EPA-453-R-02-010
United States
Environmental
Protection Agency
Office of Air Quality
Planning and Standards
RTP, NC 27711
June 2002
SEPA Technical Support
Document:
Printing, Coating, and
Dyeing of Fabrics and
Other Textiles
Proposed NESHAP
-------
EPA-453/R-02-010
June 2002
Technical Support Document:
Printing, Coating, and Dyeing of Fabrics and Other Textiles
Proposed NESHAP
U.S. Environmental Protectian Afmey
J«fion 5, Library (PL-12J)
12.West -tekson Boulevtrd. 12th
Cftica8o.»L 60604-3590
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Emission Standards Division
Research Triangle Park, North Carolina 27711
-------
DISCLAIMER
This report has been reviewed by the Emission Standards Division of the Office of Air Quality
Planning and Standards of the United States Environmental Protection Agency and approved for
publication. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
Table of Contents
Memorandum Subject Page
Coating and Printing Floor 1-1
Summary 1-1
Background 1-1
Approach to Estimating the MACT Floor 1-3
Data Collection for the MACT Floor 1-3
Results of Data Collection and the Coating MACT Database 1-6
Criterion for Evaluating HAP Emissions Reductions from Coating Operations 1-8
Consideration of Data Quality in Evaluating HAP Emissions Reductions from
Coating HAP Sources 1-8
MACT Floor Determination 1-11
References 1-18
MACT Floor for Dyeing and Finishing Compounds 2-1
Summary 2-1
Background 2-1
Approach to Estimating the MACT Floor 2-7
Data Collection for the MACT Floor 2-7
Results of Data Collection and the Dyeing and Finishing MACT Database 2-9
Criterion for Evaluating HAP Emission Reductions from Dyeing and
Finishing Operations 2-9
MACT Floor Determination 2-10
References 2-14
MACT Floor for Slashing 3-1
Coating Model Plants 4-1
References 4-9
Summary of Printing, Coating and Dyeing of Fabrics and Other Textiles NESHAP
Baseline Organic HAP Emissions and Emission Reductions 5-1
Coating and Printing Baseline Organic HAP Emissions and Emission
Reductions 5-1
Dyeing Baseline Organic HAP Emissions and Emission Reduction 5-2
Finishing Baseline Organic HAP Emissions and Emission Reduction 5-2
Slashing Baseline Organic HAP Emissions and Emission Reduction 5-3
References 5-3
Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP Nationwide Energy and
111
-------
Table of Contents (continued)
Secondary Environmental Impacts 6-1
Energy Impacts 6-1
Water Impacts 6-3
Solid Waste Impacts 6-3
Compliance Costs for Coating Model Plants 7-1
Permanent Total Enclosure Costs 7-1
Oxidizer Costs 7-4
Carbon Adsorber Costs 7-7
Methylene Chloride Control Costs 7-10
References 7-12
Incremental Cost of Non-Formaldehyde Permanent Press Finish Versus Permanent Press
Finish With Formaldehyde 8-1
Summary of Evaluation of Estimated Compliance Costs Incurred by Coating
Facilities Owned by Small Businesses 9-1
Identify Major Facilities with Coating Operations Owned by Small
Businesses 9-1
Collect Information Needed to Estimate Compliance Costs 9-1
Estimate Compliance Costs 9-8
References 9-9
Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP Nationwide Costs . . 10-1
Coating and Printing Control Costs 10-1
Dyeing and Finishing Compliance Costs 10-5
Monitoring, Reporting, and Recordkeeping Costs 10-6
Nationwide Compliance Costs of the Printing, Coating and Dyeing of Fabrics and Other
Textiles NESHAP 10-6
References 10-7
IV
-------
Table of Contents (continued)
Charts
Chart 1-1. Nationwide Coating Industry Emissions by HAP 1-7
Chart 4-1. Facility-Wide Lbs. Coating Solids Used per MACT Database Facility 4-2
Tables
Table 1-1. Coating Average Facility OCE 1-12
Table 1-2. Coating Facility Average Emission Rate 1-16
Table 2-1. Major Dye Classes and Substrate Fibers 2-3
Table 2-2. Dyeing MACT Floor 2-12
Table 2-3. Finishing MACT Floor 2-13
Table 4-1. Model Plant Parameters for Model Plant No. 1 4-4
Table 4-2. Model Plant Parameters for Model Plant No. 2 4-5
Table 4-3. Model Plant Parameters for Model Plant No. 3 4-6
Table 4-4. Model Plant Parameters for Model Plant No. 4 4-7
Table 5-1. Summary of Printing, Coating, and Dyeing of Fabrics and Other Textiles
Source Category Baseline Organic HAP Emissions and Emission Reductions ... 5-4
Table 6-1. Summary of Coating and Printing Subcategory Model and Nationwide Energy
Impacts 6-2
Table 7-1. Model Plant Specifications Used for Compliance Costing 7-2
Table 7-2. Summary of Coating Room Costs 7-3
Table 7-3. Summary of New Oxidizer Costs for Coating Model Plants 7-5
Table 7-4. Summary of Catalytic Oxidizer Upgrade Costs for Coating Model Plants 7-6
Table 7-5. Summary of New Carbon Adsorber Costs for Coating Model Plants 7-8
Table 7-6. Summary of Carbon Adsorber Upgrade Costs for Coating Model Plants 7-8
Table 7-7. Summary of New Oxidizer Costs for Control of Methylene Chloride Emissions 7-11
Table 7-8. Summary of New Carbon Adsorber Costs for Control of Methylene Chloride
Emissions 7-11
Table 9-1. Coating Facilities Owned by Small Businesses 9-2
Table 9-2. Estimated Compliance Costs for Coating Facilities Owned by Small
Businesses 9-10
Table 10-1. Summary of Coating and Printing Subcategory Model and Nationwide
Control Costs 10-3
Table 10-2. Summary of Printing, Coating, and Dyeing of Fabrics and Other Textiles
NESHAP Compliance Costs 10-7
-------
MEMORANDUM
June 12, 2002
From: G. V. Hellwig
To: Printing, Coating and Dyeing of Fabrics and Other Textiles File
Subject: COATING AND PRINTING FLOOR
SUMMARY
This memorandum describes the methodology and conclusions of the maximum achievable
control technology (MACT) floor analysis for the Coating and Printing subcategory of the
Printing, Coating, and Dyeing of Fabrics and other Textiles NESHAP. The analysis is based on
overall control efficiency (OCE) data from coating lines at 22 major or synthetic minor fabric
coating facilities that were obtained from survey data. The MACT floor for existing sources was
determined to be a 97 percent facility-wide coating line application and curing OCE for
hazardous air pollutants (HAP), which is achievable with add-on control technology. The
MACT floor for new sources was determined to be a 98 percent facility-wide coating line
application and curing OCE for hazardous air pollutants (HAP), which is achievable with add-on
control technology.
BACKGROUND12
The Printing, Coating, and Dyeing of Fabrics industry was identified as a source category of HAP
under section 112(c) of the Clean Air Act, as amended in 1990 (the Act), to be regulated by a
National Emission Standard for HAP (NESHAP) under section 112(d) of the Act. Section
112(d) of the Act directs the EPA to develop standards that require the maximum degree of
reduction in emissions of HAP that is achievable, which are commonly referred to as MACT
standards. For existing major sources, the Act requires MACT to be no less stringent than the
average emission limitation achieved by the best performing 12 percent of existing sources
among the data available to the Administrator. For new major sources, the Act requires MACT
to be no less stringent than the emission control that is achieved in practice by the best controlled
similar source. These minimum stringency levels are often referred to as the "MACT floor."
Coating and Printing was determined to be a subcategory of Printing, Coating and Dyeing of
Fabrics. The manufacturing processes, HAP emissions, and types of controls in use set it apart
from the other processes that are used in the manufacture of textile products. Coating is a web
coating operation, and the physical operations and most facilities performing coating are separate
and distinct from the other textile operations. Printing is a web process very similar to coating
and uses some of the same equipment. This memo is to explain the basis for the MACT Floor for
this subcategory.
1-1
-------
Coating is a specialized chemical finishing technique designed to produce textiles to meet high
performance requirements, e.g., for end products such as tents, roofing, soft baggage, marine
fabric, drapery linings, flexible hoses, hot-air balloons, and awnings. Coatings generally impart
elasticity to substrates, as well as resistance to one or more elements such as abrasion, water,
chemicals, heat, fire, and oil. The substrate itself provides strength (such as tear strength) and can
include wovens, nonwovens, knits, yarn, cord, and thread, although woven fabrics are most
commonly used.
Printing is the application of color to a fabric in a design or pattern. In some cases the printing
material is chemically the same as coating material only thinned to a lower viscosity. There are
typically four types of printing used for mass production, rotary screen, engraved roller, flat-bed
screen, and heat transfer. Rotary screen and engraved roller closely resemble coating and use
principally the same type of equipment as fabric coating. Flat-bed screen is typically not a high
production technique and does not emit large quantities of HAPs over a period of time given the
limits of production. Heat transfer emits little or no HAPs in the transfer of the print to the
fabric.
Both the substrates coated and printed as well as the coating itself vary. A number of different
textile substrates can be coated including rayon, nylon, polyester, cotton, and blends. Coating
chemicals used vary depending on end use of the coated fabric. Examples of coating chemicals
include vinyl, urethane, silicone, and styrene-butadiene rubber. The polymer can be bought in
various forms such as chunks, blocks, chips pellets or fine powder. However, beside the polymer
resins, several other chemicals can also be included in the prepared coating. These include
plasticizers to increase pliability (e.g., fatty acids, alcohols), solvents to disperse solids and adjust
viscosity (e.g., toluene, xylene, dimethyl formamide, and MEK), pigments, curing agents, and
fillers (e.g., carbon black and teflon). Rubber coating materials are frequently compounded in
the facility performing the coating. Manmade fibers coated with epoxy or phenolic resins are
often not immediately cured following application, but are first laid in a mold and then cured
under pressure to form a composite structure.
The coating or printing process generally comprises the following unit operations: mixing the
coating materials (including the solvents), conditioning the substrate, applying the coating to the
substrate, evaporating the solvent in a drying oven, and sometimes curing or vulcanizing. The
application and drying processes and emission controls used by facilities in the industry are
similar and therefore lend themselves well to grouping into a subcategory. The application
processes are similar in that they use continuous web coating techniques, but they include
several types of coating and substrates. The coating industry treats coating as a surface applied
coating in which a distinct layer of coating is applied to the substrate surface. Therefore, the
mass of solids applied is a measure of coated or printed production. This leads to a production-
weighted mass limit for HAP emissions, i.e., mass of HAP per mass of solids applied.
The MACT database for this subcategory consisted of a sample of seventeen facilities that EPA
had complete non-CBI emissions and control information from responses to survey
questionnaires. Although the MACT database contained information from 22 facilities, only
1-2
-------
seventeen of these are presented in this memo in order to maintain the confidential business
information request. The coating and printing subcategory consists of more than thirty operating
facilities; therefore, the MACT floor is based upon the best performing 12 percent of existing
sources among the available data, in this case 3 facilities. The control option for all of the floor
facilities in the coating and printing subcategory is capture and control by either thermal
oxidation or carbon adsorption.
Printing is sometimes performed at the same facilities as other textile wet finishing operations
such as dyeing, finishing, and coating. Printing was not a major contributor of HAPs in the
surveys and plant visits EPA conducted. In the past this was a major source of HAP emissions
and operations can emit large quantities of HAP if the formulations change from low HAP
materials. The EPA has information on only one major source of HAP emissions from printing.
The processes, application, and drying of printing are identical or nearly identical to coating, and
therefore the control options and limits would be identical as well. For this reason printing
operations were included in the fabric coating and printing subcategory. Wherever this memo
discusses coating in the process description or control option, it also applies to printing.
APPROACH TO ESTIMATING THE MACT FLOOR
The term "average," as it pertains to MACT floor determinations for existing sources, described
in section 112(d)(3) of the Act, is not defined in the statute. In a Federal Register notice
published on June 6, 1994 (59 FR 29196), the EPA announced its conclusion that Congress
intended "average" as used in section 112(d)(3) to mean a measure of mean, median, mode, or
some other measure of central tendency. The EPA concluded that it retains substantial discretion
within the statutory framework to set MACT floors at appropriate levels, and that it construes the
word "average" (as used in section 112(d)(3)) to authorize the EPA to use any reasonable
method, in a particular factual context, of determining the central tendency of a data set.
In addition, in the June 6, 1994, Federal Register notice, the EPA stated that it has discretion to
use "best engineering judgement" in collecting and analyzing data relevant to a MACT floor
determination, and in assessing the data comprehensiveness, accuracy, and variability in order to
determine which sources achieve the best emission reductions.
DATA COLLECTION FOR THE MACT FLOOR
The American Textile Manufacturers Institute (ATMI) member companies represent about 80
percent of manufacturing capacity in the textile industry. In the Spring of 1997, ATMI mailed a
MACT survey to member companies and to members of other Industry and State associations
that agreed to collaborate on the survey effort. Responses were received from almost 400
facilities, including 4 facilities with solvent-based pigment printing, 17 facilities with water-
based pigment printing, and 5 facilities with other printing 3. Only one of the facilities with
printing operations (solvent-based pigment printing) reported major source HAP potential to emit
from printing. All but 3 of the facilities reported HAP potential to emit less than 5 tons per year,
with 7 facilities reporting less than 1 ton per year HAP potential to emit.
1-3
-------
The ATMI database does not contain information about the materials used in printing. The EPA
and ATMI agreed that it would not be reasonable to resurvey printing facilities for detailed
process information, considering the low HAP emissions and potential to emit reported by
facilities in the ATMI MACT survey. However, ATMI noted that coating might not be well
represented in the survey 4. Therefore, EPA undertook a survey effort to collect additional
information from coating facilities. The EPA sent two different information collection requests
to coaters, each to 9 companies: the first group of questionnaires was sent to companies that coat
industrial fabrics5; and the second group of questionnaires was sent to companies that perform
cord treating and surface coating operations for rubber-coated textiles 6.
To develop the two lists of companies to receive the questionnaires, the 1996 toxic release
inventory (TRI) was used to identify facilities in the relevant SIC codes ( 2295 for industrial
fabrics and 2296, 3052, and 3069 for cord treating and surface coating) that were major sources
based on reported HAP releases to the air. Literature sources and stakeholders were consulted to
obtain information about number of employees, products, and whether facilities had undertaken
pollution prevention (P2) efforts. Companies were chosen for the mailing list to ensure
representation of different sizes of companies and a range of products. To obtain a sample that is
representative of the better performing facilities, preference was given to facilities that reported
taking P2 actions; hence, the EPA believes that a larger sample would not result in a substantially
different floor.
Responses were received from 22 facilities7. Five of the responses were classified largely as
confidential business information, which limited the usefulness of these responses in
characterizing the coating and printing subcategory. The results of the quantitative data
collection efforts provided the technical database used for the MACT floor determination.
In addition to quantitative information obtained from the survey, the EPA made four site visits to
coating facilities and two site visits to facilities with printing processes. The industry members
that participated in the stakeholder process included members of the American Textile
Manufacturer's Institute (ATMI), the American Yam Spinners Association (AYSA), the
Industrial Fabrics Association International (IFAI), the Northern Textile Association (NTA), and
the Rubber Manufacturer's Association (RMA), representatives of individual companies in the
regulated industry, and representatives of companies that supply coatings to the industry. States
that participated in the stakeholder process included Alabama, Florida, Georgia, North Carolina,
South Carolina, and Virginia. The U.S. EPA was represented by the Office of Air Quality and
Standards (OAQPS), the Office of Enforcement and Compliance Assurance (OECA), the Office
of Pollution Prevention and Toxic Substances (OPPTS), the Office of Research and
Development, and an EPA Small Business Ombudsman.
During stakeholder meetings, qualitative information from the Polymeric Coating of Supporting
Substrates - Background Information for Proposed Standards (EPA-450/3-85-022a, April 1987)
was presented. Comments on the qualitative information presented as well as additional
qualitative information were solicited from the stakeholders. The qualitative information
reviewed and discussed with the stakeholders is contained in the following memoranda:
1-4
-------
• Memorandum from Melissa Malkin and Steve York, RTI to Paul Almodovar,
EPA/OAQPS/ESD/CCPG. December 15, 1997 Final. Second PMACT Meeting for
Fabric Printing, Coating, and Dyeing.
Memorandum from Steve York, RTI to Paul Almodovar, EPA/OAQPS/ESD/CCPG.
February 2, 1998 Final. Initial Regulatory Subgroup PMACT Meeting for Fabric
Printing, Coating, and Dyeing.
Memorandum from Steve York, RTI to Paul Almodovar, EPA/OAQPS/ESD/CCPG.
March 2, 1998 Draft. Meeting with the American Yarn Spinners Association (AYSA)
Environmental Services Committee to discuss the status of the Fabric Printing, Coating,
and Dyeing MACT.
• Memorandum from Aarti Sharma and Steve York, RTI to Paul Almodovar,
EPA/OAQPS/ESD/CCPG. September 11,1998 Draft. EPA and Rubber Manufacturers
Association (RMA) meeting.
• Memorandum from Melissa Malkin and Steve York, RTI to Paul Almodovar,
EPA/OAQPS/ESD/CCPG. September 11, 1998 Draft. Summary of Northern Textile
Association (NTA)/U.S. Environmental Protection Agency (EPA) meeting to review the
MACT/PMACT status.
Qualitative information from these sources provided descriptions of coating and printing
processes, HAP control technologies, and process and control technology concerns. These data
verified that the coating processes and HAP emission sources are similar for all coating types and
that similar HAP control technologies are used. Therefore, the qualitative data provide a
representation of the coating industry and the control technology used by the industry. The
database is reflective of the variety of products that contain coated fabrics and the facilities that
will be subject to this rule.
Examples of the products manufactured from textiles coated by the facilities in the database
include:
• rubber belts and hoses for automotive use
• coated fabrics for use as tarps, hot air balloons, awnings, and outer wear (raincoats)
• commercial aircraft evacuation slides
• geomembranes
• speaker diaphragm surrounds
• luggage
• hot air balloons
• tennis and racquet balls
The floor facilities comprised the following types of production facilities:
• Urethane fabric coating and fabric laminating
• PVC and polyurethane coating of nylon and polyester fabrics
• Rubber and vinyl coating of textile substrates
1-5
-------
Coated fabrics produced by the floor facilities are used in manufacture of the following products:
truck tarps, geomembranes, roofing, tents, pillow tanks, architectural structures, billboards, hot
air balloons, inflatables, military fabric, air bag material for cars, and diaphragms for gas meters
and fuel pumps.
RESULTS OF DATA COLLECTION AND THE COATING MACT DATABASE
The quantitative information collected from the coating industry 8 was entered into a database
created to help determine MACT subcategory floor and to analyze impacts of regulatory options.
The coating MACT subcategory database presented in this memo contains a total of 17 facilities,
excluding 5 facilities that have classified most of the ICR response as confidential business
information (CBI). Information from the 5 facilities claiming CBI (with the exception of
emissions data, which were not claimed CBI) was not used in developing the summary data
presented in this section. In performing the MACT floor analysis, the relevant information from
the 5 facilities claiming CBI was examined to determine if any of the facilities qualified as
MACT-floor facilities. None was determined to be a MACT-floor facility.
The surveyed facilities were asked to provide facility HAP emissions from coating operations as
well as HAP emissions from the specific unit operations associated with coating. The total HAP
emissions for the 21 facilities reporting facility HAP emissions in the ICR response (one facility
did not report HAP emissions on the forms, but included sufficient HAP-containing materials
information to calculate the HAP emissions) were calculated to be 1,242 tons in 1997. Chart 1-1
presents a breakdown of the facility emissions by HAP. Unit operations associated with coating
for which HAP emissions estimates were requested including coating application, drying and
curing; substrate preparation; storage tanks; mixing; parts and equipment cleaning; and waste and
wastewater. Facilities in the MACT database reported only 4.3 percent of facility HAP
emissions from unit operations other than coating application and drying/curing (ancillary
operations), with mixing accounting for almost half of the emissions from ancillary operations.
This is roughly in line with a previous estimate of the split of VOC emissions from coating
operations made during development of the new source performance standards (NSPS) for
polymeric coating of supporting substrates 9.
Of the 21 coating MACT-database facilities that provided detailed information including
emissions and controls, thirteen facilities responded that they operate controls on their coating
lines; seven facilities reported operating with no controls. There are 29 controlled coating lines
in the MACT database. Of the 29 controlled lines, 16 lines are controlled with thermal oxidizers,
3 lines with catalytic oxidizers, 9 lines with carbon adsorbers, and one line with an electrostatic
precipitator. The reported data on capture and control device destruction efficiency consisted of
source test data, mass balance comparisons, vendor guarantees, and engineering judgement.
1-6
-------
Chart 1-1. Nationwide Coating Industry Emissions by HAP
Dimethyl f ormarride
3%
All Others
8%
Methyl ethyl ketone
34%
Toluene
47%
n Toluene
• Methyl ethyl ketone
D Hexane
D Dimethyl formamide
• All Others
1-7
-------
CRITERION FOR EVALUATING HAP EMISSION REDUCTIONS FROM COATING
OPERATIONS
The MACT floor for coating and printing was evaluated on the basis of the collection of all
operations at a facility associated with the surface coating of a textile; because, in general, the
facilities in the coating source category floor capture and control emissions from their coating
lines in this same manner. Surface coating and printing operations include preparation of a
coating for application (e.g., mixing with thinners); substrate preparation; coating application
and flash-off; drying and/or curing of applied coatings; cleaning of equipment used in surface
coating; storage of coatings, thinners, and cleaning materials; and handling and conveyance of
waste materials from the surface coating or printing operations. Coatings include such materials
as adhesives and protective or decorative coatings.
From analysis of the coating survey responses, it was found that coating application and curing
are the largest contributors of HAP emissions at coating facilities. On a nationwide basis, the
portion of total facility HAP emissions attributed to coating application and curing by
respondents to the coating MACT survey was approximately 95 percent Other operations and
activities that may create HAP emissions associated with coating include storage tanks, substrate
preparation, coating mixing/thinning operations, parts and equipment cleaning, and waste and
wastewater operations. In a facility with a permanent total enclosure (PTE) to capture fugitive
HAP emissions, at least some of the associated coating operations and activities (e.g., substrate
preparation, coating mixing/thinning operations, and parts and equipment cleaning) are
performed in the PTE. Fugitive HAP emissions from operations in the PTE are controlled at the
facility overall control efficiency (OCE).
The information concerning the level of HAP emissions from coating application and
drying/curing collected in the coating MACT survey included the capture efficiency for each
coating application area or for the entire coating line and the destruction efficiency of the control
device receiving the HAP emissions. The OCE for the coating line application and drying/curing
could be calculated from this information. Because this information was the value that was most
common among all the data available, and because it was determined that the coating application
and drying/curing OCE was the value that was most correlated with HAP emissions, coating
application and drying/curing OCE was used as the basis for the MACT floor calculations for
coating lines. The application and drying/curing OCE for the facilities in the MACT floor was
calculated as a facility-wide average of all coating lines, to incorporate the effects of averaging
across coating lines in facilities with more than one coating line.
CONSIDERATION OF DATA QUALITY IN EVALUATING HAP EMISSION
REDUCTIONS FROM COATING HAP SOURCES
There are a number of data quality issues that were considered in determining the MACT floor
for the coating industry. These issues raised questions concerning the representativeness of the
data in terms of what OCE the facilities can achieve in daily operations and over the entire year
versus what facilities report and in terms of the quality of the coating capture efficiency data.
1-8
-------
Representativeness of the Control Device Performance Data in the Coating MACT
Database
Representatives of two other web surface coating industries have noted that reported destruction
efficiencies can differ from those actually achieved in daily operation. These industries are the
metal coil surface coating and the paper and other web coating (POWC) industries, both of which
use web coating lines consisting of one or more work stations that apply the coating to the web
and subsequent drying stations, similarly to the coating industry. In fact, some coating lines are
used to coat both POWC and textile substrates.
The metal coil coating industry reports that efficiencies determined by testing are generally
measured during the initial compliance test, when the control device is new 10. Destruction
efficiency will gradually degrade with age (e.g., because of leaking heat exchangers or leaking
isolation valves), so that the reported destruction efficiency may not be representative of the
efficiency actually being achieved by control devices that have been in operation several years.
Furthermore, the metal coil coating industry notes that when a facility reports an efficiency based
on testing, it is usually based on test methods that call for averaging the results of three source
tests of the inlet and outlet emissions from the control device. These tests are generally relatively
short in duration (approximately one hour). Therefore, depending on the conditions of operation
during these tests, e.g., inlet HAP loading to the control device, the control efficiency data
acquired from the coating industry may not be representative of control device performance over
the entire range of normal facility operation and over the entire year.
An important operating parameter at coating facilities that can cause control device test results to
differ from control device performance during normal operation is the variation in loading rates.
It is possible that during compliance tests, the inlet HAP loading (i.e., the amount of HAP
volatilized from the surface and exhausted to the control device) is much higher than it is during
normal operations. This situation may result in artificially high destruction efficiency rates
achieved during testing. For example, thermal oxidizers are known to only achieve high levels of
control, such as the greater than 99 percent destruction efficiencies reported by some facilities in
the MACT database, when their inlet loadings are high n. Therefore, it is possible that
differences in reported destruction efficiencies in the coating database may only be a result of
variation in test conditions. The wide range of inlet loadings (from less than 100 ppmv to 8,500
ppmv) reported by coating facilities indicate that inlet loadings do fluctuate because of the batch
nature of the coating process (i.e., different products with different coating specifications are
often produced on the same line throughout the day). Therefore, inlet loadings will likely often be
lower than the inlet loading when the facility undergoes source testing for compliance purposes.
As a step in the data validation process, available literature was reviewed and thermal oxidizer
vendors were contacted to determine maximum destruction efficiencies that could be expected
for thermal oxidizers 12. The literature review on thermal oxidizers indicated that 99 percent
destruction efficiency is achievable under ideal conditions, but that lower efficiencies are
typically achieved under normal operating conditions. For example, the alternation between beds
1-9
-------
in a regenerative thermal oxidizer typically results in somewhat lower destruction efficiencies
than are achieved in a conventional recuperative thermal incinerator, generally below 99
percent13. The lower destruction efficiency for regenerative thermal incinerators has been
attributed in part to valve leaks within the system. In addition, a study conducted by EPA "*
concluded that 98 percent VOC reduction, or 20 ppmv by compound exit concentration is the
highest control level achievable by all new incinerators. This level is expressed as both percent
reduction and ppmw to account for the leveling off of exit concentrations as inlet concentrations
drop below 2000 ppmw.
Telephone surveys of thermal oxidizer manufacturers indicated that 98 percent is the routine
guarantee for regenerative or recuperative thermal oxidizers. Typically, this guarantee only
covers the first year of operation due to potential destruction efficiency degradation caused by
operational factors 1S. Vendors confirmed that long-term performance likely degrades because of
leakage problems. Typically, vendors reported that untreated gas leaks into the treated gas
stream through deterioration of heat exchange systems or leakage through isolation valves used
on multiple chamber regenerative units.
Because of the practical limitations of the coating survey and other industry research, information
on the specific test conditions for the control efficiency data collected was not available. For this
reason and the various factors described above, the determination of the MACT floor for coating
took into account the likelihood that the coating survey responses included only "best case" data,
which do not reflect degradation in performance over time or normal variations in coating
operations over an entire year.
Quality of Coating Capture Efficiency Data
For a coating line controlling HAP emissions by capturing the emissions and venting to a solvent
recovery device, coating line OCE is typically determined through liquid-liquid material balance
by measuring the volatile matter being applied on the coating line and the volatile matter
recovered and calculating the recovery efficiency. However, for a coating line controlling HAP
emissions by capturing the emissions and venting to a thermal oxidizer, coating line OCE is
calculated as the product of the capture and destruction efficiencies. A source can only report
100 percent capture if it meets the criteria of Method 24 of 40 CFR Part 51, Appendix M for total
enclosures. If the criteria are met and all gases from the enclosure are vented to a control device,
then capture efficiency is assumed to be 100 percent.
With regard to the database information on capture of HAP emissions from coating application
and drying/curing, it was clear that all the determinations of capture efficiencies were not
performed in the same manner. In evaluating the data for 5 facilities claiming 100 percent
capture, we found only one of these facilities reported the basis for the capture efficiency to be
permanent total enclosure (PTE) as determined by Method 204. Three facilities cited testing as
the basis for the 100 percent capture with the test method unspecified and one facility cited
testing and engineering judgement. A sixth facility that reported capture by PTE claimed 99
percent capture based on source testing. Follow up phone contacts of the 3 facilities citing
1-10
-------
testing as the basis for 100 percent capture with the test method unspecified revealed that Method
204 criteria had been met by two of the facilities. Therefore, of the 6 facilities claiming PTE,
only the data from the 3 facilities determining capture efficiency using Method 204 were used in
the MACT floor data base 16.
MACT FLOOR DETERMINATION
For this analysis, EPA determined that all 22 facilities in the coating MACT data base (including
facilities claiming responses to the questionnaire CBI) were major or synthetic minor facilities
with coating lines. Therefore, this set of 22 facilities was used to identify the top performing
facilities for coating line control as the basis for the MACT floor determination.
The coating line overall control efficiency (OCE) was calculated for all of the facilities with
sufficient information in the database as a facility-wide average, i.e., as an average of all of the
coating lines at a facility (that accounts for the effect of averaging across coating lines.) The
calculation procedure consisted of calculating an arithmetic average facility capture efficiency
(arithmetic average for all lines), an arithmetic average facility destruction (for facilities with
thermal oxidizers) or recovery (for facilities with carbon adsorbers) efficiency (arithmetic
average for all control devices receiving emissions from coating lines in the facility), and an
average facility OCE (product of average facility capture efficiency and average facility
destruction or recovery efficiency.) Arithmetic average facility capture and destruction or
recovery efficiencies were calculated because insufficient data were available to determine the
quantities and characteristics of coatings being applied on specific coating lines or stations.
Therefore, we don't know the contribution of the different lines to the total facility emissions.
Table 1-1 presents a ranking based on the average facility OCE of all facilities in the MACT
database with sufficient non-CBI information to calculate average facility OCE. For facilities
listed in the tablewithout an average facility OCE, the reason the OCE was not calculated (no
controls, information not available, or CBI) is noted.
1-11
-------
Table 1-1. Fabric Coating Average Facility OCE'
Facility
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Facility SIC
Code
2295
3052
NA
2295
2295, 3069
2295
3949
2295
3052
2295, 3052
2295
2295
2295
2295
3052
3069
3052
CB1
CBI
CBI
CBI
CBI
Type of Add-on
Control Device b
RTO
TO
TO
RTO
CA
TO
CA
TO
CA
CO
CO
NCf
NC
NC
NC
NC
NC
CBIg
CBI
CBI
CBI
CBI
Facility OCE
(%r
99.3
99.0
98.9
97.2
96.0
95.3
93.1
91.9
90.8
NAe
NA
NC
NC
NC
NC
NC
NC
CBI
CBI
CBI
CBI
CBI
Capture Efficiency
(%)"
100.0
100.0
100.0
100.0
100.0
99.0
98.0
93.8
99.8
NA
NA
NC
NC
NC
NC
NC
NC
CBI
CBI
CBI
CBI
CBI
Control Device
Efficiency
(%)"
99.3
99.0
98.9
97.2
96.0
96.3
95.0
98.0
91.0
94.0
90.0
NC
NC
NC
NC
NC
NC
CBI
CBI
CBI
CBI
CBI
a Includes average facility OCE for all facilities in the MACT database with sufficient non-CBI information to calculate average
facility OCE. For facilities without an average facility OCE, the reason the OCE was not calculated is noted.
b RTO = Regenerative Thermal Oxidizer; TO = Thermal Oxidizer; CA = Carbon Adsorber; CO = Catalytic Oxidizer.
c Product of average facility capture and control efficiencies as calculated from data reported by facility.
d Arithmetic average of data reported by facility if different efficiencies reported for different lines.
e NA = Not Available
f NC = No Control
g CBI = Confidential Business Information
NOTE: The 3 MACT floor facilities are highlighted.
1-12
-------
MACT Floor Determination for Existing Sources
As indicated previously in the BACKGROUND section of this memorandum, the MACT floor
for existing sources is determined based on the average emission limitation achieved by the best
performing twelve percent of existing sources. For the coating industry, OCE for the collection
of all coating lines at a facility is the emission limitation that reflects the best controlled sources.
The best performing 12 percent of the 22 facilities in the MACT database constitutes a set of 3
facilities.
As has been described previously, some facilities reported OCE's that could not be substantiated
based on the data provided supporting reported capture efficiency. Facilities with
unsubstantiated OCE's were not used in the MACT floor determination. Removing facilities
with unsubstantiated OCE's from the MACT floor resulted in the removal of two facilities,
which were replaced with the next best performing facilities with OCE's substantiated by
Method 204 or Procedure T verification of capture efficiency. The resulting top performing 12
percent of the facilities are the 3 facilities identified in Table 1-1 as MACT-floor facilities.
All of the top performing facilities use capture systems and control devices including both
thermal oxidizers and carbon adsorbers. The two facilities using thermal oxidizers are achieving
100 percent capture of application station emissions through the use of permanent total
enclosures. Table 1-1 shows that the range of reported OCE for the top 12 percent was 93.1 to
99.3 percent.
The reported coating values show that controls on some specific coating operations may be
capable of achieving greater than 99 percent HAP destruction based on 100 percent capture and
thermal oxidizer destruction efficiency greater than 99 percent. The average OCE of the MACT
floor facilities is 98.1 percent. However, to determine the level of emission control consistently
achievable with thermal oxidation, it is important to consider not only the level of control
reported, but also the previously cited data quality concerns and the control levels that EPA has
generally found to be achievable for this type of control technology. This approach ensures that
factors that affect control levels, such as variations in source operating conditions and inlet
loadings to the control device, are accommodated in the selection of the MACT floor.
The study conducted by EPA 17 indicated that a 98-percent reduction is the control efficiency
achievable by all new oxidizers. Information from vendor guarantees supports the determination
of a destruction efficiency of 98 percent for thermal oxidizers. Adjusting the destruction
efficiencies of the 2 facilities using thermal oxidizers in the MACT floor to 98 percent results in
the calculation of an average 97 percent facility-wide coating line OCE for the 3 facilities that
make up the best controlled twelve percent of the industry. Therefore, the MACT floor for
existing sources is 97 percent reduction of organic HAP emissions from the coating lines.
An OCE of 97 percent is attainable by all of the facilities in the MACT floor considering
available information regarding the capture and control technologies currently used at existing
sources in the coating industry. A facility using carbon adsorption for control can achieve 97
1-13
-------
percent by installing a PTE around the coating application station. A facility using a thermal
incinerator for control can achieve 97 percent with less efficient capture efficiency, e.g., 99
percent capture efficiency and 98 percent destruction efficiency.
MACT Floor Determination for New Sources
As indicated previously in the BACKGROUND section of this memorandum, the MACT floor
for new sources must reflect the emission control achieved in practice by the best-controlled
similar source. The OCE data in Table 1-1 show that the best-controlled source for which we
have data is using a permanent total enclosure to achieve 100 percent capture and a thermal
oxidizer to achieve a destruction efficiency greater than 99 percent.
As has been noted above in the description of the determination of the MACT floor for existing
sources, it is important to consider not only the level of control reported by the single best-
controlled coating facility (99+ percent facility-wide coating line OCE), but also the control
levels that EPA has generally found to be achievable for this type of control technology. As
described above, 98-percent reduction is the control efficiency achievable by all new oxidizers.
Furthermore, new solvent recovery systems can also be designed to achieve 98 percent control
efficiency 18. Therefore, these types of control devices used to reduce organic HAP emissions at
new coating facilities can be expected to achieve at least 98 percent emission reduction.
Consequently, a 98-percent facility-wide coating line OCE was determined to be the MACT
floor for new sources in the fabric coating industry.
Calculation of Alternative Emission Rates for Existing and New Sources
Data from the coating MACT database were used to calculate alternative facility emission rate
limits for existing and new sources. The alternative facility HAP emission rate for existing
sources was calculated based on applying the 97 percent MACT floor OCE to a pre-controlled
facility HAP emission rate representative for this industry. Similarly, the alternative facility
HAP emission rate for new sources was calculated based on applying the 98 percent MACT floor
OCE to a pre-controlled facility HAP emission rate representative for this industry. The
rationale for this is that an alternative facility HAP emission rate limit should not be more
stringent than the controlled HAP emission rate that can be attained by a coating facility using a
representative coating formulation and applying MACT floor control.
The calculation procedure consisted of defining a representative coating for this industry by
calculating the average pounds of HAP per pounds of solids for all of the facilities in the MACT
database with sufficient coating information. Fourteen of the 22 facilities in the MACT database
submitted detailed information about coating materials sufficient to calculate a facility average
coating in terms of pounds of HAP per pounds of solids. The pounds of HAP used to define the
representative coating included HAP used in thinning and HAP used as a cleaning solvent. All
of the HAP is assumed to be emitted; therefore, the coating composition also represents the pre-
controlled facility HAP emission rate.
1-14
-------
As shown in Table 1-2, the pre-controlled facility HAP emission rate was calculated as 4.16
pounds of HAP emitted per pound of solids. The pre-controlled facility HAP emission rate was
then factored by the 97 percent facility OCE MACT floor for existing sources to derive the
alternative facility HAP emission rate limit for existing sources of 0.12 pounds of HAP emitted
per pound of solids. The pre-controlled facility HAP emission rate was factored by the 98
percent facility OCE MACT floor for new sources to derive the alternative facility HAP emission
rate limit for new sources of 0.08 pounds of HAP emitted per pound of solids.
This equivalent emission rates were established in order to afford the complying facilities with
control options including low HAP coatings and a combination of low HAP coatings and add-on
controls. The units used in the equivalent emission limits are based on the units commonly used
in the industry and the format submitted on replies to questionnaires for this rulemaking.
Consideration of Beyond-the-Floor Technology for Existing and New Sources
The above the floor levels of control for coating and printing, to be considered, must be greater
than an overall control efficiency of 97 percent for existing sources. The floor for existing
sources was based on the use of control equipment with a control efficiency of 97 percent and a
capture efficiency of 100 percent. In addition, the 97 percent MACT floor overall control
efficiency was applied to a pre-controlled facility HAP emission rate representative for this
industry to calculate an alternative facility emission rate limit.
Two regulatory alternatives were identified that are more stringent than the existing source
MACT floor level of control for organic HAP and the alternative emission rate limit. These
alternatives were conversion to coating and printing materials that have a very low, or no,
organic HAP content and use of add-on capture systems and add-on control devices to achieve an
overall control efficiency of 98 percent.
Lower organic HAP liquid coatings fall into two primary categories. The most common
category is waterborne coatings, which allow the mixing of certain materials that would be
incompatible in organic solvent borne coatings. The second category is those higher solids
coatings that result from alternate technologies such as ultraviolet (UV)-curable coatings and
electron beam (EB)-curable coatings. Some urethane coatings can be applied with a thermal
process. These coatings do not employ organic HAP or VOC to keep the pigment and other
components of the coating in solution until curing. Therefore, organic HAP emissions are very
small.
These lower organic HAP coatings are currently in production use for some products in the
coating industry, but their applicability is limited in that, for some products, these coatings are
not able to achieve the desired final product characteristics. Similarly, low organic HAP or
waterborne printing
1-15
-------
Table 1-2. Coating Facility Average Emission Rate 1
Facility
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
AVG.
Total Pounds
of HAP in
Coating
Materials 2
598,393
72,946
626,980
643,217
459,780
894,252
939,155
848,199
16,043
35,301
CBI
CBI
CBI
CBI
Total Pounds of
Solids in
Coating
Materials 2
171,733
11,875
126,370
111,558
113,200
251,847
340,521
265,326
6,509
8,548
CBI
CBI
CBI
CBI
LbsofHAP/
Lbs of Solids3
3.48
6.14
4.96
5.77
4.06
3.55
2.76
3.20
2.46
4.13
CBI
CBI
CBI
CBI
4.16
Emission
Rate at 97 %
Facility OCE
0.10
0.18
0.15
0.17
0.12
0.11
0.08
0.10
0.07
0.12
CBI
CBI
CBI
CBI
0.12
Emission
Rate at 98 %
Facility OCE
0.07
0.12
0.10
0.12
0.08
0.07
0.06
0.06
0.05
0.08
CBI
CBI
CBI
CBI
0.08
Lists all facilities in the MACT database with sufficient information to calculate average
facility emission rate in terms of pounds of HAP emitted per pounds of solids applied.
Calculated from coating/coating component, thinning solvent, and cleaning solvent
materials reported by facility.
Calculated by dividing total pounds of HAP (including thinning and cleaning solvents) in
coating materials by total pounds of solids in coating materials.
1-16
-------
materials are used for the majority of printed products, but these printing materials are not able to
achieve the desired final product characteristics for certain products, such as designer and fashion
apparel, requiring the use of higher organic HAP printing materials. Given the limited
applicability of waterborne, UV-curable, EB-curable, and thermal ("hot-melt") coating and
waterborne printing materials, it was determined not to be feasible to require the use of these
coating and printing materials, therefore they were rejected as a beyond-the-floor option for
organic HAP.
It is technically feasible to reduce emissions from existing facilities by at least 98 percent
through the use of capture systems and add-on control devices. Based on the model plants
analysis used to estimate the impacts of the proposed rule, the incremental HAP reductions that
could be achieved by using capture systems and add-on control devices to comply with a
"beyond-the-floor" alternative of 98 percent reduction would range from about 0.09 Mg (0.1
tons) to about 3.8 Mg (4.2 tons) per facility. The effect of the alternative 98 percent reduction
would result in an estimated reduction of an additional 32 tons of HAP per year. To achieve this
small incremental HAP emission reduction, existing affected facilities would have to upgrade or
replace most existing add-on control systems. The incremental emissions reductions that would
be achieved at this time are not supported by the additional cost that many existing facilities
would incur to upgrade or replace existing add-on control systems. Therefore, requiring 98
percent overall control was rejected as a beyond-the-floor option for organic HAP at existing
sources in the coating and printing subcategory.
The above the floor levels of control for coating and printing, to be considered, must be greater
than an overall control efficiency of 98 percent for new or reconstructed affected sources. The
new source floor was based on the use of control equipment with a destruction efficiency of 98
percent and a capture efficiency of 100 percent. Vendors could not guarantee greater than 98
percent destruction efficiency for the operating conditions experienced in coating and printing
and over the life of the equipment.
The use of low HAP containing coating and printing materials was considered for an above the
floor option for new or reconstructed sources. However, as is explained above for existing
sources, it was determined that some products in the coating and printing industry cannot meet
certain performance characteristics with low-organic HAP coating and printing materials.
For these reasons it was determined that requiring above the floor emission limits for new or
reconstructed sources is not practicable for this subcategory.
1-17
-------
REFERENCES
1. Polymeric Coating of Supporting Substrates - Background Information for Proposed
Standards. Office of Air Quality Planning and Standards, U. S. EPA. EPA-450/3-85-
022a, April 1987. P. 3-18.
2. Preliminary Industry Characterization: Fabric Printing. Coating, and Dyeing. Office of
Air Quality Planning and Standards, U. S. EPA, September 1998.
3. Memorandum and Attachment from York, S. and A. Sharma, RTI to P. Almodovar,
EPA/OAQPS/ESD/CCPG. November 13, 1998 Final. Summary of meeting at which
ATMI presented the results of the ATMI MACT Survey to EPA.
4. Memorandum and 5 Attachments from Sharma, A. and S. York, RTI to P. Almodovar,
EPA/OAQPS/ESD/CCPG. August 8, 1997 Final. Summary of Initial PMACT Meeting
for Fabric Printing, Coating and Dyeing.
5. Letter and attachment, B. Jordan, EPA:OAQPS:ESD, to Bradford Industries, August 5,
1998, Enclosing Fabric Coating Information Collection Request for completion (identical
letters sent to other companies, mailing list attached).
6. Letter and attachment, B. Jordan, EPA:OAQPS:ESD, to M. Trembly, The Goodyear Tire
and Rubber Company, August 10, 1998, Enclosing 10-year MACT Surface Coating
Categories Information Collection Request for completion (identical letters sent to other
companies, mailing list attached).
7. U.S. Environmental Protection Agency. Fabric Printing, Coating, and Dyeing NESHAP.
ICR Responses. Office of Air Quality Planning and Standards. Research Triangle Park,
NC. Responses received September 1998 - October 1998.
8. Reference 7.
9. Reference 1, P. 3-18.
10. Environmental Resources Management. Metal Coil Surface Coating ICR Data Analysis
and MACT Floor Proposals. St Charles, Missouri. June 2, 1999. p. 9.
11. U. S. Environmental Protection Agency. Control Technologies for Hazardous Air
Pollutants Handbook. EPA/625/6-91/014. Office of Research and Development.
Washington, DC. June 1991. p. 3-2.
12. Reference 10.
13. USEPA. Survey of Control Technologies for Low Concentration Organic Vapor Gas
1-18
-------
Streams. USEPA, Office of Air Quality Planning and Standards. May 1995. p. 28.
14. Memorandum (and attachments) from Farmer, J. R., U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, to distribution. August 22, 1980.
Thermal incinerator and flare removal efficiency.
15. Reference 10, p. 10.
16. Memorandum from York, S., RTI to V. Hellwig, EPA/OAQPS/ESD/CCPG. October 10,
2000. Summary of add-on control device data from the fabric coating MACT database
facilities.
17. Reference 12.
18. Carbon Adsorption for Control of VOC Emissions: Theory and Full Scale System
Performance (EPA-450/3-88-012). U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina. June 1988.
1-19
-------
MEMORANDUM
January 11,2002
To: Printing, Coating and Dyeing of Fabrics and Other Textiles File
From: Steve York and Alton Peters, RTI
Subject: MACT Floor for Dyeing and Finishing Compounds
SUMMARY
This memorandum describes the methodology and conclusions of the maximum achievable
control technology (MACT) floor analysis for the dyeing and finishing subcategory of the
Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP. The analysis is based on
dyeing material information from dyeing operations at 30 major or synthetic minor fabric dyeing
facilities and finishing material information from finishing operations at 12 major or synthetic
minor fabric finishing facilities that were obtained from survey data. The dyeing MACT floor
for existing and new sources was determined to be 1.58 weight percent organic HAP in dyeing
materials as purchased. The finishing MACT floor for existing and new sources was determined
to be 0.03 weight percent organic HAP in finishing materials as purchased. For the purpose of
determining the mass fraction of organic HAP in a finishing material, each organic HAP that is
not an OSHA-defined carcinogen as specified in 29 CFR 1910.1200(d)(4) that is measured to be
present at less than 1 percent is counted as zero. Therefore, the floor for finishing is zero organic
HAP. A facility with both dyeing and finishing operations is allowed to average between the
floors, with the total mass of organic HAP in dyeing and finishing materials as purchased not to
exceed the sum of the organic HAP allowed in dyeing materials and finishing materials as
purchased.
BACKGROUND'
The Coating, Printing, and Dyeing of Fabric industry was identified as a source category of HAP
under section 112(c) of the Clean Air Act, as amended in 1990 (the Act), to be regulated by a
National Emission Standard for HAP (NESHAP) under section 112(d) of the Act. Section
112(d) of the Act directs the EPA to develop standards that require the maximum degree of
reduction in emissions of HAP that is achievable, which are commonly referred to as MACT
standards. For existing major sources, the Act requires MACT to be no less stringent than the
average emission limitation achieved by the best performing 12 percent of existing sources
among the data available to the Administrator. For new major sources, the Act requires MACT
to be no less stringent than the emission control that is achieved in practice by the best controlled
similar source. These minimum stringency levels are often referred to as the "MACT floor."
Dyeing and Finishing was determined to be a subcategory of Coating, Printing, and Dyeing of
Fabric. The manufacturing processes and materials and the HAP emissions set these processes
2-1
-------
apart from the other processes that are used in the manufacture of fabric products. Dyeing and
finishing processes both use various types of aqueous materials, the choice of which depends on
the type of substrate and the desired properties in the end product. Many facilities perform both
dyeing and finishing and use some common equipment (e.g., tenter frames) for unit operations in
both processes. In some cases the finishes are applied to fabric wet from the dyeing process and
no drying is done until after the finish application. No add-on HAP emission controls are known
to be in use on dyeing processes and very few on finishing processes. The few add-on emission
controls used on finishing processes were installed to control opacity and are not effective at
controlling HAP emissions. This memo is to explain the basis for the MACT Floor for this
subcategory.
Dyeing
Dyeing is the application of color to the whole body of a textile material with some degree of
color fastness. Textiles are dyed using continuous and batch processes and dyeing may take
place at any of several stages in the manufacturing process (i.e., prior to fiber extrusion, fiber in
staple form, yarn, fabric, garment). Most of textile dyeing is done in finishing departments of
basic textile manufacturing facilities, although there are also several commission dyehouses.
From an environmental perspective, dyeing has typically been viewed as a wastewater issue due
to large quantities of water, chemicals, and auxiliaries (such as salt) used.1- 2> 3
Dyeing is essentially a mass transfer process where the dye diffuses in solution, adsorbs onto the
fiber surface, and finally, within the fiber. Dyeing is complicated by the fact that there are many
sources of color variations, such as dyes, substrate, preparation of substrate, dyeing auxiliaries
used, and water. Processing variables such as time, temperature, and dye liquor ratio (pounds of
dyebath to pounds of cloth) also affect dyeing results. There are hundreds of dyes within several
dye classes (see Table 2-1), each of which exhibits different results when applied to different
types of fabric.
Various types of dyeing machines are used for both continuous and batch processes. Every dye
system has different characteristics in terms of versatility, cost, tension of fabric, use of carriers,
weight limitations, etc. Dyeing systems can be aqueous, non-aqueous (in organic solvents), or
use sublimation (thermosal, heat transfer). Hydrophilic fibers such as cotton, rayon, wool, and
silk, are typically easier to dye as compared with hydrophobic fibers such as acetate, polyesters,
polyamides, and polyacrylonotriles. 2
The four basic steps in the dyeing process are: dissolving or dispersing dye; diffusing dye onto
the fiber surface; absorbing dye onto the fiber surface; and diffusing dye into the fiber. Batch
dyeing involves moving the dye liquor through the goods or moving the goods through the dye
liquor. The textile material is immersed in the dyebath during the entire period of dyeing. In
batch dyeing, a certain amount of textile substrate, usually 220 to 2200 pounds, is loaded onto a
dyeing machine and is brought to equilibrium or near equilibrium with a solution containing the
dye. Once immersed in the dyebath, because the dyes have an affinity for the fibers, the dye
molecules leave the dye solution and enter the fibers over a period of minutes to hours.
2-2
-------
Table 2-1 Major Dye Classes and Substrate Fibers
Class
Acid
Azoic
Basic
Chrome
Direct
Disperse
Fiber Reactive
Naphthol (azoic)
Pigment
Sulfur
Vat
Fibers
Wool, silk, and nylon
Cotton and cellulose
Acrylic, certain polyesters
Wool, silk, nylon
Cotton, rayon, other cellulosic
Polyester, acetate, other synthetic
Cotton and other cellulosic, wool
Cotton, rayon, other cellulosic
All (requires binders)
Cotton and other cellulosic
Cotton and other cellulosic
Reference 1.
Auxiliary chemicals and controlled dyebath conditions (mainly temperature) accelerate and
optimize the action. The dye is fixed in the fiber using heat and/or chemicals after which the
substrate is washed to remove unfixed dyes and chemicals. There is a trend to use of lower
liquor ratios (pounds of dyebath to pounds of cloth) in batch dyeing, which lends benefits such as
faster heating/cooling and less waste. Batch equipment can usually be purchased as atmospheric
(operated below 212 °F) or pressurized (operated to about 280 °F) machines. 2'3'4 Most batch
dyeing is being done using pressurized machines, although some facilities use atmospheric
machines, especially for fabric dyeing.5 Atmospheric dyeing might be required for fleeces and
stretch fabrics, such as Lycra®, which typically cannot be dyed using jet equipment.6 Dyeing
processes in pressurized machines release no HAP emissions to the atmosphere since the process
is totally enclosed and the pressure is released at the end of the dyeing process by cooling the dye
bath which is subsequently drained before opening the dyeing machine.7 However, in some
cases, the drying of the pressure-dyed substrate releases HAP emissions.
Continuous processes typically consist of dye application, dye fixation with chemicals or heat,
and washing. Almost all continuous dyeing is done at atmospheric pressure.5 Continuous
dyeing is usually used for long runs of polyester/cotton fabrics and involves immersing fabrics in
a relatively concentrated dyebath for short periods. Textiles are fed continuously into a dye
range at speeds usually between 540 and 2690 feet per minute and a concentrated solution of
dyes and chemicals (held in pads) is moved evenly and uniformly to the goods with thorough
penetration. A pad mangle helps apply pressure to squeeze dye solution into the fabric and the
dye is usually diffused or fixed by heating in a steamer or oven. Dye fixation on fiber occurs
much more rapidly in continuous dyeing as compared to batch dyeing. After fabrics are dyed,
they are dried in ovens or tenter frames after washing to remove un-reacted chemical or loose
dye.1'2-3'4-6 Fabric that is processed through atmospheric batch dyeing is not dried at the dye
range; it is sent to finishing and may be finished wet or dry. 6
2-3
-------
Various classes of dyes can be used, e.g, disperse for synthetics and direct for cellulosics (see
Table 2-1). Dyes used in the textile industry are mostly synthetic and are derived from coal tar
and petroleum-based derivatives. Dyes are sold as powders, granules, pastes, liquid dispersions,
and solutions. Not only are dyes applied in different ways, they also impart color using different
mechanisms.2 Dyes can be classified according to chemical constitution or method of
application. Dyestuffs can work on principles of electrostatic bonding, covalent bonding, or
physical entrapment. For example, acid dyes work through the mechanism of electrostatic
bonding, whereas disperse dyes work by physical entrapment.4 Different dye classes exhibit
different affinities depending on the type of fiber, although even dyes within the same classes can
show wide affinity variations. They also exhibit different properties such as their fastness under
end use conditions such as light, laundering, or dry cleaning.
Various combinations of chemical auxiliaries and process conditions (temperature and pressure)
may be used to better fix the dye on the fabric or impart specific characteristics. For example, a
dye bath may contain the dyestuffs along with appropriate auxiliaries such as wetting agents and
also specific chemicals such as acetic acid or sodium hydroxide.4 The use of higher
temperatures and superatmospheric pressures have reduced the need for dye carriers (chemical
accelerants) that were required at lower temperatures for the use of disperse dyes on synthetic
substrates, such as polyester.'
The sources of HAP emissions from dyeing are the HAP constituents that are contained in
dyestuffs and auxiliary chemicals as purchased. The HAP constituents are needed to impart
certain desirable characteristics to the dyed substrate (e.g., certain colors can only be attained
through the use of HAP-containing dyestuffs or auxiliaries.) No HAP is known to be added by
the users. The fraction of HAP contained in dye materials that is emitted to the atmosphere is
generally estimated to range from zero to 10 percent or greater, depending on the characteristics
of the specific HAP constituents and the pressures and temperatures that the HAP are exposed to
in the dyeing process operations. One source test showed emissions of almost 19 percent of the
incoming HAPs at one emission point. Although some of the HAPs from these operations
remain in the waste water, there are no partition data, nor data on the atmospheric emissions from
the waste water treatment aeration basins at the textile mills. The HAP content of the material
usage is the best data available and the basis for the MACT floor is the input of HAPs in the dye
materials.
Most HAP constituents are believed to be rinsed from the substrate before the substrate is dried,
because drying a substrate with unattached dye would adversely affect the quality of the dyed
product. Because users of the dye materials do not add HAP to the purchased materials, the
amount of HAP in the dye formulations is generally much less than 1 percent, the point in the
process where the HAP are emitted depends on the types and configurations of dye equipment
and unit operations used, and no add-on emission controls are known to be used on dyeing
processes, a mass limit on the amount of HAP contained in dyeing materials (i.e. weight percent)
"as purchased" was chosen as the format of the standard.
The MACT database for dyeing consisted of a sample of 41 facilities for which EPA had
complete dyeing materials usage data from responses to survey questionnaires. Since the dyeing
2-4
-------
and finishing subcategory consists of more than 30 operating facilities, the MACT floor is based
upon the best performing 12 percent of existing sources among available data. All of the
information in the MACT database is confidential business information (CBI), therefore, no
individual facility data are presented in this memorandum. The control option for all of the floor
facilities is to limit the HAP content "as purchased" of the dyestuffs and auxiliary chemicals used
in dyeing.
Finishing
Finishing refers to any process operation performed after bleaching, dyeing, or printing that
improves the appearance and/or usefulness of a textile substrate. Finishing encompasses any of
several mechanical (e.g., texturizing, napping) and chemical processes (e.g., optical finishes,
softeners, urea-formaldehyde resins for crease resistance) performed on fiber, yarn, or fabric to
improve its appearance, texture, or performance. *'2 Since the HAP emission sources from
finishing are specific chemical compounds that may be applied and released during subsequent
drying and curing operations, the MACT floor for finishing compounds is derived from available
information on chemical finishing processes. Chemical finishing is also referred to as wet
finishing. No chemicals are used in mechanical, or dry, finishing.
The fabric is usually dried prior to chemical finishing using either convective (hot air) or
conductive (heated cans) methods.3 Chemical finishing is commonly done on a continuous
finishing range (pad and tenter frame). Fabric is passed through an aqueous solution containing
the finishing chemical(s) and auxiliaries. After treatment, the fabric is typically passed through
an oven to drive off water and activate/cure finishing chemicals. It is important to note that
there is no set recipe for the chemical finishes or mechanical finishing processes applied to any
given substrate. Finishing methods are used according to desired characteristics of the end
product (which vary widely and are market driven) and the firms themselves have some amount
of flexibility in the specific processes or chemicals they choose to use for a particular function.
The textile industry uses numerous categories of proprietary chemical speciality products that are
used as chemical finishes. Some examples of chemical finish classes include 1?4'6 :
• Resin finishes (permanent press) are used on cotton or rayon to minimize the need to
ironing by keeping the fabric smooth after washing and drying. Most resins contain
formaldehyde; resins without formaldehyde are typically much costlier and adversely
affect product quality.
• Softeners are used with resins to improve the way the fabric feels by breaking down
hardness or stiffness.
• Stain resist finishes are used extensively on carpets and upholstery fabrics. Soil release
finishes allow soils and stains to be removed by laundering
• Water repellants used to prevent fabrics from being wet out (breathable, unlike
waterproofing agents) include but are not limited to wax, silicone compounds, and
fluorine compounds.
• Flame retardant qualities can be achieved by using special fibers or phosphorus-based
finishes.
2-5
-------
• Antistatic agents decrease or eliminate static electricity in textiles.
• Stiffeners give the fabrics body or stiffness.
Other examples of types of chemical finishes include anticreasing agents, deodorants, moth
resisting agents, oil repellants, rust preventatives, and shrinkage controllers. Some companies
use more specialized finishes like electrical finishes and Teflon®. Because there are typically a
wide variety of choices of chemical finishes that can be used within each finish class, it is often
difficult to tag finishes used in certain classes as always toxic or nontoxic. In certain cases, as in
the case of permanent press finishes, most of the resins used contain formaldehyde, although low
or non-formaldehyde finishes are being developed to suit certain applications.5
There are also several different types of mechanical finishing techniques. For example,
heatsetting can be done to improve dimensional stability in synthetic fabrics. Shearing involves
using rotary blade(s) to trim raised surfaces and reduce pilling. Other examples include
embossing, glazing, sueding, and polishing.
Many chemical and mechanical alternatives are available for every finishing operation, but the
specific nature and applicability of these is unclear. Some mechanical finishes and design
alternatives can avoid chemical processing. For example for softness, enzyme softening of
cotton and other mechanical alternatives can be used. Proper use and application of N-methylol
crosslinkers can minimize formaldehyde releases. Mechanical finishing (compacting) can also
eliminate use of the crosslinker. Some crosslinkers that eliminate formaldehyde are available,
but much more expensive. The industry has made a lot of efforts to reduce amount of free
formaldehyde in resins, however good substitutes that do not adversely affect the quality of the
product are difficult to find.5 Formaldehyde contents can vary anywhere from less than one half
of one percent for light weight fabrics to 4 percent for heavy fabrics (melamine-formaldehyde
resins), and there is a lot of variability in types of resins. Formaldehyde itself does not affect the
product, however it does affect the properties of the resin itself (manufacturing). Acrylic
handbuilders and stiffeners can replace formaldehyde-based handbuilders.
The sources of HAP emissions from finishing are the HAP constituents that are contained in
finishing materials as purchased. As is the case with dyeing, the HAP constituents are needed to
impart certain desirable characteristics to the finished substrate (e.g., a resin finish containing
HAP might be applied to a cotton/polyester blend for durable press and dimensional stability.)
No HAP is known to be added by the users. In finishing, unlike in dyeing, the fraction of HAP
contained in finishes that is emitted to the atmosphere is generally assumed to be 100 percent
with the exception of HAP that cross-link to the fiber, such as formaldehyde. This is because
finished fabric is generally dried and cured at relatively high temperatures over 300 °F. Because
users of the finishing materials do not add HAP to the purchased materials, the amount of HAP
in the finish formulations is generally much less than 1 percent, and very few add-on emission
controls are known to be used on finishing processes, a mass limit on the amount of HAP
contained in finishing materials (i.e. weight percent) "as purchased" was chosen as the format of
the standard.
The MACT database for finishing consisted of a sample of 31 facilities for which EPA had
2-6
-------
complete finishing materials usage data from responses to survey questionnaires. Since the
dyeing and finishing subcategory consists of more than 30 operating facilities, the MACT floor is
based upon the best performing 12 percent of existing sources among available data. As is the
case with dyeing, all of the information in the MACT database is confidential business
information (CBI), therefore, no individual facility data are presented in this memorandum. The
control option for all of the floor facilities is to limit the HAP content "as purchased" of the
chemicals used in finishing.
APPROACH TO ESTIMATING THE MACT FLOOR
The term "average," as it pertains to MACT floor determinations for existing sources, described
in section 112(d)(3) of the Act, is not defined in the statute. In a Federal Register notice
published on June 6, 1994 (59 FR 29196), the EPA announced its conclusion that Congress
intended "average" as used in section 112(d)(3) to mean a measure of mean, median, mode, or
some other measure of central tendency. The EPA concluded that it retains substantial discretion
within the statutory framework to set MACT floors at appropriate levels, and that it construes the
word "average" (as used in section 112(d)(3)) to authorize the EPA to use any reasonable
method, in a particular factual context, of determining the central tendency of a data set.
In addition, in the June 6, 1994, Federal Register notice, the EPA stated that it has discretion to
use "best engineering judgement" in collecting and analyzing data relevant to a MACT floor
determination, and in assessing the data comprehensiveness, accuracy, and variability in order to
determine which sources achieve the best emission reductions.
DATA COLLECTION FOR THE MACT FLOOR
The American Textile Manufacturers Institute (ATMI) member companies represent about 80
percent of manufacturing capacity in the textile industry. In the Spring of 1997, ATMI mailed a
MACT survey to member companies and to members of other Industry and State associations
that agreed to collaborate on the survey effort. Responses were received from almost 400
facilities, including 8 facilities that continuous dye fiber, 24 facilities that continuous dye yarn,
36 facilities that continuous dye fabric, 8 facilities that batch dye fiber, 31 facilities that batch
dye yarn, 49 facilities that batch dye fabric and 81 facilities with wet finishing operations.
The ATMI MACT survey database 8 does not contain information about the materials used in
dyeing and finishing. However, ATMI conducted dyeing and finishing surveys of member
companies to collect information on the annual usage of dyeing and finishing materials by dye or
finish class, average and maximum HAP contents as purchased and as formulated, and actual and
potential annual HAP emissions. Responses to the ATMI dyeing survey 9 were received from 41
facilities; 31 facilities responded to the ATMI finishing survey I0. Because of the production-
related data collected in the surveys, the responses to both surveys were classified CBI by ATMI.
Therefore, no specific facility data are presented in this memorandum; only general descriptions
of the database and aggregated data related to the HAP content of materials "as purchased." The
results of the quantitative data collection efforts provided the technical database used for the
2-7
-------
MACT floor analysis.
In addition to quantitative information obtained from the surveys, the EPA made eight site visits
to facilities with dyeing and finishing operations. The industry members that participated in the
stakeholder process included members of the American Textile Manufacturer's Institute (ATMI),
the American Yam Spinners Association (AYSA), and the Northern Textile Association (NTA),
representatives of individual companies in the regulated industry, and representatives of
companies that supply dyeing and finishing materials to the industry. States that participated in
the stakeholder process included Alabama, Florida, Georgia, North Carolina, South Carolina, and
Virginia. The U.S. EPA was represented by the Office of Air Quality and Standards (OAQPS),
the Office of Enforcement and Compliance Assurance (OECA), the Office of Pollution
Prevention and Toxic Substances (OPPTS), the Office of Research and Development, and an
EPA Small Business Ombudsman.
During stakeholder meetings, qualitative information concerning dyeing and finishing process
operations, associated HAP emissions, and control options including pollution prevention
measures was presented. Comments on the qualitative information presented as well as
additional qualitative information were solicited from the stakeholders. The qualitative
information reviewed and discussed with the stakeholders is contained in the following
memoranda:
• Memorandum from Melissa Malkin and Steve York, RTI to Paul Almodovar,
EPA/OAQPS/ESD/CCPG. December 15, 1997 Final. Second PMACT Meeting for
Fabric Printing, Coating, and Dyeing.
Memorandum from Steve York, RTI to Paul Almodovar, EPA/OAQPS/ESD/CCPG.
February 2, 1998 Final. Initial Regulatory Subgroup PMACT Meeting for Fabric
Printing, Coating, and Dyeing.
Memorandum from Steve York, RTI to Paul Almodovar, EPA/OAQPS/ESD/CCPG.
March 2, 1998 Draft. Meeting with the American Yarn Spinners Association (AYSA)
Environmental Services Committee to discuss the status of the Fabric Printing, Coating,
and Dyeing MACT.
• Memorandum from Melissa Malkin and Steve York, RTI to Paul Almodovar,
EPA/OAQPS/ESD/CCPG. September 11, 1998 Draft. Summary of Northern Textile
Association (NTA)/U.S. Environmental Protection Agency (EPA) meeting to review the
MACT/PMACT status.
• Memorandum from Steve York and Aarti Sharma, RTI to Paul Almodovar,
EPA/OAQPS/ESD/CCPG. November 13, 1998 Final. Summary of meeting at which
ATMI presented the results of the ATMI MACT survey to EPA.
• Memorandum from Melissa Malkin and Steve York, RTI to Paul Almodovar,
EPA/OAQPS/ESD/CCPG. November 13, 1998 Final. Summary of ATMI Task
Force/EPA information gathering meeting.
Qualitative information from these sources provided descriptions of fabric dyeing and finishing
processes, pollution prevention opportunities and verified that HAP control technologies are not
2-8
-------
used on dyeing and finishing HAP emission sources except in a few cases to control opacity from
finishing processes. The qualitative data provide a representation of the fabric dyeing and
finishing industry. The database is reflective of the variety of dyeing and finishing processes that
are used by the facilities that will be subject to this rule.
RESULTS OF DATA COLLECTION AND THE DYEING AND FINISHING MACT
DATABASE
The quantitative information collected from the dyeing9 and finishing 10 industry was entered
into a database created to help determine MACT subcategory floor and to analyze impacts of
regulatory options. The dyeing and finishing MACT subcategory database from which
information was extracted and summarized in this memo contains a total of 30 facilities that are
major or synthetic minor HAP emission sources with dyeing processes and 12 facilities that are
major or synthetic minor HAP emission sources with finishing processes. See the "MACT
FLOOR DETERMINATION" section of this memo for a description of the reasons eleven
facilities with dyeing processes and 19 facilities with finishing processes could not be used in the
MACT floor analysis.
The surveyed facilities were asked to provide annual facility HAP emissions from dyeing and
finishing operations. The HAP contained in dyeing and finishing materials was speciated, but
emissions were reported as total HAP. The organic HAP reported in dyeing materials (dyes and
auxiliaries) at levels of at least 5 weight percent included ethylene glycol, glycol ethers,
methanol, biphenyl, 1,2,4-trichlorobenzene, and dimethyl phthalate. The total HAP emissions
from dyeing for the 30 facilities reporting facility HAP emissions were calculated to be 86 tons
in 1999. The organic HAP reported in finishing materials at levels of at least 5 weight percent
included methanol, ethylene glycol, and glycol ethers. The total HAP emissions from finishing
for the 12 facilities reporting facility HAP emissions were calculated to be 120 tons in 1999. The
HAP emissions estimates were based on the quantity of HAP in materials used in dyeing and
finishing processes in 1999 and were not broken down by process operation (i.e., storage,
mixing, substrate preparation, application, drying, curing, cleaning, waste and wastewater).
CRITERION FOR EVALUATING HAP EMISSION REDUCTIONS FROM DYEING
AND FINISHING OPERATIONS
The MACT floors for dyeing and finishing were evaluated on the basis of the HAP content of the
purchased materials used in the dyes and finishes applied at a facility. There are currently no
emission controls used to reduce HAP emissions from dyeing operations. The few emission
controls used on finishing operations were installed to reduce opacity and most are not efficient
at reducing HAP emissions. Furthermore, no emission factors have been developed for dyeing or
finishing operations and the split of emissions, particularly from dyeing, are dependent on site
specific conditions such as the unit operations the fabric passes through in the process range, the
types of equipment used for the process, the dye or finish chemistry, and the process conditions,
e.g., the points in the process where the fabric is subjected to heat. Finally, the available data
include information on the HAP content of the dyeing or finishing materials used annually at a
facility and HAP emission estimates based on the mass of HAP contained in the materials used
2-9
-------
in the process. Defining the MACT floor in terms of the mass of HAP per mass of purchased
materials (weight percent HAP in the purchased materials) correlates directly to HAP emissions,
serves to reduce the HAP emissions at the source, and is not dependent on the split of emissions
between different unit operations in the process range or between media (air and water).
MACT FLOOR DETERMINATION
For this analysis, EPA determined that a total of 30 of the 41 facilities in the ATMI dyeing
MACT database 9 are major or synthetic minor HAP emission sources and 12 of the 29 facilities
in the ATMI finishing MACT database 10 are major or synthetic minor HAP emission sources.
Eleven facilities with dyeing processes could not be used in the MACT floor analysis for the
following reasons: one facility has been shut down, 9 are area sources, and the Title V HAP
status of one facility has not been determined. Similarly, 19 facilities with finishing process
information could not be used in the MACT floor analysis for the following reasons: one facility
has been shut down, one reported only coating processes, 15 are area sources of HAP emissions,
and the Title V HAP status of 2 facilities has not been determined. Information from the
facilities with indeterminate Title V HAP status was examined to determine if any of the
facilities could potentially be MACT floor facilities. None was determined to be a MACT-floor
facility. Separate MACT floor analyses were done for dyeing and finishing, as described in the
following paragraphs.
MACT Floor Determination for Dyeing
Two different approaches were taken to calculate the MACT floor weight percent organic HAP
in purchased materials for dyeing. In the first approach, the weight percent organic HAP in
purchased materials for dyeing was calculated for each facility in the ATMI dyeing MACT
database 9. The dyeing survey collected information on the organic HAP content of dyes and of
auxiliary chemicals. To calculate the weight percent organic HAP, the mass of organic HAP in
dyes as purchased and the mass of organic HAP in auxiliaries as purchased were calculated.
Then the total mass of organic HAP in dye materials as purchased (mass of organic HAP in dyes
plus mass of organic HAP in auxiliaries) was calculated and divided by the total mass of dye
materials purchased (mass of dyes plus mass of auxiliaries) and multiplied by 100 to calculate
the weight percent HAP in dye materials purchased by each facility. Four floor facilities were
chosen (12 percent of 30), each of which reported zero organic HAP in dye materials as
purchased, therefore, the calculated MACT floor was zero weight percent organic HAP.
However, under this approach only 3 of the 11 dye classes reported in the dyeing survey were
represented in the MACT floor.
Since the choice of a dye class depends on many factors including substrate, color (market
driven), end use of the dyed fabric, and quality (e.g., dye fastness) and can not be made purely on
the basis of organic HAP content of the materials, EPA chose a second approach to calculating
the MACT floor that would represent all of the dye classes reported in the dyeing survey. Under
this second approach, a MACT floor analysis was done for each dye class in the database. For
each dye class used by each facility, the weight percent organic HAP in dye materials purchased
was calculated by calculating the total mass of organic HAP in dye materials as purchased for the
2-10
-------
dye class (mass of organic HAP in dyes plus mass of organic HAP in auxiliaries) divided by the
total mass of dye materials purchased for the dye class (mass of dyes plus mass of auxiliaries)
multiplied by 100. The number of facilities reporting use of each dye class ranged from 2 to 14
facilities. Taking 12 percent of each of these groups resulted in choosing one or two floor
facilities reporting the lowest weight percent organic HAP in dye materials for each dye class.
Table 2-2 presents the MACT floor organic HAP content calculated for each dye class in the
database. To determine the MACT floor for dyeing, a weighted average organic HAP content of
dye materials as purchased was calculated from the dye class MACT floors, using the total mass
of dye materials used by the MACT floor facility or facilities for each dye class to weight the dye
class MACT floor organic HAP contents. As shown in Table 2-2, the dyeing MACT floor
organic HAP content in materials as purchased was determined to be 1.58 weight percent for
existing sources. No technology has been identified that could achieve a lower organic HAP
content in materials as purchased, therefore the dyeing MACT floor organic HAP average
content in materials as purchased for new sources was also determined to be 1.58 weight percent.
MACT Floor Determination for Finishing
Since the choice of a finish class depends on the desired characteristics of the finished substrate
and can not be made solely on the basis of the HAP content of the finish, EPA chose the
approach of calculating the MACT floor that would represent all of the finish classes reported in
the ATMI finishing MACT database 10. As was the case for dyeing, a MACT floor analysis was
done for each finish class in the database. The finishing survey collected information on the
organic HAP content of each finish class as purchased. In some cases, facilities reported
different chemistry for finishes within the same finish class for use on different products.
Therefore, for each finish class used by each facility, the weight percent organic HAP in finish
materials purchased was calculated by determining the total mass of organic HAP in finish
materials as purchased for the finish class (sum of the mass of organic HAP in different
formulations within the finish class) divided by the total mass of finish materials purchased for
the finish class (sum of mass of finish materials purchased within the finish class) multiplied by
100. The one facility (12 percent of the number of facilities reporting use of the finish class,
which ranged from 1 to 8) reporting the lowest weight percent organic HAP in finish materials
for each finish class was chosen as the floor facility.
2-11
-------
Table 2-2. Dyeing MACT Floor
Dye Class
Acid
Basic
Develop
Direct
Disperse
Napthol
Neutral Premetalized
Pigment
Reactive
Sulfur
Vat
Dyeing MACT Floor a
Weighted Average % Organic HAP in Dye
Class Floor
0.0
0.0
0.0
0.51
0.0
0.0
0.01
0.03
0.0
5.02
0.0
1.58
a Weighted average of dye class floors.
Table 2-3 presents the MACT floor organic HAP content calculated for each finish class in the
database. To determine the MACT floor for finishing, a weighted average organic HAP content
of finish materials as purchased was calculated from the finish class MACT floors, using the total
mass of finish materials used by the MACT floor facility for each finish class to weight the finish
class MACT floor organic HAP contents. As shown in Table 2-3, the finishing MACT floor
organic HAP content in materials as purchased was determined to be 0.03 weight percent for
existing sources. For the purpose of determining the mass fraction of organic HAP in a finishing
material, each organic HAP that is not an OSHA-defined carcinogen as specified in 29 CFR
1910.1200(d)(4) that is measured to be present at less than 1 percent is counted as zero.
Therefore, the floor for finishing is zero organic HAP. No technology has been identified that
could achieve a lower organic HAP content in materials as purchased, therefore the finishing
MACT floor organic HAP content in materials as purchased for new sources was also
determined to be zero.
2-12
-------
Table 2-3. Finishing MACT Floor
Finish Class
Melamine
Non-Melamine
Water Repellants
Soil/Stain Resistant
Hand Softening
Hand Building
Flame Retardant
Other3
Finishing MACT Floor b
Weighted Average % Organic HAP in
Finish Class Floor
0.20
0.05
0.0
0.12
0.0
0.01
0.0
0.0
0.03
3 Other finishes reported include lubricants, wetting agents, anti-stick, and dressing.
b Weighted average of finishing class floors.
MACT Floor for Dyeing and Finishing Subcategory
The dyeing and finishing MACT floors represent planks in the MACT floor for the dyeing and
finishing subcategory. In a textile finishing facility with both dyeing and finishing processes,
averaging of organic HAP in materials as purchased for dyeing and finishing may be done within
the total mass of HAP allowable under the MACT floors for dyeing and finishing. For example,
if a facility uses dye materials with no organic HAP, the mass of organic HAP allowed by the
MACT floor (1.58 weight percent of the dye materials purchased) may be contained in the
finishing materials as purchased. Therefore, a facility with both dyeing and finishing operations
can choose to meet the MACT floors for each process individually, or can limit the mass of
organic HAP contained in dyeing and finishing materials as purchased to the sum of the
allowable mass of HAP under the dyeing and finishing MACT floors.
It should be noted that the reportable quantity of HAPs in the dyeing and finishing material is
limited to more than 0.1 percent by mass for carcinogenic compounds as specified in 29 CFR
1910.1200(d)(4) and more than 1.0 percent by mass for other HAP compounds. This is
consistent with the data in the MACT database; several facilities reported no HAP in purchased
materials on the basis of the HAP being less than reportable quantities in material safety data
sheets (MSDS).
Consideration of Beyond-the-Floor Technology for Existing and New Dyeing and
Finishing Sources
2-13
-------
The MACT floors for existing and new or reconstructed sources in the dyeing and finishing
subcategory are based on the best information available. The floors represent pollution
prevention options yielding the "best performing" and achievable emission rates for new or
reconstructed and existing sources in each subcategory. No "above the floor" technology has
been identified that could achieve a lower organic HAP content in materials as purchased and
would be applicable to all products for dyeing operations and zero percent HAP is the lowest
organic HAP content in materials as purchased for finishing operations that can be achieved.
REFERENCES:
1. Best Management Practices for Pollution Prevention in the Textile Industry. USEPA
Office of Research and Development. Washington, DC. EPA/625/R-96/004. September
1996.
2. North Carolina State University College of Textiles. Department of Extension and
Applied Research. Short Course Office. Fundamental Series - Basic Textiles, short
course materials.
3. "Summary of Textile Manufacturing Operations". Submitted to the USEPA by the
American Textile Manufacturers Institute. Prepared by the Institute of Textile
Technology. July 1991.
4. Memorandum and 5 Attachments from Sharma, A. and S. York, RTI, to P. Almodovar,
EPA/OAQPS/ESD/CCPG. August 8, 1997 Final. Summary of Initial PMACT Meeting
for Fabric Printing, Coating, and Dyeing.
5. Memorandum and Attachment from York, S. and A. Sharma, RTI to P. Almodovar,
EPA/OAQPS/ESD/CCPG. November 13, 1998 Final. Summary of meeting at which
ATMI presented the results of the ATMI MACT survey to EPA.
6. Memorandum and 4 Attachments from Malkin, M. and S. York, RTI, to P. Almodovar,
EPA/OAQPS/ESD/CCPG. December 15, 1997 Final. Summary of Second PMACT
Meeting for Fabric Printing, Coating, and Dyeing.
7. Memorandum and 2 Attachments from York, S. and A. Peters, RTI, to V. Hellwig and P.
Almodovar, EPA/OAQPS/ESD/CCPG. January 8, 2001 Final. Summary of American
Textile Manufacturers Institute (ATMI)/US Environmental Protection Agency (EPA)
MACT Development Meeting for Fabric Printing, Coating, and Dyeing.
8. U.S. Environmental Protection Agency. Fabric Printing, Coating, and Dyeing NESHAP.
ICR Responses. Office of Air Quality Planning and Standards. Research Triangle Park,
NC. Responses received September 1998 - October 1998.
9. Letter, J. Fleming, ATMI, to G. V. Hellwig, EPA: OAQPS: CCPG, July 27, 2000. ATMI
2-14
-------
MACT Development Support and Data Submission (textile dyeing).
10. Letter, J. Fleming, ATMI, to G. V. Hellwig, EPA: OAQPS: CCPG, November 2, 2000.
ATMI MACT Development Support (textile finishing).
2-15
-------
MEMORANDUM
January 10, 2002
To: Printing, Coating and Dyeing of Fabrics and Other Textiles File
From: G. V. Hellwig
Subject: MACT Floor for Slashing
Slashing is a yarn preparation process performed on warp yarn prior to weaving. Warp yarns
need to sustain their elongation and flexibility during the weaving process, which necessitates the
slashing process. In the slashing process, large rolls (beams) of warp yarn are passed through a
size box containing the aqueous sizing compound. Squeeze rolls remove excess solution and
the yarn then passes through a drying unit that usually consists of steam filled dry cans (rollers)
or an oven and then through a series of separator bars to prevent the ends from sticking together.1
After the separation process, the warp is then wound onto the loom beam.2 Some mills perform
desizing. During the desizing step, at the end of the textile process, most of the sizing (slashing
material) is removed from the textile by washing and the sizing is present in the wastewater.3
The objectives of slashing are to strengthen, smooth the outer surface, and lubricate the yarn. The
chemical nature of the size applied is dependent on the yarn substrate and the type of weaving
being used. The three main types of size currently used are natural products (starch), fully
synthetic products [e.g., polyvinyl alcohol (PVA)], and semisynthetic blends (e.g., modified
starches and carboxymethyl cellulose or CMC).2 When starch or modified starch is the sizing
compound there is water but no HAPs emitted from the slashing process. Starch is used
principally on cotton, but does not work well on synthetic fabrics. Also, starch is not more
widely used, and is not a good substitute for synthetic sizing, because of water pollution
concerns. Starch greatly increases the BOD and cannot be partially recycled. The PVA and
CMC are typically recycled when possible to reduce water treatment and water pollution. CMC
is not as widely used as starch and PVA because of the cost of the material. CMC is not as
effective in the slashing process on cotton and synthetic textiles as starch, modified starches or
PVA, respectively.4
'Cone. White Oak Plant. Cone Mills Corporation, Greensboro, NC, 1998
2A Dictionary of Textile Terms. Dan River, Inc. New York, New York, 1992
3The Basics of Textiles, NC State University College of Textiles, Raleigh, NC, May 1,
2000.
4Reference 3.
3-1
-------
The primary source of HAP emissions from slashing is methanol from (PVA) size, typically
applied to synthetics (although it adheres to and is used for natural fibers as well). The methanol
is present in the PVA size as a contaminant, and is not needed for the slashing process. The
methanol emissions can arise either from the size cooking operation and/or from the application
or slashing process - the distribution is unclear, although it will depend upon the temperature at
which the size is cooked, the cooking time, and how often mixing containers (cookers) are
opened.5 These processes are not presently regulated by federal, state or local agencies, and there
are no known HAP emission capture or control systems in use on size cooking or slashing
processes. Slashing operations are not controlled with air pollution control equipment. This was
confirmed by state and federal agency representatives at a PMACT meeting.6 This fact was also
confirmed by plant visits and information compiled by EPA and shared with stakeholders for
review and comment during the PMACT process.
Based on information submitted by the American Textile Manufacturers Institute (ATMI) on
September 17, 1999, it was demonstrated that the majority of the domestic textile market, in
1998, was using PVA for slashing with less than 1 percent by weight, methanol in the PVA "as
purchased." Methanol is a contaminant in the PVA that is a residual material from the
manufacture of the PVA. The typical PVA sizing compound previously contained from 4 to 10
percent methanol. As a result of efforts by the suppliers, the amount of methanol contained in
the PVA can be reduced from the four percent to ten percent in previous years to less than one
percent. Therefore, the methanol content of size "as applied" is below one percent. The ATMI
submittal included letters from suppliers representing approximately 74 percent of the domestic
market for PVA. These letters indicated that the "less than 1 percent methanol" is readily
available and these suppliers are now changing their production to supply the lower HAP
material. These letters provide detailed information from the PVA suppliers, and are located in
the Confidential Business Information files at EPA.7 Information collected from the world wide
web on two domestic suppliers of PVA confirms that PVA with "less than 1 per cent methanol"
is available from suppliers.8
The basis for the MACT Floor for the slashing subcategory was demonstrated to be the use of
5Preliminarv Industry Characterization: Fabric Printing. Coating, and Dyeing. Office of
Air Quality Planning and Standards, U. S. EPA, September 1998. Docket No. A-97-51.
6Memorandum from S. York to Paul Almodovar. Docket No. A-97-51. Initial Regulatory
Subgroup PMACT Meeting for Fabric Printing, Coating and Dyeing. January 8, 1998.
Confidential Business Information Files, OAQPS, U. S. EPA, RTP, North Carolina.
8Memorandum from S. L. Turner to Docket No. A-97-51 regarding methanol content in
slashing PVA compounds, September 27, 2000.
3-2
-------
low HAP PVA containing less than 1 percent HAP, by weight, "as purchased". Because this is
the best information available and because of the availability of low HAP PVA and a large
percentage of the operating facilities using the low HAP material in 1998, this establishes the
floor for slashing at a PVA HAP content limit of less than 1 percent, by weight, "as purchased".
For the purpose of determining the mass fraction of organic HAP in a slashing material, each
organic HAP that is not an OSHA-defmed carcinogen as specified in 29 CFR 1910.1200(d)(4)
that is measured to be present at less than 1 percent, counted as zero. Therefore, the floor for
slashing is zero organic HAP. Since the slashing is performed without the benefit of air pollution
control equipment, and the distribution of emissions is between mixing, application, and drying is
unknown, the pollution prevention option of zero HAP in the PVA "as purchased" is the preferred
limit. Other synthetic organic sizing compounds in use also contain HAP, but the HAP content of
these sizing compounds is well below 1 percent. Therefore, the emission rate limit based on the
use of slashing materials with zero organic HAP for all organic HAP compounds is the average
being achieved by all existing affected sources with slashing operations.
Because PVA sizing is available with zero organic HAP, and this represents the "best performing"
and achievable emission rate for this subcategory, the new and reconstructed source MACT floor
also is the pollution prevention option of zero organic HAP in the sizing material "as purchased".
The MACT floors for existing sources and new or reconstructed sources in the slashing
subcategory are based on the best information available. The floors represent pollution prevention
options yielding the "best performing" and achievable emission rates for existing and new or
reconstructed sources in the slashing subcategory. There is no "above the floor" technology that
could achieve a lower organic HAP content in materials "as purchased" than zero percent.
3-3
-------
MEMORANDUM
TO: Vinson Hellwig, EPA/OAQPS/ESD/CCPG
FROM: Alton Peters, Jim Turner, and Steve York, RTI
DATE: October 12, 2000
SUBJECT: Coating Model Plants
The purpose of this memorandum is to present coating model plants for the printing,
coating, and dyeing of fabrics and other textiles source category. Each model plant is a
representation of the drying/curing operations in a coating facility. The model plants will be used
to estimate add-on control device control costs and resource requirements resulting from
compliance with regulatory options. Emission control systems needed to comply with the
proposed MACT standard also include coating rooms (permanent total enclosures) to capture
fugitive HAP emissions from coating application stations. Coating room specifications are
presented in the October 12, 2001 memorandum entitled Compliance Costs for Coating Model
Plants.
The coating MACT database ' consists of twenty-one facilities of which seventeen are
non-CBI. Process, emissions, and control information is available from responses to survey
questionnaires. There is sufficient process information available from eleven of the twenty-one
facilities to provide a basis for the coating model plants.
The coatings applied by facilities in the coating MACT database can be classified as
solvent-borne and water-borne, with the vast majority of the coatings applied being solvent borne.
Most of the facilities in the MACT database apply solvent-borne coatings with either urethane or
rubber polymer resins. Some facilities in the MACT database using mostly urethane coatings
reported a small amount of vinyl coatings being used on the same lines as the urethane coatings.
This vinyl coating use represents a very small proportion of the coatings used relative to urethane
coatings; therefore, the model plants used for urethane coatings are sufficiently representative of
the plants using vinyl coatings.
Mass of coating solids applied annually could be calculated from coating materials usage
data and correlates well with the production of coated fabric. Therefore, mass of coatings solids
applied annually was determined to be the best parameter in the data base to serve as the basis for
the size of the coating facility. Chart 4-1 presents a plot of the mass of coatings solids applied per
facility in the MACT database for which sufficient non-CBI coatings materials data were
available. Interjection in this process of plants claiming coatings materials usage CBI was
evaluated, but this did not significantly change the distribution of solids used per year. Therefore,
only non-CBI data were used to specify model plants.
4-1
-------
Chart 4-1. Facility-Wide Lbs. Coating Solids Used Per MACT Database Facility
3SOOOO 00 -
300000 00
0)
(A
= 2^0000 00
tn
o>
!\ ?00000 00 -
in
_J
200,000, Lta
' "•' --"'^ '-'-.-,"'- '- V's .''
"•";' -•-.-. •
FC01 FC»5 FC02
Total Total Total
4-2
-------
The information in Chart 4-1 was used to define three different sizes of model plants as
follow: plants applying less than 50,000 pounds of solids per year, plants applying between
50,000 and 200,000 pounds of solids per year, and plants applying greater than 200,000 pounds of
solids per year.
The facilities applying less than 50,000 pounds of solids per year included facilities
applying only urethane coatings and facilities applying only rubber coatings. Hence, two model
plants were specified for this size category. Similarly, all of the facilities applying between
50,000 and 200,000 pounds of solids per year were using only rubber coatings and all of the
facilities applying greater than 200,000 pounds per year were using only urethane coatings.
Consequently, the following four model plants were specified:
• Model Plant No. 1, less than 50,000 pounds of solids applied per year in rubber
coatings
• Model Plant No. 2, less than 50,000 pounds of solids applied per year in urethane
coatings
• Model Plant No. 3, between 50,000 and 200,000 pounds of solids applied per year
in rubber coatings
• Model Plant No. 4, more than 200,000 pounds of solids applied per year in
urethane coatings.
Tables 4-1 through 4-4 present the model plant parameters. The basis for each model
plant parameter is presented in the following paragraphs.
Since there was no information in the coating MACT database on operating time, two
operating schedules were assumed; 2,000 hours per year (8 hours per day, 5 days per week, 50
weeks per year) for the small model plants and 4,000 hours per year (16 hours per day, 5 days per
week, 50 weeks per year) for the medium and large model plants. These operating schedules were
based on the operating schedules for model plants specified in the background information
document2 supporting the NSPS for polymeric coating of supporting substrates (hereafter referred
to as the fabric coating NSPS). The annual coating time was also based on the model plants
specified in development of the fabric coating NSPS.
As has already been described, the annual pounds of solids applied was calculated from
information in the MACT database for each facility with sufficient non-CBI information. For
each model plant, average values across the facilities in the MACT database in that size and
coating category were calculated for the annual pounds of solids applied.
Similarly to the calculation of annual pounds of solids applied, for each facility in the
MACT database with sufficient coatings materials information, the average coating composition
was calculated in terms of weight percent HAP, solids, and non-HAP VOC. The HAP were
speciated; only total VOC information for each coating material was collected. None of the
facilities serving as the basis for the model plants reported water in coating materials. Regarding
the HAP speciation, toluene was the predominant organic solvent reported for solvent-borne
4-3
-------
Table 4-1. Model Plant Parameters for Model Plant No. 1
Annual operating time:
Annual coating timea:
Annual pounds of solids applied:
Coating:
Ovens:
2000 hours
1000 hours
13,410 pounds
Solvent-borne rubber coating, 87% HAP (toluene) by
weight; 13% solids by weight
Number of ovens
Maximum solvent concentration
Quantity of toluene controlled
Solvent capacity
Air flow
1
25% LEL
89.7 Ib/hr
12.4 gallons/hr
2234 ACFM
Inlet temperature to control device b 120 °F
Annual coating time is estimated to be 50% of annual operating hours.
Estimated as 20 °F less than the average exhaust temperature from the oven of 140 °F that
was calculated for facilities from the MACT database in this size and coating category.
4-4
-------
Table 4-2. Model Plant Parameters for Model Plant No. 2
Annual operating time:
Annual coating timea:
Annual pounds of solids applied:
Coating:
2000 hours
1000 hours
10,775 pounds
Solvent-borne urethane coating, 51% HAP (64/36 ratio of
DMF/toluene) by weight; 29% solids by weight; 20% non-
HAP VOC by weight
Ovens:
Number of ovens
Maximum solvent concentration
Quantity of tolueneb controlled
Quantity of DMF15 controlled
Solvent capacity
Air flow
1
25% LEL
9.5 Ib/hr
17 Ib/hr
3.4 gallons/hr
8570 ACFM
Inlet temperature to control device c 300 °F
Annual coating time is estimated to be 50% of annual operating hours.
Includes VOCs of unknown composition.
Estimated as 20 °F less than the average exhaust temperature from the oven of 320 °F that
was calculated for facilities from the MACT database in this size and coating category.
4-5
-------
Table 4-3. Model Plant Parameters for Model Plant No. 3
Annual operating time:
Annual coating time":
Annual pounds of solids applied:
Coating:
4000 hours
2000 hours
136,375 pounds
Solvent-borne rubber coating, 81% HAP (toluene) by
weight; 19% solids by weight
Ovens:
Number of ovens
Maximum solvent concentration
Quantity of toluene controlled
Solvent capacity
Air flow
25% LEL
291 Ib/hr
40 gallons/hr
8465 ACFM
Inlet temperature to control device 242 °F
Annual coating time is estimated to be 50% of annual operating hours.
4-6
-------
Table 4-4. Model Plant Parameters for Model Plant No. 4
Annual operating time:
Annual coating time":
Annual pounds of solids applied:
Coating:
Ovens:
4000 hours
2000 hours
285,900 pounds
Solvent-borne urethane coating, 70% HAP (64/36 ratio of
DMF/toluene) by weight; 24% solids by weight; 6% non-
HAP VOC by weight
Number of ovens
Maximum solvent concentration
Quantity of tolueneb controlled
Quantity of OMF1" controlled
Solvent capacity
Air flow
25% LEL
163 Ib/hr
290 Ib/hr
59 gallons/hr
14,341 ACFM
Inlet temperature to control device 228 °F
Annual coating time is estimated to be 50% of annual operating hours.
Includes VOCs of unknown composition.
4-7
-------
rubber coatings. Urethane coatings were reported to contain predominately toluene and N,N-
dimethylformamide (DMF) with a small amount of non-HAP VOC. The model plant coating
compositions represent average values across facilities in the MACT database in that size and
coating category. It should be noted that the types of solvent-borne coatings and the coating
compositions are consistent with the coatings specified for the model plants developed for the
fabric coating NSPS.
The number of ovens per facility represents the average across facilities in the MACT
database in that size and coating category.
Fire insurance regulations require that combustible gases in air not be at concentrations
greater than 25 percent of the lower explosive limit (LEL)3. Exceptions can be made up to 50
percent LEL, but only with continuous monitoring of the combustible content. Gas flow rates for
modeling are based on maintaining combustible concentration at or below 25 percent LEL.
Dilution air is commonly added to the gas stream and was required for Models 1 and 3.
The LEL for toluene is 1.27 percent. Twenty-five percent of the LEL is 0.3175 percent, or
3175 ppmv. As an example of the calculation of the air flow needed to maintain combustible gas
concentrations below 25 percent of the LEL, for Model 3, 6,000 acfm is the average air flow
calculated from the MACT database for the facilities in the size and coating category represented
by Model 3. The concentration of HAP, which represents all of the combustible material in the
air stream, can be estimated from the quantity of solids applied annually, the concentration of
solids in the coating mix, and the time over which the coating mix is applied. The parameters for
Model Plant 3 are used in the equations below:
„ . fIJAD 136,375 Ib solidsly 0.81 fraction as HAP
Quantity of HAP = x
0.19 fraction of solids 2,000 coatinghours
(1)
x 453'6 8/lb = 1,431.2 g mols HAPIh
92.13 glg-mol
The minimum quantity of gas for 25 percent of the LEL can be found by dividing the
quantity of HAP by 0.3175.
_ , _ , 1,431.2 g mols HAP 1 h 24.0 Llg mol
Total gas flow, scfin -— x x
0.003175 60 min 28.32 Lift3 (2)
= 6,366.9 scfin
Converted to acfm, the minimum gas flow rate becomes:
4-8
-------
6366.9 scfm x (46° + 242) °F = g 465 ac^ ,-
(460 + 68) °F (
This value was used for the gas flow for Model 3 in place of the gas flow calculated from
the MACT database of 6,000 scfm. Similar methodology was used for the other model plants.
Models 2 and 4 had sufficiently low concentrations based on MACT database values that no
dilution air was required.
The average temperature entering the control device was calculated from data in the
MACT database for Models 3 and 4. No such data were available for use with Models 1 and 2.
However, the database did provide temperatures at the exhaust from the ovens. It was assumed
that temperature losses of 20 °F occurred between the oven exhaust and the control device inlet.
This value is consistent with model plants specified for the fabric coating NSPS.
The quantity of combustible material entering the control device (HAP and VOC) is
estimated as in Equation 1. As shown there, 1,431.3 Ib mols of HAP (toluene)/h (or 1,431.3 x
92.13 = 131,866 Ib/h) enter the control device in Model 3. Model 1 is treated similarly. No
VOCs are present in the gas stream for either Model 1 or Model 3. For Models 2 and 4, DMF
quantities are also estimated from the quantity of HAP and the ratio of toluene to DMF suggested
by the MACT database. Because the quantities of VOCs are relatively small and their
constituents are not known, the VOCs are treated as additional quantities of HAPs in the same
ratio as found for the toluene and DMF.
Solvent capacity is found from the quantity of HAP leaving the oven and the room
temperature density of the liquid HAP. For example, in Model 3:
_„„ _ Ib 1 gal .„ gal
290.7 — x — - 40 -— (A)
h 7.26 Ib h ^ '
References
1. U.S. Environmental Protection Agency. Fabric Printing, Coating, and Dyeing NESHAP.
ICR Responses. Office of Air Quality Planning and Standards. Research Triangle Park,
NC. Responses received September 1998 - October 1998.
2. Polymeric Coating of Supporting Substrates - Background Information for Proposed
Standards. Office of Air Quality Planning and Standards, U. S. EPA. EPA-450/3-85-
022a, April 1987. pp. 6-5 thru 6-10.
3. U. S. Environmental Protection Agency. Control Technologies for Hazardous Air
Pollutants Handbook. EPA/625/6-91/014. Office of Research and Development.
Washington, DC. June 1991. p. 4-3.
4-9
-------
-------
MEMORANDUM
TO: Vinson Hellwig, EPA/OAQPS/ESD/CCPG
FROM: Steve York and Alton Peters, RTI
DATE: January 7, 2002
SUBJECT: Summary of Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP
Baseline Organic HAP Emissions and Emission Reductions
Baseline organic HAP emissions data and calculations of emission reductions for coating
and printing, dyeing, finishing, and slashing operations are presented in the following paragraphs.
Though dyeing and finishing constitute a subcategory of the printing, coating, and dyeing of
fabrics and other textiles source category, the detailed estimates of baseline HAP emissions and
emission reductions are broken out in the text because the emission reductions are based on
information from separate surveys of dyeing and finishing facilities. Also attached is a table
summarizing the baseline organic HAP emissions and emission reductions by subcategory.
Coating and Printing Baseline Organic HAP Emissions and Emission Reduction
The baseline organic HAP emissions for coating were derived from 1997 Toxics Release
Inventory (TRI) data. The Standard Industrial Classification (SIC) codes used were as follows:
2262 - Finishing Plants, Synthetics
2269 - Finishing Plants, NEC
2284 - Thread Mills
2295 - Coated Fabrics, Not Rubberized
2298 - Cordage and Twine
• 3052 - Rubber and Plastics Hose and Belting
• 3069 - Fabricated Rubber Products, Not Elsewhere Classified
Baseline organic HAP emissions for printing were determined from data collected in the original
ATMIMACT survey1.
Baseline organic HAP emissions for coating were calculated to be 5537 tons per year and
for printing were calculated to be 34 tons per year, yielding a total of 5571 tons of organic HAP
emissions per year for the coating and printing subcategory. Of the 5571 tons of organic HAP
emissions, 560 tons were determined to be emitted by area sources that would not be required to
reduce organic HAP emissions to comply with the NESHAP. Of the 5,011 tons of organic HAP
emissions from major sources, 214 tons were reported to be methylene chloride emissions.
5-1
-------
Each facility in the coating MACT database 2 was examined to determine if it would be in
compliance with the proposed OCE limit or the equivalent emission rate limit based on MACT
database capture and control efficiency data and coatings use data reported in response to the
coating ICR. Similarly, information collected as described in the memorandum at page 9-1 of this
document regarding coating major facilities owned by small businesses was evaluated to
determine which facilities owned by small businesses would be required to take measures to
reduce HAP emissions to comply with the proposed emission limits. Emission reductions were
calculated for each coating MACT database facility and each major facility owned by a small
business that was determined to be required to take measures to reduce emissions to comply with
either the OCE limit or the equivalent emission rate limit. The total emission reduction for the
coating MACT database facilities and major facilities owned by small businesses was calculated
to be 62 percent.
Methylene chloride emissions were assumed to be uncontrolled, since methylene chloride
is not a VOC, and therefore, has not been required to be controlled under existing VOC
regulations. Consequently, the emission reduction calculated for methylene chloride emissions
would be 97 percent, i.e., the proposed OCE limit for existing sources of HAP emissions.
Dyeing Baseline Organic HAP Emissions and Emission Reduction
The baseline organic HAP emissions for dyeing were determined from data collected in
the original ATMI MACT survey '. Baseline organic HAP emissions were calculated to be 384
tons per year.
The emission reduction was calculated from the ATMI survey of dyeing facilities 3 as the
reduction from the average HAP content in dyeing materials as purchased for the entire dyeing
database of 12.37 percent to the HAP content in dyeing materials as purchased for the dyeing floor
of 1.58 percent, yielding a reduction of 87 percent.
Finishing Baseline Organic HAP Emissions and Emission Reduction
The baseline organic HAP emissions for finishing were determined from data collected in
the original ATMI MACT survey '. Baseline organic HAP emissions were calculated to be 517
tons per year.
The emission reduction was calculated from the ATMI survey of finishing facilities 4 as
the reduction from the average HAP content in finishing materials as purchased for the entire
finishing database of 4.9 percent to the HAP content in finishing materials as purchased for the
finishing floor of 0.03 percent. For the purpose of estimating the emission reduction, a floor of 1
percent was assumed, based on the Occupational Safety and Health Administration (OSHA)
Material Safety Data Sheet (MSDS) minimum reportable quantity of ingredients. For a non-
carcinogen, a mass fraction of less than 1 percent is not quantified further, but reported as < 1.
Using the 1 percent floor yields a reduction of 80 percent.
5-2
-------
Slashing Baseline Organic HAP Emissions and Emission Reduction
The baseline organic HAP emissions for slashing were determined from data collected in
the original ATMIMACT survey '. Baseline organic HAP emissions for slashing were calculated
to be 348 tons per year.
The emission reduction was calculated to be 50 percent, representing a reduction in weight
percent methanol content in PVA size from 2 percent to the slashing floor of 1 percent. The 2
percent baseline weight percent methanol content in PVA size is based on information provided
by ATMI and presented for review and comment to stakeholders in a PMACT briefing package.
Table 5-1 summarizes the baseline organic HAP emissions and the emission reductions for
the coating and printing, slashing, and dyeing and finishing subcategories and for the printing,
coating, and dyeing of fabrics and other textiles source category.
References
1. Memorandum, S. York and A. Sharma, RTI, to P. Almodovar, EPA/OAQPS/ESD/CCPG.
October 31,1997. Summary of meeting at which ATMI presented the results of the ATMI
MACT survey to EPA.
2. U.S. Environmental Protection Agency. Fabric Printing, Coating, and Dyeing NESHAP.
ICR Responses. Office of Air Quality Planning and Standards. Research Triangle Park,
NC. Responses received September 1998 - October 1998.
3. Letter, J. Fleming, ATMI, to G. V. Hellwig, EPA: OAQPS: CCPG, July 27, 2000. ATMI
MACT Development Support and Data Submission (textile dyeing).
4. Letter, J. Fleming, ATMI, to G. V. Hellwig, EPA: OAQPS: CCPG, November 2, 2000.
ATMI MACT Development Support (textile finishing).
5. Memorandum and attachments, M. Malkin and S. York, RTI to P. Almodovar,
EPA/OAQPS/ESD/CCPG. December 15, 1997. Summary of September 4, 1997 P-
MACT meeting, Attachment 3, p. 6 of 18.
5-3
-------
Table 5-1. Summary of Printing, Coating, and Dyeing of Fabrics and Other Textiles Source
Category Baseline Organic HAP Emissions and Emission Reductions
Subcategory
Coating and Printing
Dyeing and
Finishing
Slashing
Source Category
Nationwide Total
Emissions Before
NESHAP(tpy)
557 11
9013
3483
6820
Emissions After
NESHAP(tpy)
2389
153
174
2716
Emission
Reduction
(tpy)
3182
748
174
4104
Percent
Reduction
(%)
572
834
505
60
1 TRI data for 1997 and printing data from Reference 1.
2 Based on estimated emission reduction of 62 percent required for major sources in the coating
MACT database (Reference 2) and major sources owned by small businesses (see memorandum
at page 9-1 of this document) to comply with the proposed emission limits applied to total organic
HAP emissions from major sources (with the exception of 214 tons of methylene chloride
emissions) calculated for the coating and printing subcategory. The methylene chloride emissions
were assumed to be uncontrolled and would be reduced 97 percent by the proposed OCE limit.
3 Reference 1.
4 Based on detailed ATMI surveys of dyeing and finishing facilities (References 3 and 4).
3 Reference 5.
5-4
-------
MEMORANDUM
TO: Vinson Hellwig, EPA/OAQPS/ESD/CCPG
FROM: Steve York, Jim Turner and Jeff Coburn, RTI
DATE: January 7, 2002
SUBJECT: Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP Nationwide
Energy and Secondary Environmental Impacts
The purpose of this memorandum is to present estimates of the nationwide energy and
secondary environmental impacts resulting from compliance with the proposed printing, coating,
and dyeing of fabrics and other textiles NESHAP. The energy and secondary environmental
impacts will result from the installation of new and upgrade of existing add-on controls by
facilities in the coating and printing subcategory. Model plants and the criteria used to choose
them are described in the October 12, 2000 memorandum entitled Coating Model Plants (see page
4-1 of this document). The assignment of model plants to facilities in the coating MACT database
for the purpose of estimating impacts is described in the January 8, 2002 memorandum entitled
Printing, Coating, and Dyeing of Fabrics and Other Textiles Nationwide Compliance Costs (see
page 10-1 of this document). Similarly, the assignment of model plants to coating major facilities
owned by small businesses is described in the December 20, 2001 memorandum entitled
Summary of Evaluation of Estimated Compliance Costs Incurred by Coating Facilities Owned by
Small Businesses (see page 9-1 of this document).
Energy Impacts
Energy requirements for implementation of the compliance options for coating and
printing facilities include electricity to collect and treat ventilation air, electricity to light
permanent total enclosures and natural gas to provide supplemental fuel for stable operation of
oxidizers and to generate the steam required for carbon regeneration. Table 6-1 presents a
summary of increased coating and printing model plant and nationwide energy requirements
associated with implementation of the compliance options. It should be noted that no incremental
electricity usage is estimated for the upgrade of catalytic oxidizer model plants. This is because
the air flow does not change. Similarly, no incremental energy usage is estimated for the upgrade
of carbon adsorber Models 3 and Model 4. For each model plant, the increased efficiency comes
from the addition of a carbon bed, reducing the cycle time between carbon bed regenerations, and
therefore, reducing the HAP released to the atmosphere from breakthrough. There is no change in
air flow or in the amount of steam used for regeneration, which is a function of the organic HAP
load entering the carbon bed.
6-1
-------
Table 6-1. Summary of Coating and Printing Subcategory Model and
Nationwide Energy Impacts
Model
New Add-on Control Device
Model 1 , carbon adsorber
Model 1, catalytic oxidizer
Model 2, thermal oxidizer
Model 3, carbon adsorber
Upgrade of Add-on Control Device
Model 2, catalytic oxidizer
Model 3, catalytic oxidizer
Model 3, carbon adsorber
Model 4, catalytic oxidizer
Model 4, carbon adsorber
New Coating Room (PTE)
Small
Medium
Large
Total Energy Impacts for Model Plants Except
Methylene Chloride Model Plants
Nationwide Total Energy Impacts Except
Methvlene Chloride Enerev Impacts b
New Add-on Control System for Methylene
Chloride Emissions '
Model 1, carbon adsorber
Model 3, carbon adsorber
Total Methylene Chloride Control Energy Impacts
Nationwide Total Energy Impacts with Methylene
Chloride K.npruv Imnnrts d
Number of
plants *
2
1
2
4
1
2
3
2
1
14
13
29
1
1
Model
incremental
electricity
usage. kWh/v
8,933
11,293
28,857
119.517
0
0
0
0
0
11,200
12,250
12,600
15,742
186,588
Nationwide
total
electricity
usace* k\Vh/v
17,866
11,293
57,714
478,068
0
0
0
0
0
156,800
159,250
365,400
1,246,391
2,567,565
15,742
186,588
202,330
2 769 895
Model
incremental
natural gas
usage, scf/v
418,941
2,360,755
36,332,289
2,714,142
691,592
1,090,910
0
1,723,795
0
0
0
0
418,941
2,714,142
Nationwide
total natural
gas usage,
scf/v
837,882
2,360,755
72,664,578
10,856,568
691,592
2,181,820
0
3,447,590
0
0
0
0
93,040,785
191,664,017
418,941
2,714,142
3,133,083
194797 100
Number of model plants assigned to 14 facilities in the coating MACT database and to 12 coating major facilities owned by small businesses
to estimate the incremental energy requirement of achieving the proposed emission limits with add-on controls.
Nationwide totals for all plants in the coating and printing industry, except plants with methylene chloride emissions, are based on factoring
the total energy usage for model plants except methylene chloride model plants by the ratio of HAP emissions estimated for major HAP
emission sources in the coating and printing subcategory (minus methylene chloride emissions) to the HAP emissions reported by facilities
in the coating MACT database and major facilities owned by small businesses (the ratio is 2.06).
Includes energy usage of add-on control system and coating room.
Sum of nationwide total energy impacts except methylene chloride energy impacts and total methylene chloride control energy impacts.
6-2
-------
Water Impacts
Nationwide water impacts resulting from implementation of the compliance options are
insignificant. Facilities adding carbon adsorber systems will require increased cooling water
usage for the condenser used to recover organic HAP from the regenerated carbon and for the
spray tower specified to cool the gas entering the Model 3 carbon adsorber used to recover
methylene chloride. The cooling water for the condenser does not contact the HAP-laden stream
and is assumed to be recycled. Similarly, only enough cooling water should be used in the spray
tower to cool, but not saturate, the gas entering the Model 3 carbon adsorber, so the cooling water
is assumed not to result in wastewater. Nationwide cooling water usage is estimated to be
70,292,992 gallons per year.
There is a small increase in water usage for steam to regenerate carbon. The steam used
for regeneration will yield water requiring wastewater treatment. Nationwide total wastewater
generation is estimated to be 3,766,369 gallons per year.
Solid Waste Impacts
Facilities using existing catalytic oxidizers to comply with the emission limits probably
will be required to install larger volumes of catalysts and to replace the catalysts more frequently
than current replacement cycles to maintain high performance levels, resulting in a small increase
in solid waste generation. Similarly, facilities that currently do not operate emission control
systems and that install catalytic oxidizers to comply with the emission limits will result in an
increase in solid waste generation. Sometimes the spent catalyst will be regenerated by the
manufacturer for reuse. Activated carbon used in carbon adsorbers is returned to the manufacturer
at the end of its useful life and converted to other salable products. Little solid waste impact is
expected from this source.
6-3
-------
MEMORANDUM
TO: Vinson Hellwig, EPA/OAQPS/ESD/CCPG
FROM: Steve York, RTI
DATE: October 12, 2001
SUBJECT: Compliance Costs for Coating Model Plants
The purpose of this memorandum is to present compliance costs for the coating model
plants for the printing, coating, and dyeing of fabrics and other textiles source category. Model
plant specifications used in estimating compliance costs are summarized in Table 7-1. Emission
control systems needed to comply include coating rooms (permanent total enclosures) to capture
fugitive HAP emissions from coating application stations and either oxidizers with 97 percent
destruction efficiencies or carbon adsorbers with 97 percent recovery efficiencies.
PERMANENT TOTAL ENCLOSURE COSTS
Table 7-2 presents a summary of permanent total enclosure (PTE) costs. As shown in
Table 7-2, PTEs are costed in three sizes: 8,000 ft3; 13,000 ft3; and 18,000 ft3. Floor areas for the
three enclosures are taken as 800 ft2, 875 ft2, and 900 ft2, respectively, based on typical coating
application station sizes for the model plants. To estimate compliance costs for a coating line
needing to upgrade capture efficiency, the costs of a small PTE are applied to Model Plants 1 and
2, the costs of a medium PTE to Model Plant 3, and the costs of a large PTE to Model Plant 4.
Each PTE is assumed to have two swing doors and four windows. Costing on a square-
foot basis plus doors and windows, is taken from Reference 1. The structure is assumed to be
constructed of steel. Auxiliary costs that contribute to the purchased equipment cost (PEC) are
assumed to add 50 percent to the purchase price. Total capital investment (TCI) is taken as 1.6
times the PEC. Annual costs are charged for maintenance ($6/ft2 y) and electricity for lighting (14
kWh/ft2 y). Indirect annual costs are based on typical values in the OAQPS Control Cost Manual
2 (Manual) , i.e., 60 percent labor and materials overhead, other indirect costs of 4 percent of TCI,
and capital recovery based on 7 percent interest and a 15-year life for the enclosure.
In estimating the costs of a PTE, it has been assumed that existing process exhaust airflow
will be adequate to satisfy the EPA Method 204 criteria and to provide for worker safety and
comfort. This assumption is based on experience cited by several engineering contractors 3'4'5 that
install PTEs. For example, Pacific Environmental Services reported that of more than 100 PTE
designs completed, none has required an increase in the size of the air pollution control device in
order to maintain worker comfort.
7-1
-------
Table 7-1. Model Plant Specifications Used for Compliance Costing
Model Plant 123
Annual operating time (hr)
Annual coating time a (hr)
Solids applied annually (Ibs)
Coating type
Coating formulation b :
Weight percent HAP
Weight percent solids
Weight percent non-HAP
VOC
Ovens c :
Number
Maximum solvent
concentration (% LEL)
Solvent capacity (gal/hr)
Air flow (ACFM)
Inlet temperature to control
device (°F)
2000
1000
13,410
Rubber
87
13
0
1
25
12.4
2234
120 d
2000
1000
10,775
Urethane
51
29
20
1
25
3.4
8570
300 d
4000
2000
136,375
Rubber
81
19
0
2
25
40
8465
242
4000
2000
285,900
Urethane
70
24
6
4
25
59
14,341
228
HAP = hazardous air pollutant, LEL = lower explosive limit.
a Annual coating time is estimated to be 50 percent of annual operating hours.
b Solvent-borne rubber coating contains toluene as the solvent: solvent-borne urethane coating
contains dimethyl formamide and toluene in ratio of 35 to 20 by weight as the solvent.
c Parameters are given on a per facility basis; emissions from multiple ovens are routed to one
add-on control device.
d Estimated as 20 °F less than the average exhaust temperature from the oven that was
calculated for the facilities from the coating MACT database in the size and coating category.
7-2
-------
Table 7-2. Summary of Coating Room Costs
Model
Floor area, ft 2
Cost/ft 2, $
Cost, $
Swing doors (2), $
Windows (4), $
Sum, $
Auxiliaries (at 50 %), $
Purchased equipment cost (PEC), $
Total capital investment (TCI, 1.6 x PEC), $
Maintenance (6$/ft 2 y), $/y
Maintenance supervision (1 5 % of maintenance), $/y
Materials (50 % of maintenance labor), $/y
Electricity (lighting, 14 kWh/ft a y and $.06/kWh), $/y
Direct costs, $/y
Labor/materials overhead (60 % of labor and materials), $/y
Other indirect costs (4 % of TCI), $/y
Capital recovery (7 % interest rate, 15-year life), $/y
Indirect costs, $/y
Total annual costs, TAG, $/y
Small (R.OOOft3}
800
15
12,000
5,000
800
17,800
8,900
26,700
42,720
4,800
720
2,400
672
8,592
4,752
1,709
4,691
11,151
19,743
Medium (1 3.000 ft3}
875
18
15,313
5,000
800
21,113
10,556
31,669
50,670
5,250
788
2,625
735
9,398
5,198
2,027
5,564
12,788
22,186
Laraef 18.000 ft3}
900
20
18,000
5,000
800
23,800
11,900
35,700
57,120
5,400
810
2,700
756
9,666
5,346
2,285
6,272
13,903
23,569
Note: Costs for enclosure, doors, and windows based on cost factors presented in Reference 1.
7-3
-------
OXIDIZER COSTS
For each model plant, costs are estimated for installing a 97-percent efficient thermal or
catalytic oxidizer and for upgrading an existing catalytic oxidizer from 92 to 97 percent
destruction efficiency. Every thermal incinerator in the coating MACT database 6 is reported to
have a destruction efficiency of at least 96.3 percent (the average is greater that 98 percent),
therefore, upgrade costs are not needed for thermal oxidizers. Table 7-3 presents a summary of
the new oxidizer installation costs; Table 7-4 presents a summary of the catalytic oxidizer upgrade
costs. The costs are estimated based on the Manual. Costs estimated from the Manual are
expected to be within about 30 percent of the cost a buyer might pay for the equipment being
costed. However, much larger deviations can be found if the input parameters for the model differ
from values found in practice.
To estimate incremental costs of upgrading existing catalytic oxidizers, costs of baseline
catalytic oxidizers are subtracted from the costs of upgraded units. The cost of a new oxidizer
system includes the costs of ductwork, butterfly dampers, fans, motors, and stacks. Costs are
estimated and are summarized in Tables 7-3 and 7-4 in three areas: TCI, total annual cost (TAG),
and operation and maintenance costs (O&M). The TCI includes purchased equipment costs
(incinerator and auxiliary equipment, instrumentation, sales tax, and freight), direct installation
costs (foundation and supports, handling and erection, electrical, piping, insulation for duct work,
and painting where not included in auxiliary costs), and indirect installation costs (engineering,
construction or field expenses, contractor fees, start-up, performance test, and contingencies). The
TAC includes indirect annual costs (overhead, administrative charges, property taxes, insurance,
and capital recovery) and direct annual costs (O&M). The O&M costs are made up of electricity,
natural gas, operating labor, and maintenance labor and materials.
The Manual is designed so that the user supplies information for a variety of model
parameters. For oxidizers, some of these parameters are gas flow rate, gas temperatures at the
inlet and outlet, HAP concentration, heats of combustion and heat capacities for the HAPs, and
amount of heat recovery for oxidizers so equipped. Some of the model parameters come directly
from the model plants, e.g., values for gas flow, temperature, annual hours of operation, and
quantity of solvent are consistent with each of the model plants. For other model parameters,
assumptions are required, as are explained in the following paragraphs.
Solvents assumed to be in the oxidizer inlet for Model Plants 2 and 4 are approximately 64
percent N,N-dimethylformamide (DMF) and 36 percent toluene. The solvent assumed to be in the
oxidizer inlet for Model Plants 1 and 3 is toluene. Heats of combustion for the two compounds
are taken as 2,161 Btu/scf for DMF and 4,522 Btu/scf for toluene. Auxiliary fuel is assumed to be
natural gas with a heat of combustion of 21,502 Btu/lb. Temperature dependent chemical
property data (e.g., vapor pressures and heat capacities) were estimated from correlations and data
presented in Perry's Chemical Engineers' Handbook, 7th Edition.
7-4
-------
Table 7-3. Summary of New Oxidizer Costs for Coating Model Plants
Model Plant
Model 1, thermal
Model 1, catalytic
Model 2, thermal
Model 2, catalytic
Model 3, thermal
Model 3, catalytic
Model 4, thermal
Model 4. catalytic
Total
capital
investment,
$
434,562
Total
annual
cost, $/v
130,972
300,140 90,888
576,551
544,819
588,505
569,135
699,230
790.010
241,585
149,905
303,215
204,066
348,546
291.709
O&M
cost, $/v
58,469
23,361
^147,663
41,706
199,946
84,371
228,601
128.399
Assumptions: Units operate at 1,420 °F (thermal) or 1,200 °F (catalytic),
have 70 % heat recovery and have a retrofit factor of 1.4.
Efficiency is 97 percent for all oxidizers, which requires 1.5 x operating
labor cost and double the maintenance of existing units.
For all cases, costs include ductwork, dampers, fan, motor, and stack.
All costs are in 1997 $.
Total capital investment is annualized at 7 percent interest for 15 years.
7-5
-------
Table 7-4. Summary of Catalytic Oxidizer Upgrade Costs for Coating Model Plants
Model
Baseline
Model 1, catalytic
Model 2, catalytic
Model 3, catalytic
Model 4. catalytic
Total
capital
investment,
$
219,908
397,790
413,629
560.341
Total
annual
cost, $/y
64,913
112,214
152,020
216.723
O&M
cost, $/v
17682
36,028
69,175
106.007
Capital cost
above
baseline^ $
Annual
cost above
baseline,
$/v
O&M cost
above
baseline,
$/v
Assumptions: Baseline units are catalytic oxidizers operating at 830 °F.
Efficiency is 91 percent. Heat recovery is 50 % and retrofit factor is 1.2.
Upgrade of Baseline Unit
Model 1, catalytic
Model 2, catalytic
Model 3, catalytic
Model 4, catalytic
293,755
528,757
549,665
742,659
90,136
148,516
199,933
275,369
23,600
43,233
84,045
128,399
73,847
130,967
136,036
182,319
25,222
36,302
47,914
58,646
5,918
7,205
14,870
22,392
Assumptions: Upgraded units operate at 1,200 °F, have 70 % heat recovery and have a retrofit factor of 1.4.
Efficiency is 97 percent for upgraded oxidizers, which requires 1.5 x operating labor cost and double the
maintenance of existing units.
Baseline and Upgrade Assumptions: Costs exclude ductwork, dampers, fan, motor, and stack.
All costs are in 1997 $.
Total capital investment is annualized at 7 percent interest for 15 years.
7-6
-------
For baseline catalytic oxidizers, oxidizer efficiency is assumed to be 91 percent and outlet
temperature is assumed to be 830 °F, based on information in the coating MACT database
reported by facilities with catalytic oxidizers. Heat recovery is assumed to be 50 percent. Retrofit
costs are assumed to add 20 percent to the TCI.
Costs for upgraded and new oxidizers are based on an efficiency of 97 percent for all units.
Outlet temperatures are assumed to be 1,420 °F and 1,200 °F for thermal and catalytic units,
respectively. Heat recovery is assumed to be 70 percent. Retrofit costs are assumed to add 40
percent to the TCI, and the need for operating and maintaining the oxidizer system at constant
high efficiency is assumed to require an additional 50 percent in operating labor and double the
maintenance labor and maintenance materials of existing units.
For all cases representing the upgrade of an existing control system, costs exclude
ductwork, butterfly dampers, fans, motors, and stacks. For all cases representing the installation
of a control system in a facility with no existing controls, these auxiliaries are costed using
Chapter 10 of the Manual for ductwork, dampers, and stack. Information in Chapter 4.12 of the
Handbook - Control Technologies for Hazardous Air Pollutants7 is used for costing fans and
motors and also for sizing ductwork. Ductwork is assumed to be cold-rolled, spiral-wound steel
with three inches of insulation. Labor costs are derived from tables provided by the Bureau of
Labor Statistics at its Internet website (http://www.bls.gov/home.htm). All costs are in 1997
dollars.
The Manual provides equipment sizing equations based on simplifying assumptions. The
equations can be altered if the underlying assumptions are changed. One such change is the
assumed system heat loss. Because the waste-gas streams entering the oxidizers are at relatively
high temperatures, heat losses are assumed to be from 35 to 55 percent, depending on inlet
temperature assigned to the model plant being costed. For cases in which the model predicts
auxiliary gas consumption to be less than five percent of total gas, additional auxiliary gas is
provided for flame stabilization.
CARBON ADSORBER COSTS
For each model plant, costs are estimated for installing a 97-percent efficient carbon
adsorber and for upgrading an existing carbon adsorber from 93 to 97 percent destruction
efficiency. Table 7-5 presents a summary of the new adsorber installation costs; Table 7-6
presents a summary of the adsorber upgrade costs. The costs are estimated based on the Manual8.
The cost of a new carbon adsorber system includes the costs of carbon, adsorbers,
condensers, fan, motor, stack, and 25 feet of ductwork and damper per carbon bed (1 bed for
Model Plant No. 1, 2 beds for Model Plant No. 2 , and 3 beds for Model Plant Nos. 3 and 4).
Adsorption systems were designed to accommodate 8 hours of exhaust at the average projected
vent stream concentration over the operating time of the unit. This assumes relatively uniform
application quantities for any given operating day.
7-7
-------
Table 7-5. Summary of New Carbon Adsorber Costs for
Coating Model Plants
Model
Plant
No.
1
2
3
4
TCI.$
104,183
223,521
501,693
1.158,663
TAG w/credits,
$/vr
31,068
58,135
87,350
195.757
TAG w/o credits,
$/vr
39,773
60,694
143,163
283.576
O&M w/credits,
$/vr
16,929
27,662
22,409
51.487
O&M w/o credits,
$/vr
25,634
30,221
78,222
139.306
Assumptions: TCI includes a retrofit factor of 1.2, removal efficiency is 97 percent for all
adsorbers, recovery credits assume a value of $ 0.10/lb of HAP recovered, all costs are in 1997 $,
and TCI is annualized at 7 percent interest for 10 years.
Model
Plant
No.
1
2
3
4
Table 7-6. Summary of Carbon Adsorber
Coating Model Plants
TCI, TAG,
$ $/vr
50,347 9,633
102,822 19,191
159,504 30,492
218.447 42.523
Upgrade Costs for
O&M,
$/vr
2,781
5,166
9,748
15.184
Assumptions: TCI includes a retrofit factor of 1.2, removal efficiency is 97 percent for all
adsorbers, recovery credits are not calculated for the small amount of additional recovered HAP
resulting from the increase in efficiency from 93 to 97 percent, all costs are in 1997 $,TCI is
annualized at 7 percent interest for 10 years, and upgrade systems are based on adding one
additional carbon bed to the adsorber system (one bed half the size of the original for Model Plant
No. 1; one originally-sized bed for Model Plant Nos 2, 3, and 4).
7-8
-------
The costs of the upgraded systems are based on the following assumptions. For Model
Plant 1, a carbon bed one half the size of the original (approximately 90% efficient) bed was
added in series to the system to estimate the upgrade costs of the small model plant. For all other
model plants, one additional (originally sized) carbon bed was added to upgrade the system.
Costs are estimated and are summarized in Tables 7-5 and 7-6 in three areas: TCI, TAG,
and operation and maintenance costs (O&M). The TCI includes purchased equipment costs
(adsorber and auxiliary equipment, instrumentation, sales tax, and freight), direct installation costs
(foundation and supports, handling and erection, electrical, piping, insulation for duct work, and
painting where not included in auxiliary costs), and indirect installation costs (engineering,
construction or field expenses, contractor fees, start-up, performance test, and contingencies). The
TAG includes indirect annual costs (overhead, administrative charges, property taxes, insurance,
and capital recovery) and direct annual costs (O&M). The O&M costs are made up of electricity,
steam, cooling water, carbon replacement, operating labor, and maintenance labor and
materials.
The Manual is designed so that the user supplies information for a variety of model
parameters. For adsorbers, some of these parameters are gas flow rate, inlet gas temperatures,
HAP concentration, and adsorption coefficients for the HAPs. Some of the model parameters
come directly from the model plants, e.g., values for gas flow, temperature, annual hours of
operation, and quantity of solvent are consistent with each of the model plants. For other model
parameters, assumptions are required, as are explained in the following paragraphs.
Solvents assumed to be in the adsorber inlet for Model Plants 2 and 4 are approximately
64 percent DMF and 36 percent toluene. The solvent assumed to be in the adsorber inlet for
Model Plants 1 and 3 is toluene. The equilibrium adsorptive capacity of carbon is dependent on
the specific constituent and the operating temperature and concentrations. The Calgon fifth-order
polynomial equation presented in the Manual was used to estimate the equilibrium adsorptive
capacity of carbon for both toluene and DMF based on inlet conditions. The equilibrium
adsorptive capacities were also calculated at the adsorber outlet concentration (based on 97
percent removal) to assess the phenomena of "tailing" and whether or not the working capacity
assumption (i.e., being equal to 50 percent of the equilibrium adsorptive capacity) was sufficient
to achieve the desired removal efficiency.
For Model Plant No. 2, the exhaust stream concentration is too high for effective carbon
adsorption. Consequently, the exhaust stream was cooled to 212 °F prior to the carbon adsorption
system. This cooling was assumed to be accomplished by radiant cooling. The cost of radiant
cooling ductwork was estimated as 100 ft of normal duct work. An additional 1 inch of water
pressure drop was added to the system's pressure drop to account for the energy required to pull
the exhaust stream through the additional radiant cooling ductwork. The actual flow rate to model
plant 1 at 212 °F is 7,575 acfm.
For the toluene only systems, the 50 percent working capacity appeared sufficient for
Model Plant 1, but the calculated working capacity for Model Plant 3 was reduced by an
7-9
-------
additional factor of 1.2 based on the low equilibrium adsorptive capacity at the design outlet
toluene concentration. The total carbon amounts required for the toluene/DMF systems (Model
Plants 2 and 4) were calculated by assessing the amount of carbon needed for each chemical
independently and adding the 2 quantities together. Because of the low equilibrium adsorptive
capacities at the design outlet concentrations, adjustments to the 50 percent working capacity
assumptions were made. The working capacity for assessing the amount of carbon required for
toluene adsorption was reduced by an additional factor of 1.25 and the working capacity for
assessing the amount of carbon required for DMF adsorption was reduced by a factor of 2 for both
Model Plants 2 and 4.
Retrofit costs were assumed to add 20 percent to the TCI. For new carbon adsorbers, total
annual costs were calculated with and without a recovery credit; to calculate the recovery credit a
value of $ .10/lb was assumed for the recovered HAP. Recovery credits were not calculated for
the upgrade costs because the additional amount of HAP recovered by increasing the recovery
efficiency from 93 to 97 percent is a very small quantity.
METHYLENE CHLORIDE CONTROL COSTS
During the MACT floor data collection effort, information was collected from two
facilities that emit methylene chloride. Because the cost of controlling methylene chloride
emissions will be greater than the cost of controlling other organic HAP emissions, additional cost
analysis has been done for this specific case. The models with methylene chloride emissions
assume the coating with methylene chloride is a single-solvent coating. Model Plants 1 and 3,
with methylene chloride as the coating HAP rather than toluene, were used as the basis for
estimating the costs of installing, operating and maintaining add-on control systems for methylene
chloride emissions.
For each model plant, costs are estimated for installing a 97-percent efficient thermal or
catalytic oxidizer and a 97-percent efficient carbon adsorber. Since methylene chloride has a
higher LEL than the organic HAP specified for compliance costing, the air flow for Model Plant 1
was reduced to 250 acfm at 120 °F. No adjustment was needed to the Model Plant 3 flow rate.
As has been described in previous sections of this memorandum, TCI, TAG, and O&M costs were
estimated, using the Manual.
The costs for controlling methylene chloride emissions with oxidizers includes the costs of
a post-oxidation scrubber system needed to remove the hydrogen chloride gas and neutralize the
scrubber water and additional costs of the auxiliaries (ductwork, butterfly dampers, fans and
stacks) which must be constructed of materials able to withstand the corrosive acid gas. Also, the
heat of combustion of methylene chloride is only 662 BTU/scf, considerably lower than the heat
of combustion of toluene or DMF, therefore, more auxiliary fuel will be required.
With respect to the carbon adsorber costs, carbon's adsorptive capacity for methylene
chloride at the 242 °F inlet temperature specified for Model 3 is very low, necessitating the
installation of a spray tower to cool the gas. In addition, because of carbon's low adsorptive
7-10
-------
capacities at the target outlet concentration, a working capacity of one-third the working capacity
of the inlet was used instead of the one-half that was used in costing carbon systems to control
organic HAP emissions.
Table 7-7 presents the summary of oxidizer costs and Table 7-8 presents the summary of
carbon adsorber costs for controlling methylene chloride emissions. For Model 1, the increase in
TCI associated with controlling methylene chloride emissions ranges from around 21 percent for
thermal oxidation up to 61 percent for carbon adsorption and the increase in TAC ranges from
almost 20 percent for thermal oxidation up to almost 38 percent for carbon adsorption. For Model
3, the increase in TCI associated with controlling methylene chloride emissions ranges from
approximately 30 percent for carbon adsorption up to 38 percent for catalytic oxidation and the
increase in TAC ranges from over 20 percent for thermal oxidation to more than 59 percent for
carbon adsorption.
Table 7-7. Summary of New Oxidizer Costs for Control of Methylene Chloride Emissions
Model Plant
Model 1, thermal
Model 1, catalytic
Model 3, thermal
Model 3. catalytic
Total capital
investment, $
525,552
387,495
805,600
785.529
Total annual cost,
$/v
156,699
116,226
365,223
265.985
O&M
cost, $/y
66,602
31,474
221,135
105.448
Assumptions: Units operate at 1,420 °F (thermal) or 1,200 °F (catalytic),
have 70 % heat recovery and have a retrofit factor of 1.4. Efficiency is 97
percent for all oxidizers, which requires 1.5 x operating labor cost and
double the maintenance of existing units. For all cases, costs include
ductwork, dampers, fan, motor, and stack. All costs are in 1997 $. Total
capital investment is annualized at 7 percent interest for 15 years.
Table 7-8. Summary of New Carbon Adsorber Costs
for Control of Methvlene Chloride Emissions
Model
Plant No.
1
3
TCI.S
167,848
650.061
TAC w/credits,
$/vr
42,734
139.032
TAC w/o credits, $/yr
51,439
194.846
O&M w/credits,
$/vr
20,051
43.956
O&M w/o credits, $/yr
28,756
99.769
Assumptions: TCI includes a retrofit factor of 1.2, removal efficiency is 97 percent for all adsorbers, recovery credits
assume a value of $ 0.10/lb of HAP recovered, all costs are in 1997 $, and TCI is annualized at 7 percent interest for
10 years.
7-11
-------
REFERENCES
1. Lukey, Michael E., P.E. Permanent Total Enclosures Needed in Response to Subpart KK
and Changes in Test Procedures. Paper No. 97-TA4B.05, presented at the Air and Waste
Management Association Annual Meeting & Exhibition. Toronto, Ontario, Canada. June
1997. Table 2.
2. U. S. Environmental Protection Agency. OAQPS Control Cost Manual, Fifth Edition.
EPA-453/B-96-001. February 1996. Chapter 3.
3. Reference 1, page 3 of 4.
4. Turner, Thomas K. Local Capture or Total Enclosure? The Answer is Yes! Paper No.
94-RA111.01, presented at the Air and Waste Management Association Annual Meeting
& Exhibition. Cincinnati, OH. June 1994.
5. Bemi, Dan. "Demonstrating VOC Capture Efficiency Using Permanent Total Enclosure
Technology: Common Practices, Challenges and Rewards." Paper No. 97-TA4B.04,
presented at the Air and Waste Management Association Annual Meeting & Exhibition.
Toronto, Ontario, Canada. June 1997.
6. U.S. Environmental Protection Agency. Fabric Printing, Coating, and Dyeing NESHAP.
ICR Responses. Office of Air Quality Planning and Standards. Research Triangle Park,
NC. Responses received September 1998 - October 1998.
7. U.S. Environmental Protection Agency. Control Technologies for Hazardous Air
Pollutants. EPA/625/6-91/014. Office of Research and Development. Washington, DC.
June 1991. Pages 4-98 thru 4-101.
8. Reference 2, Chapter 4.
7-12
-------
MEMORANDUM
From: Steve York, Research Triangle Institute
To: Printing, Coating, and Dyeing of Fabrics and Other Textiles File
Subject: Incremental Cost of Non-Formaldehyde Permanent Press Finish Versus Permanent
Press Finish with Formaldehyde
Date: August 1,2001
Information was collected from three sources: (1) Cotton Incorporated, (2) BF Goodrich Textile
Performance Chemicals, and (3) Vulcan Performance Chemicals. Cotton Incorporated is a
research and marketing company representing cotton producers and importers. BF Goodrich
Textile Performance Chemicals and Vulcan Performance Chemicals supply specialty chemicals to
the textile industry, including non-formaldehyde or very low formaldehyde permanent press
resins. Information used in estimating incremental costs is summarized in the following
paragraphs and the contact summaries for each information source are attached to this
memorandum.
Mr. John Turner of Cotton Incorporated stated that finishing with a formaldehyde-containing resin
costs from 5 to 15 cents per pound of finished fabric. According to Bill Rarick of Cotton
Incorporated and Mr. Turner, the cost of the cross-linking agent runs between 50 and 65 cents per
pound and there is additional cost for chemical auxiliaries. Mr. Turner stated that BTCA is not
commercially available, but he knows of a manufacturer that will supply BTCA for $2.50 per
pound for a minimum order of 1 million pounds. For smaller quantities, the cost is $13 per
pound. Mr. Turner also stated that resins without formaldehyde do not cross link as well as resins
with formaldehyde.
Ms. Jennifer Grabowski of BF Goodrich Textile Performance Chemicals provided information
about two permanent press resins, one that contains less than 1 percent formaldehyde and the
second that contains less than 100 ppm formaldehyde (below MSDS reportable quantities). The
cost of each depends on the quantity ordered and for the less than 1 percent formaldehyde resin
ranges from $1.06 per pound for an order of 1 to 3 drums down to $0.79 per pound for an order of
50 to 80 drums. The less than 100 ppm formaldehyde resin costs 3 cents per pound more than the
1 percent formaldehyde resin at each size range.
Mr. Jerry Setzer of Vulcan Performance Chemicals provided information regarding the cost of a
non-formaldehyde cross-link system Vulcan has developed and is marketing at a cost of $1.38 per
8-1
-------
pound for the resin and $0.78 per pound for the catalyst. Mr. Setzer claimed that the Vulcan
Performance Chemicals non-formaldehyde cross-link system yields comparable results to a
formaldehyde cross-link system, when cured at the proper temperature, and additionally has less
adverse effect on the strength of the fabric than formaldehyde cross-link systems. Mr. Setzer has
estimated the cost of one pair of twill pants would be about $0.40 to 0.45 more expensive than a
pair of twill pants finished with a formaldehyde-containing permanent press resin.
The table below presents estimates of the increased cost (the cost increment above the baseline
finishing cost of 10 cents per pound of finished fabric, the midpoint of the range of 5 to 15 cents
per pound of finished fabric cited by Mr. Turner of Cotton Incorporated) of a pound of finished
fabric resulting from using a non-formaldehyde or compliant resin instead of a common low-
formaldehyde DMDHEU permanent press resin.
To estimate the increased cost, it was necessary to make certain assumptions. Information from
Cotton Incorporated regarding the cost of finishing with a formaldehyde-containing resin (i.e., 10
cents per pound of finished fabric) was assumed to be the baseline cost for permanent press
finishing. The increased cost of finishing with compliant materials was assumed to be only a
function of the incremental cost in resins and catalysts; i.e., the cost of auxiliary chemicals was
assumed to be the same for formaldehyde and non-formaldehyde systems. In addition, the
baseline cost of formaldehyde-containing resin was assumed to be 57.5 cents, the midpoint of the
range of 50 to 65 cents cited by Mr. Rarick and Mr. Turner for the formaldehyde-containing
cross-linking agent.
Information Source a
Cotton Incorporated
Vulcan Performance
Chemicals
BF Goodrich
Textile Performance
Chemicals
Cost of
Formaldehyde-
Containing Resin
($/lb of resin)
0.58
NA
NA
Cost of Non-
Formaldehyde or
Compliant Resin
($/lb of resin)
2.5
1.38
+$ 0.26/lb catalyst
0.82
Increased Finishing
Cost
($/lb of finished
fabric)
0.33
0.18
0.04
a Attachment 1 is the Cotton Incorporated contact report, Attachment 2 is the Vulcan
Performance Chemicals contact report, and Attachment 3 is the BF Goodrich Textile
Performance Chemicals contact report.
NA = Not applicable.
As can be seen in the table, the incremental cost of finishing with a non-formaldehyde or
compliant resin is estimated to range from 4 cents per pound of finished fabric up to 33 cents per
pound of finished fabric. Regarding the high end of this range, it should be noted that the BTCA
8-2
-------
non-formaldehyde finish that the non-formaldehyde resin costs from Cotton Incorporated is based
on is not currently in commercial production. The estimated incremental cost of 33 cents per
pound that is presented is based on the price per pound of an order for at least one million pounds.
This is probably representative of the incremental cost that would be incurred if BTCA were in
commercial production.
8-3
-------
ATTACHMENT 1
CONTACT REPORT
From: Steve York (919-990-8629 ), Research Triangle Institute project lead for the Printing,
Coating,and Dyeing of Fabrics and Other Textiles NESHAP
Date of Contact: March 8, 2001
Contact: Bill Rarick and John Turner
Company/Agency : Cotton Incorporated
Telephone Number: (919) 678-2220, Bill Rarick 678-2416, John Turner 678-2455
Location; Raleigh, North Carolina
CONTACT SUMMARY:
Mr. Rarick was contacted by telephone to solicit information regarding the cost of permanent press
resins containing formaldehyde versus the cost of permanent press resins without formaldehyde. Mr.
Rarick offered that for a common low formaldehyde DMDHEU permanent press resin the cost runs
about 60 to 65 cents per pound. This could be contrasted with BTCA, a non-formaldehyde finish that
is currently not in commercial production and is of limited availability. Mr. Rarick stated that prices
of BTCA range from possibly as low as $2 up to $6 per pound. Mr. Rarick also offered that as a rule
of thumb, the chemical cost can be multiplied by a factor ranging from 7 to 20 to estimate the affect
on the retail price of a product, such as a pair of pants, made from the finished fabric. Regarding
trying to set general limits on HAP content in finish materials, Mr. Rarick commented that finish
chemistry and the amount used is very complex, depending on the desired properties and the substrate
being finished, e.g., a durable press finish might be used on a knit fabric to provide dimensional
stability. Mr. Rarick transferred me to John Turner for more information on the costs of permanent
press finishes.
Mr. Turner stated that finishing with a formaldehyde-containing resin costs from 5 to 15 cents per
pound of finished fabric. The cost of the cross-linking agent runs between 50 and 60 cents per pound
and there is additional cost for chemical auxiliaries. In response to my question regarding BTCA,
Mr. Turner responded that BTCA is not commercially available, but he knows of a manufacturer that
will supply BTCA for $2.50 per pound for a minimum order of 1 million pounds. For smaller
quantities, the cost is $13 per pound. Mr. Turner also stated that resins without formaldehyde do not
cross link as well as formaldehyde. Mr. Turner mentioned there are a number of polycyclic acids
without formaldehyde that can be used for durable press finishes, e.g., citric acid, though citric acid
has a tendency to yellow and is not durable to home laundering. Mr. Turner suggested that I call
David Shank of Vulcan Performance Chemicals in Columbus, GA for cost information about the
company's polycyclic acid polymer for durable press.
8-4
-------
ATTACHMENT 2
CONTACT REPORT
From: Steve York (919-990-8629), Research Triangle Institute project lead for the Printing, Coating,
and Dyeing of Fabrics and Other Textiles NESHAP
Date of Contact: March 12, 2001
Contact: Jennifer Grabowski
Company/Agency : BF Goodrich Textile Performance Chemicals
Telephone Number: (704) 399-0216
Location; Charlotte, NC
CONTACT SUMMARY:
Ms.Grabowski was contacted by telephone to solicit information regarding the cost of permanent
press resins containing formaldehyde versus the cost of permanent press resins without formaldehyde.
Ms. Grabowski returned the call and provided information about two 40 percent glyoxal products,
i.e., Freechem 40D and Freechem 40 DL. Freechem 40D contains less than 1 percent formaldehyde
while Freechem 40DL is BF Goodrich's low-formaldehyde product and contains less than 100 ppm
formaldehyde. The cost of each depends on the quantity ordered and for Freechem 40D ranges from
$ 1.06 per pound for an order of 1 to 3 drums down to $0.79 per pound for an order of 50 to 80 drums.
Freechem 40DL costs 3 cents per pound more than Freechem 40 D at each size range.
8-5
-------
ATTACHMENT 3
CONTACT REPORT
From: Steve York (919-990-8629), Research Triangle Institute project lead for the Printing, Coating,
and Dyeing of Fabrics and Other Textiles NESHAP
Date of Contact: March 22, 2001
Contact: Jerry Setzer
Company/Agency : Vulcan Performance Chemicals
Telephone Number: (706) 576-6403
Location: Columbus, Georgia
CONTACT SUMMARY:
Mr. Setzer was contacted by telephone to solicit information regarding the cost of permanent press
resins containing formaldehyde versus the cost of permanent press resins without formaldehyde. Mr.
Setzer offered that for a non-formaldehyde cross-link system developed and marketed by Vulcan, the
cost of the resin is $1.38 per pound and the cost of the catalyst is $.78 per pound. The resin and
catalyst are added to a finishing formulation at a 3 to 1 ratio. A typical formulation also contains a
silicone softener ($1.65/lb), a wetting agent ($.36/lb) and a lubricant ($.44/lb), but these would also
probably be included in a formaldehyde cross-link system. Based on typical usage quantities, Mr.
Setzer has estimated that the cost of one pair of twill pants finished with the non-formaldehyde cross-
link system would be about $.40 - .45 more expensive than a pair of twill pants finished with a
formaldehyde-containing permanent press finish.
Mr. Setzer claimed that the Vulcan Performance Chemicals non-formaldehyde cross-link system
yields comparable results to a formaldehyde cross-link system, when cured at the proper temperature.
The system was first tested by several textile companies in 1998 and seemed to be producing erratic
results. Vulcan measured the cure temperatures and found that the fabric should reach a temperature
of 170 °F in the curing process for optimal results. In addition to yielding comparable results to
formaldehyde cross-link systems, the cross-link material used in Vulcan's system has less adverse
affect on the strength of the finished fabric.
Mr. Setzer stated that, to date, no U.S. textile company is using the Vulcan non-formaldehyde cross-
link system. The U.S. market is driven by large customers such as LL Beane and there is resistance
to change for fear of losing a customer. The Vulcan non-formaldehyde cross-link system is being
used successfully by one textile manufacturing company in Europe.
8-6
-------
MEMORANDUM
TO: Vinson Hellwig, EPA/OAQPS/ESD/CCPG
FROM: Steve York and Alton Peters, RTI
DATE: December 20, 2001
SUBJECT: Summary of Evaluation of Estimated Compliance Costs Incurred by Coating Facilities
Owned by Small Businesses
The approach followed to estimate the compliance costs that will be incurred by coating
facilities subject to the printing, coating, and dyeing of fabrics and other textiles NESHAP that are
owned by small businesses is presented in the following paragraphs. The approach consisted of the
following three steps: (1) identify the major facilities with coating operations that are owned by small
businesses, (2) collect information needed to estimate compliance costs, and (3) estimate compliance
costs.
Identify Major Facilities with Coating Operations Owned by Small Businesses
A list of major facilities with coating operations was developed using information from the
coating MACT database ' and 1997 HAP emissions data from the Toxics Release Inventory (TRI)
database. The year 1997 was chosen because it is the base year for the coating MACT database. The
coating MACT database has information on the facility Title V classification for HAP (area/minor,
synthetic minor, or major source). For facilities from the TRI database, we assumed that if reported
emissions were less than 10 tons per year (TPY) of any one HAP or 25 TPY of total HAP, then the
facility is an area source or synthetic minor and is not subject to the control requirements of the
NESHAP. In order to verify major source status, years later than 1997 were checked, and the facility
was considered a major source if emissions were 10/25 tons per year or greater for any of those years.
A list of coating facilities owned by small businesses (i.e, businesses with fewer than 1,000
employees) was provided by the Innovative Strategies and Economics Group (ISEG). This list was
used to identify the major facilities with coating operations owned by small businesses. Table 9-1
presents the list of the 25 coating facilities owned by small businesses.
Collect Information Needed to Estimate Compliance Costs
A key factor in estimating the compliance cost that will be incurred by a facility is whether
the facility currently operates add-on emission controls. The first information we collected to help
determine current compliance status was the VOC attainment status of the county in which each of
the small facilities is located.
9-1
-------
Table 9-1. Coating Facilities Owned by Small Businesses
Facility Name
Amerbelle Corp.
Athol Corp.
Bando Mfg. of America Inc
Bradford Industnes Inc.
Brownell & Co. Inc.
Delatex Processing Corp.
Duraco, Inc.
Duro Industries Inc.
Eddington Thread Mfg. Co.
Excello Fabric Finishers Inc.
Fil-Tec, Inc.
General Clothing Co. Inc.
Haartz Corp.
Holliston Mills Inc.
Hub Fabric Leather Co Inc
J. Charles Saunders Co. Inc.
Kenyon Ind. Inc.
Ouimet Corp
Par Products
Perm Racquet Sports
Robin Industries Inc.
Seaman Corp.
Schneller
Textileather Corp.
Textile Tapes Corp.
City
Vernon
Burner
Bowling Green
Lowell
Moodus
Clifton
Chicago
Fall River
Worcester
Coshocton
Cavetown
Smyrna
Acton
Church Hill
Everett
Gastonia
Kenyon
Nashville
Wylie
Phoenix
Cleveland
Bristol
Kent
Toledo
Gome
State
Connecticut
North Carolina
Kentucky
Massachusetts
Connecticut
New Jersey
Illinois
Massachusetts
Massachusetts
Ohio
Maryland
Delaware
Massachusetts
Tennessee
Massachusetts
North Carolina
Rhode Island
Tennessee
Texas
Arizona
Ohio
Tennessee
Ohio
Ohio
New Hampshire
County
Tolland
Granville
Warren
Middlesex
Middlesex
Passaic
Cook
Bristol
Worcester
Coshocton
Washington
Kent
Middlesex
Hawkins
Middlesex
Gaston
Washington
Davidson
Collin
Maricopa
Cayahoga
Sullivan
Portage
Lucas
Strafford
Total HAP
Emissions (TPY)
26.3
269.3
56.6
59.3
46.5
12.9
47.8
42
19.2
191.6
46.3
11.5
12.8
553
84.7
17.6
123
25.4
53.4
73
11.6
17
28.9
396
31.2
Sales
(Smillion)
26
47
53
53
7.5
0.75
27.5
199
5
3.75
14
7.5
100
21.2
47
15
115.3
15
1.75
75
15
68
28.3
12.9
1.75
9-2
-------
Most of the designations were taken from the website
http://www.epa.gov/oar/oaqps/greenbk/oncs.html. Also, designations for counties in some states
(e.g., Ohio) were taken from 40 CFR Part 81.
Three of the facilities owned by small businesses are located in severe non-attainment areas.
These facilities are Delatex Processing Corporation, Duraco Incorporated, and General Clothing
Company Incorporated. For each of these facilities, we assumed that the SIP RACT requirements
would be imposed and efficient controls would be in place, and therefore, no upgrade of add-on
controls will be required.
For the remaining facilities in Table 9-1, we used available data in the coating MACT
database ' or ATMI MACT database 2 and information obtained through telephone contacts of state
and local permitting authorities and the facilities to determine applicability of the printing, coating,
and dyeing of fabrics and other textiles NESHAP, current Title V permit status and level of HAP
emission control. The following paragraphs summarize information collected for each facility that
served as the basis for the estimates of compliance costs.
Amerbelie Corporation. Information in Reference 2 indicates that two coating lines are tied into
a thermal oxidizer with less than 97 percent overall control efficiency (OCE). RACT compliant
coatings are used.
Athol Corporation. Information in Reference 1 indicates that the HAP emissions reported by Athol
are from finishing machines that are subject to the Printing and Publishing NESHAP. A control
device was planned by 1999 to comply with the Printing and Publishing NESHAP.
Bando Manufacturing of America, Inc. Information in Reference 1 indicates that there are no
emission control devices used by this facility. However, drying is not done in ovens, therefore, the
coating operations do not meet the applicability criteria of the Printing, Coating, and Dyeing of
Fabrics and Other Textiles NESHAP to be proposed.
Bradford Industries. Inc. Information in Reference 1 indicates HAP emissions are controlled by
a regenerative thermal oxidizer with an OCE greater than 97 percent.
Brownell & Company. Inc. The information in this paragraph was collected in a telephone contact
on December 6, 2001 between Steve York, RTI and Nicholas Vasile, Brownell & Co., telephone
(860) 873-8625. Brownell was founded in 1846 and is currently the only mill of 10 to 12 that
formerly operated that is still operating in Moodus, Connecticut. The facility operates 10 twine, net,
and rope (cord) treating tables. The cord is fed from creels through dip tanks for coating (plasticizer-
based formulations) and then is pulled vertically through a lighted attic where flash off takes place
and is then rewound. There are no emission controls. The facility has a permit limit of 50 TPY VOC
and a 5 Ib/hr limit on highly photochemically reactive compounds and 40 Ib/hr on other VOC. To
comply with the 5 Ib/hr limit, jobs with coatings with highly photochemically reactive components
are run at a slow production rate, e.g., a job that would take 1 hour is run at a rate that will take 8
hours.
9-3
-------
Since drying is not done in ovens, the coating operations do not meet the applicability criteria of the
NESHAP to be proposed.
Duro Industries. Inc. The following information was collected in a telephone contact on December
6, 2001 between Vinson Hellwig, EPA/CCPG and Bill Bailey, Environmental Manager, Duro
Industries, telephone (508) 675-0101, x 1603. Mr. Bailey stated that the fabric coating operations that
emit HAP are controlled by a thermal oxidizer and the OCE is 94 %. He further stated that they have
aqueous coating lines with no HAPs (that can be averaged with the controlled lines). Mr. Baily also
stated that they perform printing, the print past is high solids but low HAP, and the paste diluent is
a non-HAP material. This would allow averaging of the print paste into the overall compliance
determination. Mr. Bailey was very familiar with the MACT development process, he was a
stakeholder on a prior MACT, and he was aware of the printing, coating, and dyeing of fabrics and
other textiles NESHAP. He was not certain at this time that Duro could meet the 0.12 Ib HAP/lb
solids limit, but it is possible. On the issue of recordkeeping, he anticipated no increased costs over
his Title V permit recordkeeping requirements to meet the NESHAP MRR requirements. Duro's
Title V permit encompasses all the recordkeeping that will be in the NESHAP monitoring, reporting
and recordkeeping requirements.
Eddington Thread Manufacturing Company. The following information was collected in a
telephone contact on December 5, 2001 between Vinson Hellwig, EPA/CCPG and Dana Nickel,
Massachusetts DEP, Central Region, telephone (508) 767-2772. Ms. Nickel stated that Edington has
restricted emissions of methanol from thread coating/bonding. They operate and oxidizer with an
OCE of 90 to 95 percent. They were subject to a State BACT that was more stringent than RACT.
Excello Fabric Finishers. Inc. The following information was collected in a telephone contact on
December 4, 2001 between Steve York, RTI and Kay Gilmer, Ohio EPA, Southeast Region,
Telephone (740) 380-5257. Excello uses VOC compliant coatings and has no add-on controls. The
facility operates 1 coating line which coats canvas for tents. Excello has not been able to reformulate
to reduce HAP and maintain product specifications. HAP emissions for the past 3 years (RY 2000
are preliminary data) are as follow:
Chemical
MEK
Toluene
Total Air Release (Ibs)
RY1998
39,675
222,709
RY1999
21,026
190,962
RY2000
22,226
176,634
Fil-Tech, Inc. The following information was collected in a December 5, 2001 fax transmittal from
Laramie Daniel, Maryland Air Quality Compliance Program, Air and Radiation Management
Administration (ARMA) to Alton Peters, RTI, in response to a telephone contact on December 4,
2001 between Alton Peters, RTI and Bill Reamy, Maryland Department of the Environment,
telephone (410) 631 -3504. Fil-Tec is Title V but does not have a Part 70 permit to operate. They will
have to submit a Part 71 PTO application. Three bonder/coaters exhaust to a thermal oxidizer.
9-4
-------
Destruction efficiency = 98% @ 1150 °F per an 8/97 stack test. Capture efficiency = 98.2% (also
8/97 test). OCE = 96.2%. Products currently manufactured include polyester and nomex threads,
glass insulation wrap, fiberglass wicks, and fiberglass yam. Former products include climbing rope,
rip cords, fiber optic lines, and dental floss.
Haartz Corporation. The following information was collected in a telephone contact on December
5,2001 between Vinson Hell wig, EPA/CCPG and Dana Nickel, Massachusetts DEP, Central Region,
telephone (508) 767-2772. Ms. Nickel stated that Haartz emits MEK and has two oxidizers that
control emissions. The controls meet the state RACT requirements (a minimum of 85 percent OCE)
and may be operating at a higher level of control.
Holliston Mills. Inc. The following information was collected in a telephone contact on December
6,2001 between Steve York, RTI and Dan Cochran, Holliston Mills, Telephone (423) 357-6141. Mr.
Cochran stated that he does not like giving out information over the phone, and asked that I submit
a written questionnaire. I told him we were gathering information under a tight deadline and did not
have time to send a letter. He confirmed that the plant is a coating facility subject to the Printing,
Coating, and Dyeing of Fabrics and Other Textiles NESHAP and has no emission controls. Holliston
is in debt, and was bought out about 3 and one half years ago. Mr. Cochran offered that the company
has put a lot of effort in the last 3 years into converting to waterborne coatings and is in compliance
with the State's 2.5 Ib/gal VOC limit. I talked to Mr. Cochran about the proposed limits and some
of the compliance requirements and gave him the address of the Air Toxics CCCR website.
Hub Fabric Leather Company, Inc. The following information was collected in a telephone
contact on December 5, 2001 between Vinson Hellwig, EPA/CCPG and Mon Wong, Massachusetts
DEP, Northeast Region, telephone (978) 661-7677. Hub Fabric has applied for and will shortly be
issued a Synthetic Minor Title V permit with restrictions on production using HAP-emitting materials
that will keep Hub below the major source threshold. Therefore, Hub will not be subject to the
Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP.
.1. Charles Saunders Company. Inc. The following information was collected in a telephone
contact on December 6, 2001 between Steve York, RTI and Mike Landis, NCDEHNR, Mooresville
Office, Telephone (704) 663-1699. J. Charles Saunders Co. produces thread and has 5 thread bonding
machines. There are no emission controls. The thread bonding process is basically the same as the
process at A&E that is controlled by thermal oxidizer. In 2000, Saunders reported 25,000 pounds of
VOC (methanol) emissions.
Kenyon Industries, Inc. Confidential business information in Reference 1 indicates that the coating
operations at Kenyon are controlled, but the control systems will not comply with the MACT floor
and will require upgrades.
Ouimet Corporation. The information in this paragraph was collected in a telephone contact on
December 10, 2001 between Vinson Hellwig, EPA/CCPG and Don Greeson, Ouimet Corporation,
telephone (615) 242-5478. Mr. Greeson confirmed that Ouimet manufactures synthetic leather
products. They form the material on paper then release it from the paper. It is then laminated to
9-5
-------
fabric in a heat process with no HAPs or other solvents used. It is then printed on the synthetic side,
not the fabric side, at a separate station.
The process would therefore not be subject to the Printing, Coating, and Dyeing of Fabrics and Other
Textiles NESHAP. It is probably subject to the Paper and Other Web Coating NESHAP, and the
facility is aware of that applicability.
Par Products. The following information was collected in a telephone contact on December 5,2001
between Steve York, RTI and Craig Richardson; TNRCC, Air Permits, Coating Team, telephone
(512) 239-1309. Par Products has installed a carbon adsorption system that is supposed to be 95
percent efficient. However, the facility has not been able to demonstrate the NSPS level of 90 percent
OCE, probably because of poor capture efficiency. The plant is permitted at 22.6 T of VOC and is
in violation. Par Products produces plugs for repairing tubeless tires. All of the emissions are n-
hexane.
Penn Racquet Sports. Information in the Coating MACT database indicates that this facility
operates a carbon adsorber system with three 13,000 pound carbon beds in parallel. Two beds are
in service and one is regenerating at any one time. The system OCE is 93.1 percent (98 percent
capture and 95 percent control).
Robin Industries, Inc. Steve York, RTI left messages December 6 and 10, 2001 requesting a return
call from Mr. Mike Olderman, plant manager, telephone (216) 961-5810. No return call has been
received. Steve York also left a message December 3,2001, requesting a return call from Mr. David
Hearne, Cuyahoga County Ohio EPA, Cleveland Air Pollution Control (Permits), telephone (216)
664-2178. No return call has been received. Steve York contacted Jenneta Adams of Cleveland
Department of Health and Welfare, Division of the Environment, telephone (216) 664-2457 on
December 18 and 19, 2001. Ms. Adams was able to pull the file for Robin Industries, but required
a written request for information and several levels of signature approvals to release the information,
which could not be accomplished until after January 1, 2002.
Seaman Corporation. The information in this paragraph was collected in a telephone contact on
December 4, 2001 between Steve York, RTI and Andrew Shimko, Seaman Corporation, telephone
(330) 262-1111. In response to a request for information on the Bristol, Tennessee plant, which
reported no add-on controls in 1997 in an information collection request response, Mr. Shimko stated
that the facility has 3 coating lines, only one (Line 3) with HAP emissions. Line 3 runs only a few
days a month. Mr. Shimko has estimated the cost of installing PTE/RTO to be around $1 million.
In addition, process modifications would be required costing around $200,000. The web loops
around in the dryer on Line 3 so it enters and exits the same end. The company will evaluate closing
the line. If a control system is installed, it would be used to control VOC emissions from Lines 1 and
2 as well as HAP emissions from Line 3. I asked Mr. Shimko about annual sales, noting that he had
reported $60 million in sales in 1997 and our estimate for 2000 is $22 million. He replied that their
sales have increased and were $68 million last year. Seaman is currently doing a lot of fabric coating
for military applications, which is not related to the conflict in Afghanistan.
9-6
-------
In an ICR response that is in Reference 1, Seaman has provided information showing that less than
one percent by mass of the coating materials used by the facility contain HAP. Therefore, the facility
will be able to comply with the emission rate limit without add-on controls that will be in the
proposed Printing, Coating and Dyeing of Fabrics and Other Textiles NESHAP by averaging the
HAP content from the HAP-containing materials across the solids content of all coating materials
applied by the facility.
Schneller. Inc. Confidential business information in Reference 1 indicates that the coating
operations at Schneller are controlled and the control systems will comply with the MACT floor
requirements.
Textileather Corporation. The following information was collected in a telephone contact on
December 4, 2001 between Steve York, RTI and Bob Kossow; Toledo Environmental Services,
telephone (419) 936-3015. Textileather has submitted a Title V permit application and the permit is
in process. The engineer working on the permit is on leave through the end of this year; Mr. Kossow
is not familiar with the facility but had access to the permit application. Textileather applies a
textured coating to cloth, fabric and plastic substrates. The facility has 9 vinyl coating lines (print and
finish), 3 calendering lines (high solids, restricted to 2.9 Ib VOC per gallon of coating) and 2 plastisol
lines. The vinyl coating lines are controlled by a 90 percent efficient carbon adsorption system, the
calendering lines are uncontrolled, one plastisol line is controlled by ESP to remove condensate and
the second vents to a thermal oxidizer that is permitted at 95 percent destruction, though Mr. Kossow
is certain he has seen a performance test demonstrating 99 percent (combustion T is 1400 °F).
Capture efficiency appears to be 75 percent.
The following additional information was collected in a telephone contact on December 10, 2001
between Steve York, RTI and Rick Scott, Textileather, Toledo, Ohio, telephone (419) 729-7557.
Mr. Scott confirmed that the vinyl coating lines and plastisol line controlled by thermal oxidizer are
subject to the Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP. A HAP-
containing top finish is applied on the vinyl coating lines. He stated that the calendering lines are
applying high solids materials with minimal HAP emissions. Mr. Scott stated that the carbon
adsorption system has 90 percent capture and 95 percent control and confirmed the capture efficiency
of 75 percent and destruction efficiency of 95 percent for the thermal oxidizer. Textileather is
evaluating options for complying with the NESHAP. One option that is being considered is
converting to a non-HAP finishing material with acetone substituted for the HAP. This is not an
attractive option because the finish would have a VOC content very close to the State's organic limit
of 4.8 Ib/gal of coating. An adjustment in coating viscosity would result in a violation of the State
limit. Textileather has estimated a cost of $ 1.5 million for installing PTE on all of the coating lines.
This would include air conditioning needed in the summer and extra ventilation because of employee
exposure concerns. Textileather has also evaluated replacing the carbon adsorption system with
RTO; this would cost between $2 and $3 million. In response to my question of Textileather being
an ESOP, Mr. Scott replied that employees bought the company from Gencorp in 1990 and sold to
Canadian General Tower in 1995.
Textile Tapes Corporation. The following information was collected in a telephone contact on
9-7
-------
December 6,2001 between Steve York, RTI and Danuta Royes, New Hampshire DES, Air Resources
Division, telephone (603) 271-1987. Textile Tapes is permitted as a Title V major source. The
permit is being modified to add a thermal oxidizer. The oxidizer will be permitted with a minimum
OCE of 95%. The plant has demonstrated PTE using Method 204. Two coating lines are operated,
one with a dryer and one without a dryer. The plant produces shoe laces and fabric tape that is used
to line the inside of leather shoes.
Estimate Compliance Costs
Based on the information described in the previous section of this memorandum, we
determined that six of the facilities listed in Table 9-1 will not incur compliance costs. The HAP
emissions reported by Athol Corporation are from operations subject to the Printing and Publishing
NESHAP and will be controlled accordingly. Similarly, the HAP-emitting operations at Ouimet
Corporation are subject to control under the Paper and Other Web Coating NESHAP. Hub Fabric
Leather Company, Inc. is being permitted as a synthetic minor and therefore will not be subject to the
Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP. Neither Bando Manufacturing
of America, Inc. nor Brownell & Company, Inc. dries the coated textile substrate in an oven after
coating application, therefore, the coating operations at these facilities do not meet the applicability
criteria that will be proposed in the Printing, Coating, and Dyeing of Fabrics and Other Textiles
NESHAP. Seaman Corporation's Bristol, Tennessee facility will be able to comply with the emission
rate limit without add-on controls option that will be in the proposed NESHAP.
Table 9-2 presents the estimated compliance costs for the remaining 19 coating facilities
owned by small businesses. The costs were estimated by assigning the applicable model plant control
costs (see the memorandum at page 7-1 of this document) and monitoring, recordkeeping and
reporting (MRR) costs 3 based on the following assumptions:
• The most cost effective add-on control option that would bring a facility into compliance was
costed, e.g., adding permanent total enclosures if 100 percent capture efficiency combined
with the existing control device destruction or removal efficiency would result in 97 percent
OCE, upgrading an existing control device rather than installing a new control device, or for
an uncontrolled facility, assigning the cost of the most cost effective applicable control
system;
• Sizes of Model Plants used to determine the costs to assign were chosen on the basis of
uncontrolled facility HAP emissions;
• Facilities for which we did not know the number of coating lines were assumed to have 2
lines (the average number of coating lines per facility in the coating MACT database) for the
purpose of assigning PTE costs with the exception of Holliston Mills, Inc. which was
assumed to have 4 coating lines because of the magnitude of HAP emissions;
• With regard to MRR costs, facilities in serious and severe VOC non-attainment areas were
assigned only the MRR costs associated with performance testing control devices and PTE
since other records are already maintained as part of their Title V requirements;
• Robin Industries was assumed to be uncontrolled and assigned the cost of a carbon adsorption
system with 2 PTE; and
9-8
-------
• Facilities in severe non-attainment areas were assumed to have SIP required emission
controls, and therefore, be in compliance with the MACT floor requirements.
References
1 U.S. Environmental Protection Agency. Fabric Printing, Coating, and Dyeing NESHAP.
ICR Responses. Office of Air Quality Planning and Standards. Research Triangle Park, NC.
Responses received September 1998 - October 1998.
2. Memorandum and Attachment from York, S. and A. Sharma, RTI to P. Almodovar,
EPA/OAQPS/ESD/CCPG. November 13, 1998 Final. Summary of meeting at which ATMI
presented the results of the ATMI MACT survey to EPA.
3. U.S. Environmental Protection Agency. Standard Form 83-1 Supporting Statement for OMB
Review of ICR No. Information Collection Request for the Fabric Printing, Coating, and
Dyeing Source Category. Emission Standards Division. Research Triangle Park, NC.
December 18,2001.
9-9
-------
1/1
v>
$
e
en
3
CO
1
;s Owned h
rs
P*N
OJD
"S
0
U
-2
VI
0
U
C
"a
S
o
•a
es
a>
A
n
H
o -c
u R
< 55
H
*
u <&
u
H
o —
w I I
U u
1J
voc
Attainment
Status
s
•a e C
m
w
>.
o
a
3
n
6-
1
2
ON
SO
8
so
CM
I
S
Serious
r*i
S
Tolland
Connecticut
Vernon
Amerbelle Corp.
oo
f}
^
ts
!2
I
8
Serious
oo
2
Middlese
«>
Massachuse
u
Haartz Corp '
i
oo
r-
r-
oo
m
S
™*
o
c
rM
1
m
Hawkins
Tennessee
Church Hill
CJ
c
o
~o
X
S
SO
ON
r-
^
Tf
m
S
O
c
2
Moderate
VO
C
0
a
O
c
North Carol
Gastonia
«
c
i. Charles Saunders Co
oo
1 — .
—
^
s
%
in
I
i
O
•c
OJ
^J
O
00
c
%
Rhode Islan
c
o
c
Kenyon Industries Inc.
S
m
ON
OO
00
o
oo
E
£
Senous
^
in
c
5
X
H
n
Serious(p)
r-
o.
o
i
s
Anzona
Phoenix
Penn Racquet Sports
S
o^
r-
^
Tf
m
a
o
c
m
Moderate
VO
1
C4
O
Cleveland
Robin Industries Inc.
o
0
o
c
s
1
—
Sullivan
Tennessee
o
03
1 Seaman Corp. r
Wl
r-
o
OV
„
1^
fn
S.
06
Moderate
ON
oo
-------
Table 9-2 Footnotes
Facilities in severe non-attainment areas are assumed to have SIP required controls. Robin
Industries is assumed to be uncontrolled and has been assigned the cost of a carbon adsorption
system and two PTE.
Total capital compliance costs (TCCC) and total annual compliance costs (TACC) include costs
associated with upgrade or installation of engineering control systems, where applicable, and
MRR costs.
Two coating lines assumed.
Four coating lines assumed.
PTE on two coating lines assumed.
Information in Reference 1 indicates this facility is in compliance with the emission rate without
add-on controls option that will be proposed.
9-11
-------
MEMORANDUM
TO: Vinson Hellwig, EPA/OAQPS/ESD/CCPG
FROM: Steve York and Alton Peters, RTI
DATE: June 12, 2002
SUBJECT: Printing, Coating, and Dyeing of Fabrics and Other Textiles NESHAP Nationwide
Compliance Costs
The purpose of this memorandum is to present estimates of the nationwide costs resulting
from compliance with the proposed printing, coating, and dyeing of fabrics and other textiles
NESHAP. The compliance costs consist of the costs of add-on controls for the coating and printing
subcategory; compliant, low-formaldehyde permanent press finishes for the dyeing and finishing
subcategory; and monitoring, reporting and recordkeeping (MRR) costs for all major sources in the
printing, coating, and dyeing of fabrics and other textiles source category.
Coating and Printing Control Costs
The coating MACT database' contains sufficient non-CBI information from 16 facilities that
are major sources of HAP emissions to calculate a facility organic HAP overall control efficiency
(OCE). Two of the facilities report OCE of greater than 97 percent determined using EPA test
methods, and therefore, are in compliance with the proposed OCE limit. The remaining 14 facilities
will be required to take measures to reduce organic HAP emissions either through coatings
reformulation or through adding or upgrading emission control systems.
Information needed to estimate the compliance costs that would be incurred by coating
facilities subject to the printing, coating, and dyeing of fabrics and other textiles NESHAP that are
owned by small businesses (hereafter referred to as the small business database) has also been
collected (see memorandum at page 9-1 of this document). The small business database includes
information on 20 facilities (3 of which are also in the coating MACT database). Of the 17 small
business database facilities that are not also in the coating MACT database, 5 have been determined
to be in compliance with one of the proposed emission limits. The remaining 12 facilities owned by
small businesses will be required to take measures to reduce HAP emissions.
Because 73 percent of the facilities in the coating MACT and small business databases (24
of the 33 facilities) already have controls in place, and because of the likelihood that organic HAP
are required in certain coatings to achieve desired performance characteristics, we assume facilities
needing to reduce HAP emissions to comply with one of the compliance options will do so either by
upgrading existing controls or installing controls if emissions are currently uncontrolled.
10-1
-------
We have examined the capture and control efficiencies reported by each facility with existing
add-on control systems that do not achieve the emission limits to determine the most cost-effective
measure needed to reach compliance, e.g., a facility with a 97 percent efficient control device but less
than 100 percent capture efficiency will need to install coating rooms on application stations to meet
a facility OCE of 97 percent. Similarly, for the 9 facilities that are currently uncontrolled, we have
evaluated applicable controls (facilities in the MACT database applying various coatings to industrial
fabrics report using thermal oxidizers; facilities in the MACT database applying coatings with only
one or two solvents report using catalytic oxidizers or carbon adsorbers) to determine the most cost-
effective add-on control device that could be installed to attain compliance.
Table 10-1 presents a summary of coating and printing model and nationwide control costs.
The nationwide compliance costs for model plants are based on the total number of small, medium
and large coating rooms needed to upgrade capture efficiency, the total number of control device
upgrades needed for each model plant assigned to represent a facility, and the number of new
emission control systems needed for facilities that are currently uncontrolled. In addition, two
facilities with methylene chloride emissions that will incur additional control costs (see the October
12, 2001 memorandum at page 7-1 of this document regarding compliance costs for coating model
plants) have been identified from ATMI MACT survey2 and TRI database information. Based on
the total methylene chloride emissions reported by each facility, one was assigned Model 1 carbon
adsorber control costs and one was assigned Model 3 carbon adsorber control costs.
For the 26 facilities in the coating MACT and small business databases to which model plants
are assigned, the total capital investment is $8,089,006 and the total annual cost is $2,617,336 per
year in 1997 dollars. The total HAP emissions for these facilities in 1997 were 2,326 tons. The total
nationwide organic HAP emissions in 1997 from coating and printing facilities were estimated to be
5571 tons (see January 7, 2002 memorandum at page 5-1 of this document summarizing printing,
coating, and dyeing of fabrics and other textiles NESHAP baseline organic HAP emissions and
emission reductions), of which 5,011 tons were from major sources of HAP that will be subject to the
control requirements of the NESHAP and 214 tons were methylene chloride emissions from two
coating facilities. To estimate the control costs for all coating and printing facilities, the control costs
for the coating MACT and small business database facilities represented by the model plants were
factored by the ratio of HAP emissions from major sources for the subcategory (minus methylene
chloride emissions for which control costs were estimated separately) to HAP emissions reported by
facilities represented by model plants (i.e., 4,797/2,326 = 2.06) and the control costs for methylene
chloride emissions were added. Therefore, the estimated nationwide total capital investment is
$17,574,651 and the nationwide total annual control cost is $5,615,407 per year in 1997 dollars.
10-2
-------
Table 10-1. Summary of Coating and Printing Subcategory Model and
Nationwide Control Costsa
Model
New Add-on Control Device "
Model 1 , carbon adsorber
Model 1, catalytic oxidizer
Model 2, thermal oxidizer
Model 3. carbon adsorber
Upgrade of Add-on Control Device
Model 2, catalytic oxidizer
Model 3, catalytic oxidizer
Model 3, carbon adsorber
Model 4, catalytic oxidizer
Model 4, carbon adsorber
New Coating Room (PTE)
Small
Medium
Large
Total Control Costs for Model Plants Except
Methylene Chloride Model Plants
Nationwide Total Control Costs Except Methylene
Chloride Control Costs '
New Add-on Control System for Methylene
Chloride Emissions '
Model 1 , carbon adsorber
Model 3, carbon adsorber
Total Methylene Chloride Control Costs
Nationwide Total Control Costs with Methylene
Chloride Control Costs «
Number of
plants b
|
2
1
2
4
1
2
1 3
9
1
14
13
29
1
1
Model total
capital
investment '. $
104,183
300,140
576,551
501,693
130,967
136,036
159,504
182,319
218,447
42,720
50,670
57,120
J
210,568
700,731
Nationwide
total capital
investment. $
208,366
300,140
1,153,102
2,006,772
130,967
272,072
478,512
364,638
218,447
640,800
658,710
1,656,480
8,089,006
16,663,352
210,568
700,731
911.299
17,574,651
Model total
annual cost %
$/vr
31,068
90,888
241,585
87,350
36,302
47,914
30,492
58,646
42,523
L
19,743
22,186
23,569
62,477 1
161,218
i
1
Nationwide
total annual
cost, $/vr
62,136
90,888
483,170
349,400
36,302
95,828
91,476
117,292
42,523
276,402
288,418
683,501
2,617,336
5,391,712
62,477
161,218
223,695
5,615,407
10-3
-------
Table 10-1 Footnotes
The nationwide costs were calculated using model plants to estimate the costs of bringing each of
14 coating MACT database facilities and 12 small business database facilities into compliance
with the proposed emission limits, extrapolating this to a nationwide cost based on organic HAP
emissions from major sources for the subcategory, and adding costs for controlling methylene
chloride emissions from the 2 major facilities reporting methylene chloride emissions in the TRI
database (neither of which is in the coating MACT database or is owned by a small business). For
each of the 26 coating MACT and small business database facilities, the most cost-effective add-on
control measure (e.g., upgrading capture efficiency by adding PTE to application stations, or if no
add-on controls are in place, the installation of a complete system including PTE and add-on
control device) was applied to bring the facility into compliance with one of the proposed emission
limits. The model plant costs include costs of installing, upgrading, operating and maintaining
add-on control systems. MRR costs are presented in Table 10-2. All costs are in 1997 $.
Number of model plants assigned to the 26 facilities in the coating MACT and small business
databases requiring organic HAP emission reductions to estimate the compliance cost of achieving
the MACT floor compliance options with add-on controls.
From October 12,2001 memorandum regarding compliance costs for coating model plants. Note
that the upgrade costs represent incremental costs above the costs of a baseline unit.
Model plant costs represent the costs of a new add-on control device and auxiliaries, including
ductwork, butterfly dampers, fans, motors, and stacks. Coating room costs are presented
separately.
Nationwide total control costs for all facilities in the coating and printing industry", except plants
with methylene chloride emissions are based on factoring the total control costs for model plants
except methylene chloride model plants by the ratio of HAP emissions estimated for major HAP
emission sources in the coating and printing subcategory (minus methylene chloride emissions)
to the HAP emissions reported by facilities for which control costs have been estimated (the ratio
is 2.06)
Includes cost of add-on control system and coating room.
Nationwide total control costs for all affected facilities in the coating and printing industry are the
sum of the nationwide total control costs except methylene chloride control costs and the total
methylene chloride control costs.
10-4
-------
Dyeing and Finishing Compliance Costs
The dyeing and finishing compliance options are based on the use of low-HAP materials.
During the data collection effort to support the MACT floor determination, EPA held numerous
stakeholder meetings and made eight site visits to facilities with dyeing and finishing operations.
Qualitative information concerning pollution prevention measures gathered from the stakeholder
meetings and site visits indicated that there would be substantial costs incurred in reducing the
formaldehyde content of permanent press resins. No concerns were expressed about the cost of
reformulating other dyes and finishes. Therefore, we collected information from Cotton Incorporated,
a research and marketing company, and two textile chemical suppliers regarding the incremental cost
of non-formaldehyde permanent press finish versus permanent press finish with formaldehyde (see
August 1, 2001 memorandum at page 8-1 of this document summarizing incremental cost
information).
Information collected from Cotton Incorporated indicates that the cost of finishing with a
formaldehyde-containing resin ranges from 5 to 15 cents per pound of finished fabric and the cost of
the cross-linking agent runs between 50 and 65 cents per pound. According to Cotton Incorporated,
BTCA, a non-formaldehyde finish that is not commercially available can be purchased for $2.50 per
pound for a minimum order of 1 million pounds. However, through contacts with textile chemical
suppliers we found a permanent press resin on the market that contains less than 100 ppm
formaldehyde (below MSDS reportable quantities and in compliance with the proposed emission limit
for finishing) for 82 cents per pound for an order of 50 to 80 drums. Thus the cost of the compliant
cross-linking agent is about 43 percent higher than the cost of a formaldehyde cross-linking agent.
Assuming that the cost of finishing is directly proportional to the cost of the cross-linking agent, the
cost of finishing with the compliant resin would range from about 7 to 21 cents per pound of finished
fabric, an average of approximately 4 cents per pound of finished fabric more than the cost of
finishing with a formaldehyde resin.
The ATMI MACT survey database 2 contains information about facility Title V status for
HAP, wet finishing operations with formaldehyde emissions, and the quantity (pounds) of fabric
processed in each finishing operation. Facilities that are major sources for HAP in the ATMI MACT
database reported finishing over 1.44 billion pounds of fabric per year in operations with associated
formaldehyde emissions. Actual formaldehyde emissions reported by the facilities ranged from 0.01
to 13.9 tons, with most of the facility emissions less than 1 ton per year.
The ATMI MACT finishing database 3 contains information about facility Title V status for
HAP, the HAP content of finishing materials used, and the annual production of finished fabric and
provides a basis for estimating the quantity of fabric currently finished with compliant materials by
facilities that are major sources of HAP emissions. Analysis of the database indicated that 87 percent
of the fabric finished in operations at major sources using formaldehyde-containing materials was
finished with compliant materials (in terms of formaldehyde content). Therefore, to estimate the
nationwide cost of converting to compliant finishing materials, we assumed that 13 percent of the
1.44 billion pounds of fabric per year (i.e., 186 million pounds of fabric per year) reported to be
finished at major facilities for HAP emissions in operations with associated formaldehyde emissions
10-5
-------
would incur the cost of reformulating to low-formaldehyde compliant finishing materials.
Applying the 4 cents incremental cost per pound of finished fabric to use a compliant resin
versus a formaldehyde resin to the estimated 186 million pounds of fabric currently finished with non-
compliant materials yields a nationwide annual cost of $7.5 million per year. The cost of working
with chemical suppliers to reformulate the finish is accounted for in the estimate of the MRR burden
described in the next section of this memorandum.
Monitoring. Reporting and Recordkeeping Costs
Respondents subject to national emission standards for hazardous air pollutants (NESHAP)
are required by law (40 CFR Part 63, Subpart A) to submit one-time notifications and one-time
reports on compliance status and performance test results. Respondents also must develop and
implement a Startup, Shutdown, and Malfunction Plan and make semiannual reports if an event is
inconsistent with the plan. Semiannual reports are required for periods of operation during which
measured emissions exceed an applicable limit or control device operating parameters are outside of
the established ranges. General recordkeeping requirements applicable to all NESHAP require
records of applicability determinations; test results; startup, shutdown, or malfunction events;
exceedances; performance test reports, monitoring records, and all other information needed to
determine compliance with the applicable standard.
Respondents are owners and operators of the 135 printing, coating and dyeing facilities subject
to the requirements of this rule. We estimate that the public MRR burden associated with this
proposed rule will average 213 hours per year per facility for each year after the date of promulgation
of the rule. The total annualized costs associated with MRR have been estimated at $1,403,670; the
total capital costs have been estimated at $1,156,442. Details of the cost estimates are presented in
Reference 4.
Nationwide Compliance Costs of the Printing, Coating, and Dyeing of Fabrics and Other
Textiles NESHAP
Table 10-2 presents a summary of the nationwide compliance costs of the printing, coating,
and dyeing of fabrics and other textiles NESHAP, including the control costs for affected facilities
in the coating and printing subcategory, the finishing reformulation costs for affected facilities in the
dyeing and finishing subcategory, and the MRR costs for all affected facilities in the source category.
We have estimated nationwide capital costs, in 1997 dollars, of approximately $18.8 million and
annual costs of approximately $14.5 million.
10-6
-------
Table 10-2. Summary of Printing, Coating, and Dyeing of Fabrics and Other Textiles
NESHAP Compliance Costs
Nationwide cost component
Coating and printing subcategory control costs
Dyeing and finishing subcategory reformulation
costs
Source category MRR costs
Nationwide total compliance costs
Nationwide total
capital investment,
17.6
1.2
18.8
Nationwide
total annual
cost,
$ x 10 6
5.6
7.5
1.4
14.5
References
1. U.S. Environmental Protection Agency. Fabric Printing, Coating, and Dyeing NESHAP.
ICR Responses. Office of Air Quality Planning and Standards. Research Triangle Park, NC.
Responses received September 1998 - October 1998.
2. Memorandum and Attachment from York, S. and A. Sharma, RTI to P. Almodovar,
EPA/OAQPS/ESD/CCPG. November 13,1998 Final. Summary of meeting at which ATMI
presented the results of the ATMI MACT survey to EPA.
3. Letter, J. Fleming, ATMI, to G. V. Hellwig, EPA: OAQPS: CCPG, November 2, 2000.
ATMI MACT Development Support (textile finishing).
4. U.S. Environmental Protection Agency. Standard Form 83-1 Supporting Statement for OMB
Review of ICR No. Information Collection Request for the Printing, Coating, and Dyeing of
Fabrics and Other Textiles Source Category. Emission Standards Division. Research
Triangle Park, NC. February 27, 2002.
10-7
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1 . REPORT NO. 2.
EPA-453/R-02-010
4. TITLE AND SUBTITLE
Technical Support Document: Printing, Coating, and Dyeing of
Fabrics and Other Textiles Proposed NESHAP
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Coatings and Consumer Products Group (C539-03)
Research Triangle Park, NC 2771 1
12 SPONSORING AGENCY NAME AND ADDRESS
Director
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 2771 1
3 RECIPIENT'S ACCESSION NO.
5. REPORT DATE
June 2002
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO
10 PROGRAM ELEMENT NO.
11 CONTRACT/GRANT NO
13 TYPE OF REPORT AND PERIOD COVERED
Proposed Regulation
14 SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16 ABSTRACT
Under Section 112(d) of the Clean Air Act (the Act), the U.S. Environmental Protection Agency (EPA) is
developing national emission standards for hazardous air pollutants (NESHAP) for the printing, coating, and
dyeing of fabrics and other textiles source category. The purpose of this document is to summarize the
technical background information that supports the proposed NESHAP for the printing, coating, and dyeing
of fabrics and other textiles industry.
1 7 KEY WORDS AND DOCUMENT ANALYSIS
a DESCRIPTORS
Environmental Protection
Administrative Practices and'Prbcedures *
Intergovernmental Relations ' '
Reporting and RecordkeefHrigl&djuirements
18 DISTRIBUTION STATEMENT
Release Unlimited
b IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Hazardous Air Pollutants
Printing; Boating, and Dyeing of
Fabrics tat^d Other Textiles
NES^A'P
19 SECURITY CLASS (Report)
Unclassified
20 SECURITY CLASS (Page)
Unclassified
c COSATI Field/Group
21 NO. OF PAGES
96
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
-------
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Bwlevard, 12th f\m
Oicafo, It 60604-3590
------- |