£EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
Air
Flexible Vinyl and
Urethane Coating
And Printing—
Background
Information for
Promulgated
Standards
EPA-450/3-81-016b
June 1984
Final
EIS
-------
EPA-450/3-81-016b
Flexible Vinyl and Urethane Coating and
Printing—Background Information for
Promulgated Standards
Emission Standards and Engineering Division
U.S. Environment-! Protection Agency
Region V i :•..-..-», 5 *
230 SoLth L: , ',- .) jtreet .^'
Chicago, ii;.r.c.i& OCo04
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
June 1984
-------
This report has been reviewed by the Emission Standards and Engineering Division of the Office of Air Quality Planning
and Standards, EPA, and approved for publication. Mention of trade names or commercial products is not intended to
constitute endorsement or recommendation for use. Copies of this report are available through the Library Services
Office (MD-35), U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711; or, for a fee, from
the National Technical Information Services, 5285 Port Royal Road, Springfield, Virginia 22161.
-------
for
ENVIRONMENTAL PROTECTION AGENCY
Background Information
Final Environmental Impact Statement
Flexible Vinyl and Urethane Coating and Printing
Prepared
'Jack R. Farmer
Director Emission Standards and Engineering Division
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
(Date/
1. The promulgated standards of performance will limit emissions of
volatile organic compounds (VOC) from new, modified, and reconstructed
flexible vinyl and urethane coating and printing operations.
Section 111 of the Clean Air Act (42 U.S.C. 7411), as amended, directs
the Administrator to establish standards of performance for any
category of new stationary source of air pollution that "... causes
or contributes significantly to air pollution which may reasonably be
anticipated to endanger public health or welfare."
2. Copies of this document have been sent to the following Federal
Departments: Office of Management and Budget; Labor, Health and Human
Services, Defense, Transportation, Agriculture, Commerce, Interior, and
Energy; the National Science Foundation; the Council on Environmental
Quality; members of the State and Territorial Air Pollution Program
Administrators; the Association of Local Air Pollution Control
Officials; EPA Regional Administrators; and other interested parties.
3. For additional information contact:
C. Douglas Bell, Section Chief
Standards Development Branch (MD-13)
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Telephone: (919) 541-5624
4. Copies of this document may be obtained from:
U. S. EPA Library (MD-35)
Research Triangle Park, North Carolina 27711
National Technical Information Service
5285 Port Royal Road
Springfield, Virginia 22161
111
-------
TABLE OF CONTENTS
Section Page
1.0 SUMMARY 1-1
1.1 SUMMARY OF CHANGES SINCE PROPOSAL 1-1
1.2 SUMMARY OF IMPACTS OF PROMULGATED ACTION 1-3
2.0 SUMMARY OF PUBLIC COMMENTS 2-1
2.1 SELECTION OF SOURCE CATEGORY 2-1
2.2 EMISSION CONTROL TECHNOLOGY 2-24
2.3 MODIFICATION AND RECONSTRUCTION 2-42
2,4 ECONOMIC IMPACT 2-44
2.5 SELECTION OF BEST DEMONSTRATED TECHNOLOGY 2-53
2.6 SELECTION OF FORMAT 2-55
2.7 SELECTION OF EMISSION LIMITS 2-56
2.8 LEGAL CONSIDERATIONS 2-60
2.9 TEST METHODS AND MONITORING 2-62
2.10 REPORTING AND RECORDKEEPING 2-65
v
-------
LIST OF TABLES
Table Page
2-1 List of Commenters on the Proposed Standards of
Performance for Flexible Vinyl and Urethane Coating
and Printing Operations 2-2
2-2 Excess Capacity Starting With Different Capacity
Utilization Rates 2-11
2-3 Estimated Number of New Sources by 1987 2-11
2-4 Average Sales Per Respondent: 1976 - 1982 2-12
2-5 Projections of the WMA Data 2-16
LIST OF FIGURES
Figure Page
2-1 Average Sales Per Respondent Over Time 2-13
-------
1.0 SUMMARY
On January 18, 1983, the Environmental Protection Agency (EPA) proposed
standards of performance for flexible vinyl coating and printing operations
(48 FR 2276) under authority of Section 111 of the Clean Air Act. On
October 11, 1983, the proposal was extended (48 FR 46224) to cover coating
and printing of urethane webs. At promulgation, the source category was
renamed "flexible vinyl and urethane coating and printing." Both Federal
Register notices requested public comments. Twelve comment letters were
received on the original proposal. Four were from vinyl printers, one was
from an ink manufacturer, one from a printer of urethane products, one from
a manufacturer of flexible packaging, one from a manufacturer of carbon
adsorbers, and two from vinyl printing trade associations. The comments
that were submitted, along with responses to these comments, are summarized
in this document. The summary of comments and responses serves as the basis
for the revisions made to the standard between proposal and promulgation.
No comments were received concerning extension of the proposal to cover
coating and printing of urethane webs.
1.1 SUMMARY OF CHANGES SINCE PROPOSAL
In response to public comments received on the proposed rulemaking and
as a result of EPA revaluation, six major changes were made in the proposed
standard prior to promulgation. The changes involved (1) expanding the
applicability of the standard to include printing on urethane sheets or
urethane coated webs as well as on vinyl sheets or vinyl coated webs,
(2) the addition of considerations for petitioning the Administrator to show
compliance using an inventory-type system to account for VOC content of
waterborne inks, (3) the addition of compliance provisions for incineration,
(4) the revision of the averaging period for compliance using waterborne
inks, (5) the addition of semiannual reporting requirements for conditions
when monitored parameters exceed levels observed during a performance test,
and (6) redefining "flexible vinyl and urethane products" to exempt flexible
packaging.
1-1
-------
The title of the standard, the definition of affected facility in
Section 60.580, and various definitions in Section 60.581 were revised for
clarification and to reflect the application of the standard to printing of
urethane sheets or urethane coated webs. A major urethane manufacturer
commented that there were no-technical or scientific reasons for the
printing of flexible urethane coated fabrics or urethane sheets to be
regulated any differently than the printing of flexible vinyl coated fabrics
or vinyl sheets. Urethane and vinyl products are commonly printed as a
continuous web on rotogravure equipment. The standard, which reflects the
best demonstrated technology (BDT) for the vinyl industry, also reflects BDT
for the urethane industry. Thus, the applicability of the standard has been
expanded to include the printing of flexible urethane coated fabrics or
urethane sheets.
A list of considerations for petitioning the Administrator for
permission to use an inventory-type system to account for volatile organic
compounds (VOC) in waterborne inks was added in Section 60.583(c). Such a
system would allow the operator to account for the VOC and ink solids
content using inventory and purchase records for the entire averaging period
rather than having to account for each ink as used at the print head. Such
a system would also facilitate the use of recycled waterborne inks by
allowing the operator to determine VOC and ink solids content for new inks
and diluent solvent directed to the print line during the averaging period
and to subtract the VOC discarded with unused ink.
For compliance using waterborne inks, the averaging period was
redefined in Section 60.583(b). This change was made to allow plants that
keep records on a quarterly system where each quarter consists of two 4-week
periods and one 5-week period to use a 5-week averaging period four times
per year.
Provisions have been added in Section 60.583(d) for compliance with the
VOC percent reduction standard using incineration. Accordingly, monitoring
requirements for both thermal and catalytic incineration have been added to
Section 60.584(b) and (c). The inclusion of provisions for incineration
give vinyl and urethane printers the option to choose an incinerator as a
1-2
-------
control device for operations using solvents that are not compatible with
carbon adsorption.
Finally, Section 60.585 was added containing semiannual reporting
requirements. The section states that owners or operators must report
semiannually any exceedances of the waterborne ink limitations or deviations
from the monitored parameters for carbon adsorption or incineration listed
in Section 60.584. Reporting requirements were added to facilitate
enforcement of the standard.
1.2 SUMMARY OF IMPACTS OF PROMULGATED ACTION
1.2.1 Alternatives to the Promulgated Action
Regulatory alternatives considered in the development of the standard
are discussed in Chapter 6 of "Flexible Vinyl Coating and Printing
Operations - Background Information for Proposed Standards," EPA-450/3-81-
016a, January 1983. [This document is also referred to as the Background
Information Document (BID).] These regulatory alternatives reflect
different levels of emission control. One was selected that represents the
best demonstrated technology, considering costs, nonair quality health, and
environmental and economic impacts for flexible vinyl coating and printing
operations. These alternatives remain the same.
1.2.2 Environmental Impacts of Promulgated Action
The environmental impacts of the proposed standard are discussed in
Chapter 7 of the BID. A review of these environmental impacts indicated no
changes were necessary, and therefore, the impacts remain unchanged since
proposal. The Agency has not performed an extensive investigation of the
environmental impacts of including printing on urethane products. However,
considering the similarities in the printing processes, the Agency has no
information which indicates that significant adverse impacts will occur as a
result of including urethane products. Public comments were solicited on
this issue, but none were received.
The review of the environmental impacts discussed in the proposal BID
constitutes the final Environmental Impact Statement.
1-3
-------
1.2.3 Energy and Economic Impacts of the Promulgated Action
The energy and economic impacts of the proposed standard are discussed
in Chapters 7 and 9, respectively, of the BID. These impacts reflect the
use of carbon adsorption control equipment. A review of the energy and
economic impacts indicated no changes were necessary, and therefore, these
impacts remain unchanged since proposal. Compliance provisions have been
added for the use of incineration. For facilities that meet the standard
using incineration, energy requirements and costs will increase but are
still considered reasonable (IV-B-8).
As discussed in the previous section, EPA has not performed an
extensive investigation of the energy and economic impacts of expanding the
applicability of the standard to urethane products. Again, considering the
similarities of the processes and the industries, the Agency has no
indication of significant adverse impacts. Public comments were solicited
on this issue.
1.2.4 Other Considerations
1.2.4.1 Irreversible and Irretrievable Commitment of Resources. This
impact is discussed in Chapter 7 of the BID and remains unchanged since
proposal.
1.2.4.2 Environmental and Energy Impacts of Delayed Standards. These
impacts are discussed in Chapter 1 of the BID and remain unchanged since
proposal.
1-4
-------
2. SUMMARY OF PUBLIC COMMENTS
The list of commenters, their affiliations, and the EPA docket number
for each of the comments are shown in Table 2-1. Twelve letters commenting
on the proposed standard and the Background Information Document (BID) were
received. A public hearing was not held because it was not requested. The
comments have been combined into the following 10 categories:
2.1 Selection of Source Category
2.2 Emission Control Technology
2.3 Modification and Reconstruction
2.4 Economic Impact
2.5 Selection of Best Demonstrated Technology
2.6 Selection of Format
2.7 Selection of Emission Limits
2.8 Legal Considerations
2.9 Test Methods and Monitoring
2.10 Reporting and Recordkeeping
The comments and responses are discussed in the following sections of
this chapter. A summary of changes made in the regulation is included in
Chapter 1. The docket reference is indicated in parentheses in each
comment.
2.1 SELECTION OF SOURCE CATEGORY
2.1.1 Comment: One commenter (IV-D-7) writing on behalf of the Chemical
Film and Fabrics Association (CFFA), a trade association representing
60 percent of the flexible vinyl coating and printing industry, felt that
EPA had failed to show that emission controls for the vinyl industry are
CFFA's comments were adopted by the Wallcovering Manufacturers Association
(IV-D-9) and by Uniroyal, Inc. (IV-D-10).
2-1
-------
TABLE 2-1. LIST OF COMMENTERS ON THE PROPOSED STANDARD OF
PERFORMANCE FOR FLEXIBLE VINYL AND URETHANE COATING
AND PRINTING OPERATIONS
Commenter Docket Entry Number9
Mr. Henry R. Lasman IV-D-1
Compo Industries, Inc.
200 Market Street
Lowell, Massachusetts 01852
Mr. D.R. Roeing IV-D-2
Croda Inks Corporation
777 North Merimac Avenue
Miles, Illinois 60648
Mr. Robert K. Pfeiffer IV-D-3
Union Carbide Corporation
P.O. Box 49287
Atlanta, Georgia 30359
Mr. L.C. Tropea, Jr. IV-D-4
Reynolds Aluminum
Reynolds Metals Company
Richmond, Virginia 23261
Mr. Donald Dean IV-D-5
Imperial Manufacturing Company
P.O. Box 280
Pittsburgh, New York 12901
Mr. C. Richard Springer IV-D-6
Borden, Inc.
165 N. Washington Avenue
Columbus, Ohio 43215
Mr. W. Caffey Norman IV-D-7
Chemical Film and Fabrics Association
Cleary, Gottlieb, Steen and Hamilton
1752 N Street, N.W.
Washington, D.C. 20036
Mr. F.P. Partee IV-D-8
Ford Motor Company
One Parklane Boulevard
Dearborn, Michigan 48126
(continued)
2-2
-------
TABLE 2-1. CONTINUED
Commenter Docket Entry Number9
Mr. Robert E. Hebda (adopted comments in D-7) IV-D-9
Mr. David C. Murchison
Wallcovering Manufacturers Association
Howrey and Simon
1730 Pennsylvania Avenue, N.W.
Washington, D. C. 20006
Mr. Robert C. Niles (adopted comments in D-7) IV-D-10
Uniroyal, Inc.
Middlebury, Connecticut 06749
Mr. Richard T. Paul IV-D-13
Wallcovering Manufacturers Association
666 Morris Avenue
Springfield, New Jersey 07081
Mr. Charles M. Stockinger IV-D-14
Chemical Film and Fabrics Association, Inc.
1230 Keith Building
Cleveland, Ohio 44115
These designators represent docket entry numbers for docket number A-80-8.
These documents are available for public inspection at: U.S. Environmental
Protection Agency, Central Docket Section, West Tower Lobby, Gallery 1
Waterside Mall, 401 M Street, S.W., Washington, D.C. Docket entries
IV-D-11 and IV-D-12 are not included in this list because these comments
were received and responded to before proposal of the standard.
2-3
-------
needed or will result in any significant reduction of air pollution which
may reasonably be anticipated to endanger public health or welfare. He felt
that there is no statutory authority for proceeding with the development of
this new source performance standard (NSPS) because little or no new growth
is expected and the source is not a "significant contributor" to air
pollution.
The association (IV-D-7) made several points in their letter regarding
growth of the industry. The association's representative noted that EPA's
action in discontinuing the development of an NSPS for the mineral wool
industry provides a precedent for discontinuing action on the vinyl printing
NSPS. He explained that the overall growth trend in the industry is
negative and U.S. market shipments [Standard Industrial Code (SIC) 2295]
declined steadily from 1972 to 1980. He felt that the BID's estimate of six
new sources in the wallcovering segment which accounts for 7 percent of the
industry sales was based on the questionable assumptions that the 60 percent
1979 average capacity utilization would not change and that the annual
growth rate would be 8.9 percent. He stated that it would have been more
reasonable to assume that the 1979 excess capacity would have been utilized
as has been the case from 1979 to 1982. His members have reported a decline
in wallcovering production and sales during this period. No new sources
have been constructed and several advanced print lines in the industry have
been permanently shut down.
The CFFA's representative (IV-D-7) also pointed out that, according to
the Control Techniques Guideline (CTG) for fabric coating, VOC emissions for
vinyl coating currently represent less than 0.2 percent of nationwide VOC
emissions. He noted that this is a negligible contribution to nationwide
VOC emissions, and the increase in VOC emissions of even six new sources in
the vinyl industry over the next 5 years would appear to be insignificant
for the purpose of EPA's determination of whether NSPS should be developed
for the industry.
For these reasons, CFFA suggested that EPA reconsider the need for the
proposed standard.
2-4
-------
After the public comment period closed, the Wallcovering Manufacturers
Association (WMA) commented (IV-D-13) regarding perceived inaccuracies in
the growth projections. Their comments, with which CFFA later agreed
(IV-D-14), were based on EPA's post-proposal docket entry (IV-B-2) adding
new data on growth projections supplied by WMA and EPA's analysis of these
data. WMA addressed six areas of concern (IV-D-13).
1. EPA's background information document (BID) overstated the number
of firms which might enter the wallcovering industry in the next 4 years
because it inappropriately used past sales data to project industry capacity
for the period 1979-1987.
2. The 8.9 percent growth rate hypothesized in the BID is overstated.
WMA felt that the growth rate was based on reported industry sales between
1971 and 1979 and thus ignored the downward trend in sales between 1980 and
1982.
3. In determining a projected growth rate, the BID made no effort to
analyze how projected changes in important economic variables might affect
the demand for wallcovering in the 1980s. WMA listed variables such as
competition from other types of wall finishes, growth in imported products,
competitive viability of existing producers, housing starts, nonresidential
construction, and the remodeling market.
4. The updated growth projections (IV-B-2) used an inappropriate
method to calculate the growth in sales over the 1975-1982 period. That
report used an average of each individual percentage change in sales over
the 1975-1982 period. WMA added that the standard compound growth rate is
more accurate for forecasting purposes and noted that the compound method
was used in the BID.
5. The updated growth projections (IV-B-2) misused the data supplied
by WMA. That report incorrectly assumed that the total number of companies
reporting to WMA all produced vinyl wallcovering. They did not. Therefore,
WMA felt that the projections (IV-B-2) incorrectly calculated the average
sales of vinyl wallcovering per respondent. WMA included a table showing
the number of companies which reported vinyl wallcovering sales and the
actual average sales per respondent.
2-5
-------
6. Inconsistencies exist between the BID and the updated growth
projections (IV-B-2). Specifically, WMA listed the number of wallcovering
firms, the wallcovering production capacity for 1982, and the average
capacity per firm in the industry.
WMA concluded that both the BID and the updated growth projections
(IV-B-2) were incorrect in projecting the entry of six new firms from 1982
through 1987. Rather, WMA felt that it is unlikely that any new firms will
enter the wallcovering industry in the next few years. WMA felt that
regulatory action should not be based on either the BID or the updated
growth projections (IV-B-2).
The CFFA's second letter (IV-D-14) voiced concurrence with WMA, adding
that an error was made in the BID regarding statistical information provided
by CFFA on the aggregate value of wallcovering shipments. CFFA noted that
they provided EPA with the aggregate value of wallcovering shipments by
reporting members, and estimated that its members represented approximately
60 percent of the overall vinyl coating and printing industry. CFFA stated
that EPA had twice increased the value of shipments to reflect the total
market, first on page 9-13 of the BID and again on page 9-20. CFFA added
that this error would change the forecast of the number of new sources
projected for the industry. They projected an increase of only one new
source. CFFA concluded that they joined with WMA in the feeling that
regulatory action should not be based on the growth projections in the BID
or the analysis in IV-B-2.
Response: There is statutory authority and, indeed, a requirement to
proceed with the promulgation of this NSPS. Section 111 of the Clean Air
Act requires that the Administrator establish standards of performance for
categories of new, modified, or reconstructed stationary sources that in the
Administrator's judgment cause or contribute significantly to air pollution
that may reasonably be anticipated to endanger public health or welfare.
Standards of performance prevent new air pollution problems from developing
by requiring application of the best technological system of continuous
emission reduction that the Administrator determines to be adequately
demonstrated. The 1977 Amendments to the Clean Air Act added the words, "in
2-6
-------
the Administrator's judgment," and the words, "may reasonably be
anticipated," to the statutory test. The legislative history for these
changes stresses two points.
The Act is preventive, and regulatory action should be taken to
•
prevent harm before it occurs; and
the Administrator should consider the cumulative impact of all
sources and not just the harm from a single class of sources.
[H.R. Rep. No. 95-294, 95th Congress, 1st Session, 49-50 (1977).]
The 1977 Amendments to the Clean Air Act also required that the
Administrator promulgate a priority list of source categories for which
standards of performance are to be promulgated. The priority list,
40 CFR 60.16, was promulgated in the Federal Register August 21, 1979
(44 FR 49225). Development of the priority list was initiated by compiling
data on a large number of source categories from literature sources. Major
stationary source categories were then subjected to a priority ranking
procedure using the three criteria specified in Section lll(f) of the Act.
The procedure ranks source categories on a pollutant-by-pollutant basis. In
this ranking, first priority was given to the quantity of emissions, second
priority was given to the potential impact on health or welfare, and third
priority was given to the mobility and competitive nature of the source
category.
In light of the considerations stated above, the Administrator found
that the fabric coating industry is a "significant contributor." (Applying
the criteria for prioritizing such contributors, the Administrator ranked
the fabric coating industry, of which vinyl printing is the largest organic
solvent using segment, 10th of 59 source categories on the priority list.)
This listing decision requires the Agency to promulgate standards of
performance for new sources in the category.
In regard to growth anticipated in the flexible vinyl printing
industry, EPA evaluated the growth rate using the WMA data (IV-B-2) and
still maintains the position described in a letter to CFFA in March 1982
(II-C-89); that is, the short-term growth picture for wallcovering plants
has worsened during the last couple of years, as is true for many
2-7
-------
manufacturing industries in today's economy. Decline in housing starts and
construction of commercial buildings has certainly been the major factor in
slowing the production of wallpaper. However, before the market began to
slump, there was considerable growth in wallcovering shipments; and it can
be expected to continue when the Nation's economy recovers. Therefore, the
Agency believes that growth projections based on today's economic conditions
and on utilization of existing plant capacity are inappropriate. Instead,
longer-term estimates should more accurately reflect the growth potential
for this industry. In NSPS development, a 5-year period is generally
selected on which to base environmental, economic, and energy analyses.
This standardizes the reporting of impacts in the regulation development
process, but should not be viewed as a limiting factor in EPA's decision as
to whether a particular standard is warranted. Therefore, predicting on the
basis of today's economy the number of wallcovering plants which will be
built in the next 5 years, is not as relevant as the fact that long-term
growth is expected to be strong, indicating that some number of plants or
new lines will definitely be constructed in the future.
The mineral wool industry was deleted from the priority list
(47 FR 950, January 8, 1982) because, after a rapid industry expansion in
the late 1970's, market conditions in the early 1980's indicated that an
anticipated further increase in demand had not developed and that the
industry was operating at about 60 percent of capacity. EPA estimated that
only one new mineral wool plant might be built before 1985. In addition,
there were no expansions or modifications expected to occur in the mineral
wool industry during the 5-year period. On the other hand, some commenters
(IV-D-7, IV-D-8) gave an indication that process modifications and
replacement of certain pieces of equipment (e.g., drying ovens) will occur
in the vinyl printing industry, thus submitting these sources to the NSPS.
Therefore, the Administrator does not agree with the commenter's statement
that the mineral wool industry provides a precedent for discontinuing action
on the proposed NSPS for flexible vinyl coating and printing.
In regard to the use of excess capacity in the vinyl industry, EPA has
no evidence to indicate that technological improvement will significantly
2-8
-------
reduce downtime or that any resulting increase in production will offset
production ascribed to new plants. The Agency expects, as the economic
recovery and the demand for wallcovering increases, the wallcovering
industry will continue to experience a considerable amount of downtime
•
ranging from 40 to 75 percent. This downtime is required for color changes,
quality control, and cleanup.
The background information document (BID) quotes the CTG stating that
uncontrolled vinyl printing emissions are about 36,000 megagrams/year
(40,000 tons/year). During the development of the BID, potential uncon-
trolled emissions were estimated as 61,700 megagrams/year (68,000 tons/
year). Nationwide in the year 1987, the NSPS would reduce VOC emissions
from new, modified, and reconstructed print lines by 790 megagrams
(870 tons) beyond the emission level required by State regulations. No
stationary source category accounts for more than about 2 or 3 percent of
the Nation's total VOC emissions, and most contribute 1 percent or less.
Thus, nationwide reductions in VOC emissions can be achieved only by
controlling many different sources which may be small sources in absolute
terms, but large relative to other VOC emitters.
The following discussion responds to the six areas of concern expressed
in WMA's letter (IV-D-13):
1. EPA incorrectly projected six new firms. EPA did not project six
new firms, but rather projected six new sources. Neither the background
information document (BID) nor the updated growth projections (IV-B-2) ever
mentions firms. A new source means a new rotogravure printing line, not a
new firm or even necessarily a new plant of an existing firm.
The actual number of new sources could vary between two and six,
depending on the assumptions one accepts. The number of new sources is a
function of two critical assumptions: the starting capacity utilization
rate for the industry and the average capacity utilization rate of a new
source. In the BID, the starting rate and average rate were assumed equal
based on engineering estimates of the downtime of a new line for color and
ink changes, etc. (see page 6-3 of the BID). Nonetheless, varying the
assumptions does not significantly alter the conclusions as a simple
sensitivity analysis reveals.
2-9
-------
Table 2-2 displays excess capacity as a function of different capacity
utilization rates for the projection period. Assuming that the average new
source would have a capacity of 15 million square meters; that the average
capacity utilization of a new source would be 100 percent (the assumption
made by WMA); and that the starting capacity utilization rate is 60 percent;
then two new sources would be needed by 1987. A further assumption that the
average capacity utilization rate can vary as well as the starting rate
results in the number of new sources for each scenario by 1987 as shown in
Table 2-3.
Given that the data supplied by WMA imply a starting capacity
utilization rate of 72 percent in 1979 (Table 2-4), then the number of new
sources is likely to be between three and five (i.e., a starting capacity
utilization rate between 70 and 80 and a average capacity utilization rate
per new source between 80 and 100 (Table 2-3). Although this interval
estimate of three to five new sources may be preferable to the point
estimate of six new sources reported in the BID, the basic conclusion that
there will be new sources is unaltered.
2. The growth rate of 8.9 percent used in the BID is overstated. WMA
claims that the growth rate is overstated because EPA failed to take account
of the downward trend in 1980-82. First, data for 1980-82 were not
available when the BID was submitted. Second, the data supplied by WMA
actually yield a higher growth rate of 9.12 percent when 1980-82 data are
included, as shown in Table 2-4.
Table 2-4 indicates how volatile or changeable the WMA data are,
suggesting that the recent decline can reverse itself rather quickly. For
this reason, the average annual rate of change (9.12 percent) is used rather
than a compound rate. A compound rate is appropriate only when the data are
relatively smooth and change in an exponential fashion. In this case, the
data are too volatile to be properly represented in this manner, and the
average annual percentage change is more appropriate.
Figure 2-1, which graphically portrays the average sales per respondent
over the 1975-82 period, illustrates the different results obtained when a
compound rate is used. The compound rate of 6.7 percent computed by WMA
2-10
-------
TABLE 2-2. EXCESS CAPACITY STARTING WITH DIFFERENT
CAPACITY UTILIZATION RATES
Year
1979
1980
1981
1982
1983
1984
1985
1986
1987
Output Excess Capacity
(Millions Of Starting With A Capacity Utilization Of
Sq. Meters) 60%
77.78 51.8
84.70 44.9
92.24 37.4
100.45 29.2
109.39 20.2
119.13 10.5
129.73 (0.1)
141.27 (11.7)
153.85 (24.3)
TABLE 2-3. ESTIMATED NUMBER
70% 80%
33.3 19.4
26.4 12.5
18.9 5.0
10.7 (3.3)
1.7 (12.2)
(8.0) (21.9)
(18.6) (32.5)
(30.2) (44.1)
(42.8) (56.7)
OF NEW SOURCES BY 1987
90%
8.6
1.7
(5.8)
(14.1)
(23.0)
(32.7)
(43.3)
(54.9)
(67.5)
Average
Starting
Capacity Utilization 60%
80% 2
85% 2
90% 2
100% 2
Excess Capacity
At A Capacity Utilization Rate Of
70% 80%
4 5
3 5
3 4
3 4
90%
6
5
5
5
aThe average new source is assumed to have a capacity of 15 million square
meters.
2-11
-------
TABLE 2-4. AVERAGE SALES PER RESPONDENT: 1976 - 1982
Average Sales Per Percent Change
Year Respondent (In Rolls) From Previous Year
1976
1977
1978
1979
1980
1981
1982
2,131,894
2,943,969
2,960,263
2,963,664
3,820,194
3,053,089
2,485,833
Average
34.75%
38.09
0.55
0.12
28.90
-20.08
-18.58
9.12%
Implied Capacity
Util ization
52%
71
71
72
92
74
60
79%
2-12
-------
Figure 2-1. Average Sales Per Respondent Over Time
3,900,000
3,660,000
3,420,000
3,180,000
2,940,000
2,700,000
2,460,000
2,220,000
1,980,000
1,740,000
1,500,000
An Alternative Compound
Growth Rate of 19.3%
f
/ WMA Compound
Growth Rate of 6.7%
1975 1976 1977 1978 1979 1980 1981 1982
Time
2-13
-------
completely ignores all but two data points and, therefore, is misleading.
WMA estimated the rate by taking the first data point (1975) and the last
data point (1982) and solving for the compound rate. Such a method ignores
the critical information in between, e.g., the 1980 high of 3,820,194. This
rate cannot be taken as representative of the 1975-82 period. An alterna-
tive would be to compute the compound rate between 1975 and 1980. This rate
of 19.3 percent captures more of the trend than the WMA formulation.
However, this is not representative either. The most representative rate to
employ is the average annual percentage change, a method which incorporates
all the data points including the relatively flat performance in 1977-1979.
The average annual percentage change is 9.12 percent (see Table 2-4).
3. EPA made no effort to analyze important economic variables which
might affect the demand for wallcovering in the 1980's. The type of
analysis suggested by WMA would require a rather sophisticated econometric
modeling effort. In addition, the costs of such an undertaking are likely
to exceed the benefits considerably. Research on short-run forecasting
models and experience suggest that sophisticated econometric forecasting
models are often outperformed by simple linear projections. The simple
projection used in the BID has the advantage of incorporating most of the
variables WMA listed as important indirectly by employing a growth rate
based on past data.
4. The updated growth projections (IV-B-2) used an inappropriate
method to compute the growth rate. WMA indicated that the standard compound
growth rate method was more accurate in computing a growth rate for fore-
casting purposes and noted that this was the method used in the BID.
However, as previously discussed in paragraph 2, due to the volatility of
the data submitted by WMA, EPA feels the average annual percentage change
method used in the updated projections is appropriate.
5. The growth projections (IV-B-2) misused the data supplied by WMA.
The data supplied by WMA were not misused. Rather, WMA did not provide all
the necessary caveats concerning their data until after the growth
projections were revised. When the data were first provided, the only
caveat issued was that the data were based on a voluntary survey, and thus,
-------
the number of respondents fluctuates from year to year. (This is the reason
for using an average per respondent.) No mention was made of the
reliability problem, i.e., the data are unreliable until the 1984 survey
provides tentative 1983 data along with a more reliable 1982 sales figure.
Accordingly, perhaps the 1982 figure should be disregarded. If it is, and
if the compound growth rate is computed using the method espoused by WMA,
then the growth rate would be 11.6 percent (or 13.7 percent using EPA's
averaging method) rather than 9.12 percent. Also, no information was
provided concerning how many WMA members were vinyl wallcovering producers.
At the time WMA supplied the data, the Agency assumed that most of its
members were vinyl wallcovering producers since there were no indications to
the contrary. Their letter (IV-B-13) indicates, however, there are only
about 20, not 55, producers.
Because the WMA data are highly sensitive to the number of reporting
companies and because the mix of the reporting companies (i.e., large vs.
small, those having a banner year vs. those not having a banner year, etc.)
may be a function of the business cycle (and thus result in a volatile data
set), EPA has severe reservations about using the WMA data. Nevertheless,
if the data are used, results show between two and seven new sources, as
demonstrated below.
Since no data with respect to operating rates have been provided to EPA
since the BID was developed, a compromise might be to use the averages of
the implied capacity utilization rates in Table 2-4. These rates are
implied if one starts with 60 percent in 1982. The implied average is
70 percent and this is taken as the more likely rate in 1982.
Table 2-5 shows the projected WMA data. Assuming 20 (rather than 55)
producers of vinyl wallcoverings with an average capacity of 3,551,190 rolls
(total supply = 72,023,800) and that demand (2,485,833 x 20 = 49,716,660)
increases at 9.12 percent, then there will be a short fall of 5.9 million
rolls in 1987. This translates into two new sources. With the 11.6 percent
growth rate (since the 1982 data are tentative), between four and five new
sources are predicted. With the 13.7 percent growth rate (using the more
appropriate averaging method), about seven new sources are predicted. Thus,
2-15
-------
TABLE 2-5. PROJECTIONS OF THE WMA DATA
Year
1982
1983
1984
1985
1986
1987
Demand
(In Rolls)
49,716,660
54,250,819
59,198,494
64,597,397
70,488,679
76,917,247
Supply
(In Rolls)
71,023,800
71,023,800
71,023,800
71,023,800
71,023,800
71,023,800
Excess
Demand
5,893,447
2-16
-------
even with these rather volatile data, between two and seven new sources
would still be forecast.
6. Inconsistencies exist between the BID and the updated growth
projections (IV-B-2). There are a few inconsistencies between the BID and
the updated growth projections because two different data sets for two
different time periods were supplied by two different trade groups. The BID
used deflated dollar sales data for the 1971-1979 period supplied by the
Chemical Film and Fabric Association, while the updated growth projections
(IV-B-2) used the number of rolls of vinyl wallcovering for the 1975-1982
period supplied by WMA,
The other apparent inconsistencies that WMA believes exist, however,
result from WMA's confusion between a new firm and a new source or plant.
WMA pointed out that the updated growth projections (IV-B-2) used 55
producers while the BID used 11. The reasons for using 55 producers are
discussed previously in paragraph 5. Moreover, the BID did not say there
were only 11 producers. The number 11 was inferred by WMA by incorrectly
dividing total capacity by the average size of a new source. This same
confusion explains the apparent inconsistencies regarding average and total
industry capacity.
CFFA noted in a second comment letter (IV-D-14) that statistical
information on the aggregate value of wallcovering shipments presented in
the BID had been increased incorrectly to reflect the entire market.
Indeed, the value of shipments supplied by CFFA and reflecting 60 percent of
the total market was increased to reflect the total market and then
incorrectly increased a second time in a subsequent calculation presented in
the BID. The total 1979 value of shipments (p. 9-13 and 9-20 in the BID)
should be $210 million'.
The error does not change the 8.9 percent growth rate, but it does
result in a revised estimate of new sources. Depending on the capacity
utilization rate used, the estimate of new sources ranges from one to three.
The estimate of one new source is based on a very high capacity utilization
rate of 97 percent. Three new sources would be forthcoming if a conserva-
tive capacity utilization rate of between 70 and 80 percent is used as
2-17
-------
suggested by WMA. However, an average capacity utilization rate of
60 percent was used during the development of the BID and the Administrator
feels this rate is still appropriate. A capacity utilization rate of
60 percent would result in the installation of more than three new sources.
In summary, the basic conclusions of the BID are unaltered. Varying
both the starting and the average capacity utilization rates do nothing more
than refine the estimate to be three to six new sources rather than a point
estimate of six. If the WMA data were used given the accompanying caveats
and volatility, between two and seven new sources would still be predicted.
Despite the error made in the BID regarding the value of shipments, the
analyses show that there will be growth in this industry. The prediction of
how many wallcovering lines will be built in the next 5 years is not as
relevant as the fact that long-term growth is expected to be strong,
indicating that some number of sources will be constructed in the future.
In addition to new sources, the industry itself estimates that some existing
sources anticipate making process and equipment modifications and replacing
certain pieces of equipment; thus, these sources would be subject to the
standard.
Also, the standard has other benefits in addition to reducing emissions
beyond those levels required by current State regulations. Standards of
performance establish a degree of national uniformity, which precludes
situations in which some States may attract industries by relaxing air
pollution standards relative to other States. They improve the efficiency
of case-by-case determinations of Best Available Control Technology (BACT)
for facilities located in attainment areas and Lowest Achievable Emission
Rate (LAER) for facilities located in nonattainment areas, by providing
documentation and a starting point for the basis of these determinations.
This documentation results from the process of developing a standard of
performance, which involves identification and comprehensive analysis of
alternative emission control technologies, development of associated costs,
evaluation and verification of applicable emission test methods, and identi-
fication of specific emission limits achievable with alternate technologies.
For these reasons, VOC emissions from rotogravure printing of flexible vinyl
and urethane have been selected for regulation under an NSPS.
2-1
o
-------
2.1.2 Comment: One commenter (IV-D-6) requested that EPA reconsider the
need for the regulation due to the substantial costs being imposed. He felt
the costs would be in excess of those mentioned in the BID, but did not
provide any information to support this statement.
Response: Cost analyses and impacts based on carbon adsorption control
technology are summarized in the BID and the preamble to the proposed
standard. Other comments (2.4.6 and 2.4.2.3) asked that EPA consider the
effect of potential wastewater discharge surcharges and charges for off-site
solvent recovery. Responses to these comments explain that these factors
were considered. EPA has no additional cost information which would change
the economic analysis or the assessment that costs of carbon adsorption
control are reasonable.
As a result of several comments, EPA has analyzed the cost
effectiveness of incineration as an alternative means of VOC control in this
industry. This analysis appears at (IV-B-8). Costs of incineration were
determined to be reasonable and compliance provisions for incineration were
added to the regulation.
2.1.3 Comment: One commenter (IV-D-10) recommended that the NSPS be
renamed "vinyl wallcovering" with the vinyl printing and coating industry
addressed in a separate and applicable series of documents at a later date.
He listed five reasons for this position. First, the only print line tested
to demonstrate capture and control technology was printing wallpaper.
Second, the only projected growth is in the wallpaper area. Third, there
are fundamental differences between print lines utilized for wallpaper and
other vinyl coated products such as print width and line speeds. Similarly,
another commenter (IV-D-5) stated that the model plant parameters used to
determine emission control standards were not representative of the entire
wallcovering industry. Fourth, the control system of choice for wallpaper
systems is carbon adsorbers and is the only system discussed in the BID.
Other vinyl products operators would probably use an incinerator as a
control device of choice because of multiple solvents, carbon blinding due
2-19
-------
to plasticizers and operational flexibility. Finally, State Implementation
Plan (SIP) regulations would require control substantially equal to the
NSPS. He went on to explain that the use of a 65 percent emission reduction
for the baseline is in error because 95 percent of existing facilities are
required by SIPs to reduce emissions by 81 percent or better. A proper
baseline would be 81 percent reduction, as reflected in most SIPs. Using an
81 percent emission reduction baseline requires that the environmental,
energy and economic impacts be recalculated as they are grossly over-
estimated as presented.
Response: The standard appropriately applies to the entire vinyl
printing and coating industry. In regard to the commenter's remark that the
only print line tested was printing wallcovering, the rotogravure printing
processes are substantially the same with respect to fugitive emission
generation and capture, regardless of the vinyl printed product produced.
Factors which affect the emissions from a printing line include equipment
design parameters such as configurations of the dryers and print stations,
air velocities at the fugitive capture points, and process and product
parameters such as solvent volatilities, resin types, web widths, line
speeds, and run lengths. The most important equipment design factor in
reducing VOC emissions by improved capture efficiency is design of the print
heads and dryer to minimize exposure of the wet web as it travels from the
print head to the dryer. To accomplish this, the print head can be almost
completely enclosed by moveable doors or panels that can be readily opened
for operator access. If these doors are panels made of transparent
material, the print head will still be visible to the operator. The dryer
can be designed so that the opening is very close to the point where the wet
web leaves the print head. Air flows must be balanced so that significant
velocity is maintained where the web leaves the print head and enters the
dryer to assure that emissions enter the dryer rather than escape at this
point. EPA believes that proper design of the print head - dryer relation-
ship and associated air velocities are the key to achieving the demonstrated
control efficiencies and that other parameters such as product type, web
widths, line speeds, etc., have little effect on performance.
2-20
-------
In a meeting of the Chemical Film and Fabrics Association (CFFA) and
EPA (II-E-101), there appeared to be agreement that the main factor which
affects the emissions from any flexible vinyl printing and coating print
line is the design of the capture system. If the capture system is properly
designed and operated, other equipment design and product parameters become
much less important. Since the affected facility which prints wallcovering
is the same rotogravure print line used to print other flexible vinyl
products, a wallcovering rotogravure print line is representative of the
industry. Most CFFA members present at the meeting apparently agreed with
this position although disagreement was later expressed by CFFA (II-D-89).
(EPA's response to the CFFA letter II-D-89 is found at II-C-73.) No test
data have been presented to show that EPA's conclusion is inappropriate.
EPA visited several flexible vinyl coating plants (II-B-7, II-B-9,
II-B-11, II-B-13 through 19, II-B-21, II-B-22, II-B-24, II-B-25) that
manufactured a variety of vinyl printed products including wallcovering,
furniture upholstery, automobile upholstery, window shades, and awnings.
Information from these plant visits supported the conclusions that capture
system design and operation is indeed the key to achieving 90 percent
capture and that tests of a wallcovering print line are representative of
the entire vinyl printing industry. Furthermore, EPA's experience with
similar industries such as publication rotogravure printing indicates that
capture efficiencies of 90 percent and greater have been demonstrated
(IV-B-5).
The commenter's second point on industry growth has been discussed in
the response to Comment 2.1.1. The commenter's third point addressed the
effect of variation in product parameters such as print width and line
speeds. As discussed above, the proper design of the capture system, rather
than parameters such as web width and line speed, is the most important
factor to achieving the demonstrated control efficiency. The Agency judges
that such parameters do not significantly affect the overall VOC control
efficiency demonstrated in the test on the wallcovering print line.
2-21
-------
In response to the fourth point raised by the commenter, incineration
and carbon adsorption were both described in Chapter 4 of the BID. Since
carbon adsorption was the only control method in use in the industry during
the development of the standard, the cost analyses in Chapter 8 were based
on carbon adsorption. According to comments received after proposal of the
NSPS, incineration is being considered by some vinyl printers and therefore
a more detailed discussion of the use of incineration has been addressed in
other sections, and compliance provisions for incineration have been added
to the standard.
In regard to the commenter's fifth point, EPA does not agree that the
baseline was improperly defined causing the impacts of the standard to be
overestimated. The baseline for the proposed standard, however, was defined
at 65 percent overall control based on the level of control most States,
that must achieve the ambient air quality standard for ozone, were actually
adopting. Although some States had the authority to enforce higher
reduction requirements, it appeared that most were enforcing the 65 percent
level. The 65 percent control level was based on the CTG for graphic arts -
rotogravure and flexography, which States were allowed to apply to control
emissions from the vinyl printing industry.
After studying each of the commenter's five points, the Administrator
still believes that the standard appropriately applies to the entire vinyl
printing and coating industry. The standard will not be limited to vinyl
wallcovering as requested.
2.1.4 Comment: One commenter (IV-D-1) felt that it was a serious oversight
that the proposed regulation covered the printing of flexible vinyl coated
webs without reference to printing or finishing flexible urethane coated
webs. The commenter, whose company produces both vinyl and urethane coated
products, pointed out that the major difference between the two webs is the
chemistry of the coating compounds composing the web. In one instance,
these are plasticized polyvinyl chloride (PVC) compounds and in the other,
flexible thermoplastic urethane polymers. The commenter noted that,
regardless of the chemical composition of the coating, the webs are finished
2-22
-------
by printing with solvent solutions of PVC, urethanes, and/or acrylic resins.
The commenter stated that his company knew of no technical or scientific
reasons why the printing of polyurethane coated fabrics should be regulated
differently from the printing of flexible polyvinyl coated fabrics. He
concluded by suggesting that the proposed standards for printing of flexible
vinyl be revised to specifically include the printing of urethane coated
fabrics.
Response: After some discussion of this issue with the commenter
(IV-E-1, IV-E-7) to get more information on the process of printing on
urethane webs, the Administrator believes that urethane product manufac-
turers can achieve the proposed emission standard at reasonable costs.
Since the Agency had not investigated the environmental, energy, and
economic impacts of the standard upon urethane product manufacturers, notice
of this proposed change appeared in the Federal Register on October 11, 1983
at 48 FR 46224. This notice allowed for a 30-day public comment period and
specifically requested comments regarding the impacts of the standard on
urethane printers. No comments were received. The Administrator has
concluded that the promulgated standard should include the coating and
printing of both flexible vinyl and flexible urethane.
2.1.5 Comment: One commenter (IV-D-4, IV-E-4), a producer of flexible
packaging, felt that the proposed regulation could be erroneously viewed to
apply to sources producing flexible packaging products such as products
consisting of a thin vinyl coating applied to aluminum or paper based
substrates on rotogravure printing equipment. He gave facial tissue boxes
as an example of this. He noted that EPA staff members had told him that
the standard did not apply to flexible packaging. He said that the
definitions of "affected facility" and "flexible vinyl products" hinged on a
minimum thickness of 50 micrometers (0.002 inches), but that some of his
company's flexible packaging products exceeded that thickness. He urged EPA
to revise the final regulation to clearly state that it does not apply to
the manufacture of flexible vinyl packaging. He suggested that this be done
2-23
-------
by specifically excluding "flexible packaging" in the definition of
"flexible vinyl products."
Response: After receiving this comment, EPA contacted the Flexible
Packaging Association (IV-E-5) to ask for a definition of "flexible
packaging." Although the Association representative explained that there
was some disagreement within the Association about an exact definition, he
gave the following definition: "Flexible packaging is the use of paper,
film, or foil to package a product in such a manner that the packaging
material itself generally takes the shape of its contents." The Association
representative stated that most flexible packaging is less than 50 micro-
meters (0.002 inches) thick, but noted that some Association members have
products up to 0.010 inches thick.
Although it appears that the flexible vinyl coating and printing
process is similar to the flexible packaging process described by the
commenter (TV-E-4), EPA did not obtain the technical or economic information
to determine whether best demonstrated technology (BDT) for the flexible
vinyl industry is also BDT for this segment of the flexible packaging
industry. Therefore, the definition of "flexible vinyl and urethane
products" has been revised to specifically exempt flexible packaging from
this standard. However, the Agency is beginning to develop new source
performance standards for the flexible packaging industry and the packaging
process described by the commenter will be investigated in the near future.
If the investigation shows that BDT for flexible vinyl is also applicable to
this segment of the flexible packaging industry, then this standard will be
revised to include this segment of the packaging industry.
2.2 EMISSION CONTROL TECHNOLOGY
2.2.1 Comment: One commenter (IV-D-3), a manufacturer of fluidized-bed
carbon adsorbers, explained some features of his company's product noting
that it could achieve recovery efficiencies of 95 percent or greater. In a
telephone conversation (IV-E-2), this commenter expressed concern that the
2-24
-------
preamble referred to fixed-bed carbon adsorbers as the basis of the
standard.
Response: During the development of the standard, most existing vinyl
printing facilities with carbon adsorbers used multiple fixed-bed adsorbers
and therefore fixed-bed carbon adsorbers were described in the BID. It was
not the Agency's intent to imply that fluidized-bed carbon adsorbers were
not applicable. Methods used for controlling emissions of VOC from affected
facilities are not limited to those discussed in the BID. Other control
technologies may be used to control emissions as long as the emission limit
is met. The Agency interprets this comment as further evidence of the
recovery efficiency achievable by carbon adsorbers.
2.2.2 Comment: After having checked with carbon adsorber manufacturers,
one commenter (IV-D-6) discussed several points with regard to the ability
of carbon adsorbers to function effectively and achieve 95 percent control
efficiency. First, he noted that a combination of methyl ethyl ketone (MEK)
and methyl isobutyl ketone (MIBK) going to an adsorption system is
significantly harder to adsorb and control than a single solvent, such as
toluene. He added that additional controls would be needed to separate
these solvents from water and that for these two solvents, the system must
be constructed of stainless steel in lieu of carbon steel.
The commenter made three observations about manufacturers' efficiency
guarantees of carbon adsorbers. First, he noted that since numerous
operating variables such as humidity, air stream temperature and concentra-
tion, and length of time the carbon bed is in use, are encountered in the
operation of a carbon adsorption system, one company would only guarantee
its unit at 75 percent efficiency. Next, he noted that carbon adsorption
systems would typically be oversized by 50 percent in order for another
company to state that they would be willing to give a guarantee of
95 percent. According to the commenter, that manufacturer said that to
maintain 95 percent efficiency, the carbon bed would need to be replaced in
1 to 1.5 years. He went on to say that because of the size of the carbon
2-25
-------
adsorption system required for a possible approach to 95 percent efficiency,
EPA's cost estimate was not realistic. He felt a realistic installed cost
for a new carbon adsorption system, including the drier, would be about
$1,500,000. This price would be for a system that could recover solvent
which would be reusable in the process at the facility. The commenter's
third observation about a manufacturer's carbon adsorber guarantee was that
if the concentration of MIBK and MEK varies, as it does in actual operation,
the initial guarantee would not be valid.
Response: After reexamination, it remains EPA's conclusion that carbon
adsorbers function effectively and can achieve 95 percent control efficiency
in this industry. The Agency agrees with the comment that a combination of
MEK and MIBK would require multi-column distillation equipment if separation
of these similar solvents is required. One example of a large plant used in
the cost analysis presented in the BID did include multi-column distillation
equipment. In the examples of smaller plants, however, the blend was used
directly as virgin solvent.
Additional steps are required to separate most solvents from water, and
for these two solvents, EPA agrees with the comment that the system must be
constructed of stainless steel. The cost analysis done during the develop-
ment of this standard and presented in the BID included costs for distilla-
tion equipment with a moisture removal system for the separation of water.
Similarly, the costs included a system constructed of stainless steel.
The proposed standard was based on a carbon adsorber efficiency of
95 percent primarily because the efficiency of the carbon adsorber tested
during the development of the standard ranged from 98.5 to 99.6 percent.
The facts that the manufacturer guaranteed the unit to be 95 percent
efficient and carbon adsorber efficiencies of 95 percent have been
demonstrated in two other web coating industries (publication rotogravure
printing and pressure sensitive tapes and labels) also support the basis of
95 percent.
Humidity, air stream temperature and concentration, and length of time
the carbon bed is in use are bed design parameters that can all be
2-26
-------
controlled so as to enable the carbon adsorber to achieve 95 percent
efficiency. Chapter 4 of the BID describes how these parameters are
controlled at existing carbon adsorbers in use by vinyl printers.
The commenter stated that in order to achieve 95 percent control
efficiency the carbon bed must be oversized by 50 percent and that the
carbon bed would need to be replaced in 1 to 1.5 years. Indeed, sizing of
carbon beds is an important design parameter. In regard to bed life, an
estimate of 3 years was used in the BID cost analysis based on industry
experience with carbon adsorbers. However, more frequent bed replacement
would not appreciably affect the estimated cost effectiveness of carbon
adsorbers.
The commenter also noted that a realistic installed cost for a carbon
adsorption system capable of recovering solvent that would be reusable in
the process at the facility would be $1,500,000. The commenter did not give
any size information. The BID cost analysis included a system capable of
recovering reusable solvent for a cost of $1,250,000. Thus, if the
commenter is referring to a carbon adsorption system sized for Model
Plant C, the commenter's cost estimate supports the BID estimate within the
Agency's level of accuracy.
The commenter observed that carbon adsorber guarantees would not be
valid when the concentration of MIBK and MEK varies. It was not clear if
the commenter was referring to relative concentrations of MIBK and MEK in a
combination consisting only of these two solvents or if he was referring to
the total concentration of the gas stream directed to the carbon adsorber.
In regard to the relative concentrations of MIBK and MEK, a carbon adsorber
would be designed to adequately handle the highest expected concentration of
the component which is the more difficult to adsorb and desorb. If this
"worst-case" concentration is not exceeded, the carbon adsorber should
maintain optimum control.
The Agency found little data on what happens to carbon adsorption
efficiency with extreme variations in concentrations. However, the cost
analysis was based on typical ranges of concentration. The carbon
adsorption system in the model plant analysis was sufficiently large to
2-27
-------
handle the widest range of concentrations expected to occur. The fact that
carbon adsorbers can maintain 95 percent efficiency is borne out by a
statement made at the meeting of the National Air Pollution Control
Techniques Advisory Council (NAPCTAC) (II-B-50) held to review this
standard. At this meeting, one participant noted that his company
"... has had extensive experience in the operation of
carbon adsorption solvent recovery systems. At our
Toledo, Ohio facility there is a system that has been
operating for over 30 years. A recent analysis of the
system efficiency showed that 95 percent control was
obtained."
Similarly, the carbon adsorber manufacturer quoted above in
Comment 2.2.1 noted that his company's equipment could achieve recovery
efficiencies of 95 percent or greater.
Should a vinyl printer decide that carbon adsorption will not be
feasible or desirable based upon some parameters of his operation,
compliance provisions for incineration have been added to the regulation.
2.2.3 Comment: One commenter (IV-D-7) discussed four problems associated
with the applicability of carbon adsorbers to the entire vinyl industry.
The first was the potential problem of fouling of the carbon beds by
plasticizers released from vinyl coatings when heated above 300°F and by
solvents such as dimethyl formamide which are incompatible with carbon
adsorption. Reduced carbon bed life and special desorption techniques would
add to the cost of the process. The next area of his concern included other
technical problems associated with carbon adsorbers such as space require-
ments, maintenance, carbon replacement, distillation, risk of bed fires, and
risk of total shutdown. The third problem was the cost and technical
difficulty of distilling mixed solvents with close distillation tempera-
tures. The fourth problem with carbon adsorption mentioned by the commenter
was the potential inability of carbon adsorbers to achieve 95 percent
efficiency in recovering emissions of a number of solvents used in this
industry. He said that to maintain that efficiency where plasticizers and
2-28
-------
solvents such as dimethyl formamide enter the bed, more frequent cycling,
increased steam flow, and more frequent bed replacement would be necessary
and would result in higher operating costs. The commenter concluded that
other methods of emission control, particularly incineration, may be the
only feasible technology in certain segments of the industry.
Response: This commenter presents process-specific and site-specific
conditions that were not seen during the development of the standard.
During the development of this standard, carbon adsorption was the only
control method in use in this industry. Since several comments were
received on incineration, the Agency examined the current status of this
control technology in the vinyl printing industry by contacting the three
plants listed in the commenter's letter. Of these, one (IV-E-11) is
considering the use of incineration, one (IV-E-10) has considered it but has
no current plans to install an incinerator, and one (IV-E-12) has firm plans
to install an incinerator.
The Clean Air Act requires that each NSPS reflect "application of the
best technological system of continuous emission reduction which (taking
into consideration the cost of achieving such emission reduction, . . .) the
Administrator determines has been adequately demonstrated." Such technology
is often referred to as "best demonstrated technology" or simply BDT. After
studying the comments received, the Agency had to determine if there were
some circumstances described where carbon adsorption could not be used and
could not be considered BDT as it had been during the development of the
standard. Because EPA found no vinyl printers using a control technology
other than carbon adsoprtion, the only information available on which to
base such a determination has come from contacts with vinyl printers who
were or had considered incineration in favor of carbon adsorption and the
comments received concerning incineration.
In the case of the plant considering incineration (IV-E-11), the
contact explained that they have an incinerator, formerly used to control
emissions from another process, that they may retrofit for control of
emissions from a vinyl print line. As an example of conditions where
2-29
-------
incineration would be preferable to carbon adsorption, this contact cited
the case where a plant used waterborne inks on all but one Printline. He
felt carbon adsorption would be too costly for controlling a single line,
and incineration therefore would be selected.
The second plant (IV-E-10) was no longer considering incineration.
Incineration had been studied because the printlines were far apart, making
duct and fan costs high enough to consider using two control devices rather
than one. They felt that two incinerators would be less costly than two
carbon adsorbers.
The third plant (IV-E-12) had firm plans to install an incinerator
rather than a carbon adsorber because the plant is surrounded by close
neighbors and does not have enough space for a carbon adsorber.
Addressing each of the commenter's four points; first, EPA feels that
since new, modern equipment normally does not rely on oven temperatures as
high as 300°F, it is unlikely that excessive plasticizers would be emitted
in the drying oven. However, if such plasticizers were emitted in the
printing and coating operation, bed fouling would be a potential problem and
would require the provisions listed by the commenter such as more frequent
bed cycling. Plasticizers are emitted in the embossing operation, and,
since such embosser emissions are difficult to control, especially with
carbon adsorption, embossers were not considered part of the affected
facility for the purposes of this standard.
The commenter cited several technical problems associated with carbon
adsorbers, including maintenance, risk of bed fires, risk of total shut-
downs, carbon replacement, space requirements, and distillation. Problems
associated with maintenance, bed fires, shutdowns, and carbon replacement
can be minimized by careful operation as has been the case at existing
facilities, such as the one described in the response to Comment 2.2.2, that
have had long-term success with carbon adsorption and found that it is
capable of consistently achieving 95 percent control efficiency.
The BID estimated that, at a typical plant, 3200 square feet would be
necessary to accommodate a carbon adsorber, cooling tower, solvent distilla-
tion equipment, solvent drying equipment, and recovered solvent storage
2-30
-------
tanks. Newly constructed facilities would have the opportunity to plan
space to accommodate their selected control device. Some existing plants
may not have the necessary space for a carbon adsorber. Applicability of
reconstruction provisions is judged on a case-by-case basis by the
Administrator based on economic or technical limitations. This case-by-case
approach gives some additional leeway if space requirements appear to be a
factor that would limit a source's ability to meet the standard.
Next the commenter addressed the cost and technical difficulty of
distilling mixed solvents with close distillation temperatures. EPA
acknowledges this difficulty and notes that some solvents may indeed be
incompatible with carbon adsorption. However, during the development of the
standard, it was determined that such solvent distillation problems would be
rare. The types of solvents studied in the model plant analysis of carbon
adsorption, where the solvents were distilled and reused as new solvent,
make up over 95 percent of the solvent volume used in this industry.
The commenter's final issue addressed the potential inability of carbon
adsorbers to achieve and maintain 95 percent efficiency in recovering a
number of solvents at the estimated cost. The test program clearly demon-
strated 95 percent efficiency on a carbon adsorber processing the types of
solvents used in most cases in this industry. As mentioned above, these
solvents make up over 95 percent of the solvent used in this industry.
Based on the contacts with plants who are or have considered using incinera-
tion and on the commenter's discussion, the Administrator has concluded that
there are certain site-specific and process-specific conditions where carbon
adsorption would not be BDT. As described above, some of those site-
specific conditions are when only a single print line is to be controlled,
when print line configuration prohibits control by a single carbon adsorber,
and when the available space is insufficient for » carbon adsorber. Some of
the possible process-specific conditions are when high oven temperatures
cause plasticizer emissions and when solvents with close distillation
temperatures are used.
Based on these indications that carbon adsorption would not be BDT in
some circumstances, EPA has added compliance provisions for incineration.
2-31
-------
It has been adequately demonstrated that incineration can achieve a
comparable control efficiency to that of carbon adsorption at a higher, yet
reasonable cost. Thus, an owner of a facility facing various site-specific
or process-specific problems that make carbon adsorption too costly or
unable to meet the control needs of the process might choose to achieve
compliance using incineration.
2.2.4 Comment: Two commenters (IV-D-6, IV-D-8), both representing
producers of vinyl coated products and vinyl printed products, disagreed
with EPA's treatment of the test data on capture efficiency. One commenter
(IV-D-6) explained that during the second test which indicated a capture
efficiency in excess of 90 percent, the room air supply fan was not
operating and the wall exhaust fan was only operational during times that
the print head fans were off. He stated that machines cannot be run with
the room air supply fans off, and therefore he felt that the test was not
valid or representative. The other commenter (IV-D-8) felt that the
"turbulence" or "disruption of air distribution" cited in the BID due to
operation of the room air supply fan and the wall exhaust fan were
legitimate operating conditions, and he felt their effects should not be
ignored in determining capture efficiency. He felt that EPA had "discarded
all lower but real world" capture efficiency data. He urged EPA to discard
the test data cited in the BID and conduct a new series of tests on a new
printer. A trade association (IV-D-7) concurred with these views. In
addition, the trade association stated that the emission testing program was
not conducted under normal operating conditions and referred to the General
Provisions, 40 CFR 60.8(c), which require that performance tests be "based
on representative performance of the affected facility." The association
noted that techniques such as turning off the fans or total enclosure of the
print line may be necessary during the performance test. The association,
however, maintained that enclosure of a print line creates conditions which
are not representative of the capture efficiency achieved at a print line
under normal operating conditions. In other words, it is not clear that any
new source in this industry could achieve 90 percent capture efficiency
2-32
-------
during ordinary operations. The commenter elaborated that total print line
enclosure is not practicable during operation, and the efficiency of air
management around the print line depends on a number of factors over which
the operator may not have control, even in the design of a new print line.
For example, where a new print line is installed in a large building
containing other processes, major capital expenditures could be required to
achieve 90 percent capture efficiency during normal operation. Another
commenter (IV-D-6) agreed that a total enclosure of the print line is not
practical because of operator accessibility and the high concentrations of
solvent encountered with the enclosures that in some places could exceed the
lower explosive level (LEL). The trade association (IV-D-7) , therefore,
concluded that EPA should conduct further capture efficiency tests for a new
vinyl print line during normal operation. Another commenter (IV-D-10)
stated that the emission test that verified 90 percent capture efficiency'
was conducted under contrived conditions, and therefore questioned the
results and appropriateness of a 90 percent capture regulation.
Response: In order to respond to the comments on the performance tests
and the representativeness of test conditions, it is important first to make
clear that the performance test procedure is, in effect, an integral part of
the standard. For this reason, the test procedures and test conditions used
by EPA during the development of the standard must be the same as or have a
known relationship to the test procedures which will be used and the
performance test conditions which may be encountered during the testing of
sources which are subject to this standard.
In this standard, the test procedures to be used by a plant operator
for determining compliance will be the same as or equivalent to those used
by EPA in developing the standard. The performance conditions may differ
depending, for example, on the configuration of the plant and the number of
lines. However, the EPA tests were, and subsequent performance tests will
be, conducted in conformance with the same criteria, thereby assuring that
the standard is achievable during any performance test.
2-33
-------
In conducting a performance test, the plant owner must demonstrate that
his new line can achieve the performance standard when tested under
operating conditions which are representative of normal production and/or
which reflect the air change conditions during the EPA test (as specified
below). One important feature of the performance test is that all fugitive
VOC emissions from the printing area must be captured and vented through
stacks suitable for measurement. For a new line(s) in a new plant,
emissions from the room in which the line(s) is (are) located can be
measured with no special arrangements. For a new line in a plant with
existing lines, the owner or operator may desire to isolate emissions
associated with the new line from those associated with existing lines. In
this case, he has three options:
(1) build a permanent enclosure around the affected facility;
(2) build a temporary enclosure around the affected facility (the
enclosure has to be ventilated at a rate which approximates the rate to
which the affected facility is exposed without the enclosure up to 20 air
changes per hour); or
(3) shut down all other sources of VOC and continue to exhaust
fugitive emissions from the affected facility through any building
ventilation system and other room exhausts such as print line ovens and
embossers.
The industry has argued that a total enclosure restricts operator
access and would allow VOC concentrations to increase to unacceptable
levels. It is the Agency's position that it is possible to construct a
total enclosure which provides the necessary operator access while also
preventing the buildup of high concentrations of solvent. The EPA test
results show that an increase in solvent concentration is not the inevitable
product of a total enclosure.
The two tests performed during the development of the standard were
conducted at a new facility which had installed only the first of a total of
three print lines within a building. Thus, the building itself served as a
permanent total enclosure housing only one line.
2-34
-------
The industry would disregard results of the test at this plant because
it alleges that EPA "contrived" the operating conditions. Air management is
a critical aspect of any pollution control system. EPA did indeed alter the
old normal air management scheme within the building in an attempt to attain
the best operating conditions for pollution control. As a consequence of
what was learned, the plant adopted new normal operating conditions, the
conditions under which the EPA tests were actually conducted.
During the first test, the room containing the printing line had two
ventilation fans, one in the ceiling and one on the wall. Only the wall fan
was operating prior to the first test. (The test data revealed its
discharge was much less than design.) A large intake ventilation fan
mounted directly above the operating line was turned on at EPA's request in
an attempt to assure that the test took place under normal operating
conditions. It was assumed that the ventilation fan would normally be
operated when the plant ultimately had three lines. It was not until after
the first test was complete that the owner explained that normal operation
was usually with the overhead ventilation fan off. Because of the fan's
location, it discharged directly on the print heads, causing premature
drying of the inks on the gravure cylinders and adversely affecting the
quality of much of the material printed during the test. The plant
personnel subsequently had to turn the fan off in order to produce a product
within specification. It was because EPA had inadvertently tested the plant
under abnormal conditions that the decision was made to return for a second
test.
The wall exhaust fan was not operated during the second test based on
the Agency's judgment that the best air management for capture of VOC
emissions from the plant required that the fan not operate. That judgment
was based on the following reasoning. The fan was originally installed to
ensure that the threshold limit value (TLV) would not be exceeded within the
building. On the first day of the second test, however, air flow measure-
ments revealed that although the fan motor was running, the fan was
essentially operating at a zero discharge rate because of loose drive belts.
2-35
-------
After EPA arranged to have the belts adjusted, the fan's operation caused
turbulent drafts of air through and around the printing press.
Results of the second day's test showed that the drafts interfered with
the print line's fugitive capture system. The data showed that signifi-
cantly less VOC was captured than during the tests conducted the previous
day when the belts were loose. The plant personnel and the test team
jointly concluded that the fan's location (directly adjacent to and above
the press) created drafts which adversely affected the capture system.
Since the fan had previously been operating ineffectively (because of the
maladjusted drive belts), there obviously was some question regarding the
need for it. It was then decided to run the print line with the fan off to
confirm that it was not critical to the operation of the line. If the VOC
concentrations did not increase and exceed the TLV, the fan was unnecessary.
The subsequent test confirmed that the exhaust systems associated with the
press and embosser provided sufficient ventilation to keep the VOC concen-
trations below the TLV, and the wall exhaust fan was not needed during
normal operation. Plant personnel concluded that the wall fan would
subsequently be used only when the print line fans were off or in the event
of some unusual occurrence, such as a solvent spill.
This mode of operation was advantageous for the plant because it
increased the amount of solvent captured, thereby improving the return on
the investment for the carbon adsorber. Obviously, this printing press
could, did (because of the loose belts), and would continue to operate with
ventilation provided only by the evacuation systems associated with the
press and embosser exhaust systems. The results of monitoring conducted
during the second test showed that with the wall exhaust fan off, the
11,700 SCFM evacuated from the vicinity of the printing press by the print
head ovens and embosser ventilation system was sufficient to ensure that
concentrations did not increase to unacceptable levels and that the 8-hour
average concentration in working areas did not exceed the TLV. This draft
would provide air changes approximately every 3 minutes to a room with
dimensions of 20' x 20' x 90', more than adequate to house the print line,
and coincidental^, about one-third the volume of the room which housed this
2-36
-------
print line. This is consistent with good ventilation practice for a clean
room as discussed in Industrial Ventilation - A Manual of Recommended
Practices (Committee on Industrial Ventilation), 1976, p. 5-112.
Should the operator decide to conduct his performance test with the
print line contained in a total enclosure, he has two options: to construct
either a permanent enclosure or a temporary enclosure which would be removed
upon completion of the performance test. The cost of a permanent enclosure
would be relatively small, less than 2 percent of the total capital required
for the new line (IV-B-9). A temporary enclosure would be even less
expensive.
If an operator chooses to conduct the performance test with a temporary
enclosure, the only constraint the Agency places on the test is that the
plant operator must approximate the ventilation conditions expected to be in
effect when the enclosure is removed or create conditions that would
represent good ventilation practice as recognized by authoritative sources
such as Industrial Ventilation. However, it is not necessary to ventilate
the enclosed space at a rate greater than 20 air changes per hour. This
rate is consistent with that of the test conducted by the Agency.
As an alternative to the total enclosure, a plant operator may elect to
shut down other sources of VOC in the room which contains the affected
facility. The fugitive emissions from the print line must continue to be
exhausted through any area or building ventilation system and other room
exhausts, such as print line ovens and embossers, that would operate under
normal conditions.
At some plants, the work area may be ventilated only by the drafts of
ovens and embossers. At other plants, high levels of fugitive emissions
from existing facilities may require that separate ventilation fans be
operated. In this latter situation, the operator should consider two of the
advantages of reducing the amount of ventilation used for the existing
facilities. First, the costs of circulating, pumping, heating, and
conditioning large quantities of air through a plant can be significant.
Secondly, poorly designed ventilation systems (as evidenced during testing
of the reference plant) can interfere with the VOC capture system, increase
2-37
-------
the fugitive emissions within the building, and perhaps thereby create a
need for even more ventilation air.
If a new line is installed in the same room with existing lines, there
are ways to design and install or alter building ventilation systems .which
would minimize its effect on the efficiency of the capture system of all
lines, but especially the line subject to the NSPS. For example, if a
building ventilation draft fan must be used, the intake can be manifolded
with the suction inlets located around the work stations of the existing
lines where the fugitive VOC concentrations could be expected to be highest.
If a forced air fan is used, the ventilation air can be manifolded to a
number of outlets located such that they do not compete with the solvent
recovery system but may actually improve it. Such a thoughtfully designed
dispersion system would likely permit operating with much smaller volumes of
ventilation air.
Another improvement in air management techniques was developed by the
plant that EPA tested when a second line was installed after the test
program was completed. The duct work for a second new print line was
constructed so that oven make-up air is collected from between the print
heads by "floor sweeps" rather than above the print heads as was done by the
first print line which EPA tested. This should further reduce fugitive
emissions and improve overall capture efficiency. The cost of making the
type of air management changes discussed above would likely not exceed
2 percent of the control cost for the new line. The potential savings to
the owner in reduced utility costs and increased solvent recovery would
offset some and perhaps all of the incremental control cost.
The final standard gives the plant operator a great deal of latitude in
deciding how to demonstrate compliance. A new line could be constructed
within a permanent enclosure, thereby simulating the conditions under which
EPA conducted its tests. Alternatively, the line could be installed in a
room with existing sources. If the ventilation system in the existing room
is expected to affect the performance of the new line's capture system, the
performance test may be conducted with the existing sources shut off or a
temporary enclosure may be installed to isolate the new line from the
2-38
-------
existing plant equipment. If a temporary enclosure is constructed, the
building ventilation system (to which the new line will normally be subject)
must be approximated, but only up to a maximum of 20 air changes an hour.
This is consistent with conditions under which EPA tests were conducted and
representative of good ventilation practices.
For the reasons previously discussed, it is the Administrator's
judgment that the results from the second test show that high capture
efficiency can be attained and worker exposure limited if a print line and
its adjunct vapor capture and work space ventilation systems are properly
designed. The adjustments made to the plant's air management system during
testing were needed in order to best represent the plant's normal operating
conditions, but more importantly, the adjustments were made to ensure that
the test was conducted under conditions which were representative of the
best demonstrated technology. Also, the capture efficiency data were
obtained under conditions that could be normally and routinely maintained by
properly designed new print lines and were representative of capture
efficiency that could be achieved by any new print line even when installed
in existing plants. Therefore, further capture efficiency tests were not
conducted as requested.
2.2.5 Comment: The commenter (IV-D-7) representing the trade association
stated that, while his organization supports the percent reduction format of
the proposed standard, they were concerned that because EPA had only
assessed the environmental, energy, and economic impacts of carbon adsorp-
tion, the Agency may intend that carbon adsorption technology must be used
for compliance. He felt that such requirement would be environmentally and
economically unsound and, indeed, infeasible for some segments of the
industry. He asked that EPA clarify that any combination of capture and
control which achieves the required level of VOC reduction is acceptable.
Response: It is not the Agency's intent that carbon adsorption be the
only technology available for compliance. Indeed, compliance provisions for
waterborne inks were proposed with the standard in addition to provisions
2-39
-------
for carbon adsorption. EPA has also added compliance procedures for
incineration to the regulation.
As stated in the preamble (48 FR 2283), the 85 percent reduction
requirement was based on a capture efficiency of 90 percent and a control
device efficiency of 95 percent. However, any combination of capture
efficiencies and control device efficiencies would be acceptable as long as
the 85 percent reduction is achieved.
2.2.6 Comment: One commenter (IV-D-6) felt that the NSPS should include a
provision for the use of waterborne and solventborne inks on the same piece
of equipment. He noted that this way, when a product is run through this
source, the total emission from this source would meet the required
emissions limitation. He added that this method of compliance is allowed in
the NSPS for publication rotogravure printing and that it should be
incorporated into the flexible vinyl coating and printing NSPS.
Response: The standard does not prohibit the use of waterborne and
solventborne inks on different units of the same print line. However, if
the VOC-to-ink solids ratio exceeds 1.0 over the averaging period, emissions
must be directed to a control device. Compliance with the publication
rotogravure standard is determined based on a liquid material balance which
is relatively simple to demonstrate because of the types of solvents
prevalent in that industry. The average VOC emission percentage in that
standard is calculated as: (total mass of VOC used minus total mass of VOC
recovered) divided by (total mass of VOC used plus total mass of water
used). The water used includes dilution water and water in waterborne inks.
If a vinyl printer were able to demonstrate compliance using a liquid
material balance approach, he could petition the Administrator under the
General Provisions to conduct this type of performance test.
2.2.7 Comment: One commenter (IV-D-2), an ink manufacturer, stated that he
was not familiar with EPA's definition of VOC. He pointed out that
2-40
-------
different states use different definitions of VOC and he was therefore
concerned with calculating the VOC content of his inks.
As a compliance alternative, the trade association representative
(IV-D-7) said that one member had expressed interest in using a combination
of waterborne inks and coatings formulated with nonphotochemically reactive
halogenated solvents as a means of compliance. He added that since this
would appear acceptable based on EPA's definition of "VOC," he felt Agency
consideration was desirable.
Response: EPA's letter acknowledging the receipt of the first
commenter's letter (IV-C-1) explained that the definition of VOC appears in
40 CFR 60.2. This section explains that
"'Volatile Organic Compound1 means any organic compound
which participates in atmospheric photochemical reactions; or
which is measured by a reference method, an equivalent
method, an alternative method, or which is determined by
procedures specified under any subpart."
EPA has determined that the following organic chemicals do not
participate in atmospheric photochemical reactions and are therefore
excluded from the definition of VOC:
methane
ethane
1,1,1-trichloroethane (methyl chloroform)
methylene chloride
trichlorofluoromethane (CFC-11)
dichlorodifluoromethane (CFC-12)
chlorodifluoromethane (CFC-22)
trifluoromethane (CFC-23)
trichlorotrifluoroethane (CFC-113)
dichlorotetrafluoroethane (CFC-114)
chloropentafluoroethane (CFC-115)
This list of nonphotochemically reactive chemicals may be updated in the
future. In a telephone call to the commenter (IV-E-20), EPA learned that at
least one vinyl printer is using methylene chloride, tricholorotrifluoro-
methane, and 1,1,1-trichloroethane. Nonphotochemically reactive solvents
should be excluded when determining the VOC content.
2-41
-------
The procedures in Reference Method 24 determine total volatile organics
in the ink and cannot distinguish between photochemically and nonphoto-
chemically reactive compounds. The results from Reference Method 24 would
have to be adjusted using standard gas chromatographic techniques to
subtract any nonphotochemically reactive halogenated solvents that were in
the ink.
2.2.8 Comment: One commenter (IV-D-10) asked why, if compliance testing
required the average of three sets of data, did the standard not also
require three tests?
Response: Based on the General Provisions [40 CFR 60.8(f)], the
regulation requires that each performance test shall consist of three test
runs. The test program used in developing the standard consisted of seven
test runs, and capture efficiency was 90 percent or higher for all seven
runs. There is no requirement that a specific number of tests be conducted
in a testing program for standard setting.
2.3 MODIFICATION AND RECONSTRUCTION
2.3.1 Comment: One commenter (IV-D-8) recommended that physical changes
for the purpose of energy conservation not be considered as a reconstruction
of an existing source. He disagreed with EPA's assumption that extensive
equipment modifications would not occur. He gave an example of oven
replacement, explaining that as energy costs continue to rise, the likeli-
hood of oven repair/replacements for energy conservation increases.
Expenses for oven replacement could be over 50 percent of the installed cost
of the affected facility, triggering the application of the NSPS. While
such a project would conserve energy and reduce VOC emissions, the VOC
reduction would be less than the 85 percent required. He added that
application of the NSPS may cause a delay in such energy conservation
projects "until energy costs rise even higher and the cost of additional VOC
controls to achieve 85 percent emission reduction can be justified." The
commenter strongly recommended that the costs for energy conservation be
2-42
-------
excluded from reconstruction costs and recommended a change in the
regulation to accomplish this.
Response: During the development of this regulation, EPA believed it
was unlikely that existing facilities would become affected facilities under
the provisions for modification and reconstruction. The preamble to the
proposed regulation explained that extensive modifications and reconstruc-
tions were not expected to occur in this industry. The preamble explained
that there were few changes constituting a reconstruction that could be made
to the print line and that plant owners would be likely to chose to build an
entirely new line since the cost of a new line would be comparable to the
cost of retrofitting an existing line.
Section 111 requires that standards reflecting best demonstrated
technology (BDT) apply to new sources, and these include reconstructed
sources. To justify a categorical exemption from the standard for sources
rebuilt for energy conservation purposes, EPA would have to conclude that
there was no BDT for such sources. This is not the case. On the contrary,
EPA judges that in general BDT for such sources is the same as for other new
sources, and that they should therefore be subject to the same standard. In
the course of installing new ovens, the facility owner would have an excel-
lent opportunity to improve the capture efficiency and thus raise the
overall control efficiency. In particular, EPA considered the commenter's
suggestion that application of the NSPS to such sources would delay their
rebuilding. The decision to undertake an energy conservation project is an
economic decision made when the costs of not implementing the project
(higher energy costs) exceed the costs of implementing the project (retrofit
plus controls). The purpose of the reconstruction provisions is to ensure
that an owner or operator does not rebuild an existing facility without
consideration of the achievability of the standard. Without such provi-
sions, circumvention of the standard could be attempted by replacing all but
minor components. Reconstruction costs are highly site-specific. However,
if there are cases where the cost of compliance for a rebuilt source is
unusually high, that is considered on a case-by-case basis by the
2-43
-------
Administrator in determining whether the rebuilding constitutes reconstruc-
tion subject to NSPS, under 40 CFR §60.15(b)(2).
2.4 ECONOMIC IMPACT
2.4.1 Comment: A representative of a trade association for wallcovering
manufacturers (IV-D-9) stated that, as proposed, the NSPS would impose a
serious financial burden on an industry which is substantially depressed as
a result of economic conditions. He added that unless there is a stronger
showing that the proposed NSPS is the only way to enhance environmental
quality significantly , the wallcovering industry should instead be given
the flexibility to investigate and implement a variety of new technologies,
such as waterborne inks, which may provide long-term control of undesirable
emissions at a much lower cost.
Response: The economic analysis of the proposed standard indicates
that the costs and economic impacts are reasonable. As proposed, the
standard contained compliance provisions for waterborne inks as well as for
carbon adsorption. Also, compliance provisions for incineration have been
added to the promulgated standard. With or without an NSPS, the industry
always has the flexibility to investigate and implement new technologies.
Indeed, the Agency supports and encourages the development and use of new
innovative technologies that provide for the control of air emissions.
2.4.2 Comment: Several commenters (IV-D-71, IV-D-8, IV-D-10) pointed out
what they felt to be inadequacies in EPA's economic analysis. The trade
association representative (IV-D-7)1 stated that the costs and benefits of
the NSPS are not accurately reflected in EPA's analysis of environmental,
energy, and economic impacts. He discussed four Agency assumptions he felt
were questionable to support this statement. Other commenters (IV-D-8,
IV-D-10) also mentioned some of these assumptions. The four assumptions in
question are:
2-44-
-------
2.4.2.1 Comparison between costs of compliance with the NSPS and SIPs. The
percent reduction required under the proposed NSPS is not directly
comparable to SIP requirements because the SIPs generally do not take
fugitive emissions into account, as does the proposed standard. Thus, the
commenter (IV-D-7) noted that the percent reduction requirement in the NSPS
is considerably more stringent than a percent reduction computed from SIP
emission limits. He went on to say that it should not be assumed that all
affected facilities would achieve the SIP level of control through the use
of add-on control technology. In some cases plants would achieve SIP
compliance through the use of waterborne inks and process modifications.
He concluded that the potential economic impact of the NSPS was
substantially understated because of the overly simplistic comparison of
compliance costs for the NSPS with compliance costs of current applicable
requirements. Therefore, the commenter urged EPA to determine actual and
projected costs of compliance with SIP regulations and compare those with
the costs of compliance with the NSPS.
Response: The BID states that current State regulations do not cover
the fugitive VOC losses from vinyl coating and printing operations. This
statement was not intended to imply that State regulations necessarily
exempt fugitive emissions. The statement meant that it may be possible to
meet the 65 percent overall VOC reduction listed in the CTG by controlling
dryer emissions only.
EPA did not assume that all affected facilities would achieve the SIP
level of control through the use of add-on control technology. Add-on
control technology, specifically carbon adsorption, was used in the SIP/NSPS
incremental cost analysis because it was the only control technology in use
during the development of the standard. There are two important points with
regard to the incremental differences between the SIP and NSPS levels of
control. First, regardless of how a facility met the SIP control level,
some costs would be involved whether they were for development of waterborne
inks, process modifications, or installation of carbon adsorption equipment.
Since carbon adsorption was the only technology in use, those costs were the
2-45
-------
most appropriate for analysis. Second, during development of the standard,
there was no indication that any facilities would become subject to the
standard as a result of modification and reconstruction and, hence, that no
facilities would actually be faced with increasing control from the SIP
level to the NSPS level. Although two commenters (IV-D-71 and IV-D-8) have
indicated that reconstruction or modification might be considered, it would
still be unusual for a plant to be faced with having to increase control
from the SIP level to the NSPS level. Instead, prior to construction of a
new facility, the owners would decide how best to meet the required level of
control.
The estimated costs of compliance compiled during the development of
the standard are reasonable and realistically reflect the costs of
compliance with SIP regulations based on carbon adsorption. Since EPA does
not know exactly how plants will choose to meet SIP requirements, the Agency
cannot make the actual comparison requested by the commenter.
2.4.2.2 Cost and the applicability of carbon adsorption to all segments of
the vinyl industry. The commenter (IV-D-7)1 noted that EPA's estimate of
$1.2 million for a carbon adsorption system was low by several hundred
thousand dollars. He added that the potential savings from recovered
solvent do not offset the initial capital cost of the carbon adsorption
system.
The commenter said that carbon adsorption is not feasible in all
segments of the vinyl industry and that it is not economical or practicable
in a declining industry to invest large amounts of capital required for
carbon adsorption systems.
The technical problems that the commenter associated with the
applicability of carbon adsorbers are discussed in Section 2.2. Briefly, he
noted the costs of reduced carbon bed life and special adsorption
techniques, the cost of distilling mixed solvents with close distillation
temperatures, and higher operating costs resulting from maintaining
95 percent efficiency in recovering emissions of a number of solvents used
in the industry.
2-46
-------
The commenter concluded his remarks on carbon adsorption by saying that
other methods of emission control, particularly incineration, may be the
only feasible technology in certain segments of the industry but may cost
more than carbon adsorption.
Response: EPA agrees with the commenter that other methods of emission
control, such as incineration, may be desirable in certain segments of the
industry. (This is explained in the response to Comment 2.2.3.) If
incineration were used to meet the standard, EPA estimates (IV-B-8) indicate
that the costs would be higher ($600 per ton VOC reduction) than carbon
adsorption ($75 per ton VOC reduction), yet still reasonable. Compliance
provisions for incineration have been added to the regulation.
EPA's cost estimates for carbon adsorption equipment were based on
industry cost data for installed control systems as explained in Chapter 8
of the BID. In regard to the comment that the cost estimate for the carbon
adsorber was low by several hundred thousand dollars, EPA notes that for a
system estimated to cost $1.2 million, this difference is low enough that
the costs are considered comparable.
Potential savings from recovered solvent offset annual operating costs,
but do not offset the initial capital costs. EPA estimates show, however,
that even if solvent was recovered, but no value was assigned to it, the
cost effectiveness of carbon adsorption would be about $800 per ton VOC
reduction. EPA considers the costs of even this "worst case" situation to
be reasonable.
The technical issues mentioned by the commenter are addressed in the
response to Comment 2.2.3. The Agency feels that, with the addition of
compliance provisions for incineration, the operator of a new facility will
have the option to select the most practical and cost effective means of
control for his print line.
2.4.2.3 Accuracy of estimated costs and benefits of the proposed standard.
In regard to the estimated model plant cost factors, the commenter (IV-D-7)1
noted that the cost of electrical consumption attributable to the control
2-47
-------
equipment at the plant tested was about $20,000 annually. He also added
that model plant costs such as operator costs, utility costs, and interest
rates did not reflect current costs in the industry. He explained that use
of company- or division-wide financial statements, from which profit
allocable to vinyl coating and printing operations cannot be discerned, may
have led to overstated profits for this segment.
A second commenter (IV-D-8) agreed that costs were understated through-
out the BID with respect to labor, capital, and energy. Specifically, the
commenter cited the credit for recovered solvent. He noted that the
material recovered from a carbon adsorber would require processing to remove
water and separate solvents. He felt that it was unlikely that operators
would be able to justify the cost of installing and operating distillation
equipment because of the relatively small volumes of solvent which may be
recovered. He said it was more likely that the printer would pay a fee to
have this done. He concluded that the BID should determine actual costs for
the various model plants.
Response: The commenters addressed several model plant cost factors
they considered to be low. The assumptions and references for these costs
were presented in Chapter 8 of the BID which states a probable accuracy of
±30 percent. Electrical consumption for Model Plant C for the NSPS level of
control was estimated at $14,000 which is within the lower bounds of this
level of accuracy and based on a nationwide average of utility costs. Costs
were current when the BID was prepared. The response to Comment 2.4.5
discusses the sensitivity of the analysis to labor (operator) costs.
The use of company- or division-wide financial statements was necessary
because most companies are reluctant to submit detailed financial
information for the segment of interest, due to the proprietary nature of
this information and other considerations. Consequently, it was necessary
to rely upon publicly available financial data which possessed a greater
degree of aggregation than would have been ideal. This commenter offered no
additional information for the analysis. The important point is that such
2-48
-------
aggregated data could just as easily understate the financial parameters of
the segment of interest as it could overstate them.
As a result of this comment, EPA analyzed (IV-B-4) the economic impacts
associated with paying a fee for solvent recovery rather than installing and
operating distillation equipment. In this analysis, the estimated cost of-
the distillation system was subtracted from the capital costs. This was
assumed to be 50 percent of the total capital costs. The estimated cost of
having a jobber reclaim the solvent was added based on a fee of 70 percent
of the full solvent recovery credit. For a typical model plant, the maximum
percentage price increase would be less than 2 percent under this scenario.
The Administrator considers this maximum percentage price increase
reasonable.
2.4.2.4 Declining economic condition of domestic producers of vinyl
materials and the markets. The commenter (IV-D-7) felt that the declining
market condition had not been taken into account, noting that for the past
5 years, the U.S. coated fabrics industry has suffered declining production,
sales, profitability, and employment while competition in certain market
areas from imports has increased substantially. The commenter urged that
existing industry condition and prospects for future growth, especially in
the northeastern and midwestern regions of the U.S., be considered more
carefully.
Response: Declining market conditions were taken into account in the
BID. In fact, because growth within the flexible vinyl coating and printing
industry has been uneven, the industry was divided into three subgroups for
analysis: supported vinyl materials, unsupported vinyl film, and wall
coverings. An NSPS applies to new sources and not to existing sources
unless they are considerably modified or reconstructed. As the BID points
out in Section 9.1.4 (page 9-21), no new sources are anticipated in the
supported and unsupported film segments primarily due to the declining
market conditions as noted in Section 9.1.3.1 (page 9-10). Wall coverings,
on the other hand, have experienced considerable growth in the past and
2-49
-------
three to six new sources are forecast. The response to Comment 2.1.1
discusses anticipated growth of the industry. Because the incremental costs
of compliance are minimal and, in some cases, there are actually savings
which arise from a solvent recovery credit, future growth will not be
hindered by the NSPS.
There are several reasons why the current, short-term state of the
economy should not control NSPS standard setting activity. An NSPS takes
three to four years to develop and considers long-term air quality improve-
ments from growth and also allows for the planning of the growth to be
accomplished by making specific air pollution control requirements.
Recessionary conditions can improve rapidly in an industry over the four-
year period that it takes before the next review of an NSPS is conducted by
EPA. In addition, there are some industries where the advent of new
technology (hence new sources) will occur even during recessionary times.
This depends upon the economics of the new technology in comparison to the
replaced technologies; i.e., it is generally less expensive to install
control systems in new plants than to retrofit them to existing plants.
2.4.3 Comment: Two commenters (IV-D-6, IV-D-10) felt that the proposed
NSPS does not adequately address incineration as a control alternative. One
of those commenters (IV-D-10) stated that the economic analysis was based
upon fixed-bed carbon adsorption systems with no consideration for use of an
incinerator. Full credit was taken for recovered solvent. Recent indica-
tions are that when a full analysis of the operational requirements is
completed, an incinerator may be the control device of choice. Under these
circumstances, he concluded, an analysis of incinerator costs for vinyl
coated fabric, other than wall coverings, should be completed. Another
commenter (IV-D-7) suggested that the economic analysis for this proposal
address the costs and energy and environmental impacts associated with
incineration, inert air condensation, and other systems. That commenter
(IV-D-7) felt that the use of incineration may make compliance with the
NSPS considerably more costly than projected on the basis of the cost
analysis of carbon adsorption.
2-50
-------
Response: During the development of this standard, carbon adsorption
was the only control device in use in this industry and, thus, the analysis
for development of the NSPS was based on carbon adsorption. Incineration
was described in the BID and was mentioned as a possible control technique
although it was not in use at the time. (The response to Comment 2.2.3
explains the current status of incineration in this industry.)
The commenter is correct that the economic analysis was based on
fixed-bed carbon adsorption systems and that full credit was taken for
recovered solvent. Because most manufacturers who have installed solvent
recovery equipment are reusing the recovered solvent as if it were new
solvent, control device equipment costs were estimated for equipment that
was capable of purifying the recovered solvent so that it could be reused.
EPA developed estimates (IV-B-8) of capital and annualized costs and
cost effectiveness for thermal incineration. Thermal incineration was
selected for the analysis rather than catalytic because that is the type
under consideration by the three plants contacted and because, since thermal
incineration is more costly than catalytic, the test for reasonableness
would be based on the "worst case." At a typical plant, Model Plant C
described in the BID, the results of the analysis showed an annualized
operating cost for control using thermal incineration of $358,000 and a
total capital investment of $669,000. The cost effectiveness of control
using incineration was determined to be approximately $600/ton VOC
reduction. Although this is higher than the cost of carbon adsorption
($75/ton VOC reduction), the Administrator considers this a reasonable cost
and, therefore, compliance provisions for thermal and catalytic incineration
have been added to the standard.
Energy demands of an incineration control system are greater than for
carbon adsorption. At a typical plant such as Model Plant C in the BID,
n
energy demand would be about 42 x 10 Btu per year greater for incineration
than for carbon adsorption. In terms of environmental impacts, both systems
can reduce VOC emissions by the required 85 percent. Wastewater and spent
carbon are not generated when incineration is used.
2-51
-------
No applications of the use of inert air condensation in this industry
were found and it is not expected to be used. However, should this
technology develop prior to the review of this standard, the owner of the
affected facility has the option of using this control technology if it will
meet the standard.
2.4.4 Comment: One commenter (IV-D-1Q) stated that the basis for
annualized costs included data from a 1978 Agency publication. He felt that
cost data that are at least five years old do not seem to reflect current
conditions or provide a strong base for current economic projections.
Response: The basis for some of the annualized costs presented in
Chapter 8 of the BID did include data from a 1978 EPA publication ("Capital
and Operating'Costs of Selected Air Pollution Control Systems,"
EPA-450/5-80-002). Specifically, these costs were operating labor,
maintenance, and indirect operating costs including overhead, taxes and
insurance, and administration.
As a result of this comment and Comment 2.4.2.3, the Agency examined
the sensitivity of the analysis to these estimates. Even if the wage rate
of $10.00 per hour were doubled, the amounts involved are not significant
(less than 1 percent of capital costs). Furthermore, the overhead rate of
80 percent of labor may, in fact, be high since some overhead charges such
as taxes, insurance, and administration are included separately.
2.4.5 Comment: One commenter (IV-D-10) stated that he felt the costs
associated with the wastewater from a carbon adsorption system had not been
fully investigated and were understated. His plant's solvent systems
contain mixtures of acetates, ketones, and alcohols. He stated that the
water distillate from multiple solvent distillation units may be 10 percent
organic azeotrope and that the biochemical oxygen demand (BOD) could be
about 114,740 mg/1. He noted that, with publicly owned treatment works
(POTWs) developing user charges for wastewater with over 100,000 mg/1 BOD,
vinyl coaters and printers would be faced with such a surcharge. He
2-52
-------
concluded that this issue needed to be addressed and the appropriate changes
made in the economic analysis. A second commenter (IV-D-5) resubmitted
earlier comments (II-D-100) that also requested that EPA consider the cost
of pretreatment for residual solvent prior to discharge to a POTW.
Response: Dissolved solvent in the condensate from the carbon adsorber
represents a potential water pollution problem. Although the majority of
the solvent is removed during distillation and in the second condensation
step, varying amounts could remain in the wastewater depending on the type
of solvent. When the BID was written, no local POTWs were imposing
surcharges on vinyl printers with carbon adsorption distillation systems.
This issue was raised previously (II-D-100) and, as previously stated
(II-C-86), if a municipality should require an owner to pretreat, the proper
cost attributable to the NSPS would be for the incremental amount of water
treatment required due to the NSPS beyond that of the CTG (i.e., State
regulations). This incremental cost will not be a major cost item because,
from a typical plant, this incremental difference in the amount of
wastewater generated would be about 780,000 gallons per year. Although the
Agency feels it is unlikely, should the cost of a possible POTW surcharge
cause the owner or operator of a new facility to conclude that carbon
adsorption would be too costly, he can now choose incineration as a means of
control.
2.5 SELECTION OF BEST DEMONSTRATED TECHNOLOGY
2.5.1 Comment: The trade association commenter (IV-D-7) stated that the
technology required to achieve the standard had not been demonstrated for
most segments of the industry for the following reasons: (1) the emission
testing program was not conducted under normal operating conditions and
(2) the test data on residential wallcovering production did not adequately
demonstrate that the control technology was applicable to the industry as a
whole.
In regard to the applicability of the test data to the industry as a
whole, the commenter questioned whether data and technology applicable to a
2-53
-------
residential wallcovering facility are transferable to or characteristic of
other parts of the industry such as automotive trim, upholstery, and
apparel. He cited several variables including solvents, raw materials, line
speed, web widths, and coating applications which change depending on
end-use of the product, stating that it is inadequate to use tests from one
facility using one process to print one type of product as representative of
the entire industry. Another commenter (IV-D-6) added that air management
and testing of lines wider than 30 inches has not been demonstrated. He
noted that the BID states that 30 inches would be a typical width for a new
source but that the typical width for his lines is 57 to 60 inches. The
trade association representative (IV-D-7) noted that solvents used to
produce some products such as apparel are incompatible with carbon
adsorption. He explained that incineration is probably the most flexible
control option where process parameters are variable. Such variability
would be desirable to enable a print line to produce different types of
products. He listed three trade association members that use or plan to use
incineration. The commenter felt that further study by EPA would be
necessary to establish standards that reflect emission reductions achievable
through application of the best technological system demonstrated for the
different categories of sources in the industry.
Response: The discussion of EPA's emission testing program is included
in the response to Comment 2.2.4. The applicability of test data to other
segments of the industry has been discussed in the response to
Comment 2.1.3.
In regard to the applicability of carbon adsorption to the industry as
a whole, EPA recognizes that there may be some technological limitations to
carbon adsorption. Therefore, the Agency contacted the three plants this
commenter listed as having incinerators in use or being installed (IV-E-10,
11, 12). Only one of these plants has firm plans to install an incinerator.
Based on these contacts and on comments outlining potential technical
advantages, EPA has included compliance provisions for incineration. It
appears that there may be potential situations where carbon adsorption could
2-54
-------
not be used or would not be the control device of choice, and, in these
cases, incineration can be used. Incineration control efficiencies of
98 percent have been demonstrated (IV-B-8). Since incinerators can achieve
higher control efficiencies than carbon adsorbers, incinerators should be
able to meet the standard. For those cases in which incineration would be
the preferred control method, the standard includes compliance provisions
for incineration.
2.6 SELECTION OF FORMAT
2.6.1 Comment: The trade association representative (IV-D-7) suggested
that EPA establish equipment design or engineering specifications as an
alternative to demonstrating compliance with the percent reduction require-
ments. This would allow a new facility to be designed with the assurance
that it would achieve compliance before a capital outlay. Such a format
would reduce testing and monitoring costs substantially. The commenter
added that such a provision would be compatible with the proposed NSPS for
rubber tire manufacturing.
Response: Section lll(h)(l) and (2) of the Clean Air Act allows the
Administrator to promulgate a design, equipment, work practice, or
operational standard rather than a standard of performance only when a
pollutant or pollutants cannot be emitted through a conveyance designed and
constructed to emit or capture that pollutant; when a requirement for such a
conveyance would be inconsistent with any Federal, State or local law; or
when the application of measurement methodology to a particular class of
sources is not practicable due to technological or economic limitations.
In the proposed NSPS for the rubber tire manufacturing industry
(January 20, 1983, 48 FR 2676), an alternative compliance method was
proposed that was applicable to three of the nine operations covered by the
standard. The main compliance provisions were based on a percent reduction
of the maximum allowable mass of VOC emissions per unit of production. The
alternative compliance method let the owner or operator elect to meet
equipment specifications for capture systems used in conjunction with a
2-55
-------
95 percent efficleat control device. For the vinyl and urethane printing
standard, however, EPA does not believe that equipment specifications give
reasonable assurance of compliance with the performance standard.
Since test data are available and there are no technical or economic
limitations involved, the Administrator has promulgated a performance
standard for the vinyl printing industry as required by Section lll(h)(l)
and (2).
2.7 SELECTION OF EMISSION LIMITS
2.7.1 Comment: One commenter (IV-D-5) felt that the proposed 85 percent
VOC reduction requirement could not be achieved by all affected facilities
on a continuous basis. The commenter said that carbon adsorption system
manufacturers have guaranteed 95 percent efficiency only under optimum
process and recovery conditions. The commenter thought it unlikely that a
company's management would risk a one million dollar or greater expenditure
for an expansion or reconstruction with no certainty it could meet the
85 percent reduction requirement. He felt that the 85 percent reduction
requirement would halt the limited growth projected for the industry. The
commenter had also noted in comments he resubmitted (II-D-100) on an earlier
draft of the standard that the 85 percent overall VOC reduction may not be
reproducible or attainable for new installations in existing facilities. He
based this opinion on the statements that air requirements of other presses
and processes may prevent the use of the building as a "total enclosure,"
the varying load to the emission control device from multiple printing lines
will reduce efficiency, and the carbon bed may lose efficiency under
full-time use faster than anticipated in the BID. The commenter concluded
in both letters that, since the 85 percent reduction would not be achievable
by all affected facilities, Regulatory Alternative II, the 75 percent
reduction requirement, should be adopted.
Response: The 95 percent carbon adsorber efficiency was based on test
data showing efficiencies in excess of 98 percent, technology transfer from
other industries such as publication rotogravure printing and pressure
2-56
-------
sensitive tapes and labels showing carbon adsorber efficiencies of
95 percent, and on the manufacturer's guarantee of 95 percent efficiency of
the carbon adsorber tested. As explained in the response to Comment 2.2.2,
one plant confirmed the ability of its carbon adsorber to achieve 95 percent
control after 30 years of operation. For these reasons, EPA feels confident
that 95 percent control efficiency can be achieved consistently throughout
this industry.
The commenter favored the 75 percent reduction alternative, stating
that a requirement for 85 percent emission reduction would halt industry
growth. The Administrator finds no reason, nor did the commenter offer
data, to support the idea that a 75 percent reduction requirement would have
any different effect on industry growth than would an 85 percent reduction
requirement. Total annualized cost is less under 85 percent control due to
the greater amount of recovered solvent. Capital cost of control at a
typical plant is less than 10 percent greater at the 85 percent control
level, and the economic analysis presented in Chapter 9 of the BID did not
indicate any significant impacts on the industry.
EPA has not changed its evaluation of the facts since its response
(II-C-86) to the commenter's earlier letter (II-D-100). The response to
Comment 2.2.4 discusses enclosure of the print line during testing. The
regulation provides for making specific alterations during compliance
testing. The following two paragraphs are from the previous response
(II-C-86) to these issues.
Varying load to the control device from multiple sources
For the purposes of determining compliance with a new source
performance standard, a performance test is to be conducted on each affected
facility and that is defined as a single print line. According to the
General Provisions [40 CFR 60.8(c)1, performance tests must be conducted
under conditions that are representative of performance of the affected
facility. Since each line is tested independently under representative
operating conditions, the effect of "varying loads from multiple lines" is
never a subject of issue. If operating conditions change substantially
2-57
-------
after a performance test, another performance test must be conducted under
the new representative conditions.
Loss of carbon bed efficiency
At the National Air Pollution Control Techniques Advisory Committee
meeting (II-B-50) on September 22, 1981, an industry representative
described a carbon adsorber that had been operating at one of this company's
plants for over 30 years. He noted that a recent analysis of system
efficiency showed that 95 percent control was obtained. Carbon life for
this adsorber is about 3 years, the same as was used in our cost estimates.
For these reasons, the Administrator maintains that the 85 percent
reduction requirement can be met by all affected facilities.
2.7.2 Comment: The trade association representative (IV-D-7) supported
the concept exempting waterborne inks from the emission standard, but stated
that the proposed standard was not sufficiently flexible in that area. As
an example, he noted that the proposed NSPS does not provide for the use of
waterborne and solventborne inks on the same line. The association believes
that changing the ratio of VOC to ink solids and the averaging period used
in determining the VOC content of the ink would provide the desired flexi-
bility. Specifically, they recommended that the cutoff point for waterborne
inks be raised to 2.3 kg VOC per kg ink solids to enable a plant to use
solventborne inks (e.g., 8.0 kg VOC per kg ink solids) for a longer time
period. The commenter felt this would increase the development and use of
waterborne inks.
The commenter felt that the proposed 30-day averaging period is too
short and that a longer averaging period would not affect total potential
allowable emissions.
Response: In a telephone call to clarify this comment (IV-E-16), the
commenter explained that the proposed standard did not allow for the use of
solventborne inks and waterborne inks at different print heads on the same
print line at the same time. He noted in that telephone conversation that
provisions for use of both types of inks at the same time are included in
2-58
-------
the publication rotogravure NSPS. The response to Comment 2.2.6 explains
EPA's position on this issue.
The commenter recommended increasing the ratio of VOC to ink solids
used to define "waterborne inks." As the preamble to the proposed
regulation explains, industry information has shown that the waterborne inks
under development contain less than 1 kg VOC per kg ink solids, the level
proposed below which the percent reduction standard would not apply.
Increasing this cutoff to 2.3 kg as requested would not represent best
demonstrated technology since such a change would allow an increase in
emissions that could be avoided at a reasonable cost. It is unlikely that
there are any waterborne inks in the 1.0 to 2.3 kg range. EPA does not
agree that raising this cutoff would encourage the development and use of
waterborne inks.
EPA examined alternatives (II-B-49) to the 30-day averaging period and
concluded that the 30-day averaging period was sufficient to allow some use
of high solvent inks and to encourage the development and use of waterborne
inks. In response to Comment 2.10.1, however, the averaging period was
increased to a maximum of 35 consecutive calendar days to add flexibility in
accounting procedures.
The concept of an averaging period allows a facility that uses
predominantly waterborne inks to use inks with a higher VOC content
occasionally without installing an emission control system. The question of
which averaging period to choose for waterborne inks is a question of
whether the limit of 1.0 kg VOC per kg ink solids can be exceeded for some
period of time. This period can be visualized as a peak level during the
averaging period. While lengthening the averaging period would not affect
total emissions, it would allow longer concentrated peak periods. These
peak periods possibly would cause local odor and smog problems depending on
local climatic conditions. In addition to such problems, lengthening the
averaging period would make enforcement more difficult. Thus, EPA has not
extended the averaging period as requested.
2-59
-------
2.8 LEGAL CONSIDERATIONS
2.8.1 Comment: The trade association representative (IV-D-7) commented
that EPA should consider incorporating a provision that would permit over-
control of existing facilities controlled by the same control device as a
new source in lieu of achieving the percent reduction required by the
proposed standard. He noted that such a compliance option would achieve the
same degree of emissions reduction as enforcing the SIP requirements for the
existing sources and the NSPS for the new sources, while providing needed
flexibility. He cited the rotogravure printing NSPS UO CFR § 60.433(e)l as
having such a compliance option.
Response: The suggestion presented by the commenter would involve
obtaining emission reduction credits from an existing facility to offset the
amount by which the new facility exceeds the level attainable with BDT.
This is not appropriate for several reasons. First, this is inconsistent
with the primary purpose of Section 111, which is to reduce emissions to a
minimum through the use of BDT. In addition, this alternative creates a
number of enforcement uncertainties, one of which is the question of what
would happen if the new facility has a longer life than the facility from
which the emission reduction credits were obtained. Moreover, implementa-
tion of such an approach would require (1) precise pre-construction
information on expected actual emission rates and (2) assumptions about the
future relative inputs of the multiple units to the common control device.
In the first instance, although it is presumed that the facilities in
question would be meeting State standards, inherent variability in
performance of the existing control device would make it difficult to pick a
baseline emission level against which to measure improvement. This
variability would make it very difficult to differentiate between actual
emission reductions and calculated reductions that may or may not actually
occur. In the second instance, where several different types of equipment
were ducted in common, it would be difficult not only to apportion emissions
among them as a base for determining reduction but also to assess the
long-term constancy of that apportionment once it is made.
2-60
-------
The publication rotogravure printing NSPS does not allow overcontrol of
one existing facility in lieu of achieving BDT at a new, affected facility.
Rather, it has provisions for showing compliance of the affected facility by
testing the combined affected and existing facilities controlled by the same
solvent recovery systems. The need for this compliance provision results
from the use of a single control device for more than one facility and the
inseparable coming!ing of solvent input and recycle streams. This precludes
single facility mass balances. An operator choosing this compliance option
would be required to show that the control achieved by all facilities,
affected and existing, is equal to or better than the BDT level of control
required by the NSPS. In the Agency's judgment, this assures that the
affected facilities will achieve the degree of control reflecting the
application of BDT. The rotogravure standard is based on a liquid material
balance calculated over a 30-day performance test period. Neither the
material balance nor the 30-day test period is appropriate for the vinyl and
urethane industry.
For the reasons stated above, such alternative compliance provisions as
suggested by the commenter will not be incorporated into the regulation.
2.8.2 Comment: The trade association representative (IV-D-7) recommended
that EPA reconsider the proposed standard because it does not comply with
Executive Order 12291. The NSPS fails to comply, he said, because the
environmental benefits are negligible, the costs of compliance are
substantial, and stringent control of negligible emissions from new sources
(if any) in this industry is not a regulatory objective which will "maximize
the net benefits to society."
Response: Executive Order 12291 requires a Regulatory Impacts Analysis
for "major rules" and requires that all agencies adhere to the following
general requirements:
(a) base administrative decisions on adequate information concerning
the need for and consequences of the proposed action;
2-61
-------
(b) take regulatory action only when potential benefits outweigh
potential costs to society;
(c) choose regulatory objectives that maximize net benefits to
society;
(d) choose the regulatory alternative involving the least net cost to
society; and
(e) set regulatory priorities that maximize the aggregate net benefits
to society, taking into account the condition of the particular
industries affected by the regulations, the condition of the
nation's economy, and other regulatory actions contemplated in the
future.
The preamble to the proposed regulation (48 FR 2287, col. 3) explained
that this regulation is not a major rule because it would not result in any
of the adverse economic effects described in Section 1 of the Executive
Order. Thus, no Regulatory Impacts Analysis is warranted.
The general requirements of the Order are somewhat more subjective than
the criteria for defining a "major rule." However, the Administrator feels
that the careful analysis of environmental, economic, and energy impacts
presented in the BID and discussed in the preamble indeed verify that the
general requirements of the Executive Order have been met. As explained in
the preamble, this regulation was submitted to the Office of Management and
Budget (OMB) as required by the Order and was proposed after OMB review and
approval. Finally, the requirements of the Executive Order do not override
the requirements of Section 111 of the Clean Air Act. As discussed under
Comment 2.1.1, Section 111 requires the promulgation of this NSPS.
2.9 TEST METHODS AND MONITORING
2.9.1 Comment: The trade association representative (IV-D-7) suggested
EPA amend the testing provisions to make them parallel to those for
rotogravure printing (40 CFR § 60.433). He felt that such procedures would
allow plants to prove compliance in a variety of ways and minimize the need
to shut down one facility or to build temporary structures for testing.
Such procedures might result in cost savings.
2-62
-------
Response: The response to Comment 2.8.1 explains why obtaining
emission credits from an existing facility to offset control requirements at
an affected facility is unacceptable. The response also explains that in
the rotogravure printing NSPS both existing and affected facilities would
achieve control equal to or greater than the NSPS requirement.
The General Provisions (40 CFR 60.8) give the owner or operator the
authority to use an alternative performance test method if it can be
demonstrated to the Administrator to be adequate for indicating compliance.
2.9.2 Comment: Two commenters (IV-D-2, IV-D-5) addressed compliance
provisions for facilities using waterborne inks. One commenter (IV-D-2), an
ink manufacturer, felt that calculating the VOC content of their inks would
be meaningless, add to overhead, and thus decrease profits because they have
thousands of formulations and because they change formulations to meet
customers' specifications. He agreed with the need to measure or calculate
VOC content in the inks and suggested that ink manufacturers be allowed to
submit general type formulations for families of inks and use approximate
figures (plus or minus a certain percent) for VOC content. The second
commenter (IV-D-5) felt that it was physically and economically infeasible
for a facility using inks with an average VOC content of less than 1 kg VOC
per kg ink solids to demonstrate compliance by determining and recording the
VOC content of each ink used at the print head. He had explained this in
previous comments dated October 20, 1981 (II-D-100) and resubmitted those
comments with this letter. He pointed out that the inks applied are a blend
of recycled excess inks, solvent, water, and new inks. Due to recycling and
addition of water and solvent, the commenter felt that ink manufacturer's
data would be of little use to his company. He added that it would also be
practically impossible to use Reference Method 24 for the large number of
inks his company would apply per day. This would be 250 to 350 tests in one
day for each of two plants.
The commenter suggested what he considered a more reasonable and more
accurate compliance system where the average VOC content of the inks applied
would be determined through the use of ink manufacturer's formulations,
2-63
-------
inventory and purchase records for new inks and diluents, the acceptance of
recycled inks as being previously accounted for as new ink and diluent, and
the measure of VOC content of discarded inks. This system was expressed by
the following equation:
Emission Ink Solvent Discard Blending and Storage
VOC VOC VOC " VOC " VOC
VOC diluent added at the press would be accounted for in the solvent
inventory and purchase records. The commenter noted that a small amount of
VOC would be lost from storage and blending of recycled inks, but that this
loss would be of little significance to the compliance determination as the
evaporation rate of the low VOC content inks would be slight and the
resulting error would inflate the calculated VOC-to-solids ratio. He felt
that this limited number of samples could be analyzed in outside laborato-
ries and eliminate the creation of a need for analytical laboratories in
printing facilities. He also pointed out that short-term use of solvent-
based inks on primarily waterborne facilities would be accountable because
the two systems are incompatible and will be easily differentiated in
purchasing, inventory, and manufacturing records.
The commenter concluded by requesting that the requirement in
§60.583(c) (Test methods and procedures) to determine the VOC content of
each ink as applied be deleted because it is not attainable for all
facilities. He urged that the use of the weighted average be modified to
provide credit for unused excess inks for determination of compliance. The
commenter felt that the unreasonable cost of determining the VOC content of
each waterborne ink, as applied, to comply with the proposed standard would
discourage further development of waterborne inks.
Response: EPA agrees that in some cases the use of ink manufacturer's
formulation data in combination with inventory or purchase records for new
inks and dilution solvent might be appropriate in determining the weighted
average VOC content of waterborne inks. Therefore, points to be covered in
petitioning the Administrator to use such a method have been added to the
regulation. Under such a method, the operator would account for the VOC and
2-64
-------
solids content of new inks and dilution solvent directed to the print line
during the averaging period. Credit would be given for VOC in discarded
ink. Credit would also be given for inks at the end of the averaging period
that are to be recycled but are not yet recycled. Credit for these recycled
inks is not to be taken again during the next averaging period. As the
commenter stated, however, no credit can be given for VOC lost during
blending and storage of recycled inks since this quantity cannot be easily
measured. This quantity would be of little significance in the compliance
determination.
The ink manufacturer (IV-D-2) explained that calculating VOC content of
every formulation would be an excessive requirement and asked that ink
manufacturers be allowed to submit general type formulations and to
approximate VOC content. This standard makes no requirements on ink
manufacturers. However, it is anticipated that, by allowing operators of
affected facilities to use inventory and purchase records to account for VOC
content and by easing the requirement to account for the content of each ink
as used at the print head, this commenter (IV-D-5) may find some relief.
2.10 REPORTING AND RECORDKEEPING
2.10.1 Comment: One commenter (IV-D-10) requested that the performance
averaging period for waterborne inks be defined as "a calendar month or
other comparable period on a routine basis." He enclosed a 1983 calendar
illustrating his company's 4-4-5 calendar for all accounting, production,
and marketing records. Using this system, each quarter consists of 2 months
with four weeks and one month with five weeks. He pointed out that this
standard 4-4-5 cycle permitted a uniform system for statistical analysis
although it does not follow a true "calendar" month, and the emission rates
are not affected since the rate number is not time dependent but based on
production or hours of operation. He felt that under this system, only the
reporting cycle is changed and record retention is facilitated and is more
readily available by accommodating the 4-4-5 calendar rather than a strict
calendar month system.
2-65
-------
Response: EPA agrees that it would be desirable to facilitate a
plant's recordkeeping by accommodating this suggestion. The regulation has
been changed so that the weighted average "shall be calculated over a period
that does not exceed one calendar month, or four consecutive weeks. A
facility that uses an accounting system based on quarters consisting'of two
28 calendar day periods and one 35 calendar day period may use an averaging
period of 35 calendar days four times per year provided the use of such an
accounting system is documented in the initial performance test results."
2-66
-------
TECHNICAL REPORT DATA
(Plfast rtad Instruction* on the riven* before completing)
EPA-450/3-81-Q16b
3. R6CIPIENT-S ACCESSION NO.
Flexible Vinyl and Urethane Coating and Printing
Background Information for Promulgated Standards
5. REPORT DATE
June 1984
«. PERFORMING ORGANIZATION COOC
8. PERFORMING ORGANIZATION REPORT NO
I ESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO7
11. CONTRACT/GRANT NO.
68-02-3Q58
2. SPONSORING AGENCY NAMC AND ADDRESS
DAA for Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park. North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY COO6
EPA/200/04
IOTES
Standards of performance to control emissions of volatile organic compounds
from flexible vinyl and urethane coating and printing operations are being
promulgated under the Authority of Section 111 of the Clean Air Act. These '
standards apply to flexible vinyl and urethane printing lines for which
construction, modification, or reconstruction began on or after January 18, 1983,
This document contains a summary of public comments, EPA responses, and a
discussion of differences between the proposed and promulgated standards of
performance.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b-IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Croup
Air Pollution
Industrial Fabric Coating
Pollution Control
Standards of Performance
Urethane Coating and Printing
Vinyl Coating and Printing
Volatile Organic Compounds
Air Pollution Control
13-B
iTEMENT
Unlimited
19. SECURITY CLASS (Tlta Report)
Unclassified
21. NO. OF PAGES
74
2O. SECURITY CLASS (Tilts page)
Unclassified
22. PRICE
EPA Fan* 2220.1 (R«». 4—77) PREVIOUS EDITION is OBSOV.KTC
-------
U.S. Environmental Protection Aeencv
Region V, Library y'
230 Sourh Doar^c-n Street
Chicago, Illinois 60604
------- |