow rate monitoring system for each vent
would typically cost about $3000 plus $3000  annually  to operate,
maintain, periodically calibrate,  and reduce  the data  into the
desired format.  The monitored flow rate values are then compared
to the monitored value during the lass successful performance test.
     Proper  flow  rates and  air distribution in a vapor capture
system could be ensured by an inspection and. maintenance program,
which generally would not create any additional cost burden for a
plant.  In that case, the  additional  value of information provided
by flow rate monitors  would  probably be  minimal.   Routine visual
inspections of the fan's  operation would indicate  whether or not
capture efficiencies remain at the performance test level, and no
formal monitoring of the air distribution system would be required.
D.3.3  Monitoring of  Overall  Control  System Efficiency on a Liquid
Basis
     If  a  plant uses  a vapor  recovery  control  device,  the
efficiency of  the overall plant control  (combined vapor capture
device and vapor recovery systems)  can be monitored using a liquid
material balance.  The amount  of solvent used  is  compared to the
amount of  solvent recovered.  These values  are obtained from a
plant's  inventory records.   For  this  monitoring approach,  the
averaging time or monitoring period usually needs to be longer than
a 24-hour period, and should  reflect  the  cycle time of the solvent
recovery device.   This longer averaging period is necessary to
coordinate  with  a  plant's  inventory accounting  system  and  to
eliminate short-term  variations  due to process upsets,  solvent
spills, and variable amounts of solvent in use in the process.

                               D-10

-------
     Because  most plants  already keep  good  solvent usage  and
inventory  records,  no  additional cost  to  the  plant would  be
incurred for this monitoring approach.
                              D-ll

-------
          APPENDIX  E.

ENVIRONMENTAL AND ENERGY IMPACTS
     OF THE CONTROL OPTIONS

-------

-------
                           APPENDIX E.

                 ENVIRONMENTAL AND  ENERGY IMPACTS
                      OF  THE  CONTROL  OPTIONS


     The environmental and energy impacts of the regulatory
alternatives on individual emission sources at existing magnetic
tape manufacturing facilities and new model lines are presented
in this Appendix.  The emission points and associated control
options identified in Table E-l are included in this analysis.
For a detailed description of the control options in Table E-l,
refer to Chapters 4 and 6.  The assumptions used to calculate
these impacts were presented in Chapter 7.  The environmental and
energy impacts of the regulatory alternatives that result from
combining the impacts of the various emission points were also
presented in Chapter 7.

     The environmental impacts of controlling each emission point
at existing major sources are presented in Table E-2.  (A major
source is defined in Chapter 7,  Section 7.0.)  The environmental
impact of controlling each emission point in each of the five
model lines is provided in Table E-3.  (Model line parameters are
provided in chapter 6.)

     The natural gas requirements of the control options and
associated secondary pollutants are presented in Table E-4 for
existing major sources that will use incineration to control
emission points.  No new lines are assumed to use incineration so
there are no natural gas requirements for new lines.

     The fuel oil requirements of the control options and
associated secondary pollutants are presented in Table E-5 for
existing major sources.  The same information is presented in
Tables E-6 and E-7 for new sources.

     The quantity of wastewater discharge and associated HAP
solvent emissions from each emission point are provided in
Table E-8 for existing major sources.  Wastewater discharges and
emissions from new lines are shown in Table E-9.

     The solid waste impacts from existing major sources are a
result of new control devices at one facility.  The facility, a
small facility, would need to control emissions from storage


                               E-l

-------
tanks, mix preparation, and coating application/drying.  The
solid waste impacts are 75 Kg/yr (160 Ib/yr).  For new lines,
only the small line would require a new control device.  It would
be used to control emissions from storage tanks, mix preparation,
and coating application and drying.  The solid waste impacts are
75 Kg/yr (160 Ib/yr).

     The total energy requirements associated with controlling
each emission point are presented in Table E-10.  The energy
requirements include natural gas, steam, and electricity
requirements.  There are no natural gas requirements for new
lines since none are assumed to use incineration as a control
method.  Steam and electricity requirements for new lines are
presented in Table E-ll.  Total energy requirements for new lines
are presented in Table E-12.

-------
   TABLE E-l.   CONTROL  OPTIONS FOR HAZARDOUS  AIR  POLLUTANT
    EMISSIONS  FROM MAGNETIC  TAPE MANUFACTURING FACILITIES
Emission point
^ •, >. -. . v f
"
Storage tanks
Mix preparation
Coating application/drying
Waste handling
Piping leaks (plantwide)
Participates
,.

Removable parts
Flushing of fixed lines
Control
»»„_*>__ ____,i{_.__
taflifiBg. qpetattoat:
Vent to 95% efficient control device
Cover/vent mix vessels to 95% efficient
control device
Total enclosure vented to 95% efficient
control
Direct the waste handling device process
vent to a 95% efficient control device
Implement LDAR program
Enclosed transfer
•••'--. ••••::
Cfe*a»ag aetmtfe*
75% Freeboard ratio or equivalent through
venting to control
Closed system or vent supply and waste
collection vessels to 95% efficient control
device
Overall HAP
control,
percent*


95
95
95
95
89
99.9


88
95
aThe control efficiency indicated is the control for emissions from that specific emission source, not the
 entire facility.
                                  2-3

-------
   TABLE B-2.   ANNUAL ENVIRONMENTAL  IMPACT OF THE CONTROL
                OPTIONS ON EXISTING MAJOR  SOURCES
Emission point*
Storage tanks
Mix preparation
Coating application/drying
Waste handling
Piping leaks
Removable parts cleaning
Flushing of fixed lines
Particulates
Emissions after implementation,
Mg/yr (tons/yr)
0.33 (0.36)
105 (115)
1,033 (1,136)
112(123)
48(53)
58 (64)
0(0)
0.12(0.13)
Percent reduction in
industrywide HAP emissions
75
86
44
65
89
88
Ob
70
aFor a description of the control options, refer to Table E-l.
 All facilities that conduct this cleaning activity currently have this level of control.
                                   3-4

-------
co
R
S3

§

1
o

o
8
w
fa
o
z
H
§
H
W
PS


J

         1
         fiJ
ii
•a |
            "a
             I
             .I
               ss
               e o
               SS
               b o
               S!
               6 S
               §
               o S
                   s;
                   6 s,
                   S!
                   i
                      a C

                         ss
                         6 2,
                         si
                         I
                             «
                             !
                             |
                             s
                                «
                                as
                                 '
                                3.

                                1.
                                    g

                                    O
                                    e
                                    a
                                    o
                                    s
                                    o
                                    0
                                    g
                                    s
                                    9
                                    1
                                    
-------
TABLE E-4.
                 NATURAL GAS  CONSUMPTION AND  ASSOCIATED  POLLUTANTS
                        FOR  EXISTING MAJOR SOURCES*
Emission point
Storage tanks
Mix preparation
Coating application/drying
Waste handling
Piping leaks
Removable parts cleaning
Flushing of fixed lines
Particulates
Natural gas
consumption,
GJ/yr (10° Btu/yr)
0(0)
70(66)
0(0)
0(0)
0(0)
0(0)
0(0)
0(0)
PM,
Kg/yr (lb/yr)
0(0)
0.07(0.16)
0(0)
0(0)
0(0)
0(0)
0(0)
0(0)
NOX,
Kg/yr (lb/yr)
0(0)
4 (8.8)
0(0)
0(0)
0(0)
0(0)
0(0)
0(0)
CO,
Kg/yr (lb/yr)
0(0)
1 (2.2)
0(0)
0(0)
0(0)
0(0)
0(0)
0(0)
                                           [0.35(1400 - 160)] + 160 = 594°F
 aNatural gas is used as fuel for incinerators. New lines are assumed to use carbon adsorbers to control
  solvent HAP emissions; therefore, there is no natural gas consumption for new lines.
 "For a description of the control options, refer to Table E-l.
Calculation method:
•  Assume 35% heat recovery
•  Temperature of gas entering incinerator:  Tmc
•  Mass flow rate of gas, lb/hr
   [(_scfm)(60 min/hr)]/[13.1 f^/lb air]
•  Heat required,  106 Btu/hr:
   Enthalpy 590 = 253.46
   Enthalpy 1500 = 493.64

   Q = (_lb/hr)(493.64 - 253.46)

   Assuming 10% heat loss:  Q1 = Q x 1.1
•  Heat available from VOC, 106 Btu/hr:
   -25% LEL = !3Btu/scf
   QVOC = (13 Btu/scfX	scfm)(60 min/hr)
•  Heat to be supplied, 106 Btu/hr:

   QJf = Q1 - Qvoc
•  Natural gas, Btu/yr
   [(_hr/yr)(_Btu/hr)]
                                         3-6

-------
   TABLE E-5.
FUEL  OIL  CONSUMPTION  AND ASSOCIATED  POLLUTANTS
     FOR  EXISTING MAJOR  SOURCESa
Emission point''
Storage tanks
Mix preparation
Coating application/
drying
Waste handling
Piping leaks
Removable parts
cleaning
Flushing of fixed lines
Particulates
Fuel oil
consumption,
103 L/yr (103 gal/yr)
0.5 (0.1)
390 (103)
142 (38)
71 (19)
0(0)
0(0)
0(0)
0(0)
PM,
Kg/yr (lb/yr)
1.1 (2.4)
840 (1,850)
310 (680)
150 (340)
0(0)
0(0)
0(0)
0(0)
sox,
Kg/yr (U>/yr)
15 (32)
11,000 (24,240)
4,020 (8,850)
2,000 (4,410)
0(0)
0(0)
0(0)
0(0)
NOX,
Kg/yr (lb/yr)
3(7)
2,570 (5,660)
940 (2,070)
470 (1,030)
0(0)
0(0)
0(0)
0(0)
aFuel oil is used to generate steam for stripping of solvents from the carbon beds during regeneration of a
 carbon adsorber.  The fuel oil requirement is based on the steam requirement and a boiler thermal
 efficiency of 80 percent.
       	Btu/yr steam x     eal fuel oil     =	gal/yr fuel oil
                        150,000 Btux 0.8

       For the steam requirement estimating methodology, refer to Table E-10.

"For a description of the control options, refer to Table E-l.

-------
         TABLE  E-6.    FUEL  OIL  CONSUMPTION  FOR  NEW  LINESa
Emission point
Storage tanks
Mix preparation
Coating application/
drying
Waste handling
Piping leaks
Removable parts
cleaning
Flushing of fixed
lines
All cleaning
activities
Particulates
Fuel oil consumption, IcP L/yr (1(P gal/yr)
Small
0.004 (0.001)
0.6 (0.16)
5 (1.4)
N/Ad
0(0)
0(0)
0(0)
0(0)
0(0)
Medium lines0
A
0.01 (0.003)
24(6)
0(0)
N/Ad
0(0)
0(0)
0(0)
0(0)
0(0)
B
0.01 (0.003)
0(0)
0(0)
N/Ad
0(0)
0(0)
0(0)
0(0)
0(0)
Large lines0
A
0.3 (0.07)
40(10)
0(0)
0.7 (0.2)
0(0)
0(0)
0(0)
0(0)
0(0)
B
0.3 (0.07)
0(0)
0(0)
0.7 (0.2)
0(0)
0(0)
0(0)
0(0)
0(0)
aFuel oil is used to generate steam for stripping of solvents from the carbon beds during regeneration
 of a carbon adsorber.
''For a description of the control options, refer to Table E-l.
cline A is a new line that was not built concurrently with a VOC control device.  Line B is a new line
 that was built concurrently with a VOC control device.
 N/A means that the model line is assumed not to perform waste Handling operations.
                                         S-3

-------
Q
H
CQ
D
O

J
CO
p
U
  H
CO
o
04
r-
w

J

I

3
&


S
•J
T

^
J
a
1

_,,
J
a
1


__
in


I"
*„
81
1
"i"
•o
0-
M
i
i-
o*1
<»
>
i"
V
1
i"
V
i
a
il
-s
„ J

d 3,
-§

r- 2

0 5
52
o g.
9 a
3S
9 a
6 o

-------
  TABLE E-8.   WASTEWATER DISCHARGES  AND ASSOCIATED  SOLVENT
              EMISSIONS FROM EXISTING MAJOR SOURCES3
Emission point''
Storage tanks
Mix preparation
Coating application/drying
Waste handling
Piping leaks
Removable parts cleaning
Flushing of fixed lines
Particulates
Wastewater discharge,
103 L/yr (103 gal/yr)
0(0)
3,690 (975)
1,065 (280)
840 (220)
0(0)
0(0)
0(0)
0(0)
HAP solvent emissions,
Kg/yr (lb/yr)
0(0)
185 (405)
55 (120)
40(90)
0(0)
0(0)
0(0)
0(0)
aThe quantity of wastewater is based on the use of 4 Ib of steam to desorb 1 Ib of HAP solvent.  The
 emission estimates are based on a wastewater concentration of 100 ppm.
''For a description of the control options, refer to Table E-l.
                                   E-10

-------
CO

o
H
CO
CO
H
O
CO

CU
Q
a

  ta
CO
w
u
CO
H
Q

OS
a
Ei
en

a

a
    &
    i

    !
     J
     !
     I


           *
               23
               ss
               « o,
               23
                  eg
                  2§
                  si
                  o§
                  i
                      o
                       l
                          o S
                                 Ji
                                     0
                                     o g
                                     o g
                                     o g
                                         o g
                                         31
   8
   •
   €
   I
   i
   I
                                                    -1
                                                 «-3


                                                 I

                                                IJM
                                                1 o J 3
                                                J 1.3 i
                                                3 a • 3
 lll
« -O •_) -3
                             3-11

-------
TABLE  E-10.   TOTAL ENERGY  REQUIREMENTS  AT  EXISTING
                            MAJOR SOURCES
Emission point*
Storage tanks
Mix preparation
Coating application/drying
Waste handling
Piping leaks
Removable parts cleaning
Flushing of fixed lines
Pariiculates
Natural gas
requirements,
GJ/yr
(106 Btu/yr)
0(0)
70(65)
0(0)
0(0)
0(0)
0(0)
0(0)
0(0)
Stettn rccgmrctnofilB*
GJ/yr
(106 Bni/yr)b
17 (16)
13,Q20(12,350).
4,755 (4,510)
2,370 (2,250)
0(0)
0<0) '
0(0)
0(0)
Electricity
requirements,
GJ/yr
(10° Btu/yr)c
0(0)
21(20)
580 (550)
0.03 (0.03)
0(0)
0(0)
0(0)
3(3)
Total energy
requirements,
GJ/yr
(10° Btu/yr)
17 (16)
13,110(12,435)
5,335 (5,060)
2,370 (2,250)
0(0)
0(0)
0(0)
3(3)
"For a description of the control options, refer to Table E-l .
 Energy requirements for steam and electricity were calculated as follows:
 Steam requirementst                                                      /i
       (   solvent HAP emissions, Ib/yrK  control efficiency)(4 Ib steam/1 Ib solvent) \ ib
cElectriciry requirements:
       (   airflow. cfinX5 HP/1,000 CFM)(2,544 Btu/hr/HPX    operating hr)

For the calculation of natural gas requirements, refer to Table E-4.
                                                                   steam
                                   E-12

-------
CO
W
pa
pel
O
fc

CO
H
D

Ed
c*
H
O
H
Ed
CO
Ed

Ed
     j
            *'
              -

          !°i»
           i .?•

                  o
                    I
                        «
                              0 §
                              00
                              o
                                     3
                                     3 §
                                     i
                                           e
                                           o
                                                        o
                                                                0 g
                                                        o
                                                                og
                                                                        og
                                                 og
                                                        o g
                                                                                1
                                                 o
                                                                0
                                                 o g
                                                                                I
                                                                                I
                                                                                I
                                                                                .3
                                                                                at


                                                                                j

                                                                                o
                                                                                .a

                                                                                "3

                                                                                *§
                                                                              15
Remo

put*


-------
 TABLE E-12.   TOTAL ENERGY REQUIREMENTS  FOR NEW LINES, GJ/yr
                               (106 Btu/yr)
Emission point*
Storage tanks
Mix preparation
Coating application/drying
Waste handling
Piping leaks
Removable parts cleaning
Flushing of fixed lines
Particulates
Small
0.16 (0.15)
20(19)
194 (185)
N/A
0(0)
0(0)
0(0)
0.3 (0.3)
Mediumb
A
0.3 (0.3)
792 (752)
0(0)
N/A
0(0)
0(0)
0(0)
1.6 (1.5)
B
0.3 (0.3)
0(0)
• 0(0)
N/A
0(0)
0(0)
0(0)
1.6 (1.5)
Largeb
A
8.5 (8)
1,327 (1,262)
0(0)
25(23)
0(0)
0(0)
0(0)
1.6 (1.5)
B
8.5 (8)
0(0)
0(0)
25(23)
0(0)
0(0)
0(0)
1.6 (1.5)
aFor a description of the control options, refer to Table E-l.
TJne A is a new line that was not built concurrently with a VOC control device.  Line B is a new
 line that was built concurrently with a VOC control device.
                                    3-14

-------
            APPENDIX F.



COST IMPACTS OF THE CONTROL OPTIONS

-------

-------
                           APPENDIX F.

     SAMPLE CALCULATIONS OF COST IMPACTS FOR CONTROL OPTIONS


     This appendix contains sample calculations to supplement the
explanation of cost calculations in Chapter 8.  This appendix
also describes in detail the methodology used to calculate costs
to control equipment leaks from piping.  Sample calculations are
provided to demonstrate how control costs were calculated under
each of the five scenarios identified in Chapter 8.  These
scenarios are:

     1.  The facility has an existing carbon adsorber with
sufficient capacity to control any emission points that may
require control under the NESHAP and are not currently
controlled;

     2.  The facility has an existing incinerator with sufficient
capacity to control any emission points that may require control
under the NESHAP and are not currently controlled;

     3.  The facility does not have any control device and will
need a new control device to meet the requirements of the NESHAP;

     4.  The facility only uses HAP solvents for cleaning; the
cost to control will be associated only with the cost of the
cleaning control options; and

     5.  The facility only requires control of equipment leak and
cleaning emissions to meet the requirements of the NESHAP.

The small model line is described by Scenario 3.   The medium and
large model lines (built with and without concurrent construction
of a control device) are described by Scenario l.

F.I  SAMPLE CALCULATIONS

F.l.l  Scenario l

     F.I.1.1  Basis.  A facility is adding one coating line that
could be classified as medium-sized based on the total annual
quantity of solvent to be used.  The facility has an existing
carbon adsorber to control coating application and drying
emissions from other lines and recovers the solvent onsite.  The
NSPS regulations require that the facility capture and control
93 percent of emissions from the new line as well.  The facility

-------
plans to do this by building a total enclosure around the coater
and oven and venting the enclosure to the existing 95 percent
efficient control device.  In order to comply with Regulatory
Alternative (RA) II of the NESHAP, the facility must control HAP
emissions from storage tanks, mix preparation vessels, equipment
leaks from piping, particulates, and cleaning activities.  The
existing carbon adsorber will be used to control emissions from
the mix preparation vessels and storage tanks.  The sample
calculation includes calculations of the total capital
investment, total annual cost, and cost effectiveness.
     F.I.1.2  Parameters.

      1.  Coating line classification

      2.  Baseline HAP emissions from:
            Mix room  (includes 40 percent control
              through use of covers on mix
              equipment)
            Solvent storage
            Coating application/drying
            Cleaning
            Equipment leaks in solvent recovery
            Equipment leaks in mix room

          Total baseline solvent HAP emissions

      3.  Solvent HAP emissions with RA II

      4.  Additional HAP to adsorber as a result of
            controlling
          Mix room emissions  [equals baseline
            divided by  (1 - 0.4)]
          Storage tank emissions

          Total additional HAP to adsorber

      5.  Additional airflow  (acfm) to
            adsorber  (mix equipment--one line)

      6.  Airflow needed to control
            particulate emissions  (one line)

      7.  Operating hours

      8.  HAP emissions controlled by LDAR and
            equipment modifications  (89 percent
            of uncontrolled emissions):
          Mix room
          Solvent recovery

      9.  Number of wash sinks for cleaning
      Medium


 4.2 tons/yr


 0.1 tons/yr
 3.2 tons/yr
  0.7 ton/yr
 1.2 tons/yr
  0.3 ton/yr

 9.6 tons/yr

 5.2 tons/yr
 7.0 tons/yr
  0.1 ton/yr

 7.1 tons/yr


19.8 ft3/min


19.8 ft3/min

 6,000 hr/yr
      1 ton
    1.1 tons
                               F-2

-------
  Nqte:  The amount of HAP's  controlled by  LDAR  and equipment
         modifications in  the mix  room is actually
         0.3 ton/yr; however, applying the  cost  effectiveness
         factor  to this amount  results in an  unrealistically
         low control cost.  As  discussed in Chapter 8,  the
         minimum cost of controlling  equipment leak emissions
         corresponds to controlled emissions  of  l  ton/yr.

  10.  Total capital investment

       TABLE F-l.  TOTAL CAPITAL INVESTMENT (TCI)--
               RA II APPLIED  TO SCENARIO 1
Emission point
Storage tanks
Mix preparation
Equipment leaks
Particulates
Removable parts
Closed
containers
TCI
Capital cost, $
27,650
5,300
Included in total annual
cost since overall
factor was used
80,000
340
400
113,690
Source in Chapter 8
Table 8-4
Table 8-5
Section 8.5;
Appendix F, Part F.2
Section 8.3.1.5
Section 8.3.1.6
Section 8.3.1.7

  11.  Total annual  cost

TABLE F-2.  TOTAL ANNUAL COST--RA II APPLIED TO SCENARIO 1
COM factor
Operating Ubor
Utilities:
Electricity: (19.8 fi?/nua)(.S hp/1,000 fi?/aaa)<<3.746 kWVhp)
(6,000 hr/yr)($0.049/kWh)
Steam: (4 Ib iteun/lb HAP)(7 ton* x 2,000 Ib HAP/ton)
($6/1,000 Ib .team)
Water (3 .43 gal HjO/Ib ste«m)(56,000 Ib steam)
($0.936/1, 000 gal HjO)
Raw materials
Maintenance
Annual cost
0
$43
$336
$180
0
0
Comments
It is assumed that no additional labor is needed
simply because more emission points an being
controlled with the existing adsorber.
Sources:
References 1,2
References
Reference 3
It is assumed thst additional emission ooinu will ;
not decrease the life of the carbon. Thus, the
incremental raw material cost is zero.
It is assumed that no additional maintenance is
needed simply because more emission poults are
being controlled with the existing adsorber.

-------
                     TABLE F-2.  (.continued)
Con factor
Distillation system coitt
Original heat requireroeoii: 6,240,000 BTU/hr
New heat requirement*: (6,240,000 BTU/hr)((63. 3 + 7)/63.3)]
Incremental heat requirement*: 6,930,000 - 6,240,000 BTTU/hr
Heat coit: (690,000 BTU/hr)/(915.5 BTU/Ib)(6,000 hr/yr)
($6/1, 000 Ib steam)
Indirect com
Administrative (2% of TCI): ($U3,690)(0.02)
Property tax (1 % of TCI): ($113,690)(0.01)
Insurance (1% of TCI): ($1 13,690)(0.01)
Capital recovery (TCI)[CRF (i=7%, n- 10 yn)]:
($113,690)(0.1424)
Solvent credit
Torn HAP recovered: (7 tona/yr)(0.9S controUed)(0.9 distilled)
Credit: (6 ton*/yr)($372/ton)
Equipment leak* (mix room and solvent recovery)
($5,343/ton)(l ton controUed) +
($931/ton)(l.l ton* controUed)
Total annual cost
Annual coat
6,930,000
690,000
$27,130
$2,259
$1,130
$1,130
$16,189
6
($2,226)
$6,367
$52,538
Comment*
Additional heat requirements of the distillation
system are calculated by multiplying the original
heat requirement* by the new HAP loading to the
system, divided by the original HAP loading to the
system. The original heat requirement of
6,240,000 wa* obtained from the cost analysis
done for the NSPS.4
Indirect COM* are a percentage of the total capital
investment associated with controlling emission*
from solvent storage, mix preparation, coating
application/drying, waste handling, and paniculate
transfer.^
See Chapter 8, Section 8.3.2.1 for explanation.
Calculation is baaed on an 89% efficiency for
LDAR, a cost of J5,343/ton controUed for the mix
room, and a cost of $93 I/ton controlled for the
solvent recovery area. Refer to Chapter 8,
Section 8.5 for an explanation of the methodology
used to calculate the cost effectiveness factors.

12 .   Cost effectiveness

     Total annual cost (see Tables F-l and F-2)
     Emission reduction
     Cost effectiveness
                                                         $52,538
                                                      4.4 tons/yr
                                                      $ll,940/ton
     Capital recovery costs were calculated assuming a 10 -year
useful life (n) of all equipment and an average interest rate  (i)
of 7 percent . °  The capital recovery factor is calculated as
follows:
     CRFC - [

The capital recovery factor is 0.1424.

F.I. 2  Scenario 2

     F.l. 2.1  Basis.  One of the existing facilities that falls
under this scenario requires control of its mix vessels, cleaning
operations, and equipment leaks in the mix room.  Based on the
annual solvent usage per line, the facility is classified as
small.  The facility currently controls coating application and
drying emissions with an incinerator.  The facility does not use
particulate HAP's.  The existing incinerator will be used to
control emissions from the mix preparation vessels and storage
tanks.  The sample calculation includes calculations of the total
capital investment, total annual cost, and cost effectiveness.

-------
P.1.2.2
parameters

1.  Coating line classification
          2.
          3,

          4,
          6,

          7,

          8


     *Note:
    Baseline HAP emissions from:
    Mix room (includes 40 percent control
    through use of covers on mix equipment)
    Coating application/drying
    Cleaning
    Equipment leaks in mix room
    Total baseline solvent HAP
    emissions

    Solvent HAP emissions with RA II

    Additional HAP to incinerator as a
    result of controlling:
    Mix room emissions [equals baseline
    divided by (1 - 0.4)]

    Additional airflow (acfm)  to the
    incinerator (mix equipment--one
    line)

    Operating hours

    Number of wash sinks

    HAP emissions controlled by LDAR in mix
    room
                                                            Small
                                                       0.9 ton/yr

                                                       0.2 ton/yr
                                                       0.4 ton/yr
                                                       0.1 ton/yr
                                                      1.6 tons/yr
 0.5 tons/yr



 1.4 tons/yr


13.2 ft3/min



 2,000 hr/yr

           2


      1 ton*
   The amount of HAP's controlled by LDAR and equipment
   modifications in the mix room is actually
   0.1 ton/yr; however, applying the cost effectiveness
   factor to this amount results in an unrealistically
   low control cost.  The minimum cost of LDAR
   corresponds to controlled emissions of 1 ton/yr.
                               P-5

-------
  9.  Total  capital  investment
          TABLE F-3.  TOTAL CAPITAL INVESTMENT-
                RA II APPLIED  TO SCENARIO 2
Emission point
Mix preparation
Equipment leaks
Removable parts
Closed
containers
TCI
Capital cost, $
5,400
Included in total annual
cost since overall factor
was used
940
400
6,740
Source in Chapter
8
Table 8-5
Section 8.5;
Appendix F Part F.
2
Section 8.3.1.6
Section 8.3.1.7

  10.   Total  annual  cost

TABLE F-4.  TOTAL ANNUAL COST--RA II APPLIED TO SCENARIO 2
Cost factor
Operating labor
Utilities:
Electricity: (1.7 x 10-4)($0.049/kWh)(dP « 8 in. HjO)
(13.2 ft3/min)(2,000hr/yr)/(60%
efficiency)
Natural gas: (O.OllxlO6 tt?/yr)($2. 77/1, 000 ft3)
Maintenance
Indirect costs
Administrative (2% of TCI): ($6,740)(0.02)
Property tax (1% of TCI): ($6,740)(0.01)
Insurance (1% of TCI): ($6,740)(0.01)
Capital recovery (TCI)[CRF (i=7%, a=10 yrs)]:
($6,740)(0.1424)
Annual cost
0
$2
$32
0
$108
$54
$54
$960
Comments
It is assumed that no additional
labor is needed simply because
more emission points are being
controlled with the existing
incinerator.
Sources:
Reference 7
Chapter 8, Table 8-11; Reference
8
It is assumed that no additional
maintenance is needed simply
because more emission points are
being controlled with the existing
incinerator.
Indirect costs are a percentage of
the total capital cost to control
emissions from solvent storage,
mix preparation, coating
application/drying, waste
handling^ and paniculate
transfer.-5
                            F-6

-------
                     TABLE F-4.   (continued)
Cost factor
Equipment leaks (mix room and solvent recovery)
($5,343/ton controUed)(l ton controlled)
Total annual cost
Annual cost
$5,343
$6,553
Comments
Calculation is based on an 89%
efficiency for LDAR and a cost
of $5,343/ton controlled for the
mix room. For an explanation of
how the cost effectiveness factors
were calculated, refer to
Chapter 8, Section 8.5 and
Part F.2 of this appendix.

     11.  Cost effectiveness:

          Total annual cost
          Emission reduction
          Cost effectiveness

F.I.3  Scenario 3
       $6,553
  1.1 tons/yr
   $6,082/ton
     F.I.3.1  Basis.  An existing facility that is described by
Scenario 3 has two coating lines that could be classified as
medium based on the total quantity of solvent used per line.  The
facility does not currently operate a control device.  Therefore,
emission points requiring control under RA II of the NESHAP
include solvent storage tanks, mix preparation, coating
application and drying, particulate transfer, and equipment leaks
in the mix room.  The facility does not use HAP solvents for
cleaning; therefore, none of the costs for controlling cleaning
emissions are incurred.  The sample calculation includes
calculations of the total capital investment, total annual cost,
and cost effectiveness.
     F.I.3.2  Parameters.

      1.  Coating lines classification

      2.  Baseline HAP emissions from:
            Mix room (includes 40 percent control
              through use of covers on mix
              equipment):
            Solvent storage
            Coating application/drying
            Equipment leaks in mix room

          Total baseline solvent HAP emissions

      3.  Solvent HAP emissions with RA II
       Medium


 12.2 tons/yr


   0.1 ton/yr
182.8 tons/yr
   0.7 ton/yr

195.8 tons/yr

 10.3 tons/yr
                               F-7

-------
    Airflow (acfm) to adsorber (2 drying    1,239.6 ft3/min
      ovens, 2 sets of mix vessels)
8

9,
Number of enclosed transfer devices
  needed

Additional airflow (acfm) for each
enclosed transfer device

Additional HAP to adsorber as a result
  of controlling:
  Mix room emissions [equals baseline
    divided by (1 - 0.4)]
  Solvent storage
  Coating application/drying

  Total additional HAP to adsorber

Operating hours  (24 hr/day, 250 d/yr)

HAP emissions controlled by LDAR
  in mix room:
                                               19.8 ft3/min
 20.3 tons/yr

   0.1 ton/yr
182.8 tons/yr

203.2 tons/yr

6,000 hr/yr

    1 ton/yr*
Note;  The amount of HAP's controlled by LDAR and equipment
       modifications in the mix room is actually
       0.7 ton/yr; however, applying the cost effectiveness
       factor to this amount results in an unrealistically
       low control cost.  The minimum cost of LDAR
       corresponds to controlled emissions of l ton/yr.
                         F-a

-------
     10.  Total capital  investment
              TABLE F-5.  TOTAL CAPITAL INVESTMENT--
                   RA II APPLIED TO SCENARIO 3
Emission point
Storage tanks
Mix preparation
(2 lines)
Coating
Application/drying
(1 carbon adsorber)
Equipment leaks
Particulates
(2 lines)
TCI
Capital cost, $
38,710
14,800
271,153
Included in total
annual cost since
overall factor was
used
160,000
$484,663
Source in Chapter 8
Table 8-4
Table 8-5
See Table P-6 for a
detailed breakdown of
how the costs from the
NSPS analysis were
used.
Section 8.5;
Appendix F, Part F.2
Section 8.2.1.5

     TABLE P-6.  TOTAL  CAPITAL INVESTMENT TO CONTROL COATING
          APPLICATION/DRYING EMISSIONS UNDER SCENARIO 3
Component
• Ductwork from oven
to adsorber
• Ductwork at oven
• Ductwork at adsorber
• Ductwork from
enclosure to oven
• Control device plus
ancillary equipment
• Total enclosure
No.
needed
2
2
1
2
1
2
NSPS unit
cost, $a
5,064
5,975
14,150
2,533
106,503
13,500
NESHAP cost - adjusted for
inflation and retrofit
(2)($5,064)(362.2/326.8)(1.4) = $15,715
(2)($5,975)(1. 108)(1.4) = $18,542
(1)($14,105)(1. 108)(1.4) = $21,886
(2)($2,533)(1. 108)(1.4) = $7,860
(1)($106,503)(1. 108)(1.4) - $165,255
(2)($13,500)(1,108)(1.4) = $41,895 1
Total = $271,153
aTotal installed cost, 1983 $.

-------
  11.  Total  annual  costs
TABLE P-7.  TOTAL ANNUAL COST--RA II APPLIED TO SCENARIO  3
Coat factor
Operating labor
Shift* per day
Hour* per shift
(Operating schedule is 24 hr/day, 5 d/wk, 50 wk/yr)
Labor coat ® $11 .38/hr
Supervisor (15% of labor coat)
Utilities:
Hectricity: (1,240 ft3/min)(5 hp/1,000 ft3/min)(0.746 kW/hp)
(6,000 hr/yr)($0.049/kWh)
Steam: (4 Ib steam/lb HAP)(203 tona x 2,000 Ib HAP/ton)
($6/1, 000 Ib steam)
Water: (3.43 gal HgO/lb steam)(l, 624,000 Ib steam)
($0.936/1, 000 gal H20)
Raw material*
Initial carbon charge
Initial Carbon coat (784 Ib x $2.18/Ib carbon)
Cost of labor ($0.05/Ib x 784 Ib carbon)
Carbon replacement coat: 0.2439(1 .08 x $1,709 + $39)
Maintenance
Labor (1 10 percent of operating labor)
Materials (100 percent of maintenance labor)
Indirect cost*
Administrative (2% of TCI): ($484,663)(0.02)
Property tax (1 % of TCI): ($484,663)(0.01)
Insurance (1 % of TCI): ($484,663)(0.01)
Overhead (60% of operating and maintenance labor and
maintenance materials)
Capital recovery (TCDfCRF 0*7%, n-10 yr*)]:
($484,663)(0.1424)
Solvent disposal
Quantity disposed (4.75 x 203 tona/yr)
Disposal cost ($69/ton x 964 tona)
Equipment leak* (mix room)
($5,343/ton)(l ton controlled)
Total annual cost
Annual coat
3
0.5
$4,268
$640
$1,360
$9,744
$5,214
784
$1,709
$39
$460
$4,695
$4,695
$9,693
$4,847
$4,847
$8479
$69,016
964
$66,533
$5,343
$199,934
Source
Referenced
Reference 9
Referenced
References 1, 2
Reference 3
Reference 3
Reference 10
Reference 6
Referenced
Indirect costs are a percentage of the total capital
cost to control emissions from solvent storage,
mix preparation, coating application/drying,
waste handling, and particulate transfer.
See Chapter 8, Section 8.3.2.1 for explanation.
Calculation ia based on an 89% efficiency for
LDAR and a cost of $5,343/ton controlled for
the mix room. For an explanation of how the
cost effectiveness factor* were calculated, refer
to Section 8.5 of Chapter 8 and Part B of this
appendix.

  12.   Cost effectiveness:

       Total annual cost
         (see Tables P-5,
       Emission reduction
       Cost effectiveness
F-6,  and F-7)
     $199,934

185.5 tons/yr
   $l,078/ton
                           F-10

-------
F.I.4  Scenario 4

     A facility only uses HAP solvents for cleaning.  One of the
existing magnetic tape manufacturing facilities is described by
this scenario.  The facility has one wash sink.  Therefore the
total cost to the facility is:

Total Capital Investment--RA II
Emission point
Removable parts cleaning
Closed containers
TCI
Cost, $
470
400
870
Source in Chapter 8
Section 8.2.1.6
Section 8.2.1.7

     As described in Chapter 8, there are no direct costs
associated with control of removable parts cleaning emissions
through a freeboard ratio or the use of closed containers for all
cleaning activities.  Thus, the total annual cost is simply the
capital recovery cost for the controls mentioned above  (other
indirect costs are negligible).
     Total annual cost ($870) (0.1424)
     Emission reduction
     Cost effectiveness

F.I.5  Scenario 5
     $124
29.6 tons
   $4/ton
     A facility described by Scenario 5 only requires control of
HAP emissions from cleaning activities and from piping leaks.
One of the existing magnetic tape facilities described by this
scenario has three wash sinks requiring control.  Also, the
facility must control equipment leak emissions in the mix room,
solvent recovery, and waste handling areas.  The total capital
investment for the facility is:

Total capital investment--RA II
Emission point
Removable parts cleaning
Closed containers
TCI
Cost, $
1,410
400
1,810
Source in Chapter 8 1
Section 8.2.1.6
Section 8.2.1.7

     The total annual cost to the facility is the cost incurred
from controlling equipment leaks plus the capital recovery charge
associated with the TCI calculated above.  These costs are
summarized below:
                               F-ll

-------
Total annual coat--RA II
Emission point
Mix room
(0.89) ($5,343/ton)
(12.0 tons/yr)
Solvent recovery
(0.89) ($931/ton)
(57.3 tons/yr)
Waste handling
Capital recovery
($1,810) (0.1424)
TOTAL ANNUAL COST
EMISSION REDUCTION
COST EFFECTIVENESS
Cost, $
57,062
47,478
3,137
258
107,935
366 tons/yr
$295/ton
Basis
Assumes 89% control with
the LDAR program and a cost
factor of $5, 343 /ton
controlled.
Assumes 89% control with
the LDAR program and a cost
factor of $931/ton
controlled.
See Part B of this appendix
for calculation method.
See Table F-8 for inputs
(the model line size in
this case is medium) .




        TABLE F-8.  INPUT PARAMETERS FOR EQUIPMENT LEAKS
                       FROM WASTE HANDLING


Solvent reclamation rate,
Mg/yr
Equipment count
Pump seals (light liquid)
Compressors
Flanges
Gas valves
Liquid valves
Gas pressure relief devices
Liquid pressure relief
devices
•
Small
32
5
0
0
34
87
3
0
Model unit size
Medium
160
3
0
0
20
52
2
0

Large
8,000
i_
0
0 j
7
17
1
0

-------
                     TABLE F-8.   (continued)

Sampling connections
Open-ended lines
Total
Model unit size
Small
9
35
173
Medium
5
21
103
Large
2
7
35
F.2  CALCULATION OF COSTS TO CONTROL EQUIPMENT LEAK EMISSIONS

     Emissions from equipment leaks were calculated for those
facilities that provided adequate information on equipment
counts, the percent HAP in the fluid contacting the equipment,
and the time the HAP is in contact with the fitting.  This
section identifies the scope of a leak detection and repair
(LDAR) program to control emission leaks and explains the bases
and calculation methods used to estimate the costs, emission
reductions, and cost effectiveness of controlling equipment leaks
from piping.  The calculations are based on the requirements of
the Negotiated Regulation for equipment leaks that is contained
in the proposed Hazardous Organic NESHAP (HON) .1:L

     The results of the analysis indicated that the average cost
effectiveness of an LDAR program is $5,343/ton equipment leak
emissions controlled in the mix room and $931/ton equipment leak
emissions controlled in the solvent recovery area.  Three plants
provided information used to arrive at the LDAR cost
effectiveness factor for the mix room and two plants provided
information used to calculate the LDAR cost effectiveness factor
for the solvent recovery area.

     For facilities and model lines for which plant-specific
information was not available, the above LDAR cost effectiveness
values were applied to the respective value of tons controlled to
calculate the total annual cost of an LDAR program in each area.

     The Negotiated Regulation program requires control of
equipment leak emissions through a combination of a leak
detection and repair (LDAR)  program and equipment modifications.
Table F-9 lists equipment modifications that were chosen for the
Negotiated Regulation cost analysis to control leaks from open-
ended lines, sample connections, and pressure relief valves.11

-------
               TABLE F-9.  EQUIPMENT MODIFICATIONS
Equipment type
Open-ended lines
sample connections
Pressure relief valves
Method of control
Second valve on open end
Closed- loop system
Rupture disk assembly
     These equipment modifications are assumed to provide
100 percent control of equipment leaks.

     For equipment types that are difficult to modify due to the
nature of their operation or construction, emission reductions
are best accomplished through an LDAR program.  First-year
inspection frequencies spfcified in the Negotiated Regulation are
summarized in Table F-10.
.1
 TABLE  F-10.  NEGOTIATED REGULATION--LDAR  INSPECTION FREQUENCIES
Equipment type
Pumps
Connectors ( flanges , threaded
connections)
Valves (gas and light liquid)
LDAR inspection frequency
Monthly
Annually
Quarterly
     The Negotiated Regulation LDAR program allows for reduced
inspection frequency after the first year if certain criteria are
met.  However, to be conservative in estimating costs, this
analysis assumes that facilities will not meet the criteria for
reduced monitoring frequency.  Because the Negotiated Regulation
requires facilities that do not meet the criteria for valves to
increase monitoring frequency from quarterly to monthly, the cost
of LDAR for valves is based on monthly monitoring, not quarterly
monitoring. - This approach was used to calculate cost impacts for
the HON.12

F.2.1  Cost Estimating Methodology

     The total annual cost of controlling equipment leaks
includes direct and indirect costs, less the savings chat result
from decreased solvent usage.  The calculation of individual cos.t
elements is discussed below.  Costs are dated January 1992.1^~1°

     F.2.1.1  Indirect Costs.  Indirect costs are a function of
the total capital investment (TCI) associated with implementing a
control option.  The TCI includes equipment modification costs
and start-up costs associated with LDAR.  Information necessary
to calculate indirect costs for equipment modifications and
start-up of an LDAR program is presented in Table F-ll.13"16
                               F-14

-------
CO
2
0
H
H
H
Q
i
a.
H
S
Ed
EH
J£J
H
E-"
CO
M— .
CO Q

^
0 H
Wf ...
r*
ft! H
H^9
f*
Q H

H Q
OJ §
0
to
CO
ft!
w
EH
Q,
M4
g
2
H
H
H
to
TABLE
I'K
H .S S
•3 2 2
01 S >-a "^ a
'S81JJ
iilli
2 2 u g u

i
o
U



2



§


(2


00

d




<


•?
iM *3
<§, *
0 §•
ao E
^^ Q«




Z

1

1



*



a


in


S
*M
d




<
Z




Z




Z

*
o

m
u
O



*



»


3


§

d




<
Z




«*»
Z




z

I
5

1 Light liquid
valves



«



§


2


3

d




^
Z




Z




z

1

1 Connectors
(flanges)


•<
Z



Z


Z


r>
^*
d


I
*

d

^
o
b
*
d




S
1
N/A (equipn
modification

1 Open-ended
Unes


•<;
Z



z


z


3

d


1
*

d

^
U
b
*
d




a
3
N/A (equipn
modification

1 Sample
connections


•<
Z



Z


Z


00 ^
(O ^*
d d


I I
* *

d d

s- s-

b b
* *
S 3
d d




00
£ *
!~
3, o
11
< '-|
Z S
X
1 3 >.
s 1 «s
|||!I


4j
z



z


z


n
PJ
d


I
*
3
d

S?
O
b
-o
d



•
3
vo"

Z

|| Monitoring
|| instrument
                          I
                          o
                          S
                             .g
                             <
F-1S

-------
     F. 2. 1.1.1  Equipment modifications.  The TCI is calculated
by multiplying the unit capital cost of modifying each type of
equipment by the equipment count.12  The capital recovery cost
(CRC) associated with each type of equipment modification is
determined by multiplying the TCI by a capital recovery factor
(CRF) ,  which is a function of the useful life of the equipment
and the interest rate.


               Equipment) /    Unit    \ /Capital recovery
                 count  / \capital cost/ \     factor


     Pressure relief devices.  Rupture disks in pressure relief
devices are assumed to have a 2 year life.  The corresponding
capital recovery factor at an interest rate of 10 percent is
0.58.  Each rupture disk costs $79. 13

          CRC - (Cprv) ($79/disk) (0.58)

     where Cprv is the equipment count for pressure relief
devices .

     All other costs incurred when installing a rupture disk
assembly are annualized over a 10 year period (CRF = 0.163) .

          CRC - (C.) ($3,960 - $79) (0.163)
     Open-ended lines.  The TCI associated with equipment
modifications to open-ended lines is the cost to install a second
valve to prevent emissions through the end of the line.  Valves
are assumed to have a 10 year life (CRF - 0.163) .

          CRC = (Coel) (102) (0.163)

     where CQej_ is the equipment counts for open-ended lines.

     Sample connections.  The TCI associated with equipment
modifications to sample connections includes the cost to install
a closed- loop purge system, consisting of 6 meters of piping and
three 2.5 centimeter ball valves.13  The installed equipment is
assumed to have a 10 year life.12

          CRC » (C3C) ($413/sample connection) (0.163)

     Additional indirect equipment modification costs.
Additional indirect costs associated with equipment modifications
include maintenance and miscellaneous costs.   Maintenance costs
are estimated as 5 percent of the capital cost of equipment
modifications.  Miscellaneous costs (taxes, insurance, and
administrative expenses) are estimated as 4 percent of the
capital cost of equipment modifications.12
                               F-16

-------
     F.2.1.1.2  Initial LDAR Costs.  The remaining equipment
types are controlled by LDAR.  The TCI associated with LDAR
includes the cost of monitoring equipment and start-up costs of
implementing the program.12  Examples of start-up costs include
locating all components, numbering and tagging equipment, and  the
initial monitoring and repair of a high number of leaking
components  (which decreases after the first year).  Monitoring
costs for the initial year of an LDAR program are higher to
account for these additional tasks.  Estimates of initial and
subsequent monitoring costs were taken from the cost analysis
done for the Negotiated Regulation, which estimated a monitoring
cost of $2.50 per component screened for the initial LDAR survey
and $2.00 per component screened for subsequent monitorings.16
For pressure relief devices, the cost of replacement rupture
disks required during the first year is included in the initial
cost of LDAR.  The cost of replacement pump seals needed during
the first year is included in the initial cost of LDAR.12

     Monitoring instrument.  It is assumed that one instrument is
needed per plant.  The cost of one monitoring instrument is
$6,500.  Monitoring instruments are assumed to have a 6 year
life.  The capital recovery factor at a 10 percent interest rate
is 0.23.12

          CRC - (1) ($6,500) (0.23)

     Additional indirect costs associated with the monitoring
instrument are the cost of maintaining and calibrating the
monitoring equipment, and miscellaneous monitoring equipment
costs.  As shown in Table F-5, maintenance and calibration of
monitoring equipment is estimated as 65 percent of the instrument
cost; miscellaneous expenses are 4 percent of the instrument
cost.12

     Monitoring and repair costs.  Initial monitoring costs
(annualized) are calculated by multiplying the initial cost to
monitor one component ($2.50) by the equipment count for each
type of equipment and the appropriate capital recovery factor:12


          _ /Equipment] [initial monitoring] /Capital recovery
          ~\  count  /[    fee  ($2.50)    J\     factor


     Repair costs are for repairs necessary if the first attempt
at repair by the monitoring team (made at the time that a leak is
detected) is unsuccessful.  Repair costs are estimated by
multiplying the following parameters for each equipment type:12

     1.  The number of leaks detected by the monitoring team;

     2.  The percentage of leaking components that could not be
repaired on the first attempt;

-------
     3.  The time required to repair the equipment piece; and

     4.  The hourly labor charge.

     The number of leaks detected in the initial LDAR survey for
each equipment type is estimated by multiplying the number of
components monitored by the leak frequency. • Thus, calculation  of
repair costs (annualized) is represented by the following
equation:12'16

 n?r    .   = /Equipment \ /    Leak     \ /% Requiring\/Repair\/Labor
 CRCrepair = \  count  / \ frequency (%) / \ rRepair  / \ time / \ rate
     The labor rate for repair is $22.50 per hour.  This cost  is
in January 1992 dollars.

     Calculations of monitoring and repair costs for each
equipment type in the initial LDAR program are illustrated below.
Monitoring and repair costs are multiplied by the appropriate
capital recovery factor.  Costs are also multiplied by a factor
of 1.4 to account for the cost of administration and support of
the LDAR program (record keeping, reporting, etc.), which is
estimated to be 40 percent of monitoring and repair costs.12

     Gas valves.

     Monitoring :

               CRC  -  (C   ) ($2.50/valve) (1.4) (0.163)

          where C—, is the equipment count for gas valves.

     Repair:

          CRC - (C_J (0.075) (0.25) (4 hr/repair)
                ($22.50/hr) (1.4) (0.1S3)

     Liquid ' valves .

     Monitoring:

          CRC » (Clv) ($2.50/valve) (1.4) (0.163)

          where C^v is the equipment count for liquid valves.
     Repair:
          CRC =  (Clv) (0.043) (0.25) (4 hours/repair)
                 ($22.50/hour) (1.4) (0.163)
                               p-ia

-------
     Pumps.  In addition to monitoring and repair costs,  the  cost
of replacement seals for pumps is included in the TCI  associated
with LDAR start-up costs.  Each seal costs $180, which includes  a
50 percent credit for the old seal. °  The 40 percent  increase
for administrative and support costs is not applied  to the  cost
of replacement seals because it is not a labor cost.   Also,
replacement seals are assumed to have a 2-year life.
corresponding to a capital recovery factor of 0.58.12'16

     Monitoring:

          CRC - (Cp) ($2.50/pump) (1.4) (0.163)

          where Cp is the equipment count for pumps.

     Repair:

          CRC = (Cn)(0.075)(1)(16 hours/repair)
                ($22.50/hour)(1.4)(0.163)

     Replacement seals:

          CRC - (Cp)(0.075)(1)($180/seal)(0.58)

     Connectors.

     Monitoring:

               CRC - (Ccon) ($2.50/con)  (1.4) (0.163)

          where Ccon is the equipment count for connectors.

     Repair:

          CRC - (Ccon)(0.039)(1)(2 hours)($22.50/hour)
               • (1.4) (0.163)

     F.2.1.2  Direct Costs.  Direct costs associated with
controlling equipment leak emissions include repair  costs for
components under the LDAR program.  These costs are  discussed
below.  The input parameters for the cogt equations  explained
below are summarized in Table F-12.12"16  Annual costs of
monitoring and repair are calculated in the same way as for
initial LDAR costs, with two exceptions:   (1) a capital recovery
factor is not applied;  and   (2)  there is an additional factor to
account for the frequency with which each equipment  type  is
monitored  (monitoring frequency).

     F.2.1.2.1  Monitoring and repair cost.  The cost  of
monitoring a piece of equipment after the initial monitoring  is
$2.00 per event.16  The calculation of annual monitoring  and
repair expenses is summarized in equation form below:

-------


63
H
EH
03
H
EH
03
0
O
B
H
Q
0*
O
fe
03
1

04
M

CN
H
i
&4

S
CO
H









S











1*
w
O
te
"T5
2
-
S
.£•
1





to
II
00
3 3
I
g >,^

S" § S
JJU

00
.S *>
2 §
I!
2
01
j
s

en
O
§
O
i
C3
«
'3
3


1
a
a


i




0



t
a


1
oo
6

o ^
Z =*

* 1
t J 1
&> "5
o 9"3
00 5 a.
S a. 2


at
I"



«




N


*
2



Z







<
Z


3
"3
>



,,




N


*
g



Z







<
Z
1
J
•a
1
I
3


S




o



1



z







<
z


2
2
|
3
Z

1




Z



z



z







<
z
s
_B
i
•o
g
g
0
Z

Z




1



1



z







<:
z
|
S
1
jj
f—

1

g




1



1



z







.<
z

•s -a
1 |^
1 i a* § 1
tfl'ffffl
z

z




z



z



z







<
z


3"a
•C 3
2 1

ll
                     I
                     8
                        _Q.


                         I
                     •c  £
                     I  I
                     I  I
                     _

                     1
                     1
    O.

s-   I
•g   HJ

I   I
I   i
_   «
                         3
                         ar
                         o
                         13
                      .
                     S-J-2
                     -Sis
                     1 *! 3
                     a ^ <&
                     3 <^ «
                     .  **
                     1 2 3
                     !!•*
F-20

-------
      Annual      _ /Equipment] [subsequent monitoring] /Monitoring
  monitoring cost ~\  count  /[     fee ($2.00)      J \ frequency ,

     Annual repair  costs are a function of equipment counts,
percent requiring repair,  leak frequency, monitoring frequency,
time required  for 'repair,  and labor rates.  Different regulations
specify expected leak frequencies after the program is
implemented  ("subsequent"  leak frequencies).  Subsequent  leak
frequencies specified in the HON Negotiated Regulation are
presented in Table  F-12.  Monitoring frequencies are those-
specified in the Negotiated Regulation.12* 16  Monthly monitoring
of valves is based  on the  assumption that facilities will not
meet the requirements of the Negotiated Regulation.  The
calculation of annual repair expenses is summarized below:
/Equipment     se    % requiring\ /tnonitoring\ /Repair\ fLabor\
I  count
                              % requiring\ /tnonitoring\ /Repair\ fLabor\
                                rePair   /\ frequency j I tune }\ rate j
                  frequncy(%)


     F. 2. 1.2. 2  Gas valves

     Monitoring  (monthly) :

          COST »  (Cgy) (12) ($2.00/valve) (1.4)

          where Cgv is  the equipment count for gas valves.

     Repair  (monthly) :

          COST -  (Cg,.) (0.02) (12) (0.25) (4 hours/repair)
                  ($22. 50/hour) (1.4)

     F. 2. 1.2. 3  Liquid  valves.

     Monitoring  (monthly) :

          COST -  (Clv) (12) ($2.00/valve) (1.4)

          where C^v is  the equipment count for liquid valves.

     Repair :

          Cost -  (Clv) (0.02) (12) (0.25) (4 hours/repair)
                  ($22. 50/hour) (1.4) (0.163)

     F. 2. 1.2. 4  Pumps.

     In addition  to monthly monitoring with a monitoring device,
the Negotiated Regulation  requires weekly visual inspection of
pumps.  Each inspection is estimated to last approximately
30 seconds.1^
                               F-21

-------
     In addition to monitoring and repair costs,  the cost of
replacement seals is included in the direct  cost  of  LDAR for
pumps.  Each seal costs $180, which includes a  50 percent credit
for the old seal.  The 40 percent increase for  administrative and
support costs is not applied to the cost of  replacement seals.
There are. also miscellaneous expenses associated  with replacing
pump seals; these are estimated as 80 percent of  the annual cost
of replacement seals. 2

     Monitoring:

          COST- (CD) (12) ($2.00/pump) (1.4) +
                 (CJ) (52) (30/3600) ($22,50) (1.4)

          where C_ is the equipment count for pumps.

     Repair:

          COST » (CLJ (12)(0.10)(1)(16 hours/repair)
                 ($22.50/hour)(1.4)

     Replacement seals:

          COST » (Cp)(12)(0.10)(1)($180/seal)


     Miscellaneous costs:

          COST - (Cp) (12) (0.10) (1) ($180/seal) (0.80)

     F.2.1.2.5  Connectors.

     Monitoring:

          COST - (Ccon)(1)($2.00/con)(1.4)

          where Ccon is the equipment count  for connectors.

     Repair:

          COST - (Crnn) (0.005) (1) (1) (2 hours) ($22,50/hour)
                 (1?4T

     F.2.1.3  Total Annual, Cost.  The total  annual-cost is the
3um of the indirect costs and direct costs explained above, less
any annual savings resulting from reduced solvent usage.  The
calculation of annual savings and the total  annual cost is
described below.  Total annual costs are calculated for a typical
unit operation at a magnetic tape facility in Attachment A, which
is a printout of the spreadsheet used to calculate equipment leak
control costs using the methods explained in this appendix.
                               F-22

-------
     F.2.1.3.1  Emission reductions.  Uncontrolled emissions from
equipment leaks are calculated using average emission factors  for
the synthetic organic chemical manufacturing industry (SOCMI),
which estimate typical leakage rates for each equipment type.
The factors used in this analysis are listed in Table F-13.16

           TABLE F-13.  SOCMI AVERAGE EMISSION FACTORS
Equipment type
Gas valves
Light liquid valves
Pump seals (light liquid)
Pressure relief devices
Connectors
Open-ended lines
Sampling connections
Average emission factor,
(kg/hr/source)
0.00597
0.00403
0.0199
0.104
0.00183
0.0017
0.0150
     Annual emissions from each equipment type are calculated as
follows:
       Emissions1
Emission!
 factor /
Equipment]
  count  /
Annual hours! /Weight
of operation/ \% HAP
     Equipment counts and the weight percentage of HAP's in the
fluid are specific to each facility and unit operation.  Total
uncontrolled emissions for a particular operation are the sum of
emissions calculated for each equipment type.

     As mentioned previously, the equipment modifications
specified for the Negotiated Regulation are assumed to provide
100 percent control of emissions from the equipment types that
are modified.  The percent emission reductions for equipment
types controlled by the Negotiated Regulation LDAR program are
summarized in Table F-14.15
      TABLE F-14.  EMISSION REDUCTIONS ASSOCIATED WITH LDAR
                  PROGRAM-NEGOTIATED REGULATION
Equipment type
Gas valves
Light liquid valves
Pump seals (light liquid)
Connectors
Percent emission reduction
92
88
69
93
                               F-23

-------
     The annual emission reduction resulting from LDAR  for  each
equipment type is calculated by multiplying annual uncontrolled
emissions by the appropriate percentage:


                Emission _ /Uncontrolled! ( Percent \
               reduction" 1  emissions  / Ireduction/
     The total emission reduction is the sum of the emission
reductions for each equipment type.

     F.2.1.3.2  Total annual cost.  Controlling equipment leaks
results in reduced solvent usage and a subsequent cost savings.
The annual savings equals the cost of the solvent multiplied by
the amount of solvent controlled annually.  The total annual
cost, then, is the sum of indirect and direct costs, less the
annual savings resulting from reduced solvent usage.


   Net annual _ /   Total   ] _ [Emission reduction] /Solvent cost]
      cost    ~ \annual cost/ ~ [    (Megagrams)    J \per megagram/
     The cost of solvent for magnetic tape facilities was
calculated using the representative solvent usage proportions
defined for model lines in Chapter 3 of the background
information document (BID):  40 percent methyl ethyl ketone
(MEK),  40 percent toluene,  and 20 percent methyl isobutyl ketone
(MIBK).   The solvent cost was calculated as follows, using
current solvent prices obtained from a vendor:
                                 Toluene*0.4
                                                 Mg  /
*0.2
     The cost of solvent was calculated as $657 per megagram.

F.2.2  Cost Effectiveness

     Cost effectiveness is calculated by dividing the net annual
cost of controlling equipment leaks  (in dollars) by the annual
emission reduction (in megagrams).

                                   [Net annual
                    Cost     _     [ cost  ($)
               effectiveness ~ 7Z :   :     ~—  .
                               lEmission reduction
                                   (Megagrams)
                               P-24

-------
     Sample Calculation

     Attachment A shows the calculation of direct and indirect
costs, uncontrolled and controlled equipment leak emissions,
solvent credit, total annual cost, and cost effectiveness for a
medium-sized waste handling unit within a magnetic tape facility.
The relevant parameters for this example are stated below.

                         Equipment counts

          Pump seals:  '       *                                  3
          Connectors: '                                          0
          Liquid valves:'                                      52
          Gas valves:                                          20
          Sample connections:                                   5
          Liquid pressure^ relief devices:                       0
          Gas pressure relief devices:                          2
          Open-ended lines:                                    21

     Operating hours per year (all equipment types):        2,080
     Weight percent HAP's in stream:                           70

F.3  REFERENCES FOR APPENDIX F

 1.  Neveril, R.B., GARD, Inc.  Capital and Operating Costs of
     Selected Air Pollution Control Systems.  U. S. Environmental
     Protection Agency.  Research Triangle Park, NC.  Publication
     No. EPA-450/5-80-002.  December 1978.

 2.  U. S. Department of Energy.  Monthly Energy Review.
     February 1992.  p. 116.

 3.  U. S. Environmental Protection Agency.  Office of Air
     Quality Planning and Standards (OAQPS) Cost Manual.
     Publication No. EPA-450/3-90-006.  January 1990.
     pp. 4-28 - 4-29.

 4.  Memorandum and attachments from Beall, C., and J. Glanville,
     MRI, to Johnson, W., EPA/CPB.  March 15, 1985.  Revised
     final tabular costs.

 5.  Reference 3.  p. 2-29.

 6.  U. S. Environmental Protection Agency.  Office of Air
     Quality Planning and Standards (OAQPS} Cost Manual.
     Publication No. EPA-450/3-90-006.  January" 1990.
     pp. 4-33 - 4-35.

 7.  Reference 3.  pp. 3-54 - 3-55.

 8.  Reference 2.  p. 119.
                               F-25

-------
 9.  U. S. Department of Labor.  Monthly Labor Review.
     April 1992.  p. 69.

10.  Memorandum from McManus, S., MRI, to Project File.
     February 1, 1993.  Documentation of methods used to
     determine costs of control options and regulatory
     alternatives for the magnetic tape manufacturing industry.

11.  U. S. Environmental Protection Agency.  Equipment Leaks
     Negotiated Regulation.  Washington, DC.  U.S. Government
     Printing Office '(Federal Register Notice).  March 6, 1991.
     (9,2)44:9315.

12.  Memorandum from Whitt, D., Radian, to Markwordt, D.,
     EPA/CPB.  June 5, 1991.  Impacts from the control of VHAP
     emissions from equipment leaks in non-SOCMI process units
     for HON.

13.  Fugitive Emission Sources of Organic Compounds-- Additional
     Information on Emissions, Emission Reductions, .and Costs
     (Section 2).  EPA-450/3-82-010.  U.S. Environmental
     Protection Agency, Research Triangle Park, North Carolina.
     April 1982.

14.  Economic Indicators.  Chemical Engineering.  99(4);210.
     April 1992.

15.  Richardson Engineering Services, Inc.  Richardson Process
     Plant Construction Estimation Standards.  Mechanical and
     Electrical.  Volume 3.  Mesa, Arizona.  1988.
     Sections 15-0, 15-42, 15-43, and 15-55.

16.  Memorandum from Hausle, K. J., and D. J. Whitt, Radian
     Corporation, to Markwordt, D., EPA/CPB.  February 28, 1992.
     Final cost impacts analysis for HON equipment leaks.
                               F-26

-------
ATTACHMENT A

-------

-------
   Direct COME
   Equipment modificaooai:
     PRO (S):                                                   $394.00
     Open-ended Una (S):                                         $107.10
     Sample comMcooM (S):                                       $103.25

   LDAR:
     Monitoring imtrumeoB ($):                                     JO.OO
     Pump scab (S):                                              $64100

  TOTAL annual maintenance charge* ($):                         $1,254.35

  Annual auKcUaoeoitt charge*

  Equipment modification:
    PRD <$):                                                   $316.10
    Open-ended lines (S):                                          S*S.«I
    Sample coanccaoH (S):                                       K2.«0


  LOAR:
    Monitoiinf iattnuneaai (S):                                        0
    Pump leak (S):                                              S51140

  TOTAL annual auKeUaneow caarfo:                          Jl,003.«

       ai Labor charge*
  CaivalraK
    -Momtorimj($):                                       *     J47ZOO
    •Repair (S):                                                 S15UO
  Liquid valve* (S):
    •M6ailona((S):                                           JL7<7.20
    •Repair ($):                                                 S393.12
  Pumpa($):
    -Moniloraj($):                                             S141.7S
    •Repair (S):                                               SUU40
  Conoecton:
    •Monitoring (S):                                              JO.OO
    •Repair (S):                                                  JO.OO

 TOTAL annual Uborcaarta(S):                               $4,919.67


          SUM OF DIRECT AND INDIRECT COSTS:            S9.33S.S9



 Avenge Emauoa Faeton (unooanotted):

 Pump teak (kg/hr/iourcc):                                        0.0199
 Vaivec
 •liquid (ktYBrtource):                                           0.00403
 •pa(k|/Bi/iouic«):                                            0.005*7
 Couecton (kt/nr/Mutce):                                       0.001S3
 Opta rnded line*  (tt/mViouree):                                   a0017
 Sample CoaMctJow (ktynBtooree):                                0.015
 Pmran relief deviota (ktybBtaMM):                               0.104
Preauire relief devKwdtg^T,:
Re|Ne|camrollevck

Oaivalvw
U«nt liquid varva

-------
Pump ttafcditftt liquid)                                             «9*
Conneeton                        .                               93%
Opta^KkdlfaM                                                 100%
Sample couecaoas                                               100%
Pressure relief devices                                             100%
Eminioa reductions resulting from R«t Neg control

Pump seals (kiyyr):                                               59.9»
Valves:
  -liquid (kgfrr):                                                 268,51
  •gas(k«yyr):                                                   159.94
Coonecton (k«/yr):                                                0.00
Ope»«ndeduiie*(ktyyr):                                          S1.9«
Sample Connecooni (k»yjrr):                                       109.20
Proture relief devices (kg/yr):                                .      302.SS

Toul(ktyyr)                                                     95Z4
   (Mg/yr)                                                      0.952

Solvent cost (J/Mg):                                              M57
Solvent credit (J/yr):                                              S626

TOTAL ANNUAL COST:                                       M.703

Co»t-«{fectiv«ne» (S/Mg):                                        S9.13S

-------
                    APPENDIX G.




ANNUAL OUTPUT OF COATED SUBSTRATE FOR A MODEL LINE

-------
                            APPENDIX G

     The output of coated substrate per year for a model line is
calculated based on the parameters reported in Table 6-5.  The
first step in determining model line output is to calculate the
amount of substrate coated per second.  This is done by
multiplying the substrate width by the model line speed.  The
resultant value is then used to determine the total amount of
substrate coated per hour.  Once the per-hour value is known,
total yearly output can be calculated by multiplying the model
line's operating hours per year by the total amount of substrate
it is capable of coating in an hour.  Total yearly output of
coated substrate is measured in meters squared per year.
     On the following.page, the output calculations for a small
model line are represented.
Output of Coated Substrate From a Small Model Line:

Data from Table 6-5:
     substrate width          0.15 m
     line speed               1.3  m/sec
     operating hours         2,000  hr/yr

Calculations:
     Substrate Coated Per Second
        = 1.3 m/sec x 0.15 m » 0.195 m2/sec

     Substrate Coated Per Hour
        = (0.195 m2/sec x 60 sec/min) (60 min/hr)
        = 702 m2/hr
                               G-l

-------
     Substrate Coated Per Year
        - (702 m2/hr) (2,000 hr/yr)
        - 1,404,000 m2/yr

Notes:    m - meter
        sec - second
         hr - hour
         yr - year
                               G-2

-------
                                    TECHNICAL REPORT DATA
                            (Please read tnstructions on the reverse before completing)
 REPORT NO.
    EPA-453/R-93-059
                                                             3. RECIPIENT'S ACCESSION NO.
 TITLE AND SUBTITLE
    Hazardous Air  Pollutants from  Magnetic Tape
    Manufacturing  — Background  Information for
    Proposed Standards
             5. REPORT DATE
               February 1994
             6. PERFORMING ORGANIZATION CODE
 AUTHOR(S)
             8. PERFORMING ORGANIZATION REPORT NO.
. PERFORMING ORGANIZATION NAME AND ADDRESS
  Office of Air Quality Planning and Standards
  U.S.  Environmental  Protection Agency
  Research Triangle Park,  North Carolina  27711
                                                              10. PROGRAM ELEMENT NO.
              11. CONTRACT/GRANT NO.
                                                               68-D1-0115
12. SPONSORING AGENCY NAME AND ADDRESS
  Director,  Office of  Air Quality Planning  and Standards
  Office of  Air and Radiation
  U.S.  Environmental Protection Agency
  Research Triangle Park. North Carolina  27711	
              13. TYPE OF REPORT AND PERIOD COVERED
              14. SPONSORING AGENCY CODE

               EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
  A draft rule for  the  regulation of hazardous air pollutant emission
  from magnetic tape manufacturing operations is being  proposed
  under the authority of  sections 112,  114,  116 and 301  of the
  Clean Air Act, as amended in 1990.  This  document presents back-
  ground information and  the results of  the  national impacts
  assessment for the proposed rule.
                                 KEY WORDS AND DOCUMENT ANALYSIS
                   DESCRIPTORS
                                                b.lDENTIFIERS/OPEN ENDED TERMS
                            c. COSATI Field/Group
  Air  pollution
  Hazardous Air Pollutants
  Magnetic Tape Manufacturing
  National emission standards
  Air Pollution Control
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (ThisReport)
  Unclassified
21. NO. OF PAGES
     414
                                                20. SECURITY CLASS (This page!
                                                  Unclassified
                                                                            22. PRICE
EPA Form 2220-1 (R.x. 4-77)
                       PREVIOUS EDITION IS OBSOLETE

-------
                                                       INSTRUCTIONS

  1.   REPORT NUMBER
      Insert the EPA report number as it appears on the cover of the publication.

  2.   LEAVE BLANK

  3.   RECIPIENTS ACCESSION NUMBER
      Reserved for use by each report recipient.

  4.   TITLE AND SUBTITLE
      Title should indicate  clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, it'used, in smaller
      type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
      number and include subtitle for the specific title.

  5.   REPORT DATE
      Each report shall cany a date indicating at least month and year.  Indicate the basis on which it was selected (e.g.. Jalc ofisutv. Jan- <>/
      approval, date of preparation, etc.).

  6.   PERFORMING ORGANIZATION CODE
      Leave blank.

  7.   AUTHOR(S)
      Give name(s) in conventional order (Jolm R. Doe, J. Robert  Dm: etc.).  List author's al filiation if it differs I mm the performing organi-
      zation.

  8.   PERFORMING ORGANIZATION REPORT NUMBER
      Insert if performing organization wishes to assign this number.

  9.   PERFORMING ORGANIZATION NAME AND ADDRESS
      Give name, street, city, state, and  ZIP code.  List no more than two levels of an organisational hircarchy.

  10. PROGRAM ELEMENT NUMBER
      Use the program element number  under which the report was prepared. Subordinate numbers may be included m parentheses.

  11. CONTRACT/GRANT NUMBER
      Insert contract or grant number under which report was prepared.

  12. SPONSORING AGENCY NAME AND ADDRESS
      Include ZIP code.                             -   '

  13. TYPE OF REPORT AND PERIOD COVERED
      Indicate interim final, etc., and if  applicable, dates covered.

  14. SPONSORING AGkNCY CODE
      Insert appropriate code.

  15. SUPPLEMENTARY  NOTES
      Enter information not included elsewhere but useful, such as:  Prepared in cooperation with, I ranslalum <>l. Presented at cnnlcrcmc nl.
      To be published in. Supersedes, Supplements, etc.

  16. ABSTRACT
      Include a brief (200 words or less/ factual summary of the most significant information contained in the report.  It tin- report ioniums j
      significant bibliography or literature survey, mention it here.

  17. KEY WORDS AND DOCUMENT ANALYSIS
      (a) DESCRIPTORS - Select from  the Thesaurus of Engineering and Scientific Terms the proper authon/cd terms that identity  the major
      concept of the research and are sufficiently specific and precise to be used as index entries lor cataloging.

      (b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc.  Use open-
      ended terms written  m descriptor form for those subjects for which no descriptor exists.

      (c) COSATI I-'IELD GROUP • Held and group assignments are to be taken from the 1965 COSA11 Subject Category List. Since the ma-
      jority of documents are multidisciplinary in nature, the Primary I ield/Group assignment!^ will be specific discipline, area of human
      endeavor, or type of physical object. The application)s) will be crosi-rcfcrenccd with secondary I leld/Croup assignments  that will lolln*
      the primary postingls).

  18.  DISTRIBUTION STATEMENT
       Denote releasability  to the public or limitation for reasons other  than security for example "Release Unlimited."  ( ile anv availability i<>
       the public, with address and price.

  19. & 20.  SECURITY CLASSIFICATION
       DO NOT submit classified reports to the National Technical Information service.

  21.  NUMBER OF PAGES
       Insert the total number of pages,  including this one and unnumbered pages, but exclude distribution list, il any.

  22.  PRICE
       Insert the price set by the National Technical Information Service or the Government Printing Office, il known.
EPA Form 2220-1 (R«». 4-77) (R.v.rte)

-------